From 0f69ce1e4474e5d5e266457e8a1f4166cf71f6c7 Mon Sep 17 00:00:00 2001 From: Jack Morgenstein Date: Fri, 4 Nov 2005 16:03:32 -0800 Subject: [PATCH 001/121] [IB] mthca: report page size capability Report the device's real page size capability in mthca_query_device(). Signed-off-by: Jack Morgenstein Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_dev.h | 1 + drivers/infiniband/hw/mthca/mthca_main.c | 1 + drivers/infiniband/hw/mthca/mthca_provider.c | 1 + 3 files changed, 3 insertions(+) diff --git a/drivers/infiniband/hw/mthca/mthca_dev.h b/drivers/infiniband/hw/mthca/mthca_dev.h index e7e5d3b4f004..808037f25c78 100644 --- a/drivers/infiniband/hw/mthca/mthca_dev.h +++ b/drivers/infiniband/hw/mthca/mthca_dev.h @@ -154,6 +154,7 @@ struct mthca_limits { int reserved_mcgs; int num_pds; int reserved_pds; + u32 page_size_cap; u32 flags; u8 port_width_cap; }; diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c index 45c6328e780c..16594d1342df 100644 --- a/drivers/infiniband/hw/mthca/mthca_main.c +++ b/drivers/infiniband/hw/mthca/mthca_main.c @@ -181,6 +181,7 @@ static int __devinit mthca_dev_lim(struct mthca_dev *mdev, struct mthca_dev_lim mdev->limits.reserved_uars = dev_lim->reserved_uars; mdev->limits.reserved_pds = dev_lim->reserved_pds; mdev->limits.port_width_cap = dev_lim->max_port_width; + mdev->limits.page_size_cap = ~(u32) (dev_lim->min_page_sz - 1); mdev->limits.flags = dev_lim->flags; /* IB_DEVICE_RESIZE_MAX_WR not supported by driver. diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c index 6b0166668269..e78259b2664b 100644 --- a/drivers/infiniband/hw/mthca/mthca_provider.c +++ b/drivers/infiniband/hw/mthca/mthca_provider.c @@ -90,6 +90,7 @@ static int mthca_query_device(struct ib_device *ibdev, memcpy(&props->node_guid, out_mad->data + 12, 8); props->max_mr_size = ~0ull; + props->page_size_cap = mdev->limits.page_size_cap; props->max_qp = mdev->limits.num_qps - mdev->limits.reserved_qps; props->max_qp_wr = mdev->limits.max_wqes; props->max_sge = mdev->limits.max_sg; From 8b37b94721533f2729c79bcb6fa0bb3e2bc2f400 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Sun, 6 Nov 2005 15:47:02 -0800 Subject: [PATCH 002/121] [IB] umad: two small fixes Two small fixes for the umad module: - set kobject name for issm device properly - in ib_umad_add_one(), s is subtracted from the index i when initializing ports, so s should be subtracted from the index when freeing ports in the error path as well. Signed-off-by: Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/core/user_mad.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index aed5ca23fb22..6aefeed42ab3 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -801,7 +801,7 @@ static int ib_umad_init_port(struct ib_device *device, int port_num, goto err_class; port->sm_dev->owner = THIS_MODULE; port->sm_dev->ops = &umad_sm_fops; - kobject_set_name(&port->dev->kobj, "issm%d", port->dev_num); + kobject_set_name(&port->sm_dev->kobj, "issm%d", port->dev_num); if (cdev_add(port->sm_dev, base_dev + port->dev_num + IB_UMAD_MAX_PORTS, 1)) goto err_sm_cdev; @@ -913,7 +913,7 @@ static void ib_umad_add_one(struct ib_device *device) err: while (--i >= s) - ib_umad_kill_port(&umad_dev->port[i]); + ib_umad_kill_port(&umad_dev->port[i - s]); kref_put(&umad_dev->ref, ib_umad_release_dev); } From 54b79248b24610663298e122c6ba3804468271fe Mon Sep 17 00:00:00 2001 From: Mike Kravetz Date: Mon, 7 Nov 2005 16:25:48 -0800 Subject: [PATCH 003/121] [PATCH] revised Memory Add Fixes for ppc64 Add the create_section_mapping() routine to create hptes for memory sections dynamically added after system boot. Signed-off-by: Mike Kravetz Signed-off-by: Paul Mackerras --- arch/powerpc/mm/hash_utils_64.c | 9 +++++++++ arch/powerpc/mm/mem.c | 3 +++ include/asm-powerpc/sparsemem.h | 4 ++++ 3 files changed, 16 insertions(+) diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c index 22e474876133..c8c9c2df97e9 100644 --- a/arch/powerpc/mm/hash_utils_64.c +++ b/arch/powerpc/mm/hash_utils_64.c @@ -386,6 +386,15 @@ static unsigned long __init htab_get_table_size(void) return pteg_count << 7; } +#ifdef CONFIG_MEMORY_HOTPLUG +void create_section_mapping(unsigned long start, unsigned long end) +{ + BUG_ON(htab_bolt_mapping(start, end, start, + _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_COHERENT | PP_RWXX, + mmu_linear_psize)); +} +#endif /* CONFIG_MEMORY_HOTPLUG */ + void __init htab_initialize(void) { unsigned long table, htab_size_bytes; diff --git a/arch/powerpc/mm/mem.c b/arch/powerpc/mm/mem.c index 6f55efd9be95..2ead0951823a 100644 --- a/arch/powerpc/mm/mem.c +++ b/arch/powerpc/mm/mem.c @@ -127,6 +127,9 @@ int __devinit add_memory(u64 start, u64 size) unsigned long start_pfn = start >> PAGE_SHIFT; unsigned long nr_pages = size >> PAGE_SHIFT; + start += KERNELBASE; + create_section_mapping(start, start + size); + /* this should work for most non-highmem platforms */ zone = pgdata->node_zones; diff --git a/include/asm-powerpc/sparsemem.h b/include/asm-powerpc/sparsemem.h index 1c95ab99deb3..58d2aab416f8 100644 --- a/include/asm-powerpc/sparsemem.h +++ b/include/asm-powerpc/sparsemem.h @@ -11,6 +11,10 @@ #define MAX_PHYSADDR_BITS 38 #define MAX_PHYSMEM_BITS 36 +#ifdef CONFIG_MEMORY_HOTPLUG +extern void create_section_mapping(unsigned long start, unsigned long end); +#endif /* CONFIG_MEMORY_HOTPLUG */ + #endif /* CONFIG_SPARSEMEM */ #endif /* _ASM_POWERPC_SPARSEMEM_H */ From dd7ccbd3eed52a42dc8bba57717dac3ece39b58e Mon Sep 17 00:00:00 2001 From: Mike Kravetz Date: Fri, 4 Nov 2005 15:19:32 -0800 Subject: [PATCH 004/121] [PATCH] Memory Add Fixes for ppc64 memmap_init_zone() sets page count to 1. Before 'freeing' the page, we need to clear the count. This is the same that is done on free_all_bootmem_core() for memory discovered at boot time. Signed-off-by: Mike Kravetz Signed-off-by: Paul Mackerras --- arch/powerpc/mm/mem.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/powerpc/mm/mem.c b/arch/powerpc/mm/mem.c index 2ead0951823a..1dd3cc69a490 100644 --- a/arch/powerpc/mm/mem.c +++ b/arch/powerpc/mm/mem.c @@ -110,6 +110,7 @@ EXPORT_SYMBOL(phys_mem_access_prot); void online_page(struct page *page) { ClearPageReserved(page); + set_page_count(page, 0); free_cold_page(page); totalram_pages++; num_physpages++; From 82dd26a971848efafc6c8f12803ae012d268d2eb Mon Sep 17 00:00:00 2001 From: Mike Kravetz Date: Fri, 4 Nov 2005 15:20:24 -0800 Subject: [PATCH 005/121] [PATCH] Memory Add Fixes for ppc64 This is a temporary kludge that supports adding all new memory to node 0. I will provide a more complete solution similar to that used for dynamically added CPUs in a few days. Signed-off-by: Mike Kravetz Signed-off-by: Paul Mackerras --- include/asm-ppc64/mmzone.h | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/include/asm-ppc64/mmzone.h b/include/asm-ppc64/mmzone.h index 80a708e7093a..15e777ce0f4a 100644 --- a/include/asm-ppc64/mmzone.h +++ b/include/asm-ppc64/mmzone.h @@ -33,6 +33,9 @@ extern int numa_cpu_lookup_table[]; extern char *numa_memory_lookup_table; extern cpumask_t numa_cpumask_lookup_table[]; extern int nr_cpus_in_node[]; +#ifdef CONFIG_MEMORY_HOTPLUG +extern unsigned long max_pfn; +#endif /* 16MB regions */ #define MEMORY_INCREMENT_SHIFT 24 @@ -45,6 +48,11 @@ static inline int pa_to_nid(unsigned long pa) { int nid; +#ifdef CONFIG_MEMORY_HOTPLUG + /* kludge hot added sections default to node 0 */ + if (pa >= (max_pfn << PAGE_SHIFT)) + return 0; +#endif nid = numa_memory_lookup_table[pa >> MEMORY_INCREMENT_SHIFT]; #ifdef DEBUG_NUMA From 7e9191daf643255c24893427f966795e4167ca2c Mon Sep 17 00:00:00 2001 From: Mike Kravetz Date: Mon, 7 Nov 2005 09:39:48 -0800 Subject: [PATCH 006/121] [PATCH] Memory Add Fixes for ppc64 ppc64 needs a special sysfs probe file for adding new memory. Signed-off-by: Mike Kravetz Signed-off-by: Paul Mackerras --- arch/powerpc/Kconfig | 4 ++++ arch/ppc64/Kconfig | 4 ++++ 2 files changed, 8 insertions(+) diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig index 1493c7896fe3..ed31062029f7 100644 --- a/arch/powerpc/Kconfig +++ b/arch/powerpc/Kconfig @@ -599,6 +599,10 @@ config HAVE_ARCH_EARLY_PFN_TO_NID def_bool y depends on NEED_MULTIPLE_NODES +config ARCH_MEMORY_PROBE + def_bool y + depends on MEMORY_HOTPLUG + # Some NUMA nodes have memory ranges that span # other nodes. Even though a pfn is valid and # between a node's start and end pfns, it may not diff --git a/arch/ppc64/Kconfig b/arch/ppc64/Kconfig index 29552348e581..c9d32db9d76a 100644 --- a/arch/ppc64/Kconfig +++ b/arch/ppc64/Kconfig @@ -297,6 +297,10 @@ config HAVE_ARCH_EARLY_PFN_TO_NID def_bool y depends on NEED_MULTIPLE_NODES +config ARCH_MEMORY_PROBE + def_bool y + depends on MEMORY_HOTPLUG + # Some NUMA nodes have memory ranges that span # other nodes. Even though a pfn is valid and # between a node's start and end pfns, it may not From a852daa00ad91350fe603da47becaf3d5af4f2bd Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Wed, 2 Nov 2005 21:42:48 +0100 Subject: [PATCH 007/121] [PATCH] i2c: writing-client doc update complement My latest update to the writing-clients i2c documentation file was incomplete, here's the complement. Large parts of this file are still way out-of-date, but at least now the memory allocation and freeing instructions are consistent. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- Documentation/i2c/writing-clients | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/Documentation/i2c/writing-clients b/Documentation/i2c/writing-clients index cff7b652588a..d19993cc0604 100644 --- a/Documentation/i2c/writing-clients +++ b/Documentation/i2c/writing-clients @@ -412,7 +412,7 @@ For now, you can ignore the `flags' parameter. It is there for future use. release_region(address,FOO_EXTENT); /* SENSORS ONLY END */ ERROR1: - kfree(new_client); + kfree(data); ERROR0: return err; } @@ -443,7 +443,7 @@ much simpler than the attachment code, fortunately! release_region(client->addr,LM78_EXTENT); /* HYBRID SENSORS CHIP ONLY END */ - kfree(data); + kfree(i2c_get_clientdata(client)); return 0; } From ef9627464490fe67235bbd5724d55345b92c0315 Mon Sep 17 00:00:00 2001 From: James Chapman Date: Sun, 6 Nov 2005 23:07:38 +0100 Subject: [PATCH 008/121] [PATCH] i2c: ds1337 BCD conversion fix Fix BCD value errors when month=9, moving the increment inside the BIN2BCD macro. Fix similar code for the weekday value, just for consistency. This bug was reported by Michael Burian . Signed-off-by: James Chapman Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- drivers/i2c/chips/ds1337.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/i2c/chips/ds1337.c b/drivers/i2c/chips/ds1337.c index 01b037007410..02682fb794c8 100644 --- a/drivers/i2c/chips/ds1337.c +++ b/drivers/i2c/chips/ds1337.c @@ -164,9 +164,9 @@ static int ds1337_set_datetime(struct i2c_client *client, struct rtc_time *dt) buf[1] = BIN2BCD(dt->tm_sec); buf[2] = BIN2BCD(dt->tm_min); buf[3] = BIN2BCD(dt->tm_hour); - buf[4] = BIN2BCD(dt->tm_wday) + 1; + buf[4] = BIN2BCD(dt->tm_wday + 1); buf[5] = BIN2BCD(dt->tm_mday); - buf[6] = BIN2BCD(dt->tm_mon) + 1; + buf[6] = BIN2BCD(dt->tm_mon + 1); val = dt->tm_year; if (val >= 100) { val -= 100; From 4a1c4447e523003019a2bf9b972ed6fe411e84d2 Mon Sep 17 00:00:00 2001 From: Yuan Mu Date: Mon, 7 Nov 2005 22:19:04 +0100 Subject: [PATCH 009/121] [PATCH] hwmon: Fix two w83627hf bugs * Fix in4 reads for W83627THF and W83637HF chips. * Use the correct register for alarm flags. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- drivers/hwmon/w83627hf.c | 16 ++++------------ 1 file changed, 4 insertions(+), 12 deletions(-) diff --git a/drivers/hwmon/w83627hf.c b/drivers/hwmon/w83627hf.c index 70ef926c3bd8..4e9a04e1f08e 100644 --- a/drivers/hwmon/w83627hf.c +++ b/drivers/hwmon/w83627hf.c @@ -180,11 +180,10 @@ superio_exit(void) #define W83781D_REG_BANK 0x4E #define W83781D_REG_CONFIG 0x40 -#define W83781D_REG_ALARM1 0x41 -#define W83781D_REG_ALARM2 0x42 -#define W83781D_REG_ALARM3 0x450 +#define W83781D_REG_ALARM1 0x459 +#define W83781D_REG_ALARM2 0x45A +#define W83781D_REG_ALARM3 0x45B -#define W83781D_REG_IRQ 0x4C #define W83781D_REG_BEEP_CONFIG 0x4D #define W83781D_REG_BEEP_INTS1 0x56 #define W83781D_REG_BEEP_INTS2 0x57 @@ -1370,13 +1369,6 @@ static void w83627hf_init_client(struct i2c_client *client) W83781D_REG_TEMP3_CONFIG, tmp & 0xfe); } } - - /* enable comparator mode for temp2 and temp3 so - alarm indication will work correctly */ - i = w83627hf_read_value(client, W83781D_REG_IRQ); - if (!(i & 0x40)) - w83627hf_write_value(client, W83781D_REG_IRQ, - i | 0x40); } /* Start monitoring */ @@ -1400,7 +1392,7 @@ static struct w83627hf_data *w83627hf_update_device(struct device *dev) /* skip missing sensors */ if (((data->type == w83697hf) && (i == 1)) || ((data->type == w83627thf || data->type == w83637hf) - && (i == 4 || i == 5))) + && (i == 5 || i == 6))) continue; data->in[i] = w83627hf_read_value(client, W83781D_REG_IN(i)); From 8750197f0e8f5467297d72e11444cf32f29d790f Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Mon, 31 Oct 2005 18:51:21 +0100 Subject: [PATCH 010/121] [PATCH] i2c-viapro: Some adjustments The big i2c-viapro SMBus driver update which went into 2.6.14-git1 introduced a few minor issues. Nothing critical, but I would like a few adjustments to be merged in to fix the following problems: * VIA should not be spelled Via. * Frodo Looijaard and Philip Edelbrock did not write the i2c-viapro driver. * When debugging is disabled, half of messages would be logged. * Drop an unneeded masking. * Some port reads can be avoided now that the transaction size is passed as a parameter to vt596_transaction(). * SMBus Receive Byte transactions are used for probing too (for EEPROMs), so hide errors on these too. Signed-off-by: Jean Delvare Signed-off-by: Greg Kroah-Hartman --- Documentation/i2c/busses/i2c-viapro | 6 ++---- drivers/i2c/busses/i2c-viapro.c | 27 ++++++++++++++------------- 2 files changed, 16 insertions(+), 17 deletions(-) diff --git a/Documentation/i2c/busses/i2c-viapro b/Documentation/i2c/busses/i2c-viapro index 9363b8bd6109..16775663b9f5 100644 --- a/Documentation/i2c/busses/i2c-viapro +++ b/Documentation/i2c/busses/i2c-viapro @@ -7,12 +7,10 @@ Supported adapters: * VIA Technologies, Inc. VT82C686A/B Datasheet: Sometimes available at the VIA website - * VIA Technologies, Inc. VT8231, VT8233, VT8233A, VT8235, VT8237 - Datasheet: available on request from Via + * VIA Technologies, Inc. VT8231, VT8233, VT8233A, VT8235, VT8237R + Datasheet: available on request from VIA Authors: - Frodo Looijaard , - Philip Edelbrock , Kyösti Mälkki , Mark D. Studebaker , Jean Delvare diff --git a/drivers/i2c/busses/i2c-viapro.c b/drivers/i2c/busses/i2c-viapro.c index c9366b504833..a2237d4b2cf2 100644 --- a/drivers/i2c/busses/i2c-viapro.c +++ b/drivers/i2c/busses/i2c-viapro.c @@ -142,19 +142,18 @@ static int vt596_transaction(u8 size) /* Make sure the SMBus host is ready to start transmitting */ if ((temp = inb_p(SMBHSTSTS)) & 0x1F) { dev_dbg(&vt596_adapter.dev, "SMBus busy (0x%02x). " - "Resetting... ", temp); + "Resetting...\n", temp); outb_p(temp, SMBHSTSTS); if ((temp = inb_p(SMBHSTSTS)) & 0x1F) { - printk("Failed! (0x%02x)\n", temp); + dev_err(&vt596_adapter.dev, "SMBus reset failed! " + "(0x%02x)\n", temp); return -1; - } else { - printk("Successful!\n"); } } /* Start the transaction by setting bit 6 */ - outb_p(0x40 | (size & 0x3C), SMBHSTCNT); + outb_p(0x40 | size, SMBHSTCNT); /* We will always wait for a fraction of a second */ do { @@ -171,7 +170,7 @@ static int vt596_transaction(u8 size) if (temp & 0x10) { result = -1; dev_err(&vt596_adapter.dev, "Transaction failed (0x%02x)\n", - inb_p(SMBHSTCNT) & 0x3C); + size); } if (temp & 0x08) { @@ -180,11 +179,13 @@ static int vt596_transaction(u8 size) } if (temp & 0x04) { + int read = inb_p(SMBHSTADD) & 0x01; result = -1; - /* Quick commands are used to probe for chips, so - errors are expected, and we don't want to frighten the - user. */ - if ((inb_p(SMBHSTCNT) & 0x3C) != VT596_QUICK) + /* The quick and receive byte commands are used to probe + for chips, so errors are expected, and we don't want + to frighten the user. */ + if (!((size == VT596_QUICK && !read) || + (size == VT596_BYTE && read))) dev_err(&vt596_adapter.dev, "Transaction error!\n"); } @@ -462,9 +463,9 @@ static void __exit i2c_vt596_exit(void) } } -MODULE_AUTHOR( - "Frodo Looijaard and " - "Philip Edelbrock "); +MODULE_AUTHOR("Kyosti Malkki , " + "Mark D. Studebaker and " + "Jean Delvare "); MODULE_DESCRIPTION("vt82c596 SMBus driver"); MODULE_LICENSE("GPL"); From 3825ac0ee66b50cb0208ee74796fe65f3040e67c Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Tue, 8 Nov 2005 22:48:08 +1100 Subject: [PATCH 011/121] powerpc: Fix crash in early boot on some powermacs Some powermac machines were crashing in the quiesce firmware call in prom_init.c because we have just closed the OF stdin device; notably my 1999 G3 powerbook does this. To avoid this, don't close the OF stdin device on powermacs. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/prom_init.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c index 6dc33d19fc2a..58f0917bd6b6 100644 --- a/arch/powerpc/kernel/prom_init.c +++ b/arch/powerpc/kernel/prom_init.c @@ -2081,8 +2081,13 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, prom_printf("copying OF device tree ...\n"); flatten_device_tree(); - /* in case stdin is USB and still active on IBM machines... */ - prom_close_stdin(); + /* + * in case stdin is USB and still active on IBM machines... + * Unfortunately quiesce crashes on some powermacs if we have + * closed stdin already (in particular the powerbook 101). + */ + if (RELOC(of_platform) != PLATFORM_POWERMAC) + prom_close_stdin(); /* * Call OF "quiesce" method to shut down pending DMA's from From fca5dcd4835ed09bb1a48a355344aff7a25c76e0 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Tue, 8 Nov 2005 22:55:08 +1100 Subject: [PATCH 012/121] powerpc: Simplify and clean up the xmon terminal I/O This factors out the common bits of arch/powerpc/xmon/start_*.c into a new nonstdio.c, and removes some stuff that was supposed to make xmon's I/O routines somewhat stdio-like but was never used. It also makes the parsing of the xmon= command line option common, so that ppc32 can now use xmon={off,on,early} also. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/setup-common.c | 21 +++ arch/powerpc/kernel/setup_32.c | 11 +- arch/powerpc/kernel/setup_64.c | 20 --- arch/powerpc/xmon/Makefile | 2 +- arch/powerpc/xmon/nonstdio.c | 134 +++++++++++++++ arch/powerpc/xmon/nonstdio.h | 32 ++-- arch/powerpc/xmon/start_32.c | 235 +++----------------------- arch/powerpc/xmon/start_64.c | 167 +------------------ arch/powerpc/xmon/start_8xx.c | 255 +---------------------------- arch/powerpc/xmon/subr_prf.c | 54 ------ arch/powerpc/xmon/xmon.c | 37 ++++- include/asm-powerpc/xmon.h | 1 - 12 files changed, 238 insertions(+), 731 deletions(-) create mode 100644 arch/powerpc/xmon/nonstdio.c delete mode 100644 arch/powerpc/xmon/subr_prf.c diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c index e22856ecb5a0..26bb1fe6384b 100644 --- a/arch/powerpc/kernel/setup-common.c +++ b/arch/powerpc/kernel/setup-common.c @@ -51,6 +51,7 @@ #include #include #include +#include #undef DEBUG @@ -559,3 +560,23 @@ void __init smp_setup_cpu_maps(void) #endif /* CONFIG_PPC64 */ } #endif /* CONFIG_SMP */ + +#ifdef CONFIG_XMON +static int __init early_xmon(char *p) +{ + /* ensure xmon is enabled */ + if (p) { + if (strncmp(p, "on", 2) == 0) + xmon_init(1); + if (strncmp(p, "off", 3) == 0) + xmon_init(0); + if (strncmp(p, "early", 5) != 0) + return 0; + } + xmon_init(1); + debugger(NULL); + + return 0; +} +early_param("xmon", early_xmon); +#endif diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c index 3af2631e3fab..0fbc73f7f828 100644 --- a/arch/powerpc/kernel/setup_32.c +++ b/arch/powerpc/kernel/setup_32.c @@ -303,14 +303,9 @@ void __init setup_arch(char **cmdline_p) pmac_feature_init(); /* New cool way */ #endif -#ifdef CONFIG_XMON - xmon_map_scc(); - if (strstr(cmd_line, "xmon")) { - xmon_init(1); - debugger(NULL); - } -#endif /* CONFIG_XMON */ - if ( ppc_md.progress ) ppc_md.progress("setup_arch: enter", 0x3eab); +#ifdef CONFIG_XMON_DEFAULT + xmon_init(1); +#endif #if defined(CONFIG_KGDB) if (ppc_md.kgdb_map_scc) diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index 0471e843b6c5..54c606f680b3 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -858,26 +858,6 @@ int check_legacy_ioport(unsigned long base_port) } EXPORT_SYMBOL(check_legacy_ioport); -#ifdef CONFIG_XMON -static int __init early_xmon(char *p) -{ - /* ensure xmon is enabled */ - if (p) { - if (strncmp(p, "on", 2) == 0) - xmon_init(1); - if (strncmp(p, "off", 3) == 0) - xmon_init(0); - if (strncmp(p, "early", 5) != 0) - return 0; - } - xmon_init(1); - debugger(NULL); - - return 0; -} -early_param("xmon", early_xmon); -#endif - void cpu_die(void) { if (ppc_md.cpu_die) diff --git a/arch/powerpc/xmon/Makefile b/arch/powerpc/xmon/Makefile index 79a784f0e7a9..b20312e5ed27 100644 --- a/arch/powerpc/xmon/Makefile +++ b/arch/powerpc/xmon/Makefile @@ -8,4 +8,4 @@ obj-$(CONFIG_8xx) += start_8xx.o obj-$(CONFIG_6xx) += start_32.o obj-$(CONFIG_4xx) += start_32.o obj-$(CONFIG_PPC64) += start_64.o -obj-y += xmon.o ppc-dis.o ppc-opc.o subr_prf.o setjmp.o +obj-y += xmon.o ppc-dis.o ppc-opc.o setjmp.o nonstdio.o diff --git a/arch/powerpc/xmon/nonstdio.c b/arch/powerpc/xmon/nonstdio.c new file mode 100644 index 000000000000..78765833f4c0 --- /dev/null +++ b/arch/powerpc/xmon/nonstdio.c @@ -0,0 +1,134 @@ +/* + * Copyright (C) 1996-2005 Paul Mackerras. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#include +#include +#include "nonstdio.h" + +int xmon_putchar(int c) +{ + char ch = c; + + if (c == '\n') + xmon_putchar('\r'); + return xmon_write(&ch, 1) == 1? c: -1; +} + +static char line[256]; +static char *lineptr; +static int lineleft; + +int xmon_expect(const char *str, unsigned long timeout) +{ + int c; + unsigned long t0; + + /* assume 25MHz default timebase if tb_ticks_per_sec not set yet */ + timeout *= tb_ticks_per_sec? tb_ticks_per_sec: 25000000; + t0 = get_tbl(); + do { + lineptr = line; + for (;;) { + c = xmon_read_poll(); + if (c == -1) { + if (get_tbl() - t0 > timeout) + return 0; + continue; + } + if (c == '\n') + break; + if (c != '\r' && lineptr < &line[sizeof(line) - 1]) + *lineptr++ = c; + } + *lineptr = 0; + } while (strstr(line, str) == NULL); + return 1; +} + +int xmon_getchar(void) +{ + int c; + + if (lineleft == 0) { + lineptr = line; + for (;;) { + c = xmon_readchar(); + if (c == -1 || c == 4) + break; + if (c == '\r' || c == '\n') { + *lineptr++ = '\n'; + xmon_putchar('\n'); + break; + } + switch (c) { + case 0177: + case '\b': + if (lineptr > line) { + xmon_putchar('\b'); + xmon_putchar(' '); + xmon_putchar('\b'); + --lineptr; + } + break; + case 'U' & 0x1F: + while (lineptr > line) { + xmon_putchar('\b'); + xmon_putchar(' '); + xmon_putchar('\b'); + --lineptr; + } + break; + default: + if (lineptr >= &line[sizeof(line) - 1]) + xmon_putchar('\a'); + else { + xmon_putchar(c); + *lineptr++ = c; + } + } + } + lineleft = lineptr - line; + lineptr = line; + } + if (lineleft == 0) + return -1; + --lineleft; + return *lineptr++; +} + +char *xmon_gets(char *str, int nb) +{ + char *p; + int c; + + for (p = str; p < str + nb - 1; ) { + c = xmon_getchar(); + if (c == -1) { + if (p == str) + return NULL; + break; + } + *p++ = c; + if (c == '\n') + break; + } + *p = 0; + return str; +} + +void xmon_printf(const char *format, ...) +{ + va_list args; + int n; + static char xmon_outbuf[1024]; + + va_start(args, format); + n = vsnprintf(xmon_outbuf, sizeof(xmon_outbuf), format, args); + va_end(args); + xmon_write(xmon_outbuf, n); +} diff --git a/arch/powerpc/xmon/nonstdio.h b/arch/powerpc/xmon/nonstdio.h index 84211a21c6f4..47cebbd2b1b1 100644 --- a/arch/powerpc/xmon/nonstdio.h +++ b/arch/powerpc/xmon/nonstdio.h @@ -1,22 +1,14 @@ -typedef int FILE; -extern FILE *xmon_stdin, *xmon_stdout; #define EOF (-1) -#define stdin xmon_stdin -#define stdout xmon_stdout -#define printf xmon_printf -#define fprintf xmon_fprintf -#define fputs xmon_fputs -#define fgets xmon_fgets -#define putchar xmon_putchar -#define getchar xmon_getchar -#define putc xmon_putc -#define getc xmon_getc -#define fopen(n, m) NULL -#define fflush(f) do {} while (0) -#define fclose(f) do {} while (0) -extern char *fgets(char *, int, void *); -extern void xmon_printf(const char *, ...); -extern void xmon_fprintf(void *, const char *, ...); -extern void xmon_sprintf(char *, const char *, ...); -#define perror(s) printf("%s: no files!\n", (s)) +#define printf xmon_printf +#define putchar xmon_putchar + +extern int xmon_putchar(int c); +extern int xmon_getchar(void); +extern char *xmon_gets(char *, int); +extern void xmon_printf(const char *, ...); +extern void xmon_map_scc(void); +extern int xmon_expect(const char *str, unsigned long timeout); +extern int xmon_write(void *ptr, int nb); +extern int xmon_readchar(void); +extern int xmon_read_poll(void); diff --git a/arch/powerpc/xmon/start_32.c b/arch/powerpc/xmon/start_32.c index 69b658c0f760..c2464df4217e 100644 --- a/arch/powerpc/xmon/start_32.c +++ b/arch/powerpc/xmon/start_32.c @@ -11,7 +11,6 @@ #include #include #include -#include #include #include #include @@ -22,10 +21,11 @@ #include #include #include +#include +#include "nonstdio.h" static volatile unsigned char __iomem *sccc, *sccd; unsigned int TXRDY, RXRDY, DLAB; -static int xmon_expect(const char *str, unsigned int timeout); static int use_serial; static int use_screen; @@ -33,16 +33,6 @@ static int via_modem; static int xmon_use_sccb; static struct device_node *channel_node; -#define TB_SPEED 25000000 - -static inline unsigned int readtb(void) -{ - unsigned int ret; - - asm volatile("mftb %0" : "=r" (ret) :); - return ret; -} - void buf_access(void) { if (DLAB) @@ -91,23 +81,7 @@ static unsigned long chrp_find_phys_io_base(void) } #endif /* CONFIG_PPC_CHRP */ -#ifdef CONFIG_MAGIC_SYSRQ -static void sysrq_handle_xmon(int key, struct pt_regs *regs, - struct tty_struct *tty) -{ - xmon(regs); -} - -static struct sysrq_key_op sysrq_xmon_op = -{ - .handler = sysrq_handle_xmon, - .help_msg = "Xmon", - .action_msg = "Entering xmon", -}; -#endif - -void -xmon_map_scc(void) +void xmon_map_scc(void) { #ifdef CONFIG_PPC_MULTIPLATFORM volatile unsigned char __iomem *base; @@ -217,8 +191,6 @@ xmon_map_scc(void) RXRDY = 1; DLAB = 0x80; #endif /* platform */ - - register_sysrq_key('x', &sysrq_xmon_op); } static int scc_initialized = 0; @@ -238,8 +210,7 @@ static inline void do_poll_adb(void) #endif /* CONFIG_ADB_CUDA */ } -int -xmon_write(void *handle, void *ptr, int nb) +int xmon_write(void *ptr, int nb) { char *p = ptr; int i, c, ct; @@ -311,8 +282,7 @@ static unsigned char xmon_shift_keytab[128] = "\0.\0*\0+\0\0\0\0\0/\r\0-\0" /* 0x40 - 0x4f */ "\0\0000123456789\0\0\0"; /* 0x50 - 0x5f */ -static int -xmon_get_adb_key(void) +static int xmon_get_adb_key(void) { int k, t, on; @@ -350,32 +320,21 @@ xmon_get_adb_key(void) } #endif /* CONFIG_BOOTX_TEXT */ -int -xmon_read(void *handle, void *ptr, int nb) +int xmon_readchar(void) { - char *p = ptr; - int i; - #ifdef CONFIG_BOOTX_TEXT - if (use_screen) { - for (i = 0; i < nb; ++i) - *p++ = xmon_get_adb_key(); - return i; - } + if (use_screen) + return xmon_get_adb_key(); #endif - if (!scc_initialized) - xmon_init_scc(); - for (i = 0; i < nb; ++i) { + if (!scc_initialized) + xmon_init_scc(); while ((*sccc & RXRDY) == 0) - do_poll_adb(); + do_poll_adb(); buf_access(); - *p++ = *sccd; - } - return i; + return *sccd; } -int -xmon_read_poll(void) +int xmon_read_poll(void) { if ((*sccc & RXRDY) == 0) { do_poll_adb(); @@ -395,8 +354,7 @@ static unsigned char scc_inittab[] = { 3, 0xc1, /* rx enable, 8 bits */ }; -void -xmon_init_scc(void) +void xmon_init_scc(void) { if ( _machine == _MACH_chrp ) { @@ -410,6 +368,7 @@ xmon_init_scc(void) else if ( _machine == _MACH_Pmac ) { int i, x; + unsigned long timeout; if (channel_node != 0) pmac_call_feature( @@ -424,8 +383,12 @@ xmon_init_scc(void) PMAC_FTR_MODEM_ENABLE, channel_node, 0, 1); printk(KERN_INFO "Modem powered up by debugger !\n"); - t0 = readtb(); - while (readtb() - t0 < 3*TB_SPEED) + t0 = get_tbl(); + timeout = 3 * tb_ticks_per_sec; + if (timeout == 0) + /* assume 25MHz if tb_ticks_per_sec not set */ + timeout = 75000000; + while (get_tbl() - t0 < timeout) eieio(); } /* use the B channel if requested */ @@ -447,164 +410,19 @@ xmon_init_scc(void) scc_initialized = 1; if (via_modem) { for (;;) { - xmon_write(NULL, "ATE1V1\r", 7); + xmon_write("ATE1V1\r", 7); if (xmon_expect("OK", 5)) { - xmon_write(NULL, "ATA\r", 4); + xmon_write("ATA\r", 4); if (xmon_expect("CONNECT", 40)) break; } - xmon_write(NULL, "+++", 3); + xmon_write("+++", 3); xmon_expect("OK", 3); } } } -void *xmon_stdin; -void *xmon_stdout; -void *xmon_stderr; - -int xmon_putc(int c, void *f) -{ - char ch = c; - - if (c == '\n') - xmon_putc('\r', f); - return xmon_write(f, &ch, 1) == 1? c: -1; -} - -int xmon_putchar(int c) -{ - return xmon_putc(c, xmon_stdout); -} - -int xmon_fputs(char *str, void *f) -{ - int n = strlen(str); - - return xmon_write(f, str, n) == n? 0: -1; -} - -int -xmon_readchar(void) -{ - char ch; - - for (;;) { - switch (xmon_read(xmon_stdin, &ch, 1)) { - case 1: - return ch; - case -1: - xmon_printf("read(stdin) returned -1\r\n", 0, 0); - return -1; - } - } -} - -static char line[256]; -static char *lineptr; -static int lineleft; - -int xmon_expect(const char *str, unsigned int timeout) -{ - int c; - unsigned int t0; - - timeout *= TB_SPEED; - t0 = readtb(); - do { - lineptr = line; - for (;;) { - c = xmon_read_poll(); - if (c == -1) { - if (readtb() - t0 > timeout) - return 0; - continue; - } - if (c == '\n') - break; - if (c != '\r' && lineptr < &line[sizeof(line) - 1]) - *lineptr++ = c; - } - *lineptr = 0; - } while (strstr(line, str) == NULL); - return 1; -} - -int -xmon_getchar(void) -{ - int c; - - if (lineleft == 0) { - lineptr = line; - for (;;) { - c = xmon_readchar(); - if (c == -1 || c == 4) - break; - if (c == '\r' || c == '\n') { - *lineptr++ = '\n'; - xmon_putchar('\n'); - break; - } - switch (c) { - case 0177: - case '\b': - if (lineptr > line) { - xmon_putchar('\b'); - xmon_putchar(' '); - xmon_putchar('\b'); - --lineptr; - } - break; - case 'U' & 0x1F: - while (lineptr > line) { - xmon_putchar('\b'); - xmon_putchar(' '); - xmon_putchar('\b'); - --lineptr; - } - break; - default: - if (lineptr >= &line[sizeof(line) - 1]) - xmon_putchar('\a'); - else { - xmon_putchar(c); - *lineptr++ = c; - } - } - } - lineleft = lineptr - line; - lineptr = line; - } - if (lineleft == 0) - return -1; - --lineleft; - return *lineptr++; -} - -char * -xmon_fgets(char *str, int nb, void *f) -{ - char *p; - int c; - - for (p = str; p < str + nb - 1; ) { - c = xmon_getchar(); - if (c == -1) { - if (p == str) - return NULL; - break; - } - *p++ = c; - if (c == '\n') - break; - } - *p = 0; - return str; -} - -void -xmon_enter(void) +void xmon_enter(void) { #ifdef CONFIG_ADB_PMU if (_machine == _MACH_Pmac) { @@ -613,8 +431,7 @@ xmon_enter(void) #endif } -void -xmon_leave(void) +void xmon_leave(void) { #ifdef CONFIG_ADB_PMU if (_machine == _MACH_Pmac) { diff --git a/arch/powerpc/xmon/start_64.c b/arch/powerpc/xmon/start_64.c index e50c158191e1..712552c4f242 100644 --- a/arch/powerpc/xmon/start_64.c +++ b/arch/powerpc/xmon/start_64.c @@ -6,182 +6,29 @@ * as published by the Free Software Foundation; either version * 2 of the License, or (at your option) any later version. */ -#include -#include -#include -#include -#include -#include #include -#include -#include -#include -#include #include -#include #include "nonstdio.h" -#ifdef CONFIG_MAGIC_SYSRQ - -static void sysrq_handle_xmon(int key, struct pt_regs *pt_regs, - struct tty_struct *tty) +void xmon_map_scc(void) { - /* ensure xmon is enabled */ - xmon_init(1); - debugger(pt_regs); } -static struct sysrq_key_op sysrq_xmon_op = -{ - .handler = sysrq_handle_xmon, - .help_msg = "Xmon", - .action_msg = "Entering xmon", -}; - -static int __init setup_xmon_sysrq(void) -{ - register_sysrq_key('x', &sysrq_xmon_op); - return 0; -} -__initcall(setup_xmon_sysrq); -#endif /* CONFIG_MAGIC_SYSRQ */ - -int -xmon_write(void *handle, void *ptr, int nb) +int xmon_write(void *ptr, int nb) { return udbg_write(ptr, nb); } -int -xmon_read(void *handle, void *ptr, int nb) +int xmon_readchar(void) { - return udbg_read(ptr, nb); + if (udbg_getc) + return udbg_getc(); + return -1; } -int -xmon_read_poll(void) +int xmon_read_poll(void) { if (udbg_getc_poll) return udbg_getc_poll(); return -1; } - -FILE *xmon_stdin; -FILE *xmon_stdout; - -int -xmon_putc(int c, void *f) -{ - char ch = c; - - if (c == '\n') - xmon_putc('\r', f); - return xmon_write(f, &ch, 1) == 1? c: -1; -} - -int -xmon_putchar(int c) -{ - return xmon_putc(c, xmon_stdout); -} - -int -xmon_fputs(char *str, void *f) -{ - int n = strlen(str); - - return xmon_write(f, str, n) == n? 0: -1; -} - -int -xmon_readchar(void) -{ - char ch; - - for (;;) { - switch (xmon_read(xmon_stdin, &ch, 1)) { - case 1: - return ch; - case -1: - xmon_printf("read(stdin) returned -1\r\n", 0, 0); - return -1; - } - } -} - -static char line[256]; -static char *lineptr; -static int lineleft; - -int -xmon_getchar(void) -{ - int c; - - if (lineleft == 0) { - lineptr = line; - for (;;) { - c = xmon_readchar(); - if (c == -1 || c == 4) - break; - if (c == '\r' || c == '\n') { - *lineptr++ = '\n'; - xmon_putchar('\n'); - break; - } - switch (c) { - case 0177: - case '\b': - if (lineptr > line) { - xmon_putchar('\b'); - xmon_putchar(' '); - xmon_putchar('\b'); - --lineptr; - } - break; - case 'U' & 0x1F: - while (lineptr > line) { - xmon_putchar('\b'); - xmon_putchar(' '); - xmon_putchar('\b'); - --lineptr; - } - break; - default: - if (lineptr >= &line[sizeof(line) - 1]) - xmon_putchar('\a'); - else { - xmon_putchar(c); - *lineptr++ = c; - } - } - } - lineleft = lineptr - line; - lineptr = line; - } - if (lineleft == 0) - return -1; - --lineleft; - return *lineptr++; -} - -char * -xmon_fgets(char *str, int nb, void *f) -{ - char *p; - int c; - - for (p = str; p < str + nb - 1; ) { - c = xmon_getchar(); - if (c == -1) { - if (p == str) - return NULL; - break; - } - *p++ = c; - if (c == '\n') - break; - } - *p = 0; - return str; -} diff --git a/arch/powerpc/xmon/start_8xx.c b/arch/powerpc/xmon/start_8xx.c index a48bd594cf61..4c17b0486ad5 100644 --- a/arch/powerpc/xmon/start_8xx.c +++ b/arch/powerpc/xmon/start_8xx.c @@ -15,273 +15,30 @@ #include #include #include +#include "nonstdio.h" -extern void xmon_printf(const char *fmt, ...); extern int xmon_8xx_write(char *str, int nb); extern int xmon_8xx_read_poll(void); extern int xmon_8xx_read_char(void); -void prom_drawhex(uint); -void prom_drawstring(const char *str); -static int use_screen = 1; /* default */ - -#define TB_SPEED 25000000 - -static inline unsigned int readtb(void) +void xmon_map_scc(void) { - unsigned int ret; - - asm volatile("mftb %0" : "=r" (ret) :); - return ret; -} - -void buf_access(void) -{ -} - -void -xmon_map_scc(void) -{ - cpmp = (cpm8xx_t *)&(((immap_t *)IMAP_ADDR)->im_cpm); - use_screen = 0; - - prom_drawstring("xmon uses serial port\n"); } -static int scc_initialized = 0; - void xmon_init_scc(void); -int -xmon_write(void *handle, void *ptr, int nb) +int xmon_write(void *ptr, int nb) { - char *p = ptr; - int i, c, ct; - - if (!scc_initialized) - xmon_init_scc(); - return(xmon_8xx_write(ptr, nb)); } -int xmon_wants_key; - -int -xmon_read(void *handle, void *ptr, int nb) +int xmon_readchar(void) { - char *p = ptr; - int i; - - if (!scc_initialized) - xmon_init_scc(); - - for (i = 0; i < nb; ++i) { - *p++ = xmon_8xx_read_char(); - } - return i; + return xmon_8xx_read_char(); } -int -xmon_read_poll(void) +int xmon_read_poll(void) { return(xmon_8xx_read_poll()); } - -void -xmon_init_scc() -{ - scc_initialized = 1; -} - -#if 0 -extern int (*prom_entry)(void *); - -int -xmon_exit(void) -{ - struct prom_args { - char *service; - } args; - - for (;;) { - args.service = "exit"; - (*prom_entry)(&args); - } -} -#endif - -void *xmon_stdin; -void *xmon_stdout; -void *xmon_stderr; - -void -xmon_init(void) -{ -} - -int -xmon_putc(int c, void *f) -{ - char ch = c; - - if (c == '\n') - xmon_putc('\r', f); - return xmon_write(f, &ch, 1) == 1? c: -1; -} - -int -xmon_putchar(int c) -{ - return xmon_putc(c, xmon_stdout); -} - -int -xmon_fputs(char *str, void *f) -{ - int n = strlen(str); - - return xmon_write(f, str, n) == n? 0: -1; -} - -int -xmon_readchar(void) -{ - char ch; - - for (;;) { - switch (xmon_read(xmon_stdin, &ch, 1)) { - case 1: - return ch; - case -1: - xmon_printf("read(stdin) returned -1\r\n", 0, 0); - return -1; - } - } -} - -static char line[256]; -static char *lineptr; -static int lineleft; - -#if 0 -int xmon_expect(const char *str, unsigned int timeout) -{ - int c; - unsigned int t0; - - timeout *= TB_SPEED; - t0 = readtb(); - do { - lineptr = line; - for (;;) { - c = xmon_read_poll(); - if (c == -1) { - if (readtb() - t0 > timeout) - return 0; - continue; - } - if (c == '\n') - break; - if (c != '\r' && lineptr < &line[sizeof(line) - 1]) - *lineptr++ = c; - } - *lineptr = 0; - } while (strstr(line, str) == NULL); - return 1; -} -#endif - -int -xmon_getchar(void) -{ - int c; - - if (lineleft == 0) { - lineptr = line; - for (;;) { - c = xmon_readchar(); - if (c == -1 || c == 4) - break; - if (c == '\r' || c == '\n') { - *lineptr++ = '\n'; - xmon_putchar('\n'); - break; - } - switch (c) { - case 0177: - case '\b': - if (lineptr > line) { - xmon_putchar('\b'); - xmon_putchar(' '); - xmon_putchar('\b'); - --lineptr; - } - break; - case 'U' & 0x1F: - while (lineptr > line) { - xmon_putchar('\b'); - xmon_putchar(' '); - xmon_putchar('\b'); - --lineptr; - } - break; - default: - if (lineptr >= &line[sizeof(line) - 1]) - xmon_putchar('\a'); - else { - xmon_putchar(c); - *lineptr++ = c; - } - } - } - lineleft = lineptr - line; - lineptr = line; - } - if (lineleft == 0) - return -1; - --lineleft; - return *lineptr++; -} - -char * -xmon_fgets(char *str, int nb, void *f) -{ - char *p; - int c; - - for (p = str; p < str + nb - 1; ) { - c = xmon_getchar(); - if (c == -1) { - if (p == str) - return 0; - break; - } - *p++ = c; - if (c == '\n') - break; - } - *p = 0; - return str; -} - -void -prom_drawhex(uint val) -{ - unsigned char buf[10]; - - int i; - for (i = 7; i >= 0; i--) - { - buf[i] = "0123456789abcdef"[val & 0x0f]; - val >>= 4; - } - buf[8] = '\0'; - xmon_fputs(buf, xmon_stdout); -} - -void -prom_drawstring(const char *str) -{ - xmon_fputs(str, xmon_stdout); -} diff --git a/arch/powerpc/xmon/subr_prf.c b/arch/powerpc/xmon/subr_prf.c deleted file mode 100644 index b48738c6dd33..000000000000 --- a/arch/powerpc/xmon/subr_prf.c +++ /dev/null @@ -1,54 +0,0 @@ -/* - * Written by Cort Dougan to replace the version originally used - * by Paul Mackerras, which came from NetBSD and thus had copyright - * conflicts with Linux. - * - * This file makes liberal use of the standard linux utility - * routines to reduce the size of the binary. We assume we can - * trust some parts of Linux inside the debugger. - * -- Cort (cort@cs.nmt.edu) - * - * Copyright (C) 1999 Cort Dougan. - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - */ - -#include -#include -#include -#include -#include "nonstdio.h" - -extern int xmon_write(void *, void *, int); - -void xmon_vfprintf(void *f, const char *fmt, va_list ap) -{ - static char xmon_buf[2048]; - int n; - - n = vsprintf(xmon_buf, fmt, ap); - xmon_write(f, xmon_buf, n); -} - -void xmon_printf(const char *fmt, ...) -{ - va_list ap; - - va_start(ap, fmt); - xmon_vfprintf(stdout, fmt, ap); - va_end(ap); -} -EXPORT_SYMBOL(xmon_printf); - -void xmon_fprintf(void *f, const char *fmt, ...) -{ - va_list ap; - - va_start(ap, fmt); - xmon_vfprintf(f, fmt, ap); - va_end(ap); -} - diff --git a/arch/powerpc/xmon/xmon.c b/arch/powerpc/xmon/xmon.c index 1124f1146202..b43a57425ea9 100644 --- a/arch/powerpc/xmon/xmon.c +++ b/arch/powerpc/xmon/xmon.c @@ -1,7 +1,7 @@ /* * Routines providing a simple monitor for use on the PowerMac. * - * Copyright (C) 1996 Paul Mackerras. + * Copyright (C) 1996-2005 Paul Mackerras. * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -18,6 +18,7 @@ #include #include #include +#include #include #include @@ -144,15 +145,10 @@ static void xmon_print_symbol(unsigned long address, const char *mid, static const char *getvecname(unsigned long vec); extern int print_insn_powerpc(unsigned long, unsigned long, int); -extern void printf(const char *fmt, ...); -extern void xmon_vfprintf(void *f, const char *fmt, va_list ap); -extern int xmon_putc(int c, void *f); -extern int putchar(int ch); extern void xmon_enter(void); extern void xmon_leave(void); -extern int xmon_read_poll(void); extern long setjmp(long *); extern void longjmp(long *, long); extern void xmon_save_regs(struct pt_regs *); @@ -748,7 +744,6 @@ cmds(struct pt_regs *excp) printf("%x:", smp_processor_id()); #endif /* CONFIG_SMP */ printf("mon> "); - fflush(stdout); flush_input(); termch = 0; cmd = skipbl(); @@ -2151,7 +2146,6 @@ memzcan(void) ok = mread(a, &v, 1); if (ok && !ook) { printf("%.8x .. ", a); - fflush(stdout); } else if (!ok && ook) printf("%.8x\n", a - mskip); ook = ok; @@ -2372,7 +2366,7 @@ int inchar(void) { if (lineptr == NULL || *lineptr == 0) { - if (fgets(line, sizeof(line), stdin) == NULL) { + if (xmon_gets(line, sizeof(line)) == NULL) { lineptr = NULL; return EOF; } @@ -2526,4 +2520,29 @@ void xmon_init(int enable) __debugger_dabr_match = NULL; __debugger_fault_handler = NULL; } + xmon_map_scc(); } + +#ifdef CONFIG_MAGIC_SYSRQ +static void sysrq_handle_xmon(int key, struct pt_regs *pt_regs, + struct tty_struct *tty) +{ + /* ensure xmon is enabled */ + xmon_init(1); + debugger(pt_regs); +} + +static struct sysrq_key_op sysrq_xmon_op = +{ + .handler = sysrq_handle_xmon, + .help_msg = "Xmon", + .action_msg = "Entering xmon", +}; + +static int __init setup_xmon_sysrq(void) +{ + register_sysrq_key('x', &sysrq_xmon_op); + return 0; +} +__initcall(setup_xmon_sysrq); +#endif /* CONFIG_MAGIC_SYSRQ */ diff --git a/include/asm-powerpc/xmon.h b/include/asm-powerpc/xmon.h index ace2072d4a83..43f7129984c7 100644 --- a/include/asm-powerpc/xmon.h +++ b/include/asm-powerpc/xmon.h @@ -7,7 +7,6 @@ struct pt_regs; extern int xmon(struct pt_regs *excp); extern void xmon_printf(const char *fmt, ...); extern void xmon_init(int); -extern void xmon_map_scc(void); #endif #endif From 66ba135c5a398df5c3a4b43d84d9df80cbc87c61 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 9 Nov 2005 11:01:06 +1100 Subject: [PATCH 013/121] powerpc: create kernel/setup.h for functions defined by setup-common.c and used in setup_xx.c Signed-off-by: Stephen Rothwell --- arch/powerpc/kernel/setup-common.c | 2 ++ arch/powerpc/kernel/setup.h | 6 ++++++ arch/powerpc/kernel/setup_32.c | 2 ++ arch/powerpc/kernel/setup_64.c | 2 ++ 4 files changed, 12 insertions(+) create mode 100644 arch/powerpc/kernel/setup.h diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c index 26bb1fe6384b..17c46a2e356d 100644 --- a/arch/powerpc/kernel/setup-common.c +++ b/arch/powerpc/kernel/setup-common.c @@ -53,6 +53,8 @@ #include #include +#include "setup.h" + #undef DEBUG #ifdef DEBUG diff --git a/arch/powerpc/kernel/setup.h b/arch/powerpc/kernel/setup.h new file mode 100644 index 000000000000..2ebba755272e --- /dev/null +++ b/arch/powerpc/kernel/setup.h @@ -0,0 +1,6 @@ +#ifndef _POWERPC_KERNEL_SETUP_H +#define _POWERPC_KERNEL_SETUP_H + +void check_for_initrd(void); + +#endif /* _POWERPC_KERNEL_SETUP_H */ diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c index 0fbc73f7f828..55f485f78c00 100644 --- a/arch/powerpc/kernel/setup_32.c +++ b/arch/powerpc/kernel/setup_32.c @@ -40,6 +40,8 @@ #include #include +#include "setup.h" + #define DBG(fmt...) #if defined CONFIG_KGDB diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index 54c606f680b3..fe39aac4f24d 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -61,6 +61,8 @@ #include #include +#include "setup.h" + #ifdef DEBUG #define DBG(fmt...) udbg_printf(fmt) #else From b709c0832824da20a32a3119911b9f0cf6970452 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 9 Nov 2005 13:28:33 +1100 Subject: [PATCH 014/121] ppc64: move stack switching up in interrupt processing This will make the ppc64 multiplatform irq handling more like the generic handling. Signed-off-by: Stephen Rothwell --- arch/powerpc/kernel/misc_64.S | 8 +++---- arch/ppc64/kernel/irq.c | 40 +++++++++++++++++------------------ arch/ppc64/kernel/misc.S | 8 +++---- include/asm-powerpc/irq.h | 4 ++-- 4 files changed, 30 insertions(+), 30 deletions(-) diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S index ae1433da09b2..9d09f0ad6efe 100644 --- a/arch/powerpc/kernel/misc_64.S +++ b/arch/powerpc/kernel/misc_64.S @@ -89,12 +89,12 @@ _GLOBAL(call_do_softirq) mtlr r0 blr -_GLOBAL(call_handle_IRQ_event) +_GLOBAL(call_ppc_irq_dispatch_handler) mflr r0 std r0,16(r1) - stdu r1,THREAD_SIZE-112(r6) - mr r1,r6 - bl .handle_IRQ_event + stdu r1,THREAD_SIZE-112(r5) + mr r1,r5 + bl .ppc_irq_dispatch_handler ld r1,0(r1) ld r0,16(r1) mtlr r0 diff --git a/arch/ppc64/kernel/irq.c b/arch/ppc64/kernel/irq.c index 87474584033f..b61497d5be28 100644 --- a/arch/ppc64/kernel/irq.c +++ b/arch/ppc64/kernel/irq.c @@ -157,9 +157,6 @@ void ppc_irq_dispatch_handler(struct pt_regs *regs, int irq) int cpu = smp_processor_id(); irq_desc_t *desc = get_irq_desc(irq); irqreturn_t action_ret; -#ifdef CONFIG_IRQSTACKS - struct thread_info *curtp, *irqtp; -#endif kstat_cpu(cpu).irqs[irq]++; @@ -227,20 +224,7 @@ void ppc_irq_dispatch_handler(struct pt_regs *regs, int irq) for (;;) { spin_unlock(&desc->lock); -#ifdef CONFIG_IRQSTACKS - /* Switch to the irq stack to handle this */ - curtp = current_thread_info(); - irqtp = hardirq_ctx[smp_processor_id()]; - if (curtp != irqtp) { - irqtp->task = curtp->task; - irqtp->flags = 0; - action_ret = call_handle_IRQ_event(irq, regs, action, irqtp); - irqtp->task = NULL; - if (irqtp->flags) - set_bits(irqtp->flags, &curtp->flags); - } else -#endif - action_ret = handle_IRQ_event(irq, regs, action); + action_ret = handle_IRQ_event(irq, regs, action); spin_lock(&desc->lock); if (!noirqdebug) @@ -310,6 +294,9 @@ void do_IRQ(struct pt_regs *regs) void do_IRQ(struct pt_regs *regs) { int irq; +#ifdef CONFIG_IRQSTACKS + struct thread_info *curtp, *irqtp; +#endif irq_enter(); @@ -330,9 +317,22 @@ void do_IRQ(struct pt_regs *regs) irq = ppc_md.get_irq(regs); - if (irq >= 0) - ppc_irq_dispatch_handler(regs, irq); - else + if (irq >= 0) { +#ifdef CONFIG_IRQSTACKS + /* Switch to the irq stack to handle this */ + curtp = current_thread_info(); + irqtp = hardirq_ctx[smp_processor_id()]; + if (curtp != irqtp) { + irqtp->task = curtp->task; + irqtp->flags = 0; + call_ppc_irq_dispatch_handler(regs, irq, irqtp); + irqtp->task = NULL; + if (irqtp->flags) + set_bits(irqtp->flags, &curtp->flags); + } else +#endif + ppc_irq_dispatch_handler(regs, irq); + } else /* That's not SMP safe ... but who cares ? */ ppc_spurious_interrupts++; diff --git a/arch/ppc64/kernel/misc.S b/arch/ppc64/kernel/misc.S index 914632ec587d..d82a30dc26f8 100644 --- a/arch/ppc64/kernel/misc.S +++ b/arch/ppc64/kernel/misc.S @@ -78,12 +78,12 @@ _GLOBAL(call_do_softirq) mtlr r0 blr -_GLOBAL(call_handle_IRQ_event) +_GLOBAL(call_ppc_irq_dispatch_handler) mflr r0 std r0,16(r1) - stdu r1,THREAD_SIZE-112(r6) - mr r1,r6 - bl .handle_IRQ_event + stdu r1,THREAD_SIZE-112(r5) + mr r1,r5 + bl .ppc_irq_dispatch_handler ld r1,0(r1) ld r0,16(r1) mtlr r0 diff --git a/include/asm-powerpc/irq.h b/include/asm-powerpc/irq.h index b3935ea28fff..2a768e096067 100644 --- a/include/asm-powerpc/irq.h +++ b/include/asm-powerpc/irq.h @@ -488,8 +488,8 @@ extern struct thread_info *softirq_ctx[NR_CPUS]; extern void irq_ctx_init(void); extern void call_do_softirq(struct thread_info *tp); -extern int call_handle_IRQ_event(int irq, struct pt_regs *regs, - struct irqaction *action, struct thread_info *tp); +extern int call_ppc_irq_dispatch_handler(struct pt_regs *regs, int irq, + struct thread_info *tp); #define __ARCH_HAS_DO_SOFTIRQ From 5a7b3ff4670be3330842558dc5ae46ec3fc448e5 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 9 Nov 2005 15:07:16 +1100 Subject: [PATCH 015/121] ppc64: allow iSeries to use IRQSTACKS again Signed-off-by: Stephen Rothwell --- arch/powerpc/platforms/iseries/irq.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/platforms/iseries/irq.c b/arch/powerpc/platforms/iseries/irq.c index a06603d84a45..a0ff7d95fdf3 100644 --- a/arch/powerpc/platforms/iseries/irq.c +++ b/arch/powerpc/platforms/iseries/irq.c @@ -103,6 +103,9 @@ static void intReceived(struct XmPciLpEvent *eventParm, struct pt_regs *regsParm) { int irq; +#ifdef CONFIG_IRQSTACKS + struct thread_info *curtp, *irqtp; +#endif ++Pci_Interrupt_Count; @@ -110,7 +113,20 @@ static void intReceived(struct XmPciLpEvent *eventParm, case XmPciLpEvent_SlotInterrupt: irq = eventParm->hvLpEvent.xCorrelationToken; /* Dispatch the interrupt handlers for this irq */ - ppc_irq_dispatch_handler(regsParm, irq); +#ifdef CONFIG_IRQSTACKS + /* Switch to the irq stack to handle this */ + curtp = current_thread_info(); + irqtp = hardirq_ctx[smp_processor_id()]; + if (curtp != irqtp) { + irqtp->task = curtp->task; + irqtp->flags = 0; + call_ppc_irq_dispatch_handler(regsParm, irq, irqtp); + irqtp->task = NULL; + if (irqtp->flags) + set_bits(irqtp->flags, &curtp->flags); + } else +#endif + ppc_irq_dispatch_handler(regsParm, irq); HvCallPci_eoi(eventParm->eventData.slotInterrupt.busNumber, eventParm->eventData.slotInterrupt.subBusNumber, eventParm->eventData.slotInterrupt.deviceId); From d4be4f37d9d2a5afc8e79a95beafbac4b83f20c5 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 9 Nov 2005 16:19:53 +1100 Subject: [PATCH 016/121] ppc64: remove ppc_irq_dispatch_handler Use __do_IRQ instead. The only difference is that every controller is now assumed to have an end() routine (only xics_8259 did not). Signed-off-by: Stephen Rothwell --- arch/powerpc/kernel/misc_64.S | 4 +- arch/powerpc/platforms/iseries/irq.c | 10 +-- arch/powerpc/platforms/pseries/xics.c | 1 + arch/ppc64/kernel/irq.c | 108 +------------------------- arch/ppc64/kernel/misc.S | 4 +- include/asm-powerpc/hw_irq.h | 1 - include/asm-powerpc/irq.h | 2 +- 7 files changed, 12 insertions(+), 118 deletions(-) diff --git a/arch/powerpc/kernel/misc_64.S b/arch/powerpc/kernel/misc_64.S index 9d09f0ad6efe..ae48a002f81a 100644 --- a/arch/powerpc/kernel/misc_64.S +++ b/arch/powerpc/kernel/misc_64.S @@ -89,12 +89,12 @@ _GLOBAL(call_do_softirq) mtlr r0 blr -_GLOBAL(call_ppc_irq_dispatch_handler) +_GLOBAL(call___do_IRQ) mflr r0 std r0,16(r1) stdu r1,THREAD_SIZE-112(r5) mr r1,r5 - bl .ppc_irq_dispatch_handler + bl .__do_IRQ ld r1,0(r1) ld r0,16(r1) mtlr r0 diff --git a/arch/powerpc/platforms/iseries/irq.c b/arch/powerpc/platforms/iseries/irq.c index a0ff7d95fdf3..01090e9ce0cf 100644 --- a/arch/powerpc/platforms/iseries/irq.c +++ b/arch/powerpc/platforms/iseries/irq.c @@ -120,13 +120,13 @@ static void intReceived(struct XmPciLpEvent *eventParm, if (curtp != irqtp) { irqtp->task = curtp->task; irqtp->flags = 0; - call_ppc_irq_dispatch_handler(regsParm, irq, irqtp); + call___do_IRQ(irq, regsParm, irqtp); irqtp->task = NULL; if (irqtp->flags) set_bits(irqtp->flags, &curtp->flags); } else #endif - ppc_irq_dispatch_handler(regsParm, irq); + __do_IRQ(irq, regsParm); HvCallPci_eoi(eventParm->eventData.slotInterrupt.busNumber, eventParm->eventData.slotInterrupt.subBusNumber, eventParm->eventData.slotInterrupt.deviceId); @@ -326,10 +326,8 @@ static void iSeries_disable_IRQ(unsigned int irq) } /* - * Need to define this so ppc_irq_dispatch_handler will NOT call - * enable_IRQ at the end of interrupt handling. However, this does - * nothing because there is not enough information provided to do - * the EOI HvCall. This is done by XmPciLpEvent.c + * This does nothing because there is not enough information + * provided to do the EOI HvCall. This is done by XmPciLpEvent.c */ static void iSeries_end_IRQ(unsigned int irq) { diff --git a/arch/powerpc/platforms/pseries/xics.c b/arch/powerpc/platforms/pseries/xics.c index c72c86f05cb6..405c4f3229b3 100644 --- a/arch/powerpc/platforms/pseries/xics.c +++ b/arch/powerpc/platforms/pseries/xics.c @@ -567,6 +567,7 @@ nextnode: xics_8259_pic.enable = i8259_pic.enable; xics_8259_pic.disable = i8259_pic.disable; + xics_8259_pic.end = i8259_pic.end; for (i = 0; i < 16; ++i) get_irq_desc(i)->handler = &xics_8259_pic; for (; i < NR_IRQS; ++i) diff --git a/arch/ppc64/kernel/irq.c b/arch/ppc64/kernel/irq.c index b61497d5be28..bd6a95a5914d 100644 --- a/arch/ppc64/kernel/irq.c +++ b/arch/ppc64/kernel/irq.c @@ -144,110 +144,6 @@ void fixup_irqs(cpumask_t map) } #endif -extern int noirqdebug; - -/* - * Eventually, this should take an array of interrupts and an array size - * so it can dispatch multiple interrupts. - */ -void ppc_irq_dispatch_handler(struct pt_regs *regs, int irq) -{ - int status; - struct irqaction *action; - int cpu = smp_processor_id(); - irq_desc_t *desc = get_irq_desc(irq); - irqreturn_t action_ret; - - kstat_cpu(cpu).irqs[irq]++; - - if (desc->status & IRQ_PER_CPU) { - /* no locking required for CPU-local interrupts: */ - ack_irq(irq); - action_ret = handle_IRQ_event(irq, regs, desc->action); - desc->handler->end(irq); - return; - } - - spin_lock(&desc->lock); - ack_irq(irq); - /* - REPLAY is when Linux resends an IRQ that was dropped earlier - WAITING is used by probe to mark irqs that are being tested - */ - status = desc->status & ~(IRQ_REPLAY | IRQ_WAITING); - status |= IRQ_PENDING; /* we _want_ to handle it */ - - /* - * If the IRQ is disabled for whatever reason, we cannot - * use the action we have. - */ - action = NULL; - if (likely(!(status & (IRQ_DISABLED | IRQ_INPROGRESS)))) { - action = desc->action; - if (!action || !action->handler) { - ppc_spurious_interrupts++; - printk(KERN_DEBUG "Unhandled interrupt %x, disabled\n", irq); - /* We can't call disable_irq here, it would deadlock */ - if (!desc->depth) - desc->depth = 1; - desc->status |= IRQ_DISABLED; - /* This is not a real spurrious interrupt, we - * have to eoi it, so we jump to out - */ - mask_irq(irq); - goto out; - } - status &= ~IRQ_PENDING; /* we commit to handling */ - status |= IRQ_INPROGRESS; /* we are handling it */ - } - desc->status = status; - - /* - * If there is no IRQ handler or it was disabled, exit early. - Since we set PENDING, if another processor is handling - a different instance of this same irq, the other processor - will take care of it. - */ - if (unlikely(!action)) - goto out; - - /* - * Edge triggered interrupts need to remember - * pending events. - * This applies to any hw interrupts that allow a second - * instance of the same irq to arrive while we are in do_IRQ - * or in the handler. But the code here only handles the _second_ - * instance of the irq, not the third or fourth. So it is mostly - * useful for irq hardware that does not mask cleanly in an - * SMP environment. - */ - for (;;) { - spin_unlock(&desc->lock); - - action_ret = handle_IRQ_event(irq, regs, action); - - spin_lock(&desc->lock); - if (!noirqdebug) - note_interrupt(irq, desc, action_ret, regs); - if (likely(!(desc->status & IRQ_PENDING))) - break; - desc->status &= ~IRQ_PENDING; - } -out: - desc->status &= ~IRQ_INPROGRESS; - /* - * The ->end() handler has to deal with interrupts which got - * disabled while the handler was running. - */ - if (desc->handler) { - if (desc->handler->end) - desc->handler->end(irq); - else if (desc->handler->enable) - desc->handler->enable(irq); - } - spin_unlock(&desc->lock); -} - #ifdef CONFIG_PPC_ISERIES void do_IRQ(struct pt_regs *regs) { @@ -325,13 +221,13 @@ void do_IRQ(struct pt_regs *regs) if (curtp != irqtp) { irqtp->task = curtp->task; irqtp->flags = 0; - call_ppc_irq_dispatch_handler(regs, irq, irqtp); + call___do_IRQ(irq, regs, irqtp); irqtp->task = NULL; if (irqtp->flags) set_bits(irqtp->flags, &curtp->flags); } else #endif - ppc_irq_dispatch_handler(regs, irq); + __do_IRQ(irq, regs); } else /* That's not SMP safe ... but who cares ? */ ppc_spurious_interrupts++; diff --git a/arch/ppc64/kernel/misc.S b/arch/ppc64/kernel/misc.S index d82a30dc26f8..492bca6137eb 100644 --- a/arch/ppc64/kernel/misc.S +++ b/arch/ppc64/kernel/misc.S @@ -78,12 +78,12 @@ _GLOBAL(call_do_softirq) mtlr r0 blr -_GLOBAL(call_ppc_irq_dispatch_handler) +_GLOBAL(call___do_IRQ) mflr r0 std r0,16(r1) stdu r1,THREAD_SIZE-112(r5) mr r1,r5 - bl .ppc_irq_dispatch_handler + bl .__do_IRQ ld r1,0(r1) ld r0,16(r1) mtlr r0 diff --git a/include/asm-powerpc/hw_irq.h b/include/asm-powerpc/hw_irq.h index c37b31b96337..26b89d859c56 100644 --- a/include/asm-powerpc/hw_irq.h +++ b/include/asm-powerpc/hw_irq.h @@ -12,7 +12,6 @@ #include extern void timer_interrupt(struct pt_regs *); -extern void ppc_irq_dispatch_handler(struct pt_regs *regs, int irq); #ifdef CONFIG_PPC_ISERIES diff --git a/include/asm-powerpc/irq.h b/include/asm-powerpc/irq.h index 2a768e096067..225dc182ef3b 100644 --- a/include/asm-powerpc/irq.h +++ b/include/asm-powerpc/irq.h @@ -488,7 +488,7 @@ extern struct thread_info *softirq_ctx[NR_CPUS]; extern void irq_ctx_init(void); extern void call_do_softirq(struct thread_info *tp); -extern int call_ppc_irq_dispatch_handler(struct pt_regs *regs, int irq, +extern int call___do_IRQ(int irq, struct pt_regs *regs, struct thread_info *tp); #define __ARCH_HAS_DO_SOFTIRQ From 756e7104fefc82e3ebaa5f1da5ba6659c9c1cae5 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Wed, 9 Nov 2005 18:07:45 +1100 Subject: [PATCH 017/121] powerpc: merge irq.c Signed-off-by: Stephen Rothwell --- arch/powerpc/kernel/Makefile | 2 +- arch/{ppc64 => powerpc}/kernel/irq.c | 125 ++++++++++++++----- arch/powerpc/platforms/powermac/pic.c | 3 + arch/ppc/kernel/Makefile | 5 +- arch/ppc/kernel/irq.c | 165 -------------------------- arch/ppc/kernel/ppc_ksyms.c | 1 - arch/ppc/platforms/pmac_pic.c | 3 + arch/ppc64/kernel/Makefile | 2 +- include/asm-powerpc/irq.h | 1 - 9 files changed, 104 insertions(+), 203 deletions(-) rename arch/{ppc64 => powerpc}/kernel/irq.c (76%) delete mode 100644 arch/ppc/kernel/irq.c diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile index b3ae2993efb8..cbdc14261655 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -11,7 +11,7 @@ CFLAGS_btext.o += -fPIC endif obj-y := semaphore.o cputable.o ptrace.o syscalls.o \ - signal_32.o pmc.o + irq.o signal_32.o pmc.o obj-$(CONFIG_PPC64) += setup_64.o binfmt_elf32.o sys_ppc32.o \ signal_64.o ptrace32.o systbl.o obj-$(CONFIG_ALTIVEC) += vecemu.o vector.o diff --git a/arch/ppc64/kernel/irq.c b/arch/powerpc/kernel/irq.c similarity index 76% rename from arch/ppc64/kernel/irq.c rename to arch/powerpc/kernel/irq.c index bd6a95a5914d..4b7940693f3d 100644 --- a/arch/ppc64/kernel/irq.c +++ b/arch/powerpc/kernel/irq.c @@ -5,12 +5,12 @@ * Copyright (C) 1992 Linus Torvalds * Adapted from arch/i386 by Gary Thomas * Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org) - * Updated and modified by Cort Dougan (cort@cs.nmt.edu) - * Copyright (C) 1996 Cort Dougan + * Updated and modified by Cort Dougan + * Copyright (C) 1996-2001 Cort Dougan * Adapted for Power Macintosh by Paul Mackerras * Copyright (C) 1996 Paul Mackerras (paulus@cs.anu.edu.au) * Amiga/APUS changes by Jesper Skov (jskov@cygnus.co.uk). - * + * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License * as published by the Free Software Foundation; either version @@ -21,6 +21,14 @@ * instead of just grabbing them. Thus setups with different IRQ numbers * shouldn't result in any weird surprises, and installing new handlers * should be easier. + * + * The MPC8xx has an interrupt mask in the SIU. If a bit is set, the + * interrupt is _enabled_. As expected, IRQ0 is bit 0 in the 32-bit + * mask register (of which only 16 are defined), hence the weird shifting + * and complement of the cached_irq_mask. I want to be able to stuff + * this right into the SIU SMASK register. + * Many of the prep/chrp functions are conditional compiled on CONFIG_8xx + * to reduce code space and undefined function references. */ #include @@ -29,6 +37,7 @@ #include #include #include +#include #include #include #include @@ -40,9 +49,13 @@ #include #include #include -#include +#include +#include #include #include +#ifdef CONFIG_PPC64 +#include +#endif #include #include @@ -52,35 +65,54 @@ #include #include #include -#include #include +#ifdef CONFIG_PPC64 +#include #include - -#ifdef CONFIG_SMP -extern void iSeries_smp_message_recv( struct pt_regs * ); #endif -extern irq_desc_t irq_desc[NR_IRQS]; +static int ppc_spurious_interrupts; + +#if defined(CONFIG_PPC_ISERIES) && defined(CONFIG_SMP) +extern void iSeries_smp_message_recv(struct pt_regs *); +#endif + +#ifdef CONFIG_PPC32 +#define NR_MASK_WORDS ((NR_IRQS + 31) / 32) + +unsigned long ppc_cached_irq_mask[NR_MASK_WORDS]; +atomic_t ppc_n_lost_interrupts; + +#ifdef CONFIG_TAU_INT +extern int tau_initialized; +extern int tau_interrupts(int); +#endif + +#if defined(CONFIG_SMP) && !defined(CONFIG_PPC_MERGE) +extern atomic_t ipi_recv; +extern atomic_t ipi_sent; +#endif +#endif /* CONFIG_PPC32 */ + +#ifdef CONFIG_PPC64 EXPORT_SYMBOL(irq_desc); int distribute_irqs = 1; int __irq_offset_value; -int ppc_spurious_interrupts; u64 ppc64_interrupt_controller; +#endif /* CONFIG_PPC64 */ int show_interrupts(struct seq_file *p, void *v) { - int i = *(loff_t *) v, j; - struct irqaction * action; + int i = *(loff_t *)v, j; + struct irqaction *action; irq_desc_t *desc; unsigned long flags; if (i == 0) { - seq_printf(p, " "); - for (j=0; jhandler) - seq_printf(p, " %s ", desc->handler->typename ); + seq_printf(p, " %s ", desc->handler->typename); else - seq_printf(p, " None "); + seq_puts(p, " None "); seq_printf(p, "%s", (desc->status & IRQ_LEVEL) ? "Level " : "Edge "); - seq_printf(p, " %s",action->name); - for (action=action->next; action; action = action->next) + seq_printf(p, " %s", action->name); + for (action = action->next; action; action = action->next) seq_printf(p, ", %s", action->name); seq_putc(p, '\n'); skip: spin_unlock_irqrestore(&desc->lock, flags); - } else if (i == NR_IRQS) + } else if (i == NR_IRQS) { +#ifdef CONFIG_PPC32 +#ifdef CONFIG_TAU_INT + if (tau_initialized){ + seq_puts(p, "TAU: "); + for (j = 0; j < NR_CPUS; j++) + if (cpu_online(j)) + seq_printf(p, "%10u ", tau_interrupts(j)); + seq_puts(p, " PowerPC Thermal Assist (cpu temp)\n"); + } +#endif +#if defined(CONFIG_SMP) && !defined(CONFIG_PPC_MERGE) + /* should this be per processor send/receive? */ + seq_printf(p, "IPI (recv/sent): %10u/%u\n", + atomic_read(&ipi_recv), atomic_read(&ipi_sent)); +#endif +#endif /* CONFIG_PPC32 */ seq_printf(p, "BAD: %10u\n", ppc_spurious_interrupts); + } return 0; } @@ -194,7 +241,7 @@ void do_IRQ(struct pt_regs *regs) struct thread_info *curtp, *irqtp; #endif - irq_enter(); + irq_enter(); #ifdef CONFIG_DEBUG_STACKOVERFLOW /* Debugging check for stack overflow: is there less than 2KB free? */ @@ -211,6 +258,13 @@ void do_IRQ(struct pt_regs *regs) } #endif + /* + * Every platform is required to implement ppc_md.get_irq. + * This function will either return an irq number or -1 to + * indicate there are no more pending. + * The value -2 is for buggy hardware and means that this IRQ + * has already been handled. -- Tom + */ irq = ppc_md.get_irq(regs); if (irq >= 0) { @@ -229,15 +283,19 @@ void do_IRQ(struct pt_regs *regs) #endif __do_IRQ(irq, regs); } else - /* That's not SMP safe ... but who cares ? */ - ppc_spurious_interrupts++; - - irq_exit(); +#ifdef CONFIG_PPC32 + if (irq != -2) +#endif + /* That's not SMP safe ... but who cares ? */ + ppc_spurious_interrupts++; + irq_exit(); } + #endif /* CONFIG_PPC_ISERIES */ void __init init_IRQ(void) { +#ifdef CONFIG_PPC64 static int once = 0; if (once) @@ -245,10 +303,14 @@ void __init init_IRQ(void) once++; +#endif ppc_md.init_IRQ(); +#ifdef CONFIG_PPC64 irq_ctx_init(); +#endif } +#ifdef CONFIG_PPC64 #ifndef CONFIG_PPC_ISERIES /* * Virtual IRQ mapping code, used on systems with XICS interrupt controllers. @@ -413,3 +475,4 @@ static int __init setup_noirqdistrib(char *str) } __setup("noirqdistrib", setup_noirqdistrib); +#endif /* CONFIG_PPC64 */ diff --git a/arch/powerpc/platforms/powermac/pic.c b/arch/powerpc/platforms/powermac/pic.c index 83a49e80ac29..90040c49494d 100644 --- a/arch/powerpc/platforms/powermac/pic.c +++ b/arch/powerpc/platforms/powermac/pic.c @@ -74,6 +74,9 @@ static DEFINE_SPINLOCK(pmac_pic_lock); #define GATWICK_IRQ_POOL_SIZE 10 static struct interrupt_info gatwick_int_pool[GATWICK_IRQ_POOL_SIZE]; +#define NR_MASK_WORDS ((NR_IRQS + 31) / 32) +static unsigned long ppc_lost_interrupts[NR_MASK_WORDS]; + /* * Mark an irq as "lost". This is only used on the pmac * since it can lose interrupts (see pmac_set_irq_mask). diff --git a/arch/ppc/kernel/Makefile b/arch/ppc/kernel/Makefile index 76a55a438f23..17a4da65e275 100644 --- a/arch/ppc/kernel/Makefile +++ b/arch/ppc/kernel/Makefile @@ -12,7 +12,7 @@ extra-$(CONFIG_6xx) += idle_6xx.o extra-$(CONFIG_POWER4) += idle_power4.o extra-y += vmlinux.lds -obj-y := entry.o traps.o irq.o idle.o time.o misc.o \ +obj-y := entry.o traps.o idle.o time.o misc.o \ process.o align.o \ setup.o \ ppc_htab.o @@ -38,8 +38,7 @@ endif # These are here while we do the architecture merge else -obj-y := irq.o idle.o \ - align.o +obj-y := idle.o align.o obj-$(CONFIG_6xx) += l2cr.o cpu_setup_6xx.o obj-$(CONFIG_SOFTWARE_SUSPEND) += swsusp.o obj-$(CONFIG_MODULES) += module.o diff --git a/arch/ppc/kernel/irq.c b/arch/ppc/kernel/irq.c deleted file mode 100644 index fbb2b9f8922c..000000000000 --- a/arch/ppc/kernel/irq.c +++ /dev/null @@ -1,165 +0,0 @@ -/* - * arch/ppc/kernel/irq.c - * - * Derived from arch/i386/kernel/irq.c - * Copyright (C) 1992 Linus Torvalds - * Adapted from arch/i386 by Gary Thomas - * Copyright (C) 1995-1996 Gary Thomas (gdt@linuxppc.org) - * Updated and modified by Cort Dougan - * Copyright (C) 1996-2001 Cort Dougan - * Adapted for Power Macintosh by Paul Mackerras - * Copyright (C) 1996 Paul Mackerras (paulus@cs.anu.edu.au) - * Amiga/APUS changes by Jesper Skov (jskov@cygnus.co.uk). - * - * This file contains the code used by various IRQ handling routines: - * asking for different IRQ's should be done through these routines - * instead of just grabbing them. Thus setups with different IRQ numbers - * shouldn't result in any weird surprises, and installing new handlers - * should be easier. - * - * The MPC8xx has an interrupt mask in the SIU. If a bit is set, the - * interrupt is _enabled_. As expected, IRQ0 is bit 0 in the 32-bit - * mask register (of which only 16 are defined), hence the weird shifting - * and complement of the cached_irq_mask. I want to be able to stuff - * this right into the SIU SMASK register. - * Many of the prep/chrp functions are conditional compiled on CONFIG_8xx - * to reduce code space and undefined function references. - */ - -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#include -#include -#include -#include -#include -#include -#include -#include -#include - -#define NR_MASK_WORDS ((NR_IRQS + 31) / 32) - -extern atomic_t ipi_recv; -extern atomic_t ipi_sent; - -#define MAXCOUNT 10000000 - -int ppc_spurious_interrupts = 0; -struct irqaction *ppc_irq_action[NR_IRQS]; -unsigned long ppc_cached_irq_mask[NR_MASK_WORDS]; -unsigned long ppc_lost_interrupts[NR_MASK_WORDS]; -atomic_t ppc_n_lost_interrupts; - -#ifdef CONFIG_TAU_INT -extern int tau_initialized; -extern int tau_interrupts(int); -#endif - -int show_interrupts(struct seq_file *p, void *v) -{ - int i = *(loff_t *) v, j; - struct irqaction * action; - unsigned long flags; - - if (i == 0) { - seq_puts(p, " "); - for (j=0; jhandler ) - goto skip; - seq_printf(p, "%3d: ", i); -#ifdef CONFIG_SMP - for (j = 0; j < NR_CPUS; j++) - if (cpu_online(j)) - seq_printf(p, "%10u ", - kstat_cpu(j).irqs[i]); -#else - seq_printf(p, "%10u ", kstat_irqs(i)); -#endif /* CONFIG_SMP */ - if (irq_desc[i].handler) - seq_printf(p, " %s ", irq_desc[i].handler->typename); - else - seq_puts(p, " None "); - seq_printf(p, "%s", (irq_desc[i].status & IRQ_LEVEL) ? "Level " : "Edge "); - seq_printf(p, " %s", action->name); - for (action = action->next; action; action = action->next) - seq_printf(p, ", %s", action->name); - seq_putc(p, '\n'); -skip: - spin_unlock_irqrestore(&irq_desc[i].lock, flags); - } else if (i == NR_IRQS) { -#ifdef CONFIG_TAU_INT - if (tau_initialized){ - seq_puts(p, "TAU: "); - for (j = 0; j < NR_CPUS; j++) - if (cpu_online(j)) - seq_printf(p, "%10u ", tau_interrupts(j)); - seq_puts(p, " PowerPC Thermal Assist (cpu temp)\n"); - } -#endif -#if defined(CONFIG_SMP) && !defined(CONFIG_PPC_MERGE) - /* should this be per processor send/receive? */ - seq_printf(p, "IPI (recv/sent): %10u/%u\n", - atomic_read(&ipi_recv), atomic_read(&ipi_sent)); -#endif - seq_printf(p, "BAD: %10u\n", ppc_spurious_interrupts); - } - return 0; -} - -void do_IRQ(struct pt_regs *regs) -{ - int irq, first = 1; - irq_enter(); - - /* - * Every platform is required to implement ppc_md.get_irq. - * This function will either return an irq number or -1 to - * indicate there are no more pending. But the first time - * through the loop this means there wasn't and IRQ pending. - * The value -2 is for buggy hardware and means that this IRQ - * has already been handled. -- Tom - */ - while ((irq = ppc_md.get_irq(regs)) >= 0) { - __do_IRQ(irq, regs); - first = 0; - } - if (irq != -2 && first) - /* That's not SMP safe ... but who cares ? */ - ppc_spurious_interrupts++; - irq_exit(); -} - -void __init init_IRQ(void) -{ - ppc_md.init_IRQ(); -} diff --git a/arch/ppc/kernel/ppc_ksyms.c b/arch/ppc/kernel/ppc_ksyms.c index e0ca61b37f4f..c4d5b5988430 100644 --- a/arch/ppc/kernel/ppc_ksyms.c +++ b/arch/ppc/kernel/ppc_ksyms.c @@ -78,7 +78,6 @@ EXPORT_SYMBOL(program_check_exception); EXPORT_SYMBOL(single_step_exception); EXPORT_SYMBOL(sys_sigreturn); EXPORT_SYMBOL(ppc_n_lost_interrupts); -EXPORT_SYMBOL(ppc_lost_interrupts); EXPORT_SYMBOL(ISA_DMA_THRESHOLD); EXPORT_SYMBOL(DMA_MODE_READ); diff --git a/arch/ppc/platforms/pmac_pic.c b/arch/ppc/platforms/pmac_pic.c index 9f2d95ea8564..4742bf609357 100644 --- a/arch/ppc/platforms/pmac_pic.c +++ b/arch/ppc/platforms/pmac_pic.c @@ -75,6 +75,9 @@ static DEFINE_SPINLOCK(pmac_pic_lock); #define GATWICK_IRQ_POOL_SIZE 10 static struct interrupt_info gatwick_int_pool[GATWICK_IRQ_POOL_SIZE]; +#define NR_MASK_WORDS ((NR_IRQS + 31) / 32) +static unsigned long ppc_lost_interrupts[NR_MASK_WORDS]; + /* * Mark an irq as "lost". This is only used on the pmac * since it can lose interrupts (see pmac_set_irq_mask). diff --git a/arch/ppc64/kernel/Makefile b/arch/ppc64/kernel/Makefile index c441aebe7648..13e479846249 100644 --- a/arch/ppc64/kernel/Makefile +++ b/arch/ppc64/kernel/Makefile @@ -11,7 +11,7 @@ obj-y := misc.o prom.o endif -obj-y += irq.o idle.o dma.o \ +obj-y += idle.o dma.o \ align.o pacaData.o \ udbg.o ioctl32.o \ rtc.o \ diff --git a/include/asm-powerpc/irq.h b/include/asm-powerpc/irq.h index 225dc182ef3b..c9fbcede0ef9 100644 --- a/include/asm-powerpc/irq.h +++ b/include/asm-powerpc/irq.h @@ -429,7 +429,6 @@ extern u64 ppc64_interrupt_controller; #define NR_MASK_WORDS ((NR_IRQS + 31) / 32) /* pedantic: these are long because they are used with set_bit --RR */ extern unsigned long ppc_cached_irq_mask[NR_MASK_WORDS]; -extern unsigned long ppc_lost_interrupts[NR_MASK_WORDS]; extern atomic_t ppc_n_lost_interrupts; #define virt_irq_create_mapping(x) (x) From ac4d5f74a9b243d9f3f123fe5ce609478df208d8 Mon Sep 17 00:00:00 2001 From: Uwe Zeisberger Date: Wed, 9 Nov 2005 15:54:08 +0100 Subject: [PATCH 018/121] [PATCH] kbuild: make kernelrelease in unconfigured kernel prints an error Do not include .config for target kernelrelease Signed-off-by: Uwe Zeisberger Signed-off-by: Sam Ravnborg --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 2dac8010c142..3152d6377eb2 100644 --- a/Makefile +++ b/Makefile @@ -407,7 +407,7 @@ outputmakefile: # of make so .config is not included in this case either (for *config). no-dot-config-targets := clean mrproper distclean \ - cscope TAGS tags help %docs check% + cscope TAGS tags help %docs check% kernelrelease config-targets := 0 mixed-targets := 0 From 4349d5cdf28d46c22d467079e80ee9695dcec0ad Mon Sep 17 00:00:00 2001 From: Hannes Reinecke Date: Wed, 9 Nov 2005 22:47:18 +0100 Subject: [PATCH 019/121] [PATCH] ide: incorrect device link for ide-cs Devices driven by ide-cs will appear under /sys/devices instead of the appropriate PCMCIA device. To fix this I had to extend the hw_regs_t structure with a 'struct device' field, which allows us to set the parent link for the appropriate hwif. Signed-off-by: Hannes Reinecke Signed-off-by: Jens Axboe Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/ide/ide.c | 1 + drivers/ide/legacy/ide-cs.c | 7 ++++--- include/linux/ide.h | 1 + 3 files changed, 6 insertions(+), 3 deletions(-) diff --git a/drivers/ide/ide.c b/drivers/ide/ide.c index 9fe19808d815..8af179b531c3 100644 --- a/drivers/ide/ide.c +++ b/drivers/ide/ide.c @@ -803,6 +803,7 @@ found: hwif->irq = hw->irq; hwif->noprobe = 0; hwif->chipset = hw->chipset; + hwif->gendev.parent = hw->dev; if (!initializing) { probe_hwif_init_with_fixup(hwif, fixup); diff --git a/drivers/ide/legacy/ide-cs.c b/drivers/ide/legacy/ide-cs.c index 1dafffa7e513..ef79805218e4 100644 --- a/drivers/ide/legacy/ide-cs.c +++ b/drivers/ide/legacy/ide-cs.c @@ -182,13 +182,14 @@ static void ide_detach(dev_link_t *link) } /* ide_detach */ -static int idecs_register(unsigned long io, unsigned long ctl, unsigned long irq) +static int idecs_register(unsigned long io, unsigned long ctl, unsigned long irq, struct pcmcia_device *handle) { hw_regs_t hw; memset(&hw, 0, sizeof(hw)); ide_init_hwif_ports(&hw, io, ctl, NULL); hw.irq = irq; hw.chipset = ide_pci; + hw.dev = &handle->dev; return ide_register_hw_with_fixup(&hw, NULL, ide_undecoded_slave); } @@ -327,12 +328,12 @@ static void ide_config(dev_link_t *link) /* retry registration in case device is still spinning up */ for (hd = -1, i = 0; i < 10; i++) { - hd = idecs_register(io_base, ctl_base, link->irq.AssignedIRQ); + hd = idecs_register(io_base, ctl_base, link->irq.AssignedIRQ, handle); if (hd >= 0) break; if (link->io.NumPorts1 == 0x20) { outb(0x02, ctl_base + 0x10); hd = idecs_register(io_base + 0x10, ctl_base + 0x10, - link->irq.AssignedIRQ); + link->irq.AssignedIRQ, handle); if (hd >= 0) { io_base += 0x10; ctl_base += 0x10; diff --git a/include/linux/ide.h b/include/linux/ide.h index 3461abc1e854..e0cf4811da08 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -230,6 +230,7 @@ typedef struct hw_regs_s { int dma; /* our dma entry */ ide_ack_intr_t *ack_intr; /* acknowledge interrupt */ hwif_chipset_t chipset; + struct device *dev; } hw_regs_t; /* From 83ae20c849f1bff32100fa8584408dca99ce4a62 Mon Sep 17 00:00:00 2001 From: Bjorn Helgaas Date: Wed, 9 Nov 2005 22:58:07 +0100 Subject: [PATCH 020/121] [PATCH] ide: move CONFIG_IDE_MAX_HWIFS into linux/ide.h CONFIG_IDE_MAX_HWIFS is a generic thing, no need to have it duplicated by every arch that uses it. Signed-off-by: Bjorn Helgaas Signed-off-by: Bartlomiej Zolnierkiewicz --- include/asm-alpha/ide.h | 4 ---- include/asm-sh/ide.h | 4 ---- include/asm-sh64/ide.h | 4 ---- include/linux/ide.h | 4 ++++ 4 files changed, 4 insertions(+), 12 deletions(-) diff --git a/include/asm-alpha/ide.h b/include/asm-alpha/ide.h index 68934a25931f..6126afe27380 100644 --- a/include/asm-alpha/ide.h +++ b/include/asm-alpha/ide.h @@ -15,10 +15,6 @@ #include -#ifndef MAX_HWIFS -#define MAX_HWIFS CONFIG_IDE_MAX_HWIFS -#endif - #define IDE_ARCH_OBSOLETE_DEFAULTS static inline int ide_default_irq(unsigned long base) diff --git a/include/asm-sh/ide.h b/include/asm-sh/ide.h index f42cf3977a57..711dad4cb48b 100644 --- a/include/asm-sh/ide.h +++ b/include/asm-sh/ide.h @@ -16,10 +16,6 @@ #include -#ifndef MAX_HWIFS -#define MAX_HWIFS CONFIG_IDE_MAX_HWIFS -#endif - #define ide_default_io_ctl(base) (0) #include diff --git a/include/asm-sh64/ide.h b/include/asm-sh64/ide.h index 6fd514daa1ba..852f50afe39c 100644 --- a/include/asm-sh64/ide.h +++ b/include/asm-sh64/ide.h @@ -17,10 +17,6 @@ #include -#ifndef MAX_HWIFS -#define MAX_HWIFS CONFIG_IDE_MAX_HWIFS -#endif - /* Without this, the initialisation of PCI IDE cards end up calling * ide_init_hwif_ports, which won't work. */ #ifdef CONFIG_BLK_DEV_IDEPCI diff --git a/include/linux/ide.h b/include/linux/ide.h index e0cf4811da08..77ae55d4c13c 100644 --- a/include/linux/ide.h +++ b/include/linux/ide.h @@ -267,6 +267,10 @@ static inline void ide_std_init_ports(hw_regs_t *hw, #include +#ifndef MAX_HWIFS +#define MAX_HWIFS CONFIG_IDE_MAX_HWIFS +#endif + /* needed on alpha, x86/x86_64, ia64, mips, ppc32 and sh */ #ifndef IDE_ARCH_OBSOLETE_DEFAULTS # define ide_default_io_base(index) (0) From e851b620e775c9f0e00b1c7e35072d6b98097cb9 Mon Sep 17 00:00:00 2001 From: Adrian Bunk Date: Wed, 9 Nov 2005 23:07:56 +0100 Subject: [PATCH 021/121] [PATCH] ide: possible cleanups This patch contains the following possible cleanups: - pci/cy82c693.c: make a needlessly global function static - remove the following unneeded EXPORT_SYMBOL's: - ide-taskfile.c: do_rw_taskfile - ide-iops.c: default_hwif_iops - ide-iops.c: default_hwif_transport - ide-iops.c: wait_for_ready Signed-off-by: Adrian Bunk Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/ide/ide-iops.c | 6 ------ drivers/ide/ide-taskfile.c | 2 -- drivers/ide/pci/cy82c693.c | 2 +- 3 files changed, 1 insertion(+), 9 deletions(-) diff --git a/drivers/ide/ide-iops.c b/drivers/ide/ide-iops.c index 0b0aa4f51628..af7af958ab3e 100644 --- a/drivers/ide/ide-iops.c +++ b/drivers/ide/ide-iops.c @@ -104,8 +104,6 @@ void default_hwif_iops (ide_hwif_t *hwif) hwif->INSL = ide_insl; } -EXPORT_SYMBOL(default_hwif_iops); - /* * MMIO operations, typically used for SATA controllers */ @@ -329,8 +327,6 @@ void default_hwif_transport(ide_hwif_t *hwif) hwif->atapi_output_bytes = atapi_output_bytes; } -EXPORT_SYMBOL(default_hwif_transport); - /* * Beginning of Taskfile OPCODE Library and feature sets. */ @@ -529,8 +525,6 @@ int wait_for_ready (ide_drive_t *drive, int timeout) return 0; } -EXPORT_SYMBOL(wait_for_ready); - /* * This routine busy-waits for the drive status to be not "busy". * It then checks the status for all of the "good" bits and none diff --git a/drivers/ide/ide-taskfile.c b/drivers/ide/ide-taskfile.c index 7ec18fa3b5ff..54f9639c2a8c 100644 --- a/drivers/ide/ide-taskfile.c +++ b/drivers/ide/ide-taskfile.c @@ -161,8 +161,6 @@ ide_startstop_t do_rw_taskfile (ide_drive_t *drive, ide_task_t *task) return ide_stopped; } -EXPORT_SYMBOL(do_rw_taskfile); - /* * set_multmode_intr() is invoked on completion of a WIN_SETMULT cmd. */ diff --git a/drivers/ide/pci/cy82c693.c b/drivers/ide/pci/cy82c693.c index 5a33513f3dd1..9f41ecd56338 100644 --- a/drivers/ide/pci/cy82c693.c +++ b/drivers/ide/pci/cy82c693.c @@ -469,7 +469,7 @@ static void __devinit init_hwif_cy82c693(ide_hwif_t *hwif) static __devinitdata ide_hwif_t *primary; -void __devinit init_iops_cy82c693(ide_hwif_t *hwif) +static void __devinit init_iops_cy82c693(ide_hwif_t *hwif) { if (PCI_FUNC(hwif->pci_dev->devfn) == 1) primary = hwif; From 0046b06a367cd853efd3223ce60143f3a7952522 Mon Sep 17 00:00:00 2001 From: Willem Riede Date: Wed, 9 Nov 2005 23:18:22 +0100 Subject: [PATCH 022/121] [PATCH] ide: ide-scsi fails to call idescsi_check_condition for things like "Medium not present" This patch started life as a response to fedora specific ide subsystem changes that made error handling of my ATAPI tape drive fail; the specifics are in https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=160868 The insertion of the statement rq->errors = err; near the end of ide_end_drive_cmd() in drivers/ide/ide-io.c means that rq->errors does not contain what it needs to in idescsi_end_request() in drivers/scsi/ide-scsi.c anymore. Recent mainline kernels now also have this change. The patch below makes ide-scsi whole. Signed-off-by: Willem Riede Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/scsi/ide-scsi.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/ide-scsi.c b/drivers/scsi/ide-scsi.c index c888af4a4562..3553da0e1cd5 100644 --- a/drivers/scsi/ide-scsi.c +++ b/drivers/scsi/ide-scsi.c @@ -395,6 +395,7 @@ static int idescsi_end_request (ide_drive_t *drive, int uptodate, int nrsecs) int log = test_bit(IDESCSI_LOG_CMD, &scsi->log); struct Scsi_Host *host; u8 *scsi_buf; + int errors = rq->errors; unsigned long flags; if (!(rq->flags & (REQ_SPECIAL|REQ_SENSE))) { @@ -421,11 +422,11 @@ static int idescsi_end_request (ide_drive_t *drive, int uptodate, int nrsecs) printk (KERN_WARNING "ide-scsi: %s: timed out for %lu\n", drive->name, pc->scsi_cmd->serial_number); pc->scsi_cmd->result = DID_TIME_OUT << 16; - } else if (rq->errors >= ERROR_MAX) { + } else if (errors >= ERROR_MAX) { pc->scsi_cmd->result = DID_ERROR << 16; if (log) printk ("ide-scsi: %s: I/O error for %lu\n", drive->name, pc->scsi_cmd->serial_number); - } else if (rq->errors) { + } else if (errors) { if (log) printk ("ide-scsi: %s: check condition for %lu\n", drive->name, pc->scsi_cmd->serial_number); if (!idescsi_check_condition(drive, rq)) From 7fab773de16ccaeb249acdc6e956a9759c68225d Mon Sep 17 00:00:00 2001 From: Jordan Crouse Date: Wed, 9 Nov 2005 23:26:09 +0100 Subject: [PATCH 023/121] [PATCH] ide: AMD Geode GX/LX support From: "Jordan Crouse" The core IDE engine on the CS5536 is the same as the other AMD southbridges, so unlike the CS5535, we can simply add the appropriate PCI headers to the existing amd74xx code. Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/ide/pci/amd74xx.c | 3 +++ include/linux/pci_ids.h | 2 ++ 2 files changed, 5 insertions(+) diff --git a/drivers/ide/pci/amd74xx.c b/drivers/ide/pci/amd74xx.c index 844a6c9fb949..21965e5ef25e 100644 --- a/drivers/ide/pci/amd74xx.c +++ b/drivers/ide/pci/amd74xx.c @@ -74,6 +74,7 @@ static struct amd_ide_chip { { PCI_DEVICE_ID_NVIDIA_NFORCE_MCP04_IDE, 0x50, AMD_UDMA_133 }, { PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_IDE, 0x50, AMD_UDMA_133 }, { PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_IDE, 0x50, AMD_UDMA_133 }, + { PCI_DEVICE_ID_AMD_CS5536_IDE, 0x40, AMD_UDMA_100 }, { 0 } }; @@ -491,6 +492,7 @@ static ide_pci_device_t amd74xx_chipsets[] __devinitdata = { /* 14 */ DECLARE_NV_DEV("NFORCE-MCP04"), /* 15 */ DECLARE_NV_DEV("NFORCE-MCP51"), /* 16 */ DECLARE_NV_DEV("NFORCE-MCP55"), + /* 17 */ DECLARE_AMD_DEV("AMD5536"), }; static int __devinit amd74xx_probe(struct pci_dev *dev, const struct pci_device_id *id) @@ -527,6 +529,7 @@ static struct pci_device_id amd74xx_pci_tbl[] = { { PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP04_IDE, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 14 }, { PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP51_IDE, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 15 }, { PCI_VENDOR_ID_NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP55_IDE, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 16 }, + { PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_CS5536_IDE, PCI_ANY_ID, PCI_ANY_ID, 0, 0, 17 }, { 0, }, }; MODULE_DEVICE_TABLE(pci, amd74xx_pci_tbl); diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 9a96f0588393..63dd184ec808 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -487,6 +487,8 @@ #define PCI_DEVICE_ID_AMD_8151_0 0x7454 #define PCI_DEVICE_ID_AMD_8131_APIC 0x7450 +#define PCI_DEVICE_ID_AMD_CS5536_IDE 0x209A + #define PCI_VENDOR_ID_TRIDENT 0x1023 #define PCI_DEVICE_ID_TRIDENT_4DWAVE_DX 0x2000 #define PCI_DEVICE_ID_TRIDENT_4DWAVE_NX 0x2001 From 6730c3c14421b7c924d06e31bb66e0adad225547 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Wed, 9 Nov 2005 14:56:00 -0800 Subject: [PATCH 024/121] Fix AGP compile on non-x86 architectures AGP shouldn't use "global_flush_tlb()" to flush the AGP mappings, that i spurely an x86'ism. The proper AGP mapping flusher that should be used is "flush_agp_mappings()", which on x86 obviously happens to do a global TLB flush. This makes AGP (or at least the config _I_ happen to use) compile again on ppc64. Signed-off-by: Linus Torvalds --- drivers/char/agp/backend.c | 6 +++--- drivers/char/agp/generic.c | 4 +--- 2 files changed, 4 insertions(+), 6 deletions(-) diff --git a/drivers/char/agp/backend.c b/drivers/char/agp/backend.c index 4d5ed18dad00..27bca34b4a65 100644 --- a/drivers/char/agp/backend.c +++ b/drivers/char/agp/backend.c @@ -147,7 +147,7 @@ static int agp_backend_initialize(struct agp_bridge_data *bridge) printk(KERN_ERR PFX "unable to get memory for scratch page.\n"); return -ENOMEM; } - global_flush_tlb(); + flush_agp_mappings(); bridge->scratch_page_real = virt_to_gart(addr); bridge->scratch_page = @@ -191,7 +191,7 @@ err_out: if (bridge->driver->needs_scratch_page) { bridge->driver->agp_destroy_page( gart_to_virt(bridge->scratch_page_real)); - global_flush_tlb(); + flush_agp_mappings(); } if (got_gatt) bridge->driver->free_gatt_table(bridge); @@ -217,7 +217,7 @@ static void agp_backend_cleanup(struct agp_bridge_data *bridge) bridge->driver->needs_scratch_page) { bridge->driver->agp_destroy_page( gart_to_virt(bridge->scratch_page_real)); - global_flush_tlb(); + flush_agp_mappings(); } } diff --git a/drivers/char/agp/generic.c b/drivers/char/agp/generic.c index 19f242b50781..5567ce8d72b0 100644 --- a/drivers/char/agp/generic.c +++ b/drivers/char/agp/generic.c @@ -155,7 +155,7 @@ void agp_free_memory(struct agp_memory *curr) for (i = 0; i < curr->page_count; i++) { curr->bridge->driver->agp_destroy_page(gart_to_virt(curr->memory[i])); } - global_flush_tlb(); + flush_agp_mappings(); } agp_free_key(curr->key); vfree(curr->memory); @@ -213,8 +213,6 @@ struct agp_memory *agp_allocate_memory(struct agp_bridge_data *bridge, new->memory[i] = virt_to_gart(addr); new->page_count++; } - global_flush_tlb(); - new->bridge = bridge; flush_agp_mappings(); From f5b2d8b4b5146fa2d70fec7d514fa0bd64636958 Mon Sep 17 00:00:00 2001 From: Jaya Kumar Date: Wed, 9 Nov 2005 23:58:16 +0100 Subject: [PATCH 025/121] [PATCH] ide: CS5535 driver Signed-off-by: Jaya Kumar Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/ide/Kconfig | 9 ++ drivers/ide/pci/Makefile | 1 + drivers/ide/pci/cs5535.c | 305 +++++++++++++++++++++++++++++++++++++++ include/linux/pci_ids.h | 1 + 4 files changed, 316 insertions(+) create mode 100644 drivers/ide/pci/cs5535.c diff --git a/drivers/ide/Kconfig b/drivers/ide/Kconfig index a737886e39d1..42e5b8175cbf 100644 --- a/drivers/ide/Kconfig +++ b/drivers/ide/Kconfig @@ -539,6 +539,15 @@ config BLK_DEV_CS5530 It is safe to say Y to this question. +config BLK_DEV_CS5535 + tristate "AMD CS5535 chipset support" + depends on X86 && !X86_64 + help + Include support for UDMA on the NSC/AMD CS5535 companion chipset. + This will automatically be detected and configured if found. + + It is safe to say Y to this question. + config BLK_DEV_HPT34X tristate "HPT34X chipset support" help diff --git a/drivers/ide/pci/Makefile b/drivers/ide/pci/Makefile index af46226c1796..f35d684edc25 100644 --- a/drivers/ide/pci/Makefile +++ b/drivers/ide/pci/Makefile @@ -6,6 +6,7 @@ obj-$(CONFIG_BLK_DEV_ATIIXP) += atiixp.o obj-$(CONFIG_BLK_DEV_CMD64X) += cmd64x.o obj-$(CONFIG_BLK_DEV_CS5520) += cs5520.o obj-$(CONFIG_BLK_DEV_CS5530) += cs5530.o +obj-$(CONFIG_BLK_DEV_CS5535) += cs5535.o obj-$(CONFIG_BLK_DEV_SC1200) += sc1200.o obj-$(CONFIG_BLK_DEV_CY82C693) += cy82c693.o obj-$(CONFIG_BLK_DEV_HPT34X) += hpt34x.o diff --git a/drivers/ide/pci/cs5535.c b/drivers/ide/pci/cs5535.c new file mode 100644 index 000000000000..6eb305197f3c --- /dev/null +++ b/drivers/ide/pci/cs5535.c @@ -0,0 +1,305 @@ +/* + * linux/drivers/ide/pci/cs5535.c + * + * Copyright (C) 2004-2005 Advanced Micro Devices, Inc. + * + * History: + * 09/20/2005 - Jaya Kumar + * - Reworked tuneproc, set_drive, misc mods to prep for mainline + * - Work was sponsored by CIS (M) Sdn Bhd. + * Ported to Kernel 2.6.11 on June 26, 2005 by + * Wolfgang Zuleger + * Alexander Kiausch + * Originally developed by AMD for 2.4/2.6 + * + * Development of this chipset driver was funded + * by the nice folks at National Semiconductor/AMD. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + * + * Documentation: + * CS5535 documentation available from AMD + */ + +#include +#include +#include +#include + +#include "ide-timing.h" + +#define MSR_ATAC_BASE 0x51300000 +#define ATAC_GLD_MSR_CAP (MSR_ATAC_BASE+0) +#define ATAC_GLD_MSR_CONFIG (MSR_ATAC_BASE+0x01) +#define ATAC_GLD_MSR_SMI (MSR_ATAC_BASE+0x02) +#define ATAC_GLD_MSR_ERROR (MSR_ATAC_BASE+0x03) +#define ATAC_GLD_MSR_PM (MSR_ATAC_BASE+0x04) +#define ATAC_GLD_MSR_DIAG (MSR_ATAC_BASE+0x05) +#define ATAC_IO_BAR (MSR_ATAC_BASE+0x08) +#define ATAC_RESET (MSR_ATAC_BASE+0x10) +#define ATAC_CH0D0_PIO (MSR_ATAC_BASE+0x20) +#define ATAC_CH0D0_DMA (MSR_ATAC_BASE+0x21) +#define ATAC_CH0D1_PIO (MSR_ATAC_BASE+0x22) +#define ATAC_CH0D1_DMA (MSR_ATAC_BASE+0x23) +#define ATAC_PCI_ABRTERR (MSR_ATAC_BASE+0x24) +#define ATAC_BM0_CMD_PRIM 0x00 +#define ATAC_BM0_STS_PRIM 0x02 +#define ATAC_BM0_PRD 0x04 +#define CS5535_CABLE_DETECT 0x48 + +/* Format I PIO settings. We seperate out cmd and data for safer timings */ + +static unsigned int cs5535_pio_cmd_timings[5] = +{ 0xF7F4, 0x53F3, 0x13F1, 0x5131, 0x1131 }; +static unsigned int cs5535_pio_dta_timings[5] = +{ 0xF7F4, 0xF173, 0x8141, 0x5131, 0x1131 }; + +static unsigned int cs5535_mwdma_timings[3] = +{ 0x7F0FFFF3, 0x7F035352, 0x7f024241 }; + +static unsigned int cs5535_udma_timings[5] = +{ 0x7F7436A1, 0x7F733481, 0x7F723261, 0x7F713161, 0x7F703061 }; + +/* Macros to check if the register is the reset value - reset value is an + invalid timing and indicates the register has not been set previously */ + +#define CS5535_BAD_PIO(timings) ( (timings&~0x80000000UL) == 0x00009172 ) +#define CS5535_BAD_DMA(timings) ( (timings & 0x000FFFFF) == 0x00077771 ) + +/**** + * cs5535_set_speed - Configure the chipset to the new speed + * @drive: Drive to set up + * @speed: desired speed + * + * cs5535_set_speed() configures the chipset to a new speed. + */ +static void cs5535_set_speed(ide_drive_t *drive, u8 speed) +{ + + u32 reg = 0, dummy; + int unit = drive->select.b.unit; + + + /* Set the PIO timings */ + if ((speed & XFER_MODE) == XFER_PIO) { + u8 pioa; + u8 piob; + u8 cmd; + + pioa = speed - XFER_PIO_0; + piob = ide_get_best_pio_mode(&(drive->hwif->drives[!unit]), + 255, 4, NULL); + cmd = pioa < piob ? pioa : piob; + + /* Write the speed of the current drive */ + reg = (cs5535_pio_cmd_timings[cmd] << 16) | + cs5535_pio_dta_timings[pioa]; + wrmsr(unit ? ATAC_CH0D1_PIO : ATAC_CH0D0_PIO, reg, 0); + + /* And if nessesary - change the speed of the other drive */ + rdmsr(unit ? ATAC_CH0D0_PIO : ATAC_CH0D1_PIO, reg, dummy); + + if (((reg >> 16) & cs5535_pio_cmd_timings[cmd]) != + cs5535_pio_cmd_timings[cmd]) { + reg &= 0x0000FFFF; + reg |= cs5535_pio_cmd_timings[cmd] << 16; + wrmsr(unit ? ATAC_CH0D0_PIO : ATAC_CH0D1_PIO, reg, 0); + } + + /* Set bit 31 of the DMA register for PIO format 1 timings */ + rdmsr(unit ? ATAC_CH0D1_DMA : ATAC_CH0D0_DMA, reg, dummy); + wrmsr(unit ? ATAC_CH0D1_DMA : ATAC_CH0D0_DMA, + reg | 0x80000000UL, 0); + } else { + rdmsr(unit ? ATAC_CH0D1_DMA : ATAC_CH0D0_DMA, reg, dummy); + + reg &= 0x80000000UL; /* Preserve the PIO format bit */ + + if (speed >= XFER_UDMA_0 && speed <= XFER_UDMA_7) + reg |= cs5535_udma_timings[speed - XFER_UDMA_0]; + else if (speed >= XFER_MW_DMA_0 && speed <= XFER_MW_DMA_2) + reg |= cs5535_mwdma_timings[speed - XFER_MW_DMA_0]; + else + return; + + wrmsr(unit ? ATAC_CH0D1_DMA : ATAC_CH0D0_DMA, reg, 0); + } +} + +static u8 cs5535_ratemask(ide_drive_t *drive) +{ + /* eighty93 will return 1 if it's 80core and capable of + exceeding udma2, 0 otherwise. we need ratemask to set + the max speed and if we can > udma2 then we return 2 + which selects speed_max as udma4 which is the 5535's max + speed, and 1 selects udma2 which is the max for 40c */ + if (!eighty_ninty_three(drive)) + return 1; + + return 2; +} + + +/**** + * cs5535_set_drive - Configure the drive to the new speed + * @drive: Drive to set up + * @speed: desired speed + * + * cs5535_set_drive() configures the drive and the chipset to a + * new speed. It also can be called by upper layers. + */ +static int cs5535_set_drive(ide_drive_t *drive, u8 speed) +{ + speed = ide_rate_filter(cs5535_ratemask(drive), speed); + ide_config_drive_speed(drive, speed); + cs5535_set_speed(drive, speed); + + return 0; +} + +/**** + * cs5535_tuneproc - PIO setup + * @drive: drive to set up + * @pio: mode to use (255 for 'best possible') + * + * A callback from the upper layers for PIO-only tuning. + */ +static void cs5535_tuneproc(ide_drive_t *drive, u8 xferspeed) +{ + u8 modes[] = { XFER_PIO_0, XFER_PIO_1, XFER_PIO_2, XFER_PIO_3, + XFER_PIO_4 }; + + /* cs5535 max pio is pio 4, best_pio will check the blacklist. + i think we don't need to rate_filter the incoming xferspeed + since we know we're only going to choose pio */ + xferspeed = ide_get_best_pio_mode(drive, xferspeed, 4, NULL); + ide_config_drive_speed(drive, modes[xferspeed]); + cs5535_set_speed(drive, xferspeed); +} + +static int cs5535_config_drive_for_dma(ide_drive_t *drive) +{ + u8 speed; + + speed = ide_dma_speed(drive, cs5535_ratemask(drive)); + + /* If no DMA speed was available then let dma_check hit pio */ + if (!speed) { + return 0; + } + + cs5535_set_drive(drive, speed); + return ide_dma_enable(drive); +} + +static int cs5535_dma_check(ide_drive_t *drive) +{ + ide_hwif_t *hwif = drive->hwif; + struct hd_driveid *id = drive->id; + u8 speed; + + drive->init_speed = 0; + + if ((id->capability & 1) && drive->autodma) { + if (ide_use_dma(drive)) { + if (cs5535_config_drive_for_dma(drive)) + return hwif->ide_dma_on(drive); + } + + goto fast_ata_pio; + + } else if ((id->capability & 8) || (id->field_valid & 2)) { +fast_ata_pio: + speed = ide_get_best_pio_mode(drive, 255, 4, NULL); + cs5535_set_drive(drive, speed); + return hwif->ide_dma_off_quietly(drive); + } + /* IORDY not supported */ + return 0; +} + +static u8 __devinit cs5535_cable_detect(struct pci_dev *dev) +{ + u8 bit; + + /* if a 80 wire cable was detected */ + pci_read_config_byte(dev, CS5535_CABLE_DETECT, &bit); + return (bit & 1); +} + +/**** + * init_hwif_cs5535 - Initialize one ide cannel + * @hwif: Channel descriptor + * + * This gets invoked by the IDE driver once for each channel. It + * performs channel-specific pre-initialization before drive probing. + * + */ +static void __devinit init_hwif_cs5535(ide_hwif_t *hwif) +{ + int i; + + hwif->autodma = 0; + + hwif->tuneproc = &cs5535_tuneproc; + hwif->speedproc = &cs5535_set_drive; + hwif->ide_dma_check = &cs5535_dma_check; + + hwif->atapi_dma = 1; + hwif->ultra_mask = 0x1F; + hwif->mwdma_mask = 0x07; + + + hwif->udma_four = cs5535_cable_detect(hwif->pci_dev); + + if (!noautodma) + hwif->autodma = 1; + + /* just setting autotune and not worrying about bios timings */ + for (i = 0; i < 2; i++) { + hwif->drives[i].autotune = 1; + hwif->drives[i].autodma = hwif->autodma; + } +} + +static ide_pci_device_t cs5535_chipset __devinitdata = { + .name = "CS5535", + .init_hwif = init_hwif_cs5535, + .channels = 1, + .autodma = AUTODMA, + .bootable = ON_BOARD, +}; + +static int __devinit cs5535_init_one(struct pci_dev *dev, + const struct pci_device_id *id) +{ + return ide_setup_pci_device(dev, &cs5535_chipset); +} + +static struct pci_device_id cs5535_pci_tbl[] = +{ + { PCI_VENDOR_ID_NS, PCI_DEVICE_ID_NS_CS5535_IDE, PCI_ANY_ID, + PCI_ANY_ID, 0, 0, 0}, + { 0, }, +}; + +MODULE_DEVICE_TABLE(pci, cs5535_pci_tbl); + +static struct pci_driver driver = { + .name = "CS5535_IDE", + .id_table = cs5535_pci_tbl, + .probe = cs5535_init_one, +}; + +static int __init cs5535_ide_init(void) +{ + return ide_pci_register_driver(&driver); +} + +module_init(cs5535_ide_init); + +MODULE_AUTHOR("AMD"); +MODULE_DESCRIPTION("PCI driver module for AMD/NS CS5535 IDE"); +MODULE_LICENSE("GPL"); diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 63dd184ec808..4e06eb0f4451 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -387,6 +387,7 @@ #define PCI_DEVICE_ID_NS_SC1100_SMI 0x0511 #define PCI_DEVICE_ID_NS_SC1100_XBUS 0x0515 #define PCI_DEVICE_ID_NS_87410 0xd001 +#define PCI_DEVICE_ID_NS_CS5535_IDE 0x002d #define PCI_VENDOR_ID_TSENG 0x100c #define PCI_DEVICE_ID_TSENG_W32P_2 0x3202 From 25000c2c8a6cbf9bba2de6560370ee222b4c613d Mon Sep 17 00:00:00 2001 From: Alan Cox Date: Thu, 10 Nov 2005 00:10:37 +0100 Subject: [PATCH 026/121] [PATCH] ide: explain the PCI bus test we do in Matthew Wilcox asked that this got a comment explaining why it is done so here it is. Signed-off-by: Alan Cox Signed-off-by: Bartlomiej Zolnierkiewicz --- include/asm-i386/ide.h | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/include/asm-i386/ide.h b/include/asm-i386/ide.h index 79dfab87135d..454440193eac 100644 --- a/include/asm-i386/ide.h +++ b/include/asm-i386/ide.h @@ -41,6 +41,12 @@ static __inline__ int ide_default_irq(unsigned long base) static __inline__ unsigned long ide_default_io_base(int index) { + /* + * If PCI is present then it is not safe to poke around + * the other legacy IDE ports. Only 0x1f0 and 0x170 are + * defined compatibility mode ports for PCI. A user can + * override this using ide= but we must default safe. + */ if (pci_find_device(PCI_ANY_ID, PCI_ANY_ID, NULL) == NULL) { switch(index) { case 2: return 0x1e8; From d868dd19ad0828b1b3b56f3b06bd4a39971847be Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Thu, 10 Nov 2005 00:19:14 +0100 Subject: [PATCH 027/121] [PATCH] siimage: enable interrupts on Adaptec SA-1210 card The siimage driver proports to support the Adaptec SA-1210 SATA controller. However, at least some of those cards boot-up with their interrupts disabled internally. The siimage driver currently ignores that fact, so that driver does not actually work with those cards. This patch enables those interrupts on cards that need it. [ This is implemented based on similar code in the libata-based sata_sil driver. ] Signed-off-by: John W. Linville Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/ide/pci/siimage.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/drivers/ide/pci/siimage.c b/drivers/ide/pci/siimage.c index 2b9961b88135..022d244f2eb0 100644 --- a/drivers/ide/pci/siimage.c +++ b/drivers/ide/pci/siimage.c @@ -701,6 +701,7 @@ static unsigned int setup_mmio_siimage (struct pci_dev *dev, const char *name) unsigned long barsize = pci_resource_len(dev, 5); u8 tmpbyte = 0; void __iomem *ioaddr; + u32 tmp, irq_mask; /* * Drop back to PIO if we can't map the mmio. Some @@ -726,6 +727,14 @@ static unsigned int setup_mmio_siimage (struct pci_dev *dev, const char *name) pci_set_drvdata(dev, (void *) ioaddr); if (pdev_is_sata(dev)) { + /* make sure IDE0/1 interrupts are not masked */ + irq_mask = (1 << 22) | (1 << 23); + tmp = readl(ioaddr + 0x48); + if (tmp & irq_mask) { + tmp &= ~irq_mask; + writel(tmp, ioaddr + 0x48); + readl(ioaddr + 0x48); /* flush */ + } writel(0, ioaddr + 0x148); writel(0, ioaddr + 0x1C8); } From 07203f6471a11f4eda48acbe528c0ce4704ad42c Mon Sep 17 00:00:00 2001 From: Ondrej Zary Date: Thu, 10 Nov 2005 00:25:15 +0100 Subject: [PATCH 028/121] [PATCH] ide-floppy: software eject not working with LS-120 drive The problem (eject not working on ATAPI LS-120 drive) is caused by idefloppy_ioctl() function which *first* tries generic_ide_ioctl() and *only* if it fails with -EINVAL, proceeds with the specific ioctls. The generic eject command fails with something other than -EINVAL and the specific one is never executed. This patch fixes it by first going through the internal ioctls and only trying generic_ide_ioctl() if none of them matches. Signed-off-by: Ondrej Zary Signed-off-by: Bartlomiej Zolnierkiewicz --- drivers/ide/ide-floppy.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/drivers/ide/ide-floppy.c b/drivers/ide/ide-floppy.c index e83f54d37f96..f615ab759962 100644 --- a/drivers/ide/ide-floppy.c +++ b/drivers/ide/ide-floppy.c @@ -2038,11 +2038,9 @@ static int idefloppy_ioctl(struct inode *inode, struct file *file, struct ide_floppy_obj *floppy = ide_floppy_g(bdev->bd_disk); ide_drive_t *drive = floppy->drive; void __user *argp = (void __user *)arg; - int err = generic_ide_ioctl(drive, file, bdev, cmd, arg); + int err; int prevent = (arg) ? 1 : 0; idefloppy_pc_t pc; - if (err != -EINVAL) - return err; switch (cmd) { case CDROMEJECT: @@ -2094,7 +2092,7 @@ static int idefloppy_ioctl(struct inode *inode, struct file *file, case IDEFLOPPY_IOCTL_FORMAT_GET_PROGRESS: return idefloppy_get_format_progress(drive, argp); } - return -EINVAL; + return generic_ide_ioctl(drive, file, bdev, cmd, arg); } static int idefloppy_media_changed(struct gendisk *disk) From a47ab9371e664952b1104a70ec8e9b74db3f7a5f Mon Sep 17 00:00:00 2001 From: "Chen, Kenneth W" Date: Wed, 9 Nov 2005 15:45:29 -0800 Subject: [PATCH 029/121] [PATCH] optimize activate_task() recalc_task_prio() is called from activate_task() to calculate dynamic priority and interactive credit for the activating task. For real-time scheduling process, all that dynamic calculation is thrown away at the end because rt priority is fixed. Patch to optimize recalc_task_prio() away for rt processes. Signed-off-by: Ken Chen Acked-by: Ingo Molnar Cc: Nick Piggin Cc: Con Kolivas Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- kernel/sched.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/kernel/sched.c b/kernel/sched.c index ac3f5cc3bb51..b6506671b2be 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -815,7 +815,8 @@ static void activate_task(task_t *p, runqueue_t *rq, int local) } #endif - p->prio = recalc_task_prio(p, now); + if (!rt_task(p)) + p->prio = recalc_task_prio(p, now); /* * This checks to make sure it's not an uninterruptible task From e4d76e1c0b15590f2ad9bba89426c2520cd22ca6 Mon Sep 17 00:00:00 2001 From: Andrew Morton Date: Wed, 9 Nov 2005 15:45:30 -0800 Subject: [PATCH 030/121] [PATCH] powerpc: sched fixups - Re-add a hunk lost during merge: ppc64 is missing the hunk that disables preempt on the secondary CPUs before they call cpu_idle(). - ppc's cpu_idle() had the need_resched() test wrong. Cc: Nick Piggin CC: Benjamin Herrenschmidt Cc: Paul Mackerras Cc: Anton Blanchard Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- arch/powerpc/kernel/smp.c | 1 + arch/ppc/kernel/idle.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c index 5c330c3366e4..36d67a8d7cbb 100644 --- a/arch/powerpc/kernel/smp.c +++ b/arch/powerpc/kernel/smp.c @@ -510,6 +510,7 @@ int __devinit start_secondary(void *unused) smp_store_cpu_info(cpu); set_dec(tb_ticks_per_jiffy); + preempt_disable(); cpu_callin_map[cpu] = 1; smp_ops->setup_cpu(cpu); diff --git a/arch/ppc/kernel/idle.c b/arch/ppc/kernel/idle.c index 3c4e4cb61074..821a75e45602 100644 --- a/arch/ppc/kernel/idle.c +++ b/arch/ppc/kernel/idle.c @@ -63,7 +63,7 @@ void cpu_idle(void) int cpu = smp_processor_id(); for (;;) { - while (need_resched()) { + while (!need_resched()) { if (ppc_md.idle != NULL) ppc_md.idle(); else From c5ff700116a56a870ef40cc4ac6f19bf2530b466 Mon Sep 17 00:00:00 2001 From: David Gibson Date: Wed, 9 Nov 2005 11:21:07 +1100 Subject: [PATCH 031/121] [PATCH] powerpc: Merge signal.h Having already merged the ppc and ppc64 versions of signal.c, this patch finishes the job by merging signal.h. The two versions were almost identical already. Notable changes: - We use BITS_PER_LONG to correctly size sigset_t - Remove some uneeded #includes and struct forward declarations. This does mean adding an include to signal_32.c which relied on the indirect inclusion of sigcontext.h - As the ppc64 version, the merged signal.h has prototypes for do_signal() and do_signal32(). Thus remove extra prototypes from ppc_ksyms.c which had them directly. Built and booted on POWER5 LPAR (ARCH=ppc64 and ARCH=powerpc). Built for 32-bit powermac (ARCH=ppc and ARCH=powerpc) and Walnut (ARCH=ppc). Signed-off-by: David Gibson Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/ppc_ksyms.c | 2 +- arch/powerpc/kernel/signal_32.c | 1 + arch/ppc/kernel/ppc_ksyms.c | 2 +- include/{asm-ppc => asm-powerpc}/signal.h | 41 ++++--- include/asm-ppc64/signal.h | 132 ---------------------- 5 files changed, 22 insertions(+), 156 deletions(-) rename include/{asm-ppc => asm-powerpc}/signal.h (82%) delete mode 100644 include/asm-ppc64/signal.h diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c index 47d6f7e2ea9f..5d9fd0369aad 100644 --- a/arch/powerpc/kernel/ppc_ksyms.c +++ b/arch/powerpc/kernel/ppc_ksyms.c @@ -44,6 +44,7 @@ #include #include #include +#include #ifdef CONFIG_8xx #include @@ -56,7 +57,6 @@ extern void machine_check_exception(struct pt_regs *regs); extern void alignment_exception(struct pt_regs *regs); extern void program_check_exception(struct pt_regs *regs); extern void single_step_exception(struct pt_regs *regs); -extern int do_signal(sigset_t *, struct pt_regs *); extern int pmac_newworld; extern int sys_sigreturn(struct pt_regs *regs); diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c index 081d931eae48..a7c4515f320f 100644 --- a/arch/powerpc/kernel/signal_32.c +++ b/arch/powerpc/kernel/signal_32.c @@ -42,6 +42,7 @@ #include #include +#include #ifdef CONFIG_PPC64 #include "ppc32.h" #include diff --git a/arch/ppc/kernel/ppc_ksyms.c b/arch/ppc/kernel/ppc_ksyms.c index e0ca61b37f4f..fc0d05c69895 100644 --- a/arch/ppc/kernel/ppc_ksyms.c +++ b/arch/ppc/kernel/ppc_ksyms.c @@ -46,6 +46,7 @@ #include #include #include +#include #ifdef CONFIG_8xx #include @@ -57,7 +58,6 @@ extern void machine_check_exception(struct pt_regs *regs); extern void alignment_exception(struct pt_regs *regs); extern void program_check_exception(struct pt_regs *regs); extern void single_step_exception(struct pt_regs *regs); -extern int do_signal(sigset_t *, struct pt_regs *); extern int pmac_newworld; extern int sys_sigreturn(struct pt_regs *regs); diff --git a/include/asm-ppc/signal.h b/include/asm-powerpc/signal.h similarity index 82% rename from include/asm-ppc/signal.h rename to include/asm-powerpc/signal.h index caf6ede3710f..694c8d2dab87 100644 --- a/include/asm-ppc/signal.h +++ b/include/asm-powerpc/signal.h @@ -1,18 +1,11 @@ -#ifndef _ASMPPC_SIGNAL_H -#define _ASMPPC_SIGNAL_H +#ifndef _ASM_POWERPC_SIGNAL_H +#define _ASM_POWERPC_SIGNAL_H -#ifdef __KERNEL__ #include -#endif /* __KERNEL__ */ - -/* Avoid too many header ordering problems. */ -struct siginfo; - -/* Most things should be clean enough to redefine this at will, if care - is taken to make libc match. */ +#include #define _NSIG 64 -#define _NSIG_BPW 32 +#define _NSIG_BPW BITS_PER_LONG #define _NSIG_WORDS (_NSIG / _NSIG_BPW) typedef unsigned long old_sigset_t; /* at least 32 bits */ @@ -77,19 +70,19 @@ typedef struct { * SA_ONESHOT and SA_NOMASK are the historical Linux names for the Single * Unix names RESETHAND and NODEFER respectively. */ -#define SA_NOCLDSTOP 0x00000001 -#define SA_NOCLDWAIT 0x00000002 -#define SA_SIGINFO 0x00000004 -#define SA_ONSTACK 0x08000000 -#define SA_RESTART 0x10000000 -#define SA_NODEFER 0x40000000 -#define SA_RESETHAND 0x80000000 +#define SA_NOCLDSTOP 0x00000001U +#define SA_NOCLDWAIT 0x00000002U +#define SA_SIGINFO 0x00000004U +#define SA_ONSTACK 0x08000000U +#define SA_RESTART 0x10000000U +#define SA_NODEFER 0x40000000U +#define SA_RESETHAND 0x80000000U #define SA_NOMASK SA_NODEFER #define SA_ONESHOT SA_RESETHAND -#define SA_INTERRUPT 0x20000000 /* dummy -- ignored */ +#define SA_INTERRUPT 0x20000000u /* dummy -- ignored */ -#define SA_RESTORER 0x04000000 +#define SA_RESTORER 0x04000000U /* * sigaltstack controls @@ -127,10 +120,13 @@ typedef struct sigaltstack { } stack_t; #ifdef __KERNEL__ -#include +struct pt_regs; +extern int do_signal(sigset_t *oldset, struct pt_regs *regs); +extern int do_signal32(sigset_t *oldset, struct pt_regs *regs); #define ptrace_signal_deliver(regs, cookie) do { } while (0) #endif /* __KERNEL__ */ +#ifndef __powerpc64__ /* * These are parameters to dbg_sigreturn syscall. They enable or * disable certain debugging things that can be done from signal @@ -149,5 +145,6 @@ struct sig_dbg_op { /* Enable or disable branch tracing. The value sets the state. */ #define SIG_DBG_BRANCH_TRACING 2 +#endif /* ! __powerpc64__ */ -#endif +#endif /* _ASM_POWERPC_SIGNAL_H */ diff --git a/include/asm-ppc64/signal.h b/include/asm-ppc64/signal.h deleted file mode 100644 index 432df7dd355d..000000000000 --- a/include/asm-ppc64/signal.h +++ /dev/null @@ -1,132 +0,0 @@ -#ifndef _ASMPPC64_SIGNAL_H -#define _ASMPPC64_SIGNAL_H - -#include -#include -#include - -/* Avoid too many header ordering problems. */ -struct siginfo; - -#define _NSIG 64 -#define _NSIG_BPW 64 -#define _NSIG_WORDS (_NSIG / _NSIG_BPW) - -typedef unsigned long old_sigset_t; /* at least 32 bits */ - -typedef struct { - unsigned long sig[_NSIG_WORDS]; -} sigset_t; - -#define SIGHUP 1 -#define SIGINT 2 -#define SIGQUIT 3 -#define SIGILL 4 -#define SIGTRAP 5 -#define SIGABRT 6 -#define SIGIOT 6 -#define SIGBUS 7 -#define SIGFPE 8 -#define SIGKILL 9 -#define SIGUSR1 10 -#define SIGSEGV 11 -#define SIGUSR2 12 -#define SIGPIPE 13 -#define SIGALRM 14 -#define SIGTERM 15 -#define SIGSTKFLT 16 -#define SIGCHLD 17 -#define SIGCONT 18 -#define SIGSTOP 19 -#define SIGTSTP 20 -#define SIGTTIN 21 -#define SIGTTOU 22 -#define SIGURG 23 -#define SIGXCPU 24 -#define SIGXFSZ 25 -#define SIGVTALRM 26 -#define SIGPROF 27 -#define SIGWINCH 28 -#define SIGIO 29 -#define SIGPOLL SIGIO -/* -#define SIGLOST 29 -*/ -#define SIGPWR 30 -#define SIGSYS 31 -#define SIGUNUSED 31 - -/* These should not be considered constants from userland. */ -#define SIGRTMIN 32 -#define SIGRTMAX _NSIG - -/* - * SA_FLAGS values: - * - * SA_ONSTACK is not currently supported, but will allow sigaltstack(2). - * SA_INTERRUPT is a no-op, but left due to historical reasons. Use the - * SA_RESTART flag to get restarting signals (which were the default long ago) - * SA_NOCLDSTOP flag to turn off SIGCHLD when children stop. - * SA_RESETHAND clears the handler when the signal is delivered. - * SA_NOCLDWAIT flag on SIGCHLD to inhibit zombies. - * SA_NODEFER prevents the current signal from being masked in the handler. - * - * SA_ONESHOT and SA_NOMASK are the historical Linux names for the Single - * Unix names RESETHAND and NODEFER respectively. - */ -#define SA_NOCLDSTOP 0x00000001u -#define SA_NOCLDWAIT 0x00000002u -#define SA_SIGINFO 0x00000004u -#define SA_ONSTACK 0x08000000u -#define SA_RESTART 0x10000000u -#define SA_NODEFER 0x40000000u -#define SA_RESETHAND 0x80000000u - -#define SA_NOMASK SA_NODEFER -#define SA_ONESHOT SA_RESETHAND -#define SA_INTERRUPT 0x20000000u /* dummy -- ignored */ - -#define SA_RESTORER 0x04000000u - -/* - * sigaltstack controls - */ -#define SS_ONSTACK 1 -#define SS_DISABLE 2 - -#define MINSIGSTKSZ 2048 -#define SIGSTKSZ 8192 - -#include - -struct old_sigaction { - __sighandler_t sa_handler; - old_sigset_t sa_mask; - unsigned long sa_flags; - __sigrestore_t sa_restorer; -}; - -struct sigaction { - __sighandler_t sa_handler; - unsigned long sa_flags; - __sigrestore_t sa_restorer; - sigset_t sa_mask; /* mask last for extensibility */ -}; - -struct k_sigaction { - struct sigaction sa; -}; - -typedef struct sigaltstack { - void __user *ss_sp; - int ss_flags; - size_t ss_size; -} stack_t; - -struct pt_regs; -struct timespec; -extern int do_signal(sigset_t *oldset, struct pt_regs *regs); -extern int do_signal32(sigset_t *oldset, struct pt_regs *regs); -#define ptrace_signal_deliver(regs, cookie) do { } while (0) - -#endif /* _ASMPPC64_SIGNAL_H */ From 584224e4095d8abcf2bef38efacc291be9a44c20 Mon Sep 17 00:00:00 2001 From: David Gibson Date: Wed, 9 Nov 2005 13:04:06 +1100 Subject: [PATCH 032/121] [PATCH] powerpc: Merge current.h This patch merges current.h. This is a one-big-ifdef merge, but both versions are so tiny, I think we can live with it. While we're at it, we get rid of the fairly pointless redirection through get_current() in the ppc64 version. Built and booted on POWER5 LPAR (ARCH=powerpc & ARCH=ppc64). Built for 32-bit pmac (ARCH=powerpc & ARCH=ppc). Signed-off-by: David Gibson Signed-off-by: Paul Mackerras --- include/asm-powerpc/current.h | 27 +++++++++++++++++++++++++++ include/asm-ppc/current.h | 11 ----------- include/asm-ppc64/current.h | 16 ---------------- 3 files changed, 27 insertions(+), 27 deletions(-) create mode 100644 include/asm-powerpc/current.h delete mode 100644 include/asm-ppc/current.h delete mode 100644 include/asm-ppc64/current.h diff --git a/include/asm-powerpc/current.h b/include/asm-powerpc/current.h new file mode 100644 index 000000000000..82cd4a9ca99a --- /dev/null +++ b/include/asm-powerpc/current.h @@ -0,0 +1,27 @@ +#ifndef _ASM_POWERPC_CURRENT_H +#define _ASM_POWERPC_CURRENT_H + +/* + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +struct task_struct; + +#ifdef __powerpc64__ +#include + +#define current (get_paca()->__current) + +#else + +/* + * We keep `current' in r2 for speed. + */ +register struct task_struct *current asm ("r2"); + +#endif + +#endif /* _ASM_POWERPC_CURRENT_H */ diff --git a/include/asm-ppc/current.h b/include/asm-ppc/current.h deleted file mode 100644 index 8d41501ba10d..000000000000 --- a/include/asm-ppc/current.h +++ /dev/null @@ -1,11 +0,0 @@ -#ifdef __KERNEL__ -#ifndef _PPC_CURRENT_H -#define _PPC_CURRENT_H - -/* - * We keep `current' in r2 for speed. - */ -register struct task_struct *current asm ("r2"); - -#endif /* !(_PPC_CURRENT_H) */ -#endif /* __KERNEL__ */ diff --git a/include/asm-ppc64/current.h b/include/asm-ppc64/current.h deleted file mode 100644 index 52ddc60c8b65..000000000000 --- a/include/asm-ppc64/current.h +++ /dev/null @@ -1,16 +0,0 @@ -#ifndef _PPC64_CURRENT_H -#define _PPC64_CURRENT_H - -#include - -/* - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - */ - -#define get_current() (get_paca()->__current) -#define current get_current() - -#endif /* !(_PPC64_CURRENT_H) */ From 8882a4da1c932c9f311c9f739e6719adea3e25d9 Mon Sep 17 00:00:00 2001 From: David Gibson Date: Wed, 9 Nov 2005 13:38:01 +1100 Subject: [PATCH 033/121] [PATCH] powerpc: Move various ppc64 files with no ppc32 equivalent to powerpc This patch moves a bunch of files from arch/ppc64 and include/asm-ppc64 which have no equivalents in ppc32 code into arch/powerpc and include/asm-powerpc. The file affected are: abs_addr.h compat.h lppaca.h paca.h tce.h cpu_setup_power4.S ioctl32.c firmware.c pacaData.c The only changes apart from the move and corresponding Makefile changes are: - #ifndef/#define in includes updated to _ASM_POWERPC_ form - trailing whitespace removed - comments giving full paths removed - pacaData.c renamed paca.c to remove studlyCaps - Misplaced { moved in lppaca.h Built and booted on POWER5 LPAR (ARCH=powerpc and ARCH=ppc64), built for 32-bit powermac (ARCH=powerpc). Signed-off-by: David Gibson Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/Makefile | 5 ++++- arch/{ppc64 => powerpc}/kernel/cpu_setup_power4.S | 8 ++++---- arch/{ppc64 => powerpc}/kernel/firmware.c | 2 -- arch/{ppc64 => powerpc}/kernel/ioctl32.c | 4 ++-- .../kernel/pacaData.c => powerpc/kernel/paca.c} | 2 +- arch/ppc64/kernel/Makefile | 9 +++------ include/{asm-ppc64 => asm-powerpc}/abs_addr.h | 6 +++--- include/{asm-ppc64 => asm-powerpc}/compat.h | 8 ++++---- include/{asm-ppc64 => asm-powerpc}/lppaca.h | 9 ++++----- include/{asm-ppc64 => asm-powerpc}/paca.h | 15 +++++++-------- include/{asm-ppc64 => asm-powerpc}/tce.h | 6 +++--- 11 files changed, 35 insertions(+), 39 deletions(-) rename arch/{ppc64 => powerpc}/kernel/cpu_setup_power4.S (99%) rename arch/{ppc64 => powerpc}/kernel/firmware.c (97%) rename arch/{ppc64 => powerpc}/kernel/ioctl32.c (99%) rename arch/{ppc64/kernel/pacaData.c => powerpc/kernel/paca.c} (99%) rename include/{asm-ppc64 => asm-powerpc}/abs_addr.h (94%) rename include/{asm-ppc64 => asm-powerpc}/compat.h (97%) rename include/{asm-ppc64 => asm-powerpc}/lppaca.h (98%) rename include/{asm-ppc64 => asm-powerpc}/paca.h (95%) rename include/{asm-ppc64 => asm-powerpc}/tce.h (96%) diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile index b3ae2993efb8..103cb8128ef4 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -4,6 +4,7 @@ ifeq ($(CONFIG_PPC64),y) EXTRA_CFLAGS += -mno-minimal-toc +CFLAGS_ioctl32.o += -Ifs/ endif ifeq ($(CONFIG_PPC32),y) CFLAGS_prom_init.o += -fPIC @@ -13,7 +14,9 @@ endif obj-y := semaphore.o cputable.o ptrace.o syscalls.o \ signal_32.o pmc.o obj-$(CONFIG_PPC64) += setup_64.o binfmt_elf32.o sys_ppc32.o \ - signal_64.o ptrace32.o systbl.o + signal_64.o ptrace32.o systbl.o \ + paca.o ioctl32.o cpu_setup_power4.o \ + firmware.o obj-$(CONFIG_ALTIVEC) += vecemu.o vector.o obj-$(CONFIG_POWER4) += idle_power4.o obj-$(CONFIG_PPC_OF) += of_device.o diff --git a/arch/ppc64/kernel/cpu_setup_power4.S b/arch/powerpc/kernel/cpu_setup_power4.S similarity index 99% rename from arch/ppc64/kernel/cpu_setup_power4.S rename to arch/powerpc/kernel/cpu_setup_power4.S index 1fb673c511ff..cca942fe6115 100644 --- a/arch/ppc64/kernel/cpu_setup_power4.S +++ b/arch/powerpc/kernel/cpu_setup_power4.S @@ -114,11 +114,11 @@ _GLOBAL(__setup_cpu_ppc970) .data .balign L1_CACHE_BYTES,0 -cpu_state_storage: +cpu_state_storage: .space CS_SIZE .balign L1_CACHE_BYTES,0 .text - + /* Called in normal context to backup CPU 0 state. This * does not include cache settings. This function is also * called for machine sleep. This does not include the MMU @@ -151,7 +151,7 @@ _GLOBAL(__save_cpu_setup) std r3,CS_HID4(r5) mfspr r3,SPRN_HID5 std r3,CS_HID5(r5) - + 2: mtcr r7 blr @@ -213,7 +213,7 @@ _GLOBAL(__restore_cpu_setup) mtspr SPRN_HID1,r3 sync isync - + /* Restore HID4 */ ld r3,CS_HID4(r5) sync diff --git a/arch/ppc64/kernel/firmware.c b/arch/powerpc/kernel/firmware.c similarity index 97% rename from arch/ppc64/kernel/firmware.c rename to arch/powerpc/kernel/firmware.c index d8432c0fb27d..65eae752a527 100644 --- a/arch/ppc64/kernel/firmware.c +++ b/arch/powerpc/kernel/firmware.c @@ -1,6 +1,4 @@ /* - * arch/ppc64/kernel/firmware.c - * * Extracted from cputable.c * * Copyright (C) 2001 Ben. Herrenschmidt (benh@kernel.crashing.org) diff --git a/arch/ppc64/kernel/ioctl32.c b/arch/powerpc/kernel/ioctl32.c similarity index 99% rename from arch/ppc64/kernel/ioctl32.c rename to arch/powerpc/kernel/ioctl32.c index ba4a899045c2..3fa6a93adbd0 100644 --- a/arch/ppc64/kernel/ioctl32.c +++ b/arch/powerpc/kernel/ioctl32.c @@ -1,6 +1,6 @@ -/* +/* * ioctl32.c: Conversion between 32bit and 64bit native ioctls. - * + * * Based on sparc64 ioctl32.c by: * * Copyright (C) 1997-2000 Jakub Jelinek (jakub@redhat.com) diff --git a/arch/ppc64/kernel/pacaData.c b/arch/powerpc/kernel/paca.c similarity index 99% rename from arch/ppc64/kernel/pacaData.c rename to arch/powerpc/kernel/paca.c index 3133c72b28ec..179948eb0580 100644 --- a/arch/ppc64/kernel/pacaData.c +++ b/arch/powerpc/kernel/paca.c @@ -32,7 +32,7 @@ EXPORT_SYMBOL(systemcfg); * field correctly */ extern unsigned long __toc_start; -/* The Paca is an array with one entry per processor. Each contains an +/* The Paca is an array with one entry per processor. Each contains an * lppaca, which contains the information shared between the * hypervisor and Linux. Each also contains an ItLpRegSave area which * is used by the hypervisor to save registers. diff --git a/arch/ppc64/kernel/Makefile b/arch/ppc64/kernel/Makefile index c441aebe7648..eb3187f18fb0 100644 --- a/arch/ppc64/kernel/Makefile +++ b/arch/ppc64/kernel/Makefile @@ -12,11 +12,10 @@ obj-y := misc.o prom.o endif obj-y += irq.o idle.o dma.o \ - align.o pacaData.o \ - udbg.o ioctl32.o \ + align.o \ + udbg.o \ rtc.o \ - cpu_setup_power4.o \ - iommu.o sysfs.o vdso.o firmware.o + iommu.o sysfs.o vdso.o obj-y += vdso32/ vdso64/ pci-obj-$(CONFIG_PPC_MULTIPLATFORM) += pci_dn.o pci_direct_iommu.o @@ -52,8 +51,6 @@ obj-$(CONFIG_PPC_MAPLE) += udbg_16550.o obj-$(CONFIG_KPROBES) += kprobes.o -CFLAGS_ioctl32.o += -Ifs/ - ifneq ($(CONFIG_PPC_MERGE),y) ifeq ($(CONFIG_PPC_ISERIES),y) arch/ppc64/kernel/head.o: arch/powerpc/kernel/lparmap.s diff --git a/include/asm-ppc64/abs_addr.h b/include/asm-powerpc/abs_addr.h similarity index 94% rename from include/asm-ppc64/abs_addr.h rename to include/asm-powerpc/abs_addr.h index dc3fc3fefef2..18415108fc56 100644 --- a/include/asm-ppc64/abs_addr.h +++ b/include/asm-powerpc/abs_addr.h @@ -1,5 +1,5 @@ -#ifndef _ABS_ADDR_H -#define _ABS_ADDR_H +#ifndef _ASM_POWERPC_ABS_ADDR_H +#define _ASM_POWERPC_ABS_ADDR_H #include @@ -70,4 +70,4 @@ static inline unsigned long phys_to_abs(unsigned long pa) #define iseries_hv_addr(virtaddr) \ (0x8000000000000000 | virt_to_abs(virtaddr)) -#endif /* _ABS_ADDR_H */ +#endif /* _ASM_POWERPC_ABS_ADDR_H */ diff --git a/include/asm-ppc64/compat.h b/include/asm-powerpc/compat.h similarity index 97% rename from include/asm-ppc64/compat.h rename to include/asm-powerpc/compat.h index 6ec62cd2d1d1..4db4360c4d4a 100644 --- a/include/asm-ppc64/compat.h +++ b/include/asm-powerpc/compat.h @@ -1,5 +1,5 @@ -#ifndef _ASM_PPC64_COMPAT_H -#define _ASM_PPC64_COMPAT_H +#ifndef _ASM_POWERPC_COMPAT_H +#define _ASM_POWERPC_COMPAT_H /* * Architecture specific compatibility types */ @@ -49,7 +49,7 @@ struct compat_stat { compat_dev_t st_dev; compat_ino_t st_ino; compat_mode_t st_mode; - compat_nlink_t st_nlink; + compat_nlink_t st_nlink; __compat_uid32_t st_uid; __compat_gid32_t st_gid; compat_dev_t st_rdev; @@ -202,4 +202,4 @@ struct compat_shmid64_ds { compat_ulong_t __unused6; }; -#endif /* _ASM_PPC64_COMPAT_H */ +#endif /* _ASM_POWERPC_COMPAT_H */ diff --git a/include/asm-ppc64/lppaca.h b/include/asm-powerpc/lppaca.h similarity index 98% rename from include/asm-ppc64/lppaca.h rename to include/asm-powerpc/lppaca.h index 9e2a6c0649a0..c1bedab1515b 100644 --- a/include/asm-ppc64/lppaca.h +++ b/include/asm-powerpc/lppaca.h @@ -16,8 +16,8 @@ * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#ifndef _ASM_LPPACA_H -#define _ASM_LPPACA_H +#ifndef _ASM_POWERPC_LPPACA_H +#define _ASM_POWERPC_LPPACA_H //============================================================================= // @@ -28,8 +28,7 @@ //---------------------------------------------------------------------------- #include -struct lppaca -{ +struct lppaca { //============================================================================= // CACHE_LINE_1 0x0000 - 0x007F Contains read-only data // NOTE: The xDynXyz fields are fields that will be dynamically changed by @@ -129,4 +128,4 @@ struct lppaca u8 pmc_save_area[256]; // PMC interrupt Area x00-xFF }; -#endif /* _ASM_LPPACA_H */ +#endif /* _ASM_POWERPC_LPPACA_H */ diff --git a/include/asm-ppc64/paca.h b/include/asm-powerpc/paca.h similarity index 95% rename from include/asm-ppc64/paca.h rename to include/asm-powerpc/paca.h index bccacd6aa93a..92c765c35bd0 100644 --- a/include/asm-ppc64/paca.h +++ b/include/asm-powerpc/paca.h @@ -1,11 +1,8 @@ -#ifndef _PPC64_PACA_H -#define _PPC64_PACA_H - /* - * include/asm-ppc64/paca.h + * include/asm-powerpc/paca.h * - * This control block defines the PACA which defines the processor - * specific data for each logical processor on the system. + * This control block defines the PACA which defines the processor + * specific data for each logical processor on the system. * There are some pointers defined that are utilized by PLIC. * * C 2001 PPC 64 Team, IBM Corp @@ -14,7 +11,9 @@ * modify it under the terms of the GNU General Public License * as published by the Free Software Foundation; either version * 2 of the License, or (at your option) any later version. - */ + */ +#ifndef _ASM_POWERPC_PACA_H +#define _ASM_POWERPC_PACA_H #include #include @@ -118,4 +117,4 @@ struct paca_struct { extern struct paca_struct paca[]; -#endif /* _PPC64_PACA_H */ +#endif /* _ASM_POWERPC_PACA_H */ diff --git a/include/asm-ppc64/tce.h b/include/asm-powerpc/tce.h similarity index 96% rename from include/asm-ppc64/tce.h rename to include/asm-powerpc/tce.h index d40b6b42ab35..d099d5200f9b 100644 --- a/include/asm-ppc64/tce.h +++ b/include/asm-powerpc/tce.h @@ -18,8 +18,8 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#ifndef _ASM_TCE_H -#define _ASM_TCE_H +#ifndef _ASM_POWERPC_TCE_H +#define _ASM_POWERPC_TCE_H /* * Tces come in two formats, one for the virtual bus and a different @@ -61,4 +61,4 @@ union tce_entry { }; -#endif +#endif /* _ASM_POWERPC_TCE_H */ From 43cefe29d497c2a9d3c7eae89964260e7b35149e Mon Sep 17 00:00:00 2001 From: Matt Porter Date: Wed, 9 Nov 2005 06:42:05 -0700 Subject: [PATCH 034/121] [PATCH] ppc32: fix ppc44x fpu build Fixes ppc44x fpu support that broke from a bad arch/powerpc merge. Instead of adding KernelFP back in (which duplicates code) we use the same kernel fpu unavailable handler as classic PPC processors. Signed-off-by: Matt Porter Signed-off-by: Paul Mackerras --- arch/ppc/kernel/head_booke.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/ppc/kernel/head_booke.h b/arch/ppc/kernel/head_booke.h index aeb349b47af3..f3d274c6b231 100644 --- a/arch/ppc/kernel/head_booke.h +++ b/arch/ppc/kernel/head_booke.h @@ -358,6 +358,6 @@ label: NORMAL_EXCEPTION_PROLOG; \ bne load_up_fpu; /* if from user, just load it up */ \ addi r3,r1,STACK_FRAME_OVERHEAD; \ - EXC_XFER_EE_LITE(0x800, KernelFP) + EXC_XFER_EE_LITE(0x800, kernel_fp_unavailable_exception) #endif /* __HEAD_BOOKE_H__ */ From f7b9996990bccaa9f53cbea7aea8ab5355e7f10c Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Wed, 9 Nov 2005 11:00:16 -0200 Subject: [PATCH 035/121] [PATCH] fs_enet build fix Due to the recent update of the platform code, some platform device drivers fail to compile. This fix is for fs_enet, adding #include of a new header, to which a number of platform stuff has been relocated. Signed-off-by: Vitaly Bordug Signed-off-by: Paul Mackerras --- drivers/net/fs_enet/fs_enet-main.c | 1 + drivers/net/fs_enet/mac-fcc.c | 1 + drivers/net/fs_enet/mac-fec.c | 1 + drivers/net/fs_enet/mac-scc.c | 1 + 4 files changed, 4 insertions(+) diff --git a/drivers/net/fs_enet/fs_enet-main.c b/drivers/net/fs_enet/fs_enet-main.c index 9342d5bc7bb4..f5d49a110654 100644 --- a/drivers/net/fs_enet/fs_enet-main.c +++ b/drivers/net/fs_enet/fs_enet-main.c @@ -37,6 +37,7 @@ #include #include #include +#include #include #include diff --git a/drivers/net/fs_enet/mac-fcc.c b/drivers/net/fs_enet/mac-fcc.c index a940b96433c7..e67b1d06611c 100644 --- a/drivers/net/fs_enet/mac-fcc.c +++ b/drivers/net/fs_enet/mac-fcc.c @@ -34,6 +34,7 @@ #include #include #include +#include #include #include diff --git a/drivers/net/fs_enet/mac-fec.c b/drivers/net/fs_enet/mac-fec.c index 5ef4e845a387..2e8f44469699 100644 --- a/drivers/net/fs_enet/mac-fec.c +++ b/drivers/net/fs_enet/mac-fec.c @@ -34,6 +34,7 @@ #include #include #include +#include #include #include diff --git a/drivers/net/fs_enet/mac-scc.c b/drivers/net/fs_enet/mac-scc.c index d8c6e9cadcf5..a3897fda71fa 100644 --- a/drivers/net/fs_enet/mac-scc.c +++ b/drivers/net/fs_enet/mac-scc.c @@ -34,6 +34,7 @@ #include #include #include +#include #include #include From eba2fb2dac93ed01a9f85b2f623b49a93bef3ed7 Mon Sep 17 00:00:00 2001 From: Olaf Hering Date: Wed, 9 Nov 2005 20:51:03 +0100 Subject: [PATCH 036/121] [PATCH] ppc64 boot: remove local initializers Remove initialization of local variables. They get all values assigned before use. Signed-off-by: Olaf Hering Signed-off-by: Paul Mackerras --- arch/ppc64/boot/addRamDisk.c | 54 ++++++++++++++++++------------------ 1 file changed, 27 insertions(+), 27 deletions(-) diff --git a/arch/ppc64/boot/addRamDisk.c b/arch/ppc64/boot/addRamDisk.c index 7f2c09473394..c4e5de96142b 100644 --- a/arch/ppc64/boot/addRamDisk.c +++ b/arch/ppc64/boot/addRamDisk.c @@ -34,35 +34,35 @@ void death(const char *msg, FILE *fdesc, const char *fname) int main(int argc, char **argv) { char inbuf[4096]; - FILE *ramDisk = NULL; - FILE *sysmap = NULL; - FILE *inputVmlinux = NULL; - FILE *outputVmlinux = NULL; - - unsigned i = 0; - unsigned long ramFileLen = 0; - unsigned long ramLen = 0; - unsigned long roundR = 0; - - unsigned long sysmapFileLen = 0; - unsigned long sysmapLen = 0; - unsigned long sysmapPages = 0; - char* ptr_end = NULL; - unsigned long offset_end = 0; + FILE *ramDisk; + FILE *sysmap; + FILE *inputVmlinux; + FILE *outputVmlinux; - unsigned long kernelLen = 0; - unsigned long actualKernelLen = 0; - unsigned long round = 0; - unsigned long roundedKernelLen = 0; - unsigned long ramStartOffs = 0; - unsigned long ramPages = 0; - unsigned long roundedKernelPages = 0; - unsigned long hvReleaseData = 0; + unsigned i; + unsigned long ramFileLen; + unsigned long ramLen; + unsigned long roundR; + + unsigned long sysmapFileLen; + unsigned long sysmapLen; + unsigned long sysmapPages; + char *ptr_end; + unsigned long offset_end; + + unsigned long kernelLen; + unsigned long actualKernelLen; + unsigned long round; + unsigned long roundedKernelLen; + unsigned long ramStartOffs; + unsigned long ramPages; + unsigned long roundedKernelPages; + unsigned long hvReleaseData; u_int32_t eyeCatcher = 0xc8a5d9c4; - unsigned long naca = 0; - unsigned long xRamDisk = 0; - unsigned long xRamDiskSize = 0; - long padPages = 0; + unsigned long naca; + unsigned long xRamDisk; + unsigned long xRamDiskSize; + long padPages; if (argc < 2) { From 58638bffe3d594ed3a902824ea40a4196c745bbd Mon Sep 17 00:00:00 2001 From: Olaf Hering Date: Wed, 9 Nov 2005 20:52:20 +0100 Subject: [PATCH 037/121] [PATCH] ppc64 boot: remove argv usage Use a local variable for the input filenames. Signed-off-by: Olaf Hering Signed-off-by: Paul Mackerras --- arch/ppc64/boot/addRamDisk.c | 32 ++++++++++++++++++-------------- 1 file changed, 18 insertions(+), 14 deletions(-) diff --git a/arch/ppc64/boot/addRamDisk.c b/arch/ppc64/boot/addRamDisk.c index c4e5de96142b..c10746a187b6 100644 --- a/arch/ppc64/boot/addRamDisk.c +++ b/arch/ppc64/boot/addRamDisk.c @@ -39,6 +39,7 @@ int main(int argc, char **argv) FILE *inputVmlinux; FILE *outputVmlinux; + char *rd_name, *lx_name, *out_name; unsigned i; unsigned long ramFileLen; unsigned long ramLen; @@ -69,6 +70,7 @@ int main(int argc, char **argv) fprintf(stderr, "Name of RAM disk file missing.\n"); exit(1); } + rd_name = argv[1] if (argc < 3) { fprintf(stderr, "Name of System Map input file is missing.\n"); @@ -79,16 +81,18 @@ int main(int argc, char **argv) fprintf(stderr, "Name of vmlinux file missing.\n"); exit(1); } + lx_name = argv[3]; if (argc < 5) { fprintf(stderr, "Name of vmlinux output file missing.\n"); exit(1); } + out_name = argv[4]; - ramDisk = fopen(argv[1], "r"); + ramDisk = fopen(rd_name, "r"); if ( ! ramDisk ) { - fprintf(stderr, "RAM disk file \"%s\" failed to open.\n", argv[1]); + fprintf(stderr, "RAM disk file \"%s\" failed to open.\n", rd_name); exit(1); } @@ -98,15 +102,15 @@ int main(int argc, char **argv) exit(1); } - inputVmlinux = fopen(argv[3], "r"); + inputVmlinux = fopen(lx_name, "r"); if ( ! inputVmlinux ) { - fprintf(stderr, "vmlinux file \"%s\" failed to open.\n", argv[3]); + fprintf(stderr, "vmlinux file \"%s\" failed to open.\n", lx_name); exit(1); } - outputVmlinux = fopen(argv[4], "w+"); + outputVmlinux = fopen(out_name, "w+"); if ( ! outputVmlinux ) { - fprintf(stderr, "output vmlinux file \"%s\" failed to open.\n", argv[4]); + fprintf(stderr, "output vmlinux file \"%s\" failed to open.\n", out_name); exit(1); } @@ -194,7 +198,7 @@ int main(int argc, char **argv) fseek(ramDisk, 0, SEEK_END); ramFileLen = ftell(ramDisk); fseek(ramDisk, 0, SEEK_SET); - printf("%s file size = %ld/0x%lx \n", argv[1], ramFileLen, ramFileLen); + printf("%s file size = %ld/0x%lx \n", rd_name, ramFileLen, ramFileLen); ramLen = ramFileLen; @@ -248,7 +252,7 @@ int main(int argc, char **argv) /* fseek to the hvReleaseData pointer */ fseek(outputVmlinux, ElfHeaderSize + 0x24, SEEK_SET); if (fread(&hvReleaseData, 4, 1, outputVmlinux) != 1) { - death("Could not read hvReleaseData pointer\n", outputVmlinux, argv[4]); + death("Could not read hvReleaseData pointer\n", outputVmlinux, out_name); } hvReleaseData = ntohl(hvReleaseData); /* Convert to native int */ printf("hvReleaseData is at %08x\n", hvReleaseData); @@ -256,11 +260,11 @@ int main(int argc, char **argv) /* fseek to the hvReleaseData */ fseek(outputVmlinux, ElfHeaderSize + hvReleaseData, SEEK_SET); if (fread(inbuf, 0x40, 1, outputVmlinux) != 1) { - death("Could not read hvReleaseData\n", outputVmlinux, argv[4]); + death("Could not read hvReleaseData\n", outputVmlinux, out_name); } /* Check hvReleaseData sanity */ if (memcmp(inbuf, &eyeCatcher, 4) != 0) { - death("hvReleaseData is invalid\n", outputVmlinux, argv[4]); + death("hvReleaseData is invalid\n", outputVmlinux, out_name); } /* Get the naca pointer */ naca = ntohl(*((u_int32_t*) &inbuf[0x0C])) - KERNELBASE; @@ -269,13 +273,13 @@ int main(int argc, char **argv) /* fseek to the naca */ fseek(outputVmlinux, ElfHeaderSize + naca, SEEK_SET); if (fread(inbuf, 0x18, 1, outputVmlinux) != 1) { - death("Could not read naca\n", outputVmlinux, argv[4]); + death("Could not read naca\n", outputVmlinux, out_name); } xRamDisk = ntohl(*((u_int32_t *) &inbuf[0x0c])); xRamDiskSize = ntohl(*((u_int32_t *) &inbuf[0x14])); /* Make sure a RAM disk isn't already present */ if ((xRamDisk != 0) || (xRamDiskSize != 0)) { - death("RAM disk is already attached to this kernel\n", outputVmlinux, argv[4]); + death("RAM disk is already attached to this kernel\n", outputVmlinux, out_name); } /* Fill in the values */ *((u_int32_t *) &inbuf[0x0c]) = htonl(ramStartOffs); @@ -285,7 +289,7 @@ int main(int argc, char **argv) fflush(outputVmlinux); fseek(outputVmlinux, ElfHeaderSize + naca, SEEK_SET); if (fwrite(inbuf, 0x18, 1, outputVmlinux) != 1) { - death("Could not write naca\n", outputVmlinux, argv[4]); + death("Could not write naca\n", outputVmlinux, out_name); } printf("Ram Disk of 0x%lx pages is attached to the kernel at offset 0x%08x\n", ramPages, ramStartOffs); @@ -293,7 +297,7 @@ int main(int argc, char **argv) /* Done */ fclose(outputVmlinux); /* Set permission to executable */ - chmod(argv[4], S_IRUSR|S_IWUSR|S_IXUSR|S_IRGRP|S_IXGRP|S_IROTH|S_IXOTH); + chmod(out_name, S_IRUSR|S_IWUSR|S_IXUSR|S_IRGRP|S_IXGRP|S_IROTH|S_IXOTH); return 0; } From 681c774d0fb59c9b9525c2879d1cfae29c2f7866 Mon Sep 17 00:00:00 2001 From: Olaf Hering Date: Wed, 9 Nov 2005 20:53:43 +0100 Subject: [PATCH 038/121] [PATCH] ppc64 boot: remove sysmap from required filenames A stripped vmlinux does not contain enough symbols to recreate the System.map. The System.map file is only used to determine the end of the runtime memory size. This is the same value (rounded up to PAGE_SIZE) as ->memsiz in the ELF program header. Also, the target vmlinux.initrd doesnt work in 2.6.14: arch/ppc64/boot/addRamDisk arch/ppc64/boot/ramdisk.image.gz vmlinux.strip arch/ppc64/boot/vmlinux.initrd Name of vmlinux output file missing. Signed-off-by: Olaf Hering Signed-off-by: Paul Mackerras --- arch/ppc64/boot/addRamDisk.c | 131 ++++++++++++++++++----------------- 1 file changed, 69 insertions(+), 62 deletions(-) diff --git a/arch/ppc64/boot/addRamDisk.c b/arch/ppc64/boot/addRamDisk.c index c10746a187b6..d99c04a38e46 100644 --- a/arch/ppc64/boot/addRamDisk.c +++ b/arch/ppc64/boot/addRamDisk.c @@ -5,11 +5,59 @@ #include #include #include +#include #define ElfHeaderSize (64 * 1024) #define ElfPages (ElfHeaderSize / 4096) #define KERNELBASE (0xc000000000000000) +#define _ALIGN_UP(addr,size) (((addr)+((size)-1))&(~((size)-1))) +struct addr_range { + unsigned long long addr; + unsigned long memsize; + unsigned long offset; +}; + +static int check_elf64(void *p, int size, struct addr_range *r) +{ + Elf64_Ehdr *elf64 = p; + Elf64_Phdr *elf64ph; + + if (elf64->e_ident[EI_MAG0] != ELFMAG0 || + elf64->e_ident[EI_MAG1] != ELFMAG1 || + elf64->e_ident[EI_MAG2] != ELFMAG2 || + elf64->e_ident[EI_MAG3] != ELFMAG3 || + elf64->e_ident[EI_CLASS] != ELFCLASS64 || + elf64->e_ident[EI_DATA] != ELFDATA2MSB || + elf64->e_type != ET_EXEC || elf64->e_machine != EM_PPC64) + return 0; + + if ((elf64->e_phoff + sizeof(Elf64_Phdr)) > size) + return 0; + + elf64ph = (Elf64_Phdr *) ((unsigned long)elf64 + + (unsigned long)elf64->e_phoff); + + r->memsize = (unsigned long)elf64ph->p_memsz; + r->offset = (unsigned long)elf64ph->p_offset; + r->addr = (unsigned long long)elf64ph->p_vaddr; + +#ifdef DEBUG + printf("PPC64 ELF file, ph:\n"); + printf("p_type 0x%08x\n", elf64ph->p_type); + printf("p_flags 0x%08x\n", elf64ph->p_flags); + printf("p_offset 0x%016llx\n", elf64ph->p_offset); + printf("p_vaddr 0x%016llx\n", elf64ph->p_vaddr); + printf("p_paddr 0x%016llx\n", elf64ph->p_paddr); + printf("p_filesz 0x%016llx\n", elf64ph->p_filesz); + printf("p_memsz 0x%016llx\n", elf64ph->p_memsz); + printf("p_align 0x%016llx\n", elf64ph->p_align); + printf("... skipping 0x%08lx bytes of ELF header\n", + (unsigned long)elf64ph->p_offset); +#endif + + return 64; +} void get4k(FILE *file, char *buf ) { unsigned j; @@ -34,21 +82,17 @@ void death(const char *msg, FILE *fdesc, const char *fname) int main(int argc, char **argv) { char inbuf[4096]; + struct addr_range vmlinux; FILE *ramDisk; - FILE *sysmap; FILE *inputVmlinux; FILE *outputVmlinux; char *rd_name, *lx_name, *out_name; - unsigned i; + + size_t i; unsigned long ramFileLen; unsigned long ramLen; unsigned long roundR; - - unsigned long sysmapFileLen; - unsigned long sysmapLen; - unsigned long sysmapPages; - char *ptr_end; unsigned long offset_end; unsigned long kernelLen; @@ -70,24 +114,19 @@ int main(int argc, char **argv) fprintf(stderr, "Name of RAM disk file missing.\n"); exit(1); } - rd_name = argv[1] + rd_name = argv[1]; if (argc < 3) { - fprintf(stderr, "Name of System Map input file is missing.\n"); - exit(1); - } - - if (argc < 4) { fprintf(stderr, "Name of vmlinux file missing.\n"); exit(1); } - lx_name = argv[3]; + lx_name = argv[2]; - if (argc < 5) { + if (argc < 4) { fprintf(stderr, "Name of vmlinux output file missing.\n"); exit(1); } - out_name = argv[4]; + out_name = argv[3]; ramDisk = fopen(rd_name, "r"); @@ -96,12 +135,6 @@ int main(int argc, char **argv) exit(1); } - sysmap = fopen(argv[2], "r"); - if ( ! sysmap ) { - fprintf(stderr, "System Map file \"%s\" failed to open.\n", argv[2]); - exit(1); - } - inputVmlinux = fopen(lx_name, "r"); if ( ! inputVmlinux ) { fprintf(stderr, "vmlinux file \"%s\" failed to open.\n", lx_name); @@ -113,18 +146,24 @@ int main(int argc, char **argv) fprintf(stderr, "output vmlinux file \"%s\" failed to open.\n", out_name); exit(1); } - - - + + i = fread(inbuf, 1, sizeof(inbuf), inputVmlinux); + if (i != sizeof(inbuf)) { + fprintf(stderr, "can not read vmlinux file %s: %u\n", lx_name, i); + exit(1); + } + + i = check_elf64(inbuf, sizeof(inbuf), &vmlinux); + if (i == 0) { + fprintf(stderr, "You must have a linux kernel specified as argv[2]\n"); + exit(1); + } + /* Input Vmlinux file */ fseek(inputVmlinux, 0, SEEK_END); kernelLen = ftell(inputVmlinux); fseek(inputVmlinux, 0, SEEK_SET); printf("kernel file size = %d\n", kernelLen); - if ( kernelLen == 0 ) { - fprintf(stderr, "You must have a linux kernel specified as argv[3]\n"); - exit(1); - } actualKernelLen = kernelLen - ElfHeaderSize; @@ -138,39 +177,7 @@ int main(int argc, char **argv) roundedKernelPages = roundedKernelLen / 4096; printf("Vmlinux pages to copy = %ld/0x%lx \n", roundedKernelPages, roundedKernelPages); - - - /* Input System Map file */ - /* (needs to be processed simply to determine if we need to add pad pages due to the static variables not being included in the vmlinux) */ - fseek(sysmap, 0, SEEK_END); - sysmapFileLen = ftell(sysmap); - fseek(sysmap, 0, SEEK_SET); - printf("%s file size = %ld/0x%lx \n", argv[2], sysmapFileLen, sysmapFileLen); - - sysmapLen = sysmapFileLen; - - roundR = 4096 - (sysmapLen % 4096); - if (roundR) { - printf("Rounding System Map file up to a multiple of 4096, adding %ld/0x%lx \n", roundR, roundR); - sysmapLen += roundR; - } - printf("Rounded System Map size is %ld/0x%lx \n", sysmapLen, sysmapLen); - - /* Process the Sysmap file to determine where _end is */ - sysmapPages = sysmapLen / 4096; - /* read the whole file line by line, expect that it doesn't fail */ - while ( fgets(inbuf, 4096, sysmap) ) ; - /* search for _end in the last page of the system map */ - ptr_end = strstr(inbuf, " _end"); - if (!ptr_end) { - fprintf(stderr, "Unable to find _end in the sysmap file \n"); - fprintf(stderr, "inbuf: \n"); - fprintf(stderr, "%s \n", inbuf); - exit(1); - } - printf("Found _end in the last page of the sysmap - backing up 10 characters it looks like %s", ptr_end-10); - /* convert address of _end in system map to hex offset. */ - offset_end = (unsigned int)strtol(ptr_end-10, NULL, 16); + offset_end = _ALIGN_UP(vmlinux.memsize, 4096); /* calc how many pages we need to insert between the vmlinux and the start of the ram disk */ padPages = offset_end/4096 - roundedKernelPages; From c44bc68dc0d2ca95e48b7242fef48200cd03352d Mon Sep 17 00:00:00 2001 From: Olaf Hering Date: Wed, 9 Nov 2005 20:54:43 +0100 Subject: [PATCH 039/121] [PATCH] ppc64 boot: fix compile warnings Fix a few compile warnings arch/ppc64/boot/addRamDisk.c:166: warning: int format, long unsigned int arg (arg 2) arch/ppc64/boot/addRamDisk.c:170: warning: int format, long unsigned int arg (arg 2) arch/ppc64/boot/addRamDisk.c:265: warning: unsigned int format, long unsigned int arg (arg 2) arch/ppc64/boot/addRamDisk.c:302: warning: unsigned int format, long unsigned int arg (arg 3) Signed-off-by: Olaf Hering Signed-off-by: Paul Mackerras --- arch/ppc64/boot/addRamDisk.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/arch/ppc64/boot/addRamDisk.c b/arch/ppc64/boot/addRamDisk.c index d99c04a38e46..c02a99952be7 100644 --- a/arch/ppc64/boot/addRamDisk.c +++ b/arch/ppc64/boot/addRamDisk.c @@ -163,11 +163,11 @@ int main(int argc, char **argv) fseek(inputVmlinux, 0, SEEK_END); kernelLen = ftell(inputVmlinux); fseek(inputVmlinux, 0, SEEK_SET); - printf("kernel file size = %d\n", kernelLen); + printf("kernel file size = %lu\n", kernelLen); actualKernelLen = kernelLen - ElfHeaderSize; - printf("actual kernel length (minus ELF header) = %d\n", actualKernelLen); + printf("actual kernel length (minus ELF header) = %lu\n", actualKernelLen); round = actualKernelLen % 4096; roundedKernelLen = actualKernelLen; @@ -262,7 +262,7 @@ int main(int argc, char **argv) death("Could not read hvReleaseData pointer\n", outputVmlinux, out_name); } hvReleaseData = ntohl(hvReleaseData); /* Convert to native int */ - printf("hvReleaseData is at %08x\n", hvReleaseData); + printf("hvReleaseData is at %08lx\n", hvReleaseData); /* fseek to the hvReleaseData */ fseek(outputVmlinux, ElfHeaderSize + hvReleaseData, SEEK_SET); @@ -298,7 +298,7 @@ int main(int argc, char **argv) if (fwrite(inbuf, 0x18, 1, outputVmlinux) != 1) { death("Could not write naca\n", outputVmlinux, out_name); } - printf("Ram Disk of 0x%lx pages is attached to the kernel at offset 0x%08x\n", + printf("Ram Disk of 0x%lx pages is attached to the kernel at offset 0x%08lx\n", ramPages, ramStartOffs); /* Done */ From 77ac166fba909d520b9a21dcc28cb49081e31033 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 10 Nov 2005 11:12:11 +1100 Subject: [PATCH 040/121] [PATCH] ppc64: Don't panic when early __ioremap fails Early calls to __ioremap() will panic if the hash insertion fails. This patch makes them return NULL instead. It happens with some pSeries users who enabled CONFIG_BOOTX_TEXT. The later is getting an incorrect address for the fame buffer and the hash insertion fails. With this patch, it will display an error instead of crashing at boot. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Paul Mackerras --- arch/powerpc/mm/pgtable_64.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/mm/pgtable_64.c b/arch/powerpc/mm/pgtable_64.c index 900842451bd3..c7f7bb6f30b3 100644 --- a/arch/powerpc/mm/pgtable_64.c +++ b/arch/powerpc/mm/pgtable_64.c @@ -122,8 +122,11 @@ static int map_io_page(unsigned long ea, unsigned long pa, int flags) * */ if (htab_bolt_mapping(ea, ea + PAGE_SIZE, pa, flags, - mmu_virtual_psize)) - panic("Can't map bolted IO mapping"); + mmu_virtual_psize)) { + printk(KERN_ERR "Failed to do bolted mapping IO " + "memory at %016lx !\n", pa); + return -ENOMEM; + } } return 0; } From ae65a391caf1830aea9e4cadd69bb1aa69aad492 Mon Sep 17 00:00:00 2001 From: linas Date: Thu, 3 Nov 2005 18:42:26 -0600 Subject: [PATCH 041/121] [PATCH] ppc64: uniform usage of bus unit id interfaces 01-pci-dn-uniformization.patch This patch changes the rtas_pci interface to use the new struct pci_dn structure for two routines that work with pci device nodes. This patch also does some minor janitorial work: it uses some handy macros and cleans up some trailing whitespace in the affected file. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 4 --- arch/ppc64/kernel/rtas_pci.c | 47 ++++++++++++++++++----------------- include/asm-powerpc/ppc-pci.h | 4 +++ 3 files changed, 28 insertions(+), 27 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index 035d1b14a207..eedf21d5f3b7 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -71,10 +71,6 @@ * and sent out for processing. */ -/** Bus Unit ID macros; get low and hi 32-bits of the 64-bit BUID */ -#define BUID_HI(buid) ((buid) >> 32) -#define BUID_LO(buid) ((buid) & 0xffffffff) - /* EEH event workqueue setup. */ static DEFINE_SPINLOCK(eeh_eventlist_lock); LIST_HEAD(eeh_eventlist); diff --git a/arch/ppc64/kernel/rtas_pci.c b/arch/ppc64/kernel/rtas_pci.c index 3c3f19192fcc..0e5a8e116653 100644 --- a/arch/ppc64/kernel/rtas_pci.c +++ b/arch/ppc64/kernel/rtas_pci.c @@ -5,19 +5,19 @@ * Copyright (C) 2003 Anton Blanchard , IBM * * RTAS specific routines for PCI. - * + * * Based on code from pci.c, chrp_pci.c and pSeries_pci.c * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. - * + * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. - * + * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA @@ -47,7 +47,7 @@ static int write_pci_config; static int ibm_read_pci_config; static int ibm_write_pci_config; -static int config_access_valid(struct pci_dn *dn, int where) +static inline int config_access_valid(struct pci_dn *dn, int where) { if (where < 256) return 1; @@ -72,16 +72,14 @@ static int of_device_available(struct device_node * dn) return 0; } -static int rtas_read_config(struct device_node *dn, int where, int size, u32 *val) +static int rtas_read_config(struct pci_dn *pdn, int where, int size, u32 *val) { int returnval = -1; unsigned long buid, addr; int ret; - struct pci_dn *pdn; - if (!dn || !dn->data) + if (!pdn) return PCIBIOS_DEVICE_NOT_FOUND; - pdn = dn->data; if (!config_access_valid(pdn, where)) return PCIBIOS_BAD_REGISTER_NUMBER; @@ -90,7 +88,7 @@ static int rtas_read_config(struct device_node *dn, int where, int size, u32 *va buid = pdn->phb->buid; if (buid) { ret = rtas_call(ibm_read_pci_config, 4, 2, &returnval, - addr, buid >> 32, buid & 0xffffffff, size); + addr, BUID_HI(buid), BUID_LO(buid), size); } else { ret = rtas_call(read_pci_config, 2, 2, &returnval, addr, size); } @@ -100,7 +98,7 @@ static int rtas_read_config(struct device_node *dn, int where, int size, u32 *va return PCIBIOS_DEVICE_NOT_FOUND; if (returnval == EEH_IO_ERROR_VALUE(size) && - eeh_dn_check_failure (dn, NULL)) + eeh_dn_check_failure (pdn->node, NULL)) return PCIBIOS_DEVICE_NOT_FOUND; return PCIBIOS_SUCCESSFUL; @@ -118,23 +116,23 @@ static int rtas_pci_read_config(struct pci_bus *bus, busdn = bus->sysdata; /* must be a phb */ /* Search only direct children of the bus */ - for (dn = busdn->child; dn; dn = dn->sibling) - if (dn->data && PCI_DN(dn)->devfn == devfn + for (dn = busdn->child; dn; dn = dn->sibling) { + struct pci_dn *pdn = PCI_DN(dn); + if (pdn && pdn->devfn == devfn && of_device_available(dn)) - return rtas_read_config(dn, where, size, val); + return rtas_read_config(pdn, where, size, val); + } return PCIBIOS_DEVICE_NOT_FOUND; } -int rtas_write_config(struct device_node *dn, int where, int size, u32 val) +int rtas_write_config(struct pci_dn *pdn, int where, int size, u32 val) { unsigned long buid, addr; int ret; - struct pci_dn *pdn; - if (!dn || !dn->data) + if (!pdn) return PCIBIOS_DEVICE_NOT_FOUND; - pdn = dn->data; if (!config_access_valid(pdn, where)) return PCIBIOS_BAD_REGISTER_NUMBER; @@ -142,7 +140,8 @@ int rtas_write_config(struct device_node *dn, int where, int size, u32 val) (pdn->devfn << 8) | (where & 0xff); buid = pdn->phb->buid; if (buid) { - ret = rtas_call(ibm_write_pci_config, 5, 1, NULL, addr, buid >> 32, buid & 0xffffffff, size, (ulong) val); + ret = rtas_call(ibm_write_pci_config, 5, 1, NULL, addr, + BUID_HI(buid), BUID_LO(buid), size, (ulong) val); } else { ret = rtas_call(write_pci_config, 3, 1, NULL, addr, size, (ulong)val); } @@ -165,10 +164,12 @@ static int rtas_pci_write_config(struct pci_bus *bus, busdn = bus->sysdata; /* must be a phb */ /* Search only direct children of the bus */ - for (dn = busdn->child; dn; dn = dn->sibling) - if (dn->data && PCI_DN(dn)->devfn == devfn + for (dn = busdn->child; dn; dn = dn->sibling) { + struct pci_dn *pdn = PCI_DN(dn); + if (pdn && pdn->devfn == devfn && of_device_available(dn)) - return rtas_write_config(dn, where, size, val); + return rtas_write_config(pdn, where, size, val); + } return PCIBIOS_DEVICE_NOT_FOUND; } @@ -221,7 +222,7 @@ static void python_countermeasures(struct device_node *dev, /* Python's register file is 1 MB in size. */ chip_regs = ioremap(reg_struct.address & ~(0xfffffUL), 0x100000); - /* + /* * Firmware doesn't always clear this bit which is critical * for good performance - Anton */ @@ -292,7 +293,7 @@ static int phb_set_bus_ranges(struct device_node *dev, if (bus_range == NULL || len < 2 * sizeof(int)) { return 1; } - + phb->first_busno = bus_range[0]; phb->last_busno = bus_range[1]; diff --git a/include/asm-powerpc/ppc-pci.h b/include/asm-powerpc/ppc-pci.h index 13aacff755f3..302f0a9878c7 100644 --- a/include/asm-powerpc/ppc-pci.h +++ b/include/asm-powerpc/ppc-pci.h @@ -26,6 +26,10 @@ extern unsigned long find_and_init_phbs(void); extern struct pci_dev *ppc64_isabridge_dev; /* may be NULL if no ISA bus */ +/** Bus Unit ID macros; get low and hi 32-bits of the 64-bit BUID */ +#define BUID_HI(buid) ((buid) >> 32) +#define BUID_LO(buid) ((buid) & 0xffffffff) + /* PCI device_node operations */ struct device_node; typedef void *(*traverse_func)(struct device_node *me, void *data); From 693765021460f19c5d7c638eb76e2efe1cd84752 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:47:50 -0600 Subject: [PATCH 042/121] [PATCH] ppc64: misc minor cleanup 02-eeh-minor-cleanup.patch This patch performs some minor cleanup of the eeh.c file, including: -- trim some trailing whitespace -- remove extraneous #includes -- use the macro PCI_DN uniformly, instead of the void pointer chase. -- typos in comments -- improved debug printk's Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 53 +++++++++++++++++++---------------------- 1 file changed, 25 insertions(+), 28 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index eedf21d5f3b7..bb11569d2b4b 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -1,32 +1,31 @@ /* * eeh.c * Copyright (C) 2001 Dave Engebretsen & Todd Inglett IBM Corporation - * + * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. - * + * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. - * + * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ -#include #include #include -#include #include #include #include #include #include #include +#include #include #include #include @@ -49,8 +48,8 @@ * were "empty": all reads return 0xff's and all writes are silently * ignored. EEH slot isolation events can be triggered by parity * errors on the address or data busses (e.g. during posted writes), - * which in turn might be caused by dust, vibration, humidity, - * radioactivity or plain-old failed hardware. + * which in turn might be caused by low voltage on the bus, dust, + * vibration, humidity, radioactivity or plain-old failed hardware. * * Note, however, that one of the leading causes of EEH slot * freeze events are buggy device drivers, buggy device microcode, @@ -256,18 +255,17 @@ static void __pci_addr_cache_insert_device(struct pci_dev *dev) dn = pci_device_to_OF_node(dev); if (!dn) { - printk(KERN_WARNING "PCI: no pci dn found for dev=%s\n", - pci_name(dev)); + printk(KERN_WARNING "PCI: no pci dn found for dev=%s\n", pci_name(dev)); return; } /* Skip any devices for which EEH is not enabled. */ - pdn = dn->data; + pdn = PCI_DN(dn); if (!(pdn->eeh_mode & EEH_MODE_SUPPORTED) || pdn->eeh_mode & EEH_MODE_NOCHECK) { #ifdef DEBUG - printk(KERN_INFO "PCI: skip building address cache for=%s\n", - pci_name(dev)); + printk(KERN_INFO "PCI: skip building address cache for=%s - %s\n", + pci_name(dev), pdn->node->full_name); #endif return; } @@ -410,16 +408,16 @@ int eeh_unregister_notifier(struct notifier_block *nb) * @dn: device node to read * @rets: array to return results in */ -static int read_slot_reset_state(struct device_node *dn, int rets[]) +static int read_slot_reset_state(struct pci_dn *pdn, int rets[]) { int token, outputs; - struct pci_dn *pdn = dn->data; if (ibm_read_slot_reset_state2 != RTAS_UNKNOWN_SERVICE) { token = ibm_read_slot_reset_state2; outputs = 4; } else { token = ibm_read_slot_reset_state; + rets[2] = 0; /* fake PE Unavailable info */ outputs = 3; } @@ -496,7 +494,7 @@ static void eeh_event_handler(void *dummy) /** * eeh_token_to_phys - convert EEH address token to phys address - * @token i/o token, should be address in the form 0xE.... + * @token i/o token, should be address in the form 0xA.... */ static inline unsigned long eeh_token_to_phys(unsigned long token) { @@ -522,7 +520,7 @@ static inline unsigned long eeh_token_to_phys(unsigned long token) * will query firmware for the EEH status. * * Returns 0 if there has not been an EEH error; otherwise returns - * a non-zero value and queues up a solt isolation event notification. + * a non-zero value and queues up a slot isolation event notification. * * It is safe to call this routine in an interrupt context. */ @@ -542,7 +540,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) if (!dn) return 0; - pdn = dn->data; + pdn = PCI_DN(dn); /* Access to IO BARs might get this far and still not want checking. */ if (!pdn->eeh_capable || !(pdn->eeh_mode & EEH_MODE_SUPPORTED) || @@ -562,7 +560,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) atomic_inc(&eeh_fail_count); if (atomic_read(&eeh_fail_count) >= EEH_MAX_FAILS) { /* re-read the slot reset state */ - if (read_slot_reset_state(dn, rets) != 0) + if (read_slot_reset_state(pdn, rets) != 0) rets[0] = -1; /* reset state unknown */ eeh_panic(dev, rets[0]); } @@ -576,7 +574,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) * function zero of a multi-function device. * In any case they must share a common PHB. */ - ret = read_slot_reset_state(dn, rets); + ret = read_slot_reset_state(pdn, rets); if (!(ret == 0 && rets[1] == 1 && (rets[0] == 2 || rets[0] == 4))) { __get_cpu_var(false_positives)++; return 0; @@ -635,7 +633,6 @@ EXPORT_SYMBOL(eeh_dn_check_failure); * @token i/o token, should be address in the form 0xA.... * @val value, should be all 1's (XXX why do we need this arg??) * - * Check for an eeh failure at the given token address. * Check for an EEH failure at the given token address. Call this * routine if the result of a read was all 0xff's and you want to * find out if this is due to an EEH slot freeze event. This routine @@ -680,7 +677,7 @@ static void *early_enable_eeh(struct device_node *dn, void *data) u32 *device_id = (u32 *)get_property(dn, "device-id", NULL); u32 *regs; int enable; - struct pci_dn *pdn = dn->data; + struct pci_dn *pdn = PCI_DN(dn); pdn->eeh_mode = 0; @@ -732,7 +729,7 @@ static void *early_enable_eeh(struct device_node *dn, void *data) /* This device doesn't support EEH, but it may have an * EEH parent, in which case we mark it as supported. */ - if (dn->parent && dn->parent->data + if (dn->parent && PCI_DN(dn->parent) && (PCI_DN(dn->parent)->eeh_mode & EEH_MODE_SUPPORTED)) { /* Parent supports EEH. */ pdn->eeh_mode |= EEH_MODE_SUPPORTED; @@ -745,7 +742,7 @@ static void *early_enable_eeh(struct device_node *dn, void *data) dn->full_name); } - return NULL; + return NULL; } /* @@ -793,13 +790,11 @@ void __init eeh_init(void) for (phb = of_find_node_by_name(NULL, "pci"); phb; phb = of_find_node_by_name(phb, "pci")) { unsigned long buid; - struct pci_dn *pci; buid = get_phb_buid(phb); - if (buid == 0 || phb->data == NULL) + if (buid == 0 || PCI_DN(phb) == NULL) continue; - pci = phb->data; info.buid_lo = BUID_LO(buid); info.buid_hi = BUID_HI(buid); traverse_pci_devices(phb, early_enable_eeh, &info); @@ -828,11 +823,13 @@ void eeh_add_device_early(struct device_node *dn) struct pci_controller *phb; struct eeh_early_enable_info info; - if (!dn || !dn->data) + if (!dn || !PCI_DN(dn)) return; phb = PCI_DN(dn)->phb; if (NULL == phb || 0 == phb->buid) { - printk(KERN_WARNING "EEH: Expected buid but found none\n"); + printk(KERN_WARNING "EEH: Expected buid but found none for %s\n", + dn->full_name); + dump_stack(); return; } From 56b0fca3a008ddec8c00cfdddbef17bd9ce188e3 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:48:45 -0600 Subject: [PATCH 043/121] [PATCH] ppc64: PCI address cache minor fixes 03-eeh-addr-cache-cleanup.patch This is a minor patch to clean up a buglet related to the PCI address cache. (The buglet doesn't manifes itself unless there are also bugs elsewhere, which is why its minor.). Also: -- Improved debug printing. -- Declare some private routines as static -- Adds reference counting to struct pci_dn->pcidev structure Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 33 ++++++++++++++++++++++++++------- include/asm-powerpc/ppc-pci.h | 4 ---- 2 files changed, 26 insertions(+), 11 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index bb11569d2b4b..99f11b66b5a0 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -219,9 +219,9 @@ pci_addr_cache_insert(struct pci_dev *dev, unsigned long alo, while (*p) { parent = *p; piar = rb_entry(parent, struct pci_io_addr_range, rb_node); - if (alo < piar->addr_lo) { + if (ahi < piar->addr_lo) { p = &parent->rb_left; - } else if (ahi > piar->addr_hi) { + } else if (alo > piar->addr_hi) { p = &parent->rb_right; } else { if (dev != piar->pcidev || @@ -240,6 +240,11 @@ pci_addr_cache_insert(struct pci_dev *dev, unsigned long alo, piar->pcidev = dev; piar->flags = flags; +#ifdef DEBUG + printk(KERN_DEBUG "PIAR: insert range=[%lx:%lx] dev=%s\n", + alo, ahi, pci_name (dev)); +#endif + rb_link_node(&piar->rb_node, parent, p); rb_insert_color(&piar->rb_node, &pci_io_addr_cache_root.rb_root); @@ -301,7 +306,7 @@ static void __pci_addr_cache_insert_device(struct pci_dev *dev) * we maintain a cache of devices that can be quickly searched. * This routine adds a device to that cache. */ -void pci_addr_cache_insert_device(struct pci_dev *dev) +static void pci_addr_cache_insert_device(struct pci_dev *dev) { unsigned long flags; @@ -344,7 +349,7 @@ restart: * the tree multiple times (once per resource). * But so what; device removal doesn't need to be that fast. */ -void pci_addr_cache_remove_device(struct pci_dev *dev) +static void pci_addr_cache_remove_device(struct pci_dev *dev) { unsigned long flags; @@ -366,6 +371,9 @@ void __init pci_addr_cache_build(void) { struct pci_dev *dev = NULL; + if (!eeh_subsystem_enabled) + return; + spin_lock_init(&pci_io_addr_cache_root.piar_lock); while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) { @@ -837,7 +845,7 @@ void eeh_add_device_early(struct device_node *dn) info.buid_lo = BUID_LO(phb->buid); early_enable_eeh(dn, &info); } -EXPORT_SYMBOL(eeh_add_device_early); +EXPORT_SYMBOL_GPL(eeh_add_device_early); /** * eeh_add_device_late - perform EEH initialization for the indicated pci device @@ -848,6 +856,8 @@ EXPORT_SYMBOL(eeh_add_device_early); */ void eeh_add_device_late(struct pci_dev *dev) { + struct device_node *dn; + if (!dev || !eeh_subsystem_enabled) return; @@ -855,9 +865,13 @@ void eeh_add_device_late(struct pci_dev *dev) printk(KERN_DEBUG "EEH: adding device %s\n", pci_name(dev)); #endif + pci_dev_get (dev); + dn = pci_device_to_OF_node(dev); + PCI_DN(dn)->pcidev = dev; + pci_addr_cache_insert_device (dev); } -EXPORT_SYMBOL(eeh_add_device_late); +EXPORT_SYMBOL_GPL(eeh_add_device_late); /** * eeh_remove_device - undo EEH setup for the indicated pci device @@ -868,6 +882,7 @@ EXPORT_SYMBOL(eeh_add_device_late); */ void eeh_remove_device(struct pci_dev *dev) { + struct device_node *dn; if (!dev || !eeh_subsystem_enabled) return; @@ -876,8 +891,12 @@ void eeh_remove_device(struct pci_dev *dev) printk(KERN_DEBUG "EEH: remove device %s\n", pci_name(dev)); #endif pci_addr_cache_remove_device(dev); + + dn = pci_device_to_OF_node(dev); + PCI_DN(dn)->pcidev = NULL; + pci_dev_put (dev); } -EXPORT_SYMBOL(eeh_remove_device); +EXPORT_SYMBOL_GPL(eeh_remove_device); static int proc_eeh_show(struct seq_file *m, void *v) { diff --git a/include/asm-powerpc/ppc-pci.h b/include/asm-powerpc/ppc-pci.h index 302f0a9878c7..c47146f74a7b 100644 --- a/include/asm-powerpc/ppc-pci.h +++ b/include/asm-powerpc/ppc-pci.h @@ -40,10 +40,6 @@ void pci_devs_phb_init(void); void pci_devs_phb_init_dynamic(struct pci_controller *phb); void __devinit scan_phb(struct pci_controller *hose); -/* PCI address cache management routines */ -void pci_addr_cache_insert_device(struct pci_dev *dev); -void pci_addr_cache_remove_device(struct pci_dev *dev); - /* From rtas_pci.h */ void init_pci_config_tokens (void); unsigned long get_phb_buid (struct device_node *); From 177bc9367efe89a642bc9a07351726fd2c31f5ff Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:48:52 -0600 Subject: [PATCH 044/121] [PATCH] ppc64: PCI error rate statistics 04-eeh-statistics.patch This minor patch adds some statistics-gathering counters that allow the behaviour of the EEH subsystem o be monitored. While far from perfect, it does provide a rudimentary device that makes understanding of the current state of the system a bit easier. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 43 ++++++++++++++++++++++++++++++----------- 1 file changed, 32 insertions(+), 11 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index 99f11b66b5a0..926d3110dfd7 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -102,6 +102,10 @@ static DEFINE_SPINLOCK(slot_errbuf_lock); static int eeh_error_buf_size; /* System monitoring statistics */ +static DEFINE_PER_CPU(unsigned long, no_device); +static DEFINE_PER_CPU(unsigned long, no_dn); +static DEFINE_PER_CPU(unsigned long, no_cfg_addr); +static DEFINE_PER_CPU(unsigned long, ignored_check); static DEFINE_PER_CPU(unsigned long, total_mmio_ffs); static DEFINE_PER_CPU(unsigned long, false_positives); static DEFINE_PER_CPU(unsigned long, ignored_failures); @@ -493,8 +497,6 @@ static void eeh_event_handler(void *dummy) notifier_call_chain (&eeh_notifier_chain, EEH_NOTIFY_FREEZE, event); - __get_cpu_var(slot_resets)++; - pci_dev_put(event->dev); kfree(event); } @@ -546,17 +548,24 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) if (!eeh_subsystem_enabled) return 0; - if (!dn) + if (!dn) { + __get_cpu_var(no_dn)++; return 0; + } pdn = PCI_DN(dn); /* Access to IO BARs might get this far and still not want checking. */ if (!pdn->eeh_capable || !(pdn->eeh_mode & EEH_MODE_SUPPORTED) || pdn->eeh_mode & EEH_MODE_NOCHECK) { + __get_cpu_var(ignored_check)++; +#ifdef DEBUG + printk ("EEH:ignored check for %s %s\n", pci_name (dev), dn->full_name); +#endif return 0; } if (!pdn->eeh_config_addr) { + __get_cpu_var(no_cfg_addr)++; return 0; } @@ -590,6 +599,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) /* prevent repeated reports of this failure */ pdn->eeh_mode |= EEH_MODE_ISOLATED; + __get_cpu_var(slot_resets)++; reset_state = rets[0]; @@ -657,8 +667,10 @@ unsigned long eeh_check_failure(const volatile void __iomem *token, unsigned lon /* Finding the phys addr + pci device; this is pretty quick. */ addr = eeh_token_to_phys((unsigned long __force) token); dev = pci_get_device_by_addr(addr); - if (!dev) + if (!dev) { + __get_cpu_var(no_device)++; return val; + } dn = pci_device_to_OF_node(dev); eeh_dn_check_failure (dn, dev); @@ -903,12 +915,17 @@ static int proc_eeh_show(struct seq_file *m, void *v) unsigned int cpu; unsigned long ffs = 0, positives = 0, failures = 0; unsigned long resets = 0; + unsigned long no_dev = 0, no_dn = 0, no_cfg = 0, no_check = 0; for_each_cpu(cpu) { ffs += per_cpu(total_mmio_ffs, cpu); positives += per_cpu(false_positives, cpu); failures += per_cpu(ignored_failures, cpu); resets += per_cpu(slot_resets, cpu); + no_dev += per_cpu(no_device, cpu); + no_dn += per_cpu(no_dn, cpu); + no_cfg += per_cpu(no_cfg_addr, cpu); + no_check += per_cpu(ignored_check, cpu); } if (0 == eeh_subsystem_enabled) { @@ -916,13 +933,17 @@ static int proc_eeh_show(struct seq_file *m, void *v) seq_printf(m, "eeh_total_mmio_ffs=%ld\n", ffs); } else { seq_printf(m, "EEH Subsystem is enabled\n"); - seq_printf(m, "eeh_total_mmio_ffs=%ld\n" - "eeh_false_positives=%ld\n" - "eeh_ignored_failures=%ld\n" - "eeh_slot_resets=%ld\n" - "eeh_fail_count=%d\n", - ffs, positives, failures, resets, - eeh_fail_count.counter); + seq_printf(m, + "no device=%ld\n" + "no device node=%ld\n" + "no config address=%ld\n" + "check not wanted=%ld\n" + "eeh_total_mmio_ffs=%ld\n" + "eeh_false_positives=%ld\n" + "eeh_ignored_failures=%ld\n" + "eeh_slot_resets=%ld\n", + no_dev, no_dn, no_cfg, no_check, + ffs, positives, failures, resets); } return 0; From df7242b1156966c3b1aa0fd2bc63e3736099b592 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:49:01 -0600 Subject: [PATCH 045/121] [PATCH] ppc64: RTAS error reporting restructuring 05-eeh-slot-error-detail.patch This patch encapsulates a section of code that reports the EEH event. The new subroutine can be used in several places to report the error. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 46 +++++++++++++++++++++++++++-------------- 1 file changed, 30 insertions(+), 16 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index 926d3110dfd7..b23159464abb 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -397,6 +397,28 @@ void __init pci_addr_cache_build(void) /* --------------------------------------------------------------- */ /* Above lies the PCI Address Cache. Below lies the EEH event infrastructure */ +void eeh_slot_error_detail (struct pci_dn *pdn, int severity) +{ + unsigned long flags; + int rc; + + /* Log the error with the rtas logger */ + spin_lock_irqsave(&slot_errbuf_lock, flags); + memset(slot_errbuf, 0, eeh_error_buf_size); + + rc = rtas_call(ibm_slot_error_detail, + 8, 1, NULL, pdn->eeh_config_addr, + BUID_HI(pdn->phb->buid), + BUID_LO(pdn->phb->buid), NULL, 0, + virt_to_phys(slot_errbuf), + eeh_error_buf_size, + severity); + + if (rc == 0) + log_error(slot_errbuf, ERR_TYPE_RTAS_LOG, 0); + spin_unlock_irqrestore(&slot_errbuf_lock, flags); +} + /** * eeh_register_notifier - Register to find out about EEH events. * @nb: notifier block to callback on events @@ -454,9 +476,12 @@ static void eeh_panic(struct pci_dev *dev, int reset_state) * Since the panic_on_oops sysctl is used to halt the system * in light of potential corruption, we can use it here. */ - if (panic_on_oops) + if (panic_on_oops) { + struct device_node *dn = pci_device_to_OF_node(dev); + eeh_slot_error_detail (PCI_DN(dn), 2 /* Permanent Error */); panic("EEH: MMIO failure (%d) on device:%s\n", reset_state, pci_name(dev)); + } else { __get_cpu_var(ignored_failures)++; printk(KERN_INFO "EEH: Ignored MMIO failure (%d) on device:%s\n", @@ -539,7 +564,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) int ret; int rets[3]; unsigned long flags; - int rc, reset_state; + int reset_state; struct eeh_event *event; struct pci_dn *pdn; @@ -603,20 +628,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) reset_state = rets[0]; - spin_lock_irqsave(&slot_errbuf_lock, flags); - memset(slot_errbuf, 0, eeh_error_buf_size); - - rc = rtas_call(ibm_slot_error_detail, - 8, 1, NULL, pdn->eeh_config_addr, - BUID_HI(pdn->phb->buid), - BUID_LO(pdn->phb->buid), NULL, 0, - virt_to_phys(slot_errbuf), - eeh_error_buf_size, - 1 /* Temporary Error */); - - if (rc == 0) - log_error(slot_errbuf, ERR_TYPE_RTAS_LOG, 0); - spin_unlock_irqrestore(&slot_errbuf_lock, flags); + eeh_slot_error_detail (pdn, 1 /* Temporary Error */); printk(KERN_INFO "EEH: MMIO failure (%d) on device: %s %s\n", rets[0], dn->name, dn->full_name); @@ -783,6 +795,8 @@ void __init eeh_init(void) struct device_node *phb, *np; struct eeh_early_enable_info info; + spin_lock_init(&slot_errbuf_lock); + np = of_find_node_by_path("/rtas"); if (np == NULL) return; From 76e6faf7a3a3ad3e18a1b70f9e4cd96cdf58140d Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:49:15 -0600 Subject: [PATCH 046/121] [PATCH] ppc64: avoid PCI error reporting for empty slots 06-eeh-empty-slot-error.patch Performing PCI config-space reads to empty PCI slots can lead to reports of "permanent failure" from the firmware. Ignore permanent failures on empty slots. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 29 +++++++++++++++++++++++++++-- 1 file changed, 27 insertions(+), 2 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index b23159464abb..0060934dffd2 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -617,7 +617,32 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) * In any case they must share a common PHB. */ ret = read_slot_reset_state(pdn, rets); - if (!(ret == 0 && rets[1] == 1 && (rets[0] == 2 || rets[0] == 4))) { + + /* If the call to firmware failed, punt */ + if (ret != 0) { + printk(KERN_WARNING "EEH: read_slot_reset_state() failed; rc=%d dn=%s\n", + ret, dn->full_name); + __get_cpu_var(false_positives)++; + return 0; + } + + /* If EEH is not supported on this device, punt. */ + if (rets[1] != 1) { + printk(KERN_WARNING "EEH: event on unsupported device, rc=%d dn=%s\n", + ret, dn->full_name); + __get_cpu_var(false_positives)++; + return 0; + } + + /* If not the kind of error we know about, punt. */ + if (rets[0] != 2 && rets[0] != 4 && rets[0] != 5) { + __get_cpu_var(false_positives)++; + return 0; + } + + /* Note that config-io to empty slots may fail; + * we recognize empty because they don't have children. */ + if ((rets[0] == 5) && (dn->child == NULL)) { __get_cpu_var(false_positives)++; return 0; } @@ -650,7 +675,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) /* Most EEH events are due to device driver bugs. Having * a stack trace will help the device-driver authors figure * out what happened. So print that out. */ - dump_stack(); + if (rets[0] != 5) dump_stack(); schedule_work(&eeh_event_wq); return 0; From fd761fd876e4d1c0d07b6d93bc45c999fa596cb0 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:49:23 -0600 Subject: [PATCH 047/121] [PATCH] ppc64: serialize reports of PCI errors 07-eeh-report-race.patch When a PCI slot is isolated, all PCI functions under that slot are affected. If hese functions have separate device drivers, the EEH isolation event might be reported multiple times. This patch adds a lock to prevent the racing of such multiple reports. It also marks every device under the slot as having experienced an EEH event, so that multiple reports may be recognized more easily. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 98 +++++++++++++++++++++++++++++++++++------ 1 file changed, 85 insertions(+), 13 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index 0060934dffd2..e7522f6da69d 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -96,6 +96,9 @@ static int ibm_slot_error_detail; static int eeh_subsystem_enabled; +/* Lock to avoid races due to multiple reports of an error */ +static DEFINE_SPINLOCK(confirm_error_lock); + /* Buffer for reporting slot-error-detail rtas calls */ static unsigned char slot_errbuf[RTAS_ERROR_LOG_MAX]; static DEFINE_SPINLOCK(slot_errbuf_lock); @@ -544,6 +547,55 @@ static inline unsigned long eeh_token_to_phys(unsigned long token) return pa | (token & (PAGE_SIZE-1)); } +/** + * Return the "partitionable endpoint" (pe) under which this device lies + */ +static struct device_node * find_device_pe(struct device_node *dn) +{ + while ((dn->parent) && PCI_DN(dn->parent) && + (PCI_DN(dn->parent)->eeh_mode & EEH_MODE_SUPPORTED)) { + dn = dn->parent; + } + return dn; +} + +/** Mark all devices that are peers of this device as failed. + * Mark the device driver too, so that it can see the failure + * immediately; this is critical, since some drivers poll + * status registers in interrupts ... If a driver is polling, + * and the slot is frozen, then the driver can deadlock in + * an interrupt context, which is bad. + */ + +static inline void __eeh_mark_slot (struct device_node *dn) +{ + while (dn) { + PCI_DN(dn)->eeh_mode |= EEH_MODE_ISOLATED; + + if (dn->child) + __eeh_mark_slot (dn->child); + dn = dn->sibling; + } +} + +static inline void __eeh_clear_slot (struct device_node *dn) +{ + while (dn) { + PCI_DN(dn)->eeh_mode &= ~EEH_MODE_ISOLATED; + if (dn->child) + __eeh_clear_slot (dn->child); + dn = dn->sibling; + } +} + +static inline void eeh_clear_slot (struct device_node *dn) +{ + unsigned long flags; + spin_lock_irqsave(&confirm_error_lock, flags); + __eeh_clear_slot (dn); + spin_unlock_irqrestore(&confirm_error_lock, flags); +} + /** * eeh_dn_check_failure - check if all 1's data is due to EEH slot freeze * @dn device node @@ -567,6 +619,8 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) int reset_state; struct eeh_event *event; struct pci_dn *pdn; + struct device_node *pe_dn; + int rc = 0; __get_cpu_var(total_mmio_ffs)++; @@ -594,10 +648,14 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) return 0; } - /* - * If we already have a pending isolation event for this - * slot, we know it's bad already, we don't need to check... + /* If we already have a pending isolation event for this + * slot, we know it's bad already, we don't need to check. + * Do this checking under a lock; as multiple PCI devices + * in one slot might report errors simultaneously, and we + * only want one error recovery routine running. */ + spin_lock_irqsave(&confirm_error_lock, flags); + rc = 1; if (pdn->eeh_mode & EEH_MODE_ISOLATED) { atomic_inc(&eeh_fail_count); if (atomic_read(&eeh_fail_count) >= EEH_MAX_FAILS) { @@ -606,7 +664,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) rets[0] = -1; /* reset state unknown */ eeh_panic(dev, rets[0]); } - return 0; + goto dn_unlock; } /* @@ -623,7 +681,8 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) printk(KERN_WARNING "EEH: read_slot_reset_state() failed; rc=%d dn=%s\n", ret, dn->full_name); __get_cpu_var(false_positives)++; - return 0; + rc = 0; + goto dn_unlock; } /* If EEH is not supported on this device, punt. */ @@ -631,25 +690,33 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) printk(KERN_WARNING "EEH: event on unsupported device, rc=%d dn=%s\n", ret, dn->full_name); __get_cpu_var(false_positives)++; - return 0; + rc = 0; + goto dn_unlock; } /* If not the kind of error we know about, punt. */ if (rets[0] != 2 && rets[0] != 4 && rets[0] != 5) { __get_cpu_var(false_positives)++; - return 0; + rc = 0; + goto dn_unlock; } /* Note that config-io to empty slots may fail; * we recognize empty because they don't have children. */ if ((rets[0] == 5) && (dn->child == NULL)) { __get_cpu_var(false_positives)++; - return 0; + rc = 0; + goto dn_unlock; } - /* prevent repeated reports of this failure */ - pdn->eeh_mode |= EEH_MODE_ISOLATED; - __get_cpu_var(slot_resets)++; + __get_cpu_var(slot_resets)++; + + /* Avoid repeated reports of this failure, including problems + * with other functions on this device, and functions under + * bridges. */ + pe_dn = find_device_pe (dn); + __eeh_mark_slot (pe_dn); + spin_unlock_irqrestore(&confirm_error_lock, flags); reset_state = rets[0]; @@ -678,10 +745,14 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) if (rets[0] != 5) dump_stack(); schedule_work(&eeh_event_wq); - return 0; + return 1; + +dn_unlock: + spin_unlock_irqrestore(&confirm_error_lock, flags); + return rc; } -EXPORT_SYMBOL(eeh_dn_check_failure); +EXPORT_SYMBOL_GPL(eeh_dn_check_failure); /** * eeh_check_failure - check if all 1's data is due to EEH slot freeze @@ -820,6 +891,7 @@ void __init eeh_init(void) struct device_node *phb, *np; struct eeh_early_enable_info info; + spin_lock_init(&confirm_error_lock); spin_lock_init(&slot_errbuf_lock); np = of_find_node_by_path("/rtas"); From 5c1344e988c716470b5609708215dd7f135f0e3a Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:49:31 -0600 Subject: [PATCH 048/121] [PATCH] ppc64: escape hatch for spinning interrupt deadlocks 08-eeh-spin-counter.patch One an EEH event is triggers, all further I/O to a device is blocked (until reset). Bad device drivers may end up spinning in their interrupt handlers, trying to read an interrupt status register that will never change state. This patch moves that spin counter to a per-device structure, and adds some diagnostic prints to help locate the bad driver. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 21 +++++++++++++-------- 1 file changed, 13 insertions(+), 8 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index e7522f6da69d..0c52c2de92e0 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -78,14 +78,12 @@ DECLARE_WORK(eeh_event_wq, eeh_event_handler, NULL); static struct notifier_block *eeh_notifier_chain; -/* - * If a device driver keeps reading an MMIO register in an interrupt +/* If a device driver keeps reading an MMIO register in an interrupt * handler after a slot isolation event has occurred, we assume it * is broken and panic. This sets the threshold for how many read * attempts we allow before panicking. */ -#define EEH_MAX_FAILS 1000 -static atomic_t eeh_fail_count; +#define EEH_MAX_FAILS 100000 /* RTAS tokens */ static int ibm_set_eeh_option; @@ -521,7 +519,6 @@ static void eeh_event_handler(void *dummy) "%s\n", event->reset_state, pci_name(event->dev)); - atomic_set(&eeh_fail_count, 0); notifier_call_chain (&eeh_notifier_chain, EEH_NOTIFY_FREEZE, event); @@ -657,12 +654,18 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) spin_lock_irqsave(&confirm_error_lock, flags); rc = 1; if (pdn->eeh_mode & EEH_MODE_ISOLATED) { - atomic_inc(&eeh_fail_count); - if (atomic_read(&eeh_fail_count) >= EEH_MAX_FAILS) { + pdn->eeh_check_count ++; + if (pdn->eeh_check_count >= EEH_MAX_FAILS) { + printk (KERN_ERR "EEH: Device driver ignored %d bad reads, panicing\n", + pdn->eeh_check_count); + dump_stack(); + /* re-read the slot reset state */ if (read_slot_reset_state(pdn, rets) != 0) rets[0] = -1; /* reset state unknown */ - eeh_panic(dev, rets[0]); + + /* If we are here, then we hit an infinite loop. Stop. */ + panic("EEH: MMIO halt (%d) on device:%s\n", rets[0], pci_name(dev)); } goto dn_unlock; } @@ -808,6 +811,8 @@ static void *early_enable_eeh(struct device_node *dn, void *data) struct pci_dn *pdn = PCI_DN(dn); pdn->eeh_mode = 0; + pdn->eeh_check_count = 0; + pdn->eeh_freeze_count = 0; if (status && strcmp(status, "ok") != 0) return NULL; /* ignore devices with bad status */ From 18126f35f69c928af26ec1dda2cb91b23220cdd4 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:49:38 -0600 Subject: [PATCH 049/121] [PATCH] ppc64: bugfix: crash on PCI hotplug 09-hotplug-bugfix.patch In the current 2.6.14-rc2-git6 kernel, performing a Dynamic LPAR Add of a hotplug slot will crash the system, with the following (abbreviated) stack trace: cpu 0x3: Vector: 700 (Program Check) at [c000000053dff7f0] pc: c0000000004f5974: .__alloc_bootmem+0x0/0xb0 lr: c0000000000258a0: .update_dn_pci_info+0x108/0x118 c0000000000257c8 .update_dn_pci_info+0x30/0x118 (unreliable) c0000000000258fc .pci_dn_reconfig_notifier+0x4c/0x64 c000000000060754 .notifier_call_chain+0x68/0x9c The root cause was that __init __alloc_bootmem() was called long after boot had finished, resulting in a crash because this routine is undefined after boot time. The patch below fixes this crash, and adds some docs to clarify the code. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/pci_dn.c | 21 +++++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) diff --git a/arch/ppc64/kernel/pci_dn.c b/arch/ppc64/kernel/pci_dn.c index 1a443a7ada4c..12c4c9e9bbc7 100644 --- a/arch/ppc64/kernel/pci_dn.c +++ b/arch/ppc64/kernel/pci_dn.c @@ -43,7 +43,7 @@ static void * __devinit update_dn_pci_info(struct device_node *dn, void *data) u32 *regs; struct pci_dn *pdn; - if (phb->is_dynamic) + if (mem_init_done) pdn = kmalloc(sizeof(*pdn), GFP_KERNEL); else pdn = alloc_bootmem(sizeof(*pdn)); @@ -120,6 +120,14 @@ void *traverse_pci_devices(struct device_node *start, traverse_func pre, return NULL; } +/** + * pci_devs_phb_init_dynamic - setup pci devices under this PHB + * phb: pci-to-host bridge (top-level bridge connecting to cpu) + * + * This routine is called both during boot, (before the memory + * subsystem is set up, before kmalloc is valid) and during the + * dynamic lpar operation of adding a PHB to a running system. + */ void __devinit pci_devs_phb_init_dynamic(struct pci_controller *phb) { struct device_node * dn = (struct device_node *) phb->arch_data; @@ -201,9 +209,14 @@ static struct notifier_block pci_dn_reconfig_nb = { .notifier_call = pci_dn_reconfig_notifier, }; -/* - * Actually initialize the phbs. - * The buswalk on this phb has not happened yet. +/** + * pci_devs_phb_init - Initialize phbs and pci devs under them. + * + * This routine walks over all phb's (pci-host bridges) on the + * system, and sets up assorted pci-related structures + * (including pci info in the device node structs) for each + * pci device found underneath. This routine runs once, + * early in the boot sequence. */ void __init pci_devs_phb_init(void) { From f8632c822719cce08cfb128859e354007744cbba Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:49:45 -0600 Subject: [PATCH 050/121] [PATCH] ppc64: bugfix: don't silently ignore PCI errors 10-EEH-enable-bugfix.patch Bugfix: With the curent linux-2.6.14-rc2-git6, EEH errors are ignored because thier detection requires an unused, uninitialized flag to be set. This patch removes the unused flag. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/ppc64/kernel/eeh.c | 5 +++-- include/asm-ppc64/pci-bridge.h | 1 - 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/ppc64/kernel/eeh.c b/arch/ppc64/kernel/eeh.c index 0c52c2de92e0..9df1d5018363 100644 --- a/arch/ppc64/kernel/eeh.c +++ b/arch/ppc64/kernel/eeh.c @@ -631,11 +631,12 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) pdn = PCI_DN(dn); /* Access to IO BARs might get this far and still not want checking. */ - if (!pdn->eeh_capable || !(pdn->eeh_mode & EEH_MODE_SUPPORTED) || + if (!(pdn->eeh_mode & EEH_MODE_SUPPORTED) || pdn->eeh_mode & EEH_MODE_NOCHECK) { __get_cpu_var(ignored_check)++; #ifdef DEBUG - printk ("EEH:ignored check for %s %s\n", pci_name (dev), dn->full_name); + printk ("EEH:ignored check (%x) for %s %s\n", + pdn->eeh_mode, pci_name (dev), dn->full_name); #endif return 0; } diff --git a/include/asm-ppc64/pci-bridge.h b/include/asm-ppc64/pci-bridge.h index 60cf8c838af0..efbdaece0cf0 100644 --- a/include/asm-ppc64/pci-bridge.h +++ b/include/asm-ppc64/pci-bridge.h @@ -63,7 +63,6 @@ struct pci_dn { int devfn; /* for pci devices */ int eeh_mode; /* See eeh.h for possible EEH_MODEs */ int eeh_config_addr; - int eeh_capable; /* from firmware */ int eeh_check_count; /* # times driver ignored error */ int eeh_freeze_count; /* # times this device froze up. */ int eeh_is_bridge; /* device is pci-to-pci bridge */ From 7f79da7accd63a6adb84f4602f66779f6a701e7b Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:49:51 -0600 Subject: [PATCH 051/121] [PATCH] ppc64: move eeh.c to powerpc directory from ppc64 11-eeh-move-to-powerpc.patch Move arch/ppc64/kernel/eeh.c to arch//powerpc/platforms/pseries/eeh.c No other changes (except for Makefile to build it) Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/pseries/Makefile | 1 + arch/{ppc64/kernel => powerpc/platforms/pseries}/eeh.c | 0 arch/ppc64/kernel/Makefile | 1 - 3 files changed, 1 insertion(+), 1 deletion(-) rename arch/{ppc64/kernel => powerpc/platforms/pseries}/eeh.c (100%) diff --git a/arch/powerpc/platforms/pseries/Makefile b/arch/powerpc/platforms/pseries/Makefile index b9938fece781..dbdffb2fe429 100644 --- a/arch/powerpc/platforms/pseries/Makefile +++ b/arch/powerpc/platforms/pseries/Makefile @@ -3,3 +3,4 @@ obj-y := pci.o lpar.o hvCall.o nvram.o reconfig.o \ obj-$(CONFIG_SMP) += smp.o obj-$(CONFIG_IBMVIO) += vio.o obj-$(CONFIG_XICS) += xics.o +obj-$(CONFIG_EEH) += eeh.o diff --git a/arch/ppc64/kernel/eeh.c b/arch/powerpc/platforms/pseries/eeh.c similarity index 100% rename from arch/ppc64/kernel/eeh.c rename to arch/powerpc/platforms/pseries/eeh.c diff --git a/arch/ppc64/kernel/Makefile b/arch/ppc64/kernel/Makefile index eb3187f18fb0..7548968b7997 100644 --- a/arch/ppc64/kernel/Makefile +++ b/arch/ppc64/kernel/Makefile @@ -30,7 +30,6 @@ endif obj-$(CONFIG_PPC_PSERIES) += udbg_16550.o obj-$(CONFIG_KEXEC) += machine_kexec.o -obj-$(CONFIG_EEH) += eeh.o obj-$(CONFIG_PROC_FS) += proc_ppc64.o obj-$(CONFIG_MODULES) += module.o ifneq ($(CONFIG_PPC_MERGE),y) From 172ca9261800bacbbc7d320d9924d9b482dff8de Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:50:04 -0600 Subject: [PATCH 052/121] [PATCH] ppc64: PCI error event dispatcher 12-eeh-event-dispatcher.patch ppc64: EEH Recovery dispatcher thread This patch adds a mechanism to create recovery threads when an EEH event is received. Since an EEH freeze state may be detected within an interrupt context, we need to get out of the interrupt context before starting recovery. This dispatcher does this in two steps: first, it uses a workqueue to get out, and then lanuches a kernel thread, so that the recovery routine can sleep for exteded periods without upseting the keventd. A kernel thread is created with each EEH event, rather than having one long-running daemon started at boot time. This is because it is anticipated that EEH events will be very rare (very very rare, ideally) and so its pointless to cluter the process tables with a daemon that will almost never run. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/pseries/Makefile | 2 +- arch/powerpc/platforms/pseries/eeh.c | 140 +++---------------- arch/powerpc/platforms/pseries/eeh_event.c | 155 +++++++++++++++++++++ include/asm-powerpc/eeh_event.h | 52 +++++++ include/asm-ppc64/eeh.h | 46 ++---- 5 files changed, 237 insertions(+), 158 deletions(-) create mode 100644 arch/powerpc/platforms/pseries/eeh_event.c create mode 100644 include/asm-powerpc/eeh_event.h diff --git a/arch/powerpc/platforms/pseries/Makefile b/arch/powerpc/platforms/pseries/Makefile index dbdffb2fe429..27515476ad6c 100644 --- a/arch/powerpc/platforms/pseries/Makefile +++ b/arch/powerpc/platforms/pseries/Makefile @@ -3,4 +3,4 @@ obj-y := pci.o lpar.o hvCall.o nvram.o reconfig.o \ obj-$(CONFIG_SMP) += smp.o obj-$(CONFIG_IBMVIO) += vio.o obj-$(CONFIG_XICS) += xics.o -obj-$(CONFIG_EEH) += eeh.o +obj-$(CONFIG_EEH) += eeh.o eeh_event.o diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c index 9df1d5018363..1fec99d53311 100644 --- a/arch/powerpc/platforms/pseries/eeh.c +++ b/arch/powerpc/platforms/pseries/eeh.c @@ -19,7 +19,6 @@ #include #include -#include #include #include #include @@ -27,12 +26,12 @@ #include #include #include +#include #include #include -#include -#include -#include #include +#include +#include #undef DEBUG @@ -70,14 +69,6 @@ * and sent out for processing. */ -/* EEH event workqueue setup. */ -static DEFINE_SPINLOCK(eeh_eventlist_lock); -LIST_HEAD(eeh_eventlist); -static void eeh_event_handler(void *); -DECLARE_WORK(eeh_event_wq, eeh_event_handler, NULL); - -static struct notifier_block *eeh_notifier_chain; - /* If a device driver keeps reading an MMIO register in an interrupt * handler after a slot isolation event has occurred, we assume it * is broken and panic. This sets the threshold for how many read @@ -420,24 +411,6 @@ void eeh_slot_error_detail (struct pci_dn *pdn, int severity) spin_unlock_irqrestore(&slot_errbuf_lock, flags); } -/** - * eeh_register_notifier - Register to find out about EEH events. - * @nb: notifier block to callback on events - */ -int eeh_register_notifier(struct notifier_block *nb) -{ - return notifier_chain_register(&eeh_notifier_chain, nb); -} - -/** - * eeh_unregister_notifier - Unregister to an EEH event notifier. - * @nb: notifier block to callback on events - */ -int eeh_unregister_notifier(struct notifier_block *nb) -{ - return notifier_chain_unregister(&eeh_notifier_chain, nb); -} - /** * read_slot_reset_state - Read the reset state of a device node's slot * @dn: device node to read @@ -460,73 +433,6 @@ static int read_slot_reset_state(struct pci_dn *pdn, int rets[]) BUID_HI(pdn->phb->buid), BUID_LO(pdn->phb->buid)); } -/** - * eeh_panic - call panic() for an eeh event that cannot be handled. - * The philosophy of this routine is that it is better to panic and - * halt the OS than it is to risk possible data corruption by - * oblivious device drivers that don't know better. - * - * @dev pci device that had an eeh event - * @reset_state current reset state of the device slot - */ -static void eeh_panic(struct pci_dev *dev, int reset_state) -{ - /* - * XXX We should create a separate sysctl for this. - * - * Since the panic_on_oops sysctl is used to halt the system - * in light of potential corruption, we can use it here. - */ - if (panic_on_oops) { - struct device_node *dn = pci_device_to_OF_node(dev); - eeh_slot_error_detail (PCI_DN(dn), 2 /* Permanent Error */); - panic("EEH: MMIO failure (%d) on device:%s\n", reset_state, - pci_name(dev)); - } - else { - __get_cpu_var(ignored_failures)++; - printk(KERN_INFO "EEH: Ignored MMIO failure (%d) on device:%s\n", - reset_state, pci_name(dev)); - } -} - -/** - * eeh_event_handler - dispatch EEH events. The detection of a frozen - * slot can occur inside an interrupt, where it can be hard to do - * anything about it. The goal of this routine is to pull these - * detection events out of the context of the interrupt handler, and - * re-dispatch them for processing at a later time in a normal context. - * - * @dummy - unused - */ -static void eeh_event_handler(void *dummy) -{ - unsigned long flags; - struct eeh_event *event; - - while (1) { - spin_lock_irqsave(&eeh_eventlist_lock, flags); - event = NULL; - if (!list_empty(&eeh_eventlist)) { - event = list_entry(eeh_eventlist.next, struct eeh_event, list); - list_del(&event->list); - } - spin_unlock_irqrestore(&eeh_eventlist_lock, flags); - if (event == NULL) - break; - - printk(KERN_INFO "EEH: MMIO failure (%d), notifiying device " - "%s\n", event->reset_state, - pci_name(event->dev)); - - notifier_call_chain (&eeh_notifier_chain, - EEH_NOTIFY_FREEZE, event); - - pci_dev_put(event->dev); - kfree(event); - } -} - /** * eeh_token_to_phys - convert EEH address token to phys address * @token i/o token, should be address in the form 0xA.... @@ -613,8 +519,6 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) int ret; int rets[3]; unsigned long flags; - int reset_state; - struct eeh_event *event; struct pci_dn *pdn; struct device_node *pe_dn; int rc = 0; @@ -722,33 +626,12 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) __eeh_mark_slot (pe_dn); spin_unlock_irqrestore(&confirm_error_lock, flags); - reset_state = rets[0]; - - eeh_slot_error_detail (pdn, 1 /* Temporary Error */); - - printk(KERN_INFO "EEH: MMIO failure (%d) on device: %s %s\n", - rets[0], dn->name, dn->full_name); - event = kmalloc(sizeof(*event), GFP_ATOMIC); - if (event == NULL) { - eeh_panic(dev, reset_state); - return 1; - } - - event->dev = dev; - event->dn = dn; - event->reset_state = reset_state; - - /* We may or may not be called in an interrupt context */ - spin_lock_irqsave(&eeh_eventlist_lock, flags); - list_add(&event->list, &eeh_eventlist); - spin_unlock_irqrestore(&eeh_eventlist_lock, flags); - + eeh_send_failure_event (dn, dev, rets[0], rets[2]); + /* Most EEH events are due to device driver bugs. Having * a stack trace will help the device-driver authors figure * out what happened. So print that out. */ if (rets[0] != 5) dump_stack(); - schedule_work(&eeh_event_wq); - return 1; dn_unlock: @@ -793,6 +676,14 @@ unsigned long eeh_check_failure(const volatile void __iomem *token, unsigned lon EXPORT_SYMBOL(eeh_check_failure); +/* ------------------------------------------------------------- */ +/* The code below deals with enabling EEH for devices during the + * early boot sequence. EEH must be enabled before any PCI probing + * can be done. + */ + +#define EEH_ENABLE 1 + struct eeh_early_enable_info { unsigned int buid_hi; unsigned int buid_lo; @@ -850,8 +741,9 @@ static void *early_enable_eeh(struct device_node *dn, void *data) /* First register entry is addr (00BBSS00) */ /* Try to enable eeh */ ret = rtas_call(ibm_set_eeh_option, 4, 1, NULL, - regs[0], info->buid_hi, info->buid_lo, - EEH_ENABLE); + regs[0], info->buid_hi, info->buid_lo, + EEH_ENABLE); + if (ret == 0) { eeh_subsystem_enabled = 1; pdn->eeh_mode |= EEH_MODE_SUPPORTED; diff --git a/arch/powerpc/platforms/pseries/eeh_event.c b/arch/powerpc/platforms/pseries/eeh_event.c new file mode 100644 index 000000000000..92497333c2b6 --- /dev/null +++ b/arch/powerpc/platforms/pseries/eeh_event.c @@ -0,0 +1,155 @@ +/* + * eeh_event.c + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Copyright (c) 2005 Linas Vepstas + */ + +#include +#include +#include + +/** Overview: + * EEH error states may be detected within exception handlers; + * however, the recovery processing needs to occur asynchronously + * in a normal kernel context and not an interrupt context. + * This pair of routines creates an event and queues it onto a + * work-queue, where a worker thread can drive recovery. + */ + +/* EEH event workqueue setup. */ +static spinlock_t eeh_eventlist_lock = SPIN_LOCK_UNLOCKED; +LIST_HEAD(eeh_eventlist); +static void eeh_thread_launcher(void *); +DECLARE_WORK(eeh_event_wq, eeh_thread_launcher, NULL); + +/** + * eeh_panic - call panic() for an eeh event that cannot be handled. + * The philosophy of this routine is that it is better to panic and + * halt the OS than it is to risk possible data corruption by + * oblivious device drivers that don't know better. + * + * @dev pci device that had an eeh event + * @reset_state current reset state of the device slot + */ +static void eeh_panic(struct pci_dev *dev, int reset_state) +{ + /* + * Since the panic_on_oops sysctl is used to halt the system + * in light of potential corruption, we can use it here. + */ + if (panic_on_oops) { + panic("EEH: MMIO failure (%d) on device:%s\n", reset_state, + pci_name(dev)); + } + else { + printk(KERN_INFO "EEH: Ignored MMIO failure (%d) on device:%s\n", + reset_state, pci_name(dev)); + } +} + +/** + * eeh_event_handler - dispatch EEH events. The detection of a frozen + * slot can occur inside an interrupt, where it can be hard to do + * anything about it. The goal of this routine is to pull these + * detection events out of the context of the interrupt handler, and + * re-dispatch them for processing at a later time in a normal context. + * + * @dummy - unused + */ +static int eeh_event_handler(void * dummy) +{ + unsigned long flags; + struct eeh_event *event; + + daemonize ("eehd"); + + while (1) { + set_current_state(TASK_INTERRUPTIBLE); + + spin_lock_irqsave(&eeh_eventlist_lock, flags); + event = NULL; + if (!list_empty(&eeh_eventlist)) { + event = list_entry(eeh_eventlist.next, struct eeh_event, list); + list_del(&event->list); + } + spin_unlock_irqrestore(&eeh_eventlist_lock, flags); + if (event == NULL) + break; + + printk(KERN_INFO "EEH: Detected PCI bus error on device %s\n", + pci_name(event->dev)); + + eeh_panic (event->dev, event->state); + + kfree(event); + } + + return 0; +} + +/** + * eeh_thread_launcher + * + * @dummy - unused + */ +static void eeh_thread_launcher(void *dummy) +{ + if (kernel_thread(eeh_event_handler, NULL, CLONE_KERNEL) < 0) + printk(KERN_ERR "Failed to start EEH daemon\n"); +} + +/** + * eeh_send_failure_event - generate a PCI error event + * @dev pci device + * + * This routine can be called within an interrupt context; + * the actual event will be delivered in a normal context + * (from a workqueue). + */ +int eeh_send_failure_event (struct device_node *dn, + struct pci_dev *dev, + int state, + int time_unavail) +{ + unsigned long flags; + struct eeh_event *event; + + event = kmalloc(sizeof(*event), GFP_ATOMIC); + if (event == NULL) { + printk (KERN_ERR "EEH: out of memory, event not handled\n"); + return 1; + } + + if (dev) + pci_dev_get(dev); + + event->dn = dn; + event->dev = dev; + event->state = state; + event->time_unavail = time_unavail; + + /* We may or may not be called in an interrupt context */ + spin_lock_irqsave(&eeh_eventlist_lock, flags); + list_add(&event->list, &eeh_eventlist); + spin_unlock_irqrestore(&eeh_eventlist_lock, flags); + + schedule_work(&eeh_event_wq); + + return 0; +} + +/********************** END OF FILE ******************************/ diff --git a/include/asm-powerpc/eeh_event.h b/include/asm-powerpc/eeh_event.h new file mode 100644 index 000000000000..d168a30b3866 --- /dev/null +++ b/include/asm-powerpc/eeh_event.h @@ -0,0 +1,52 @@ +/* + * eeh_event.h + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Copyright (c) 2005 Linas Vepstas + */ + +#ifndef ASM_PPC64_EEH_EVENT_H +#define ASM_PPC64_EEH_EVENT_H + +/** EEH event -- structure holding pci controller data that describes + * a change in the isolation status of a PCI slot. A pointer + * to this struct is passed as the data pointer in a notify callback. + */ +struct eeh_event { + struct list_head list; + struct device_node *dn; /* struct device node */ + struct pci_dev *dev; /* affected device */ + int state; + int time_unavail; /* milliseconds until device might be available */ +}; + +/** + * eeh_send_failure_event - generate a PCI error event + * @dev pci device + * + * This routine builds a PCI error event which will be delivered + * to all listeners on the peh_notifier_chain. + * + * This routine can be called within an interrupt context; + * the actual event will be delivered in a normal context + * (from a workqueue). + */ +int eeh_send_failure_event (struct device_node *dn, + struct pci_dev *dev, + int reset_state, + int time_unavail); + +#endif /* ASM_PPC64_EEH_EVENT_H */ diff --git a/include/asm-ppc64/eeh.h b/include/asm-ppc64/eeh.h index 40c8eb57493e..89f26ab31908 100644 --- a/include/asm-ppc64/eeh.h +++ b/include/asm-ppc64/eeh.h @@ -1,4 +1,4 @@ -/* +/* * eeh.h * Copyright (C) 2001 Dave Engebretsen & Todd Inglett IBM Corporation. * @@ -6,12 +6,12 @@ * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. - * + * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. - * + * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA @@ -27,8 +27,6 @@ struct pci_dev; struct device_node; -struct device_node; -struct notifier_block; #ifdef CONFIG_EEH @@ -37,6 +35,10 @@ struct notifier_block; #define EEH_MODE_NOCHECK (1<<1) #define EEH_MODE_ISOLATED (1<<2) +/* Max number of EEH freezes allowed before we consider the device + * to be permanently disabled. */ +#define EEH_MAX_ALLOWED_FREEZES 5 + void __init eeh_init(void); unsigned long eeh_check_failure(const volatile void __iomem *token, unsigned long val); @@ -59,36 +61,14 @@ void eeh_add_device_late(struct pci_dev *); * eeh_remove_device - undo EEH setup for the indicated pci device * @dev: pci device to be removed * - * This routine should be when a device is removed from a running - * system (e.g. by hotplug or dlpar). + * This routine should be called when a device is removed from + * a running system (e.g. by hotplug or dlpar). It unregisters + * the PCI device from the EEH subsystem. I/O errors affecting + * this device will no longer be detected after this call; thus, + * i/o errors affecting this slot may leave this device unusable. */ void eeh_remove_device(struct pci_dev *); -#define EEH_DISABLE 0 -#define EEH_ENABLE 1 -#define EEH_RELEASE_LOADSTORE 2 -#define EEH_RELEASE_DMA 3 - -/** - * Notifier event flags. - */ -#define EEH_NOTIFY_FREEZE 1 - -/** EEH event -- structure holding pci slot data that describes - * a change in the isolation status of a PCI slot. A pointer - * to this struct is passed as the data pointer in a notify callback. - */ -struct eeh_event { - struct list_head list; - struct pci_dev *dev; - struct device_node *dn; - int reset_state; -}; - -/** Register to find out about EEH events. */ -int eeh_register_notifier(struct notifier_block *nb); -int eeh_unregister_notifier(struct notifier_block *nb); - /** * EEH_POSSIBLE_ERROR() -- test for possible MMIO failure. * @@ -129,7 +109,7 @@ static inline void eeh_remove_device(struct pci_dev *dev) { } #define EEH_IO_ERROR_VALUE(size) (-1UL) #endif /* CONFIG_EEH */ -/* +/* * MMIO read/write operations with EEH support. */ static inline u8 eeh_readb(const volatile void __iomem *addr) From 6dee3fb94004c43ce09f6bf5e7c0b778ec5b8cc8 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:50:10 -0600 Subject: [PATCH 053/121] [PATCH] ppc64: PCI reset support routines 13-eeh-recovery-support-routines.patch EEH Recovery support routines This patch adds routines required to help drive the recovery of EEH-frozen slots. The main function is to drive the PCI #RST signal line high for a qurter of a second, and then allow for a second & a half of settle time. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/pseries/eeh.c | 99 ++++++++++++++++++++++++++++ include/asm-powerpc/ppc-pci.h | 14 ++++ 2 files changed, 113 insertions(+) diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c index 1fec99d53311..a06c91a4852b 100644 --- a/arch/powerpc/platforms/pseries/eeh.c +++ b/arch/powerpc/platforms/pseries/eeh.c @@ -17,6 +17,7 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ +#include #include #include #include @@ -676,6 +677,104 @@ unsigned long eeh_check_failure(const volatile void __iomem *token, unsigned lon EXPORT_SYMBOL(eeh_check_failure); +/* ------------------------------------------------------------- */ +/* The code below deals with error recovery */ + +/** Return negative value if a permanent error, else return + * a number of milliseconds to wait until the PCI slot is + * ready to be used. + */ +static int +eeh_slot_availability(struct pci_dn *pdn) +{ + int rc; + int rets[3]; + + rc = read_slot_reset_state(pdn, rets); + + if (rc) return rc; + + if (rets[1] == 0) return -1; /* EEH is not supported */ + if (rets[0] == 0) return 0; /* Oll Korrect */ + if (rets[0] == 5) { + if (rets[2] == 0) return -1; /* permanently unavailable */ + return rets[2]; /* number of millisecs to wait */ + } + return -1; +} + +/** rtas_pci_slot_reset raises/lowers the pci #RST line + * state: 1/0 to raise/lower the #RST + * + * Clear the EEH-frozen condition on a slot. This routine + * asserts the PCI #RST line if the 'state' argument is '1', + * and drops the #RST line if 'state is '0'. This routine is + * safe to call in an interrupt context. + * + */ + +static void +rtas_pci_slot_reset(struct pci_dn *pdn, int state) +{ + int rc; + + BUG_ON (pdn==NULL); + + if (!pdn->phb) { + printk (KERN_WARNING "EEH: in slot reset, device node %s has no phb\n", + pdn->node->full_name); + return; + } + + rc = rtas_call(ibm_set_slot_reset,4,1, NULL, + pdn->eeh_config_addr, + BUID_HI(pdn->phb->buid), + BUID_LO(pdn->phb->buid), + state); + if (rc) { + printk (KERN_WARNING "EEH: Unable to reset the failed slot, (%d) #RST=%d dn=%s\n", + rc, state, pdn->node->full_name); + return; + } + + if (state == 0) + eeh_clear_slot (pdn->node->parent->child); +} + +/** rtas_set_slot_reset -- assert the pci #RST line for 1/4 second + * dn -- device node to be reset. + */ + +void +rtas_set_slot_reset(struct pci_dn *pdn) +{ + int i, rc; + + rtas_pci_slot_reset (pdn, 1); + + /* The PCI bus requires that the reset be held high for at least + * a 100 milliseconds. We wait a bit longer 'just in case'. */ + +#define PCI_BUS_RST_HOLD_TIME_MSEC 250 + msleep (PCI_BUS_RST_HOLD_TIME_MSEC); + rtas_pci_slot_reset (pdn, 0); + + /* After a PCI slot has been reset, the PCI Express spec requires + * a 1.5 second idle time for the bus to stabilize, before starting + * up traffic. */ +#define PCI_BUS_SETTLE_TIME_MSEC 1800 + msleep (PCI_BUS_SETTLE_TIME_MSEC); + + /* Now double check with the firmware to make sure the device is + * ready to be used; if not, wait for recovery. */ + for (i=0; i<10; i++) { + rc = eeh_slot_availability (pdn); + if (rc <= 0) break; + + msleep (rc+100); + } +} + /* ------------------------------------------------------------- */ /* The code below deals with enabling EEH for devices during the * early boot sequence. EEH must be enabled before any PCI probing diff --git a/include/asm-powerpc/ppc-pci.h b/include/asm-powerpc/ppc-pci.h index c47146f74a7b..930a606b4e35 100644 --- a/include/asm-powerpc/ppc-pci.h +++ b/include/asm-powerpc/ppc-pci.h @@ -52,4 +52,18 @@ extern unsigned long pci_probe_only; extern unsigned long pci_assign_all_buses; extern int pci_read_irq_line(struct pci_dev *pci_dev); +/* ---- EEH internal-use-only related routines ---- */ +#ifdef CONFIG_EEH +/** + * rtas_set_slot_reset -- unfreeze a frozen slot + * + * Clear the EEH-frozen condition on a slot. This routine + * does this by asserting the PCI #RST line for 1/8th of + * a second; this routine will sleep while the adapter is + * being reset. + */ +void rtas_set_slot_reset (struct pci_dn *); + +#endif + #endif /* _ASM_POWERPC_PPC_PCI_H */ From 8b553f32db3bf5d0ec0819c595932eb21cd45945 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:50:17 -0600 Subject: [PATCH 054/121] [PATCH] ppc64: Save & restore of PCI device BARS 14-eeh-device-bar-save.patch After a PCI device has been resest, the device BAR's and other config space info must be restored to the same state as they were in when the firmware first handed us this device. This will allow the PCI device driver, when restarted, to correctly recognize and set up the device. Tis patch saves the device config space as early as reasonable after the firmware has handed over the device. Te state resore funcion is inteded for use by the EEH recovery routines. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/pseries/eeh.c | 115 ++++++++++++++++++++++++++- include/asm-powerpc/ppc-pci.h | 23 ++++++ 2 files changed, 137 insertions(+), 1 deletion(-) diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c index a06c91a4852b..b760836bb9d1 100644 --- a/arch/powerpc/platforms/pseries/eeh.c +++ b/arch/powerpc/platforms/pseries/eeh.c @@ -77,6 +77,9 @@ */ #define EEH_MAX_FAILS 100000 +/* Misc forward declaraions */ +static void eeh_save_bars(struct pci_dev * pdev, struct pci_dn *pdn); + /* RTAS tokens */ static int ibm_set_eeh_option; static int ibm_set_slot_reset; @@ -366,6 +369,7 @@ static void pci_addr_cache_remove_device(struct pci_dev *dev) */ void __init pci_addr_cache_build(void) { + struct device_node *dn; struct pci_dev *dev = NULL; if (!eeh_subsystem_enabled) @@ -379,6 +383,10 @@ void __init pci_addr_cache_build(void) continue; } pci_addr_cache_insert_device(dev); + + /* Save the BAR's; firmware doesn't restore these after EEH reset */ + dn = pci_device_to_OF_node(dev); + eeh_save_bars(dev, PCI_DN(dn)); } #ifdef DEBUG @@ -775,6 +783,108 @@ rtas_set_slot_reset(struct pci_dn *pdn) } } +/* ------------------------------------------------------- */ +/** Save and restore of PCI BARs + * + * Although firmware will set up BARs during boot, it doesn't + * set up device BAR's after a device reset, although it will, + * if requested, set up bridge configuration. Thus, we need to + * configure the PCI devices ourselves. + */ + +/** + * __restore_bars - Restore the Base Address Registers + * Loads the PCI configuration space base address registers, + * the expansion ROM base address, the latency timer, and etc. + * from the saved values in the device node. + */ +static inline void __restore_bars (struct pci_dn *pdn) +{ + int i; + + if (NULL==pdn->phb) return; + for (i=4; i<10; i++) { + rtas_write_config(pdn, i*4, 4, pdn->config_space[i]); + } + + /* 12 == Expansion ROM Address */ + rtas_write_config(pdn, 12*4, 4, pdn->config_space[12]); + +#define BYTE_SWAP(OFF) (8*((OFF)/4)+3-(OFF)) +#define SAVED_BYTE(OFF) (((u8 *)(pdn->config_space))[BYTE_SWAP(OFF)]) + + rtas_write_config (pdn, PCI_CACHE_LINE_SIZE, 1, + SAVED_BYTE(PCI_CACHE_LINE_SIZE)); + + rtas_write_config (pdn, PCI_LATENCY_TIMER, 1, + SAVED_BYTE(PCI_LATENCY_TIMER)); + + /* max latency, min grant, interrupt pin and line */ + rtas_write_config(pdn, 15*4, 4, pdn->config_space[15]); +} + +/** + * eeh_restore_bars - restore the PCI config space info + * + * This routine performs a recursive walk to the children + * of this device as well. + */ +void eeh_restore_bars(struct pci_dn *pdn) +{ + struct device_node *dn; + if (!pdn) + return; + + if (! pdn->eeh_is_bridge) + __restore_bars (pdn); + + dn = pdn->node->child; + while (dn) { + eeh_restore_bars (PCI_DN(dn)); + dn = dn->sibling; + } +} + +/** + * eeh_save_bars - save device bars + * + * Save the values of the device bars. Unlike the restore + * routine, this routine is *not* recursive. This is because + * PCI devices are added individuallly; but, for the restore, + * an entire slot is reset at a time. + */ +static void eeh_save_bars(struct pci_dev * pdev, struct pci_dn *pdn) +{ + int i; + + if (!pdev || !pdn ) + return; + + for (i = 0; i < 16; i++) + pci_read_config_dword(pdev, i * 4, &pdn->config_space[i]); + + if (pdev->hdr_type == PCI_HEADER_TYPE_BRIDGE) + pdn->eeh_is_bridge = 1; +} + +void +rtas_configure_bridge(struct pci_dn *pdn) +{ + int token = rtas_token ("ibm,configure-bridge"); + int rc; + + if (token == RTAS_UNKNOWN_SERVICE) + return; + rc = rtas_call(token,3,1, NULL, + pdn->eeh_config_addr, + BUID_HI(pdn->phb->buid), + BUID_LO(pdn->phb->buid)); + if (rc) { + printk (KERN_WARNING "EEH: Unable to configure device bridge (%d) for %s\n", + rc, pdn->node->full_name); + } +} + /* ------------------------------------------------------------- */ /* The code below deals with enabling EEH for devices during the * early boot sequence. EEH must be enabled before any PCI probing @@ -977,6 +1087,7 @@ EXPORT_SYMBOL_GPL(eeh_add_device_early); void eeh_add_device_late(struct pci_dev *dev) { struct device_node *dn; + struct pci_dn *pdn; if (!dev || !eeh_subsystem_enabled) return; @@ -987,9 +1098,11 @@ void eeh_add_device_late(struct pci_dev *dev) pci_dev_get (dev); dn = pci_device_to_OF_node(dev); - PCI_DN(dn)->pcidev = dev; + pdn = PCI_DN(dn); + pdn->pcidev = dev; pci_addr_cache_insert_device (dev); + eeh_save_bars(dev, pdn); } EXPORT_SYMBOL_GPL(eeh_add_device_late); diff --git a/include/asm-powerpc/ppc-pci.h b/include/asm-powerpc/ppc-pci.h index 930a606b4e35..d86c47872bea 100644 --- a/include/asm-powerpc/ppc-pci.h +++ b/include/asm-powerpc/ppc-pci.h @@ -64,6 +64,29 @@ extern int pci_read_irq_line(struct pci_dev *pci_dev); */ void rtas_set_slot_reset (struct pci_dn *); +/** + * eeh_restore_bars - Restore device configuration info. + * + * A reset of a PCI device will clear out its config space. + * This routines will restore the config space for this + * device, and is children, to values previously obtained + * from the firmware. + */ +void eeh_restore_bars(struct pci_dn *); + +/** + * rtas_configure_bridge -- firmware initialization of pci bridge + * + * Ask the firmware to configure all PCI bridges devices + * located behind the indicated node. Required after a + * pci device reset. Does essentially the same hing as + * eeh_restore_bars, but for brdges, and lets firmware + * do the work. + */ +void rtas_configure_bridge(struct pci_dn *); + +int rtas_write_config(struct pci_dn *, int where, int size, u32 val); + #endif #endif /* _ASM_POWERPC_PPC_PCI_H */ From 9fb9cbb1082d6b31fb45aa1a14432449a0df6cf1 Mon Sep 17 00:00:00 2001 From: Yasuyuki Kozakai Date: Wed, 9 Nov 2005 16:38:16 -0800 Subject: [PATCH 055/121] [NETFILTER]: Add nf_conntrack subsystem. The existing connection tracking subsystem in netfilter can only handle ipv4. There were basically two choices present to add connection tracking support for ipv6. We could either duplicate all of the ipv4 connection tracking code into an ipv6 counterpart, or (the choice taken by these patches) we could design a generic layer that could handle both ipv4 and ipv6 and thus requiring only one sub-protocol (TCP, UDP, etc.) connection tracking helper module to be written. In fact nf_conntrack is capable of working with any layer 3 protocol. The existing ipv4 specific conntrack code could also not deal with the pecularities of doing connection tracking on ipv6, which is also cured here. For example, these issues include: 1) ICMPv6 handling, which is used for neighbour discovery in ipv6 thus some messages such as these should not participate in connection tracking since effectively they are like ARP messages 2) fragmentation must be handled differently in ipv6, because the simplistic "defrag, connection track and NAT, refrag" (which the existing ipv4 connection tracking does) approach simply isn't feasible in ipv6 3) ipv6 extension header parsing must occur at the correct spots before and after connection tracking decisions, and there were no provisions for this in the existing connection tracking design 4) ipv6 has no need for stateful NAT The ipv4 specific conntrack layer is kept around, until all of the ipv4 specific conntrack helpers are ported over to nf_conntrack and it is feature complete. Once that occurs, the old conntrack stuff will get placed into the feature-removal-schedule and we will fully kill it off 6 months later. Signed-off-by: Yasuyuki Kozakai Signed-off-by: Harald Welte Signed-off-by: Arnaldo Carvalho de Melo --- include/linux/netfilter/nf_conntrack_common.h | 159 ++ include/linux/netfilter/nf_conntrack_ftp.h | 44 + include/linux/netfilter/nf_conntrack_sctp.h | 27 + include/linux/netfilter/nf_conntrack_tcp.h | 56 + .../netfilter/nf_conntrack_tuple_common.h | 13 + include/linux/netfilter_ipv4/ip_conntrack.h | 152 +- .../linux/netfilter_ipv4/ip_conntrack_ftp.h | 39 +- .../linux/netfilter_ipv4/ip_conntrack_icmp.h | 9 +- .../linux/netfilter_ipv4/ip_conntrack_sctp.h | 21 +- .../linux/netfilter_ipv4/ip_conntrack_tcp.h | 47 +- .../linux/netfilter_ipv4/ip_conntrack_tuple.h | 10 +- include/linux/netfilter_ipv6.h | 1 + include/linux/skbuff.h | 19 + include/linux/sysctl.h | 37 + .../net/netfilter/ipv4/nf_conntrack_icmp.h | 11 + .../net/netfilter/ipv4/nf_conntrack_ipv4.h | 43 + .../net/netfilter/ipv6/nf_conntrack_icmpv6.h | 27 + include/net/netfilter/nf_conntrack.h | 354 ++++ include/net/netfilter/nf_conntrack_compat.h | 108 ++ include/net/netfilter/nf_conntrack_core.h | 76 + include/net/netfilter/nf_conntrack_helper.h | 51 + include/net/netfilter/nf_conntrack_l3proto.h | 93 + include/net/netfilter/nf_conntrack_protocol.h | 105 ++ include/net/netfilter/nf_conntrack_tuple.h | 190 ++ net/core/skbuff.c | 15 + net/ipv4/netfilter/Kconfig | 41 +- net/ipv4/netfilter/Makefile | 6 + net/ipv4/netfilter/ip_conntrack_netlink.c | 2 +- net/ipv4/netfilter/ipt_CLUSTERIP.c | 12 +- net/ipv4/netfilter/ipt_CONNMARK.c | 22 +- net/ipv4/netfilter/ipt_NOTRACK.c | 4 +- net/ipv4/netfilter/ipt_connbytes.c | 39 +- net/ipv4/netfilter/ipt_connmark.c | 10 +- net/ipv4/netfilter/ipt_conntrack.c | 96 + net/ipv4/netfilter/ipt_helper.c | 54 + net/ipv4/netfilter/ipt_state.c | 6 +- .../netfilter/nf_conntrack_l3proto_ipv4.c | 571 ++++++ net/ipv4/netfilter/nf_conntrack_proto_icmp.c | 301 ++++ net/ipv6/ip6_input.c | 5 + net/ipv6/ip6_output.c | 6 + net/ipv6/netfilter/Kconfig | 14 + net/ipv6/netfilter/Makefile | 6 + net/ipv6/netfilter/ip6t_MARK.c | 6 +- .../netfilter/nf_conntrack_l3proto_ipv6.c | 556 ++++++ .../netfilter/nf_conntrack_proto_icmpv6.c | 272 +++ net/ipv6/netfilter/nf_conntrack_reasm.c | 885 ++++++++++ net/ipv6/raw.c | 4 +- net/netfilter/Kconfig | 74 + net/netfilter/Makefile | 8 + net/netfilter/nf_conntrack_core.c | 1538 +++++++++++++++++ net/netfilter/nf_conntrack_ftp.c | 698 ++++++++ net/netfilter/nf_conntrack_l3proto_generic.c | 98 ++ net/netfilter/nf_conntrack_proto_generic.c | 85 + net/netfilter/nf_conntrack_proto_sctp.c | 670 +++++++ net/netfilter/nf_conntrack_proto_tcp.c | 1162 +++++++++++++ net/netfilter/nf_conntrack_proto_udp.c | 216 +++ net/netfilter/nf_conntrack_standalone.c | 869 ++++++++++ 57 files changed, 9710 insertions(+), 333 deletions(-) create mode 100644 include/linux/netfilter/nf_conntrack_common.h create mode 100644 include/linux/netfilter/nf_conntrack_ftp.h create mode 100644 include/linux/netfilter/nf_conntrack_sctp.h create mode 100644 include/linux/netfilter/nf_conntrack_tcp.h create mode 100644 include/linux/netfilter/nf_conntrack_tuple_common.h create mode 100644 include/net/netfilter/ipv4/nf_conntrack_icmp.h create mode 100644 include/net/netfilter/ipv4/nf_conntrack_ipv4.h create mode 100644 include/net/netfilter/ipv6/nf_conntrack_icmpv6.h create mode 100644 include/net/netfilter/nf_conntrack.h create mode 100644 include/net/netfilter/nf_conntrack_compat.h create mode 100644 include/net/netfilter/nf_conntrack_core.h create mode 100644 include/net/netfilter/nf_conntrack_helper.h create mode 100644 include/net/netfilter/nf_conntrack_l3proto.h create mode 100644 include/net/netfilter/nf_conntrack_protocol.h create mode 100644 include/net/netfilter/nf_conntrack_tuple.h create mode 100644 net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c create mode 100644 net/ipv4/netfilter/nf_conntrack_proto_icmp.c create mode 100644 net/ipv6/netfilter/nf_conntrack_l3proto_ipv6.c create mode 100644 net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c create mode 100644 net/ipv6/netfilter/nf_conntrack_reasm.c create mode 100644 net/netfilter/nf_conntrack_core.c create mode 100644 net/netfilter/nf_conntrack_ftp.c create mode 100644 net/netfilter/nf_conntrack_l3proto_generic.c create mode 100644 net/netfilter/nf_conntrack_proto_generic.c create mode 100644 net/netfilter/nf_conntrack_proto_sctp.c create mode 100644 net/netfilter/nf_conntrack_proto_tcp.c create mode 100644 net/netfilter/nf_conntrack_proto_udp.c create mode 100644 net/netfilter/nf_conntrack_standalone.c diff --git a/include/linux/netfilter/nf_conntrack_common.h b/include/linux/netfilter/nf_conntrack_common.h new file mode 100644 index 000000000000..6d39b518486b --- /dev/null +++ b/include/linux/netfilter/nf_conntrack_common.h @@ -0,0 +1,159 @@ +#ifndef _NF_CONNTRACK_COMMON_H +#define _NF_CONNTRACK_COMMON_H +/* Connection state tracking for netfilter. This is separated from, + but required by, the NAT layer; it can also be used by an iptables + extension. */ +enum ip_conntrack_info +{ + /* Part of an established connection (either direction). */ + IP_CT_ESTABLISHED, + + /* Like NEW, but related to an existing connection, or ICMP error + (in either direction). */ + IP_CT_RELATED, + + /* Started a new connection to track (only + IP_CT_DIR_ORIGINAL); may be a retransmission. */ + IP_CT_NEW, + + /* >= this indicates reply direction */ + IP_CT_IS_REPLY, + + /* Number of distinct IP_CT types (no NEW in reply dirn). */ + IP_CT_NUMBER = IP_CT_IS_REPLY * 2 - 1 +}; + +/* Bitset representing status of connection. */ +enum ip_conntrack_status { + /* It's an expected connection: bit 0 set. This bit never changed */ + IPS_EXPECTED_BIT = 0, + IPS_EXPECTED = (1 << IPS_EXPECTED_BIT), + + /* We've seen packets both ways: bit 1 set. Can be set, not unset. */ + IPS_SEEN_REPLY_BIT = 1, + IPS_SEEN_REPLY = (1 << IPS_SEEN_REPLY_BIT), + + /* Conntrack should never be early-expired. */ + IPS_ASSURED_BIT = 2, + IPS_ASSURED = (1 << IPS_ASSURED_BIT), + + /* Connection is confirmed: originating packet has left box */ + IPS_CONFIRMED_BIT = 3, + IPS_CONFIRMED = (1 << IPS_CONFIRMED_BIT), + + /* Connection needs src nat in orig dir. This bit never changed. */ + IPS_SRC_NAT_BIT = 4, + IPS_SRC_NAT = (1 << IPS_SRC_NAT_BIT), + + /* Connection needs dst nat in orig dir. This bit never changed. */ + IPS_DST_NAT_BIT = 5, + IPS_DST_NAT = (1 << IPS_DST_NAT_BIT), + + /* Both together. */ + IPS_NAT_MASK = (IPS_DST_NAT | IPS_SRC_NAT), + + /* Connection needs TCP sequence adjusted. */ + IPS_SEQ_ADJUST_BIT = 6, + IPS_SEQ_ADJUST = (1 << IPS_SEQ_ADJUST_BIT), + + /* NAT initialization bits. */ + IPS_SRC_NAT_DONE_BIT = 7, + IPS_SRC_NAT_DONE = (1 << IPS_SRC_NAT_DONE_BIT), + + IPS_DST_NAT_DONE_BIT = 8, + IPS_DST_NAT_DONE = (1 << IPS_DST_NAT_DONE_BIT), + + /* Both together */ + IPS_NAT_DONE_MASK = (IPS_DST_NAT_DONE | IPS_SRC_NAT_DONE), + + /* Connection is dying (removed from lists), can not be unset. */ + IPS_DYING_BIT = 9, + IPS_DYING = (1 << IPS_DYING_BIT), +}; + +/* Connection tracking event bits */ +enum ip_conntrack_events +{ + /* New conntrack */ + IPCT_NEW_BIT = 0, + IPCT_NEW = (1 << IPCT_NEW_BIT), + + /* Expected connection */ + IPCT_RELATED_BIT = 1, + IPCT_RELATED = (1 << IPCT_RELATED_BIT), + + /* Destroyed conntrack */ + IPCT_DESTROY_BIT = 2, + IPCT_DESTROY = (1 << IPCT_DESTROY_BIT), + + /* Timer has been refreshed */ + IPCT_REFRESH_BIT = 3, + IPCT_REFRESH = (1 << IPCT_REFRESH_BIT), + + /* Status has changed */ + IPCT_STATUS_BIT = 4, + IPCT_STATUS = (1 << IPCT_STATUS_BIT), + + /* Update of protocol info */ + IPCT_PROTOINFO_BIT = 5, + IPCT_PROTOINFO = (1 << IPCT_PROTOINFO_BIT), + + /* Volatile protocol info */ + IPCT_PROTOINFO_VOLATILE_BIT = 6, + IPCT_PROTOINFO_VOLATILE = (1 << IPCT_PROTOINFO_VOLATILE_BIT), + + /* New helper for conntrack */ + IPCT_HELPER_BIT = 7, + IPCT_HELPER = (1 << IPCT_HELPER_BIT), + + /* Update of helper info */ + IPCT_HELPINFO_BIT = 8, + IPCT_HELPINFO = (1 << IPCT_HELPINFO_BIT), + + /* Volatile helper info */ + IPCT_HELPINFO_VOLATILE_BIT = 9, + IPCT_HELPINFO_VOLATILE = (1 << IPCT_HELPINFO_VOLATILE_BIT), + + /* NAT info */ + IPCT_NATINFO_BIT = 10, + IPCT_NATINFO = (1 << IPCT_NATINFO_BIT), + + /* Counter highest bit has been set */ + IPCT_COUNTER_FILLING_BIT = 11, + IPCT_COUNTER_FILLING = (1 << IPCT_COUNTER_FILLING_BIT), +}; + +enum ip_conntrack_expect_events { + IPEXP_NEW_BIT = 0, + IPEXP_NEW = (1 << IPEXP_NEW_BIT), +}; + +#ifdef __KERNEL__ +struct ip_conntrack_counter +{ + u_int32_t packets; + u_int32_t bytes; +}; + +struct ip_conntrack_stat +{ + unsigned int searched; + unsigned int found; + unsigned int new; + unsigned int invalid; + unsigned int ignore; + unsigned int delete; + unsigned int delete_list; + unsigned int insert; + unsigned int insert_failed; + unsigned int drop; + unsigned int early_drop; + unsigned int error; + unsigned int expect_new; + unsigned int expect_create; + unsigned int expect_delete; +}; + +#endif /* __KERNEL__ */ + +#endif /* _NF_CONNTRACK_COMMON_H */ diff --git a/include/linux/netfilter/nf_conntrack_ftp.h b/include/linux/netfilter/nf_conntrack_ftp.h new file mode 100644 index 000000000000..ad4a41c9ce93 --- /dev/null +++ b/include/linux/netfilter/nf_conntrack_ftp.h @@ -0,0 +1,44 @@ +#ifndef _NF_CONNTRACK_FTP_H +#define _NF_CONNTRACK_FTP_H +/* FTP tracking. */ + +/* This enum is exposed to userspace */ +enum ip_ct_ftp_type +{ + /* PORT command from client */ + IP_CT_FTP_PORT, + /* PASV response from server */ + IP_CT_FTP_PASV, + /* EPRT command from client */ + IP_CT_FTP_EPRT, + /* EPSV response from server */ + IP_CT_FTP_EPSV, +}; + +#ifdef __KERNEL__ + +#define FTP_PORT 21 + +#define NUM_SEQ_TO_REMEMBER 2 +/* This structure exists only once per master */ +struct ip_ct_ftp_master { + /* Valid seq positions for cmd matching after newline */ + u_int32_t seq_aft_nl[IP_CT_DIR_MAX][NUM_SEQ_TO_REMEMBER]; + /* 0 means seq_match_aft_nl not set */ + int seq_aft_nl_num[IP_CT_DIR_MAX]; +}; + +struct ip_conntrack_expect; + +/* For NAT to hook in when we find a packet which describes what other + * connection we should expect. */ +extern unsigned int (*ip_nat_ftp_hook)(struct sk_buff **pskb, + enum ip_conntrack_info ctinfo, + enum ip_ct_ftp_type type, + unsigned int matchoff, + unsigned int matchlen, + struct ip_conntrack_expect *exp, + u32 *seq); +#endif /* __KERNEL__ */ + +#endif /* _NF_CONNTRACK_FTP_H */ diff --git a/include/linux/netfilter/nf_conntrack_sctp.h b/include/linux/netfilter/nf_conntrack_sctp.h new file mode 100644 index 000000000000..b8994d9fd1a9 --- /dev/null +++ b/include/linux/netfilter/nf_conntrack_sctp.h @@ -0,0 +1,27 @@ +#ifndef _NF_CONNTRACK_SCTP_H +#define _NF_CONNTRACK_SCTP_H +/* SCTP tracking. */ + +#include + +enum sctp_conntrack { + SCTP_CONNTRACK_NONE, + SCTP_CONNTRACK_CLOSED, + SCTP_CONNTRACK_COOKIE_WAIT, + SCTP_CONNTRACK_COOKIE_ECHOED, + SCTP_CONNTRACK_ESTABLISHED, + SCTP_CONNTRACK_SHUTDOWN_SENT, + SCTP_CONNTRACK_SHUTDOWN_RECD, + SCTP_CONNTRACK_SHUTDOWN_ACK_SENT, + SCTP_CONNTRACK_MAX +}; + +struct ip_ct_sctp +{ + enum sctp_conntrack state; + + u_int32_t vtag[IP_CT_DIR_MAX]; + u_int32_t ttag[IP_CT_DIR_MAX]; +}; + +#endif /* _NF_CONNTRACK_SCTP_H */ diff --git a/include/linux/netfilter/nf_conntrack_tcp.h b/include/linux/netfilter/nf_conntrack_tcp.h new file mode 100644 index 000000000000..b2feeffde384 --- /dev/null +++ b/include/linux/netfilter/nf_conntrack_tcp.h @@ -0,0 +1,56 @@ +#ifndef _NF_CONNTRACK_TCP_H +#define _NF_CONNTRACK_TCP_H +/* TCP tracking. */ + +/* This is exposed to userspace (ctnetlink) */ +enum tcp_conntrack { + TCP_CONNTRACK_NONE, + TCP_CONNTRACK_SYN_SENT, + TCP_CONNTRACK_SYN_RECV, + TCP_CONNTRACK_ESTABLISHED, + TCP_CONNTRACK_FIN_WAIT, + TCP_CONNTRACK_CLOSE_WAIT, + TCP_CONNTRACK_LAST_ACK, + TCP_CONNTRACK_TIME_WAIT, + TCP_CONNTRACK_CLOSE, + TCP_CONNTRACK_LISTEN, + TCP_CONNTRACK_MAX, + TCP_CONNTRACK_IGNORE +}; + +/* Window scaling is advertised by the sender */ +#define IP_CT_TCP_FLAG_WINDOW_SCALE 0x01 + +/* SACK is permitted by the sender */ +#define IP_CT_TCP_FLAG_SACK_PERM 0x02 + +/* This sender sent FIN first */ +#define IP_CT_TCP_FLAG_CLOSE_INIT 0x03 + +#ifdef __KERNEL__ + +struct ip_ct_tcp_state { + u_int32_t td_end; /* max of seq + len */ + u_int32_t td_maxend; /* max of ack + max(win, 1) */ + u_int32_t td_maxwin; /* max(win) */ + u_int8_t td_scale; /* window scale factor */ + u_int8_t loose; /* used when connection picked up from the middle */ + u_int8_t flags; /* per direction options */ +}; + +struct ip_ct_tcp +{ + struct ip_ct_tcp_state seen[2]; /* connection parameters per direction */ + u_int8_t state; /* state of the connection (enum tcp_conntrack) */ + /* For detecting stale connections */ + u_int8_t last_dir; /* Direction of the last packet (enum ip_conntrack_dir) */ + u_int8_t retrans; /* Number of retransmitted packets */ + u_int8_t last_index; /* Index of the last packet */ + u_int32_t last_seq; /* Last sequence number seen in dir */ + u_int32_t last_ack; /* Last sequence number seen in opposite dir */ + u_int32_t last_end; /* Last seq + len */ +}; + +#endif /* __KERNEL__ */ + +#endif /* _NF_CONNTRACK_TCP_H */ diff --git a/include/linux/netfilter/nf_conntrack_tuple_common.h b/include/linux/netfilter/nf_conntrack_tuple_common.h new file mode 100644 index 000000000000..8e145f0d61cb --- /dev/null +++ b/include/linux/netfilter/nf_conntrack_tuple_common.h @@ -0,0 +1,13 @@ +#ifndef _NF_CONNTRACK_TUPLE_COMMON_H +#define _NF_CONNTRACK_TUPLE_COMMON_H + +enum ip_conntrack_dir +{ + IP_CT_DIR_ORIGINAL, + IP_CT_DIR_REPLY, + IP_CT_DIR_MAX +}; + +#define CTINFO2DIR(ctinfo) ((ctinfo) >= IP_CT_IS_REPLY ? IP_CT_DIR_REPLY : IP_CT_DIR_ORIGINAL) + +#endif /* _NF_CONNTRACK_TUPLE_COMMON_H */ diff --git a/include/linux/netfilter_ipv4/ip_conntrack.h b/include/linux/netfilter_ipv4/ip_conntrack.h index d078bb91d9e5..b3432ab59a17 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack.h +++ b/include/linux/netfilter_ipv4/ip_conntrack.h @@ -1,132 +1,7 @@ #ifndef _IP_CONNTRACK_H #define _IP_CONNTRACK_H -/* Connection state tracking for netfilter. This is separated from, - but required by, the NAT layer; it can also be used by an iptables - extension. */ -enum ip_conntrack_info -{ - /* Part of an established connection (either direction). */ - IP_CT_ESTABLISHED, - /* Like NEW, but related to an existing connection, or ICMP error - (in either direction). */ - IP_CT_RELATED, - - /* Started a new connection to track (only - IP_CT_DIR_ORIGINAL); may be a retransmission. */ - IP_CT_NEW, - - /* >= this indicates reply direction */ - IP_CT_IS_REPLY, - - /* Number of distinct IP_CT types (no NEW in reply dirn). */ - IP_CT_NUMBER = IP_CT_IS_REPLY * 2 - 1 -}; - -/* Bitset representing status of connection. */ -enum ip_conntrack_status { - /* It's an expected connection: bit 0 set. This bit never changed */ - IPS_EXPECTED_BIT = 0, - IPS_EXPECTED = (1 << IPS_EXPECTED_BIT), - - /* We've seen packets both ways: bit 1 set. Can be set, not unset. */ - IPS_SEEN_REPLY_BIT = 1, - IPS_SEEN_REPLY = (1 << IPS_SEEN_REPLY_BIT), - - /* Conntrack should never be early-expired. */ - IPS_ASSURED_BIT = 2, - IPS_ASSURED = (1 << IPS_ASSURED_BIT), - - /* Connection is confirmed: originating packet has left box */ - IPS_CONFIRMED_BIT = 3, - IPS_CONFIRMED = (1 << IPS_CONFIRMED_BIT), - - /* Connection needs src nat in orig dir. This bit never changed. */ - IPS_SRC_NAT_BIT = 4, - IPS_SRC_NAT = (1 << IPS_SRC_NAT_BIT), - - /* Connection needs dst nat in orig dir. This bit never changed. */ - IPS_DST_NAT_BIT = 5, - IPS_DST_NAT = (1 << IPS_DST_NAT_BIT), - - /* Both together. */ - IPS_NAT_MASK = (IPS_DST_NAT | IPS_SRC_NAT), - - /* Connection needs TCP sequence adjusted. */ - IPS_SEQ_ADJUST_BIT = 6, - IPS_SEQ_ADJUST = (1 << IPS_SEQ_ADJUST_BIT), - - /* NAT initialization bits. */ - IPS_SRC_NAT_DONE_BIT = 7, - IPS_SRC_NAT_DONE = (1 << IPS_SRC_NAT_DONE_BIT), - - IPS_DST_NAT_DONE_BIT = 8, - IPS_DST_NAT_DONE = (1 << IPS_DST_NAT_DONE_BIT), - - /* Both together */ - IPS_NAT_DONE_MASK = (IPS_DST_NAT_DONE | IPS_SRC_NAT_DONE), - - /* Connection is dying (removed from lists), can not be unset. */ - IPS_DYING_BIT = 9, - IPS_DYING = (1 << IPS_DYING_BIT), -}; - -/* Connection tracking event bits */ -enum ip_conntrack_events -{ - /* New conntrack */ - IPCT_NEW_BIT = 0, - IPCT_NEW = (1 << IPCT_NEW_BIT), - - /* Expected connection */ - IPCT_RELATED_BIT = 1, - IPCT_RELATED = (1 << IPCT_RELATED_BIT), - - /* Destroyed conntrack */ - IPCT_DESTROY_BIT = 2, - IPCT_DESTROY = (1 << IPCT_DESTROY_BIT), - - /* Timer has been refreshed */ - IPCT_REFRESH_BIT = 3, - IPCT_REFRESH = (1 << IPCT_REFRESH_BIT), - - /* Status has changed */ - IPCT_STATUS_BIT = 4, - IPCT_STATUS = (1 << IPCT_STATUS_BIT), - - /* Update of protocol info */ - IPCT_PROTOINFO_BIT = 5, - IPCT_PROTOINFO = (1 << IPCT_PROTOINFO_BIT), - - /* Volatile protocol info */ - IPCT_PROTOINFO_VOLATILE_BIT = 6, - IPCT_PROTOINFO_VOLATILE = (1 << IPCT_PROTOINFO_VOLATILE_BIT), - - /* New helper for conntrack */ - IPCT_HELPER_BIT = 7, - IPCT_HELPER = (1 << IPCT_HELPER_BIT), - - /* Update of helper info */ - IPCT_HELPINFO_BIT = 8, - IPCT_HELPINFO = (1 << IPCT_HELPINFO_BIT), - - /* Volatile helper info */ - IPCT_HELPINFO_VOLATILE_BIT = 9, - IPCT_HELPINFO_VOLATILE = (1 << IPCT_HELPINFO_VOLATILE_BIT), - - /* NAT info */ - IPCT_NATINFO_BIT = 10, - IPCT_NATINFO = (1 << IPCT_NATINFO_BIT), - - /* Counter highest bit has been set */ - IPCT_COUNTER_FILLING_BIT = 11, - IPCT_COUNTER_FILLING = (1 << IPCT_COUNTER_FILLING_BIT), -}; - -enum ip_conntrack_expect_events { - IPEXP_NEW_BIT = 0, - IPEXP_NEW = (1 << IPEXP_NEW_BIT), -}; +#include #ifdef __KERNEL__ #include @@ -194,12 +69,6 @@ do { \ #define IP_NF_ASSERT(x) #endif -struct ip_conntrack_counter -{ - u_int32_t packets; - u_int32_t bytes; -}; - struct ip_conntrack_helper; struct ip_conntrack @@ -426,25 +295,6 @@ static inline int is_dying(struct ip_conntrack *ct) extern unsigned int ip_conntrack_htable_size; -struct ip_conntrack_stat -{ - unsigned int searched; - unsigned int found; - unsigned int new; - unsigned int invalid; - unsigned int ignore; - unsigned int delete; - unsigned int delete_list; - unsigned int insert; - unsigned int insert_failed; - unsigned int drop; - unsigned int early_drop; - unsigned int error; - unsigned int expect_new; - unsigned int expect_create; - unsigned int expect_delete; -}; - #define CONNTRACK_STAT_INC(count) (__get_cpu_var(ip_conntrack_stat).count++) #ifdef CONFIG_IP_NF_CONNTRACK_EVENTS diff --git a/include/linux/netfilter_ipv4/ip_conntrack_ftp.h b/include/linux/netfilter_ipv4/ip_conntrack_ftp.h index 5f06429b9047..63811934de4d 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_ftp.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_ftp.h @@ -1,43 +1,6 @@ #ifndef _IP_CONNTRACK_FTP_H #define _IP_CONNTRACK_FTP_H -/* FTP tracking. */ -#ifdef __KERNEL__ +#include -#define FTP_PORT 21 - -#endif /* __KERNEL__ */ - -enum ip_ct_ftp_type -{ - /* PORT command from client */ - IP_CT_FTP_PORT, - /* PASV response from server */ - IP_CT_FTP_PASV, - /* EPRT command from client */ - IP_CT_FTP_EPRT, - /* EPSV response from server */ - IP_CT_FTP_EPSV, -}; - -#define NUM_SEQ_TO_REMEMBER 2 -/* This structure exists only once per master */ -struct ip_ct_ftp_master { - /* Valid seq positions for cmd matching after newline */ - u_int32_t seq_aft_nl[IP_CT_DIR_MAX][NUM_SEQ_TO_REMEMBER]; - /* 0 means seq_match_aft_nl not set */ - int seq_aft_nl_num[IP_CT_DIR_MAX]; -}; - -struct ip_conntrack_expect; - -/* For NAT to hook in when we find a packet which describes what other - * connection we should expect. */ -extern unsigned int (*ip_nat_ftp_hook)(struct sk_buff **pskb, - enum ip_conntrack_info ctinfo, - enum ip_ct_ftp_type type, - unsigned int matchoff, - unsigned int matchlen, - struct ip_conntrack_expect *exp, - u32 *seq); #endif /* _IP_CONNTRACK_FTP_H */ diff --git a/include/linux/netfilter_ipv4/ip_conntrack_icmp.h b/include/linux/netfilter_ipv4/ip_conntrack_icmp.h index f1664abbe392..eed5ee3e4744 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_icmp.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_icmp.h @@ -1,11 +1,6 @@ #ifndef _IP_CONNTRACK_ICMP_H #define _IP_CONNTRACK_ICMP_H -/* ICMP tracking. */ -#include -struct ip_ct_icmp -{ - /* Optimization: when number in == number out, forget immediately. */ - atomic_t count; -}; +#include + #endif /* _IP_CONNTRACK_ICMP_H */ diff --git a/include/linux/netfilter_ipv4/ip_conntrack_sctp.h b/include/linux/netfilter_ipv4/ip_conntrack_sctp.h index 7a8d869321f7..4099a041a32a 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_sctp.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_sctp.h @@ -1,25 +1,6 @@ #ifndef _IP_CONNTRACK_SCTP_H #define _IP_CONNTRACK_SCTP_H -/* SCTP tracking. */ -enum sctp_conntrack { - SCTP_CONNTRACK_NONE, - SCTP_CONNTRACK_CLOSED, - SCTP_CONNTRACK_COOKIE_WAIT, - SCTP_CONNTRACK_COOKIE_ECHOED, - SCTP_CONNTRACK_ESTABLISHED, - SCTP_CONNTRACK_SHUTDOWN_SENT, - SCTP_CONNTRACK_SHUTDOWN_RECD, - SCTP_CONNTRACK_SHUTDOWN_ACK_SENT, - SCTP_CONNTRACK_MAX -}; - -struct ip_ct_sctp -{ - enum sctp_conntrack state; - - u_int32_t vtag[IP_CT_DIR_MAX]; - u_int32_t ttag[IP_CT_DIR_MAX]; -}; +#include #endif /* _IP_CONNTRACK_SCTP_H */ diff --git a/include/linux/netfilter_ipv4/ip_conntrack_tcp.h b/include/linux/netfilter_ipv4/ip_conntrack_tcp.h index 16da044d97a7..876b8fb17e68 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_tcp.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_tcp.h @@ -1,51 +1,6 @@ #ifndef _IP_CONNTRACK_TCP_H #define _IP_CONNTRACK_TCP_H -/* TCP tracking. */ -enum tcp_conntrack { - TCP_CONNTRACK_NONE, - TCP_CONNTRACK_SYN_SENT, - TCP_CONNTRACK_SYN_RECV, - TCP_CONNTRACK_ESTABLISHED, - TCP_CONNTRACK_FIN_WAIT, - TCP_CONNTRACK_CLOSE_WAIT, - TCP_CONNTRACK_LAST_ACK, - TCP_CONNTRACK_TIME_WAIT, - TCP_CONNTRACK_CLOSE, - TCP_CONNTRACK_LISTEN, - TCP_CONNTRACK_MAX, - TCP_CONNTRACK_IGNORE -}; - -/* Window scaling is advertised by the sender */ -#define IP_CT_TCP_FLAG_WINDOW_SCALE 0x01 - -/* SACK is permitted by the sender */ -#define IP_CT_TCP_FLAG_SACK_PERM 0x02 - -/* This sender sent FIN first */ -#define IP_CT_TCP_FLAG_CLOSE_INIT 0x03 - -struct ip_ct_tcp_state { - u_int32_t td_end; /* max of seq + len */ - u_int32_t td_maxend; /* max of ack + max(win, 1) */ - u_int32_t td_maxwin; /* max(win) */ - u_int8_t td_scale; /* window scale factor */ - u_int8_t loose; /* used when connection picked up from the middle */ - u_int8_t flags; /* per direction options */ -}; - -struct ip_ct_tcp -{ - struct ip_ct_tcp_state seen[2]; /* connection parameters per direction */ - u_int8_t state; /* state of the connection (enum tcp_conntrack) */ - /* For detecting stale connections */ - u_int8_t last_dir; /* Direction of the last packet (enum ip_conntrack_dir) */ - u_int8_t retrans; /* Number of retransmitted packets */ - u_int8_t last_index; /* Index of the last packet */ - u_int32_t last_seq; /* Last sequence number seen in dir */ - u_int32_t last_ack; /* Last sequence number seen in opposite dir */ - u_int32_t last_end; /* Last seq + len */ -}; +#include #endif /* _IP_CONNTRACK_TCP_H */ diff --git a/include/linux/netfilter_ipv4/ip_conntrack_tuple.h b/include/linux/netfilter_ipv4/ip_conntrack_tuple.h index 3232db11a4e5..2fdabdb4c0ef 100644 --- a/include/linux/netfilter_ipv4/ip_conntrack_tuple.h +++ b/include/linux/netfilter_ipv4/ip_conntrack_tuple.h @@ -2,6 +2,7 @@ #define _IP_CONNTRACK_TUPLE_H #include +#include /* A `tuple' is a structure containing the information to uniquely identify a connection. ie. if two packets have the same tuple, they @@ -88,13 +89,6 @@ struct ip_conntrack_tuple (tuple)->dst.u.all = 0; \ } while (0) -enum ip_conntrack_dir -{ - IP_CT_DIR_ORIGINAL, - IP_CT_DIR_REPLY, - IP_CT_DIR_MAX -}; - #ifdef __KERNEL__ #define DUMP_TUPLE(tp) \ @@ -103,8 +97,6 @@ DEBUGP("tuple %p: %u %u.%u.%u.%u:%hu -> %u.%u.%u.%u:%hu\n", \ NIPQUAD((tp)->src.ip), ntohs((tp)->src.u.all), \ NIPQUAD((tp)->dst.ip), ntohs((tp)->dst.u.all)) -#define CTINFO2DIR(ctinfo) ((ctinfo) >= IP_CT_IS_REPLY ? IP_CT_DIR_REPLY : IP_CT_DIR_ORIGINAL) - /* If we're the first tuple, it's the original dir. */ #define DIRECTION(h) ((enum ip_conntrack_dir)(h)->tuple.dst.dir) diff --git a/include/linux/netfilter_ipv6.h b/include/linux/netfilter_ipv6.h index edcc2c6eb5c7..53b2983f6278 100644 --- a/include/linux/netfilter_ipv6.h +++ b/include/linux/netfilter_ipv6.h @@ -59,6 +59,7 @@ enum nf_ip6_hook_priorities { NF_IP6_PRI_FIRST = INT_MIN, + NF_IP6_PRI_CONNTRACK_DEFRAG = -400, NF_IP6_PRI_SELINUX_FIRST = -225, NF_IP6_PRI_CONNTRACK = -200, NF_IP6_PRI_BRIDGE_SABOTAGE_FORWARD = -175, diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index fdfb8fe8c38c..83010231db99 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -274,6 +274,9 @@ struct sk_buff { #if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE) __u8 ipvs_property:1; #endif +#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) + struct sk_buff *nfct_reasm; +#endif #ifdef CONFIG_BRIDGE_NETFILTER struct nf_bridge_info *nf_bridge; #endif @@ -1313,10 +1316,26 @@ static inline void nf_conntrack_get(struct nf_conntrack *nfct) if (nfct) atomic_inc(&nfct->use); } +#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) +static inline void nf_conntrack_get_reasm(struct sk_buff *skb) +{ + if (skb) + atomic_inc(&skb->users); +} +static inline void nf_conntrack_put_reasm(struct sk_buff *skb) +{ + if (skb) + kfree_skb(skb); +} +#endif static inline void nf_reset(struct sk_buff *skb) { nf_conntrack_put(skb->nfct); skb->nfct = NULL; +#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) + nf_conntrack_put_reasm(skb->nfct_reasm); + skb->nfct_reasm = NULL; +#endif } #ifdef CONFIG_BRIDGE_NETFILTER diff --git a/include/linux/sysctl.h b/include/linux/sysctl.h index fc131d6602b9..22cf5e1ac987 100644 --- a/include/linux/sysctl.h +++ b/include/linux/sysctl.h @@ -205,6 +205,7 @@ enum NET_ECONET=16, NET_SCTP=17, NET_LLC=18, + NET_NETFILTER=19, }; /* /proc/sys/kernel/random */ @@ -270,6 +271,42 @@ enum NET_UNIX_MAX_DGRAM_QLEN=3, }; +/* /proc/sys/net/netfilter */ +enum +{ + NET_NF_CONNTRACK_MAX=1, + NET_NF_CONNTRACK_TCP_TIMEOUT_SYN_SENT=2, + NET_NF_CONNTRACK_TCP_TIMEOUT_SYN_RECV=3, + NET_NF_CONNTRACK_TCP_TIMEOUT_ESTABLISHED=4, + NET_NF_CONNTRACK_TCP_TIMEOUT_FIN_WAIT=5, + NET_NF_CONNTRACK_TCP_TIMEOUT_CLOSE_WAIT=6, + NET_NF_CONNTRACK_TCP_TIMEOUT_LAST_ACK=7, + NET_NF_CONNTRACK_TCP_TIMEOUT_TIME_WAIT=8, + NET_NF_CONNTRACK_TCP_TIMEOUT_CLOSE=9, + NET_NF_CONNTRACK_UDP_TIMEOUT=10, + NET_NF_CONNTRACK_UDP_TIMEOUT_STREAM=11, + NET_NF_CONNTRACK_ICMP_TIMEOUT=12, + NET_NF_CONNTRACK_GENERIC_TIMEOUT=13, + NET_NF_CONNTRACK_BUCKETS=14, + NET_NF_CONNTRACK_LOG_INVALID=15, + NET_NF_CONNTRACK_TCP_TIMEOUT_MAX_RETRANS=16, + NET_NF_CONNTRACK_TCP_LOOSE=17, + NET_NF_CONNTRACK_TCP_BE_LIBERAL=18, + NET_NF_CONNTRACK_TCP_MAX_RETRANS=19, + NET_NF_CONNTRACK_SCTP_TIMEOUT_CLOSED=20, + NET_NF_CONNTRACK_SCTP_TIMEOUT_COOKIE_WAIT=21, + NET_NF_CONNTRACK_SCTP_TIMEOUT_COOKIE_ECHOED=22, + NET_NF_CONNTRACK_SCTP_TIMEOUT_ESTABLISHED=23, + NET_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_SENT=24, + NET_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_RECD=25, + NET_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_ACK_SENT=26, + NET_NF_CONNTRACK_COUNT=27, + NET_NF_CONNTRACK_ICMPV6_TIMEOUT=28, + NET_NF_CONNTRACK_FRAG6_TIMEOUT=29, + NET_NF_CONNTRACK_FRAG6_LOW_THRESH=30, + NET_NF_CONNTRACK_FRAG6_HIGH_THRESH=31, +}; + /* /proc/sys/net/ipv4 */ enum { diff --git a/include/net/netfilter/ipv4/nf_conntrack_icmp.h b/include/net/netfilter/ipv4/nf_conntrack_icmp.h new file mode 100644 index 000000000000..3dd22cff23ec --- /dev/null +++ b/include/net/netfilter/ipv4/nf_conntrack_icmp.h @@ -0,0 +1,11 @@ +#ifndef _NF_CONNTRACK_ICMP_H +#define _NF_CONNTRACK_ICMP_H +/* ICMP tracking. */ +#include + +struct ip_ct_icmp +{ + /* Optimization: when number in == number out, forget immediately. */ + atomic_t count; +}; +#endif /* _NF_CONNTRACK_ICMP_H */ diff --git a/include/net/netfilter/ipv4/nf_conntrack_ipv4.h b/include/net/netfilter/ipv4/nf_conntrack_ipv4.h new file mode 100644 index 000000000000..25b081a730e6 --- /dev/null +++ b/include/net/netfilter/ipv4/nf_conntrack_ipv4.h @@ -0,0 +1,43 @@ +/* + * IPv4 support for nf_conntrack. + * + * 23 Mar 2004: Yasuyuki Kozakai @ USAGI + * - move L3 protocol dependent part from include/linux/netfilter_ipv4/ + * ip_conntarck.h + */ + +#ifndef _NF_CONNTRACK_IPV4_H +#define _NF_CONNTRACK_IPV4_H + +#ifdef CONFIG_IP_NF_NAT_NEEDED +#include + +/* per conntrack: nat application helper private data */ +union ip_conntrack_nat_help { + /* insert nat helper private data here */ +}; + +struct nf_conntrack_ipv4_nat { + struct ip_nat_info info; + union ip_conntrack_nat_help help; +#if defined(CONFIG_IP_NF_TARGET_MASQUERADE) || \ + defined(CONFIG_IP_NF_TARGET_MASQUERADE_MODULE) + int masq_index; +#endif +}; +#endif /* CONFIG_IP_NF_NAT_NEEDED */ + +struct nf_conntrack_ipv4 { +#ifdef CONFIG_IP_NF_NAT_NEEDED + struct nf_conntrack_ipv4_nat *nat; +#endif +}; + +/* Returns new sk_buff, or NULL */ +struct sk_buff * +nf_ct_ipv4_ct_gather_frags(struct sk_buff *skb); + +/* call to create an explicit dependency on nf_conntrack_l3proto_ipv4. */ +extern void need_ip_conntrack(void); + +#endif /*_NF_CONNTRACK_IPV4_H*/ diff --git a/include/net/netfilter/ipv6/nf_conntrack_icmpv6.h b/include/net/netfilter/ipv6/nf_conntrack_icmpv6.h new file mode 100644 index 000000000000..86591afda29c --- /dev/null +++ b/include/net/netfilter/ipv6/nf_conntrack_icmpv6.h @@ -0,0 +1,27 @@ +/* + * ICMPv6 tracking. + * + * 21 Apl 2004: Yasuyuki Kozakai @USAGI + * - separated from nf_conntrack_icmp.h + * + * Derived from include/linux/netfiter_ipv4/ip_conntrack_icmp.h + */ + +#ifndef _NF_CONNTRACK_ICMPV6_H +#define _NF_CONNTRACK_ICMPV6_H +#include + +#ifndef ICMPV6_NI_QUERY +#define ICMPV6_NI_QUERY 139 +#endif +#ifndef ICMPV6_NI_REPLY +#define ICMPV6_NI_REPLY 140 +#endif + +struct nf_ct_icmpv6 +{ + /* Optimization: when number in == number out, forget immediately. */ + atomic_t count; +}; + +#endif /* _NF_CONNTRACK_ICMPV6_H */ diff --git a/include/net/netfilter/nf_conntrack.h b/include/net/netfilter/nf_conntrack.h new file mode 100644 index 000000000000..cc4825610795 --- /dev/null +++ b/include/net/netfilter/nf_conntrack.h @@ -0,0 +1,354 @@ +/* + * Connection state tracking for netfilter. This is separated from, + * but required by, the (future) NAT layer; it can also be used by an iptables + * extension. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - generalize L3 protocol dependent part. + * + * Derived from include/linux/netfiter_ipv4/ip_conntrack.h + */ + +#ifndef _NF_CONNTRACK_H +#define _NF_CONNTRACK_H + +#include + +#ifdef __KERNEL__ +#include +#include +#include +#include + +#include +#include +#include +#include + +#include + +/* per conntrack: protocol private data */ +union nf_conntrack_proto { + /* insert conntrack proto private data here */ + struct ip_ct_sctp sctp; + struct ip_ct_tcp tcp; + struct ip_ct_icmp icmp; + struct nf_ct_icmpv6 icmpv6; +}; + +union nf_conntrack_expect_proto { + /* insert expect proto private data here */ +}; + +/* Add protocol helper include file here */ +#include + +/* per conntrack: application helper private data */ +union nf_conntrack_help { + /* insert conntrack helper private data (master) here */ + struct ip_ct_ftp_master ct_ftp_info; +}; + +#include +#include + +#ifdef CONFIG_NETFILTER_DEBUG +#define NF_CT_ASSERT(x) \ +do { \ + if (!(x)) \ + /* Wooah! I'm tripping my conntrack in a frenzy of \ + netplay... */ \ + printk("NF_CT_ASSERT: %s:%i(%s)\n", \ + __FILE__, __LINE__, __FUNCTION__); \ +} while(0) +#else +#define NF_CT_ASSERT(x) +#endif + +struct nf_conntrack_helper; + +#include +struct nf_conn +{ + /* Usage count in here is 1 for hash table/destruct timer, 1 per skb, + plus 1 for any connection(s) we are `master' for */ + struct nf_conntrack ct_general; + + /* XXX should I move this to the tail ? - Y.K */ + /* These are my tuples; original and reply */ + struct nf_conntrack_tuple_hash tuplehash[IP_CT_DIR_MAX]; + + /* Have we seen traffic both ways yet? (bitset) */ + unsigned long status; + + /* Timer function; drops refcnt when it goes off. */ + struct timer_list timeout; + +#ifdef CONFIG_NF_CT_ACCT + /* Accounting Information (same cache line as other written members) */ + struct ip_conntrack_counter counters[IP_CT_DIR_MAX]; +#endif + /* If we were expected by an expectation, this will be it */ + struct nf_conn *master; + + /* Current number of expected connections */ + unsigned int expecting; + + /* Helper. if any */ + struct nf_conntrack_helper *helper; + + /* features - nat, helper, ... used by allocating system */ + u_int32_t features; + + /* Storage reserved for other modules: */ + + union nf_conntrack_proto proto; + +#if defined(CONFIG_NF_CONNTRACK_MARK) + u_int32_t mark; +#endif + + /* These members are dynamically allocated. */ + + union nf_conntrack_help *help; + + /* Layer 3 dependent members. (ex: NAT) */ + union { + struct nf_conntrack_ipv4 *ipv4; + } l3proto; + void *data[0]; +}; + +struct nf_conntrack_expect +{ + /* Internal linked list (global expectation list) */ + struct list_head list; + + /* We expect this tuple, with the following mask */ + struct nf_conntrack_tuple tuple, mask; + + /* Function to call after setup and insertion */ + void (*expectfn)(struct nf_conn *new, + struct nf_conntrack_expect *this); + + /* The conntrack of the master connection */ + struct nf_conn *master; + + /* Timer function; deletes the expectation. */ + struct timer_list timeout; + + /* Usage count. */ + atomic_t use; + + /* Flags */ + unsigned int flags; + +#ifdef CONFIG_NF_NAT_NEEDED + /* This is the original per-proto part, used to map the + * expected connection the way the recipient expects. */ + union nf_conntrack_manip_proto saved_proto; + /* Direction relative to the master connection. */ + enum ip_conntrack_dir dir; +#endif +}; + +#define NF_CT_EXPECT_PERMANENT 0x1 + +static inline struct nf_conn * +nf_ct_tuplehash_to_ctrack(const struct nf_conntrack_tuple_hash *hash) +{ + return container_of(hash, struct nf_conn, + tuplehash[hash->tuple.dst.dir]); +} + +/* get master conntrack via master expectation */ +#define master_ct(conntr) (conntr->master) + +/* Alter reply tuple (maybe alter helper). */ +extern void +nf_conntrack_alter_reply(struct nf_conn *conntrack, + const struct nf_conntrack_tuple *newreply); + +/* Is this tuple taken? (ignoring any belonging to the given + conntrack). */ +extern int +nf_conntrack_tuple_taken(const struct nf_conntrack_tuple *tuple, + const struct nf_conn *ignored_conntrack); + +/* Return conntrack_info and tuple hash for given skb. */ +static inline struct nf_conn * +nf_ct_get(const struct sk_buff *skb, enum ip_conntrack_info *ctinfo) +{ + *ctinfo = skb->nfctinfo; + return (struct nf_conn *)skb->nfct; +} + +/* decrement reference count on a conntrack */ +static inline void nf_ct_put(struct nf_conn *ct) +{ + NF_CT_ASSERT(ct); + nf_conntrack_put(&ct->ct_general); +} + +/* call to create an explicit dependency on nf_conntrack. */ +extern void need_nf_conntrack(void); + +extern int nf_ct_invert_tuplepr(struct nf_conntrack_tuple *inverse, + const struct nf_conntrack_tuple *orig); + +extern void __nf_ct_refresh_acct(struct nf_conn *ct, + enum ip_conntrack_info ctinfo, + const struct sk_buff *skb, + unsigned long extra_jiffies, + int do_acct); + +/* Refresh conntrack for this many jiffies and do accounting */ +static inline void nf_ct_refresh_acct(struct nf_conn *ct, + enum ip_conntrack_info ctinfo, + const struct sk_buff *skb, + unsigned long extra_jiffies) +{ + __nf_ct_refresh_acct(ct, ctinfo, skb, extra_jiffies, 1); +} + +/* Refresh conntrack for this many jiffies */ +static inline void nf_ct_refresh(struct nf_conn *ct, + const struct sk_buff *skb, + unsigned long extra_jiffies) +{ + __nf_ct_refresh_acct(ct, 0, skb, extra_jiffies, 0); +} + +/* These are for NAT. Icky. */ +/* Update TCP window tracking data when NAT mangles the packet */ +extern void nf_conntrack_tcp_update(struct sk_buff *skb, + unsigned int dataoff, + struct nf_conn *conntrack, + int dir); + +/* Call me when a conntrack is destroyed. */ +extern void (*nf_conntrack_destroyed)(struct nf_conn *conntrack); + +/* Fake conntrack entry for untracked connections */ +extern struct nf_conn nf_conntrack_untracked; + +extern int nf_ct_no_defrag; + +/* Iterate over all conntracks: if iter returns true, it's deleted. */ +extern void +nf_ct_iterate_cleanup(int (*iter)(struct nf_conn *i, void *data), void *data); +extern void nf_conntrack_free(struct nf_conn *ct); +extern struct nf_conn * +nf_conntrack_alloc(const struct nf_conntrack_tuple *orig, + const struct nf_conntrack_tuple *repl); + +/* It's confirmed if it is, or has been in the hash table. */ +static inline int nf_ct_is_confirmed(struct nf_conn *ct) +{ + return test_bit(IPS_CONFIRMED_BIT, &ct->status); +} + +static inline int nf_ct_is_dying(struct nf_conn *ct) +{ + return test_bit(IPS_DYING_BIT, &ct->status); +} + +extern unsigned int nf_conntrack_htable_size; + +#define NF_CT_STAT_INC(count) (__get_cpu_var(nf_conntrack_stat).count++) + +#ifdef CONFIG_NF_CONNTRACK_EVENTS +#include +#include + +struct nf_conntrack_ecache { + struct nf_conn *ct; + unsigned int events; +}; +DECLARE_PER_CPU(struct nf_conntrack_ecache, nf_conntrack_ecache); + +#define CONNTRACK_ECACHE(x) (__get_cpu_var(nf_conntrack_ecache).x) + +extern struct notifier_block *nf_conntrack_chain; +extern struct notifier_block *nf_conntrack_expect_chain; + +static inline int nf_conntrack_register_notifier(struct notifier_block *nb) +{ + return notifier_chain_register(&nf_conntrack_chain, nb); +} + +static inline int nf_conntrack_unregister_notifier(struct notifier_block *nb) +{ + return notifier_chain_unregister(&nf_conntrack_chain, nb); +} + +static inline int +nf_conntrack_expect_register_notifier(struct notifier_block *nb) +{ + return notifier_chain_register(&nf_conntrack_expect_chain, nb); +} + +static inline int +nf_conntrack_expect_unregister_notifier(struct notifier_block *nb) +{ + return notifier_chain_unregister(&nf_conntrack_expect_chain, nb); +} + +extern void nf_ct_deliver_cached_events(const struct nf_conn *ct); +extern void __nf_ct_event_cache_init(struct nf_conn *ct); + +static inline void +nf_conntrack_event_cache(enum ip_conntrack_events event, + const struct sk_buff *skb) +{ + struct nf_conn *ct = (struct nf_conn *)skb->nfct; + struct nf_conntrack_ecache *ecache; + + local_bh_disable(); + ecache = &__get_cpu_var(nf_conntrack_ecache); + if (ct != ecache->ct) + __nf_ct_event_cache_init(ct); + ecache->events |= event; + local_bh_enable(); +} + +static inline void nf_conntrack_event(enum ip_conntrack_events event, + struct nf_conn *ct) +{ + if (nf_ct_is_confirmed(ct) && !nf_ct_is_dying(ct)) + notifier_call_chain(&nf_conntrack_chain, event, ct); +} + +static inline void +nf_conntrack_expect_event(enum ip_conntrack_expect_events event, + struct nf_conntrack_expect *exp) +{ + notifier_call_chain(&nf_conntrack_expect_chain, event, exp); +} +#else /* CONFIG_NF_CONNTRACK_EVENTS */ +static inline void nf_conntrack_event_cache(enum ip_conntrack_events event, + const struct sk_buff *skb) {} +static inline void nf_conntrack_event(enum ip_conntrack_events event, + struct nf_conn *ct) {} +static inline void nf_ct_deliver_cached_events(const struct nf_conn *ct) {} +static inline void +nf_conntrack_expect_event(enum ip_conntrack_expect_events event, + struct nf_conntrack_expect *exp) {} +#endif /* CONFIG_NF_CONNTRACK_EVENTS */ + +/* no helper, no nat */ +#define NF_CT_F_BASIC 0 +/* for helper */ +#define NF_CT_F_HELP 1 +/* for nat. */ +#define NF_CT_F_NAT 2 +#define NF_CT_F_NUM 4 + +extern int +nf_conntrack_register_cache(u_int32_t features, const char *name, size_t size, + int (*init_conntrack)(struct nf_conn *, u_int32_t)); +extern void +nf_conntrack_unregister_cache(u_int32_t features); + +#endif /* __KERNEL__ */ +#endif /* _NF_CONNTRACK_H */ diff --git a/include/net/netfilter/nf_conntrack_compat.h b/include/net/netfilter/nf_conntrack_compat.h new file mode 100644 index 000000000000..3cac19fb3648 --- /dev/null +++ b/include/net/netfilter/nf_conntrack_compat.h @@ -0,0 +1,108 @@ +#ifndef _NF_CONNTRACK_COMPAT_H +#define _NF_CONNTRACK_COMPAT_H + +#ifdef __KERNEL__ + +#if defined(CONFIG_IP_NF_CONNTRACK) || defined(CONFIG_IP_NF_CONNTRACK_MODULE) + +#include + +#ifdef CONFIG_IP_NF_CONNTRACK_MARK +static inline u_int32_t *nf_ct_get_mark(const struct sk_buff *skb, + u_int32_t *ctinfo) +{ + struct ip_conntrack *ct = ip_conntrack_get(skb, ctinfo); + + if (ct) + return &ct->mark; + else + return NULL; +} +#endif /* CONFIG_IP_NF_CONNTRACK_MARK */ + +#ifdef CONFIG_IP_NF_CT_ACCT +static inline struct ip_conntrack_counter * +nf_ct_get_counters(const struct sk_buff *skb) +{ + enum ip_conntrack_info ctinfo; + struct ip_conntrack *ct = ip_conntrack_get(skb, &ctinfo); + + if (ct) + return ct->counters; + else + return NULL; +} +#endif /* CONFIG_IP_NF_CT_ACCT */ + +static inline int nf_ct_is_untracked(const struct sk_buff *skb) +{ + return (skb->nfct == &ip_conntrack_untracked.ct_general); +} + +static inline void nf_ct_untrack(struct sk_buff *skb) +{ + skb->nfct = &ip_conntrack_untracked.ct_general; +} + +static inline int nf_ct_get_ctinfo(const struct sk_buff *skb, + enum ip_conntrack_info *ctinfo) +{ + struct ip_conntrack *ct = ip_conntrack_get(skb, ctinfo); + return (ct != NULL); +} + +#else /* CONFIG_IP_NF_CONNTRACK */ + +#include +#include + +#ifdef CONFIG_NF_CONNTRACK_MARK + +static inline u_int32_t *nf_ct_get_mark(const struct sk_buff *skb, + u_int32_t *ctinfo) +{ + struct nf_conn *ct = nf_ct_get(skb, ctinfo); + + if (ct) + return &ct->mark; + else + return NULL; +} +#endif /* CONFIG_NF_CONNTRACK_MARK */ + +#ifdef CONFIG_NF_CT_ACCT +static inline struct ip_conntrack_counter * +nf_ct_get_counters(const struct sk_buff *skb) +{ + enum ip_conntrack_info ctinfo; + struct nf_conn *ct = nf_ct_get(skb, &ctinfo); + + if (ct) + return ct->counters; + else + return NULL; +} +#endif /* CONFIG_NF_CT_ACCT */ + +static inline int nf_ct_is_untracked(const struct sk_buff *skb) +{ + return (skb->nfct == &nf_conntrack_untracked.ct_general); +} + +static inline void nf_ct_untrack(struct sk_buff *skb) +{ + skb->nfct = &nf_conntrack_untracked.ct_general; +} + +static inline int nf_ct_get_ctinfo(const struct sk_buff *skb, + enum ip_conntrack_info *ctinfo) +{ + struct nf_conn *ct = nf_ct_get(skb, ctinfo); + return (ct != NULL); +} + +#endif /* CONFIG_IP_NF_CONNTRACK */ + +#endif /* __KERNEL__ */ + +#endif /* _NF_CONNTRACK_COMPAT_H */ diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h new file mode 100644 index 000000000000..da254525a4ce --- /dev/null +++ b/include/net/netfilter/nf_conntrack_core.h @@ -0,0 +1,76 @@ +/* + * This header is used to share core functionality between the + * standalone connection tracking module, and the compatibility layer's use + * of connection tracking. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - generalize L3 protocol dependent part. + * + * Derived from include/linux/netfiter_ipv4/ip_conntrack_core.h + */ + +#ifndef _NF_CONNTRACK_CORE_H +#define _NF_CONNTRACK_CORE_H + +#include + +/* This header is used to share core functionality between the + standalone connection tracking module, and the compatibility layer's use + of connection tracking. */ +extern unsigned int nf_conntrack_in(int pf, + unsigned int hooknum, + struct sk_buff **pskb); + +extern int nf_conntrack_init(void); +extern void nf_conntrack_cleanup(void); + +struct nf_conntrack_l3proto; +extern struct nf_conntrack_l3proto *nf_ct_find_l3proto(u_int16_t pf); +/* Like above, but you already have conntrack read lock. */ +extern struct nf_conntrack_l3proto *__nf_ct_find_l3proto(u_int16_t l3proto); + +struct nf_conntrack_protocol; + +extern int +nf_ct_get_tuple(const struct sk_buff *skb, + unsigned int nhoff, + unsigned int dataoff, + u_int16_t l3num, + u_int8_t protonum, + struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_l3proto *l3proto, + const struct nf_conntrack_protocol *protocol); + +extern int +nf_ct_invert_tuple(struct nf_conntrack_tuple *inverse, + const struct nf_conntrack_tuple *orig, + const struct nf_conntrack_l3proto *l3proto, + const struct nf_conntrack_protocol *protocol); + +/* Find a connection corresponding to a tuple. */ +extern struct nf_conntrack_tuple_hash * +nf_conntrack_find_get(const struct nf_conntrack_tuple *tuple, + const struct nf_conn *ignored_conntrack); + +extern int __nf_conntrack_confirm(struct sk_buff **pskb); + +/* Confirm a connection: returns NF_DROP if packet must be dropped. */ +static inline int nf_conntrack_confirm(struct sk_buff **pskb) +{ + struct nf_conn *ct = (struct nf_conn *)(*pskb)->nfct; + int ret = NF_ACCEPT; + + if (ct) { + if (!nf_ct_is_confirmed(ct)) + ret = __nf_conntrack_confirm(pskb); + nf_ct_deliver_cached_events(ct); + } + return ret; +} + +extern void __nf_conntrack_attach(struct sk_buff *nskb, struct sk_buff *skb); + +extern struct list_head *nf_conntrack_hash; +extern struct list_head nf_conntrack_expect_list; +extern rwlock_t nf_conntrack_lock ; +#endif /* _NF_CONNTRACK_CORE_H */ diff --git a/include/net/netfilter/nf_conntrack_helper.h b/include/net/netfilter/nf_conntrack_helper.h new file mode 100644 index 000000000000..5a66b2a3a623 --- /dev/null +++ b/include/net/netfilter/nf_conntrack_helper.h @@ -0,0 +1,51 @@ +/* + * connection tracking helpers. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - generalize L3 protocol dependent part. + * + * Derived from include/linux/netfiter_ipv4/ip_conntrack_helper.h + */ + +#ifndef _NF_CONNTRACK_HELPER_H +#define _NF_CONNTRACK_HELPER_H +#include + +struct module; + +struct nf_conntrack_helper +{ + struct list_head list; /* Internal use. */ + + const char *name; /* name of the module */ + struct module *me; /* pointer to self */ + unsigned int max_expected; /* Maximum number of concurrent + * expected connections */ + unsigned int timeout; /* timeout for expecteds */ + + /* Mask of things we will help (compared against server response) */ + struct nf_conntrack_tuple tuple; + struct nf_conntrack_tuple mask; + + /* Function to call when data passes; return verdict, or -1 to + invalidate. */ + int (*help)(struct sk_buff **pskb, + unsigned int protoff, + struct nf_conn *ct, + enum ip_conntrack_info conntrackinfo); +}; + +extern int nf_conntrack_helper_register(struct nf_conntrack_helper *); +extern void nf_conntrack_helper_unregister(struct nf_conntrack_helper *); + +/* Allocate space for an expectation: this is mandatory before calling + nf_conntrack_expect_related. You will have to call put afterwards. */ +extern struct nf_conntrack_expect * +nf_conntrack_expect_alloc(struct nf_conn *master); +extern void nf_conntrack_expect_put(struct nf_conntrack_expect *exp); + +/* Add an expected connection: can have more than one per connection */ +extern int nf_conntrack_expect_related(struct nf_conntrack_expect *exp); +extern void nf_conntrack_unexpect_related(struct nf_conntrack_expect *exp); + +#endif /*_NF_CONNTRACK_HELPER_H*/ diff --git a/include/net/netfilter/nf_conntrack_l3proto.h b/include/net/netfilter/nf_conntrack_l3proto.h new file mode 100644 index 000000000000..01663e5b33df --- /dev/null +++ b/include/net/netfilter/nf_conntrack_l3proto.h @@ -0,0 +1,93 @@ +/* + * Copyright (C)2003,2004 USAGI/WIDE Project + * + * Header for use in defining a given L3 protocol for connection tracking. + * + * Author: + * Yasuyuki Kozakai @USAGI + * + * Derived from include/netfilter_ipv4/ip_conntrack_protocol.h + */ + +#ifndef _NF_CONNTRACK_L3PROTO_H +#define _NF_CONNTRACK_L3PROTO_H +#include +#include + +struct nf_conntrack_l3proto +{ + /* Next pointer. */ + struct list_head list; + + /* L3 Protocol Family number. ex) PF_INET */ + u_int16_t l3proto; + + /* Protocol name */ + const char *name; + + /* + * Try to fill in the third arg: nhoff is offset of l3 proto + * hdr. Return true if possible. + */ + int (*pkt_to_tuple)(const struct sk_buff *skb, unsigned int nhoff, + struct nf_conntrack_tuple *tuple); + + /* + * Invert the per-proto part of the tuple: ie. turn xmit into reply. + * Some packets can't be inverted: return 0 in that case. + */ + int (*invert_tuple)(struct nf_conntrack_tuple *inverse, + const struct nf_conntrack_tuple *orig); + + /* Print out the per-protocol part of the tuple. */ + int (*print_tuple)(struct seq_file *s, + const struct nf_conntrack_tuple *); + + /* Print out the private part of the conntrack. */ + int (*print_conntrack)(struct seq_file *s, const struct nf_conn *); + + /* Returns verdict for packet, or -1 for invalid. */ + int (*packet)(struct nf_conn *conntrack, + const struct sk_buff *skb, + enum ip_conntrack_info ctinfo); + + /* + * Called when a new connection for this protocol found; + * returns TRUE if it's OK. If so, packet() called next. + */ + int (*new)(struct nf_conn *conntrack, const struct sk_buff *skb); + + /* Called when a conntrack entry is destroyed */ + void (*destroy)(struct nf_conn *conntrack); + + /* + * Called before tracking. + * *dataoff: offset of protocol header (TCP, UDP,...) in *pskb + * *protonum: protocol number + */ + int (*prepare)(struct sk_buff **pskb, unsigned int hooknum, + unsigned int *dataoff, u_int8_t *protonum); + + u_int32_t (*get_features)(const struct nf_conntrack_tuple *tuple); + + /* Module (if any) which this is connected to. */ + struct module *me; +}; + +extern struct nf_conntrack_l3proto *nf_ct_l3protos[AF_MAX]; + +/* Protocol registration. */ +extern int nf_conntrack_l3proto_register(struct nf_conntrack_l3proto *proto); +extern void nf_conntrack_l3proto_unregister(struct nf_conntrack_l3proto *proto); + +static inline struct nf_conntrack_l3proto * +nf_ct_find_l3proto(u_int16_t l3proto) +{ + return nf_ct_l3protos[l3proto]; +} + +/* Existing built-in protocols */ +extern struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4; +extern struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv6; +extern struct nf_conntrack_l3proto nf_conntrack_generic_l3proto; +#endif /*_NF_CONNTRACK_L3PROTO_H*/ diff --git a/include/net/netfilter/nf_conntrack_protocol.h b/include/net/netfilter/nf_conntrack_protocol.h new file mode 100644 index 000000000000..b3afda35397a --- /dev/null +++ b/include/net/netfilter/nf_conntrack_protocol.h @@ -0,0 +1,105 @@ +/* + * Header for use in defining a given protocol for connection tracking. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - generalized L3 protocol dependent part. + * + * Derived from include/linux/netfiter_ipv4/ip_conntrack_protcol.h + */ + +#ifndef _NF_CONNTRACK_PROTOCOL_H +#define _NF_CONNTRACK_PROTOCOL_H +#include + +struct seq_file; + +struct nf_conntrack_protocol +{ + /* Next pointer. */ + struct list_head list; + + /* L3 Protocol number. */ + u_int16_t l3proto; + + /* Protocol number. */ + u_int8_t proto; + + /* Protocol name */ + const char *name; + + /* Try to fill in the third arg: dataoff is offset past network protocol + hdr. Return true if possible. */ + int (*pkt_to_tuple)(const struct sk_buff *skb, + unsigned int dataoff, + struct nf_conntrack_tuple *tuple); + + /* Invert the per-proto part of the tuple: ie. turn xmit into reply. + * Some packets can't be inverted: return 0 in that case. + */ + int (*invert_tuple)(struct nf_conntrack_tuple *inverse, + const struct nf_conntrack_tuple *orig); + + /* Print out the per-protocol part of the tuple. Return like seq_* */ + int (*print_tuple)(struct seq_file *s, + const struct nf_conntrack_tuple *); + + /* Print out the private part of the conntrack. */ + int (*print_conntrack)(struct seq_file *s, const struct nf_conn *); + + /* Returns verdict for packet, or -1 for invalid. */ + int (*packet)(struct nf_conn *conntrack, + const struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info ctinfo, + int pf, + unsigned int hooknum); + + /* Called when a new connection for this protocol found; + * returns TRUE if it's OK. If so, packet() called next. */ + int (*new)(struct nf_conn *conntrack, const struct sk_buff *skb, + unsigned int dataoff); + + /* Called when a conntrack entry is destroyed */ + void (*destroy)(struct nf_conn *conntrack); + + int (*error)(struct sk_buff *skb, unsigned int dataoff, + enum ip_conntrack_info *ctinfo, + int pf, unsigned int hooknum); + + /* Module (if any) which this is connected to. */ + struct module *me; +}; + +/* Existing built-in protocols */ +extern struct nf_conntrack_protocol nf_conntrack_protocol_tcp6; +extern struct nf_conntrack_protocol nf_conntrack_protocol_udp4; +extern struct nf_conntrack_protocol nf_conntrack_protocol_udp6; +extern struct nf_conntrack_protocol nf_conntrack_generic_protocol; + +#define MAX_NF_CT_PROTO 256 +extern struct nf_conntrack_protocol **nf_ct_protos[PF_MAX]; + +extern struct nf_conntrack_protocol * +nf_ct_find_proto(u_int16_t l3proto, u_int8_t protocol); + +/* Protocol registration. */ +extern int nf_conntrack_protocol_register(struct nf_conntrack_protocol *proto); +extern void nf_conntrack_protocol_unregister(struct nf_conntrack_protocol *proto); + +/* Log invalid packets */ +extern unsigned int nf_ct_log_invalid; + +#ifdef CONFIG_SYSCTL +#ifdef DEBUG_INVALID_PACKETS +#define LOG_INVALID(proto) \ + (nf_ct_log_invalid == (proto) || nf_ct_log_invalid == IPPROTO_RAW) +#else +#define LOG_INVALID(proto) \ + ((nf_ct_log_invalid == (proto) || nf_ct_log_invalid == IPPROTO_RAW) \ + && net_ratelimit()) +#endif +#else +#define LOG_INVALID(proto) 0 +#endif /* CONFIG_SYSCTL */ + +#endif /*_NF_CONNTRACK_PROTOCOL_H*/ diff --git a/include/net/netfilter/nf_conntrack_tuple.h b/include/net/netfilter/nf_conntrack_tuple.h new file mode 100644 index 000000000000..14ce790e5c65 --- /dev/null +++ b/include/net/netfilter/nf_conntrack_tuple.h @@ -0,0 +1,190 @@ +/* + * Definitions and Declarations for tuple. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - generalize L3 protocol dependent part. + * + * Derived from include/linux/netfiter_ipv4/ip_conntrack_tuple.h + */ + +#ifndef _NF_CONNTRACK_TUPLE_H +#define _NF_CONNTRACK_TUPLE_H + +#include + +/* A `tuple' is a structure containing the information to uniquely + identify a connection. ie. if two packets have the same tuple, they + are in the same connection; if not, they are not. + + We divide the structure along "manipulatable" and + "non-manipulatable" lines, for the benefit of the NAT code. +*/ + +#define NF_CT_TUPLE_L3SIZE 4 + +/* The l3 protocol-specific manipulable parts of the tuple: always in + network order! */ +union nf_conntrack_man_l3proto { + u_int32_t all[NF_CT_TUPLE_L3SIZE]; + u_int32_t ip; + u_int32_t ip6[4]; +}; + +/* The protocol-specific manipulable parts of the tuple: always in + network order! */ +union nf_conntrack_man_proto +{ + /* Add other protocols here. */ + u_int16_t all; + + struct { + u_int16_t port; + } tcp; + struct { + u_int16_t port; + } udp; + struct { + u_int16_t id; + } icmp; + struct { + u_int16_t port; + } sctp; +}; + +/* The manipulable part of the tuple. */ +struct nf_conntrack_man +{ + union nf_conntrack_man_l3proto u3; + union nf_conntrack_man_proto u; + /* Layer 3 protocol */ + u_int16_t l3num; +}; + +/* This contains the information to distinguish a connection. */ +struct nf_conntrack_tuple +{ + struct nf_conntrack_man src; + + /* These are the parts of the tuple which are fixed. */ + struct { + union { + u_int32_t all[NF_CT_TUPLE_L3SIZE]; + u_int32_t ip; + u_int32_t ip6[4]; + } u3; + union { + /* Add other protocols here. */ + u_int16_t all; + + struct { + u_int16_t port; + } tcp; + struct { + u_int16_t port; + } udp; + struct { + u_int8_t type, code; + } icmp; + struct { + u_int16_t port; + } sctp; + } u; + + /* The protocol. */ + u_int8_t protonum; + + /* The direction (for tuplehash) */ + u_int8_t dir; + } dst; +}; + +/* This is optimized opposed to a memset of the whole structure. Everything we + * really care about is the source/destination unions */ +#define NF_CT_TUPLE_U_BLANK(tuple) \ + do { \ + (tuple)->src.u.all = 0; \ + (tuple)->dst.u.all = 0; \ + memset(&(tuple)->src.u3, 0, sizeof((tuple)->src.u3)); \ + memset(&(tuple)->dst.u3, 0, sizeof((tuple)->dst.u3)); \ + } while (0) + +#ifdef __KERNEL__ + +#define NF_CT_DUMP_TUPLE(tp) \ +DEBUGP("tuple %p: %u %u %04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x %hu -> %04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x %hu\n", \ + (tp), (tp)->src.l3num, (tp)->dst.protonum, \ + NIP6(*(struct in6_addr *)(tp)->src.u3.all), ntohs((tp)->src.u.all), \ + NIP6(*(struct in6_addr *)(tp)->dst.u3.all), ntohs((tp)->dst.u.all)) + +/* If we're the first tuple, it's the original dir. */ +#define NF_CT_DIRECTION(h) \ + ((enum ip_conntrack_dir)(h)->tuple.dst.dir) + +/* Connections have two entries in the hash table: one for each way */ +struct nf_conntrack_tuple_hash +{ + struct list_head list; + + struct nf_conntrack_tuple tuple; +}; + +#endif /* __KERNEL__ */ + +static inline int nf_ct_tuple_src_equal(const struct nf_conntrack_tuple *t1, + const struct nf_conntrack_tuple *t2) +{ + return (t1->src.u3.all[0] == t2->src.u3.all[0] && + t1->src.u3.all[1] == t2->src.u3.all[1] && + t1->src.u3.all[2] == t2->src.u3.all[2] && + t1->src.u3.all[3] == t2->src.u3.all[3] && + t1->src.u.all == t2->src.u.all && + t1->src.l3num == t2->src.l3num && + t1->dst.protonum == t2->dst.protonum); +} + +static inline int nf_ct_tuple_dst_equal(const struct nf_conntrack_tuple *t1, + const struct nf_conntrack_tuple *t2) +{ + return (t1->dst.u3.all[0] == t2->dst.u3.all[0] && + t1->dst.u3.all[1] == t2->dst.u3.all[1] && + t1->dst.u3.all[2] == t2->dst.u3.all[2] && + t1->dst.u3.all[3] == t2->dst.u3.all[3] && + t1->dst.u.all == t2->dst.u.all && + t1->src.l3num == t2->src.l3num && + t1->dst.protonum == t2->dst.protonum); +} + +static inline int nf_ct_tuple_equal(const struct nf_conntrack_tuple *t1, + const struct nf_conntrack_tuple *t2) +{ + return nf_ct_tuple_src_equal(t1, t2) && nf_ct_tuple_dst_equal(t1, t2); +} + +static inline int nf_ct_tuple_mask_cmp(const struct nf_conntrack_tuple *t, + const struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *mask) +{ + int count = 0; + + for (count = 0; count < NF_CT_TUPLE_L3SIZE; count++){ + if ((t->src.u3.all[count] ^ tuple->src.u3.all[count]) & + mask->src.u3.all[count]) + return 0; + } + + for (count = 0; count < NF_CT_TUPLE_L3SIZE; count++){ + if ((t->dst.u3.all[count] ^ tuple->dst.u3.all[count]) & + mask->dst.u3.all[count]) + return 0; + } + + if ((t->src.u.all ^ tuple->src.u.all) & mask->src.u.all || + (t->dst.u.all ^ tuple->dst.u.all) & mask->dst.u.all || + (t->src.l3num ^ tuple->src.l3num) & mask->src.l3num || + (t->dst.protonum ^ tuple->dst.protonum) & mask->dst.protonum) + return 0; + + return 1; +} + +#endif /* _NF_CONNTRACK_TUPLE_H */ diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 95501e40100e..b7d13a4fff48 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -336,6 +336,9 @@ void __kfree_skb(struct sk_buff *skb) } #ifdef CONFIG_NETFILTER nf_conntrack_put(skb->nfct); +#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) + nf_conntrack_put_reasm(skb->nfct_reasm); +#endif #ifdef CONFIG_BRIDGE_NETFILTER nf_bridge_put(skb->nf_bridge); #endif @@ -414,9 +417,17 @@ struct sk_buff *skb_clone(struct sk_buff *skb, gfp_t gfp_mask) C(nfct); nf_conntrack_get(skb->nfct); C(nfctinfo); +#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) + C(nfct_reasm); + nf_conntrack_get_reasm(skb->nfct_reasm); +#endif #if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE) C(ipvs_property); #endif +#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) + C(nfct_reasm); + nf_conntrack_get_reasm(skb->nfct_reasm); +#endif #ifdef CONFIG_BRIDGE_NETFILTER C(nf_bridge); nf_bridge_get(skb->nf_bridge); @@ -474,6 +485,10 @@ static void copy_skb_header(struct sk_buff *new, const struct sk_buff *old) new->nfct = old->nfct; nf_conntrack_get(old->nfct); new->nfctinfo = old->nfctinfo; +#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) + new->nfct_reasm = old->nfct_reasm; + nf_conntrack_get_reasm(old->nfct_reasm); +#endif #if defined(CONFIG_IP_VS) || defined(CONFIG_IP_VS_MODULE) new->ipvs_property = old->ipvs_property; #endif diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig index 7d917e4ce1d9..9d3c8b5f327e 100644 --- a/net/ipv4/netfilter/Kconfig +++ b/net/ipv4/netfilter/Kconfig @@ -5,6 +5,20 @@ menu "IP: Netfilter Configuration" depends on INET && NETFILTER +config NF_CONNTRACK_IPV4 + tristate "IPv4 support for new connection tracking (EXPERIMENTAL)" + depends on EXPERIMENTAL && NF_CONNTRACK + ---help--- + Connection tracking keeps a record of what packets have passed + through your machine, in order to figure out how they are related + into connections. + + This is IPv4 support on Layer 3 independent connection tracking. + Layer 3 independent connection tracking is experimental scheme + which generalize ip_conntrack to support other layer 3 protocols. + + To compile it as a module, choose M here. If unsure, say N. + # connection tracking, helpers and protocols config IP_NF_CONNTRACK tristate "Connection tracking (required for masq/NAT)" @@ -209,8 +223,8 @@ config IP_NF_MATCH_PKTTYPE tristate "Packet type match support" depends on IP_NF_IPTABLES help - Packet type matching allows you to match a packet by - its "class", eg. BROADCAST, MULTICAST, ... + Packet type matching allows you to match a packet by + its "class", eg. BROADCAST, MULTICAST, ... Typical usage: iptables -A INPUT -m pkttype --pkt-type broadcast -j LOG @@ -317,7 +331,8 @@ config IP_NF_MATCH_TCPMSS config IP_NF_MATCH_HELPER tristate "Helper match support" - depends on IP_NF_CONNTRACK && IP_NF_IPTABLES + depends on IP_NF_IPTABLES + depends on IP_NF_CONNTRACK || NF_CONNTRACK_IPV4 help Helper matching allows you to match packets in dynamic connections tracked by a conntrack-helper, ie. ip_conntrack_ftp @@ -326,7 +341,8 @@ config IP_NF_MATCH_HELPER config IP_NF_MATCH_STATE tristate "Connection state match support" - depends on IP_NF_CONNTRACK && IP_NF_IPTABLES + depends on IP_NF_IPTABLES + depends on IP_NF_CONNTRACK || NF_CONNTRACK_IPV4 help Connection state matching allows you to match packets based on their relationship to a tracked connection (ie. previous packets). This @@ -336,7 +352,8 @@ config IP_NF_MATCH_STATE config IP_NF_MATCH_CONNTRACK tristate "Connection tracking match support" - depends on IP_NF_CONNTRACK && IP_NF_IPTABLES + depends on IP_NF_IPTABLES + depends on IP_NF_CONNTRACK || NF_CONNTRACK_IPV4 help This is a general conntrack match module, a superset of the state match. @@ -422,7 +439,8 @@ config IP_NF_MATCH_COMMENT config IP_NF_MATCH_CONNMARK tristate 'Connection mark match support' - depends on IP_NF_CONNTRACK_MARK && IP_NF_IPTABLES + depends on IP_NF_IPTABLES + depends on IP_NF_CONNTRACK_MARK || (NF_CONNTRACK_MARK && NF_CONNTRACK_IPV4) help This option adds a `connmark' match, which allows you to match the connection mark value previously set for the session by `CONNMARK'. @@ -433,7 +451,8 @@ config IP_NF_MATCH_CONNMARK config IP_NF_MATCH_CONNBYTES tristate 'Connection byte/packet counter match support' - depends on IP_NF_CT_ACCT && IP_NF_IPTABLES + depends on IP_NF_IPTABLES + depends on IP_NF_CT_ACCT || (NF_CT_ACCT && NF_CONNTRACK_IPV4) help This option adds a `connbytes' match, which allows you to match the number of bytes and/or packets for each direction within a connection. @@ -747,7 +766,8 @@ config IP_NF_TARGET_TTL config IP_NF_TARGET_CONNMARK tristate 'CONNMARK target support' - depends on IP_NF_CONNTRACK_MARK && IP_NF_MANGLE + depends on IP_NF_MANGLE + depends on IP_NF_CONNTRACK_MARK || (NF_CONNTRACK_MARK && NF_CONNTRACK_IPV4) help This option adds a `CONNMARK' target, which allows one to manipulate the connection mark value. Similar to the MARK target, but @@ -759,7 +779,8 @@ config IP_NF_TARGET_CONNMARK config IP_NF_TARGET_CLUSTERIP tristate "CLUSTERIP target support (EXPERIMENTAL)" - depends on IP_NF_CONNTRACK_MARK && IP_NF_IPTABLES && EXPERIMENTAL + depends on IP_NF_IPTABLES && EXPERIMENTAL + depends on IP_NF_CONNTRACK_MARK || (NF_CONNTRACK_MARK && NF_CONNTRACK_IPV4) help The CLUSTERIP target allows you to build load-balancing clusters of network servers without having a dedicated load-balancing @@ -782,7 +803,7 @@ config IP_NF_RAW config IP_NF_TARGET_NOTRACK tristate 'NOTRACK target support' depends on IP_NF_RAW - depends on IP_NF_CONNTRACK + depends on IP_NF_CONNTRACK || NF_CONNTRACK_IPV4 help The NOTRACK target allows a select rule to specify which packets *not* to enter the conntrack/NAT diff --git a/net/ipv4/netfilter/Makefile b/net/ipv4/netfilter/Makefile index dab4b58dd31e..058c48e258fc 100644 --- a/net/ipv4/netfilter/Makefile +++ b/net/ipv4/netfilter/Makefile @@ -103,3 +103,9 @@ obj-$(CONFIG_IP_NF_ARP_MANGLE) += arpt_mangle.o obj-$(CONFIG_IP_NF_ARPFILTER) += arptable_filter.o obj-$(CONFIG_IP_NF_QUEUE) += ip_queue.o + +# objects for l3 independent conntrack +nf_conntrack_ipv4-objs := nf_conntrack_l3proto_ipv4.o nf_conntrack_proto_icmp.o + +# l3 independent conntrack +obj-$(CONFIG_NF_CONNTRACK_IPV4) += nf_conntrack_ipv4.o diff --git a/net/ipv4/netfilter/ip_conntrack_netlink.c b/net/ipv4/netfilter/ip_conntrack_netlink.c index 5c1c0a3d1c4b..d2a4fec22862 100644 --- a/net/ipv4/netfilter/ip_conntrack_netlink.c +++ b/net/ipv4/netfilter/ip_conntrack_netlink.c @@ -1376,7 +1376,7 @@ ctnetlink_del_expect(struct sock *ctnl, struct sk_buff *skb, ip_conntrack_expect_put(exp); } } - write_unlock(&ip_conntrack_lock); + write_unlock_bh(&ip_conntrack_lock); } else { /* This basically means we have to flush everything*/ write_lock_bh(&ip_conntrack_lock); diff --git a/net/ipv4/netfilter/ipt_CLUSTERIP.c b/net/ipv4/netfilter/ipt_CLUSTERIP.c index 9bcb398fbc1f..45c52d8f4d99 100644 --- a/net/ipv4/netfilter/ipt_CLUSTERIP.c +++ b/net/ipv4/netfilter/ipt_CLUSTERIP.c @@ -29,7 +29,7 @@ #include #include -#include +#include #define CLUSTERIP_VERSION "0.8" @@ -316,14 +316,14 @@ target(struct sk_buff **pskb, { const struct ipt_clusterip_tgt_info *cipinfo = targinfo; enum ip_conntrack_info ctinfo; - struct ip_conntrack *ct = ip_conntrack_get((*pskb), &ctinfo); - u_int32_t hash; + u_int32_t *mark, hash; /* don't need to clusterip_config_get() here, since refcount * is only decremented by destroy() - and ip_tables guarantees * that the ->target() function isn't called after ->destroy() */ - if (!ct) { + mark = nf_ct_get_mark((*pskb), &ctinfo); + if (mark == NULL) { printk(KERN_ERR "CLUSTERIP: no conntrack!\n"); /* FIXME: need to drop invalid ones, since replies * to outgoing connections of other nodes will be @@ -346,7 +346,7 @@ target(struct sk_buff **pskb, switch (ctinfo) { case IP_CT_NEW: - ct->mark = hash; + *mark = hash; break; case IP_CT_RELATED: case IP_CT_RELATED+IP_CT_IS_REPLY: @@ -363,7 +363,7 @@ target(struct sk_buff **pskb, #ifdef DEBUG_CLUSTERP DUMP_TUPLE(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple); #endif - DEBUGP("hash=%u ct_hash=%u ", hash, ct->mark); + DEBUGP("hash=%u ct_hash=%u ", hash, *mark); if (!clusterip_responsible(cipinfo->config, hash)) { DEBUGP("not responsible\n"); return NF_DROP; diff --git a/net/ipv4/netfilter/ipt_CONNMARK.c b/net/ipv4/netfilter/ipt_CONNMARK.c index 05d66ab59424..8acac5a40a92 100644 --- a/net/ipv4/netfilter/ipt_CONNMARK.c +++ b/net/ipv4/netfilter/ipt_CONNMARK.c @@ -29,7 +29,7 @@ MODULE_LICENSE("GPL"); #include #include -#include +#include static unsigned int target(struct sk_buff **pskb, @@ -43,24 +43,24 @@ target(struct sk_buff **pskb, u_int32_t diff; u_int32_t nfmark; u_int32_t newmark; + u_int32_t ctinfo; + u_int32_t *ctmark = nf_ct_get_mark(*pskb, &ctinfo); - enum ip_conntrack_info ctinfo; - struct ip_conntrack *ct = ip_conntrack_get((*pskb), &ctinfo); - if (ct) { + if (ctmark) { switch(markinfo->mode) { case IPT_CONNMARK_SET: - newmark = (ct->mark & ~markinfo->mask) | markinfo->mark; - if (newmark != ct->mark) - ct->mark = newmark; + newmark = (*ctmark & ~markinfo->mask) | markinfo->mark; + if (newmark != *ctmark) + *ctmark = newmark; break; case IPT_CONNMARK_SAVE: - newmark = (ct->mark & ~markinfo->mask) | ((*pskb)->nfmark & markinfo->mask); - if (ct->mark != newmark) - ct->mark = newmark; + newmark = (*ctmark & ~markinfo->mask) | ((*pskb)->nfmark & markinfo->mask); + if (*ctmark != newmark) + *ctmark = newmark; break; case IPT_CONNMARK_RESTORE: nfmark = (*pskb)->nfmark; - diff = (ct->mark ^ nfmark) & markinfo->mask; + diff = (*ctmark ^ nfmark) & markinfo->mask; if (diff != 0) (*pskb)->nfmark = nfmark ^ diff; break; diff --git a/net/ipv4/netfilter/ipt_NOTRACK.c b/net/ipv4/netfilter/ipt_NOTRACK.c index a4bb9b3bc292..e3c69d072c6e 100644 --- a/net/ipv4/netfilter/ipt_NOTRACK.c +++ b/net/ipv4/netfilter/ipt_NOTRACK.c @@ -5,7 +5,7 @@ #include #include -#include +#include static unsigned int target(struct sk_buff **pskb, @@ -23,7 +23,7 @@ target(struct sk_buff **pskb, If there is a real ct entry correspondig to this packet, it'll hang aroun till timing out. We don't deal with it for performance reasons. JK */ - (*pskb)->nfct = &ip_conntrack_untracked.ct_general; + nf_ct_untrack(*pskb); (*pskb)->nfctinfo = IP_CT_NEW; nf_conntrack_get((*pskb)->nfct); diff --git a/net/ipv4/netfilter/ipt_connbytes.c b/net/ipv4/netfilter/ipt_connbytes.c index df4a42c6da22..d68a048b7176 100644 --- a/net/ipv4/netfilter/ipt_connbytes.c +++ b/net/ipv4/netfilter/ipt_connbytes.c @@ -10,7 +10,7 @@ */ #include #include -#include +#include #include #include @@ -46,60 +46,59 @@ match(const struct sk_buff *skb, int *hotdrop) { const struct ipt_connbytes_info *sinfo = matchinfo; - enum ip_conntrack_info ctinfo; - struct ip_conntrack *ct; u_int64_t what = 0; /* initialize to make gcc happy */ + const struct ip_conntrack_counter *counters; - if (!(ct = ip_conntrack_get((struct sk_buff *)skb, &ctinfo))) + if (!(counters = nf_ct_get_counters(skb))) return 0; /* no match */ switch (sinfo->what) { case IPT_CONNBYTES_PKTS: switch (sinfo->direction) { case IPT_CONNBYTES_DIR_ORIGINAL: - what = ct->counters[IP_CT_DIR_ORIGINAL].packets; + what = counters[IP_CT_DIR_ORIGINAL].packets; break; case IPT_CONNBYTES_DIR_REPLY: - what = ct->counters[IP_CT_DIR_REPLY].packets; + what = counters[IP_CT_DIR_REPLY].packets; break; case IPT_CONNBYTES_DIR_BOTH: - what = ct->counters[IP_CT_DIR_ORIGINAL].packets; - what += ct->counters[IP_CT_DIR_REPLY].packets; + what = counters[IP_CT_DIR_ORIGINAL].packets; + what += counters[IP_CT_DIR_REPLY].packets; break; } break; case IPT_CONNBYTES_BYTES: switch (sinfo->direction) { case IPT_CONNBYTES_DIR_ORIGINAL: - what = ct->counters[IP_CT_DIR_ORIGINAL].bytes; + what = counters[IP_CT_DIR_ORIGINAL].bytes; break; case IPT_CONNBYTES_DIR_REPLY: - what = ct->counters[IP_CT_DIR_REPLY].bytes; + what = counters[IP_CT_DIR_REPLY].bytes; break; case IPT_CONNBYTES_DIR_BOTH: - what = ct->counters[IP_CT_DIR_ORIGINAL].bytes; - what += ct->counters[IP_CT_DIR_REPLY].bytes; + what = counters[IP_CT_DIR_ORIGINAL].bytes; + what += counters[IP_CT_DIR_REPLY].bytes; break; } break; case IPT_CONNBYTES_AVGPKT: switch (sinfo->direction) { case IPT_CONNBYTES_DIR_ORIGINAL: - what = div64_64(ct->counters[IP_CT_DIR_ORIGINAL].bytes, - ct->counters[IP_CT_DIR_ORIGINAL].packets); + what = div64_64(counters[IP_CT_DIR_ORIGINAL].bytes, + counters[IP_CT_DIR_ORIGINAL].packets); break; case IPT_CONNBYTES_DIR_REPLY: - what = div64_64(ct->counters[IP_CT_DIR_REPLY].bytes, - ct->counters[IP_CT_DIR_REPLY].packets); + what = div64_64(counters[IP_CT_DIR_REPLY].bytes, + counters[IP_CT_DIR_REPLY].packets); break; case IPT_CONNBYTES_DIR_BOTH: { u_int64_t bytes; u_int64_t pkts; - bytes = ct->counters[IP_CT_DIR_ORIGINAL].bytes + - ct->counters[IP_CT_DIR_REPLY].bytes; - pkts = ct->counters[IP_CT_DIR_ORIGINAL].packets+ - ct->counters[IP_CT_DIR_REPLY].packets; + bytes = counters[IP_CT_DIR_ORIGINAL].bytes + + counters[IP_CT_DIR_REPLY].bytes; + pkts = counters[IP_CT_DIR_ORIGINAL].packets+ + counters[IP_CT_DIR_REPLY].packets; /* FIXME_THEORETICAL: what to do if sum * overflows ? */ diff --git a/net/ipv4/netfilter/ipt_connmark.c b/net/ipv4/netfilter/ipt_connmark.c index bf8de47ce004..5306ef293b92 100644 --- a/net/ipv4/netfilter/ipt_connmark.c +++ b/net/ipv4/netfilter/ipt_connmark.c @@ -28,7 +28,7 @@ MODULE_LICENSE("GPL"); #include #include -#include +#include static int match(const struct sk_buff *skb, @@ -39,12 +39,12 @@ match(const struct sk_buff *skb, int *hotdrop) { const struct ipt_connmark_info *info = matchinfo; - enum ip_conntrack_info ctinfo; - struct ip_conntrack *ct = ip_conntrack_get((struct sk_buff *)skb, &ctinfo); - if (!ct) + u_int32_t ctinfo; + const u_int32_t *ctmark = nf_ct_get_mark(skb, &ctinfo); + if (!ctmark) return 0; - return ((ct->mark & info->mask) == info->mark) ^ info->invert; + return (((*ctmark) & info->mask) == info->mark) ^ info->invert; } static int diff --git a/net/ipv4/netfilter/ipt_conntrack.c b/net/ipv4/netfilter/ipt_conntrack.c index c1d22801b7cf..c8d18705469b 100644 --- a/net/ipv4/netfilter/ipt_conntrack.c +++ b/net/ipv4/netfilter/ipt_conntrack.c @@ -10,7 +10,14 @@ #include #include + +#if defined(CONFIG_IP_NF_CONNTRACK) || defined(CONFIG_IP_NF_CONNTRACK_MODULE) #include +#include +#else +#include +#endif + #include #include @@ -18,6 +25,8 @@ MODULE_LICENSE("GPL"); MODULE_AUTHOR("Marc Boucher "); MODULE_DESCRIPTION("iptables connection tracking match module"); +#if defined(CONFIG_IP_NF_CONNTRACK) || defined(CONFIG_IP_NF_CONNTRACK_MODULE) + static int match(const struct sk_buff *skb, const struct net_device *in, @@ -102,6 +111,93 @@ match(const struct sk_buff *skb, return 1; } +#else /* CONFIG_IP_NF_CONNTRACK */ +static int +match(const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + const void *matchinfo, + int offset, + int *hotdrop) +{ + const struct ipt_conntrack_info *sinfo = matchinfo; + struct nf_conn *ct; + enum ip_conntrack_info ctinfo; + unsigned int statebit; + + ct = nf_ct_get((struct sk_buff *)skb, &ctinfo); + +#define FWINV(bool,invflg) ((bool) ^ !!(sinfo->invflags & invflg)) + + if (ct == &nf_conntrack_untracked) + statebit = IPT_CONNTRACK_STATE_UNTRACKED; + else if (ct) + statebit = IPT_CONNTRACK_STATE_BIT(ctinfo); + else + statebit = IPT_CONNTRACK_STATE_INVALID; + + if(sinfo->flags & IPT_CONNTRACK_STATE) { + if (ct) { + if(ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.u3.ip != + ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u3.ip) + statebit |= IPT_CONNTRACK_STATE_SNAT; + + if(ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.u3.ip != + ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.u3.ip) + statebit |= IPT_CONNTRACK_STATE_DNAT; + } + + if (FWINV((statebit & sinfo->statemask) == 0, IPT_CONNTRACK_STATE)) + return 0; + } + + if(sinfo->flags & IPT_CONNTRACK_PROTO) { + if (!ct || FWINV(ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum != sinfo->tuple[IP_CT_DIR_ORIGINAL].dst.protonum, IPT_CONNTRACK_PROTO)) + return 0; + } + + if(sinfo->flags & IPT_CONNTRACK_ORIGSRC) { + if (!ct || FWINV((ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.u3.ip&sinfo->sipmsk[IP_CT_DIR_ORIGINAL].s_addr) != sinfo->tuple[IP_CT_DIR_ORIGINAL].src.ip, IPT_CONNTRACK_ORIGSRC)) + return 0; + } + + if(sinfo->flags & IPT_CONNTRACK_ORIGDST) { + if (!ct || FWINV((ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.u3.ip&sinfo->dipmsk[IP_CT_DIR_ORIGINAL].s_addr) != sinfo->tuple[IP_CT_DIR_ORIGINAL].dst.ip, IPT_CONNTRACK_ORIGDST)) + return 0; + } + + if(sinfo->flags & IPT_CONNTRACK_REPLSRC) { + if (!ct || FWINV((ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.u3.ip&sinfo->sipmsk[IP_CT_DIR_REPLY].s_addr) != sinfo->tuple[IP_CT_DIR_REPLY].src.ip, IPT_CONNTRACK_REPLSRC)) + return 0; + } + + if(sinfo->flags & IPT_CONNTRACK_REPLDST) { + if (!ct || FWINV((ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.u3.ip&sinfo->dipmsk[IP_CT_DIR_REPLY].s_addr) != sinfo->tuple[IP_CT_DIR_REPLY].dst.ip, IPT_CONNTRACK_REPLDST)) + return 0; + } + + if(sinfo->flags & IPT_CONNTRACK_STATUS) { + if (!ct || FWINV((ct->status & sinfo->statusmask) == 0, IPT_CONNTRACK_STATUS)) + return 0; + } + + if(sinfo->flags & IPT_CONNTRACK_EXPIRES) { + unsigned long expires; + + if(!ct) + return 0; + + expires = timer_pending(&ct->timeout) ? (ct->timeout.expires - jiffies)/HZ : 0; + + if (FWINV(!(expires >= sinfo->expires_min && expires <= sinfo->expires_max), IPT_CONNTRACK_EXPIRES)) + return 0; + } + + return 1; +} + +#endif /* CONFIG_NF_IP_CONNTRACK */ + static int check(const char *tablename, const struct ipt_ip *ip, void *matchinfo, diff --git a/net/ipv4/netfilter/ipt_helper.c b/net/ipv4/netfilter/ipt_helper.c index 3e7dd014de43..bf14e1c7798a 100644 --- a/net/ipv4/netfilter/ipt_helper.c +++ b/net/ipv4/netfilter/ipt_helper.c @@ -13,9 +13,15 @@ #include #include #include +#if defined(CONFIG_IP_NF_CONNTRACK) || defined(CONFIG_IP_NF_CONNTRACK_MODULE) #include #include #include +#else +#include +#include +#include +#endif #include #include @@ -29,6 +35,7 @@ MODULE_DESCRIPTION("iptables helper match module"); #define DEBUGP(format, args...) #endif +#if defined(CONFIG_IP_NF_CONNTRACK) || defined(CONFIG_IP_NF_CONNTRACK_MODULE) static int match(const struct sk_buff *skb, const struct net_device *in, @@ -73,6 +80,53 @@ out_unlock: return ret; } +#else /* CONFIG_IP_NF_CONNTRACK */ + +static int +match(const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + const void *matchinfo, + int offset, + int *hotdrop) +{ + const struct ipt_helper_info *info = matchinfo; + struct nf_conn *ct; + enum ip_conntrack_info ctinfo; + int ret = info->invert; + + ct = nf_ct_get((struct sk_buff *)skb, &ctinfo); + if (!ct) { + DEBUGP("ipt_helper: Eek! invalid conntrack?\n"); + return ret; + } + + if (!ct->master) { + DEBUGP("ipt_helper: conntrack %p has no master\n", ct); + return ret; + } + + read_lock_bh(&nf_conntrack_lock); + if (!ct->master->helper) { + DEBUGP("ipt_helper: master ct %p has no helper\n", + exp->expectant); + goto out_unlock; + } + + DEBUGP("master's name = %s , info->name = %s\n", + ct->master->helper->name, info->name); + + if (info->name[0] == '\0') + ret ^= 1; + else + ret ^= !strncmp(ct->master->helper->name, info->name, + strlen(ct->master->helper->name)); +out_unlock: + read_unlock_bh(&nf_conntrack_lock); + return ret; +} +#endif + static int check(const char *tablename, const struct ipt_ip *ip, void *matchinfo, diff --git a/net/ipv4/netfilter/ipt_state.c b/net/ipv4/netfilter/ipt_state.c index b1511b97ea5f..4d7f16b70cec 100644 --- a/net/ipv4/netfilter/ipt_state.c +++ b/net/ipv4/netfilter/ipt_state.c @@ -10,7 +10,7 @@ #include #include -#include +#include #include #include @@ -30,9 +30,9 @@ match(const struct sk_buff *skb, enum ip_conntrack_info ctinfo; unsigned int statebit; - if (skb->nfct == &ip_conntrack_untracked.ct_general) + if (nf_ct_is_untracked(skb)) statebit = IPT_STATE_UNTRACKED; - else if (!ip_conntrack_get(skb, &ctinfo)) + else if (!nf_ct_get_ctinfo(skb, &ctinfo)) statebit = IPT_STATE_INVALID; else statebit = IPT_STATE_BIT(ctinfo); diff --git a/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c b/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c new file mode 100644 index 000000000000..8202c1c0afad --- /dev/null +++ b/net/ipv4/netfilter/nf_conntrack_l3proto_ipv4.c @@ -0,0 +1,571 @@ +/* (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - move L3 protocol dependent part to this file. + * 23 Mar 2004: Yasuyuki Kozakai @USAGI + * - add get_features() to support various size of conntrack + * structures. + * + * Derived from net/ipv4/netfilter/ip_conntrack_standalone.c + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include +#include + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +DECLARE_PER_CPU(struct nf_conntrack_stat, nf_conntrack_stat); + +static int ipv4_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff, + struct nf_conntrack_tuple *tuple) +{ + u_int32_t _addrs[2], *ap; + ap = skb_header_pointer(skb, nhoff + offsetof(struct iphdr, saddr), + sizeof(u_int32_t) * 2, _addrs); + if (ap == NULL) + return 0; + + tuple->src.u3.ip = ap[0]; + tuple->dst.u3.ip = ap[1]; + + return 1; +} + +static int ipv4_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + tuple->src.u3.ip = orig->dst.u3.ip; + tuple->dst.u3.ip = orig->src.u3.ip; + + return 1; +} + +static int ipv4_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + return seq_printf(s, "src=%u.%u.%u.%u dst=%u.%u.%u.%u ", + NIPQUAD(tuple->src.u3.ip), + NIPQUAD(tuple->dst.u3.ip)); +} + +static int ipv4_print_conntrack(struct seq_file *s, + const struct nf_conn *conntrack) +{ + return 0; +} + +/* Returns new sk_buff, or NULL */ +static struct sk_buff * +nf_ct_ipv4_gather_frags(struct sk_buff *skb, u_int32_t user) +{ + skb_orphan(skb); + + local_bh_disable(); + skb = ip_defrag(skb, user); + local_bh_enable(); + + if (skb) + ip_send_check(skb->nh.iph); + + return skb; +} + +static int +ipv4_prepare(struct sk_buff **pskb, unsigned int hooknum, unsigned int *dataoff, + u_int8_t *protonum) +{ + /* Never happen */ + if ((*pskb)->nh.iph->frag_off & htons(IP_OFFSET)) { + if (net_ratelimit()) { + printk(KERN_ERR "ipv4_prepare: Frag of proto %u (hook=%u)\n", + (*pskb)->nh.iph->protocol, hooknum); + } + return -NF_DROP; + } + + *dataoff = (*pskb)->nh.raw - (*pskb)->data + (*pskb)->nh.iph->ihl*4; + *protonum = (*pskb)->nh.iph->protocol; + + return NF_ACCEPT; +} + +int nat_module_is_loaded = 0; +static u_int32_t ipv4_get_features(const struct nf_conntrack_tuple *tuple) +{ + if (nat_module_is_loaded) + return NF_CT_F_NAT; + + return NF_CT_F_BASIC; +} + +static unsigned int ipv4_confirm(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + /* We've seen it coming out the other side: confirm it */ + return nf_conntrack_confirm(pskb); +} + +static unsigned int ipv4_conntrack_help(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + struct nf_conn *ct; + enum ip_conntrack_info ctinfo; + + /* This is where we call the helper: as the packet goes out. */ + ct = nf_ct_get(*pskb, &ctinfo); + if (ct && ct->helper) { + unsigned int ret; + ret = ct->helper->help(pskb, + (*pskb)->nh.raw - (*pskb)->data + + (*pskb)->nh.iph->ihl*4, + ct, ctinfo); + if (ret != NF_ACCEPT) + return ret; + } + return NF_ACCEPT; +} + +static unsigned int ipv4_conntrack_defrag(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ +#if !defined(CONFIG_IP_NF_NAT) && !defined(CONFIG_IP_NF_NAT_MODULE) + /* Previously seen (loopback)? Ignore. Do this before + fragment check. */ + if ((*pskb)->nfct) + return NF_ACCEPT; +#endif + + /* Gather fragments. */ + if ((*pskb)->nh.iph->frag_off & htons(IP_MF|IP_OFFSET)) { + *pskb = nf_ct_ipv4_gather_frags(*pskb, + hooknum == NF_IP_PRE_ROUTING ? + IP_DEFRAG_CONNTRACK_IN : + IP_DEFRAG_CONNTRACK_OUT); + if (!*pskb) + return NF_STOLEN; + } + return NF_ACCEPT; +} + +static unsigned int ipv4_refrag(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + struct rtable *rt = (struct rtable *)(*pskb)->dst; + + /* We've seen it coming out the other side: confirm */ + if (ipv4_confirm(hooknum, pskb, in, out, okfn) != NF_ACCEPT) + return NF_DROP; + + /* Local packets are never produced too large for their + interface. We degfragment them at LOCAL_OUT, however, + so we have to refragment them here. */ + if ((*pskb)->len > dst_mtu(&rt->u.dst) && + !skb_shinfo(*pskb)->tso_size) { + /* No hook can be after us, so this should be OK. */ + ip_fragment(*pskb, okfn); + return NF_STOLEN; + } + return NF_ACCEPT; +} + +static unsigned int ipv4_conntrack_in(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + return nf_conntrack_in(PF_INET, hooknum, pskb); +} + +static unsigned int ipv4_conntrack_local(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + /* root is playing with raw sockets. */ + if ((*pskb)->len < sizeof(struct iphdr) + || (*pskb)->nh.iph->ihl * 4 < sizeof(struct iphdr)) { + if (net_ratelimit()) + printk("ipt_hook: happy cracking.\n"); + return NF_ACCEPT; + } + return nf_conntrack_in(PF_INET, hooknum, pskb); +} + +/* Connection tracking may drop packets, but never alters them, so + make it the first hook. */ +static struct nf_hook_ops ipv4_conntrack_defrag_ops = { + .hook = ipv4_conntrack_defrag, + .owner = THIS_MODULE, + .pf = PF_INET, + .hooknum = NF_IP_PRE_ROUTING, + .priority = NF_IP_PRI_CONNTRACK_DEFRAG, +}; + +static struct nf_hook_ops ipv4_conntrack_in_ops = { + .hook = ipv4_conntrack_in, + .owner = THIS_MODULE, + .pf = PF_INET, + .hooknum = NF_IP_PRE_ROUTING, + .priority = NF_IP_PRI_CONNTRACK, +}; + +static struct nf_hook_ops ipv4_conntrack_defrag_local_out_ops = { + .hook = ipv4_conntrack_defrag, + .owner = THIS_MODULE, + .pf = PF_INET, + .hooknum = NF_IP_LOCAL_OUT, + .priority = NF_IP_PRI_CONNTRACK_DEFRAG, +}; + +static struct nf_hook_ops ipv4_conntrack_local_out_ops = { + .hook = ipv4_conntrack_local, + .owner = THIS_MODULE, + .pf = PF_INET, + .hooknum = NF_IP_LOCAL_OUT, + .priority = NF_IP_PRI_CONNTRACK, +}; + +/* helpers */ +static struct nf_hook_ops ipv4_conntrack_helper_out_ops = { + .hook = ipv4_conntrack_help, + .owner = THIS_MODULE, + .pf = PF_INET, + .hooknum = NF_IP_POST_ROUTING, + .priority = NF_IP_PRI_CONNTRACK_HELPER, +}; + +static struct nf_hook_ops ipv4_conntrack_helper_in_ops = { + .hook = ipv4_conntrack_help, + .owner = THIS_MODULE, + .pf = PF_INET, + .hooknum = NF_IP_LOCAL_IN, + .priority = NF_IP_PRI_CONNTRACK_HELPER, +}; + + +/* Refragmenter; last chance. */ +static struct nf_hook_ops ipv4_conntrack_out_ops = { + .hook = ipv4_refrag, + .owner = THIS_MODULE, + .pf = PF_INET, + .hooknum = NF_IP_POST_ROUTING, + .priority = NF_IP_PRI_CONNTRACK_CONFIRM, +}; + +static struct nf_hook_ops ipv4_conntrack_local_in_ops = { + .hook = ipv4_confirm, + .owner = THIS_MODULE, + .pf = PF_INET, + .hooknum = NF_IP_LOCAL_IN, + .priority = NF_IP_PRI_CONNTRACK_CONFIRM, +}; + +#ifdef CONFIG_SYSCTL +/* From nf_conntrack_proto_icmp.c */ +extern unsigned long nf_ct_icmp_timeout; +static struct ctl_table_header *nf_ct_ipv4_sysctl_header; + +static ctl_table nf_ct_sysctl_table[] = { + { + .ctl_name = NET_NF_CONNTRACK_ICMP_TIMEOUT, + .procname = "nf_conntrack_icmp_timeout", + .data = &nf_ct_icmp_timeout, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { .ctl_name = 0 } +}; + +static ctl_table nf_ct_netfilter_table[] = { + { + .ctl_name = NET_NETFILTER, + .procname = "netfilter", + .mode = 0555, + .child = nf_ct_sysctl_table, + }, + { .ctl_name = 0 } +}; + +static ctl_table nf_ct_net_table[] = { + { + .ctl_name = CTL_NET, + .procname = "net", + .mode = 0555, + .child = nf_ct_netfilter_table, + }, + { .ctl_name = 0 } +}; +#endif + +/* Fast function for those who don't want to parse /proc (and I don't + blame them). */ +/* Reversing the socket's dst/src point of view gives us the reply + mapping. */ +static int +getorigdst(struct sock *sk, int optval, void __user *user, int *len) +{ + struct inet_sock *inet = inet_sk(sk); + struct nf_conntrack_tuple_hash *h; + struct nf_conntrack_tuple tuple; + + NF_CT_TUPLE_U_BLANK(&tuple); + tuple.src.u3.ip = inet->rcv_saddr; + tuple.src.u.tcp.port = inet->sport; + tuple.dst.u3.ip = inet->daddr; + tuple.dst.u.tcp.port = inet->dport; + tuple.src.l3num = PF_INET; + tuple.dst.protonum = IPPROTO_TCP; + + /* We only do TCP at the moment: is there a better way? */ + if (strcmp(sk->sk_prot->name, "TCP")) { + DEBUGP("SO_ORIGINAL_DST: Not a TCP socket\n"); + return -ENOPROTOOPT; + } + + if ((unsigned int) *len < sizeof(struct sockaddr_in)) { + DEBUGP("SO_ORIGINAL_DST: len %u not %u\n", + *len, sizeof(struct sockaddr_in)); + return -EINVAL; + } + + h = nf_conntrack_find_get(&tuple, NULL); + if (h) { + struct sockaddr_in sin; + struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h); + + sin.sin_family = AF_INET; + sin.sin_port = ct->tuplehash[IP_CT_DIR_ORIGINAL] + .tuple.dst.u.tcp.port; + sin.sin_addr.s_addr = ct->tuplehash[IP_CT_DIR_ORIGINAL] + .tuple.dst.u3.ip; + + DEBUGP("SO_ORIGINAL_DST: %u.%u.%u.%u %u\n", + NIPQUAD(sin.sin_addr.s_addr), ntohs(sin.sin_port)); + nf_ct_put(ct); + if (copy_to_user(user, &sin, sizeof(sin)) != 0) + return -EFAULT; + else + return 0; + } + DEBUGP("SO_ORIGINAL_DST: Can't find %u.%u.%u.%u/%u-%u.%u.%u.%u/%u.\n", + NIPQUAD(tuple.src.u3.ip), ntohs(tuple.src.u.tcp.port), + NIPQUAD(tuple.dst.u3.ip), ntohs(tuple.dst.u.tcp.port)); + return -ENOENT; +} + +static struct nf_sockopt_ops so_getorigdst = { + .pf = PF_INET, + .get_optmin = SO_ORIGINAL_DST, + .get_optmax = SO_ORIGINAL_DST+1, + .get = &getorigdst, +}; + +struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4 = { + .l3proto = PF_INET, + .name = "ipv4", + .pkt_to_tuple = ipv4_pkt_to_tuple, + .invert_tuple = ipv4_invert_tuple, + .print_tuple = ipv4_print_tuple, + .print_conntrack = ipv4_print_conntrack, + .prepare = ipv4_prepare, + .get_features = ipv4_get_features, + .me = THIS_MODULE, +}; + +extern struct nf_conntrack_protocol nf_conntrack_protocol_tcp4; +extern struct nf_conntrack_protocol nf_conntrack_protocol_udp4; +extern struct nf_conntrack_protocol nf_conntrack_protocol_icmp; +static int init_or_cleanup(int init) +{ + int ret = 0; + + if (!init) goto cleanup; + + ret = nf_register_sockopt(&so_getorigdst); + if (ret < 0) { + printk(KERN_ERR "Unable to register netfilter socket option\n"); + goto cleanup_nothing; + } + + ret = nf_conntrack_protocol_register(&nf_conntrack_protocol_tcp4); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register tcp.\n"); + goto cleanup_sockopt; + } + + ret = nf_conntrack_protocol_register(&nf_conntrack_protocol_udp4); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register udp.\n"); + goto cleanup_tcp; + } + + ret = nf_conntrack_protocol_register(&nf_conntrack_protocol_icmp); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register icmp.\n"); + goto cleanup_udp; + } + + ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv4); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register ipv4\n"); + goto cleanup_icmp; + } + + ret = nf_register_hook(&ipv4_conntrack_defrag_ops); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register pre-routing defrag hook.\n"); + goto cleanup_ipv4; + } + ret = nf_register_hook(&ipv4_conntrack_defrag_local_out_ops); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register local_out defrag hook.\n"); + goto cleanup_defragops; + } + + ret = nf_register_hook(&ipv4_conntrack_in_ops); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register pre-routing hook.\n"); + goto cleanup_defraglocalops; + } + + ret = nf_register_hook(&ipv4_conntrack_local_out_ops); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register local out hook.\n"); + goto cleanup_inops; + } + + ret = nf_register_hook(&ipv4_conntrack_helper_in_ops); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register local helper hook.\n"); + goto cleanup_inandlocalops; + } + + ret = nf_register_hook(&ipv4_conntrack_helper_out_ops); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register postrouting helper hook.\n"); + goto cleanup_helperinops; + } + + ret = nf_register_hook(&ipv4_conntrack_out_ops); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register post-routing hook.\n"); + goto cleanup_helperoutops; + } + + ret = nf_register_hook(&ipv4_conntrack_local_in_ops); + if (ret < 0) { + printk("nf_conntrack_ipv4: can't register local in hook.\n"); + goto cleanup_inoutandlocalops; + } + +#ifdef CONFIG_SYSCTL + nf_ct_ipv4_sysctl_header = register_sysctl_table(nf_ct_net_table, 0); + if (nf_ct_ipv4_sysctl_header == NULL) { + printk("nf_conntrack: can't register to sysctl.\n"); + ret = -ENOMEM; + goto cleanup_localinops; + } +#endif + + /* For use by REJECT target */ + ip_ct_attach = __nf_conntrack_attach; + + return ret; + + cleanup: + synchronize_net(); + ip_ct_attach = NULL; +#ifdef CONFIG_SYSCTL + unregister_sysctl_table(nf_ct_ipv4_sysctl_header); + cleanup_localinops: +#endif + nf_unregister_hook(&ipv4_conntrack_local_in_ops); + cleanup_inoutandlocalops: + nf_unregister_hook(&ipv4_conntrack_out_ops); + cleanup_helperoutops: + nf_unregister_hook(&ipv4_conntrack_helper_out_ops); + cleanup_helperinops: + nf_unregister_hook(&ipv4_conntrack_helper_in_ops); + cleanup_inandlocalops: + nf_unregister_hook(&ipv4_conntrack_local_out_ops); + cleanup_inops: + nf_unregister_hook(&ipv4_conntrack_in_ops); + cleanup_defraglocalops: + nf_unregister_hook(&ipv4_conntrack_defrag_local_out_ops); + cleanup_defragops: + nf_unregister_hook(&ipv4_conntrack_defrag_ops); + cleanup_ipv4: + nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv4); + cleanup_icmp: + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_icmp); + cleanup_udp: + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_udp4); + cleanup_tcp: + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_tcp4); + cleanup_sockopt: + nf_unregister_sockopt(&so_getorigdst); + cleanup_nothing: + return ret; +} + +MODULE_LICENSE("GPL"); + +static int __init init(void) +{ + need_nf_conntrack(); + return init_or_cleanup(1); +} + +static void __exit fini(void) +{ + init_or_cleanup(0); +} + +module_init(init); +module_exit(fini); + +void need_ip_conntrack(void) +{ +} + +EXPORT_SYMBOL(need_ip_conntrack); +EXPORT_SYMBOL(nf_ct_ipv4_gather_frags); diff --git a/net/ipv4/netfilter/nf_conntrack_proto_icmp.c b/net/ipv4/netfilter/nf_conntrack_proto_icmp.c new file mode 100644 index 000000000000..7ddb5c08f7b8 --- /dev/null +++ b/net/ipv4/netfilter/nf_conntrack_proto_icmp.c @@ -0,0 +1,301 @@ +/* (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - enable working with Layer 3 protocol independent connection tracking. + * + * Derived from net/ipv4/netfilter/ip_conntrack_proto_icmp.c + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +unsigned long nf_ct_icmp_timeout = 30*HZ; + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +static int icmp_pkt_to_tuple(const struct sk_buff *skb, + unsigned int dataoff, + struct nf_conntrack_tuple *tuple) +{ + struct icmphdr _hdr, *hp; + + hp = skb_header_pointer(skb, dataoff, sizeof(_hdr), &_hdr); + if (hp == NULL) + return 0; + + tuple->dst.u.icmp.type = hp->type; + tuple->src.u.icmp.id = hp->un.echo.id; + tuple->dst.u.icmp.code = hp->code; + + return 1; +} + +static int icmp_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + /* Add 1; spaces filled with 0. */ + static u_int8_t invmap[] + = { [ICMP_ECHO] = ICMP_ECHOREPLY + 1, + [ICMP_ECHOREPLY] = ICMP_ECHO + 1, + [ICMP_TIMESTAMP] = ICMP_TIMESTAMPREPLY + 1, + [ICMP_TIMESTAMPREPLY] = ICMP_TIMESTAMP + 1, + [ICMP_INFO_REQUEST] = ICMP_INFO_REPLY + 1, + [ICMP_INFO_REPLY] = ICMP_INFO_REQUEST + 1, + [ICMP_ADDRESS] = ICMP_ADDRESSREPLY + 1, + [ICMP_ADDRESSREPLY] = ICMP_ADDRESS + 1}; + + if (orig->dst.u.icmp.type >= sizeof(invmap) + || !invmap[orig->dst.u.icmp.type]) + return 0; + + tuple->src.u.icmp.id = orig->src.u.icmp.id; + tuple->dst.u.icmp.type = invmap[orig->dst.u.icmp.type] - 1; + tuple->dst.u.icmp.code = orig->dst.u.icmp.code; + return 1; +} + +/* Print out the per-protocol part of the tuple. */ +static int icmp_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + return seq_printf(s, "type=%u code=%u id=%u ", + tuple->dst.u.icmp.type, + tuple->dst.u.icmp.code, + ntohs(tuple->src.u.icmp.id)); +} + +/* Print out the private part of the conntrack. */ +static int icmp_print_conntrack(struct seq_file *s, + const struct nf_conn *conntrack) +{ + return 0; +} + +/* Returns verdict for packet, or -1 for invalid. */ +static int icmp_packet(struct nf_conn *ct, + const struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info ctinfo, + int pf, + unsigned int hooknum) +{ + /* Try to delete connection immediately after all replies: + won't actually vanish as we still have skb, and del_timer + means this will only run once even if count hits zero twice + (theoretically possible with SMP) */ + if (CTINFO2DIR(ctinfo) == IP_CT_DIR_REPLY) { + if (atomic_dec_and_test(&ct->proto.icmp.count) + && del_timer(&ct->timeout)) + ct->timeout.function((unsigned long)ct); + } else { + atomic_inc(&ct->proto.icmp.count); + nf_conntrack_event_cache(IPCT_PROTOINFO_VOLATILE, skb); + nf_ct_refresh_acct(ct, ctinfo, skb, nf_ct_icmp_timeout); + } + + return NF_ACCEPT; +} + +/* Called when a new connection for this protocol found. */ +static int icmp_new(struct nf_conn *conntrack, + const struct sk_buff *skb, unsigned int dataoff) +{ + static u_int8_t valid_new[] + = { [ICMP_ECHO] = 1, + [ICMP_TIMESTAMP] = 1, + [ICMP_INFO_REQUEST] = 1, + [ICMP_ADDRESS] = 1 }; + + if (conntrack->tuplehash[0].tuple.dst.u.icmp.type >= sizeof(valid_new) + || !valid_new[conntrack->tuplehash[0].tuple.dst.u.icmp.type]) { + /* Can't create a new ICMP `conn' with this. */ + DEBUGP("icmp: can't create new conn with type %u\n", + conntrack->tuplehash[0].tuple.dst.u.icmp.type); + NF_CT_DUMP_TUPLE(&conntrack->tuplehash[0].tuple); + return 0; + } + atomic_set(&conntrack->proto.icmp.count, 0); + return 1; +} + +extern struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv4; +/* Returns conntrack if it dealt with ICMP, and filled in skb fields */ +static int +icmp_error_message(struct sk_buff *skb, + enum ip_conntrack_info *ctinfo, + unsigned int hooknum) +{ + struct nf_conntrack_tuple innertuple, origtuple; + struct { + struct icmphdr icmp; + struct iphdr ip; + } _in, *inside; + struct nf_conntrack_protocol *innerproto; + struct nf_conntrack_tuple_hash *h; + int dataoff; + + NF_CT_ASSERT(skb->nfct == NULL); + + /* Not enough header? */ + inside = skb_header_pointer(skb, skb->nh.iph->ihl*4, sizeof(_in), &_in); + if (inside == NULL) + return -NF_ACCEPT; + + /* Ignore ICMP's containing fragments (shouldn't happen) */ + if (inside->ip.frag_off & htons(IP_OFFSET)) { + DEBUGP("icmp_error_message: fragment of proto %u\n", + inside->ip.protocol); + return -NF_ACCEPT; + } + + innerproto = nf_ct_find_proto(PF_INET, inside->ip.protocol); + dataoff = skb->nh.iph->ihl*4 + sizeof(inside->icmp); + /* Are they talking about one of our connections? */ + if (!nf_ct_get_tuple(skb, dataoff, dataoff + inside->ip.ihl*4, PF_INET, + inside->ip.protocol, &origtuple, + &nf_conntrack_l3proto_ipv4, innerproto)) { + DEBUGP("icmp_error_message: ! get_tuple p=%u", + inside->ip.protocol); + return -NF_ACCEPT; + } + + /* Ordinarily, we'd expect the inverted tupleproto, but it's + been preserved inside the ICMP. */ + if (!nf_ct_invert_tuple(&innertuple, &origtuple, + &nf_conntrack_l3proto_ipv4, innerproto)) { + DEBUGP("icmp_error_message: no match\n"); + return -NF_ACCEPT; + } + + *ctinfo = IP_CT_RELATED; + + h = nf_conntrack_find_get(&innertuple, NULL); + if (!h) { + /* Locally generated ICMPs will match inverted if they + haven't been SNAT'ed yet */ + /* FIXME: NAT code has to handle half-done double NAT --RR */ + if (hooknum == NF_IP_LOCAL_OUT) + h = nf_conntrack_find_get(&origtuple, NULL); + + if (!h) { + DEBUGP("icmp_error_message: no match\n"); + return -NF_ACCEPT; + } + + /* Reverse direction from that found */ + if (NF_CT_DIRECTION(h) == IP_CT_DIR_REPLY) + *ctinfo += IP_CT_IS_REPLY; + } else { + if (NF_CT_DIRECTION(h) == IP_CT_DIR_REPLY) + *ctinfo += IP_CT_IS_REPLY; + } + + /* Update skb to refer to this connection */ + skb->nfct = &nf_ct_tuplehash_to_ctrack(h)->ct_general; + skb->nfctinfo = *ctinfo; + return -NF_ACCEPT; +} + +/* Small and modified version of icmp_rcv */ +static int +icmp_error(struct sk_buff *skb, unsigned int dataoff, + enum ip_conntrack_info *ctinfo, int pf, unsigned int hooknum) +{ + struct icmphdr _ih, *icmph; + + /* Not enough header? */ + icmph = skb_header_pointer(skb, skb->nh.iph->ihl*4, sizeof(_ih), &_ih); + if (icmph == NULL) { + if (LOG_INVALID(IPPROTO_ICMP)) + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, + "nf_ct_icmp: short packet "); + return -NF_ACCEPT; + } + + /* See ip_conntrack_proto_tcp.c */ + if (hooknum != NF_IP_PRE_ROUTING) + goto checksum_skipped; + + switch (skb->ip_summed) { + case CHECKSUM_HW: + if (!(u16)csum_fold(skb->csum)) + break; + if (LOG_INVALID(IPPROTO_ICMP)) + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, + "nf_ct_icmp: bad HW ICMP checksum "); + return -NF_ACCEPT; + case CHECKSUM_NONE: + if ((u16)csum_fold(skb_checksum(skb, 0, skb->len, 0))) { + if (LOG_INVALID(IPPROTO_ICMP)) + nf_log_packet(PF_INET, 0, skb, NULL, NULL, + NULL, + "nf_ct_icmp: bad ICMP checksum "); + return -NF_ACCEPT; + } + default: + break; + } + +checksum_skipped: + /* + * 18 is the highest 'known' ICMP type. Anything else is a mystery + * + * RFC 1122: 3.2.2 Unknown ICMP messages types MUST be silently + * discarded. + */ + if (icmph->type > NR_ICMP_TYPES) { + if (LOG_INVALID(IPPROTO_ICMP)) + nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, + "nf_ct_icmp: invalid ICMP type "); + return -NF_ACCEPT; + } + + /* Need to track icmp error message? */ + if (icmph->type != ICMP_DEST_UNREACH + && icmph->type != ICMP_SOURCE_QUENCH + && icmph->type != ICMP_TIME_EXCEEDED + && icmph->type != ICMP_PARAMETERPROB + && icmph->type != ICMP_REDIRECT) + return NF_ACCEPT; + + return icmp_error_message(skb, ctinfo, hooknum); +} + +struct nf_conntrack_protocol nf_conntrack_protocol_icmp = +{ + .list = { NULL, NULL }, + .l3proto = PF_INET, + .proto = IPPROTO_ICMP, + .name = "icmp", + .pkt_to_tuple = icmp_pkt_to_tuple, + .invert_tuple = icmp_invert_tuple, + .print_tuple = icmp_print_tuple, + .print_conntrack = icmp_print_conntrack, + .packet = icmp_packet, + .new = icmp_new, + .error = icmp_error, + .destroy = NULL, + .me = NULL +}; + +EXPORT_SYMBOL(nf_conntrack_protocol_icmp); diff --git a/net/ipv6/ip6_input.c b/net/ipv6/ip6_input.c index 6e3480426939..a6026d2787d2 100644 --- a/net/ipv6/ip6_input.c +++ b/net/ipv6/ip6_input.c @@ -176,6 +176,11 @@ resubmit: if (ipprot->flags & INET6_PROTO_FINAL) { struct ipv6hdr *hdr; + /* Free reference early: we don't need it any more, + and it may hold ip_conntrack module loaded + indefinitely. */ + nf_reset(skb); + skb_postpull_rcsum(skb, skb->nh.raw, skb->h.raw - skb->nh.raw); hdr = skb->nh.ipv6h; diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index dbd9767b32e4..c1fa693511a1 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -441,9 +441,15 @@ static void ip6_copy_metadata(struct sk_buff *to, struct sk_buff *from) #ifdef CONFIG_NETFILTER to->nfmark = from->nfmark; /* Connection association is same as pre-frag packet */ + nf_conntrack_put(to->nfct); to->nfct = from->nfct; nf_conntrack_get(to->nfct); to->nfctinfo = from->nfctinfo; +#if defined(CONFIG_NF_CONNTRACK) || defined(CONFIG_NF_CONNTRACK_MODULE) + nf_conntrack_put_reasm(to->nfct_reasm); + to->nfct_reasm = from->nfct_reasm; + nf_conntrack_get_reasm(to->nfct_reasm); +#endif #ifdef CONFIG_BRIDGE_NETFILTER nf_bridge_put(to->nf_bridge); to->nf_bridge = from->nf_bridge; diff --git a/net/ipv6/netfilter/Kconfig b/net/ipv6/netfilter/Kconfig index bb7ccfe33f23..971ba60bf6e9 100644 --- a/net/ipv6/netfilter/Kconfig +++ b/net/ipv6/netfilter/Kconfig @@ -278,5 +278,19 @@ config IP6_NF_RAW If you want to compile it as a module, say M here and read . If unsure, say `N'. +config NF_CONNTRACK_IPV6 + tristate "IPv6 support for new connection tracking (EXPERIMENTAL)" + depends on EXPERIMENTAL && NF_CONNTRACK + ---help--- + Connection tracking keeps a record of what packets have passed + through your machine, in order to figure out how they are related + into connections. + + This is IPv6 support on Layer 3 independent connection tracking. + Layer 3 independent connection tracking is experimental scheme + which generalize ip_conntrack to support other layer 3 protocols. + + To compile it as a module, choose M here. If unsure, say N. + endmenu diff --git a/net/ipv6/netfilter/Makefile b/net/ipv6/netfilter/Makefile index 2b2c370e8b1c..9ab5b2ca1f59 100644 --- a/net/ipv6/netfilter/Makefile +++ b/net/ipv6/netfilter/Makefile @@ -27,3 +27,9 @@ obj-$(CONFIG_IP6_NF_TARGET_LOG) += ip6t_LOG.o obj-$(CONFIG_IP6_NF_RAW) += ip6table_raw.o obj-$(CONFIG_IP6_NF_MATCH_HL) += ip6t_hl.o obj-$(CONFIG_IP6_NF_TARGET_REJECT) += ip6t_REJECT.o + +# objects for l3 independent conntrack +nf_conntrack_ipv6-objs := nf_conntrack_l3proto_ipv6.o nf_conntrack_proto_icmpv6.o nf_conntrack_reasm.o + +# l3 independent conntrack +obj-$(CONFIG_NF_CONNTRACK_IPV6) += nf_conntrack_ipv6.o diff --git a/net/ipv6/netfilter/ip6t_MARK.c b/net/ipv6/netfilter/ip6t_MARK.c index 0c7584f92172..eab8fb864ee0 100644 --- a/net/ipv6/netfilter/ip6t_MARK.c +++ b/net/ipv6/netfilter/ip6t_MARK.c @@ -56,9 +56,9 @@ checkentry(const char *tablename, return 1; } -static struct ip6t_target ip6t_mark_reg = { - .name = "MARK", - .target = target, +static struct ip6t_target ip6t_mark_reg = { + .name = "MARK", + .target = target, .checkentry = checkentry, .me = THIS_MODULE }; diff --git a/net/ipv6/netfilter/nf_conntrack_l3proto_ipv6.c b/net/ipv6/netfilter/nf_conntrack_l3proto_ipv6.c new file mode 100644 index 000000000000..e2c90b3a8074 --- /dev/null +++ b/net/ipv6/netfilter/nf_conntrack_l3proto_ipv6.c @@ -0,0 +1,556 @@ +/* + * Copyright (C)2004 USAGI/WIDE Project + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * Author: + * Yasuyuki Kozakai @USAGI + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - support Layer 3 protocol independent connection tracking. + * Based on the original ip_conntrack code which had the following + * copyright information: + * (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * + * 23 Mar 2004: Yasuyuki Kozakai @USAGI + * - add get_features() to support various size of conntrack + * structures. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +DECLARE_PER_CPU(struct ip_conntrack_stat, nf_conntrack_stat); + +static int ipv6_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff, + struct nf_conntrack_tuple *tuple) +{ + u_int32_t _addrs[8], *ap; + + ap = skb_header_pointer(skb, nhoff + offsetof(struct ipv6hdr, saddr), + sizeof(_addrs), _addrs); + if (ap == NULL) + return 0; + + memcpy(tuple->src.u3.ip6, ap, sizeof(tuple->src.u3.ip6)); + memcpy(tuple->dst.u3.ip6, ap + 4, sizeof(tuple->dst.u3.ip6)); + + return 1; +} + +static int ipv6_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + memcpy(tuple->src.u3.ip6, orig->dst.u3.ip6, sizeof(tuple->src.u3.ip6)); + memcpy(tuple->dst.u3.ip6, orig->src.u3.ip6, sizeof(tuple->dst.u3.ip6)); + + return 1; +} + +static int ipv6_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + return seq_printf(s, "src=%x:%x:%x:%x:%x:%x:%x:%x dst=%x:%x:%x:%x:%x:%x:%x:%x ", + NIP6(*((struct in6_addr *)tuple->src.u3.ip6)), + NIP6(*((struct in6_addr *)tuple->dst.u3.ip6))); +} + +static int ipv6_print_conntrack(struct seq_file *s, + const struct nf_conn *conntrack) +{ + return 0; +} + +/* + * Based on ipv6_skip_exthdr() in net/ipv6/exthdr.c + * + * This function parses (probably truncated) exthdr set "hdr" + * of length "len". "nexthdrp" initially points to some place, + * where type of the first header can be found. + * + * It skips all well-known exthdrs, and returns pointer to the start + * of unparsable area i.e. the first header with unknown type. + * if success, *nexthdr is updated by type/protocol of this header. + * + * NOTES: - it may return pointer pointing beyond end of packet, + * if the last recognized header is truncated in the middle. + * - if packet is truncated, so that all parsed headers are skipped, + * it returns -1. + * - if packet is fragmented, return pointer of the fragment header. + * - ESP is unparsable for now and considered like + * normal payload protocol. + * - Note also special handling of AUTH header. Thanks to IPsec wizards. + */ + +int nf_ct_ipv6_skip_exthdr(struct sk_buff *skb, int start, u8 *nexthdrp, + int len) +{ + u8 nexthdr = *nexthdrp; + + while (ipv6_ext_hdr(nexthdr)) { + struct ipv6_opt_hdr hdr; + int hdrlen; + + if (len < (int)sizeof(struct ipv6_opt_hdr)) + return -1; + if (nexthdr == NEXTHDR_NONE) + break; + if (nexthdr == NEXTHDR_FRAGMENT) + break; + if (skb_copy_bits(skb, start, &hdr, sizeof(hdr))) + BUG(); + if (nexthdr == NEXTHDR_AUTH) + hdrlen = (hdr.hdrlen+2)<<2; + else + hdrlen = ipv6_optlen(&hdr); + + nexthdr = hdr.nexthdr; + len -= hdrlen; + start += hdrlen; + } + + *nexthdrp = nexthdr; + return start; +} + +static int +ipv6_prepare(struct sk_buff **pskb, unsigned int hooknum, unsigned int *dataoff, + u_int8_t *protonum) +{ + unsigned int extoff; + unsigned char pnum; + int protoff; + + extoff = (u8*)((*pskb)->nh.ipv6h + 1) - (*pskb)->data; + pnum = (*pskb)->nh.ipv6h->nexthdr; + + protoff = nf_ct_ipv6_skip_exthdr(*pskb, extoff, &pnum, + (*pskb)->len - extoff); + + /* + * (protoff == (*pskb)->len) mean that the packet doesn't have no data + * except of IPv6 & ext headers. but it's tracked anyway. - YK + */ + if ((protoff < 0) || (protoff > (*pskb)->len)) { + DEBUGP("ip6_conntrack_core: can't find proto in pkt\n"); + NF_CT_STAT_INC(error); + NF_CT_STAT_INC(invalid); + return -NF_ACCEPT; + } + + *dataoff = protoff; + *protonum = pnum; + return NF_ACCEPT; +} + +static u_int32_t ipv6_get_features(const struct nf_conntrack_tuple *tuple) +{ + return NF_CT_F_BASIC; +} + +static unsigned int ipv6_confirm(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + struct nf_conn *ct; + enum ip_conntrack_info ctinfo; + + /* This is where we call the helper: as the packet goes out. */ + ct = nf_ct_get(*pskb, &ctinfo); + if (ct && ct->helper) { + unsigned int ret, protoff; + unsigned int extoff = (u8*)((*pskb)->nh.ipv6h + 1) + - (*pskb)->data; + unsigned char pnum = (*pskb)->nh.ipv6h->nexthdr; + + protoff = nf_ct_ipv6_skip_exthdr(*pskb, extoff, &pnum, + (*pskb)->len - extoff); + if (protoff < 0 || protoff > (*pskb)->len || + pnum == NEXTHDR_FRAGMENT) { + DEBUGP("proto header not found\n"); + return NF_ACCEPT; + } + + ret = ct->helper->help(pskb, protoff, ct, ctinfo); + if (ret != NF_ACCEPT) + return ret; + } + + /* We've seen it coming out the other side: confirm it */ + + return nf_conntrack_confirm(pskb); +} + +extern struct sk_buff *nf_ct_frag6_gather(struct sk_buff *skb); +extern void nf_ct_frag6_output(unsigned int hooknum, struct sk_buff *skb, + struct net_device *in, + struct net_device *out, + int (*okfn)(struct sk_buff *)); +static unsigned int ipv6_defrag(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + struct sk_buff *reasm; + + /* Previously seen (loopback)? */ + if ((*pskb)->nfct) + return NF_ACCEPT; + + reasm = nf_ct_frag6_gather(*pskb); + + /* queued */ + if (reasm == NULL) + return NF_STOLEN; + + /* error occured or not fragmented */ + if (reasm == *pskb) + return NF_ACCEPT; + + nf_ct_frag6_output(hooknum, reasm, (struct net_device *)in, + (struct net_device *)out, okfn); + + return NF_STOLEN; +} + +static unsigned int ipv6_conntrack_in(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + struct sk_buff *reasm = (*pskb)->nfct_reasm; + + /* This packet is fragmented and has reassembled packet. */ + if (reasm) { + /* Reassembled packet isn't parsed yet ? */ + if (!reasm->nfct) { + unsigned int ret; + + ret = nf_conntrack_in(PF_INET6, hooknum, &reasm); + if (ret != NF_ACCEPT) + return ret; + } + nf_conntrack_get(reasm->nfct); + (*pskb)->nfct = reasm->nfct; + return NF_ACCEPT; + } + + return nf_conntrack_in(PF_INET6, hooknum, pskb); +} + +static unsigned int ipv6_conntrack_local(unsigned int hooknum, + struct sk_buff **pskb, + const struct net_device *in, + const struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + /* root is playing with raw sockets. */ + if ((*pskb)->len < sizeof(struct ipv6hdr)) { + if (net_ratelimit()) + printk("ipv6_conntrack_local: packet too short\n"); + return NF_ACCEPT; + } + return ipv6_conntrack_in(hooknum, pskb, in, out, okfn); +} + +/* Connection tracking may drop packets, but never alters them, so + make it the first hook. */ +static struct nf_hook_ops ipv6_conntrack_defrag_ops = { + .hook = ipv6_defrag, + .owner = THIS_MODULE, + .pf = PF_INET6, + .hooknum = NF_IP6_PRE_ROUTING, + .priority = NF_IP6_PRI_CONNTRACK_DEFRAG, +}; + +static struct nf_hook_ops ipv6_conntrack_in_ops = { + .hook = ipv6_conntrack_in, + .owner = THIS_MODULE, + .pf = PF_INET6, + .hooknum = NF_IP6_PRE_ROUTING, + .priority = NF_IP6_PRI_CONNTRACK, +}; + +static struct nf_hook_ops ipv6_conntrack_local_out_ops = { + .hook = ipv6_conntrack_local, + .owner = THIS_MODULE, + .pf = PF_INET6, + .hooknum = NF_IP6_LOCAL_OUT, + .priority = NF_IP6_PRI_CONNTRACK, +}; + +static struct nf_hook_ops ipv6_conntrack_defrag_local_out_ops = { + .hook = ipv6_defrag, + .owner = THIS_MODULE, + .pf = PF_INET6, + .hooknum = NF_IP6_LOCAL_OUT, + .priority = NF_IP6_PRI_CONNTRACK_DEFRAG, +}; + +/* Refragmenter; last chance. */ +static struct nf_hook_ops ipv6_conntrack_out_ops = { + .hook = ipv6_confirm, + .owner = THIS_MODULE, + .pf = PF_INET6, + .hooknum = NF_IP6_POST_ROUTING, + .priority = NF_IP6_PRI_LAST, +}; + +static struct nf_hook_ops ipv6_conntrack_local_in_ops = { + .hook = ipv6_confirm, + .owner = THIS_MODULE, + .pf = PF_INET6, + .hooknum = NF_IP6_LOCAL_IN, + .priority = NF_IP6_PRI_LAST-1, +}; + +#ifdef CONFIG_SYSCTL + +/* From nf_conntrack_proto_icmpv6.c */ +extern unsigned long nf_ct_icmpv6_timeout; + +/* From nf_conntrack_frag6.c */ +extern unsigned long nf_ct_frag6_timeout; +extern unsigned long nf_ct_frag6_low_thresh; +extern unsigned long nf_ct_frag6_high_thresh; + +static struct ctl_table_header *nf_ct_ipv6_sysctl_header; + +static ctl_table nf_ct_sysctl_table[] = { + { + .ctl_name = NET_NF_CONNTRACK_ICMPV6_TIMEOUT, + .procname = "nf_conntrack_icmpv6_timeout", + .data = &nf_ct_icmpv6_timeout, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_FRAG6_TIMEOUT, + .procname = "nf_conntrack_frag6_timeout", + .data = &nf_ct_frag6_timeout, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_FRAG6_LOW_THRESH, + .procname = "nf_conntrack_frag6_low_thresh", + .data = &nf_ct_frag6_low_thresh, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_FRAG6_HIGH_THRESH, + .procname = "nf_conntrack_frag6_high_thresh", + .data = &nf_ct_frag6_high_thresh, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { .ctl_name = 0 } +}; + +static ctl_table nf_ct_netfilter_table[] = { + { + .ctl_name = NET_NETFILTER, + .procname = "netfilter", + .mode = 0555, + .child = nf_ct_sysctl_table, + }, + { .ctl_name = 0 } +}; + +static ctl_table nf_ct_net_table[] = { + { + .ctl_name = CTL_NET, + .procname = "net", + .mode = 0555, + .child = nf_ct_netfilter_table, + }, + { .ctl_name = 0 } +}; +#endif + +struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv6 = { + .l3proto = PF_INET6, + .name = "ipv6", + .pkt_to_tuple = ipv6_pkt_to_tuple, + .invert_tuple = ipv6_invert_tuple, + .print_tuple = ipv6_print_tuple, + .print_conntrack = ipv6_print_conntrack, + .prepare = ipv6_prepare, + .get_features = ipv6_get_features, + .me = THIS_MODULE, +}; + +extern struct nf_conntrack_protocol nf_conntrack_protocol_tcp6; +extern struct nf_conntrack_protocol nf_conntrack_protocol_udp6; +extern struct nf_conntrack_protocol nf_conntrack_protocol_icmpv6; +extern int nf_ct_frag6_init(void); +extern void nf_ct_frag6_cleanup(void); +static int init_or_cleanup(int init) +{ + int ret = 0; + + if (!init) goto cleanup; + + ret = nf_ct_frag6_init(); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't initialize frag6.\n"); + goto cleanup_nothing; + } + ret = nf_conntrack_protocol_register(&nf_conntrack_protocol_tcp6); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register tcp.\n"); + goto cleanup_frag6; + } + + ret = nf_conntrack_protocol_register(&nf_conntrack_protocol_udp6); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register udp.\n"); + goto cleanup_tcp; + } + + ret = nf_conntrack_protocol_register(&nf_conntrack_protocol_icmpv6); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register icmpv6.\n"); + goto cleanup_udp; + } + + ret = nf_conntrack_l3proto_register(&nf_conntrack_l3proto_ipv6); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register ipv6\n"); + goto cleanup_icmpv6; + } + + ret = nf_register_hook(&ipv6_conntrack_defrag_ops); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register pre-routing defrag " + "hook.\n"); + goto cleanup_ipv6; + } + + ret = nf_register_hook(&ipv6_conntrack_defrag_local_out_ops); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register local_out defrag " + "hook.\n"); + goto cleanup_defragops; + } + + ret = nf_register_hook(&ipv6_conntrack_in_ops); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register pre-routing hook.\n"); + goto cleanup_defraglocalops; + } + + ret = nf_register_hook(&ipv6_conntrack_local_out_ops); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register local out hook.\n"); + goto cleanup_inops; + } + + ret = nf_register_hook(&ipv6_conntrack_out_ops); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register post-routing hook.\n"); + goto cleanup_inandlocalops; + } + + ret = nf_register_hook(&ipv6_conntrack_local_in_ops); + if (ret < 0) { + printk("nf_conntrack_ipv6: can't register local in hook.\n"); + goto cleanup_inoutandlocalops; + } + +#ifdef CONFIG_SYSCTL + nf_ct_ipv6_sysctl_header = register_sysctl_table(nf_ct_net_table, 0); + if (nf_ct_ipv6_sysctl_header == NULL) { + printk("nf_conntrack: can't register to sysctl.\n"); + ret = -ENOMEM; + goto cleanup_localinops; + } +#endif + return ret; + + cleanup: + synchronize_net(); +#ifdef CONFIG_SYSCTL + unregister_sysctl_table(nf_ct_ipv6_sysctl_header); + cleanup_localinops: +#endif + nf_unregister_hook(&ipv6_conntrack_local_in_ops); + cleanup_inoutandlocalops: + nf_unregister_hook(&ipv6_conntrack_out_ops); + cleanup_inandlocalops: + nf_unregister_hook(&ipv6_conntrack_local_out_ops); + cleanup_inops: + nf_unregister_hook(&ipv6_conntrack_in_ops); + cleanup_defraglocalops: + nf_unregister_hook(&ipv6_conntrack_defrag_local_out_ops); + cleanup_defragops: + nf_unregister_hook(&ipv6_conntrack_defrag_ops); + cleanup_ipv6: + nf_conntrack_l3proto_unregister(&nf_conntrack_l3proto_ipv6); + cleanup_icmpv6: + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_icmpv6); + cleanup_udp: + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_udp6); + cleanup_tcp: + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_tcp6); + cleanup_frag6: + nf_ct_frag6_cleanup(); + cleanup_nothing: + return ret; +} + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Yasuyuki KOZAKAI @USAGI "); + +static int __init init(void) +{ + need_nf_conntrack(); + return init_or_cleanup(1); +} + +static void __exit fini(void) +{ + init_or_cleanup(0); +} + +module_init(init); +module_exit(fini); + +void need_ip6_conntrack(void) +{ +} + +EXPORT_SYMBOL(need_ip6_conntrack); diff --git a/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c b/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c new file mode 100644 index 000000000000..c0f1da5497a9 --- /dev/null +++ b/net/ipv6/netfilter/nf_conntrack_proto_icmpv6.c @@ -0,0 +1,272 @@ +/* + * Copyright (C)2003,2004 USAGI/WIDE Project + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * Author: + * Yasuyuki Kozakai @USAGI + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - ICMPv6 tracking support. Derived from the original ip_conntrack code + * net/ipv4/netfilter/ip_conntrack_proto_icmp.c which had the following + * copyright information: + * (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +unsigned long nf_ct_icmpv6_timeout = 30*HZ; + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +static int icmpv6_pkt_to_tuple(const struct sk_buff *skb, + unsigned int dataoff, + struct nf_conntrack_tuple *tuple) +{ + struct icmp6hdr _hdr, *hp; + + hp = skb_header_pointer(skb, dataoff, sizeof(_hdr), &_hdr); + if (hp == NULL) + return 0; + tuple->dst.u.icmp.type = hp->icmp6_type; + tuple->src.u.icmp.id = hp->icmp6_identifier; + tuple->dst.u.icmp.code = hp->icmp6_code; + + return 1; +} + +static int icmpv6_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + /* Add 1; spaces filled with 0. */ + static u_int8_t invmap[] = { + [ICMPV6_ECHO_REQUEST - 128] = ICMPV6_ECHO_REPLY + 1, + [ICMPV6_ECHO_REPLY - 128] = ICMPV6_ECHO_REQUEST + 1, + [ICMPV6_NI_QUERY - 128] = ICMPV6_NI_QUERY + 1, + [ICMPV6_NI_REPLY - 128] = ICMPV6_NI_REPLY +1 + }; + + __u8 type = orig->dst.u.icmp.type - 128; + if (type >= sizeof(invmap) || !invmap[type]) + return 0; + + tuple->src.u.icmp.id = orig->src.u.icmp.id; + tuple->dst.u.icmp.type = invmap[type] - 1; + tuple->dst.u.icmp.code = orig->dst.u.icmp.code; + return 1; +} + +/* Print out the per-protocol part of the tuple. */ +static int icmpv6_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + return seq_printf(s, "type=%u code=%u id=%u ", + tuple->dst.u.icmp.type, + tuple->dst.u.icmp.code, + ntohs(tuple->src.u.icmp.id)); +} + +/* Print out the private part of the conntrack. */ +static int icmpv6_print_conntrack(struct seq_file *s, + const struct nf_conn *conntrack) +{ + return 0; +} + +/* Returns verdict for packet, or -1 for invalid. */ +static int icmpv6_packet(struct nf_conn *ct, + const struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info ctinfo, + int pf, + unsigned int hooknum) +{ + /* Try to delete connection immediately after all replies: + won't actually vanish as we still have skb, and del_timer + means this will only run once even if count hits zero twice + (theoretically possible with SMP) */ + if (CTINFO2DIR(ctinfo) == IP_CT_DIR_REPLY) { + if (atomic_dec_and_test(&ct->proto.icmp.count) + && del_timer(&ct->timeout)) + ct->timeout.function((unsigned long)ct); + } else { + atomic_inc(&ct->proto.icmp.count); + nf_conntrack_event_cache(IPCT_PROTOINFO_VOLATILE, skb); + nf_ct_refresh_acct(ct, ctinfo, skb, nf_ct_icmpv6_timeout); + } + + return NF_ACCEPT; +} + +/* Called when a new connection for this protocol found. */ +static int icmpv6_new(struct nf_conn *conntrack, + const struct sk_buff *skb, + unsigned int dataoff) +{ + static u_int8_t valid_new[] = { + [ICMPV6_ECHO_REQUEST - 128] = 1, + [ICMPV6_NI_QUERY - 128] = 1 + }; + + if (conntrack->tuplehash[0].tuple.dst.u.icmp.type - 128 >= sizeof(valid_new) + || !valid_new[conntrack->tuplehash[0].tuple.dst.u.icmp.type - 128]) { + /* Can't create a new ICMPv6 `conn' with this. */ + DEBUGP("icmp: can't create new conn with type %u\n", + conntrack->tuplehash[0].tuple.dst.u.icmp.type); + NF_CT_DUMP_TUPLE(&conntrack->tuplehash[0].tuple); + return 0; + } + atomic_set(&conntrack->proto.icmp.count, 0); + return 1; +} + +extern int +nf_ct_ipv6_skip_exthdr(struct sk_buff *skb, int start, u8 *nexthdrp, int len); +extern struct nf_conntrack_l3proto nf_conntrack_l3proto_ipv6; +static int +icmpv6_error_message(struct sk_buff *skb, + unsigned int icmp6off, + enum ip_conntrack_info *ctinfo, + unsigned int hooknum) +{ + struct nf_conntrack_tuple intuple, origtuple; + struct nf_conntrack_tuple_hash *h; + struct icmp6hdr _hdr, *hp; + unsigned int inip6off; + struct nf_conntrack_protocol *inproto; + u_int8_t inprotonum; + unsigned int inprotoff; + + NF_CT_ASSERT(skb->nfct == NULL); + + hp = skb_header_pointer(skb, icmp6off, sizeof(_hdr), &_hdr); + if (hp == NULL) { + DEBUGP("icmpv6_error: Can't get ICMPv6 hdr.\n"); + return -NF_ACCEPT; + } + + inip6off = icmp6off + sizeof(_hdr); + if (skb_copy_bits(skb, inip6off+offsetof(struct ipv6hdr, nexthdr), + &inprotonum, sizeof(inprotonum)) != 0) { + DEBUGP("icmpv6_error: Can't get nexthdr in inner IPv6 header.\n"); + return -NF_ACCEPT; + } + inprotoff = nf_ct_ipv6_skip_exthdr(skb, + inip6off + sizeof(struct ipv6hdr), + &inprotonum, + skb->len - inip6off + - sizeof(struct ipv6hdr)); + + if ((inprotoff < 0) || (inprotoff > skb->len) || + (inprotonum == NEXTHDR_FRAGMENT)) { + DEBUGP("icmpv6_error: Can't get protocol header in ICMPv6 payload.\n"); + return -NF_ACCEPT; + } + + inproto = nf_ct_find_proto(PF_INET6, inprotonum); + + /* Are they talking about one of our connections? */ + if (!nf_ct_get_tuple(skb, inip6off, inprotoff, PF_INET6, inprotonum, + &origtuple, &nf_conntrack_l3proto_ipv6, inproto)) { + DEBUGP("icmpv6_error: Can't get tuple\n"); + return -NF_ACCEPT; + } + + /* Ordinarily, we'd expect the inverted tupleproto, but it's + been preserved inside the ICMP. */ + if (!nf_ct_invert_tuple(&intuple, &origtuple, + &nf_conntrack_l3proto_ipv6, inproto)) { + DEBUGP("icmpv6_error: Can't invert tuple\n"); + return -NF_ACCEPT; + } + + *ctinfo = IP_CT_RELATED; + + h = nf_conntrack_find_get(&intuple, NULL); + if (!h) { + DEBUGP("icmpv6_error: no match\n"); + return -NF_ACCEPT; + } else { + if (NF_CT_DIRECTION(h) == IP_CT_DIR_REPLY) + *ctinfo += IP_CT_IS_REPLY; + } + + /* Update skb to refer to this connection */ + skb->nfct = &nf_ct_tuplehash_to_ctrack(h)->ct_general; + skb->nfctinfo = *ctinfo; + return -NF_ACCEPT; +} + +static int +icmpv6_error(struct sk_buff *skb, unsigned int dataoff, + enum ip_conntrack_info *ctinfo, int pf, unsigned int hooknum) +{ + struct icmp6hdr _ih, *icmp6h; + + icmp6h = skb_header_pointer(skb, dataoff, sizeof(_ih), &_ih); + if (icmp6h == NULL) { + if (LOG_INVALID(IPPROTO_ICMPV6)) + nf_log_packet(PF_INET6, 0, skb, NULL, NULL, NULL, + "nf_ct_icmpv6: short packet "); + return -NF_ACCEPT; + } + + if (hooknum != NF_IP6_PRE_ROUTING) + goto skipped; + + /* Ignore it if the checksum's bogus. */ + if (csum_ipv6_magic(&skb->nh.ipv6h->saddr, &skb->nh.ipv6h->daddr, + skb->len - dataoff, IPPROTO_ICMPV6, + skb_checksum(skb, dataoff, + skb->len - dataoff, 0))) { + nf_log_packet(PF_INET6, 0, skb, NULL, NULL, NULL, + "nf_ct_icmpv6: ICMPv6 checksum failed\n"); + return -NF_ACCEPT; + } + +skipped: + + /* is not error message ? */ + if (icmp6h->icmp6_type >= 128) + return NF_ACCEPT; + + return icmpv6_error_message(skb, dataoff, ctinfo, hooknum); +} + +struct nf_conntrack_protocol nf_conntrack_protocol_icmpv6 = +{ + .l3proto = PF_INET6, + .proto = IPPROTO_ICMPV6, + .name = "icmpv6", + .pkt_to_tuple = icmpv6_pkt_to_tuple, + .invert_tuple = icmpv6_invert_tuple, + .print_tuple = icmpv6_print_tuple, + .print_conntrack = icmpv6_print_conntrack, + .packet = icmpv6_packet, + .new = icmpv6_new, + .error = icmpv6_error, +}; + +EXPORT_SYMBOL(nf_conntrack_protocol_icmpv6); diff --git a/net/ipv6/netfilter/nf_conntrack_reasm.c b/net/ipv6/netfilter/nf_conntrack_reasm.c new file mode 100644 index 000000000000..7640b9bb7694 --- /dev/null +++ b/net/ipv6/netfilter/nf_conntrack_reasm.c @@ -0,0 +1,885 @@ +/* + * IPv6 fragment reassembly for connection tracking + * + * Copyright (C)2004 USAGI/WIDE Project + * + * Author: + * Yasuyuki Kozakai @USAGI + * + * Based on: net/ipv6/reassembly.c + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +#define NF_CT_FRAG6_HIGH_THRESH 262144 /* == 256*1024 */ +#define NF_CT_FRAG6_LOW_THRESH 196608 /* == 192*1024 */ +#define NF_CT_FRAG6_TIMEOUT IPV6_FRAG_TIMEOUT + +int nf_ct_frag6_high_thresh = 256*1024; +int nf_ct_frag6_low_thresh = 192*1024; +int nf_ct_frag6_timeout = IPV6_FRAG_TIMEOUT; + +struct nf_ct_frag6_skb_cb +{ + struct inet6_skb_parm h; + int offset; + struct sk_buff *orig; +}; + +#define NFCT_FRAG6_CB(skb) ((struct nf_ct_frag6_skb_cb*)((skb)->cb)) + +struct nf_ct_frag6_queue +{ + struct nf_ct_frag6_queue *next; + struct list_head lru_list; /* lru list member */ + + __u32 id; /* fragment id */ + struct in6_addr saddr; + struct in6_addr daddr; + + spinlock_t lock; + atomic_t refcnt; + struct timer_list timer; /* expire timer */ + struct sk_buff *fragments; + int len; + int meat; + struct timeval stamp; + unsigned int csum; + __u8 last_in; /* has first/last segment arrived? */ +#define COMPLETE 4 +#define FIRST_IN 2 +#define LAST_IN 1 + __u16 nhoffset; + struct nf_ct_frag6_queue **pprev; +}; + +/* Hash table. */ + +#define FRAG6Q_HASHSZ 64 + +static struct nf_ct_frag6_queue *nf_ct_frag6_hash[FRAG6Q_HASHSZ]; +static rwlock_t nf_ct_frag6_lock = RW_LOCK_UNLOCKED; +static u32 nf_ct_frag6_hash_rnd; +static LIST_HEAD(nf_ct_frag6_lru_list); +int nf_ct_frag6_nqueues = 0; + +static __inline__ void __fq_unlink(struct nf_ct_frag6_queue *fq) +{ + if (fq->next) + fq->next->pprev = fq->pprev; + *fq->pprev = fq->next; + list_del(&fq->lru_list); + nf_ct_frag6_nqueues--; +} + +static __inline__ void fq_unlink(struct nf_ct_frag6_queue *fq) +{ + write_lock(&nf_ct_frag6_lock); + __fq_unlink(fq); + write_unlock(&nf_ct_frag6_lock); +} + +static unsigned int ip6qhashfn(u32 id, struct in6_addr *saddr, + struct in6_addr *daddr) +{ + u32 a, b, c; + + a = saddr->s6_addr32[0]; + b = saddr->s6_addr32[1]; + c = saddr->s6_addr32[2]; + + a += JHASH_GOLDEN_RATIO; + b += JHASH_GOLDEN_RATIO; + c += nf_ct_frag6_hash_rnd; + __jhash_mix(a, b, c); + + a += saddr->s6_addr32[3]; + b += daddr->s6_addr32[0]; + c += daddr->s6_addr32[1]; + __jhash_mix(a, b, c); + + a += daddr->s6_addr32[2]; + b += daddr->s6_addr32[3]; + c += id; + __jhash_mix(a, b, c); + + return c & (FRAG6Q_HASHSZ - 1); +} + +static struct timer_list nf_ct_frag6_secret_timer; +int nf_ct_frag6_secret_interval = 10 * 60 * HZ; + +static void nf_ct_frag6_secret_rebuild(unsigned long dummy) +{ + unsigned long now = jiffies; + int i; + + write_lock(&nf_ct_frag6_lock); + get_random_bytes(&nf_ct_frag6_hash_rnd, sizeof(u32)); + for (i = 0; i < FRAG6Q_HASHSZ; i++) { + struct nf_ct_frag6_queue *q; + + q = nf_ct_frag6_hash[i]; + while (q) { + struct nf_ct_frag6_queue *next = q->next; + unsigned int hval = ip6qhashfn(q->id, + &q->saddr, + &q->daddr); + + if (hval != i) { + /* Unlink. */ + if (q->next) + q->next->pprev = q->pprev; + *q->pprev = q->next; + + /* Relink to new hash chain. */ + if ((q->next = nf_ct_frag6_hash[hval]) != NULL) + q->next->pprev = &q->next; + nf_ct_frag6_hash[hval] = q; + q->pprev = &nf_ct_frag6_hash[hval]; + } + + q = next; + } + } + write_unlock(&nf_ct_frag6_lock); + + mod_timer(&nf_ct_frag6_secret_timer, now + nf_ct_frag6_secret_interval); +} + +atomic_t nf_ct_frag6_mem = ATOMIC_INIT(0); + +/* Memory Tracking Functions. */ +static inline void frag_kfree_skb(struct sk_buff *skb) +{ + atomic_sub(skb->truesize, &nf_ct_frag6_mem); + if (NFCT_FRAG6_CB(skb)->orig) + kfree_skb(NFCT_FRAG6_CB(skb)->orig); + + kfree_skb(skb); +} + +static inline void frag_free_queue(struct nf_ct_frag6_queue *fq) +{ + atomic_sub(sizeof(struct nf_ct_frag6_queue), &nf_ct_frag6_mem); + kfree(fq); +} + +static inline struct nf_ct_frag6_queue *frag_alloc_queue(void) +{ + struct nf_ct_frag6_queue *fq = kmalloc(sizeof(struct nf_ct_frag6_queue), GFP_ATOMIC); + + if (!fq) + return NULL; + atomic_add(sizeof(struct nf_ct_frag6_queue), &nf_ct_frag6_mem); + return fq; +} + +/* Destruction primitives. */ + +/* Complete destruction of fq. */ +static void nf_ct_frag6_destroy(struct nf_ct_frag6_queue *fq) +{ + struct sk_buff *fp; + + BUG_TRAP(fq->last_in&COMPLETE); + BUG_TRAP(del_timer(&fq->timer) == 0); + + /* Release all fragment data. */ + fp = fq->fragments; + while (fp) { + struct sk_buff *xp = fp->next; + + frag_kfree_skb(fp); + fp = xp; + } + + frag_free_queue(fq); +} + +static __inline__ void fq_put(struct nf_ct_frag6_queue *fq) +{ + if (atomic_dec_and_test(&fq->refcnt)) + nf_ct_frag6_destroy(fq); +} + +/* Kill fq entry. It is not destroyed immediately, + * because caller (and someone more) holds reference count. + */ +static __inline__ void fq_kill(struct nf_ct_frag6_queue *fq) +{ + if (del_timer(&fq->timer)) + atomic_dec(&fq->refcnt); + + if (!(fq->last_in & COMPLETE)) { + fq_unlink(fq); + atomic_dec(&fq->refcnt); + fq->last_in |= COMPLETE; + } +} + +static void nf_ct_frag6_evictor(void) +{ + struct nf_ct_frag6_queue *fq; + struct list_head *tmp; + + for (;;) { + if (atomic_read(&nf_ct_frag6_mem) <= nf_ct_frag6_low_thresh) + return; + read_lock(&nf_ct_frag6_lock); + if (list_empty(&nf_ct_frag6_lru_list)) { + read_unlock(&nf_ct_frag6_lock); + return; + } + tmp = nf_ct_frag6_lru_list.next; + fq = list_entry(tmp, struct nf_ct_frag6_queue, lru_list); + atomic_inc(&fq->refcnt); + read_unlock(&nf_ct_frag6_lock); + + spin_lock(&fq->lock); + if (!(fq->last_in&COMPLETE)) + fq_kill(fq); + spin_unlock(&fq->lock); + + fq_put(fq); + } +} + +static void nf_ct_frag6_expire(unsigned long data) +{ + struct nf_ct_frag6_queue *fq = (struct nf_ct_frag6_queue *) data; + + spin_lock(&fq->lock); + + if (fq->last_in & COMPLETE) + goto out; + + fq_kill(fq); + +out: + spin_unlock(&fq->lock); + fq_put(fq); +} + +/* Creation primitives. */ + + +static struct nf_ct_frag6_queue *nf_ct_frag6_intern(unsigned int hash, + struct nf_ct_frag6_queue *fq_in) +{ + struct nf_ct_frag6_queue *fq; + + write_lock(&nf_ct_frag6_lock); +#ifdef CONFIG_SMP + for (fq = nf_ct_frag6_hash[hash]; fq; fq = fq->next) { + if (fq->id == fq_in->id && + !ipv6_addr_cmp(&fq_in->saddr, &fq->saddr) && + !ipv6_addr_cmp(&fq_in->daddr, &fq->daddr)) { + atomic_inc(&fq->refcnt); + write_unlock(&nf_ct_frag6_lock); + fq_in->last_in |= COMPLETE; + fq_put(fq_in); + return fq; + } + } +#endif + fq = fq_in; + + if (!mod_timer(&fq->timer, jiffies + nf_ct_frag6_timeout)) + atomic_inc(&fq->refcnt); + + atomic_inc(&fq->refcnt); + if ((fq->next = nf_ct_frag6_hash[hash]) != NULL) + fq->next->pprev = &fq->next; + nf_ct_frag6_hash[hash] = fq; + fq->pprev = &nf_ct_frag6_hash[hash]; + INIT_LIST_HEAD(&fq->lru_list); + list_add_tail(&fq->lru_list, &nf_ct_frag6_lru_list); + nf_ct_frag6_nqueues++; + write_unlock(&nf_ct_frag6_lock); + return fq; +} + + +static struct nf_ct_frag6_queue * +nf_ct_frag6_create(unsigned int hash, u32 id, struct in6_addr *src, struct in6_addr *dst) +{ + struct nf_ct_frag6_queue *fq; + + if ((fq = frag_alloc_queue()) == NULL) { + DEBUGP("Can't alloc new queue\n"); + goto oom; + } + + memset(fq, 0, sizeof(struct nf_ct_frag6_queue)); + + fq->id = id; + ipv6_addr_copy(&fq->saddr, src); + ipv6_addr_copy(&fq->daddr, dst); + + init_timer(&fq->timer); + fq->timer.function = nf_ct_frag6_expire; + fq->timer.data = (long) fq; + fq->lock = SPIN_LOCK_UNLOCKED; + atomic_set(&fq->refcnt, 1); + + return nf_ct_frag6_intern(hash, fq); + +oom: + return NULL; +} + +static __inline__ struct nf_ct_frag6_queue * +fq_find(u32 id, struct in6_addr *src, struct in6_addr *dst) +{ + struct nf_ct_frag6_queue *fq; + unsigned int hash = ip6qhashfn(id, src, dst); + + read_lock(&nf_ct_frag6_lock); + for (fq = nf_ct_frag6_hash[hash]; fq; fq = fq->next) { + if (fq->id == id && + !ipv6_addr_cmp(src, &fq->saddr) && + !ipv6_addr_cmp(dst, &fq->daddr)) { + atomic_inc(&fq->refcnt); + read_unlock(&nf_ct_frag6_lock); + return fq; + } + } + read_unlock(&nf_ct_frag6_lock); + + return nf_ct_frag6_create(hash, id, src, dst); +} + + +static int nf_ct_frag6_queue(struct nf_ct_frag6_queue *fq, struct sk_buff *skb, + struct frag_hdr *fhdr, int nhoff) +{ + struct sk_buff *prev, *next; + int offset, end; + + if (fq->last_in & COMPLETE) { + DEBUGP("Allready completed\n"); + goto err; + } + + offset = ntohs(fhdr->frag_off) & ~0x7; + end = offset + (ntohs(skb->nh.ipv6h->payload_len) - + ((u8 *) (fhdr + 1) - (u8 *) (skb->nh.ipv6h + 1))); + + if ((unsigned int)end > IPV6_MAXPLEN) { + DEBUGP("offset is too large.\n"); + return -1; + } + + if (skb->ip_summed == CHECKSUM_HW) + skb->csum = csum_sub(skb->csum, + csum_partial(skb->nh.raw, + (u8*)(fhdr + 1) - skb->nh.raw, + 0)); + + /* Is this the final fragment? */ + if (!(fhdr->frag_off & htons(IP6_MF))) { + /* If we already have some bits beyond end + * or have different end, the segment is corrupted. + */ + if (end < fq->len || + ((fq->last_in & LAST_IN) && end != fq->len)) { + DEBUGP("already received last fragment\n"); + goto err; + } + fq->last_in |= LAST_IN; + fq->len = end; + } else { + /* Check if the fragment is rounded to 8 bytes. + * Required by the RFC. + */ + if (end & 0x7) { + /* RFC2460 says always send parameter problem in + * this case. -DaveM + */ + DEBUGP("the end of this fragment is not rounded to 8 bytes.\n"); + return -1; + } + if (end > fq->len) { + /* Some bits beyond end -> corruption. */ + if (fq->last_in & LAST_IN) { + DEBUGP("last packet already reached.\n"); + goto err; + } + fq->len = end; + } + } + + if (end == offset) + goto err; + + /* Point into the IP datagram 'data' part. */ + if (!pskb_pull(skb, (u8 *) (fhdr + 1) - skb->data)) { + DEBUGP("queue: message is too short.\n"); + goto err; + } + if (end-offset < skb->len) { + if (pskb_trim(skb, end - offset)) { + DEBUGP("Can't trim\n"); + goto err; + } + if (skb->ip_summed != CHECKSUM_UNNECESSARY) + skb->ip_summed = CHECKSUM_NONE; + } + + /* Find out which fragments are in front and at the back of us + * in the chain of fragments so far. We must know where to put + * this fragment, right? + */ + prev = NULL; + for (next = fq->fragments; next != NULL; next = next->next) { + if (NFCT_FRAG6_CB(next)->offset >= offset) + break; /* bingo! */ + prev = next; + } + + /* We found where to put this one. Check for overlap with + * preceding fragment, and, if needed, align things so that + * any overlaps are eliminated. + */ + if (prev) { + int i = (NFCT_FRAG6_CB(prev)->offset + prev->len) - offset; + + if (i > 0) { + offset += i; + if (end <= offset) { + DEBUGP("overlap\n"); + goto err; + } + if (!pskb_pull(skb, i)) { + DEBUGP("Can't pull\n"); + goto err; + } + if (skb->ip_summed != CHECKSUM_UNNECESSARY) + skb->ip_summed = CHECKSUM_NONE; + } + } + + /* Look for overlap with succeeding segments. + * If we can merge fragments, do it. + */ + while (next && NFCT_FRAG6_CB(next)->offset < end) { + /* overlap is 'i' bytes */ + int i = end - NFCT_FRAG6_CB(next)->offset; + + if (i < next->len) { + /* Eat head of the next overlapped fragment + * and leave the loop. The next ones cannot overlap. + */ + DEBUGP("Eat head of the overlapped parts.: %d", i); + if (!pskb_pull(next, i)) + goto err; + + /* next fragment */ + NFCT_FRAG6_CB(next)->offset += i; + fq->meat -= i; + if (next->ip_summed != CHECKSUM_UNNECESSARY) + next->ip_summed = CHECKSUM_NONE; + break; + } else { + struct sk_buff *free_it = next; + + /* Old fragmnet is completely overridden with + * new one drop it. + */ + next = next->next; + + if (prev) + prev->next = next; + else + fq->fragments = next; + + fq->meat -= free_it->len; + frag_kfree_skb(free_it); + } + } + + NFCT_FRAG6_CB(skb)->offset = offset; + + /* Insert this fragment in the chain of fragments. */ + skb->next = next; + if (prev) + prev->next = skb; + else + fq->fragments = skb; + + skb->dev = NULL; + skb_get_timestamp(skb, &fq->stamp); + fq->meat += skb->len; + atomic_add(skb->truesize, &nf_ct_frag6_mem); + + /* The first fragment. + * nhoffset is obtained from the first fragment, of course. + */ + if (offset == 0) { + fq->nhoffset = nhoff; + fq->last_in |= FIRST_IN; + } + write_lock(&nf_ct_frag6_lock); + list_move_tail(&fq->lru_list, &nf_ct_frag6_lru_list); + write_unlock(&nf_ct_frag6_lock); + return 0; + +err: + return -1; +} + +/* + * Check if this packet is complete. + * Returns NULL on failure by any reason, and pointer + * to current nexthdr field in reassembled frame. + * + * It is called with locked fq, and caller must check that + * queue is eligible for reassembly i.e. it is not COMPLETE, + * the last and the first frames arrived and all the bits are here. + */ +static struct sk_buff * +nf_ct_frag6_reasm(struct nf_ct_frag6_queue *fq, struct net_device *dev) +{ + struct sk_buff *fp, *op, *head = fq->fragments; + int payload_len; + + fq_kill(fq); + + BUG_TRAP(head != NULL); + BUG_TRAP(NFCT_FRAG6_CB(head)->offset == 0); + + /* Unfragmented part is taken from the first segment. */ + payload_len = (head->data - head->nh.raw) - sizeof(struct ipv6hdr) + fq->len - sizeof(struct frag_hdr); + if (payload_len > IPV6_MAXPLEN) { + DEBUGP("payload len is too large.\n"); + goto out_oversize; + } + + /* Head of list must not be cloned. */ + if (skb_cloned(head) && pskb_expand_head(head, 0, 0, GFP_ATOMIC)) { + DEBUGP("skb is cloned but can't expand head"); + goto out_oom; + } + + /* If the first fragment is fragmented itself, we split + * it to two chunks: the first with data and paged part + * and the second, holding only fragments. */ + if (skb_shinfo(head)->frag_list) { + struct sk_buff *clone; + int i, plen = 0; + + if ((clone = alloc_skb(0, GFP_ATOMIC)) == NULL) { + DEBUGP("Can't alloc skb\n"); + goto out_oom; + } + clone->next = head->next; + head->next = clone; + skb_shinfo(clone)->frag_list = skb_shinfo(head)->frag_list; + skb_shinfo(head)->frag_list = NULL; + for (i=0; inr_frags; i++) + plen += skb_shinfo(head)->frags[i].size; + clone->len = clone->data_len = head->data_len - plen; + head->data_len -= clone->len; + head->len -= clone->len; + clone->csum = 0; + clone->ip_summed = head->ip_summed; + + NFCT_FRAG6_CB(clone)->orig = NULL; + atomic_add(clone->truesize, &nf_ct_frag6_mem); + } + + /* We have to remove fragment header from datagram and to relocate + * header in order to calculate ICV correctly. */ + head->nh.raw[fq->nhoffset] = head->h.raw[0]; + memmove(head->head + sizeof(struct frag_hdr), head->head, + (head->data - head->head) - sizeof(struct frag_hdr)); + head->mac.raw += sizeof(struct frag_hdr); + head->nh.raw += sizeof(struct frag_hdr); + + skb_shinfo(head)->frag_list = head->next; + head->h.raw = head->data; + skb_push(head, head->data - head->nh.raw); + atomic_sub(head->truesize, &nf_ct_frag6_mem); + + for (fp=head->next; fp; fp = fp->next) { + head->data_len += fp->len; + head->len += fp->len; + if (head->ip_summed != fp->ip_summed) + head->ip_summed = CHECKSUM_NONE; + else if (head->ip_summed == CHECKSUM_HW) + head->csum = csum_add(head->csum, fp->csum); + head->truesize += fp->truesize; + atomic_sub(fp->truesize, &nf_ct_frag6_mem); + } + + head->next = NULL; + head->dev = dev; + skb_set_timestamp(head, &fq->stamp); + head->nh.ipv6h->payload_len = htons(payload_len); + + /* Yes, and fold redundant checksum back. 8) */ + if (head->ip_summed == CHECKSUM_HW) + head->csum = csum_partial(head->nh.raw, head->h.raw-head->nh.raw, head->csum); + + fq->fragments = NULL; + + /* all original skbs are linked into the NFCT_FRAG6_CB(head).orig */ + fp = skb_shinfo(head)->frag_list; + if (NFCT_FRAG6_CB(fp)->orig == NULL) + /* at above code, head skb is divided into two skbs. */ + fp = fp->next; + + op = NFCT_FRAG6_CB(head)->orig; + for (; fp; fp = fp->next) { + struct sk_buff *orig = NFCT_FRAG6_CB(fp)->orig; + + op->next = orig; + op = orig; + NFCT_FRAG6_CB(fp)->orig = NULL; + } + + return head; + +out_oversize: + if (net_ratelimit()) + printk(KERN_DEBUG "nf_ct_frag6_reasm: payload len = %d\n", payload_len); + goto out_fail; +out_oom: + if (net_ratelimit()) + printk(KERN_DEBUG "nf_ct_frag6_reasm: no memory for reassembly\n"); +out_fail: + return NULL; +} + +/* + * find the header just before Fragment Header. + * + * if success return 0 and set ... + * (*prevhdrp): the value of "Next Header Field" in the header + * just before Fragment Header. + * (*prevhoff): the offset of "Next Header Field" in the header + * just before Fragment Header. + * (*fhoff) : the offset of Fragment Header. + * + * Based on ipv6_skip_hdr() in net/ipv6/exthdr.c + * + */ +static int +find_prev_fhdr(struct sk_buff *skb, u8 *prevhdrp, int *prevhoff, int *fhoff) +{ + u8 nexthdr = skb->nh.ipv6h->nexthdr; + u8 prev_nhoff = (u8 *)&skb->nh.ipv6h->nexthdr - skb->data; + int start = (u8 *)(skb->nh.ipv6h+1) - skb->data; + int len = skb->len - start; + u8 prevhdr = NEXTHDR_IPV6; + + while (nexthdr != NEXTHDR_FRAGMENT) { + struct ipv6_opt_hdr hdr; + int hdrlen; + + if (!ipv6_ext_hdr(nexthdr)) { + return -1; + } + if (len < (int)sizeof(struct ipv6_opt_hdr)) { + DEBUGP("too short\n"); + return -1; + } + if (nexthdr == NEXTHDR_NONE) { + DEBUGP("next header is none\n"); + return -1; + } + if (skb_copy_bits(skb, start, &hdr, sizeof(hdr))) + BUG(); + if (nexthdr == NEXTHDR_AUTH) + hdrlen = (hdr.hdrlen+2)<<2; + else + hdrlen = ipv6_optlen(&hdr); + + prevhdr = nexthdr; + prev_nhoff = start; + + nexthdr = hdr.nexthdr; + len -= hdrlen; + start += hdrlen; + } + + if (len < 0) + return -1; + + *prevhdrp = prevhdr; + *prevhoff = prev_nhoff; + *fhoff = start; + + return 0; +} + +struct sk_buff *nf_ct_frag6_gather(struct sk_buff *skb) +{ + struct sk_buff *clone; + struct net_device *dev = skb->dev; + struct frag_hdr *fhdr; + struct nf_ct_frag6_queue *fq; + struct ipv6hdr *hdr; + int fhoff, nhoff; + u8 prevhdr; + struct sk_buff *ret_skb = NULL; + + /* Jumbo payload inhibits frag. header */ + if (skb->nh.ipv6h->payload_len == 0) { + DEBUGP("payload len = 0\n"); + return skb; + } + + if (find_prev_fhdr(skb, &prevhdr, &nhoff, &fhoff) < 0) + return skb; + + clone = skb_clone(skb, GFP_ATOMIC); + if (clone == NULL) { + DEBUGP("Can't clone skb\n"); + return skb; + } + + NFCT_FRAG6_CB(clone)->orig = skb; + + if (!pskb_may_pull(clone, fhoff + sizeof(*fhdr))) { + DEBUGP("message is too short.\n"); + goto ret_orig; + } + + clone->h.raw = clone->data + fhoff; + hdr = clone->nh.ipv6h; + fhdr = (struct frag_hdr *)clone->h.raw; + + if (!(fhdr->frag_off & htons(0xFFF9))) { + DEBUGP("Invalid fragment offset\n"); + /* It is not a fragmented frame */ + goto ret_orig; + } + + if (atomic_read(&nf_ct_frag6_mem) > nf_ct_frag6_high_thresh) + nf_ct_frag6_evictor(); + + fq = fq_find(fhdr->identification, &hdr->saddr, &hdr->daddr); + if (fq == NULL) { + DEBUGP("Can't find and can't create new queue\n"); + goto ret_orig; + } + + spin_lock(&fq->lock); + + if (nf_ct_frag6_queue(fq, clone, fhdr, nhoff) < 0) { + spin_unlock(&fq->lock); + DEBUGP("Can't insert skb to queue\n"); + fq_put(fq); + goto ret_orig; + } + + if (fq->last_in == (FIRST_IN|LAST_IN) && fq->meat == fq->len) { + ret_skb = nf_ct_frag6_reasm(fq, dev); + if (ret_skb == NULL) + DEBUGP("Can't reassemble fragmented packets\n"); + } + spin_unlock(&fq->lock); + + fq_put(fq); + return ret_skb; + +ret_orig: + kfree_skb(clone); + return skb; +} + +void nf_ct_frag6_output(unsigned int hooknum, struct sk_buff *skb, + struct net_device *in, struct net_device *out, + int (*okfn)(struct sk_buff *)) +{ + struct sk_buff *s, *s2; + + for (s = NFCT_FRAG6_CB(skb)->orig; s;) { + nf_conntrack_put_reasm(s->nfct_reasm); + nf_conntrack_get_reasm(skb); + s->nfct_reasm = skb; + + s2 = s->next; + NF_HOOK_THRESH(PF_INET6, hooknum, s, in, out, okfn, + NF_IP6_PRI_CONNTRACK_DEFRAG + 1); + s = s2; + } + nf_conntrack_put_reasm(skb); +} + +int nf_ct_frag6_kfree_frags(struct sk_buff *skb) +{ + struct sk_buff *s, *s2; + + for (s = NFCT_FRAG6_CB(skb)->orig; s; s = s2) { + + s2 = s->next; + kfree_skb(s); + } + + kfree_skb(skb); + + return 0; +} + +int nf_ct_frag6_init(void) +{ + nf_ct_frag6_hash_rnd = (u32) ((num_physpages ^ (num_physpages>>7)) ^ + (jiffies ^ (jiffies >> 6))); + + init_timer(&nf_ct_frag6_secret_timer); + nf_ct_frag6_secret_timer.function = nf_ct_frag6_secret_rebuild; + nf_ct_frag6_secret_timer.expires = jiffies + + nf_ct_frag6_secret_interval; + add_timer(&nf_ct_frag6_secret_timer); + + return 0; +} + +void nf_ct_frag6_cleanup(void) +{ + del_timer(&nf_ct_frag6_secret_timer); + nf_ct_frag6_evictor(); +} diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index a1265a320b11..651c79b41eeb 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -174,8 +174,10 @@ int ipv6_raw_deliver(struct sk_buff *skb, int nexthdr) struct sk_buff *clone = skb_clone(skb, GFP_ATOMIC); /* Not releasing hash table! */ - if (clone) + if (clone) { + nf_reset(clone); rawv6_rcv(sk, clone); + } } sk = __raw_v6_lookup(sk_next(sk), nexthdr, daddr, saddr, IP6CB(skb)->iif); diff --git a/net/netfilter/Kconfig b/net/netfilter/Kconfig index 8296b38bf270..a84f9221e5f0 100644 --- a/net/netfilter/Kconfig +++ b/net/netfilter/Kconfig @@ -1,3 +1,6 @@ +menu "Core Netfilter Configuration" + depends on NET && NETFILTER + config NETFILTER_NETLINK tristate "Netfilter netlink interface" help @@ -22,3 +25,74 @@ config NETFILTER_NETLINK_LOG and is also scheduled to replace the old syslog-based ipt_LOG and ip6t_LOG modules. +config NF_CONNTRACK + tristate "Layer 3 Independent Connection tracking (EXPERIMENTAL)" + depends on EXPERIMENTAL && IP_NF_CONNTRACK=n + default n + ---help--- + Connection tracking keeps a record of what packets have passed + through your machine, in order to figure out how they are related + into connections. + + Layer 3 independent connection tracking is experimental scheme + which generalize ip_conntrack to support other layer 3 protocols. + + To compile it as a module, choose M here. If unsure, say N. + +config NF_CT_ACCT + bool "Connection tracking flow accounting" + depends on NF_CONNTRACK + help + If this option is enabled, the connection tracking code will + keep per-flow packet and byte counters. + + Those counters can be used for flow-based accounting or the + `connbytes' match. + + If unsure, say `N'. + +config NF_CONNTRACK_MARK + bool 'Connection mark tracking support' + depends on NF_CONNTRACK + help + This option enables support for connection marks, used by the + `CONNMARK' target and `connmark' match. Similar to the mark value + of packets, but this mark value is kept in the conntrack session + instead of the individual packets. + +config NF_CONNTRACK_EVENTS + bool "Connection tracking events" + depends on NF_CONNTRACK + help + If this option is enabled, the connection tracking code will + provide a notifier chain that can be used by other kernel code + to get notified aboutchanges in the connection tracking state. + + If unsure, say `N'. + +config NF_CT_PROTO_SCTP + tristate 'SCTP protocol on new connection tracking support (EXPERIMENTAL)' + depends on EXPERIMENTAL && NF_CONNTRACK + default n + help + With this option enabled, the layer 3 independent connection + tracking code will be able to do state tracking on SCTP connections. + + If you want to compile it as a module, say M here and read + Documentation/modules.txt. If unsure, say `N'. + +config NF_CONNTRACK_FTP + tristate "FTP support on new connection tracking (EXPERIMENTAL)" + depends on EXPERIMENTAL && NF_CONNTRACK + help + Tracking FTP connections is problematic: special helpers are + required for tracking them, and doing masquerading and other forms + of Network Address Translation on them. + + This is FTP support on Layer 3 independent connection tracking. + Layer 3 independent connection tracking is experimental scheme + which generalize ip_conntrack to support other layer 3 protocols. + + To compile it as a module, choose M here. If unsure, say N. + +endmenu diff --git a/net/netfilter/Makefile b/net/netfilter/Makefile index b3b44f8b415a..55f019ad2c08 100644 --- a/net/netfilter/Makefile +++ b/net/netfilter/Makefile @@ -5,3 +5,11 @@ obj-$(CONFIG_NETFILTER) = netfilter.o obj-$(CONFIG_NETFILTER_NETLINK) += nfnetlink.o obj-$(CONFIG_NETFILTER_NETLINK_QUEUE) += nfnetlink_queue.o obj-$(CONFIG_NETFILTER_NETLINK_LOG) += nfnetlink_log.o + +nf_conntrack-objs := nf_conntrack_core.o nf_conntrack_standalone.o nf_conntrack_l3proto_generic.o nf_conntrack_proto_generic.o nf_conntrack_proto_tcp.o nf_conntrack_proto_udp.o + +obj-$(CONFIG_NF_CONNTRACK) += nf_conntrack.o +obj-$(CONFIG_NF_CONNTRACK_FTP) += nf_conntrack_ftp.o + +# SCTP protocol connection tracking +obj-$(CONFIG_NF_CT_PROTO_SCTP) += nf_conntrack_proto_sctp.o diff --git a/net/netfilter/nf_conntrack_core.c b/net/netfilter/nf_conntrack_core.c new file mode 100644 index 000000000000..9a67c796b385 --- /dev/null +++ b/net/netfilter/nf_conntrack_core.c @@ -0,0 +1,1538 @@ +/* Connection state tracking for netfilter. This is separated from, + but required by, the NAT layer; it can also be used by an iptables + extension. */ + +/* (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2005 Netfilter Core Team + * (C) 2003,2004 USAGI/WIDE Project + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 23 Apr 2001: Harald Welte + * - new API and handling of conntrack/nat helpers + * - now capable of multiple expectations for one master + * 16 Jul 2002: Harald Welte + * - add usage/reference counts to ip_conntrack_expect + * - export ip_conntrack[_expect]_{find_get,put} functions + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - generalize L3 protocol denendent part. + * 23 Mar 2004: Yasuyuki Kozakai @USAGI + * - add support various size of conntrack structures. + * + * Derived from net/ipv4/netfilter/ip_conntrack_core.c + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* This rwlock protects the main hash table, protocol/helper/expected + registrations, conntrack timers*/ +#define ASSERT_READ_LOCK(x) +#define ASSERT_WRITE_LOCK(x) + +#include +#include +#include +#include +#include +#include + +#define NF_CONNTRACK_VERSION "0.4.1" + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +DEFINE_RWLOCK(nf_conntrack_lock); + +/* nf_conntrack_standalone needs this */ +atomic_t nf_conntrack_count = ATOMIC_INIT(0); + +void (*nf_conntrack_destroyed)(struct nf_conn *conntrack) = NULL; +LIST_HEAD(nf_conntrack_expect_list); +struct nf_conntrack_protocol **nf_ct_protos[PF_MAX]; +struct nf_conntrack_l3proto *nf_ct_l3protos[PF_MAX]; +static LIST_HEAD(helpers); +unsigned int nf_conntrack_htable_size = 0; +int nf_conntrack_max; +struct list_head *nf_conntrack_hash; +static kmem_cache_t *nf_conntrack_expect_cachep; +struct nf_conn nf_conntrack_untracked; +unsigned int nf_ct_log_invalid; +static LIST_HEAD(unconfirmed); +static int nf_conntrack_vmalloc; + +#ifdef CONFIG_NF_CONNTRACK_EVENTS +struct notifier_block *nf_conntrack_chain; +struct notifier_block *nf_conntrack_expect_chain; + +DEFINE_PER_CPU(struct nf_conntrack_ecache, nf_conntrack_ecache); + +/* deliver cached events and clear cache entry - must be called with locally + * disabled softirqs */ +static inline void +__nf_ct_deliver_cached_events(struct nf_conntrack_ecache *ecache) +{ + DEBUGP("ecache: delivering events for %p\n", ecache->ct); + if (nf_ct_is_confirmed(ecache->ct) && !nf_ct_is_dying(ecache->ct) + && ecache->events) + notifier_call_chain(&nf_conntrack_chain, ecache->events, + ecache->ct); + + ecache->events = 0; + nf_ct_put(ecache->ct); + ecache->ct = NULL; +} + +/* Deliver all cached events for a particular conntrack. This is called + * by code prior to async packet handling for freeing the skb */ +void nf_ct_deliver_cached_events(const struct nf_conn *ct) +{ + struct nf_conntrack_ecache *ecache; + + local_bh_disable(); + ecache = &__get_cpu_var(nf_conntrack_ecache); + if (ecache->ct == ct) + __nf_ct_deliver_cached_events(ecache); + local_bh_enable(); +} + +/* Deliver cached events for old pending events, if current conntrack != old */ +void __nf_ct_event_cache_init(struct nf_conn *ct) +{ + struct nf_conntrack_ecache *ecache; + + /* take care of delivering potentially old events */ + ecache = &__get_cpu_var(nf_conntrack_ecache); + BUG_ON(ecache->ct == ct); + if (ecache->ct) + __nf_ct_deliver_cached_events(ecache); + /* initialize for this conntrack/packet */ + ecache->ct = ct; + nf_conntrack_get(&ct->ct_general); +} + +/* flush the event cache - touches other CPU's data and must not be called + * while packets are still passing through the code */ +static void nf_ct_event_cache_flush(void) +{ + struct nf_conntrack_ecache *ecache; + int cpu; + + for_each_cpu(cpu) { + ecache = &per_cpu(nf_conntrack_ecache, cpu); + if (ecache->ct) + nf_ct_put(ecache->ct); + } +} +#else +static inline void nf_ct_event_cache_flush(void) {} +#endif /* CONFIG_NF_CONNTRACK_EVENTS */ + +DEFINE_PER_CPU(struct ip_conntrack_stat, nf_conntrack_stat); +EXPORT_PER_CPU_SYMBOL(nf_conntrack_stat); + +/* + * This scheme offers various size of "struct nf_conn" dependent on + * features(helper, nat, ...) + */ + +#define NF_CT_FEATURES_NAMELEN 256 +static struct { + /* name of slab cache. printed in /proc/slabinfo */ + char *name; + + /* size of slab cache */ + size_t size; + + /* slab cache pointer */ + kmem_cache_t *cachep; + + /* allocated slab cache + modules which uses this slab cache */ + int use; + + /* Initialization */ + int (*init_conntrack)(struct nf_conn *, u_int32_t); + +} nf_ct_cache[NF_CT_F_NUM]; + +/* protect members of nf_ct_cache except of "use" */ +DEFINE_RWLOCK(nf_ct_cache_lock); + +/* This avoids calling kmem_cache_create() with same name simultaneously */ +DECLARE_MUTEX(nf_ct_cache_mutex); + +extern struct nf_conntrack_protocol nf_conntrack_generic_protocol; +struct nf_conntrack_protocol * +nf_ct_find_proto(u_int16_t l3proto, u_int8_t protocol) +{ + if (unlikely(nf_ct_protos[l3proto] == NULL)) + return &nf_conntrack_generic_protocol; + + return nf_ct_protos[l3proto][protocol]; +} + +static int nf_conntrack_hash_rnd_initted; +static unsigned int nf_conntrack_hash_rnd; + +static u_int32_t __hash_conntrack(const struct nf_conntrack_tuple *tuple, + unsigned int size, unsigned int rnd) +{ + unsigned int a, b; + a = jhash((void *)tuple->src.u3.all, sizeof(tuple->src.u3.all), + ((tuple->src.l3num) << 16) | tuple->dst.protonum); + b = jhash((void *)tuple->dst.u3.all, sizeof(tuple->dst.u3.all), + (tuple->src.u.all << 16) | tuple->dst.u.all); + + return jhash_2words(a, b, rnd) % size; +} + +static inline u_int32_t hash_conntrack(const struct nf_conntrack_tuple *tuple) +{ + return __hash_conntrack(tuple, nf_conntrack_htable_size, + nf_conntrack_hash_rnd); +} + +/* Initialize "struct nf_conn" which has spaces for helper */ +static int +init_conntrack_for_helper(struct nf_conn *conntrack, u_int32_t features) +{ + + conntrack->help = (union nf_conntrack_help *) + (((unsigned long)conntrack->data + + (__alignof__(union nf_conntrack_help) - 1)) + & (~((unsigned long)(__alignof__(union nf_conntrack_help) -1)))); + return 0; +} + +int nf_conntrack_register_cache(u_int32_t features, const char *name, + size_t size, + int (*init)(struct nf_conn *, u_int32_t)) +{ + int ret = 0; + char *cache_name; + kmem_cache_t *cachep; + + DEBUGP("nf_conntrack_register_cache: features=0x%x, name=%s, size=%d\n", + features, name, size); + + if (features < NF_CT_F_BASIC || features >= NF_CT_F_NUM) { + DEBUGP("nf_conntrack_register_cache: invalid features.: 0x%x\n", + features); + return -EINVAL; + } + + down(&nf_ct_cache_mutex); + + write_lock_bh(&nf_ct_cache_lock); + /* e.g: multiple helpers are loaded */ + if (nf_ct_cache[features].use > 0) { + DEBUGP("nf_conntrack_register_cache: already resisterd.\n"); + if ((!strncmp(nf_ct_cache[features].name, name, + NF_CT_FEATURES_NAMELEN)) + && nf_ct_cache[features].size == size + && nf_ct_cache[features].init_conntrack == init) { + DEBUGP("nf_conntrack_register_cache: reusing.\n"); + nf_ct_cache[features].use++; + ret = 0; + } else + ret = -EBUSY; + + write_unlock_bh(&nf_ct_cache_lock); + up(&nf_ct_cache_mutex); + return ret; + } + write_unlock_bh(&nf_ct_cache_lock); + + /* + * The memory space for name of slab cache must be alive until + * cache is destroyed. + */ + cache_name = kmalloc(sizeof(char)*NF_CT_FEATURES_NAMELEN, GFP_ATOMIC); + if (cache_name == NULL) { + DEBUGP("nf_conntrack_register_cache: can't alloc cache_name\n"); + ret = -ENOMEM; + goto out_up_mutex; + } + + if (strlcpy(cache_name, name, NF_CT_FEATURES_NAMELEN) + >= NF_CT_FEATURES_NAMELEN) { + printk("nf_conntrack_register_cache: name too long\n"); + ret = -EINVAL; + goto out_free_name; + } + + cachep = kmem_cache_create(cache_name, size, 0, 0, + NULL, NULL); + if (!cachep) { + printk("nf_conntrack_register_cache: Can't create slab cache " + "for the features = 0x%x\n", features); + ret = -ENOMEM; + goto out_free_name; + } + + write_lock_bh(&nf_ct_cache_lock); + nf_ct_cache[features].use = 1; + nf_ct_cache[features].size = size; + nf_ct_cache[features].init_conntrack = init; + nf_ct_cache[features].cachep = cachep; + nf_ct_cache[features].name = cache_name; + write_unlock_bh(&nf_ct_cache_lock); + + goto out_up_mutex; + +out_free_name: + kfree(cache_name); +out_up_mutex: + up(&nf_ct_cache_mutex); + return ret; +} + +/* FIXME: In the current, only nf_conntrack_cleanup() can call this function. */ +void nf_conntrack_unregister_cache(u_int32_t features) +{ + kmem_cache_t *cachep; + char *name; + + /* + * This assures that kmem_cache_create() isn't called before destroying + * slab cache. + */ + DEBUGP("nf_conntrack_unregister_cache: 0x%04x\n", features); + down(&nf_ct_cache_mutex); + + write_lock_bh(&nf_ct_cache_lock); + if (--nf_ct_cache[features].use > 0) { + write_unlock_bh(&nf_ct_cache_lock); + up(&nf_ct_cache_mutex); + return; + } + cachep = nf_ct_cache[features].cachep; + name = nf_ct_cache[features].name; + nf_ct_cache[features].cachep = NULL; + nf_ct_cache[features].name = NULL; + nf_ct_cache[features].init_conntrack = NULL; + nf_ct_cache[features].size = 0; + write_unlock_bh(&nf_ct_cache_lock); + + synchronize_net(); + + kmem_cache_destroy(cachep); + kfree(name); + + up(&nf_ct_cache_mutex); +} + +int +nf_ct_get_tuple(const struct sk_buff *skb, + unsigned int nhoff, + unsigned int dataoff, + u_int16_t l3num, + u_int8_t protonum, + struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_l3proto *l3proto, + const struct nf_conntrack_protocol *protocol) +{ + NF_CT_TUPLE_U_BLANK(tuple); + + tuple->src.l3num = l3num; + if (l3proto->pkt_to_tuple(skb, nhoff, tuple) == 0) + return 0; + + tuple->dst.protonum = protonum; + tuple->dst.dir = IP_CT_DIR_ORIGINAL; + + return protocol->pkt_to_tuple(skb, dataoff, tuple); +} + +int +nf_ct_invert_tuple(struct nf_conntrack_tuple *inverse, + const struct nf_conntrack_tuple *orig, + const struct nf_conntrack_l3proto *l3proto, + const struct nf_conntrack_protocol *protocol) +{ + NF_CT_TUPLE_U_BLANK(inverse); + + inverse->src.l3num = orig->src.l3num; + if (l3proto->invert_tuple(inverse, orig) == 0) + return 0; + + inverse->dst.dir = !orig->dst.dir; + + inverse->dst.protonum = orig->dst.protonum; + return protocol->invert_tuple(inverse, orig); +} + +/* nf_conntrack_expect helper functions */ +static void nf_ct_unlink_expect(struct nf_conntrack_expect *exp) +{ + ASSERT_WRITE_LOCK(&nf_conntrack_lock); + NF_CT_ASSERT(!timer_pending(&exp_timeout)); + list_del(&exp->list); + NF_CT_STAT_INC(expect_delete); + exp->master->expecting--; + nf_conntrack_expect_put(exp); +} + +static void expectation_timed_out(unsigned long ul_expect) +{ + struct nf_conntrack_expect *exp = (void *)ul_expect; + + write_lock_bh(&nf_conntrack_lock); + nf_ct_unlink_expect(exp); + write_unlock_bh(&nf_conntrack_lock); + nf_conntrack_expect_put(exp); +} + +/* If an expectation for this connection is found, it gets delete from + * global list then returned. */ +static struct nf_conntrack_expect * +find_expectation(const struct nf_conntrack_tuple *tuple) +{ + struct nf_conntrack_expect *i; + + list_for_each_entry(i, &nf_conntrack_expect_list, list) { + /* If master is not in hash table yet (ie. packet hasn't left + this machine yet), how can other end know about expected? + Hence these are not the droids you are looking for (if + master ct never got confirmed, we'd hold a reference to it + and weird things would happen to future packets). */ + if (nf_ct_tuple_mask_cmp(tuple, &i->tuple, &i->mask) + && nf_ct_is_confirmed(i->master)) { + if (i->flags & NF_CT_EXPECT_PERMANENT) { + atomic_inc(&i->use); + return i; + } else if (del_timer(&i->timeout)) { + nf_ct_unlink_expect(i); + return i; + } + } + } + return NULL; +} + +/* delete all expectations for this conntrack */ +static void remove_expectations(struct nf_conn *ct) +{ + struct nf_conntrack_expect *i, *tmp; + + /* Optimization: most connection never expect any others. */ + if (ct->expecting == 0) + return; + + list_for_each_entry_safe(i, tmp, &nf_conntrack_expect_list, list) { + if (i->master == ct && del_timer(&i->timeout)) { + nf_ct_unlink_expect(i); + nf_conntrack_expect_put(i); + } + } +} + +static void +clean_from_lists(struct nf_conn *ct) +{ + unsigned int ho, hr; + + DEBUGP("clean_from_lists(%p)\n", ct); + ASSERT_WRITE_LOCK(&nf_conntrack_lock); + + ho = hash_conntrack(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple); + hr = hash_conntrack(&ct->tuplehash[IP_CT_DIR_REPLY].tuple); + LIST_DELETE(&nf_conntrack_hash[ho], &ct->tuplehash[IP_CT_DIR_ORIGINAL]); + LIST_DELETE(&nf_conntrack_hash[hr], &ct->tuplehash[IP_CT_DIR_REPLY]); + + /* Destroy all pending expectations */ + remove_expectations(ct); +} + +static void +destroy_conntrack(struct nf_conntrack *nfct) +{ + struct nf_conn *ct = (struct nf_conn *)nfct; + struct nf_conntrack_l3proto *l3proto; + struct nf_conntrack_protocol *proto; + + DEBUGP("destroy_conntrack(%p)\n", ct); + NF_CT_ASSERT(atomic_read(&nfct->use) == 0); + NF_CT_ASSERT(!timer_pending(&ct->timeout)); + + nf_conntrack_event(IPCT_DESTROY, ct); + set_bit(IPS_DYING_BIT, &ct->status); + + /* To make sure we don't get any weird locking issues here: + * destroy_conntrack() MUST NOT be called with a write lock + * to nf_conntrack_lock!!! -HW */ + l3proto = nf_ct_find_l3proto(ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.l3num); + if (l3proto && l3proto->destroy) + l3proto->destroy(ct); + + proto = nf_ct_find_proto(ct->tuplehash[IP_CT_DIR_REPLY].tuple.src.l3num, + ct->tuplehash[IP_CT_DIR_REPLY].tuple.dst.protonum); + if (proto && proto->destroy) + proto->destroy(ct); + + if (nf_conntrack_destroyed) + nf_conntrack_destroyed(ct); + + write_lock_bh(&nf_conntrack_lock); + /* Expectations will have been removed in clean_from_lists, + * except TFTP can create an expectation on the first packet, + * before connection is in the list, so we need to clean here, + * too. */ + remove_expectations(ct); + + /* We overload first tuple to link into unconfirmed list. */ + if (!nf_ct_is_confirmed(ct)) { + BUG_ON(list_empty(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list)); + list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list); + } + + NF_CT_STAT_INC(delete); + write_unlock_bh(&nf_conntrack_lock); + + if (ct->master) + nf_ct_put(ct->master); + + DEBUGP("destroy_conntrack: returning ct=%p to slab\n", ct); + nf_conntrack_free(ct); +} + +static void death_by_timeout(unsigned long ul_conntrack) +{ + struct nf_conn *ct = (void *)ul_conntrack; + + write_lock_bh(&nf_conntrack_lock); + /* Inside lock so preempt is disabled on module removal path. + * Otherwise we can get spurious warnings. */ + NF_CT_STAT_INC(delete_list); + clean_from_lists(ct); + write_unlock_bh(&nf_conntrack_lock); + nf_ct_put(ct); +} + +static inline int +conntrack_tuple_cmp(const struct nf_conntrack_tuple_hash *i, + const struct nf_conntrack_tuple *tuple, + const struct nf_conn *ignored_conntrack) +{ + ASSERT_READ_LOCK(&nf_conntrack_lock); + return nf_ct_tuplehash_to_ctrack(i) != ignored_conntrack + && nf_ct_tuple_equal(tuple, &i->tuple); +} + +static struct nf_conntrack_tuple_hash * +__nf_conntrack_find(const struct nf_conntrack_tuple *tuple, + const struct nf_conn *ignored_conntrack) +{ + struct nf_conntrack_tuple_hash *h; + unsigned int hash = hash_conntrack(tuple); + + ASSERT_READ_LOCK(&nf_conntrack_lock); + list_for_each_entry(h, &nf_conntrack_hash[hash], list) { + if (conntrack_tuple_cmp(h, tuple, ignored_conntrack)) { + NF_CT_STAT_INC(found); + return h; + } + NF_CT_STAT_INC(searched); + } + + return NULL; +} + +/* Find a connection corresponding to a tuple. */ +struct nf_conntrack_tuple_hash * +nf_conntrack_find_get(const struct nf_conntrack_tuple *tuple, + const struct nf_conn *ignored_conntrack) +{ + struct nf_conntrack_tuple_hash *h; + + read_lock_bh(&nf_conntrack_lock); + h = __nf_conntrack_find(tuple, ignored_conntrack); + if (h) + atomic_inc(&nf_ct_tuplehash_to_ctrack(h)->ct_general.use); + read_unlock_bh(&nf_conntrack_lock); + + return h; +} + +/* Confirm a connection given skb; places it in hash table */ +int +__nf_conntrack_confirm(struct sk_buff **pskb) +{ + unsigned int hash, repl_hash; + struct nf_conn *ct; + enum ip_conntrack_info ctinfo; + + ct = nf_ct_get(*pskb, &ctinfo); + + /* ipt_REJECT uses nf_conntrack_attach to attach related + ICMP/TCP RST packets in other direction. Actual packet + which created connection will be IP_CT_NEW or for an + expected connection, IP_CT_RELATED. */ + if (CTINFO2DIR(ctinfo) != IP_CT_DIR_ORIGINAL) + return NF_ACCEPT; + + hash = hash_conntrack(&ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple); + repl_hash = hash_conntrack(&ct->tuplehash[IP_CT_DIR_REPLY].tuple); + + /* We're not in hash table, and we refuse to set up related + connections for unconfirmed conns. But packet copies and + REJECT will give spurious warnings here. */ + /* NF_CT_ASSERT(atomic_read(&ct->ct_general.use) == 1); */ + + /* No external references means noone else could have + confirmed us. */ + NF_CT_ASSERT(!nf_ct_is_confirmed(ct)); + DEBUGP("Confirming conntrack %p\n", ct); + + write_lock_bh(&nf_conntrack_lock); + + /* See if there's one in the list already, including reverse: + NAT could have grabbed it without realizing, since we're + not in the hash. If there is, we lost race. */ + if (!LIST_FIND(&nf_conntrack_hash[hash], + conntrack_tuple_cmp, + struct nf_conntrack_tuple_hash *, + &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple, NULL) + && !LIST_FIND(&nf_conntrack_hash[repl_hash], + conntrack_tuple_cmp, + struct nf_conntrack_tuple_hash *, + &ct->tuplehash[IP_CT_DIR_REPLY].tuple, NULL)) { + /* Remove from unconfirmed list */ + list_del(&ct->tuplehash[IP_CT_DIR_ORIGINAL].list); + + list_prepend(&nf_conntrack_hash[hash], + &ct->tuplehash[IP_CT_DIR_ORIGINAL]); + list_prepend(&nf_conntrack_hash[repl_hash], + &ct->tuplehash[IP_CT_DIR_REPLY]); + /* Timer relative to confirmation time, not original + setting time, otherwise we'd get timer wrap in + weird delay cases. */ + ct->timeout.expires += jiffies; + add_timer(&ct->timeout); + atomic_inc(&ct->ct_general.use); + set_bit(IPS_CONFIRMED_BIT, &ct->status); + NF_CT_STAT_INC(insert); + write_unlock_bh(&nf_conntrack_lock); + if (ct->helper) + nf_conntrack_event_cache(IPCT_HELPER, *pskb); +#ifdef CONFIG_NF_NAT_NEEDED + if (test_bit(IPS_SRC_NAT_DONE_BIT, &ct->status) || + test_bit(IPS_DST_NAT_DONE_BIT, &ct->status)) + nf_conntrack_event_cache(IPCT_NATINFO, *pskb); +#endif + nf_conntrack_event_cache(master_ct(ct) ? + IPCT_RELATED : IPCT_NEW, *pskb); + return NF_ACCEPT; + } + + NF_CT_STAT_INC(insert_failed); + write_unlock_bh(&nf_conntrack_lock); + return NF_DROP; +} + +/* Returns true if a connection correspondings to the tuple (required + for NAT). */ +int +nf_conntrack_tuple_taken(const struct nf_conntrack_tuple *tuple, + const struct nf_conn *ignored_conntrack) +{ + struct nf_conntrack_tuple_hash *h; + + read_lock_bh(&nf_conntrack_lock); + h = __nf_conntrack_find(tuple, ignored_conntrack); + read_unlock_bh(&nf_conntrack_lock); + + return h != NULL; +} + +/* There's a small race here where we may free a just-assured + connection. Too bad: we're in trouble anyway. */ +static inline int unreplied(const struct nf_conntrack_tuple_hash *i) +{ + return !(test_bit(IPS_ASSURED_BIT, + &nf_ct_tuplehash_to_ctrack(i)->status)); +} + +static int early_drop(struct list_head *chain) +{ + /* Traverse backwards: gives us oldest, which is roughly LRU */ + struct nf_conntrack_tuple_hash *h; + struct nf_conn *ct = NULL; + int dropped = 0; + + read_lock_bh(&nf_conntrack_lock); + h = LIST_FIND_B(chain, unreplied, struct nf_conntrack_tuple_hash *); + if (h) { + ct = nf_ct_tuplehash_to_ctrack(h); + atomic_inc(&ct->ct_general.use); + } + read_unlock_bh(&nf_conntrack_lock); + + if (!ct) + return dropped; + + if (del_timer(&ct->timeout)) { + death_by_timeout((unsigned long)ct); + dropped = 1; + NF_CT_STAT_INC(early_drop); + } + nf_ct_put(ct); + return dropped; +} + +static inline int helper_cmp(const struct nf_conntrack_helper *i, + const struct nf_conntrack_tuple *rtuple) +{ + return nf_ct_tuple_mask_cmp(rtuple, &i->tuple, &i->mask); +} + +static struct nf_conntrack_helper * +nf_ct_find_helper(const struct nf_conntrack_tuple *tuple) +{ + return LIST_FIND(&helpers, helper_cmp, + struct nf_conntrack_helper *, + tuple); +} + +static struct nf_conn * +__nf_conntrack_alloc(const struct nf_conntrack_tuple *orig, + const struct nf_conntrack_tuple *repl, + const struct nf_conntrack_l3proto *l3proto) +{ + struct nf_conn *conntrack = NULL; + u_int32_t features = 0; + + if (!nf_conntrack_hash_rnd_initted) { + get_random_bytes(&nf_conntrack_hash_rnd, 4); + nf_conntrack_hash_rnd_initted = 1; + } + + if (nf_conntrack_max + && atomic_read(&nf_conntrack_count) >= nf_conntrack_max) { + unsigned int hash = hash_conntrack(orig); + /* Try dropping from this hash chain. */ + if (!early_drop(&nf_conntrack_hash[hash])) { + if (net_ratelimit()) + printk(KERN_WARNING + "nf_conntrack: table full, dropping" + " packet.\n"); + return ERR_PTR(-ENOMEM); + } + } + + /* find features needed by this conntrack. */ + features = l3proto->get_features(orig); + read_lock_bh(&nf_conntrack_lock); + if (nf_ct_find_helper(repl) != NULL) + features |= NF_CT_F_HELP; + read_unlock_bh(&nf_conntrack_lock); + + DEBUGP("nf_conntrack_alloc: features=0x%x\n", features); + + read_lock_bh(&nf_ct_cache_lock); + + if (!nf_ct_cache[features].use) { + DEBUGP("nf_conntrack_alloc: not supported features = 0x%x\n", + features); + goto out; + } + + conntrack = kmem_cache_alloc(nf_ct_cache[features].cachep, GFP_ATOMIC); + if (conntrack == NULL) { + DEBUGP("nf_conntrack_alloc: Can't alloc conntrack from cache\n"); + goto out; + } + + memset(conntrack, 0, nf_ct_cache[features].size); + conntrack->features = features; + if (nf_ct_cache[features].init_conntrack && + nf_ct_cache[features].init_conntrack(conntrack, features) < 0) { + DEBUGP("nf_conntrack_alloc: failed to init\n"); + kmem_cache_free(nf_ct_cache[features].cachep, conntrack); + conntrack = NULL; + goto out; + } + + atomic_set(&conntrack->ct_general.use, 1); + conntrack->ct_general.destroy = destroy_conntrack; + conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple = *orig; + conntrack->tuplehash[IP_CT_DIR_REPLY].tuple = *repl; + /* Don't set timer yet: wait for confirmation */ + init_timer(&conntrack->timeout); + conntrack->timeout.data = (unsigned long)conntrack; + conntrack->timeout.function = death_by_timeout; + + atomic_inc(&nf_conntrack_count); +out: + read_unlock_bh(&nf_ct_cache_lock); + return conntrack; +} + +struct nf_conn *nf_conntrack_alloc(const struct nf_conntrack_tuple *orig, + const struct nf_conntrack_tuple *repl) +{ + struct nf_conntrack_l3proto *l3proto; + + l3proto = nf_ct_find_l3proto(orig->src.l3num); + return __nf_conntrack_alloc(orig, repl, l3proto); +} + +void nf_conntrack_free(struct nf_conn *conntrack) +{ + u_int32_t features = conntrack->features; + NF_CT_ASSERT(features >= NF_CT_F_BASIC && features < NF_CT_F_NUM); + DEBUGP("nf_conntrack_free: features = 0x%x, conntrack=%p\n", features, + conntrack); + kmem_cache_free(nf_ct_cache[features].cachep, conntrack); + atomic_dec(&nf_conntrack_count); +} + +/* Allocate a new conntrack: we return -ENOMEM if classification + failed due to stress. Otherwise it really is unclassifiable. */ +static struct nf_conntrack_tuple_hash * +init_conntrack(const struct nf_conntrack_tuple *tuple, + struct nf_conntrack_l3proto *l3proto, + struct nf_conntrack_protocol *protocol, + struct sk_buff *skb, + unsigned int dataoff) +{ + struct nf_conn *conntrack; + struct nf_conntrack_tuple repl_tuple; + struct nf_conntrack_expect *exp; + + if (!nf_ct_invert_tuple(&repl_tuple, tuple, l3proto, protocol)) { + DEBUGP("Can't invert tuple.\n"); + return NULL; + } + + conntrack = __nf_conntrack_alloc(tuple, &repl_tuple, l3proto); + if (conntrack == NULL || IS_ERR(conntrack)) { + DEBUGP("Can't allocate conntrack.\n"); + return (struct nf_conntrack_tuple_hash *)conntrack; + } + + if (!protocol->new(conntrack, skb, dataoff)) { + nf_conntrack_free(conntrack); + DEBUGP("init conntrack: can't track with proto module\n"); + return NULL; + } + + write_lock_bh(&nf_conntrack_lock); + exp = find_expectation(tuple); + + if (exp) { + DEBUGP("conntrack: expectation arrives ct=%p exp=%p\n", + conntrack, exp); + /* Welcome, Mr. Bond. We've been expecting you... */ + __set_bit(IPS_EXPECTED_BIT, &conntrack->status); + conntrack->master = exp->master; +#ifdef CONFIG_NF_CONNTRACK_MARK + conntrack->mark = exp->master->mark; +#endif + nf_conntrack_get(&conntrack->master->ct_general); + NF_CT_STAT_INC(expect_new); + } else { + conntrack->helper = nf_ct_find_helper(&repl_tuple); + + NF_CT_STAT_INC(new); + } + + /* Overload tuple linked list to put us in unconfirmed list. */ + list_add(&conntrack->tuplehash[IP_CT_DIR_ORIGINAL].list, &unconfirmed); + + write_unlock_bh(&nf_conntrack_lock); + + if (exp) { + if (exp->expectfn) + exp->expectfn(conntrack, exp); + nf_conntrack_expect_put(exp); + } + + return &conntrack->tuplehash[IP_CT_DIR_ORIGINAL]; +} + +/* On success, returns conntrack ptr, sets skb->nfct and ctinfo */ +static inline struct nf_conn * +resolve_normal_ct(struct sk_buff *skb, + unsigned int dataoff, + u_int16_t l3num, + u_int8_t protonum, + struct nf_conntrack_l3proto *l3proto, + struct nf_conntrack_protocol *proto, + int *set_reply, + enum ip_conntrack_info *ctinfo) +{ + struct nf_conntrack_tuple tuple; + struct nf_conntrack_tuple_hash *h; + struct nf_conn *ct; + + if (!nf_ct_get_tuple(skb, (unsigned int)(skb->nh.raw - skb->data), + dataoff, l3num, protonum, &tuple, l3proto, + proto)) { + DEBUGP("resolve_normal_ct: Can't get tuple\n"); + return NULL; + } + + /* look for tuple match */ + h = nf_conntrack_find_get(&tuple, NULL); + if (!h) { + h = init_conntrack(&tuple, l3proto, proto, skb, dataoff); + if (!h) + return NULL; + if (IS_ERR(h)) + return (void *)h; + } + ct = nf_ct_tuplehash_to_ctrack(h); + + /* It exists; we have (non-exclusive) reference. */ + if (NF_CT_DIRECTION(h) == IP_CT_DIR_REPLY) { + *ctinfo = IP_CT_ESTABLISHED + IP_CT_IS_REPLY; + /* Please set reply bit if this packet OK */ + *set_reply = 1; + } else { + /* Once we've had two way comms, always ESTABLISHED. */ + if (test_bit(IPS_SEEN_REPLY_BIT, &ct->status)) { + DEBUGP("nf_conntrack_in: normal packet for %p\n", ct); + *ctinfo = IP_CT_ESTABLISHED; + } else if (test_bit(IPS_EXPECTED_BIT, &ct->status)) { + DEBUGP("nf_conntrack_in: related packet for %p\n", ct); + *ctinfo = IP_CT_RELATED; + } else { + DEBUGP("nf_conntrack_in: new packet for %p\n", ct); + *ctinfo = IP_CT_NEW; + } + *set_reply = 0; + } + skb->nfct = &ct->ct_general; + skb->nfctinfo = *ctinfo; + return ct; +} + +unsigned int +nf_conntrack_in(int pf, unsigned int hooknum, struct sk_buff **pskb) +{ + struct nf_conn *ct; + enum ip_conntrack_info ctinfo; + struct nf_conntrack_l3proto *l3proto; + struct nf_conntrack_protocol *proto; + unsigned int dataoff; + u_int8_t protonum; + int set_reply = 0; + int ret; + + /* Previously seen (loopback or untracked)? Ignore. */ + if ((*pskb)->nfct) { + NF_CT_STAT_INC(ignore); + return NF_ACCEPT; + } + + l3proto = nf_ct_find_l3proto((u_int16_t)pf); + if ((ret = l3proto->prepare(pskb, hooknum, &dataoff, &protonum)) <= 0) { + DEBUGP("not prepared to track yet or error occured\n"); + return -ret; + } + + proto = nf_ct_find_proto((u_int16_t)pf, protonum); + + /* It may be an special packet, error, unclean... + * inverse of the return code tells to the netfilter + * core what to do with the packet. */ + if (proto->error != NULL && + (ret = proto->error(*pskb, dataoff, &ctinfo, pf, hooknum)) <= 0) { + NF_CT_STAT_INC(error); + NF_CT_STAT_INC(invalid); + return -ret; + } + + ct = resolve_normal_ct(*pskb, dataoff, pf, protonum, l3proto, proto, + &set_reply, &ctinfo); + if (!ct) { + /* Not valid part of a connection */ + NF_CT_STAT_INC(invalid); + return NF_ACCEPT; + } + + if (IS_ERR(ct)) { + /* Too stressed to deal. */ + NF_CT_STAT_INC(drop); + return NF_DROP; + } + + NF_CT_ASSERT((*pskb)->nfct); + + ret = proto->packet(ct, *pskb, dataoff, ctinfo, pf, hooknum); + if (ret < 0) { + /* Invalid: inverse of the return code tells + * the netfilter core what to do */ + DEBUGP("nf_conntrack_in: Can't track with proto module\n"); + nf_conntrack_put((*pskb)->nfct); + (*pskb)->nfct = NULL; + NF_CT_STAT_INC(invalid); + return -ret; + } + + if (set_reply && !test_and_set_bit(IPS_SEEN_REPLY_BIT, &ct->status)) + nf_conntrack_event_cache(IPCT_STATUS, *pskb); + + return ret; +} + +int nf_ct_invert_tuplepr(struct nf_conntrack_tuple *inverse, + const struct nf_conntrack_tuple *orig) +{ + return nf_ct_invert_tuple(inverse, orig, + nf_ct_find_l3proto(orig->src.l3num), + nf_ct_find_proto(orig->src.l3num, + orig->dst.protonum)); +} + +/* Would two expected things clash? */ +static inline int expect_clash(const struct nf_conntrack_expect *a, + const struct nf_conntrack_expect *b) +{ + /* Part covered by intersection of masks must be unequal, + otherwise they clash */ + struct nf_conntrack_tuple intersect_mask; + int count; + + intersect_mask.src.l3num = a->mask.src.l3num & b->mask.src.l3num; + intersect_mask.src.u.all = a->mask.src.u.all & b->mask.src.u.all; + intersect_mask.dst.u.all = a->mask.dst.u.all & b->mask.dst.u.all; + intersect_mask.dst.protonum = a->mask.dst.protonum + & b->mask.dst.protonum; + + for (count = 0; count < NF_CT_TUPLE_L3SIZE; count++){ + intersect_mask.src.u3.all[count] = + a->mask.src.u3.all[count] & b->mask.src.u3.all[count]; + } + + for (count = 0; count < NF_CT_TUPLE_L3SIZE; count++){ + intersect_mask.dst.u3.all[count] = + a->mask.dst.u3.all[count] & b->mask.dst.u3.all[count]; + } + + return nf_ct_tuple_mask_cmp(&a->tuple, &b->tuple, &intersect_mask); +} + +static inline int expect_matches(const struct nf_conntrack_expect *a, + const struct nf_conntrack_expect *b) +{ + return a->master == b->master + && nf_ct_tuple_equal(&a->tuple, &b->tuple) + && nf_ct_tuple_equal(&a->mask, &b->mask); +} + +/* Generally a bad idea to call this: could have matched already. */ +void nf_conntrack_unexpect_related(struct nf_conntrack_expect *exp) +{ + struct nf_conntrack_expect *i; + + write_lock_bh(&nf_conntrack_lock); + /* choose the the oldest expectation to evict */ + list_for_each_entry_reverse(i, &nf_conntrack_expect_list, list) { + if (expect_matches(i, exp) && del_timer(&i->timeout)) { + nf_ct_unlink_expect(i); + write_unlock_bh(&nf_conntrack_lock); + nf_conntrack_expect_put(i); + return; + } + } + write_unlock_bh(&nf_conntrack_lock); +} + +/* We don't increase the master conntrack refcount for non-fulfilled + * conntracks. During the conntrack destruction, the expectations are + * always killed before the conntrack itself */ +struct nf_conntrack_expect *nf_conntrack_expect_alloc(struct nf_conn *me) +{ + struct nf_conntrack_expect *new; + + new = kmem_cache_alloc(nf_conntrack_expect_cachep, GFP_ATOMIC); + if (!new) { + DEBUGP("expect_related: OOM allocating expect\n"); + return NULL; + } + new->master = me; + atomic_set(&new->use, 1); + return new; +} + +void nf_conntrack_expect_put(struct nf_conntrack_expect *exp) +{ + if (atomic_dec_and_test(&exp->use)) + kmem_cache_free(nf_conntrack_expect_cachep, exp); +} + +static void nf_conntrack_expect_insert(struct nf_conntrack_expect *exp) +{ + atomic_inc(&exp->use); + exp->master->expecting++; + list_add(&exp->list, &nf_conntrack_expect_list); + + init_timer(&exp->timeout); + exp->timeout.data = (unsigned long)exp; + exp->timeout.function = expectation_timed_out; + exp->timeout.expires = jiffies + exp->master->helper->timeout * HZ; + add_timer(&exp->timeout); + + atomic_inc(&exp->use); + NF_CT_STAT_INC(expect_create); +} + +/* Race with expectations being used means we could have none to find; OK. */ +static void evict_oldest_expect(struct nf_conn *master) +{ + struct nf_conntrack_expect *i; + + list_for_each_entry_reverse(i, &nf_conntrack_expect_list, list) { + if (i->master == master) { + if (del_timer(&i->timeout)) { + nf_ct_unlink_expect(i); + nf_conntrack_expect_put(i); + } + break; + } + } +} + +static inline int refresh_timer(struct nf_conntrack_expect *i) +{ + if (!del_timer(&i->timeout)) + return 0; + + i->timeout.expires = jiffies + i->master->helper->timeout*HZ; + add_timer(&i->timeout); + return 1; +} + +int nf_conntrack_expect_related(struct nf_conntrack_expect *expect) +{ + struct nf_conntrack_expect *i; + int ret; + + DEBUGP("nf_conntrack_expect_related %p\n", related_to); + DEBUGP("tuple: "); NF_CT_DUMP_TUPLE(&expect->tuple); + DEBUGP("mask: "); NF_CT_DUMP_TUPLE(&expect->mask); + + write_lock_bh(&nf_conntrack_lock); + list_for_each_entry(i, &nf_conntrack_expect_list, list) { + if (expect_matches(i, expect)) { + /* Refresh timer: if it's dying, ignore.. */ + if (refresh_timer(i)) { + ret = 0; + goto out; + } + } else if (expect_clash(i, expect)) { + ret = -EBUSY; + goto out; + } + } + /* Will be over limit? */ + if (expect->master->helper->max_expected && + expect->master->expecting >= expect->master->helper->max_expected) + evict_oldest_expect(expect->master); + + nf_conntrack_expect_insert(expect); + nf_conntrack_expect_event(IPEXP_NEW, expect); + ret = 0; +out: + write_unlock_bh(&nf_conntrack_lock); + return ret; +} + +/* Alter reply tuple (maybe alter helper). This is for NAT, and is + implicitly racy: see __nf_conntrack_confirm */ +void nf_conntrack_alter_reply(struct nf_conn *conntrack, + const struct nf_conntrack_tuple *newreply) +{ + write_lock_bh(&nf_conntrack_lock); + /* Should be unconfirmed, so not in hash table yet */ + NF_CT_ASSERT(!nf_ct_is_confirmed(conntrack)); + + DEBUGP("Altering reply tuple of %p to ", conntrack); + NF_CT_DUMP_TUPLE(newreply); + + conntrack->tuplehash[IP_CT_DIR_REPLY].tuple = *newreply; + if (!conntrack->master && conntrack->expecting == 0) + conntrack->helper = nf_ct_find_helper(newreply); + write_unlock_bh(&nf_conntrack_lock); +} + +int nf_conntrack_helper_register(struct nf_conntrack_helper *me) +{ + int ret; + BUG_ON(me->timeout == 0); + + ret = nf_conntrack_register_cache(NF_CT_F_HELP, "nf_conntrack:help", + sizeof(struct nf_conn) + + sizeof(union nf_conntrack_help) + + __alignof__(union nf_conntrack_help), + init_conntrack_for_helper); + if (ret < 0) { + printk(KERN_ERR "nf_conntrack_helper_reigster: Unable to create slab cache for conntracks\n"); + return ret; + } + write_lock_bh(&nf_conntrack_lock); + list_prepend(&helpers, me); + write_unlock_bh(&nf_conntrack_lock); + + return 0; +} + +static inline int unhelp(struct nf_conntrack_tuple_hash *i, + const struct nf_conntrack_helper *me) +{ + if (nf_ct_tuplehash_to_ctrack(i)->helper == me) { + nf_conntrack_event(IPCT_HELPER, nf_ct_tuplehash_to_ctrack(i)); + nf_ct_tuplehash_to_ctrack(i)->helper = NULL; + } + return 0; +} + +void nf_conntrack_helper_unregister(struct nf_conntrack_helper *me) +{ + unsigned int i; + struct nf_conntrack_expect *exp, *tmp; + + /* Need write lock here, to delete helper. */ + write_lock_bh(&nf_conntrack_lock); + LIST_DELETE(&helpers, me); + + /* Get rid of expectations */ + list_for_each_entry_safe(exp, tmp, &nf_conntrack_expect_list, list) { + if (exp->master->helper == me && del_timer(&exp->timeout)) { + nf_ct_unlink_expect(exp); + nf_conntrack_expect_put(exp); + } + } + + /* Get rid of expecteds, set helpers to NULL. */ + LIST_FIND_W(&unconfirmed, unhelp, struct nf_conntrack_tuple_hash*, me); + for (i = 0; i < nf_conntrack_htable_size; i++) + LIST_FIND_W(&nf_conntrack_hash[i], unhelp, + struct nf_conntrack_tuple_hash *, me); + write_unlock_bh(&nf_conntrack_lock); + + /* Someone could be still looking at the helper in a bh. */ + synchronize_net(); +} + +/* Refresh conntrack for this many jiffies and do accounting if do_acct is 1 */ +void __nf_ct_refresh_acct(struct nf_conn *ct, + enum ip_conntrack_info ctinfo, + const struct sk_buff *skb, + unsigned long extra_jiffies, + int do_acct) +{ + int event = 0; + + NF_CT_ASSERT(ct->timeout.data == (unsigned long)ct); + NF_CT_ASSERT(skb); + + write_lock_bh(&nf_conntrack_lock); + + /* If not in hash table, timer will not be active yet */ + if (!nf_ct_is_confirmed(ct)) { + ct->timeout.expires = extra_jiffies; + event = IPCT_REFRESH; + } else { + /* Need del_timer for race avoidance (may already be dying). */ + if (del_timer(&ct->timeout)) { + ct->timeout.expires = jiffies + extra_jiffies; + add_timer(&ct->timeout); + event = IPCT_REFRESH; + } + } + +#ifdef CONFIG_NF_CT_ACCT + if (do_acct) { + ct->counters[CTINFO2DIR(ctinfo)].packets++; + ct->counters[CTINFO2DIR(ctinfo)].bytes += + skb->len - (unsigned int)(skb->nh.raw - skb->data); + if ((ct->counters[CTINFO2DIR(ctinfo)].packets & 0x80000000) + || (ct->counters[CTINFO2DIR(ctinfo)].bytes & 0x80000000)) + event |= IPCT_COUNTER_FILLING; + } +#endif + + write_unlock_bh(&nf_conntrack_lock); + + /* must be unlocked when calling event cache */ + if (event) + nf_conntrack_event_cache(event, skb); +} + +/* Used by ipt_REJECT and ip6t_REJECT. */ +void __nf_conntrack_attach(struct sk_buff *nskb, struct sk_buff *skb) +{ + struct nf_conn *ct; + enum ip_conntrack_info ctinfo; + + /* This ICMP is in reverse direction to the packet which caused it */ + ct = nf_ct_get(skb, &ctinfo); + if (CTINFO2DIR(ctinfo) == IP_CT_DIR_ORIGINAL) + ctinfo = IP_CT_RELATED + IP_CT_IS_REPLY; + else + ctinfo = IP_CT_RELATED; + + /* Attach to new skbuff, and increment count */ + nskb->nfct = &ct->ct_general; + nskb->nfctinfo = ctinfo; + nf_conntrack_get(nskb->nfct); +} + +static inline int +do_iter(const struct nf_conntrack_tuple_hash *i, + int (*iter)(struct nf_conn *i, void *data), + void *data) +{ + return iter(nf_ct_tuplehash_to_ctrack(i), data); +} + +/* Bring out ya dead! */ +static struct nf_conntrack_tuple_hash * +get_next_corpse(int (*iter)(struct nf_conn *i, void *data), + void *data, unsigned int *bucket) +{ + struct nf_conntrack_tuple_hash *h = NULL; + + write_lock_bh(&nf_conntrack_lock); + for (; *bucket < nf_conntrack_htable_size; (*bucket)++) { + h = LIST_FIND_W(&nf_conntrack_hash[*bucket], do_iter, + struct nf_conntrack_tuple_hash *, iter, data); + if (h) + break; + } + if (!h) + h = LIST_FIND_W(&unconfirmed, do_iter, + struct nf_conntrack_tuple_hash *, iter, data); + if (h) + atomic_inc(&nf_ct_tuplehash_to_ctrack(h)->ct_general.use); + write_unlock_bh(&nf_conntrack_lock); + + return h; +} + +void +nf_ct_iterate_cleanup(int (*iter)(struct nf_conn *i, void *data), void *data) +{ + struct nf_conntrack_tuple_hash *h; + unsigned int bucket = 0; + + while ((h = get_next_corpse(iter, data, &bucket)) != NULL) { + struct nf_conn *ct = nf_ct_tuplehash_to_ctrack(h); + /* Time to push up daises... */ + if (del_timer(&ct->timeout)) + death_by_timeout((unsigned long)ct); + /* ... else the timer will get him soon. */ + + nf_ct_put(ct); + } +} + +static int kill_all(struct nf_conn *i, void *data) +{ + return 1; +} + +static void free_conntrack_hash(struct list_head *hash, int vmalloced, int size) +{ + if (vmalloced) + vfree(hash); + else + free_pages((unsigned long)hash, + get_order(sizeof(struct list_head) * size)); +} + +/* Mishearing the voices in his head, our hero wonders how he's + supposed to kill the mall. */ +void nf_conntrack_cleanup(void) +{ + int i; + + /* This makes sure all current packets have passed through + netfilter framework. Roll on, two-stage module + delete... */ + synchronize_net(); + + nf_ct_event_cache_flush(); + i_see_dead_people: + nf_ct_iterate_cleanup(kill_all, NULL); + if (atomic_read(&nf_conntrack_count) != 0) { + schedule(); + goto i_see_dead_people; + } + + for (i = 0; i < NF_CT_F_NUM; i++) { + if (nf_ct_cache[i].use == 0) + continue; + + NF_CT_ASSERT(nf_ct_cache[i].use == 1); + nf_ct_cache[i].use = 1; + nf_conntrack_unregister_cache(i); + } + kmem_cache_destroy(nf_conntrack_expect_cachep); + free_conntrack_hash(nf_conntrack_hash, nf_conntrack_vmalloc, + nf_conntrack_htable_size); +} + +static struct list_head *alloc_hashtable(int size, int *vmalloced) +{ + struct list_head *hash; + unsigned int i; + + *vmalloced = 0; + hash = (void*)__get_free_pages(GFP_KERNEL, + get_order(sizeof(struct list_head) + * size)); + if (!hash) { + *vmalloced = 1; + printk(KERN_WARNING "nf_conntrack: falling back to vmalloc.\n"); + hash = vmalloc(sizeof(struct list_head) * size); + } + + if (hash) + for (i = 0; i < size; i++) + INIT_LIST_HEAD(&hash[i]); + + return hash; +} + +int set_hashsize(const char *val, struct kernel_param *kp) +{ + int i, bucket, hashsize, vmalloced; + int old_vmalloced, old_size; + int rnd; + struct list_head *hash, *old_hash; + struct nf_conntrack_tuple_hash *h; + + /* On boot, we can set this without any fancy locking. */ + if (!nf_conntrack_htable_size) + return param_set_uint(val, kp); + + hashsize = simple_strtol(val, NULL, 0); + if (!hashsize) + return -EINVAL; + + hash = alloc_hashtable(hashsize, &vmalloced); + if (!hash) + return -ENOMEM; + + /* We have to rehahs for the new table anyway, so we also can + * use a newrandom seed */ + get_random_bytes(&rnd, 4); + + write_lock_bh(&nf_conntrack_lock); + for (i = 0; i < nf_conntrack_htable_size; i++) { + while (!list_empty(&nf_conntrack_hash[i])) { + h = list_entry(nf_conntrack_hash[i].next, + struct nf_conntrack_tuple_hash, list); + list_del(&h->list); + bucket = __hash_conntrack(&h->tuple, hashsize, rnd); + list_add_tail(&h->list, &hash[bucket]); + } + } + old_size = nf_conntrack_htable_size; + old_vmalloced = nf_conntrack_vmalloc; + old_hash = nf_conntrack_hash; + + nf_conntrack_htable_size = hashsize; + nf_conntrack_vmalloc = vmalloced; + nf_conntrack_hash = hash; + nf_conntrack_hash_rnd = rnd; + write_unlock_bh(&nf_conntrack_lock); + + free_conntrack_hash(old_hash, old_vmalloced, old_size); + return 0; +} + +module_param_call(hashsize, set_hashsize, param_get_uint, + &nf_conntrack_htable_size, 0600); + +int __init nf_conntrack_init(void) +{ + unsigned int i; + int ret; + + /* Idea from tcp.c: use 1/16384 of memory. On i386: 32MB + * machine has 256 buckets. >= 1GB machines have 8192 buckets. */ + if (!nf_conntrack_htable_size) { + nf_conntrack_htable_size + = (((num_physpages << PAGE_SHIFT) / 16384) + / sizeof(struct list_head)); + if (num_physpages > (1024 * 1024 * 1024 / PAGE_SIZE)) + nf_conntrack_htable_size = 8192; + if (nf_conntrack_htable_size < 16) + nf_conntrack_htable_size = 16; + } + nf_conntrack_max = 8 * nf_conntrack_htable_size; + + printk("nf_conntrack version %s (%u buckets, %d max)\n", + NF_CONNTRACK_VERSION, nf_conntrack_htable_size, + nf_conntrack_max); + + nf_conntrack_hash = alloc_hashtable(nf_conntrack_htable_size, + &nf_conntrack_vmalloc); + if (!nf_conntrack_hash) { + printk(KERN_ERR "Unable to create nf_conntrack_hash\n"); + goto err_out; + } + + ret = nf_conntrack_register_cache(NF_CT_F_BASIC, "nf_conntrack:basic", + sizeof(struct nf_conn), NULL); + if (ret < 0) { + printk(KERN_ERR "Unable to create nf_conn slab cache\n"); + goto err_free_hash; + } + + nf_conntrack_expect_cachep = kmem_cache_create("nf_conntrack_expect", + sizeof(struct nf_conntrack_expect), + 0, 0, NULL, NULL); + if (!nf_conntrack_expect_cachep) { + printk(KERN_ERR "Unable to create nf_expect slab cache\n"); + goto err_free_conntrack_slab; + } + + /* Don't NEED lock here, but good form anyway. */ + write_lock_bh(&nf_conntrack_lock); + for (i = 0; i < PF_MAX; i++) + nf_ct_l3protos[i] = &nf_conntrack_generic_l3proto; + write_unlock_bh(&nf_conntrack_lock); + + /* Set up fake conntrack: + - to never be deleted, not in any hashes */ + atomic_set(&nf_conntrack_untracked.ct_general.use, 1); + /* - and look it like as a confirmed connection */ + set_bit(IPS_CONFIRMED_BIT, &nf_conntrack_untracked.status); + + return ret; + +err_free_conntrack_slab: + nf_conntrack_unregister_cache(NF_CT_F_BASIC); +err_free_hash: + free_conntrack_hash(nf_conntrack_hash, nf_conntrack_vmalloc, + nf_conntrack_htable_size); +err_out: + return -ENOMEM; +} diff --git a/net/netfilter/nf_conntrack_ftp.c b/net/netfilter/nf_conntrack_ftp.c new file mode 100644 index 000000000000..65080e269f27 --- /dev/null +++ b/net/netfilter/nf_conntrack_ftp.c @@ -0,0 +1,698 @@ +/* FTP extension for connection tracking. */ + +/* (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * (C) 2003,2004 USAGI/WIDE Project + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - enable working with Layer 3 protocol independent connection tracking. + * - track EPRT and EPSV commands with IPv6 address. + * + * Derived from net/ipv4/netfilter/ip_conntrack_ftp.c + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Rusty Russell "); +MODULE_DESCRIPTION("ftp connection tracking helper"); + +/* This is slow, but it's simple. --RR */ +static char *ftp_buffer; + +static DEFINE_SPINLOCK(nf_ftp_lock); + +#define MAX_PORTS 8 +static u_int16_t ports[MAX_PORTS]; +static unsigned int ports_c; +module_param_array(ports, ushort, &ports_c, 0400); + +static int loose; +module_param(loose, int, 0600); + +unsigned int (*nf_nat_ftp_hook)(struct sk_buff **pskb, + enum ip_conntrack_info ctinfo, + enum ip_ct_ftp_type type, + unsigned int matchoff, + unsigned int matchlen, + struct nf_conntrack_expect *exp, + u32 *seq); +EXPORT_SYMBOL_GPL(nf_nat_ftp_hook); + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +static int try_rfc959(const char *, size_t, struct nf_conntrack_man *, char); +static int try_eprt(const char *, size_t, struct nf_conntrack_man *, char); +static int try_epsv_response(const char *, size_t, struct nf_conntrack_man *, + char); + +static struct ftp_search { + enum ip_conntrack_dir dir; + const char *pattern; + size_t plen; + char skip; + char term; + enum ip_ct_ftp_type ftptype; + int (*getnum)(const char *, size_t, struct nf_conntrack_man *, char); +} search[] = { + { + IP_CT_DIR_ORIGINAL, + "PORT", sizeof("PORT") - 1, ' ', '\r', + IP_CT_FTP_PORT, + try_rfc959, + }, + { + IP_CT_DIR_REPLY, + "227 ", sizeof("227 ") - 1, '(', ')', + IP_CT_FTP_PASV, + try_rfc959, + }, + { + IP_CT_DIR_ORIGINAL, + "EPRT", sizeof("EPRT") - 1, ' ', '\r', + IP_CT_FTP_EPRT, + try_eprt, + }, + { + IP_CT_DIR_REPLY, + "229 ", sizeof("229 ") - 1, '(', ')', + IP_CT_FTP_EPSV, + try_epsv_response, + }, +}; + +/* This code is based on inet_pton() in glibc-2.2.4 */ +static int +get_ipv6_addr(const char *src, size_t dlen, struct in6_addr *dst, u_int8_t term) +{ + static const char xdigits[] = "0123456789abcdef"; + u_int8_t tmp[16], *tp, *endp, *colonp; + int ch, saw_xdigit; + u_int32_t val; + size_t clen = 0; + + tp = memset(tmp, '\0', sizeof(tmp)); + endp = tp + sizeof(tmp); + colonp = NULL; + + /* Leading :: requires some special handling. */ + if (*src == ':'){ + if (*++src != ':') { + DEBUGP("invalid \":\" at the head of addr\n"); + return 0; + } + clen++; + } + + saw_xdigit = 0; + val = 0; + while ((clen < dlen) && (*src != term)) { + const char *pch; + + ch = tolower(*src++); + clen++; + + pch = strchr(xdigits, ch); + if (pch != NULL) { + val <<= 4; + val |= (pch - xdigits); + if (val > 0xffff) + return 0; + + saw_xdigit = 1; + continue; + } + if (ch != ':') { + DEBUGP("get_ipv6_addr: invalid char. \'%c\'\n", ch); + return 0; + } + + if (!saw_xdigit) { + if (colonp) { + DEBUGP("invalid location of \"::\".\n"); + return 0; + } + colonp = tp; + continue; + } else if (*src == term) { + DEBUGP("trancated IPv6 addr\n"); + return 0; + } + + if (tp + 2 > endp) + return 0; + *tp++ = (u_int8_t) (val >> 8) & 0xff; + *tp++ = (u_int8_t) val & 0xff; + + saw_xdigit = 0; + val = 0; + continue; + } + if (saw_xdigit) { + if (tp + 2 > endp) + return 0; + *tp++ = (u_int8_t) (val >> 8) & 0xff; + *tp++ = (u_int8_t) val & 0xff; + } + if (colonp != NULL) { + /* + * Since some memmove()'s erroneously fail to handle + * overlapping regions, we'll do the shift by hand. + */ + const int n = tp - colonp; + int i; + + if (tp == endp) + return 0; + + for (i = 1; i <= n; i++) { + endp[- i] = colonp[n - i]; + colonp[n - i] = 0; + } + tp = endp; + } + if (tp != endp || (*src != term)) + return 0; + + memcpy(dst->s6_addr, tmp, sizeof(dst->s6_addr)); + return clen; +} + +static int try_number(const char *data, size_t dlen, u_int32_t array[], + int array_size, char sep, char term) +{ + u_int32_t i, len; + + memset(array, 0, sizeof(array[0])*array_size); + + /* Keep data pointing at next char. */ + for (i = 0, len = 0; len < dlen && i < array_size; len++, data++) { + if (*data >= '0' && *data <= '9') { + array[i] = array[i]*10 + *data - '0'; + } + else if (*data == sep) + i++; + else { + /* Unexpected character; true if it's the + terminator and we're finished. */ + if (*data == term && i == array_size - 1) + return len; + + DEBUGP("Char %u (got %u nums) `%u' unexpected\n", + len, i, *data); + return 0; + } + } + DEBUGP("Failed to fill %u numbers separated by %c\n", array_size, sep); + + return 0; +} + +/* Returns 0, or length of numbers: 192,168,1,1,5,6 */ +static int try_rfc959(const char *data, size_t dlen, + struct nf_conntrack_man *cmd, char term) +{ + int length; + u_int32_t array[6]; + + length = try_number(data, dlen, array, 6, ',', term); + if (length == 0) + return 0; + + cmd->u3.ip = htonl((array[0] << 24) | (array[1] << 16) | + (array[2] << 8) | array[3]); + cmd->u.tcp.port = htons((array[4] << 8) | array[5]); + return length; +} + +/* Grab port: number up to delimiter */ +static int get_port(const char *data, int start, size_t dlen, char delim, + u_int16_t *port) +{ + u_int16_t tmp_port = 0; + int i; + + for (i = start; i < dlen; i++) { + /* Finished? */ + if (data[i] == delim) { + if (tmp_port == 0) + break; + *port = htons(tmp_port); + DEBUGP("get_port: return %d\n", tmp_port); + return i + 1; + } + else if (data[i] >= '0' && data[i] <= '9') + tmp_port = tmp_port*10 + data[i] - '0'; + else { /* Some other crap */ + DEBUGP("get_port: invalid char.\n"); + break; + } + } + return 0; +} + +/* Returns 0, or length of numbers: |1|132.235.1.2|6275| or |2|3ffe::1|6275| */ +static int try_eprt(const char *data, size_t dlen, struct nf_conntrack_man *cmd, + char term) +{ + char delim; + int length; + + /* First character is delimiter, then "1" for IPv4 or "2" for IPv6, + then delimiter again. */ + if (dlen <= 3) { + DEBUGP("EPRT: too short\n"); + return 0; + } + delim = data[0]; + if (isdigit(delim) || delim < 33 || delim > 126 || data[2] != delim) { + DEBUGP("try_eprt: invalid delimitter.\n"); + return 0; + } + + if ((cmd->l3num == PF_INET && data[1] != '1') || + (cmd->l3num == PF_INET6 && data[1] != '2')) { + DEBUGP("EPRT: invalid protocol number.\n"); + return 0; + } + + DEBUGP("EPRT: Got %c%c%c\n", delim, data[1], delim); + + if (data[1] == '1') { + u_int32_t array[4]; + + /* Now we have IP address. */ + length = try_number(data + 3, dlen - 3, array, 4, '.', delim); + if (length != 0) + cmd->u3.ip = htonl((array[0] << 24) | (array[1] << 16) + | (array[2] << 8) | array[3]); + } else { + /* Now we have IPv6 address. */ + length = get_ipv6_addr(data + 3, dlen - 3, + (struct in6_addr *)cmd->u3.ip6, delim); + } + + if (length == 0) + return 0; + DEBUGP("EPRT: Got IP address!\n"); + /* Start offset includes initial "|1|", and trailing delimiter */ + return get_port(data, 3 + length + 1, dlen, delim, &cmd->u.tcp.port); +} + +/* Returns 0, or length of numbers: |||6446| */ +static int try_epsv_response(const char *data, size_t dlen, + struct nf_conntrack_man *cmd, char term) +{ + char delim; + + /* Three delimiters. */ + if (dlen <= 3) return 0; + delim = data[0]; + if (isdigit(delim) || delim < 33 || delim > 126 + || data[1] != delim || data[2] != delim) + return 0; + + return get_port(data, 3, dlen, delim, &cmd->u.tcp.port); +} + +/* Return 1 for match, 0 for accept, -1 for partial. */ +static int find_pattern(const char *data, size_t dlen, + const char *pattern, size_t plen, + char skip, char term, + unsigned int *numoff, + unsigned int *numlen, + struct nf_conntrack_man *cmd, + int (*getnum)(const char *, size_t, + struct nf_conntrack_man *, char)) +{ + size_t i; + + DEBUGP("find_pattern `%s': dlen = %u\n", pattern, dlen); + if (dlen == 0) + return 0; + + if (dlen <= plen) { + /* Short packet: try for partial? */ + if (strnicmp(data, pattern, dlen) == 0) + return -1; + else return 0; + } + + if (strnicmp(data, pattern, plen) != 0) { +#if 0 + size_t i; + + DEBUGP("ftp: string mismatch\n"); + for (i = 0; i < plen; i++) { + DEBUGP("ftp:char %u `%c'(%u) vs `%c'(%u)\n", + i, data[i], data[i], + pattern[i], pattern[i]); + } +#endif + return 0; + } + + DEBUGP("Pattern matches!\n"); + /* Now we've found the constant string, try to skip + to the 'skip' character */ + for (i = plen; data[i] != skip; i++) + if (i == dlen - 1) return -1; + + /* Skip over the last character */ + i++; + + DEBUGP("Skipped up to `%c'!\n", skip); + + *numoff = i; + *numlen = getnum(data + i, dlen - i, cmd, term); + if (!*numlen) + return -1; + + DEBUGP("Match succeeded!\n"); + return 1; +} + +/* Look up to see if we're just after a \n. */ +static int find_nl_seq(u32 seq, const struct ip_ct_ftp_master *info, int dir) +{ + unsigned int i; + + for (i = 0; i < info->seq_aft_nl_num[dir]; i++) + if (info->seq_aft_nl[dir][i] == seq) + return 1; + return 0; +} + +/* We don't update if it's older than what we have. */ +static void update_nl_seq(u32 nl_seq, struct ip_ct_ftp_master *info, int dir, + struct sk_buff *skb) +{ + unsigned int i, oldest = NUM_SEQ_TO_REMEMBER; + + /* Look for oldest: if we find exact match, we're done. */ + for (i = 0; i < info->seq_aft_nl_num[dir]; i++) { + if (info->seq_aft_nl[dir][i] == nl_seq) + return; + + if (oldest == info->seq_aft_nl_num[dir] + || before(info->seq_aft_nl[dir][i], oldest)) + oldest = i; + } + + if (info->seq_aft_nl_num[dir] < NUM_SEQ_TO_REMEMBER) { + info->seq_aft_nl[dir][info->seq_aft_nl_num[dir]++] = nl_seq; + nf_conntrack_event_cache(IPCT_HELPINFO_VOLATILE, skb); + } else if (oldest != NUM_SEQ_TO_REMEMBER) { + info->seq_aft_nl[dir][oldest] = nl_seq; + nf_conntrack_event_cache(IPCT_HELPINFO_VOLATILE, skb); + } +} + +static int help(struct sk_buff **pskb, + unsigned int protoff, + struct nf_conn *ct, + enum ip_conntrack_info ctinfo) +{ + unsigned int dataoff, datalen; + struct tcphdr _tcph, *th; + char *fb_ptr; + int ret; + u32 seq; + int dir = CTINFO2DIR(ctinfo); + unsigned int matchlen, matchoff; + struct ip_ct_ftp_master *ct_ftp_info = &ct->help->ct_ftp_info; + struct nf_conntrack_expect *exp; + struct nf_conntrack_man cmd = {}; + + unsigned int i; + int found = 0, ends_in_nl; + + /* Until there's been traffic both ways, don't look in packets. */ + if (ctinfo != IP_CT_ESTABLISHED + && ctinfo != IP_CT_ESTABLISHED+IP_CT_IS_REPLY) { + DEBUGP("ftp: Conntrackinfo = %u\n", ctinfo); + return NF_ACCEPT; + } + + th = skb_header_pointer(*pskb, protoff, sizeof(_tcph), &_tcph); + if (th == NULL) + return NF_ACCEPT; + + dataoff = protoff + th->doff * 4; + /* No data? */ + if (dataoff >= (*pskb)->len) { + DEBUGP("ftp: dataoff(%u) >= skblen(%u)\n", dataoff, + (*pskb)->len); + return NF_ACCEPT; + } + datalen = (*pskb)->len - dataoff; + + spin_lock_bh(&nf_ftp_lock); + fb_ptr = skb_header_pointer(*pskb, dataoff, datalen, ftp_buffer); + BUG_ON(fb_ptr == NULL); + + ends_in_nl = (fb_ptr[datalen - 1] == '\n'); + seq = ntohl(th->seq) + datalen; + + /* Look up to see if we're just after a \n. */ + if (!find_nl_seq(ntohl(th->seq), ct_ftp_info, dir)) { + /* Now if this ends in \n, update ftp info. */ + DEBUGP("nf_conntrack_ftp_help: wrong seq pos %s(%u) or %s(%u)\n", + ct_ftp_info->seq_aft_nl_num[dir] > 0 ? "" : "(UNSET)", + ct_ftp_info->seq_aft_nl[dir][0], + ct_ftp_info->seq_aft_nl_num[dir] > 1 ? "" : "(UNSET)", + ct_ftp_info->seq_aft_nl[dir][1]); + ret = NF_ACCEPT; + goto out_update_nl; + } + + /* Initialize IP/IPv6 addr to expected address (it's not mentioned + in EPSV responses) */ + cmd.l3num = ct->tuplehash[dir].tuple.src.l3num; + memcpy(cmd.u3.all, &ct->tuplehash[dir].tuple.src.u3.all, + sizeof(cmd.u3.all)); + + for (i = 0; i < ARRAY_SIZE(search); i++) { + if (search[i].dir != dir) continue; + + found = find_pattern(fb_ptr, datalen, + search[i].pattern, + search[i].plen, + search[i].skip, + search[i].term, + &matchoff, &matchlen, + &cmd, + search[i].getnum); + if (found) break; + } + if (found == -1) { + /* We don't usually drop packets. After all, this is + connection tracking, not packet filtering. + However, it is necessary for accurate tracking in + this case. */ + if (net_ratelimit()) + printk("conntrack_ftp: partial %s %u+%u\n", + search[i].pattern, + ntohl(th->seq), datalen); + ret = NF_DROP; + goto out; + } else if (found == 0) { /* No match */ + ret = NF_ACCEPT; + goto out_update_nl; + } + + DEBUGP("conntrack_ftp: match `%.*s' (%u bytes at %u)\n", + (int)matchlen, fb_ptr + matchoff, + matchlen, ntohl(th->seq) + matchoff); + + exp = nf_conntrack_expect_alloc(ct); + if (exp == NULL) { + ret = NF_DROP; + goto out; + } + + /* We refer to the reverse direction ("!dir") tuples here, + * because we're expecting something in the other direction. + * Doesn't matter unless NAT is happening. */ + exp->tuple.dst.u3 = ct->tuplehash[!dir].tuple.dst.u3; + + /* Update the ftp info */ + if ((cmd.l3num == ct->tuplehash[dir].tuple.src.l3num) && + memcmp(&cmd.u3.all, &ct->tuplehash[dir].tuple.src.u3.all, + sizeof(cmd.u3.all))) { + /* Enrico Scholz's passive FTP to partially RNAT'd ftp + server: it really wants us to connect to a + different IP address. Simply don't record it for + NAT. */ + if (cmd.l3num == PF_INET) { + DEBUGP("conntrack_ftp: NOT RECORDING: %u,%u,%u,%u != %u.%u.%u.%u\n", + NIPQUAD(cmd.u3.ip), + NIPQUAD(ct->tuplehash[dir].tuple.src.u3.ip)); + } else { + DEBUGP("conntrack_ftp: NOT RECORDING: %x:%x:%x:%x:%x:%x:%x:%x != %x:%x:%x:%x:%x:%x:%x:%x\n", + NIP6(*((struct in6_addr *)cmd.u3.ip6)), + NIP6(*((struct in6_addr *)ct->tuplehash[dir] + .tuple.src.u3.ip6))); + } + + /* Thanks to Cristiano Lincoln Mattos + for reporting this potential + problem (DMZ machines opening holes to internal + networks, or the packet filter itself). */ + if (!loose) { + ret = NF_ACCEPT; + goto out_put_expect; + } + memcpy(&exp->tuple.dst.u3, &cmd.u3.all, + sizeof(exp->tuple.dst.u3)); + } + + exp->tuple.src.u3 = ct->tuplehash[!dir].tuple.src.u3; + exp->tuple.src.l3num = cmd.l3num; + exp->tuple.src.u.tcp.port = 0; + exp->tuple.dst.u.tcp.port = cmd.u.tcp.port; + exp->tuple.dst.protonum = IPPROTO_TCP; + + exp->mask = (struct nf_conntrack_tuple) + { .src = { .l3num = 0xFFFF, + .u = { .tcp = { 0 }}, + }, + .dst = { .protonum = 0xFF, + .u = { .tcp = { 0xFFFF }}, + }, + }; + if (cmd.l3num == PF_INET) { + exp->mask.src.u3.ip = 0xFFFFFFFF; + exp->mask.dst.u3.ip = 0xFFFFFFFF; + } else { + memset(exp->mask.src.u3.ip6, 0xFF, + sizeof(exp->mask.src.u3.ip6)); + memset(exp->mask.dst.u3.ip6, 0xFF, + sizeof(exp->mask.src.u3.ip6)); + } + + exp->expectfn = NULL; + exp->flags = 0; + + /* Now, NAT might want to mangle the packet, and register the + * (possibly changed) expectation itself. */ + if (nf_nat_ftp_hook) + ret = nf_nat_ftp_hook(pskb, ctinfo, search[i].ftptype, + matchoff, matchlen, exp, &seq); + else { + /* Can't expect this? Best to drop packet now. */ + if (nf_conntrack_expect_related(exp) != 0) + ret = NF_DROP; + else + ret = NF_ACCEPT; + } + +out_put_expect: + nf_conntrack_expect_put(exp); + +out_update_nl: + /* Now if this ends in \n, update ftp info. Seq may have been + * adjusted by NAT code. */ + if (ends_in_nl) + update_nl_seq(seq, ct_ftp_info, dir, *pskb); + out: + spin_unlock_bh(&nf_ftp_lock); + return ret; +} + +static struct nf_conntrack_helper ftp[MAX_PORTS][2]; +static char ftp_names[MAX_PORTS][2][sizeof("ftp-65535")]; + +/* don't make this __exit, since it's called from __init ! */ +static void fini(void) +{ + int i, j; + for (i = 0; i < ports_c; i++) { + for (j = 0; j < 2; j++) { + if (ftp[i][j].me == NULL) + continue; + + DEBUGP("nf_ct_ftp: unregistering helper for pf: %d " + "port: %d\n", + ftp[i][j].tuple.src.l3num, ports[i]); + nf_conntrack_helper_unregister(&ftp[i][j]); + } + } + + kfree(ftp_buffer); +} + +static int __init init(void) +{ + int i, j = -1, ret = 0; + char *tmpname; + + ftp_buffer = kmalloc(65536, GFP_KERNEL); + if (!ftp_buffer) + return -ENOMEM; + + if (ports_c == 0) + ports[ports_c++] = FTP_PORT; + + /* FIXME should be configurable whether IPv4 and IPv6 FTP connections + are tracked or not - YK */ + for (i = 0; i < ports_c; i++) { + memset(&ftp[i], 0, sizeof(struct nf_conntrack_helper)); + + ftp[i][0].tuple.src.l3num = PF_INET; + ftp[i][1].tuple.src.l3num = PF_INET6; + for (j = 0; j < 2; j++) { + ftp[i][j].tuple.src.u.tcp.port = htons(ports[i]); + ftp[i][j].tuple.dst.protonum = IPPROTO_TCP; + ftp[i][j].mask.src.u.tcp.port = 0xFFFF; + ftp[i][j].mask.dst.protonum = 0xFF; + ftp[i][j].max_expected = 1; + ftp[i][j].timeout = 5 * 60; /* 5 Minutes */ + ftp[i][j].me = THIS_MODULE; + ftp[i][j].help = help; + tmpname = &ftp_names[i][j][0]; + if (ports[i] == FTP_PORT) + sprintf(tmpname, "ftp"); + else + sprintf(tmpname, "ftp-%d", ports[i]); + ftp[i][j].name = tmpname; + + DEBUGP("nf_ct_ftp: registering helper for pf: %d " + "port: %d\n", + ftp[i][j].tuple.src.l3num, ports[i]); + ret = nf_conntrack_helper_register(&ftp[i][j]); + if (ret) { + printk("nf_ct_ftp: failed to register helper " + " for pf: %d port: %d\n", + ftp[i][j].tuple.src.l3num, ports[i]); + fini(); + return ret; + } + } + } + + return 0; +} + +module_init(init); +module_exit(fini); diff --git a/net/netfilter/nf_conntrack_l3proto_generic.c b/net/netfilter/nf_conntrack_l3proto_generic.c new file mode 100644 index 000000000000..7de4f06c63c5 --- /dev/null +++ b/net/netfilter/nf_conntrack_l3proto_generic.c @@ -0,0 +1,98 @@ +/* + * (C) 2003,2004 USAGI/WIDE Project + * + * Based largely upon the original ip_conntrack code which + * had the following copyright information: + * + * (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * Author: + * Yasuyuki Kozakai @USAGI + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +DECLARE_PER_CPU(struct nf_conntrack_stat, nf_conntrack_stat); + +static int generic_pkt_to_tuple(const struct sk_buff *skb, unsigned int nhoff, + struct nf_conntrack_tuple *tuple) +{ + memset(&tuple->src.u3, 0, sizeof(tuple->src.u3)); + memset(&tuple->dst.u3, 0, sizeof(tuple->dst.u3)); + + return 1; +} + +static int generic_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + memset(&tuple->src.u3, 0, sizeof(tuple->src.u3)); + memset(&tuple->dst.u3, 0, sizeof(tuple->dst.u3)); + + return 1; +} + +static int generic_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + return 0; +} + +static int generic_print_conntrack(struct seq_file *s, + const struct nf_conn *conntrack) +{ + return 0; +} + +static int +generic_prepare(struct sk_buff **pskb, unsigned int hooknum, + unsigned int *dataoff, u_int8_t *protonum) +{ + /* Never track !!! */ + return -NF_ACCEPT; +} + + +static u_int32_t generic_get_features(const struct nf_conntrack_tuple *tuple) + +{ + return NF_CT_F_BASIC; +} + +struct nf_conntrack_l3proto nf_conntrack_generic_l3proto = { + .l3proto = PF_UNSPEC, + .name = "unknown", + .pkt_to_tuple = generic_pkt_to_tuple, + .invert_tuple = generic_invert_tuple, + .print_tuple = generic_print_tuple, + .print_conntrack = generic_print_conntrack, + .prepare = generic_prepare, + .get_features = generic_get_features, + .me = THIS_MODULE, +}; diff --git a/net/netfilter/nf_conntrack_proto_generic.c b/net/netfilter/nf_conntrack_proto_generic.c new file mode 100644 index 000000000000..36425f6c833f --- /dev/null +++ b/net/netfilter/nf_conntrack_proto_generic.c @@ -0,0 +1,85 @@ +/* (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - enable working with L3 protocol independent connection tracking. + * + * Derived from net/ipv4/netfilter/ip_conntrack_proto_generic.c + */ + +#include +#include +#include +#include +#include + +unsigned long nf_ct_generic_timeout = 600*HZ; + +static int generic_pkt_to_tuple(const struct sk_buff *skb, + unsigned int dataoff, + struct nf_conntrack_tuple *tuple) +{ + tuple->src.u.all = 0; + tuple->dst.u.all = 0; + + return 1; +} + +static int generic_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + tuple->src.u.all = 0; + tuple->dst.u.all = 0; + + return 1; +} + +/* Print out the per-protocol part of the tuple. */ +static int generic_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + return 0; +} + +/* Print out the private part of the conntrack. */ +static int generic_print_conntrack(struct seq_file *s, + const struct nf_conn *state) +{ + return 0; +} + +/* Returns verdict for packet, or -1 for invalid. */ +static int packet(struct nf_conn *conntrack, + const struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info ctinfo, + int pf, + unsigned int hooknum) +{ + nf_ct_refresh_acct(conntrack, ctinfo, skb, nf_ct_generic_timeout); + return NF_ACCEPT; +} + +/* Called when a new connection for this protocol found. */ +static int new(struct nf_conn *conntrack, const struct sk_buff *skb, + unsigned int dataoff) +{ + return 1; +} + +struct nf_conntrack_protocol nf_conntrack_generic_protocol = +{ + .l3proto = PF_UNSPEC, + .proto = 0, + .name = "unknown", + .pkt_to_tuple = generic_pkt_to_tuple, + .invert_tuple = generic_invert_tuple, + .print_tuple = generic_print_tuple, + .print_conntrack = generic_print_conntrack, + .packet = packet, + .new = new, +}; diff --git a/net/netfilter/nf_conntrack_proto_sctp.c b/net/netfilter/nf_conntrack_proto_sctp.c new file mode 100644 index 000000000000..3a600f77b4e0 --- /dev/null +++ b/net/netfilter/nf_conntrack_proto_sctp.c @@ -0,0 +1,670 @@ +/* + * Connection tracking protocol helper module for SCTP. + * + * SCTP is defined in RFC 2960. References to various sections in this code + * are to this RFC. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 17 Oct 2004: Yasuyuki Kozakai @USAGI + * - enable working with L3 protocol independent connection tracking. + * + * Derived from net/ipv4/ip_conntrack_sctp.c + */ + +/* + * Added support for proc manipulation of timeouts. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include + +#if 0 +#define DEBUGP(format, ...) printk(format, ## __VA_ARGS__) +#else +#define DEBUGP(format, args...) +#endif + +/* Protects conntrack->proto.sctp */ +static DEFINE_RWLOCK(sctp_lock); + +/* FIXME: Examine ipfilter's timeouts and conntrack transitions more + closely. They're more complex. --RR + + And so for me for SCTP :D -Kiran */ + +static const char *sctp_conntrack_names[] = { + "NONE", + "CLOSED", + "COOKIE_WAIT", + "COOKIE_ECHOED", + "ESTABLISHED", + "SHUTDOWN_SENT", + "SHUTDOWN_RECD", + "SHUTDOWN_ACK_SENT", +}; + +#define SECS * HZ +#define MINS * 60 SECS +#define HOURS * 60 MINS +#define DAYS * 24 HOURS + +static unsigned long nf_ct_sctp_timeout_closed = 10 SECS; +static unsigned long nf_ct_sctp_timeout_cookie_wait = 3 SECS; +static unsigned long nf_ct_sctp_timeout_cookie_echoed = 3 SECS; +static unsigned long nf_ct_sctp_timeout_established = 5 DAYS; +static unsigned long nf_ct_sctp_timeout_shutdown_sent = 300 SECS / 1000; +static unsigned long nf_ct_sctp_timeout_shutdown_recd = 300 SECS / 1000; +static unsigned long nf_ct_sctp_timeout_shutdown_ack_sent = 3 SECS; + +static unsigned long * sctp_timeouts[] += { NULL, /* SCTP_CONNTRACK_NONE */ + &nf_ct_sctp_timeout_closed, /* SCTP_CONNTRACK_CLOSED */ + &nf_ct_sctp_timeout_cookie_wait, /* SCTP_CONNTRACK_COOKIE_WAIT */ + &nf_ct_sctp_timeout_cookie_echoed, /* SCTP_CONNTRACK_COOKIE_ECHOED */ + &nf_ct_sctp_timeout_established, /* SCTP_CONNTRACK_ESTABLISHED */ + &nf_ct_sctp_timeout_shutdown_sent, /* SCTP_CONNTRACK_SHUTDOWN_SENT */ + &nf_ct_sctp_timeout_shutdown_recd, /* SCTP_CONNTRACK_SHUTDOWN_RECD */ + &nf_ct_sctp_timeout_shutdown_ack_sent /* SCTP_CONNTRACK_SHUTDOWN_ACK_SENT */ + }; + +#define sNO SCTP_CONNTRACK_NONE +#define sCL SCTP_CONNTRACK_CLOSED +#define sCW SCTP_CONNTRACK_COOKIE_WAIT +#define sCE SCTP_CONNTRACK_COOKIE_ECHOED +#define sES SCTP_CONNTRACK_ESTABLISHED +#define sSS SCTP_CONNTRACK_SHUTDOWN_SENT +#define sSR SCTP_CONNTRACK_SHUTDOWN_RECD +#define sSA SCTP_CONNTRACK_SHUTDOWN_ACK_SENT +#define sIV SCTP_CONNTRACK_MAX + +/* + These are the descriptions of the states: + +NOTE: These state names are tantalizingly similar to the states of an +SCTP endpoint. But the interpretation of the states is a little different, +considering that these are the states of the connection and not of an end +point. Please note the subtleties. -Kiran + +NONE - Nothing so far. +COOKIE WAIT - We have seen an INIT chunk in the original direction, or also + an INIT_ACK chunk in the reply direction. +COOKIE ECHOED - We have seen a COOKIE_ECHO chunk in the original direction. +ESTABLISHED - We have seen a COOKIE_ACK in the reply direction. +SHUTDOWN_SENT - We have seen a SHUTDOWN chunk in the original direction. +SHUTDOWN_RECD - We have seen a SHUTDOWN chunk in the reply directoin. +SHUTDOWN_ACK_SENT - We have seen a SHUTDOWN_ACK chunk in the direction opposite + to that of the SHUTDOWN chunk. +CLOSED - We have seen a SHUTDOWN_COMPLETE chunk in the direction of + the SHUTDOWN chunk. Connection is closed. +*/ + +/* TODO + - I have assumed that the first INIT is in the original direction. + This messes things when an INIT comes in the reply direction in CLOSED + state. + - Check the error type in the reply dir before transitioning from +cookie echoed to closed. + - Sec 5.2.4 of RFC 2960 + - Multi Homing support. +*/ + +/* SCTP conntrack state transitions */ +static enum sctp_conntrack sctp_conntracks[2][9][SCTP_CONNTRACK_MAX] = { + { +/* ORIGINAL */ +/* sNO, sCL, sCW, sCE, sES, sSS, sSR, sSA */ +/* init */ {sCW, sCW, sCW, sCE, sES, sSS, sSR, sSA}, +/* init_ack */ {sCL, sCL, sCW, sCE, sES, sSS, sSR, sSA}, +/* abort */ {sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL}, +/* shutdown */ {sCL, sCL, sCW, sCE, sSS, sSS, sSR, sSA}, +/* shutdown_ack */ {sSA, sCL, sCW, sCE, sES, sSA, sSA, sSA}, +/* error */ {sCL, sCL, sCW, sCE, sES, sSS, sSR, sSA},/* Cant have Stale cookie*/ +/* cookie_echo */ {sCL, sCL, sCE, sCE, sES, sSS, sSR, sSA},/* 5.2.4 - Big TODO */ +/* cookie_ack */ {sCL, sCL, sCW, sCE, sES, sSS, sSR, sSA},/* Cant come in orig dir */ +/* shutdown_comp*/ {sCL, sCL, sCW, sCE, sES, sSS, sSR, sCL} + }, + { +/* REPLY */ +/* sNO, sCL, sCW, sCE, sES, sSS, sSR, sSA */ +/* init */ {sIV, sCL, sCW, sCE, sES, sSS, sSR, sSA},/* INIT in sCL Big TODO */ +/* init_ack */ {sIV, sCL, sCW, sCE, sES, sSS, sSR, sSA}, +/* abort */ {sIV, sCL, sCL, sCL, sCL, sCL, sCL, sCL}, +/* shutdown */ {sIV, sCL, sCW, sCE, sSR, sSS, sSR, sSA}, +/* shutdown_ack */ {sIV, sCL, sCW, sCE, sES, sSA, sSA, sSA}, +/* error */ {sIV, sCL, sCW, sCL, sES, sSS, sSR, sSA}, +/* cookie_echo */ {sIV, sCL, sCW, sCE, sES, sSS, sSR, sSA},/* Cant come in reply dir */ +/* cookie_ack */ {sIV, sCL, sCW, sES, sES, sSS, sSR, sSA}, +/* shutdown_comp*/ {sIV, sCL, sCW, sCE, sES, sSS, sSR, sCL} + } +}; + +static int sctp_pkt_to_tuple(const struct sk_buff *skb, + unsigned int dataoff, + struct nf_conntrack_tuple *tuple) +{ + sctp_sctphdr_t _hdr, *hp; + + DEBUGP(__FUNCTION__); + DEBUGP("\n"); + + /* Actually only need first 8 bytes. */ + hp = skb_header_pointer(skb, dataoff, 8, &_hdr); + if (hp == NULL) + return 0; + + tuple->src.u.sctp.port = hp->source; + tuple->dst.u.sctp.port = hp->dest; + return 1; +} + +static int sctp_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + DEBUGP(__FUNCTION__); + DEBUGP("\n"); + + tuple->src.u.sctp.port = orig->dst.u.sctp.port; + tuple->dst.u.sctp.port = orig->src.u.sctp.port; + return 1; +} + +/* Print out the per-protocol part of the tuple. */ +static int sctp_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + DEBUGP(__FUNCTION__); + DEBUGP("\n"); + + return seq_printf(s, "sport=%hu dport=%hu ", + ntohs(tuple->src.u.sctp.port), + ntohs(tuple->dst.u.sctp.port)); +} + +/* Print out the private part of the conntrack. */ +static int sctp_print_conntrack(struct seq_file *s, + const struct nf_conn *conntrack) +{ + enum sctp_conntrack state; + + DEBUGP(__FUNCTION__); + DEBUGP("\n"); + + read_lock_bh(&sctp_lock); + state = conntrack->proto.sctp.state; + read_unlock_bh(&sctp_lock); + + return seq_printf(s, "%s ", sctp_conntrack_names[state]); +} + +#define for_each_sctp_chunk(skb, sch, _sch, offset, dataoff, count) \ +for (offset = dataoff + sizeof(sctp_sctphdr_t), count = 0; \ + offset < skb->len && \ + (sch = skb_header_pointer(skb, offset, sizeof(_sch), &_sch)); \ + offset += (htons(sch->length) + 3) & ~3, count++) + +/* Some validity checks to make sure the chunks are fine */ +static int do_basic_checks(struct nf_conn *conntrack, + const struct sk_buff *skb, + unsigned int dataoff, + char *map) +{ + u_int32_t offset, count; + sctp_chunkhdr_t _sch, *sch; + int flag; + + DEBUGP(__FUNCTION__); + DEBUGP("\n"); + + flag = 0; + + for_each_sctp_chunk (skb, sch, _sch, offset, dataoff, count) { + DEBUGP("Chunk Num: %d Type: %d\n", count, sch->type); + + if (sch->type == SCTP_CID_INIT + || sch->type == SCTP_CID_INIT_ACK + || sch->type == SCTP_CID_SHUTDOWN_COMPLETE) { + flag = 1; + } + + /* Cookie Ack/Echo chunks not the first OR + Init / Init Ack / Shutdown compl chunks not the only chunks */ + if ((sch->type == SCTP_CID_COOKIE_ACK + || sch->type == SCTP_CID_COOKIE_ECHO + || flag) + && count !=0 ) { + DEBUGP("Basic checks failed\n"); + return 1; + } + + if (map) { + set_bit(sch->type, (void *)map); + } + } + + DEBUGP("Basic checks passed\n"); + return 0; +} + +static int new_state(enum ip_conntrack_dir dir, + enum sctp_conntrack cur_state, + int chunk_type) +{ + int i; + + DEBUGP(__FUNCTION__); + DEBUGP("\n"); + + DEBUGP("Chunk type: %d\n", chunk_type); + + switch (chunk_type) { + case SCTP_CID_INIT: + DEBUGP("SCTP_CID_INIT\n"); + i = 0; break; + case SCTP_CID_INIT_ACK: + DEBUGP("SCTP_CID_INIT_ACK\n"); + i = 1; break; + case SCTP_CID_ABORT: + DEBUGP("SCTP_CID_ABORT\n"); + i = 2; break; + case SCTP_CID_SHUTDOWN: + DEBUGP("SCTP_CID_SHUTDOWN\n"); + i = 3; break; + case SCTP_CID_SHUTDOWN_ACK: + DEBUGP("SCTP_CID_SHUTDOWN_ACK\n"); + i = 4; break; + case SCTP_CID_ERROR: + DEBUGP("SCTP_CID_ERROR\n"); + i = 5; break; + case SCTP_CID_COOKIE_ECHO: + DEBUGP("SCTP_CID_COOKIE_ECHO\n"); + i = 6; break; + case SCTP_CID_COOKIE_ACK: + DEBUGP("SCTP_CID_COOKIE_ACK\n"); + i = 7; break; + case SCTP_CID_SHUTDOWN_COMPLETE: + DEBUGP("SCTP_CID_SHUTDOWN_COMPLETE\n"); + i = 8; break; + default: + /* Other chunks like DATA, SACK, HEARTBEAT and + its ACK do not cause a change in state */ + DEBUGP("Unknown chunk type, Will stay in %s\n", + sctp_conntrack_names[cur_state]); + return cur_state; + } + + DEBUGP("dir: %d cur_state: %s chunk_type: %d new_state: %s\n", + dir, sctp_conntrack_names[cur_state], chunk_type, + sctp_conntrack_names[sctp_conntracks[dir][i][cur_state]]); + + return sctp_conntracks[dir][i][cur_state]; +} + +/* Returns verdict for packet, or -1 for invalid. */ +static int sctp_packet(struct nf_conn *conntrack, + const struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info ctinfo, + int pf, + unsigned int hooknum) +{ + enum sctp_conntrack newconntrack, oldsctpstate; + sctp_sctphdr_t _sctph, *sh; + sctp_chunkhdr_t _sch, *sch; + u_int32_t offset, count; + char map[256 / sizeof (char)] = {0}; + + DEBUGP(__FUNCTION__); + DEBUGP("\n"); + + sh = skb_header_pointer(skb, dataoff, sizeof(_sctph), &_sctph); + if (sh == NULL) + return -1; + + if (do_basic_checks(conntrack, skb, dataoff, map) != 0) + return -1; + + /* Check the verification tag (Sec 8.5) */ + if (!test_bit(SCTP_CID_INIT, (void *)map) + && !test_bit(SCTP_CID_SHUTDOWN_COMPLETE, (void *)map) + && !test_bit(SCTP_CID_COOKIE_ECHO, (void *)map) + && !test_bit(SCTP_CID_ABORT, (void *)map) + && !test_bit(SCTP_CID_SHUTDOWN_ACK, (void *)map) + && (sh->vtag != conntrack->proto.sctp.vtag[CTINFO2DIR(ctinfo)])) { + DEBUGP("Verification tag check failed\n"); + return -1; + } + + oldsctpstate = newconntrack = SCTP_CONNTRACK_MAX; + for_each_sctp_chunk (skb, sch, _sch, offset, dataoff, count) { + write_lock_bh(&sctp_lock); + + /* Special cases of Verification tag check (Sec 8.5.1) */ + if (sch->type == SCTP_CID_INIT) { + /* Sec 8.5.1 (A) */ + if (sh->vtag != 0) { + write_unlock_bh(&sctp_lock); + return -1; + } + } else if (sch->type == SCTP_CID_ABORT) { + /* Sec 8.5.1 (B) */ + if (!(sh->vtag == conntrack->proto.sctp.vtag[CTINFO2DIR(ctinfo)]) + && !(sh->vtag == conntrack->proto.sctp.vtag + [1 - CTINFO2DIR(ctinfo)])) { + write_unlock_bh(&sctp_lock); + return -1; + } + } else if (sch->type == SCTP_CID_SHUTDOWN_COMPLETE) { + /* Sec 8.5.1 (C) */ + if (!(sh->vtag == conntrack->proto.sctp.vtag[CTINFO2DIR(ctinfo)]) + && !(sh->vtag == conntrack->proto.sctp.vtag + [1 - CTINFO2DIR(ctinfo)] + && (sch->flags & 1))) { + write_unlock_bh(&sctp_lock); + return -1; + } + } else if (sch->type == SCTP_CID_COOKIE_ECHO) { + /* Sec 8.5.1 (D) */ + if (!(sh->vtag == conntrack->proto.sctp.vtag[CTINFO2DIR(ctinfo)])) { + write_unlock_bh(&sctp_lock); + return -1; + } + } + + oldsctpstate = conntrack->proto.sctp.state; + newconntrack = new_state(CTINFO2DIR(ctinfo), oldsctpstate, sch->type); + + /* Invalid */ + if (newconntrack == SCTP_CONNTRACK_MAX) { + DEBUGP("nf_conntrack_sctp: Invalid dir=%i ctype=%u conntrack=%u\n", + CTINFO2DIR(ctinfo), sch->type, oldsctpstate); + write_unlock_bh(&sctp_lock); + return -1; + } + + /* If it is an INIT or an INIT ACK note down the vtag */ + if (sch->type == SCTP_CID_INIT + || sch->type == SCTP_CID_INIT_ACK) { + sctp_inithdr_t _inithdr, *ih; + + ih = skb_header_pointer(skb, offset + sizeof(sctp_chunkhdr_t), + sizeof(_inithdr), &_inithdr); + if (ih == NULL) { + write_unlock_bh(&sctp_lock); + return -1; + } + DEBUGP("Setting vtag %x for dir %d\n", + ih->init_tag, !CTINFO2DIR(ctinfo)); + conntrack->proto.sctp.vtag[!CTINFO2DIR(ctinfo)] = ih->init_tag; + } + + conntrack->proto.sctp.state = newconntrack; + if (oldsctpstate != newconntrack) + nf_conntrack_event_cache(IPCT_PROTOINFO, skb); + write_unlock_bh(&sctp_lock); + } + + nf_ct_refresh_acct(conntrack, ctinfo, skb, *sctp_timeouts[newconntrack]); + + if (oldsctpstate == SCTP_CONNTRACK_COOKIE_ECHOED + && CTINFO2DIR(ctinfo) == IP_CT_DIR_REPLY + && newconntrack == SCTP_CONNTRACK_ESTABLISHED) { + DEBUGP("Setting assured bit\n"); + set_bit(IPS_ASSURED_BIT, &conntrack->status); + nf_conntrack_event_cache(IPCT_STATUS, skb); + } + + return NF_ACCEPT; +} + +/* Called when a new connection for this protocol found. */ +static int sctp_new(struct nf_conn *conntrack, const struct sk_buff *skb, + unsigned int dataoff) +{ + enum sctp_conntrack newconntrack; + sctp_sctphdr_t _sctph, *sh; + sctp_chunkhdr_t _sch, *sch; + u_int32_t offset, count; + char map[256 / sizeof (char)] = {0}; + + DEBUGP(__FUNCTION__); + DEBUGP("\n"); + + sh = skb_header_pointer(skb, dataoff, sizeof(_sctph), &_sctph); + if (sh == NULL) + return 0; + + if (do_basic_checks(conntrack, skb, dataoff, map) != 0) + return 0; + + /* If an OOTB packet has any of these chunks discard (Sec 8.4) */ + if ((test_bit (SCTP_CID_ABORT, (void *)map)) + || (test_bit (SCTP_CID_SHUTDOWN_COMPLETE, (void *)map)) + || (test_bit (SCTP_CID_COOKIE_ACK, (void *)map))) { + return 0; + } + + newconntrack = SCTP_CONNTRACK_MAX; + for_each_sctp_chunk (skb, sch, _sch, offset, dataoff, count) { + /* Don't need lock here: this conntrack not in circulation yet */ + newconntrack = new_state(IP_CT_DIR_ORIGINAL, + SCTP_CONNTRACK_NONE, sch->type); + + /* Invalid: delete conntrack */ + if (newconntrack == SCTP_CONNTRACK_MAX) { + DEBUGP("nf_conntrack_sctp: invalid new deleting.\n"); + return 0; + } + + /* Copy the vtag into the state info */ + if (sch->type == SCTP_CID_INIT) { + if (sh->vtag == 0) { + sctp_inithdr_t _inithdr, *ih; + + ih = skb_header_pointer(skb, offset + sizeof(sctp_chunkhdr_t), + sizeof(_inithdr), &_inithdr); + if (ih == NULL) + return 0; + + DEBUGP("Setting vtag %x for new conn\n", + ih->init_tag); + + conntrack->proto.sctp.vtag[IP_CT_DIR_REPLY] = + ih->init_tag; + } else { + /* Sec 8.5.1 (A) */ + return 0; + } + } + /* If it is a shutdown ack OOTB packet, we expect a return + shutdown complete, otherwise an ABORT Sec 8.4 (5) and (8) */ + else { + DEBUGP("Setting vtag %x for new conn OOTB\n", + sh->vtag); + conntrack->proto.sctp.vtag[IP_CT_DIR_REPLY] = sh->vtag; + } + + conntrack->proto.sctp.state = newconntrack; + } + + return 1; +} + +struct nf_conntrack_protocol nf_conntrack_protocol_sctp4 = { + .l3proto = PF_INET, + .proto = IPPROTO_SCTP, + .name = "sctp", + .pkt_to_tuple = sctp_pkt_to_tuple, + .invert_tuple = sctp_invert_tuple, + .print_tuple = sctp_print_tuple, + .print_conntrack = sctp_print_conntrack, + .packet = sctp_packet, + .new = sctp_new, + .destroy = NULL, + .me = THIS_MODULE +}; + +struct nf_conntrack_protocol nf_conntrack_protocol_sctp6 = { + .l3proto = PF_INET6, + .proto = IPPROTO_SCTP, + .name = "sctp", + .pkt_to_tuple = sctp_pkt_to_tuple, + .invert_tuple = sctp_invert_tuple, + .print_tuple = sctp_print_tuple, + .print_conntrack = sctp_print_conntrack, + .packet = sctp_packet, + .new = sctp_new, + .destroy = NULL, + .me = THIS_MODULE +}; + +#ifdef CONFIG_SYSCTL +static ctl_table nf_ct_sysctl_table[] = { + { + .ctl_name = NET_NF_CONNTRACK_SCTP_TIMEOUT_CLOSED, + .procname = "nf_conntrack_sctp_timeout_closed", + .data = &nf_ct_sctp_timeout_closed, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_SCTP_TIMEOUT_COOKIE_WAIT, + .procname = "nf_conntrack_sctp_timeout_cookie_wait", + .data = &nf_ct_sctp_timeout_cookie_wait, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_SCTP_TIMEOUT_COOKIE_ECHOED, + .procname = "nf_conntrack_sctp_timeout_cookie_echoed", + .data = &nf_ct_sctp_timeout_cookie_echoed, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_SCTP_TIMEOUT_ESTABLISHED, + .procname = "nf_conntrack_sctp_timeout_established", + .data = &nf_ct_sctp_timeout_established, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_SENT, + .procname = "nf_conntrack_sctp_timeout_shutdown_sent", + .data = &nf_ct_sctp_timeout_shutdown_sent, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_RECD, + .procname = "nf_conntrack_sctp_timeout_shutdown_recd", + .data = &nf_ct_sctp_timeout_shutdown_recd, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_SCTP_TIMEOUT_SHUTDOWN_ACK_SENT, + .procname = "nf_conntrack_sctp_timeout_shutdown_ack_sent", + .data = &nf_ct_sctp_timeout_shutdown_ack_sent, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { .ctl_name = 0 } +}; + +static ctl_table nf_ct_netfilter_table[] = { + { + .ctl_name = NET_NETFILTER, + .procname = "netfilter", + .mode = 0555, + .child = nf_ct_sysctl_table, + }, + { .ctl_name = 0 } +}; + +static ctl_table nf_ct_net_table[] = { + { + .ctl_name = CTL_NET, + .procname = "net", + .mode = 0555, + .child = nf_ct_netfilter_table, + }, + { .ctl_name = 0 } +}; + +static struct ctl_table_header *nf_ct_sysctl_header; +#endif + +int __init init(void) +{ + int ret; + + ret = nf_conntrack_protocol_register(&nf_conntrack_protocol_sctp4); + if (ret) { + printk("nf_conntrack_proto_sctp4: protocol register failed\n"); + goto out; + } + ret = nf_conntrack_protocol_register(&nf_conntrack_protocol_sctp6); + if (ret) { + printk("nf_conntrack_proto_sctp6: protocol register failed\n"); + goto cleanup_sctp4; + } + +#ifdef CONFIG_SYSCTL + nf_ct_sysctl_header = register_sysctl_table(nf_ct_net_table, 0); + if (nf_ct_sysctl_header == NULL) { + printk("nf_conntrack_proto_sctp: can't register to sysctl.\n"); + goto cleanup; + } +#endif + + return ret; + +#ifdef CONFIG_SYSCTL + cleanup: + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_sctp6); +#endif + cleanup_sctp4: + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_sctp4); + out: + DEBUGP("SCTP conntrack module loading %s\n", + ret ? "failed": "succeeded"); + return ret; +} + +void __exit fini(void) +{ + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_sctp6); + nf_conntrack_protocol_unregister(&nf_conntrack_protocol_sctp4); +#ifdef CONFIG_SYSCTL + unregister_sysctl_table(nf_ct_sysctl_header); +#endif + DEBUGP("SCTP conntrack module unloaded\n"); +} + +module_init(init); +module_exit(fini); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Kiran Kumar Immidi"); +MODULE_DESCRIPTION("Netfilter connection tracking protocol helper for SCTP"); diff --git a/net/netfilter/nf_conntrack_proto_tcp.c b/net/netfilter/nf_conntrack_proto_tcp.c new file mode 100644 index 000000000000..83d90dd624f0 --- /dev/null +++ b/net/netfilter/nf_conntrack_proto_tcp.c @@ -0,0 +1,1162 @@ +/* (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * Jozsef Kadlecsik : + * - Real stateful connection tracking + * - Modified state transitions table + * - Window scaling support added + * - SACK support added + * + * Willy Tarreau: + * - State table bugfixes + * - More robust state changes + * - Tuning timer parameters + * + * 27 Oct 2004: Yasuyuki Kozakai @USAGI + * - genelized Layer 3 protocol part. + * + * Derived from net/ipv4/netfilter/ip_conntrack_proto_tcp.c + * + * version 2.2 + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include + +#include +#include +#include +#include +#include + +#if 0 +#define DEBUGP printk +#define DEBUGP_VARS +#else +#define DEBUGP(format, args...) +#endif + +/* Protects conntrack->proto.tcp */ +static DEFINE_RWLOCK(tcp_lock); + +/* "Be conservative in what you do, + be liberal in what you accept from others." + If it's non-zero, we mark only out of window RST segments as INVALID. */ +int nf_ct_tcp_be_liberal = 0; + +/* When connection is picked up from the middle, how many packets are required + to pass in each direction when we assume we are in sync - if any side uses + window scaling, we lost the game. + If it is set to zero, we disable picking up already established + connections. */ +int nf_ct_tcp_loose = 3; + +/* Max number of the retransmitted packets without receiving an (acceptable) + ACK from the destination. If this number is reached, a shorter timer + will be started. */ +int nf_ct_tcp_max_retrans = 3; + + /* FIXME: Examine ipfilter's timeouts and conntrack transitions more + closely. They're more complex. --RR */ + +static const char *tcp_conntrack_names[] = { + "NONE", + "SYN_SENT", + "SYN_RECV", + "ESTABLISHED", + "FIN_WAIT", + "CLOSE_WAIT", + "LAST_ACK", + "TIME_WAIT", + "CLOSE", + "LISTEN" +}; + +#define SECS * HZ +#define MINS * 60 SECS +#define HOURS * 60 MINS +#define DAYS * 24 HOURS + +unsigned long nf_ct_tcp_timeout_syn_sent = 2 MINS; +unsigned long nf_ct_tcp_timeout_syn_recv = 60 SECS; +unsigned long nf_ct_tcp_timeout_established = 5 DAYS; +unsigned long nf_ct_tcp_timeout_fin_wait = 2 MINS; +unsigned long nf_ct_tcp_timeout_close_wait = 60 SECS; +unsigned long nf_ct_tcp_timeout_last_ack = 30 SECS; +unsigned long nf_ct_tcp_timeout_time_wait = 2 MINS; +unsigned long nf_ct_tcp_timeout_close = 10 SECS; + +/* RFC1122 says the R2 limit should be at least 100 seconds. + Linux uses 15 packets as limit, which corresponds + to ~13-30min depending on RTO. */ +unsigned long nf_ct_tcp_timeout_max_retrans = 5 MINS; + +static unsigned long * tcp_timeouts[] += { NULL, /* TCP_CONNTRACK_NONE */ + &nf_ct_tcp_timeout_syn_sent, /* TCP_CONNTRACK_SYN_SENT, */ + &nf_ct_tcp_timeout_syn_recv, /* TCP_CONNTRACK_SYN_RECV, */ + &nf_ct_tcp_timeout_established, /* TCP_CONNTRACK_ESTABLISHED, */ + &nf_ct_tcp_timeout_fin_wait, /* TCP_CONNTRACK_FIN_WAIT, */ + &nf_ct_tcp_timeout_close_wait, /* TCP_CONNTRACK_CLOSE_WAIT, */ + &nf_ct_tcp_timeout_last_ack, /* TCP_CONNTRACK_LAST_ACK, */ + &nf_ct_tcp_timeout_time_wait, /* TCP_CONNTRACK_TIME_WAIT, */ + &nf_ct_tcp_timeout_close, /* TCP_CONNTRACK_CLOSE, */ + NULL, /* TCP_CONNTRACK_LISTEN */ + }; + +#define sNO TCP_CONNTRACK_NONE +#define sSS TCP_CONNTRACK_SYN_SENT +#define sSR TCP_CONNTRACK_SYN_RECV +#define sES TCP_CONNTRACK_ESTABLISHED +#define sFW TCP_CONNTRACK_FIN_WAIT +#define sCW TCP_CONNTRACK_CLOSE_WAIT +#define sLA TCP_CONNTRACK_LAST_ACK +#define sTW TCP_CONNTRACK_TIME_WAIT +#define sCL TCP_CONNTRACK_CLOSE +#define sLI TCP_CONNTRACK_LISTEN +#define sIV TCP_CONNTRACK_MAX +#define sIG TCP_CONNTRACK_IGNORE + +/* What TCP flags are set from RST/SYN/FIN/ACK. */ +enum tcp_bit_set { + TCP_SYN_SET, + TCP_SYNACK_SET, + TCP_FIN_SET, + TCP_ACK_SET, + TCP_RST_SET, + TCP_NONE_SET, +}; + +/* + * The TCP state transition table needs a few words... + * + * We are the man in the middle. All the packets go through us + * but might get lost in transit to the destination. + * It is assumed that the destinations can't receive segments + * we haven't seen. + * + * The checked segment is in window, but our windows are *not* + * equivalent with the ones of the sender/receiver. We always + * try to guess the state of the current sender. + * + * The meaning of the states are: + * + * NONE: initial state + * SYN_SENT: SYN-only packet seen + * SYN_RECV: SYN-ACK packet seen + * ESTABLISHED: ACK packet seen + * FIN_WAIT: FIN packet seen + * CLOSE_WAIT: ACK seen (after FIN) + * LAST_ACK: FIN seen (after FIN) + * TIME_WAIT: last ACK seen + * CLOSE: closed connection + * + * LISTEN state is not used. + * + * Packets marked as IGNORED (sIG): + * if they may be either invalid or valid + * and the receiver may send back a connection + * closing RST or a SYN/ACK. + * + * Packets marked as INVALID (sIV): + * if they are invalid + * or we do not support the request (simultaneous open) + */ +static enum tcp_conntrack tcp_conntracks[2][6][TCP_CONNTRACK_MAX] = { + { +/* ORIGINAL */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*syn*/ { sSS, sSS, sIG, sIG, sIG, sIG, sIG, sSS, sSS, sIV }, +/* + * sNO -> sSS Initialize a new connection + * sSS -> sSS Retransmitted SYN + * sSR -> sIG Late retransmitted SYN? + * sES -> sIG Error: SYNs in window outside the SYN_SENT state + * are errors. Receiver will reply with RST + * and close the connection. + * Or we are not in sync and hold a dead connection. + * sFW -> sIG + * sCW -> sIG + * sLA -> sIG + * sTW -> sSS Reopened connection (RFC 1122). + * sCL -> sSS + */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*synack*/ { sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV }, +/* + * A SYN/ACK from the client is always invalid: + * - either it tries to set up a simultaneous open, which is + * not supported; + * - or the firewall has just been inserted between the two hosts + * during the session set-up. The SYN will be retransmitted + * by the true client (or it'll time out). + */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*fin*/ { sIV, sIV, sFW, sFW, sLA, sLA, sLA, sTW, sCL, sIV }, +/* + * sNO -> sIV Too late and no reason to do anything... + * sSS -> sIV Client migth not send FIN in this state: + * we enforce waiting for a SYN/ACK reply first. + * sSR -> sFW Close started. + * sES -> sFW + * sFW -> sLA FIN seen in both directions, waiting for + * the last ACK. + * Migth be a retransmitted FIN as well... + * sCW -> sLA + * sLA -> sLA Retransmitted FIN. Remain in the same state. + * sTW -> sTW + * sCL -> sCL + */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*ack*/ { sES, sIV, sES, sES, sCW, sCW, sTW, sTW, sCL, sIV }, +/* + * sNO -> sES Assumed. + * sSS -> sIV ACK is invalid: we haven't seen a SYN/ACK yet. + * sSR -> sES Established state is reached. + * sES -> sES :-) + * sFW -> sCW Normal close request answered by ACK. + * sCW -> sCW + * sLA -> sTW Last ACK detected. + * sTW -> sTW Retransmitted last ACK. Remain in the same state. + * sCL -> sCL + */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*rst*/ { sIV, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sIV }, +/*none*/ { sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV } + }, + { +/* REPLY */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*syn*/ { sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV }, +/* + * sNO -> sIV Never reached. + * sSS -> sIV Simultaneous open, not supported + * sSR -> sIV Simultaneous open, not supported. + * sES -> sIV Server may not initiate a connection. + * sFW -> sIV + * sCW -> sIV + * sLA -> sIV + * sTW -> sIV Reopened connection, but server may not do it. + * sCL -> sIV + */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*synack*/ { sIV, sSR, sSR, sIG, sIG, sIG, sIG, sIG, sIG, sIV }, +/* + * sSS -> sSR Standard open. + * sSR -> sSR Retransmitted SYN/ACK. + * sES -> sIG Late retransmitted SYN/ACK? + * sFW -> sIG Might be SYN/ACK answering ignored SYN + * sCW -> sIG + * sLA -> sIG + * sTW -> sIG + * sCL -> sIG + */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*fin*/ { sIV, sIV, sFW, sFW, sLA, sLA, sLA, sTW, sCL, sIV }, +/* + * sSS -> sIV Server might not send FIN in this state. + * sSR -> sFW Close started. + * sES -> sFW + * sFW -> sLA FIN seen in both directions. + * sCW -> sLA + * sLA -> sLA Retransmitted FIN. + * sTW -> sTW + * sCL -> sCL + */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*ack*/ { sIV, sIV, sSR, sES, sCW, sCW, sTW, sTW, sCL, sIV }, +/* + * sSS -> sIV Might be a half-open connection. + * sSR -> sSR Might answer late resent SYN. + * sES -> sES :-) + * sFW -> sCW Normal close request answered by ACK. + * sCW -> sCW + * sLA -> sTW Last ACK detected. + * sTW -> sTW Retransmitted last ACK. + * sCL -> sCL + */ +/* sNO, sSS, sSR, sES, sFW, sCW, sLA, sTW, sCL, sLI */ +/*rst*/ { sIV, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sCL, sIV }, +/*none*/ { sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV, sIV } + } +}; + +static int tcp_pkt_to_tuple(const struct sk_buff *skb, + unsigned int dataoff, + struct nf_conntrack_tuple *tuple) +{ + struct tcphdr _hdr, *hp; + + /* Actually only need first 8 bytes. */ + hp = skb_header_pointer(skb, dataoff, 8, &_hdr); + if (hp == NULL) + return 0; + + tuple->src.u.tcp.port = hp->source; + tuple->dst.u.tcp.port = hp->dest; + + return 1; +} + +static int tcp_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + tuple->src.u.tcp.port = orig->dst.u.tcp.port; + tuple->dst.u.tcp.port = orig->src.u.tcp.port; + return 1; +} + +/* Print out the per-protocol part of the tuple. */ +static int tcp_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + return seq_printf(s, "sport=%hu dport=%hu ", + ntohs(tuple->src.u.tcp.port), + ntohs(tuple->dst.u.tcp.port)); +} + +/* Print out the private part of the conntrack. */ +static int tcp_print_conntrack(struct seq_file *s, + const struct nf_conn *conntrack) +{ + enum tcp_conntrack state; + + read_lock_bh(&tcp_lock); + state = conntrack->proto.tcp.state; + read_unlock_bh(&tcp_lock); + + return seq_printf(s, "%s ", tcp_conntrack_names[state]); +} + +static unsigned int get_conntrack_index(const struct tcphdr *tcph) +{ + if (tcph->rst) return TCP_RST_SET; + else if (tcph->syn) return (tcph->ack ? TCP_SYNACK_SET : TCP_SYN_SET); + else if (tcph->fin) return TCP_FIN_SET; + else if (tcph->ack) return TCP_ACK_SET; + else return TCP_NONE_SET; +} + +/* TCP connection tracking based on 'Real Stateful TCP Packet Filtering + in IP Filter' by Guido van Rooij. + + http://www.nluug.nl/events/sane2000/papers.html + http://www.iae.nl/users/guido/papers/tcp_filtering.ps.gz + + The boundaries and the conditions are changed according to RFC793: + the packet must intersect the window (i.e. segments may be + after the right or before the left edge) and thus receivers may ACK + segments after the right edge of the window. + + td_maxend = max(sack + max(win,1)) seen in reply packets + td_maxwin = max(max(win, 1)) + (sack - ack) seen in sent packets + td_maxwin += seq + len - sender.td_maxend + if seq + len > sender.td_maxend + td_end = max(seq + len) seen in sent packets + + I. Upper bound for valid data: seq <= sender.td_maxend + II. Lower bound for valid data: seq + len >= sender.td_end - receiver.td_maxwin + III. Upper bound for valid ack: sack <= receiver.td_end + IV. Lower bound for valid ack: ack >= receiver.td_end - MAXACKWINDOW + + where sack is the highest right edge of sack block found in the packet. + + The upper bound limit for a valid ack is not ignored - + we doesn't have to deal with fragments. +*/ + +static inline __u32 segment_seq_plus_len(__u32 seq, + size_t len, + unsigned int dataoff, + struct tcphdr *tcph) +{ + /* XXX Should I use payload length field in IP/IPv6 header ? + * - YK */ + return (seq + len - dataoff - tcph->doff*4 + + (tcph->syn ? 1 : 0) + (tcph->fin ? 1 : 0)); +} + +/* Fixme: what about big packets? */ +#define MAXACKWINCONST 66000 +#define MAXACKWINDOW(sender) \ + ((sender)->td_maxwin > MAXACKWINCONST ? (sender)->td_maxwin \ + : MAXACKWINCONST) + +/* + * Simplified tcp_parse_options routine from tcp_input.c + */ +static void tcp_options(const struct sk_buff *skb, + unsigned int dataoff, + struct tcphdr *tcph, + struct ip_ct_tcp_state *state) +{ + unsigned char buff[(15 * 4) - sizeof(struct tcphdr)]; + unsigned char *ptr; + int length = (tcph->doff*4) - sizeof(struct tcphdr); + + if (!length) + return; + + ptr = skb_header_pointer(skb, dataoff + sizeof(struct tcphdr), + length, buff); + BUG_ON(ptr == NULL); + + state->td_scale = + state->flags = 0; + + while (length > 0) { + int opcode=*ptr++; + int opsize; + + switch (opcode) { + case TCPOPT_EOL: + return; + case TCPOPT_NOP: /* Ref: RFC 793 section 3.1 */ + length--; + continue; + default: + opsize=*ptr++; + if (opsize < 2) /* "silly options" */ + return; + if (opsize > length) + break; /* don't parse partial options */ + + if (opcode == TCPOPT_SACK_PERM + && opsize == TCPOLEN_SACK_PERM) + state->flags |= IP_CT_TCP_FLAG_SACK_PERM; + else if (opcode == TCPOPT_WINDOW + && opsize == TCPOLEN_WINDOW) { + state->td_scale = *(u_int8_t *)ptr; + + if (state->td_scale > 14) { + /* See RFC1323 */ + state->td_scale = 14; + } + state->flags |= + IP_CT_TCP_FLAG_WINDOW_SCALE; + } + ptr += opsize - 2; + length -= opsize; + } + } +} + +static void tcp_sack(const struct sk_buff *skb, unsigned int dataoff, + struct tcphdr *tcph, __u32 *sack) +{ + unsigned char buff[(15 * 4) - sizeof(struct tcphdr)]; + unsigned char *ptr; + int length = (tcph->doff*4) - sizeof(struct tcphdr); + __u32 tmp; + + if (!length) + return; + + ptr = skb_header_pointer(skb, dataoff + sizeof(struct tcphdr), + length, buff); + BUG_ON(ptr == NULL); + + /* Fast path for timestamp-only option */ + if (length == TCPOLEN_TSTAMP_ALIGNED*4 + && *(__u32 *)ptr == + __constant_ntohl((TCPOPT_NOP << 24) + | (TCPOPT_NOP << 16) + | (TCPOPT_TIMESTAMP << 8) + | TCPOLEN_TIMESTAMP)) + return; + + while (length > 0) { + int opcode = *ptr++; + int opsize, i; + + switch (opcode) { + case TCPOPT_EOL: + return; + case TCPOPT_NOP: /* Ref: RFC 793 section 3.1 */ + length--; + continue; + default: + opsize = *ptr++; + if (opsize < 2) /* "silly options" */ + return; + if (opsize > length) + break; /* don't parse partial options */ + + if (opcode == TCPOPT_SACK + && opsize >= (TCPOLEN_SACK_BASE + + TCPOLEN_SACK_PERBLOCK) + && !((opsize - TCPOLEN_SACK_BASE) + % TCPOLEN_SACK_PERBLOCK)) { + for (i = 0; + i < (opsize - TCPOLEN_SACK_BASE); + i += TCPOLEN_SACK_PERBLOCK) { + memcpy(&tmp, (__u32 *)(ptr + i) + 1, + sizeof(__u32)); + tmp = ntohl(tmp); + + if (after(tmp, *sack)) + *sack = tmp; + } + return; + } + ptr += opsize - 2; + length -= opsize; + } + } +} + +static int tcp_in_window(struct ip_ct_tcp *state, + enum ip_conntrack_dir dir, + unsigned int index, + const struct sk_buff *skb, + unsigned int dataoff, + struct tcphdr *tcph, + int pf) +{ + struct ip_ct_tcp_state *sender = &state->seen[dir]; + struct ip_ct_tcp_state *receiver = &state->seen[!dir]; + __u32 seq, ack, sack, end, win, swin; + int res; + + /* + * Get the required data from the packet. + */ + seq = ntohl(tcph->seq); + ack = sack = ntohl(tcph->ack_seq); + win = ntohs(tcph->window); + end = segment_seq_plus_len(seq, skb->len, dataoff, tcph); + + if (receiver->flags & IP_CT_TCP_FLAG_SACK_PERM) + tcp_sack(skb, dataoff, tcph, &sack); + + DEBUGP("tcp_in_window: START\n"); + DEBUGP("tcp_in_window: src=%u.%u.%u.%u:%hu dst=%u.%u.%u.%u:%hu " + "seq=%u ack=%u sack=%u win=%u end=%u\n", + NIPQUAD(iph->saddr), ntohs(tcph->source), + NIPQUAD(iph->daddr), ntohs(tcph->dest), + seq, ack, sack, win, end); + DEBUGP("tcp_in_window: sender end=%u maxend=%u maxwin=%u scale=%i " + "receiver end=%u maxend=%u maxwin=%u scale=%i\n", + sender->td_end, sender->td_maxend, sender->td_maxwin, + sender->td_scale, + receiver->td_end, receiver->td_maxend, receiver->td_maxwin, + receiver->td_scale); + + if (sender->td_end == 0) { + /* + * Initialize sender data. + */ + if (tcph->syn && tcph->ack) { + /* + * Outgoing SYN-ACK in reply to a SYN. + */ + sender->td_end = + sender->td_maxend = end; + sender->td_maxwin = (win == 0 ? 1 : win); + + tcp_options(skb, dataoff, tcph, sender); + /* + * RFC 1323: + * Both sides must send the Window Scale option + * to enable window scaling in either direction. + */ + if (!(sender->flags & IP_CT_TCP_FLAG_WINDOW_SCALE + && receiver->flags & IP_CT_TCP_FLAG_WINDOW_SCALE)) + sender->td_scale = + receiver->td_scale = 0; + } else { + /* + * We are in the middle of a connection, + * its history is lost for us. + * Let's try to use the data from the packet. + */ + sender->td_end = end; + sender->td_maxwin = (win == 0 ? 1 : win); + sender->td_maxend = end + sender->td_maxwin; + } + } else if (((state->state == TCP_CONNTRACK_SYN_SENT + && dir == IP_CT_DIR_ORIGINAL) + || (state->state == TCP_CONNTRACK_SYN_RECV + && dir == IP_CT_DIR_REPLY)) + && after(end, sender->td_end)) { + /* + * RFC 793: "if a TCP is reinitialized ... then it need + * not wait at all; it must only be sure to use sequence + * numbers larger than those recently used." + */ + sender->td_end = + sender->td_maxend = end; + sender->td_maxwin = (win == 0 ? 1 : win); + + tcp_options(skb, dataoff, tcph, sender); + } + + if (!(tcph->ack)) { + /* + * If there is no ACK, just pretend it was set and OK. + */ + ack = sack = receiver->td_end; + } else if (((tcp_flag_word(tcph) & (TCP_FLAG_ACK|TCP_FLAG_RST)) == + (TCP_FLAG_ACK|TCP_FLAG_RST)) + && (ack == 0)) { + /* + * Broken TCP stacks, that set ACK in RST packets as well + * with zero ack value. + */ + ack = sack = receiver->td_end; + } + + if (seq == end + && (!tcph->rst + || (seq == 0 && state->state == TCP_CONNTRACK_SYN_SENT))) + /* + * Packets contains no data: we assume it is valid + * and check the ack value only. + * However RST segments are always validated by their + * SEQ number, except when seq == 0 (reset sent answering + * SYN. + */ + seq = end = sender->td_end; + + DEBUGP("tcp_in_window: src=%u.%u.%u.%u:%hu dst=%u.%u.%u.%u:%hu " + "seq=%u ack=%u sack =%u win=%u end=%u\n", + NIPQUAD(iph->saddr), ntohs(tcph->source), + NIPQUAD(iph->daddr), ntohs(tcph->dest), + seq, ack, sack, win, end); + DEBUGP("tcp_in_window: sender end=%u maxend=%u maxwin=%u scale=%i " + "receiver end=%u maxend=%u maxwin=%u scale=%i\n", + sender->td_end, sender->td_maxend, sender->td_maxwin, + sender->td_scale, + receiver->td_end, receiver->td_maxend, receiver->td_maxwin, + receiver->td_scale); + + DEBUGP("tcp_in_window: I=%i II=%i III=%i IV=%i\n", + before(seq, sender->td_maxend + 1), + after(end, sender->td_end - receiver->td_maxwin - 1), + before(sack, receiver->td_end + 1), + after(ack, receiver->td_end - MAXACKWINDOW(sender))); + + if (sender->loose || receiver->loose || + (before(seq, sender->td_maxend + 1) && + after(end, sender->td_end - receiver->td_maxwin - 1) && + before(sack, receiver->td_end + 1) && + after(ack, receiver->td_end - MAXACKWINDOW(sender)))) { + /* + * Take into account window scaling (RFC 1323). + */ + if (!tcph->syn) + win <<= sender->td_scale; + + /* + * Update sender data. + */ + swin = win + (sack - ack); + if (sender->td_maxwin < swin) + sender->td_maxwin = swin; + if (after(end, sender->td_end)) + sender->td_end = end; + /* + * Update receiver data. + */ + if (after(end, sender->td_maxend)) + receiver->td_maxwin += end - sender->td_maxend; + if (after(sack + win, receiver->td_maxend - 1)) { + receiver->td_maxend = sack + win; + if (win == 0) + receiver->td_maxend++; + } + + /* + * Check retransmissions. + */ + if (index == TCP_ACK_SET) { + if (state->last_dir == dir + && state->last_seq == seq + && state->last_ack == ack + && state->last_end == end) + state->retrans++; + else { + state->last_dir = dir; + state->last_seq = seq; + state->last_ack = ack; + state->last_end = end; + state->retrans = 0; + } + } + /* + * Close the window of disabled window tracking :-) + */ + if (sender->loose) + sender->loose--; + + res = 1; + } else { + if (LOG_INVALID(IPPROTO_TCP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_tcp: %s ", + before(seq, sender->td_maxend + 1) ? + after(end, sender->td_end - receiver->td_maxwin - 1) ? + before(sack, receiver->td_end + 1) ? + after(ack, receiver->td_end - MAXACKWINDOW(sender)) ? "BUG" + : "ACK is under the lower bound (possible overly delayed ACK)" + : "ACK is over the upper bound (ACKed data not seen yet)" + : "SEQ is under the lower bound (already ACKed data retransmitted)" + : "SEQ is over the upper bound (over the window of the receiver)"); + + res = nf_ct_tcp_be_liberal; + } + + DEBUGP("tcp_in_window: res=%i sender end=%u maxend=%u maxwin=%u " + "receiver end=%u maxend=%u maxwin=%u\n", + res, sender->td_end, sender->td_maxend, sender->td_maxwin, + receiver->td_end, receiver->td_maxend, receiver->td_maxwin); + + return res; +} + +#ifdef CONFIG_IP_NF_NAT_NEEDED +/* Update sender->td_end after NAT successfully mangled the packet */ +/* Caller must linearize skb at tcp header. */ +void nf_conntrack_tcp_update(struct sk_buff *skb, + unsigned int dataoff, + struct nf_conn *conntrack, + int dir) +{ + struct tcphdr *tcph = (void *)skb->data + dataoff; + __u32 end; +#ifdef DEBUGP_VARS + struct ip_ct_tcp_state *sender = &conntrack->proto.tcp.seen[dir]; + struct ip_ct_tcp_state *receiver = &conntrack->proto.tcp.seen[!dir]; +#endif + + end = segment_seq_plus_len(ntohl(tcph->seq), skb->len, dataoff, tcph); + + write_lock_bh(&tcp_lock); + /* + * We have to worry for the ack in the reply packet only... + */ + if (after(end, conntrack->proto.tcp.seen[dir].td_end)) + conntrack->proto.tcp.seen[dir].td_end = end; + conntrack->proto.tcp.last_end = end; + write_unlock_bh(&tcp_lock); + DEBUGP("tcp_update: sender end=%u maxend=%u maxwin=%u scale=%i " + "receiver end=%u maxend=%u maxwin=%u scale=%i\n", + sender->td_end, sender->td_maxend, sender->td_maxwin, + sender->td_scale, + receiver->td_end, receiver->td_maxend, receiver->td_maxwin, + receiver->td_scale); +} + +#endif + +#define TH_FIN 0x01 +#define TH_SYN 0x02 +#define TH_RST 0x04 +#define TH_PUSH 0x08 +#define TH_ACK 0x10 +#define TH_URG 0x20 +#define TH_ECE 0x40 +#define TH_CWR 0x80 + +/* table of valid flag combinations - ECE and CWR are always valid */ +static u8 tcp_valid_flags[(TH_FIN|TH_SYN|TH_RST|TH_PUSH|TH_ACK|TH_URG) + 1] = +{ + [TH_SYN] = 1, + [TH_SYN|TH_ACK] = 1, + [TH_SYN|TH_ACK|TH_PUSH] = 1, + [TH_RST] = 1, + [TH_RST|TH_ACK] = 1, + [TH_RST|TH_ACK|TH_PUSH] = 1, + [TH_FIN|TH_ACK] = 1, + [TH_ACK] = 1, + [TH_ACK|TH_PUSH] = 1, + [TH_ACK|TH_URG] = 1, + [TH_ACK|TH_URG|TH_PUSH] = 1, + [TH_FIN|TH_ACK|TH_PUSH] = 1, + [TH_FIN|TH_ACK|TH_URG] = 1, + [TH_FIN|TH_ACK|TH_URG|TH_PUSH] = 1, +}; + +/* Protect conntrack agaist broken packets. Code taken from ipt_unclean.c. */ +static int tcp_error(struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info *ctinfo, + int pf, + unsigned int hooknum, + int(*csum)(const struct sk_buff *,unsigned int)) +{ + struct tcphdr _tcph, *th; + unsigned int tcplen = skb->len - dataoff; + u_int8_t tcpflags; + + /* Smaller that minimal TCP header? */ + th = skb_header_pointer(skb, dataoff, sizeof(_tcph), &_tcph); + if (th == NULL) { + if (LOG_INVALID(IPPROTO_TCP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_tcp: short packet "); + return -NF_ACCEPT; + } + + /* Not whole TCP header or malformed packet */ + if (th->doff*4 < sizeof(struct tcphdr) || tcplen < th->doff*4) { + if (LOG_INVALID(IPPROTO_TCP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_tcp: truncated/malformed packet "); + return -NF_ACCEPT; + } + + /* Checksum invalid? Ignore. + * We skip checking packets on the outgoing path + * because the semantic of CHECKSUM_HW is different there + * and moreover root might send raw packets. + */ + /* FIXME: Source route IP option packets --RR */ + if (((pf == PF_INET && hooknum == NF_IP_PRE_ROUTING) || + (pf == PF_INET6 && hooknum == NF_IP6_PRE_ROUTING)) + && skb->ip_summed != CHECKSUM_UNNECESSARY + && csum(skb, dataoff)) { + if (LOG_INVALID(IPPROTO_TCP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_tcp: bad TCP checksum "); + return -NF_ACCEPT; + } + + /* Check TCP flags. */ + tcpflags = (((u_int8_t *)th)[13] & ~(TH_ECE|TH_CWR)); + if (!tcp_valid_flags[tcpflags]) { + if (LOG_INVALID(IPPROTO_TCP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_tcp: invalid TCP flag combination "); + return -NF_ACCEPT; + } + + return NF_ACCEPT; +} + +static int csum4(const struct sk_buff *skb, unsigned int dataoff) +{ + return csum_tcpudp_magic(skb->nh.iph->saddr, skb->nh.iph->daddr, + skb->len - dataoff, IPPROTO_TCP, + skb->ip_summed == CHECKSUM_HW ? skb->csum + : skb_checksum(skb, dataoff, + skb->len - dataoff, 0)); +} + +static int csum6(const struct sk_buff *skb, unsigned int dataoff) +{ + return csum_ipv6_magic(&skb->nh.ipv6h->saddr, &skb->nh.ipv6h->daddr, + skb->len - dataoff, IPPROTO_TCP, + skb->ip_summed == CHECKSUM_HW ? skb->csum + : skb_checksum(skb, dataoff, skb->len - dataoff, + 0)); +} + +static int tcp_error4(struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info *ctinfo, + int pf, + unsigned int hooknum) +{ + return tcp_error(skb, dataoff, ctinfo, pf, hooknum, csum4); +} + +static int tcp_error6(struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info *ctinfo, + int pf, + unsigned int hooknum) +{ + return tcp_error(skb, dataoff, ctinfo, pf, hooknum, csum6); +} + +/* Returns verdict for packet, or -1 for invalid. */ +static int tcp_packet(struct nf_conn *conntrack, + const struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info ctinfo, + int pf, + unsigned int hooknum) +{ + enum tcp_conntrack new_state, old_state; + enum ip_conntrack_dir dir; + struct tcphdr *th, _tcph; + unsigned long timeout; + unsigned int index; + + th = skb_header_pointer(skb, dataoff, sizeof(_tcph), &_tcph); + BUG_ON(th == NULL); + + write_lock_bh(&tcp_lock); + old_state = conntrack->proto.tcp.state; + dir = CTINFO2DIR(ctinfo); + index = get_conntrack_index(th); + new_state = tcp_conntracks[dir][index][old_state]; + + switch (new_state) { + case TCP_CONNTRACK_IGNORE: + /* Either SYN in ORIGINAL + * or SYN/ACK in REPLY. */ + if (index == TCP_SYNACK_SET + && conntrack->proto.tcp.last_index == TCP_SYN_SET + && conntrack->proto.tcp.last_dir != dir + && ntohl(th->ack_seq) == + conntrack->proto.tcp.last_end) { + /* This SYN/ACK acknowledges a SYN that we earlier + * ignored as invalid. This means that the client and + * the server are both in sync, while the firewall is + * not. We kill this session and block the SYN/ACK so + * that the client cannot but retransmit its SYN and + * thus initiate a clean new session. + */ + write_unlock_bh(&tcp_lock); + if (LOG_INVALID(IPPROTO_TCP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_tcp: killing out of sync session "); + if (del_timer(&conntrack->timeout)) + conntrack->timeout.function((unsigned long) + conntrack); + return -NF_DROP; + } + conntrack->proto.tcp.last_index = index; + conntrack->proto.tcp.last_dir = dir; + conntrack->proto.tcp.last_seq = ntohl(th->seq); + conntrack->proto.tcp.last_end = + segment_seq_plus_len(ntohl(th->seq), skb->len, dataoff, th); + + write_unlock_bh(&tcp_lock); + if (LOG_INVALID(IPPROTO_TCP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_tcp: invalid packed ignored "); + return NF_ACCEPT; + case TCP_CONNTRACK_MAX: + /* Invalid packet */ + DEBUGP("nf_ct_tcp: Invalid dir=%i index=%u ostate=%u\n", + dir, get_conntrack_index(th), + old_state); + write_unlock_bh(&tcp_lock); + if (LOG_INVALID(IPPROTO_TCP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_tcp: invalid state "); + return -NF_ACCEPT; + case TCP_CONNTRACK_SYN_SENT: + if (old_state < TCP_CONNTRACK_TIME_WAIT) + break; + if ((conntrack->proto.tcp.seen[dir].flags & + IP_CT_TCP_FLAG_CLOSE_INIT) + || after(ntohl(th->seq), + conntrack->proto.tcp.seen[dir].td_end)) { + /* Attempt to reopen a closed connection. + * Delete this connection and look up again. */ + write_unlock_bh(&tcp_lock); + if (del_timer(&conntrack->timeout)) + conntrack->timeout.function((unsigned long) + conntrack); + return -NF_REPEAT; + } + case TCP_CONNTRACK_CLOSE: + if (index == TCP_RST_SET + && test_bit(IPS_SEEN_REPLY_BIT, &conntrack->status) + && conntrack->proto.tcp.last_index == TCP_SYN_SET + && ntohl(th->ack_seq) == conntrack->proto.tcp.last_end) { + /* RST sent to invalid SYN we had let trough + * SYN was in window then, tear down connection. + * We skip window checking, because packet might ACK + * segments we ignored in the SYN. */ + goto in_window; + } + /* Just fall trough */ + default: + /* Keep compilers happy. */ + break; + } + + if (!tcp_in_window(&conntrack->proto.tcp, dir, index, + skb, dataoff, th, pf)) { + write_unlock_bh(&tcp_lock); + return -NF_ACCEPT; + } + in_window: + /* From now on we have got in-window packets */ + conntrack->proto.tcp.last_index = index; + + DEBUGP("tcp_conntracks: src=%u.%u.%u.%u:%hu dst=%u.%u.%u.%u:%hu " + "syn=%i ack=%i fin=%i rst=%i old=%i new=%i\n", + NIPQUAD(iph->saddr), ntohs(th->source), + NIPQUAD(iph->daddr), ntohs(th->dest), + (th->syn ? 1 : 0), (th->ack ? 1 : 0), + (th->fin ? 1 : 0), (th->rst ? 1 : 0), + old_state, new_state); + + conntrack->proto.tcp.state = new_state; + if (old_state != new_state + && (new_state == TCP_CONNTRACK_FIN_WAIT + || new_state == TCP_CONNTRACK_CLOSE)) + conntrack->proto.tcp.seen[dir].flags |= IP_CT_TCP_FLAG_CLOSE_INIT; + timeout = conntrack->proto.tcp.retrans >= nf_ct_tcp_max_retrans + && *tcp_timeouts[new_state] > nf_ct_tcp_timeout_max_retrans + ? nf_ct_tcp_timeout_max_retrans : *tcp_timeouts[new_state]; + write_unlock_bh(&tcp_lock); + + nf_conntrack_event_cache(IPCT_PROTOINFO_VOLATILE, skb); + if (new_state != old_state) + nf_conntrack_event_cache(IPCT_PROTOINFO, skb); + + if (!test_bit(IPS_SEEN_REPLY_BIT, &conntrack->status)) { + /* If only reply is a RST, we can consider ourselves not to + have an established connection: this is a fairly common + problem case, so we can delete the conntrack + immediately. --RR */ + if (th->rst) { + if (del_timer(&conntrack->timeout)) + conntrack->timeout.function((unsigned long) + conntrack); + return NF_ACCEPT; + } + } else if (!test_bit(IPS_ASSURED_BIT, &conntrack->status) + && (old_state == TCP_CONNTRACK_SYN_RECV + || old_state == TCP_CONNTRACK_ESTABLISHED) + && new_state == TCP_CONNTRACK_ESTABLISHED) { + /* Set ASSURED if we see see valid ack in ESTABLISHED + after SYN_RECV or a valid answer for a picked up + connection. */ + set_bit(IPS_ASSURED_BIT, &conntrack->status); + nf_conntrack_event_cache(IPCT_STATUS, skb); + } + nf_ct_refresh_acct(conntrack, ctinfo, skb, timeout); + + return NF_ACCEPT; +} + +/* Called when a new connection for this protocol found. */ +static int tcp_new(struct nf_conn *conntrack, + const struct sk_buff *skb, + unsigned int dataoff) +{ + enum tcp_conntrack new_state; + struct tcphdr *th, _tcph; +#ifdef DEBUGP_VARS + struct ip_ct_tcp_state *sender = &conntrack->proto.tcp.seen[0]; + struct ip_ct_tcp_state *receiver = &conntrack->proto.tcp.seen[1]; +#endif + + th = skb_header_pointer(skb, dataoff, sizeof(_tcph), &_tcph); + BUG_ON(th == NULL); + + /* Don't need lock here: this conntrack not in circulation yet */ + new_state + = tcp_conntracks[0][get_conntrack_index(th)] + [TCP_CONNTRACK_NONE]; + + /* Invalid: delete conntrack */ + if (new_state >= TCP_CONNTRACK_MAX) { + DEBUGP("nf_ct_tcp: invalid new deleting.\n"); + return 0; + } + + if (new_state == TCP_CONNTRACK_SYN_SENT) { + /* SYN packet */ + conntrack->proto.tcp.seen[0].td_end = + segment_seq_plus_len(ntohl(th->seq), skb->len, + dataoff, th); + conntrack->proto.tcp.seen[0].td_maxwin = ntohs(th->window); + if (conntrack->proto.tcp.seen[0].td_maxwin == 0) + conntrack->proto.tcp.seen[0].td_maxwin = 1; + conntrack->proto.tcp.seen[0].td_maxend = + conntrack->proto.tcp.seen[0].td_end; + + tcp_options(skb, dataoff, th, &conntrack->proto.tcp.seen[0]); + conntrack->proto.tcp.seen[1].flags = 0; + conntrack->proto.tcp.seen[0].loose = + conntrack->proto.tcp.seen[1].loose = 0; + } else if (nf_ct_tcp_loose == 0) { + /* Don't try to pick up connections. */ + return 0; + } else { + /* + * We are in the middle of a connection, + * its history is lost for us. + * Let's try to use the data from the packet. + */ + conntrack->proto.tcp.seen[0].td_end = + segment_seq_plus_len(ntohl(th->seq), skb->len, + dataoff, th); + conntrack->proto.tcp.seen[0].td_maxwin = ntohs(th->window); + if (conntrack->proto.tcp.seen[0].td_maxwin == 0) + conntrack->proto.tcp.seen[0].td_maxwin = 1; + conntrack->proto.tcp.seen[0].td_maxend = + conntrack->proto.tcp.seen[0].td_end + + conntrack->proto.tcp.seen[0].td_maxwin; + conntrack->proto.tcp.seen[0].td_scale = 0; + + /* We assume SACK. Should we assume window scaling too? */ + conntrack->proto.tcp.seen[0].flags = + conntrack->proto.tcp.seen[1].flags = IP_CT_TCP_FLAG_SACK_PERM; + conntrack->proto.tcp.seen[0].loose = + conntrack->proto.tcp.seen[1].loose = nf_ct_tcp_loose; + } + + conntrack->proto.tcp.seen[1].td_end = 0; + conntrack->proto.tcp.seen[1].td_maxend = 0; + conntrack->proto.tcp.seen[1].td_maxwin = 1; + conntrack->proto.tcp.seen[1].td_scale = 0; + + /* tcp_packet will set them */ + conntrack->proto.tcp.state = TCP_CONNTRACK_NONE; + conntrack->proto.tcp.last_index = TCP_NONE_SET; + + DEBUGP("tcp_new: sender end=%u maxend=%u maxwin=%u scale=%i " + "receiver end=%u maxend=%u maxwin=%u scale=%i\n", + sender->td_end, sender->td_maxend, sender->td_maxwin, + sender->td_scale, + receiver->td_end, receiver->td_maxend, receiver->td_maxwin, + receiver->td_scale); + return 1; +} + +struct nf_conntrack_protocol nf_conntrack_protocol_tcp4 = +{ + .l3proto = PF_INET, + .proto = IPPROTO_TCP, + .name = "tcp", + .pkt_to_tuple = tcp_pkt_to_tuple, + .invert_tuple = tcp_invert_tuple, + .print_tuple = tcp_print_tuple, + .print_conntrack = tcp_print_conntrack, + .packet = tcp_packet, + .new = tcp_new, + .error = tcp_error4, +}; + +struct nf_conntrack_protocol nf_conntrack_protocol_tcp6 = +{ + .l3proto = PF_INET6, + .proto = IPPROTO_TCP, + .name = "tcp", + .pkt_to_tuple = tcp_pkt_to_tuple, + .invert_tuple = tcp_invert_tuple, + .print_tuple = tcp_print_tuple, + .print_conntrack = tcp_print_conntrack, + .packet = tcp_packet, + .new = tcp_new, + .error = tcp_error6, +}; + +EXPORT_SYMBOL(nf_conntrack_protocol_tcp4); +EXPORT_SYMBOL(nf_conntrack_protocol_tcp6); diff --git a/net/netfilter/nf_conntrack_proto_udp.c b/net/netfilter/nf_conntrack_proto_udp.c new file mode 100644 index 000000000000..3cae7ce420dd --- /dev/null +++ b/net/netfilter/nf_conntrack_proto_udp.c @@ -0,0 +1,216 @@ +/* (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - enable working with Layer 3 protocol independent connection tracking. + * + * Derived from net/ipv4/netfilter/ip_conntrack_proto_udp.c + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +unsigned long nf_ct_udp_timeout = 30*HZ; +unsigned long nf_ct_udp_timeout_stream = 180*HZ; + +static int udp_pkt_to_tuple(const struct sk_buff *skb, + unsigned int dataoff, + struct nf_conntrack_tuple *tuple) +{ + struct udphdr _hdr, *hp; + + /* Actually only need first 8 bytes. */ + hp = skb_header_pointer(skb, dataoff, sizeof(_hdr), &_hdr); + if (hp == NULL) + return 0; + + tuple->src.u.udp.port = hp->source; + tuple->dst.u.udp.port = hp->dest; + + return 1; +} + +static int udp_invert_tuple(struct nf_conntrack_tuple *tuple, + const struct nf_conntrack_tuple *orig) +{ + tuple->src.u.udp.port = orig->dst.u.udp.port; + tuple->dst.u.udp.port = orig->src.u.udp.port; + return 1; +} + +/* Print out the per-protocol part of the tuple. */ +static int udp_print_tuple(struct seq_file *s, + const struct nf_conntrack_tuple *tuple) +{ + return seq_printf(s, "sport=%hu dport=%hu ", + ntohs(tuple->src.u.udp.port), + ntohs(tuple->dst.u.udp.port)); +} + +/* Print out the private part of the conntrack. */ +static int udp_print_conntrack(struct seq_file *s, + const struct nf_conn *conntrack) +{ + return 0; +} + +/* Returns verdict for packet, and may modify conntracktype */ +static int udp_packet(struct nf_conn *conntrack, + const struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info ctinfo, + int pf, + unsigned int hooknum) +{ + /* If we've seen traffic both ways, this is some kind of UDP + stream. Extend timeout. */ + if (test_bit(IPS_SEEN_REPLY_BIT, &conntrack->status)) { + nf_ct_refresh_acct(conntrack, ctinfo, skb, + nf_ct_udp_timeout_stream); + /* Also, more likely to be important, and not a probe */ + if (!test_and_set_bit(IPS_ASSURED_BIT, &conntrack->status)) + nf_conntrack_event_cache(IPCT_STATUS, skb); + } else + nf_ct_refresh_acct(conntrack, ctinfo, skb, nf_ct_udp_timeout); + + return NF_ACCEPT; +} + +/* Called when a new connection for this protocol found. */ +static int udp_new(struct nf_conn *conntrack, const struct sk_buff *skb, + unsigned int dataoff) +{ + return 1; +} + +static int udp_error(struct sk_buff *skb, unsigned int dataoff, + enum ip_conntrack_info *ctinfo, + int pf, + unsigned int hooknum, + int (*csum)(const struct sk_buff *, unsigned int)) +{ + unsigned int udplen = skb->len - dataoff; + struct udphdr _hdr, *hdr; + + /* Header is too small? */ + hdr = skb_header_pointer(skb, dataoff, sizeof(_hdr), &_hdr); + if (hdr == NULL) { + if (LOG_INVALID(IPPROTO_UDP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_udp: short packet "); + return -NF_ACCEPT; + } + + /* Truncated/malformed packets */ + if (ntohs(hdr->len) > udplen || ntohs(hdr->len) < sizeof(*hdr)) { + if (LOG_INVALID(IPPROTO_UDP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_udp: truncated/malformed packet "); + return -NF_ACCEPT; + } + + /* Packet with no checksum */ + if (!hdr->check) + return NF_ACCEPT; + + /* Checksum invalid? Ignore. + * We skip checking packets on the outgoing path + * because the semantic of CHECKSUM_HW is different there + * and moreover root might send raw packets. + * FIXME: Source route IP option packets --RR */ + if (((pf == PF_INET && hooknum == NF_IP_PRE_ROUTING) || + (pf == PF_INET6 && hooknum == NF_IP6_PRE_ROUTING)) + && skb->ip_summed != CHECKSUM_UNNECESSARY + && csum(skb, dataoff)) { + if (LOG_INVALID(IPPROTO_UDP)) + nf_log_packet(pf, 0, skb, NULL, NULL, NULL, + "nf_ct_udp: bad UDP checksum "); + return -NF_ACCEPT; + } + + return NF_ACCEPT; +} + +static int csum4(const struct sk_buff *skb, unsigned int dataoff) +{ + return csum_tcpudp_magic(skb->nh.iph->saddr, skb->nh.iph->daddr, + skb->len - dataoff, IPPROTO_UDP, + skb->ip_summed == CHECKSUM_HW ? skb->csum + : skb_checksum(skb, dataoff, + skb->len - dataoff, 0)); +} + +static int csum6(const struct sk_buff *skb, unsigned int dataoff) +{ + return csum_ipv6_magic(&skb->nh.ipv6h->saddr, &skb->nh.ipv6h->daddr, + skb->len - dataoff, IPPROTO_UDP, + skb->ip_summed == CHECKSUM_HW ? skb->csum + : skb_checksum(skb, dataoff, skb->len - dataoff, + 0)); +} + +static int udp_error4(struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info *ctinfo, + int pf, + unsigned int hooknum) +{ + return udp_error(skb, dataoff, ctinfo, pf, hooknum, csum4); +} + +static int udp_error6(struct sk_buff *skb, + unsigned int dataoff, + enum ip_conntrack_info *ctinfo, + int pf, + unsigned int hooknum) +{ + return udp_error(skb, dataoff, ctinfo, pf, hooknum, csum6); +} + +struct nf_conntrack_protocol nf_conntrack_protocol_udp4 = +{ + .l3proto = PF_INET, + .proto = IPPROTO_UDP, + .name = "udp", + .pkt_to_tuple = udp_pkt_to_tuple, + .invert_tuple = udp_invert_tuple, + .print_tuple = udp_print_tuple, + .print_conntrack = udp_print_conntrack, + .packet = udp_packet, + .new = udp_new, + .error = udp_error4, +}; + +struct nf_conntrack_protocol nf_conntrack_protocol_udp6 = +{ + .l3proto = PF_INET6, + .proto = IPPROTO_UDP, + .name = "udp", + .pkt_to_tuple = udp_pkt_to_tuple, + .invert_tuple = udp_invert_tuple, + .print_tuple = udp_print_tuple, + .print_conntrack = udp_print_conntrack, + .packet = udp_packet, + .new = udp_new, + .error = udp_error6, +}; + +EXPORT_SYMBOL(nf_conntrack_protocol_udp4); +EXPORT_SYMBOL(nf_conntrack_protocol_udp6); diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c new file mode 100644 index 000000000000..45224db4fe2f --- /dev/null +++ b/net/netfilter/nf_conntrack_standalone.c @@ -0,0 +1,869 @@ +/* This file contains all the functions required for the standalone + nf_conntrack module. + + These are not required by the compatibility layer. +*/ + +/* (C) 1999-2001 Paul `Rusty' Russell + * (C) 2002-2004 Netfilter Core Team + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * 16 Dec 2003: Yasuyuki Kozakai @USAGI + * - generalize L3 protocol dependent part. + * + * Derived from net/ipv4/netfilter/ip_conntrack_standalone.c + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#ifdef CONFIG_SYSCTL +#include +#endif + +#define ASSERT_READ_LOCK(x) +#define ASSERT_WRITE_LOCK(x) + +#include +#include +#include +#include +#include +#include + +#if 0 +#define DEBUGP printk +#else +#define DEBUGP(format, args...) +#endif + +MODULE_LICENSE("GPL"); + +extern atomic_t nf_conntrack_count; +DECLARE_PER_CPU(struct ip_conntrack_stat, nf_conntrack_stat); + +static int kill_l3proto(struct nf_conn *i, void *data) +{ + return (i->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num == + ((struct nf_conntrack_l3proto *)data)->l3proto); +} + +static int kill_proto(struct nf_conn *i, void *data) +{ + struct nf_conntrack_protocol *proto; + proto = (struct nf_conntrack_protocol *)data; + return (i->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum == + proto->proto) && + (i->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num == + proto->l3proto); +} + +#ifdef CONFIG_PROC_FS +static int +print_tuple(struct seq_file *s, const struct nf_conntrack_tuple *tuple, + struct nf_conntrack_l3proto *l3proto, + struct nf_conntrack_protocol *proto) +{ + return l3proto->print_tuple(s, tuple) || proto->print_tuple(s, tuple); +} + +#ifdef CONFIG_NF_CT_ACCT +static unsigned int +seq_print_counters(struct seq_file *s, + const struct ip_conntrack_counter *counter) +{ + return seq_printf(s, "packets=%llu bytes=%llu ", + (unsigned long long)counter->packets, + (unsigned long long)counter->bytes); +} +#else +#define seq_print_counters(x, y) 0 +#endif + +struct ct_iter_state { + unsigned int bucket; +}; + +static struct list_head *ct_get_first(struct seq_file *seq) +{ + struct ct_iter_state *st = seq->private; + + for (st->bucket = 0; + st->bucket < nf_conntrack_htable_size; + st->bucket++) { + if (!list_empty(&nf_conntrack_hash[st->bucket])) + return nf_conntrack_hash[st->bucket].next; + } + return NULL; +} + +static struct list_head *ct_get_next(struct seq_file *seq, struct list_head *head) +{ + struct ct_iter_state *st = seq->private; + + head = head->next; + while (head == &nf_conntrack_hash[st->bucket]) { + if (++st->bucket >= nf_conntrack_htable_size) + return NULL; + head = nf_conntrack_hash[st->bucket].next; + } + return head; +} + +static struct list_head *ct_get_idx(struct seq_file *seq, loff_t pos) +{ + struct list_head *head = ct_get_first(seq); + + if (head) + while (pos && (head = ct_get_next(seq, head))) + pos--; + return pos ? NULL : head; +} + +static void *ct_seq_start(struct seq_file *seq, loff_t *pos) +{ + read_lock_bh(&nf_conntrack_lock); + return ct_get_idx(seq, *pos); +} + +static void *ct_seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + (*pos)++; + return ct_get_next(s, v); +} + +static void ct_seq_stop(struct seq_file *s, void *v) +{ + read_unlock_bh(&nf_conntrack_lock); +} + +/* return 0 on success, 1 in case of error */ +static int ct_seq_show(struct seq_file *s, void *v) +{ + const struct nf_conntrack_tuple_hash *hash = v; + const struct nf_conn *conntrack = nf_ct_tuplehash_to_ctrack(hash); + struct nf_conntrack_l3proto *l3proto; + struct nf_conntrack_protocol *proto; + + ASSERT_READ_LOCK(&nf_conntrack_lock); + NF_CT_ASSERT(conntrack); + + /* we only want to print DIR_ORIGINAL */ + if (NF_CT_DIRECTION(hash)) + return 0; + + l3proto = nf_ct_find_l3proto(conntrack->tuplehash[IP_CT_DIR_ORIGINAL] + .tuple.src.l3num); + + NF_CT_ASSERT(l3proto); + proto = nf_ct_find_proto(conntrack->tuplehash[IP_CT_DIR_ORIGINAL] + .tuple.src.l3num, + conntrack->tuplehash[IP_CT_DIR_ORIGINAL] + .tuple.dst.protonum); + NF_CT_ASSERT(proto); + + if (seq_printf(s, "%-8s %u %-8s %u %ld ", + l3proto->name, + conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple.src.l3num, + proto->name, + conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple.dst.protonum, + timer_pending(&conntrack->timeout) + ? (long)(conntrack->timeout.expires - jiffies)/HZ : 0) != 0) + return -ENOSPC; + + if (l3proto->print_conntrack(s, conntrack)) + return -ENOSPC; + + if (proto->print_conntrack(s, conntrack)) + return -ENOSPC; + + if (print_tuple(s, &conntrack->tuplehash[IP_CT_DIR_ORIGINAL].tuple, + l3proto, proto)) + return -ENOSPC; + + if (seq_print_counters(s, &conntrack->counters[IP_CT_DIR_ORIGINAL])) + return -ENOSPC; + + if (!(test_bit(IPS_SEEN_REPLY_BIT, &conntrack->status))) + if (seq_printf(s, "[UNREPLIED] ")) + return -ENOSPC; + + if (print_tuple(s, &conntrack->tuplehash[IP_CT_DIR_REPLY].tuple, + l3proto, proto)) + return -ENOSPC; + + if (seq_print_counters(s, &conntrack->counters[IP_CT_DIR_REPLY])) + return -ENOSPC; + + if (test_bit(IPS_ASSURED_BIT, &conntrack->status)) + if (seq_printf(s, "[ASSURED] ")) + return -ENOSPC; + +#if defined(CONFIG_NF_CONNTRACK_MARK) + if (seq_printf(s, "mark=%u ", conntrack->mark)) + return -ENOSPC; +#endif + + if (seq_printf(s, "use=%u\n", atomic_read(&conntrack->ct_general.use))) + return -ENOSPC; + + return 0; +} + +static struct seq_operations ct_seq_ops = { + .start = ct_seq_start, + .next = ct_seq_next, + .stop = ct_seq_stop, + .show = ct_seq_show +}; + +static int ct_open(struct inode *inode, struct file *file) +{ + struct seq_file *seq; + struct ct_iter_state *st; + int ret; + + st = kmalloc(sizeof(struct ct_iter_state), GFP_KERNEL); + if (st == NULL) + return -ENOMEM; + ret = seq_open(file, &ct_seq_ops); + if (ret) + goto out_free; + seq = file->private_data; + seq->private = st; + memset(st, 0, sizeof(struct ct_iter_state)); + return ret; +out_free: + kfree(st); + return ret; +} + +static struct file_operations ct_file_ops = { + .owner = THIS_MODULE, + .open = ct_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; + +/* expects */ +static void *exp_seq_start(struct seq_file *s, loff_t *pos) +{ + struct list_head *e = &nf_conntrack_expect_list; + loff_t i; + + /* strange seq_file api calls stop even if we fail, + * thus we need to grab lock since stop unlocks */ + read_lock_bh(&nf_conntrack_lock); + + if (list_empty(e)) + return NULL; + + for (i = 0; i <= *pos; i++) { + e = e->next; + if (e == &nf_conntrack_expect_list) + return NULL; + } + return e; +} + +static void *exp_seq_next(struct seq_file *s, void *v, loff_t *pos) +{ + struct list_head *e = v; + + ++*pos; + e = e->next; + + if (e == &nf_conntrack_expect_list) + return NULL; + + return e; +} + +static void exp_seq_stop(struct seq_file *s, void *v) +{ + read_unlock_bh(&nf_conntrack_lock); +} + +static int exp_seq_show(struct seq_file *s, void *v) +{ + struct nf_conntrack_expect *expect = v; + + if (expect->timeout.function) + seq_printf(s, "%ld ", timer_pending(&expect->timeout) + ? (long)(expect->timeout.expires - jiffies)/HZ : 0); + else + seq_printf(s, "- "); + seq_printf(s, "l3proto = %u proto=%u ", + expect->tuple.src.l3num, + expect->tuple.dst.protonum); + print_tuple(s, &expect->tuple, + nf_ct_find_l3proto(expect->tuple.src.l3num), + nf_ct_find_proto(expect->tuple.src.l3num, + expect->tuple.dst.protonum)); + return seq_putc(s, '\n'); +} + +static struct seq_operations exp_seq_ops = { + .start = exp_seq_start, + .next = exp_seq_next, + .stop = exp_seq_stop, + .show = exp_seq_show +}; + +static int exp_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &exp_seq_ops); +} + +static struct file_operations exp_file_ops = { + .owner = THIS_MODULE, + .open = exp_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release +}; + +static void *ct_cpu_seq_start(struct seq_file *seq, loff_t *pos) +{ + int cpu; + + if (*pos == 0) + return SEQ_START_TOKEN; + + for (cpu = *pos-1; cpu < NR_CPUS; ++cpu) { + if (!cpu_possible(cpu)) + continue; + *pos = cpu + 1; + return &per_cpu(nf_conntrack_stat, cpu); + } + + return NULL; +} + +static void *ct_cpu_seq_next(struct seq_file *seq, void *v, loff_t *pos) +{ + int cpu; + + for (cpu = *pos; cpu < NR_CPUS; ++cpu) { + if (!cpu_possible(cpu)) + continue; + *pos = cpu + 1; + return &per_cpu(nf_conntrack_stat, cpu); + } + + return NULL; +} + +static void ct_cpu_seq_stop(struct seq_file *seq, void *v) +{ +} + +static int ct_cpu_seq_show(struct seq_file *seq, void *v) +{ + unsigned int nr_conntracks = atomic_read(&nf_conntrack_count); + struct ip_conntrack_stat *st = v; + + if (v == SEQ_START_TOKEN) { + seq_printf(seq, "entries searched found new invalid ignore delete delete_list insert insert_failed drop early_drop icmp_error expect_new expect_create expect_delete\n"); + return 0; + } + + seq_printf(seq, "%08x %08x %08x %08x %08x %08x %08x %08x " + "%08x %08x %08x %08x %08x %08x %08x %08x \n", + nr_conntracks, + st->searched, + st->found, + st->new, + st->invalid, + st->ignore, + st->delete, + st->delete_list, + st->insert, + st->insert_failed, + st->drop, + st->early_drop, + st->error, + + st->expect_new, + st->expect_create, + st->expect_delete + ); + return 0; +} + +static struct seq_operations ct_cpu_seq_ops = { + .start = ct_cpu_seq_start, + .next = ct_cpu_seq_next, + .stop = ct_cpu_seq_stop, + .show = ct_cpu_seq_show, +}; + +static int ct_cpu_seq_open(struct inode *inode, struct file *file) +{ + return seq_open(file, &ct_cpu_seq_ops); +} + +static struct file_operations ct_cpu_seq_fops = { + .owner = THIS_MODULE, + .open = ct_cpu_seq_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release_private, +}; +#endif /* CONFIG_PROC_FS */ + +/* Sysctl support */ + +#ifdef CONFIG_SYSCTL + +/* From nf_conntrack_core.c */ +extern int nf_conntrack_max; +extern unsigned int nf_conntrack_htable_size; + +/* From nf_conntrack_proto_tcp.c */ +extern unsigned long nf_ct_tcp_timeout_syn_sent; +extern unsigned long nf_ct_tcp_timeout_syn_recv; +extern unsigned long nf_ct_tcp_timeout_established; +extern unsigned long nf_ct_tcp_timeout_fin_wait; +extern unsigned long nf_ct_tcp_timeout_close_wait; +extern unsigned long nf_ct_tcp_timeout_last_ack; +extern unsigned long nf_ct_tcp_timeout_time_wait; +extern unsigned long nf_ct_tcp_timeout_close; +extern unsigned long nf_ct_tcp_timeout_max_retrans; +extern int nf_ct_tcp_loose; +extern int nf_ct_tcp_be_liberal; +extern int nf_ct_tcp_max_retrans; + +/* From nf_conntrack_proto_udp.c */ +extern unsigned long nf_ct_udp_timeout; +extern unsigned long nf_ct_udp_timeout_stream; + +/* From nf_conntrack_proto_generic.c */ +extern unsigned long nf_ct_generic_timeout; + +/* Log invalid packets of a given protocol */ +static int log_invalid_proto_min = 0; +static int log_invalid_proto_max = 255; + +static struct ctl_table_header *nf_ct_sysctl_header; + +static ctl_table nf_ct_sysctl_table[] = { + { + .ctl_name = NET_NF_CONNTRACK_MAX, + .procname = "nf_conntrack_max", + .data = &nf_conntrack_max, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, + { + .ctl_name = NET_NF_CONNTRACK_COUNT, + .procname = "nf_conntrack_count", + .data = &nf_conntrack_count, + .maxlen = sizeof(int), + .mode = 0444, + .proc_handler = &proc_dointvec, + }, + { + .ctl_name = NET_NF_CONNTRACK_BUCKETS, + .procname = "nf_conntrack_buckets", + .data = &nf_conntrack_htable_size, + .maxlen = sizeof(unsigned int), + .mode = 0444, + .proc_handler = &proc_dointvec, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_SYN_SENT, + .procname = "nf_conntrack_tcp_timeout_syn_sent", + .data = &nf_ct_tcp_timeout_syn_sent, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_SYN_RECV, + .procname = "nf_conntrack_tcp_timeout_syn_recv", + .data = &nf_ct_tcp_timeout_syn_recv, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_ESTABLISHED, + .procname = "nf_conntrack_tcp_timeout_established", + .data = &nf_ct_tcp_timeout_established, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_FIN_WAIT, + .procname = "nf_conntrack_tcp_timeout_fin_wait", + .data = &nf_ct_tcp_timeout_fin_wait, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_CLOSE_WAIT, + .procname = "nf_conntrack_tcp_timeout_close_wait", + .data = &nf_ct_tcp_timeout_close_wait, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_LAST_ACK, + .procname = "nf_conntrack_tcp_timeout_last_ack", + .data = &nf_ct_tcp_timeout_last_ack, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_TIME_WAIT, + .procname = "nf_conntrack_tcp_timeout_time_wait", + .data = &nf_ct_tcp_timeout_time_wait, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_CLOSE, + .procname = "nf_conntrack_tcp_timeout_close", + .data = &nf_ct_tcp_timeout_close, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_UDP_TIMEOUT, + .procname = "nf_conntrack_udp_timeout", + .data = &nf_ct_udp_timeout, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_UDP_TIMEOUT_STREAM, + .procname = "nf_conntrack_udp_timeout_stream", + .data = &nf_ct_udp_timeout_stream, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_GENERIC_TIMEOUT, + .procname = "nf_conntrack_generic_timeout", + .data = &nf_ct_generic_timeout, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_LOG_INVALID, + .procname = "nf_conntrack_log_invalid", + .data = &nf_ct_log_invalid, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_minmax, + .strategy = &sysctl_intvec, + .extra1 = &log_invalid_proto_min, + .extra2 = &log_invalid_proto_max, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_TIMEOUT_MAX_RETRANS, + .procname = "nf_conntrack_tcp_timeout_max_retrans", + .data = &nf_ct_tcp_timeout_max_retrans, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec_jiffies, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_LOOSE, + .procname = "nf_conntrack_tcp_loose", + .data = &nf_ct_tcp_loose, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_BE_LIBERAL, + .procname = "nf_conntrack_tcp_be_liberal", + .data = &nf_ct_tcp_be_liberal, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, + { + .ctl_name = NET_NF_CONNTRACK_TCP_MAX_RETRANS, + .procname = "nf_conntrack_tcp_max_retrans", + .data = &nf_ct_tcp_max_retrans, + .maxlen = sizeof(unsigned int), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, + + { .ctl_name = 0 } +}; + +#define NET_NF_CONNTRACK_MAX 2089 + +static ctl_table nf_ct_netfilter_table[] = { + { + .ctl_name = NET_NETFILTER, + .procname = "netfilter", + .mode = 0555, + .child = nf_ct_sysctl_table, + }, + { + .ctl_name = NET_NF_CONNTRACK_MAX, + .procname = "nf_conntrack_max", + .data = &nf_conntrack_max, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = &proc_dointvec, + }, + { .ctl_name = 0 } +}; + +static ctl_table nf_ct_net_table[] = { + { + .ctl_name = CTL_NET, + .procname = "net", + .mode = 0555, + .child = nf_ct_netfilter_table, + }, + { .ctl_name = 0 } +}; +EXPORT_SYMBOL(nf_ct_log_invalid); +#endif /* CONFIG_SYSCTL */ + +static int init_or_cleanup(int init) +{ +#ifdef CONFIG_PROC_FS + struct proc_dir_entry *proc, *proc_exp, *proc_stat; +#endif + int ret = 0; + + if (!init) goto cleanup; + + ret = nf_conntrack_init(); + if (ret < 0) + goto cleanup_nothing; + +#ifdef CONFIG_PROC_FS + proc = proc_net_fops_create("nf_conntrack", 0440, &ct_file_ops); + if (!proc) goto cleanup_init; + + proc_exp = proc_net_fops_create("nf_conntrack_expect", 0440, + &exp_file_ops); + if (!proc_exp) goto cleanup_proc; + + proc_stat = create_proc_entry("nf_conntrack", S_IRUGO, proc_net_stat); + if (!proc_stat) + goto cleanup_proc_exp; + + proc_stat->proc_fops = &ct_cpu_seq_fops; + proc_stat->owner = THIS_MODULE; +#endif +#ifdef CONFIG_SYSCTL + nf_ct_sysctl_header = register_sysctl_table(nf_ct_net_table, 0); + if (nf_ct_sysctl_header == NULL) { + printk("nf_conntrack: can't register to sysctl.\n"); + ret = -ENOMEM; + goto cleanup_proc_stat; + } +#endif + + return ret; + + cleanup: +#ifdef CONFIG_SYSCTL + unregister_sysctl_table(nf_ct_sysctl_header); + cleanup_proc_stat: +#endif +#ifdef CONFIG_PROC_FS + proc_net_remove("nf_conntrack_stat"); + cleanup_proc_exp: + proc_net_remove("nf_conntrack_expect"); + cleanup_proc: + proc_net_remove("nf_conntrack"); + cleanup_init: +#endif /* CNFIG_PROC_FS */ + nf_conntrack_cleanup(); + cleanup_nothing: + return ret; +} + +int nf_conntrack_l3proto_register(struct nf_conntrack_l3proto *proto) +{ + int ret = 0; + + write_lock_bh(&nf_conntrack_lock); + if (nf_ct_l3protos[proto->l3proto] != &nf_conntrack_generic_l3proto) { + ret = -EBUSY; + goto out; + } + nf_ct_l3protos[proto->l3proto] = proto; +out: + write_unlock_bh(&nf_conntrack_lock); + + return ret; +} + +void nf_conntrack_l3proto_unregister(struct nf_conntrack_l3proto *proto) +{ + write_lock_bh(&nf_conntrack_lock); + nf_ct_l3protos[proto->l3proto] = &nf_conntrack_generic_l3proto; + write_unlock_bh(&nf_conntrack_lock); + + /* Somebody could be still looking at the proto in bh. */ + synchronize_net(); + + /* Remove all contrack entries for this protocol */ + nf_ct_iterate_cleanup(kill_l3proto, proto); +} + +/* FIXME: Allow NULL functions and sub in pointers to generic for + them. --RR */ +int nf_conntrack_protocol_register(struct nf_conntrack_protocol *proto) +{ + int ret = 0; + +retry: + write_lock_bh(&nf_conntrack_lock); + if (nf_ct_protos[proto->l3proto]) { + if (nf_ct_protos[proto->l3proto][proto->proto] + != &nf_conntrack_generic_protocol) { + ret = -EBUSY; + goto out_unlock; + } + } else { + /* l3proto may be loaded latter. */ + struct nf_conntrack_protocol **proto_array; + int i; + + write_unlock_bh(&nf_conntrack_lock); + + proto_array = (struct nf_conntrack_protocol **) + kmalloc(MAX_NF_CT_PROTO * + sizeof(struct nf_conntrack_protocol *), + GFP_KERNEL); + if (proto_array == NULL) { + ret = -ENOMEM; + goto out; + } + for (i = 0; i < MAX_NF_CT_PROTO; i++) + proto_array[i] = &nf_conntrack_generic_protocol; + + write_lock_bh(&nf_conntrack_lock); + if (nf_ct_protos[proto->l3proto]) { + /* bad timing, but no problem */ + write_unlock_bh(&nf_conntrack_lock); + kfree(proto_array); + } else { + nf_ct_protos[proto->l3proto] = proto_array; + write_unlock_bh(&nf_conntrack_lock); + } + + /* + * Just once because array is never freed until unloading + * nf_conntrack.ko + */ + goto retry; + } + + nf_ct_protos[proto->l3proto][proto->proto] = proto; + +out_unlock: + write_unlock_bh(&nf_conntrack_lock); +out: + return ret; +} + +void nf_conntrack_protocol_unregister(struct nf_conntrack_protocol *proto) +{ + write_lock_bh(&nf_conntrack_lock); + nf_ct_protos[proto->l3proto][proto->proto] + = &nf_conntrack_generic_protocol; + write_unlock_bh(&nf_conntrack_lock); + + /* Somebody could be still looking at the proto in bh. */ + synchronize_net(); + + /* Remove all contrack entries for this protocol */ + nf_ct_iterate_cleanup(kill_proto, proto); +} + +static int __init init(void) +{ + return init_or_cleanup(1); +} + +static void __exit fini(void) +{ + init_or_cleanup(0); +} + +module_init(init); +module_exit(fini); + +/* Some modules need us, but don't depend directly on any symbol. + They should call this. */ +void need_nf_conntrack(void) +{ +} + +#ifdef CONFIG_NF_CONNTRACK_EVENTS +EXPORT_SYMBOL_GPL(nf_conntrack_chain); +EXPORT_SYMBOL_GPL(nf_conntrack_expect_chain); +EXPORT_SYMBOL_GPL(nf_conntrack_register_notifier); +EXPORT_SYMBOL_GPL(nf_conntrack_unregister_notifier); +EXPORT_SYMBOL_GPL(__nf_ct_event_cache_init); +EXPORT_PER_CPU_SYMBOL_GPL(nf_conntrack_ecache); +EXPORT_SYMBOL_GPL(nf_ct_deliver_cached_events); +#endif +EXPORT_SYMBOL(nf_conntrack_l3proto_register); +EXPORT_SYMBOL(nf_conntrack_l3proto_unregister); +EXPORT_SYMBOL(nf_conntrack_protocol_register); +EXPORT_SYMBOL(nf_conntrack_protocol_unregister); +EXPORT_SYMBOL(nf_ct_invert_tuplepr); +EXPORT_SYMBOL(nf_conntrack_alter_reply); +EXPORT_SYMBOL(nf_conntrack_destroyed); +EXPORT_SYMBOL(need_nf_conntrack); +EXPORT_SYMBOL(nf_conntrack_helper_register); +EXPORT_SYMBOL(nf_conntrack_helper_unregister); +EXPORT_SYMBOL(nf_ct_iterate_cleanup); +EXPORT_SYMBOL(__nf_ct_refresh_acct); +EXPORT_SYMBOL(nf_ct_protos); +EXPORT_SYMBOL(nf_ct_find_proto); +EXPORT_SYMBOL(nf_ct_l3protos); +EXPORT_SYMBOL(nf_conntrack_expect_alloc); +EXPORT_SYMBOL(nf_conntrack_expect_put); +EXPORT_SYMBOL(nf_conntrack_expect_related); +EXPORT_SYMBOL(nf_conntrack_unexpect_related); +EXPORT_SYMBOL(nf_conntrack_tuple_taken); +EXPORT_SYMBOL(nf_conntrack_htable_size); +EXPORT_SYMBOL(nf_conntrack_lock); +EXPORT_SYMBOL(nf_conntrack_hash); +EXPORT_SYMBOL(nf_conntrack_untracked); +EXPORT_SYMBOL_GPL(nf_conntrack_find_get); +#ifdef CONFIG_IP_NF_NAT_NEEDED +EXPORT_SYMBOL(nf_conntrack_tcp_update); +#endif +EXPORT_SYMBOL(__nf_conntrack_confirm); +EXPORT_SYMBOL(nf_ct_get_tuple); +EXPORT_SYMBOL(nf_ct_invert_tuple); +EXPORT_SYMBOL(nf_conntrack_in); +EXPORT_SYMBOL(__nf_conntrack_attach); From a23414beb6607dfd40d3245f7df9dd97a4e2c82b Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 12:00:55 +1100 Subject: [PATCH 056/121] ppc/powerpc: workarounds for old Open Firmware versions This adds code to work around some problems with old versions of Open Firmware, such as on the early powermacs (7500 etc.) and the "Longtrail" CHRP machine. On these machines we have to claim the physical and virtual address ranges explicitly when claiming memory and then set up a V->P mapping. The Longtrail has more problems: setprop doesn't work, and we have to set an "allow-reclaim" variable to 0 in order to get claim on physical memory ranges to fail if the memory is already claimed. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/prom.c | 9 +- arch/powerpc/kernel/prom_init.c | 172 +++++++++++++++++++++------- arch/powerpc/platforms/chrp/setup.c | 4 +- arch/ppc/boot/include/of1275.h | 3 + arch/ppc/boot/of1275/Makefile | 2 +- arch/ppc/boot/of1275/call_prom.c | 74 ++++++++++++ arch/ppc/boot/of1275/claim.c | 99 ++++++++++++---- arch/ppc/boot/of1275/finddevice.c | 19 +-- 8 files changed, 297 insertions(+), 85 deletions(-) create mode 100644 arch/ppc/boot/of1275/call_prom.c diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c index f645adb57534..5af39f866735 100644 --- a/arch/powerpc/kernel/prom.c +++ b/arch/powerpc/kernel/prom.c @@ -1264,7 +1264,14 @@ static int __init early_init_dt_scan_memory(unsigned long node, unsigned long l; /* We are scanning "memory" nodes only */ - if (type == NULL || strcmp(type, "memory") != 0) + if (type == NULL) { + /* + * The longtrail doesn't have a device_type on the + * /memory node, so look for the node called /memory@0. + */ + if (depth != 1 || strcmp(uname, "memory@0") != 0) + return 0; + } else if (strcmp(type, "memory") != 0) return 0; reg = (cell_t *)of_get_flat_dt_prop(node, "reg", &l); diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c index 58f0917bd6b6..09db1bb9ec91 100644 --- a/arch/powerpc/kernel/prom_init.c +++ b/arch/powerpc/kernel/prom_init.c @@ -94,11 +94,17 @@ extern const struct linux_logo logo_linux_clut224; #ifdef CONFIG_PPC64 #define RELOC(x) (*PTRRELOC(&(x))) #define ADDR(x) (u32) add_reloc_offset((unsigned long)(x)) +#define OF_WORKAROUNDS 0 #else #define RELOC(x) (x) #define ADDR(x) (u32) (x) +#define OF_WORKAROUNDS of_workarounds +int of_workarounds; #endif +#define OF_WA_CLAIM 1 /* do phys/virt claim separately, then map */ +#define OF_WA_LONGTRAIL 2 /* work around longtrail bugs */ + #define PROM_BUG() do { \ prom_printf("kernel BUG at %s line 0x%x!\n", \ RELOC(__FILE__), __LINE__); \ @@ -128,10 +134,11 @@ struct prom_args { struct prom_t { ihandle root; - ihandle chosen; + phandle chosen; int cpu; ihandle stdout; ihandle mmumap; + ihandle memory; }; struct mem_map_entry { @@ -360,16 +367,36 @@ static void __init prom_printf(const char *format, ...) static unsigned int __init prom_claim(unsigned long virt, unsigned long size, unsigned long align) { - int ret; struct prom_t *_prom = &RELOC(prom); - ret = call_prom("claim", 3, 1, (prom_arg_t)virt, (prom_arg_t)size, - (prom_arg_t)align); - if (ret != -1 && _prom->mmumap != 0) - /* old pmacs need us to map as well */ + if (align == 0 && (OF_WORKAROUNDS & OF_WA_CLAIM)) { + /* + * Old OF requires we claim physical and virtual separately + * and then map explicitly (assuming virtual mode) + */ + int ret; + prom_arg_t result; + + ret = call_prom_ret("call-method", 5, 2, &result, + ADDR("claim"), _prom->memory, + align, size, virt); + if (ret != 0 || result == -1) + return -1; + ret = call_prom_ret("call-method", 5, 2, &result, + ADDR("claim"), _prom->mmumap, + align, size, virt); + if (ret != 0) { + call_prom("call-method", 4, 1, ADDR("release"), + _prom->memory, size, virt); + return -1; + } + /* the 0x12 is M (coherence) + PP == read/write */ call_prom("call-method", 6, 1, - ADDR("map"), _prom->mmumap, 0, size, virt, virt); - return ret; + ADDR("map"), _prom->mmumap, 0x12, size, virt, virt); + return virt; + } + return call_prom("claim", 3, 1, (prom_arg_t)virt, (prom_arg_t)size, + (prom_arg_t)align); } static void __init __attribute__((noreturn)) prom_panic(const char *reason) @@ -415,11 +442,52 @@ static int inline prom_getproplen(phandle node, const char *pname) return call_prom("getproplen", 2, 1, node, ADDR(pname)); } -static int inline prom_setprop(phandle node, const char *pname, - void *value, size_t valuelen) +static void add_string(char **str, const char *q) { - return call_prom("setprop", 4, 1, node, ADDR(pname), - (u32)(unsigned long) value, (u32) valuelen); + char *p = *str; + + while (*q) + *p++ = *q++; + *p++ = ' '; + *str = p; +} + +static char *tohex(unsigned int x) +{ + static char digits[] = "0123456789abcdef"; + static char result[9]; + int i; + + result[8] = 0; + i = 8; + do { + --i; + result[i] = digits[x & 0xf]; + x >>= 4; + } while (x != 0 && i > 0); + return &result[i]; +} + +static int __init prom_setprop(phandle node, const char *nodename, + const char *pname, void *value, size_t valuelen) +{ + char cmd[256], *p; + + if (!(OF_WORKAROUNDS & OF_WA_LONGTRAIL)) + return call_prom("setprop", 4, 1, node, ADDR(pname), + (u32)(unsigned long) value, (u32) valuelen); + + /* gah... setprop doesn't work on longtrail, have to use interpret */ + p = cmd; + add_string(&p, "dev"); + add_string(&p, nodename); + add_string(&p, tohex((u32)(unsigned long) value)); + add_string(&p, tohex(valuelen)); + add_string(&p, tohex(ADDR(pname))); + add_string(&p, tohex(strlen(RELOC(pname)))); + add_string(&p, "property"); + *p = 0; + return call_prom("interpret", 1, 1, (u32)(unsigned long) cmd); } /* We can't use the standard versions because of RELOC headaches. */ @@ -980,7 +1048,7 @@ static void __init prom_instantiate_rtas(void) rtas_inst = call_prom("open", 1, 1, ADDR("/rtas")); if (!IHANDLE_VALID(rtas_inst)) { - prom_printf("opening rtas package failed"); + prom_printf("opening rtas package failed (%x)\n", rtas_inst); return; } @@ -988,7 +1056,7 @@ static void __init prom_instantiate_rtas(void) if (call_prom_ret("call-method", 3, 2, &entry, ADDR("instantiate-rtas"), - rtas_inst, base) == PROM_ERROR + rtas_inst, base) != 0 || entry == 0) { prom_printf(" failed\n"); return; @@ -997,8 +1065,10 @@ static void __init prom_instantiate_rtas(void) reserve_mem(base, size); - prom_setprop(rtas_node, "linux,rtas-base", &base, sizeof(base)); - prom_setprop(rtas_node, "linux,rtas-entry", &entry, sizeof(entry)); + prom_setprop(rtas_node, "/rtas", "linux,rtas-base", + &base, sizeof(base)); + prom_setprop(rtas_node, "/rtas", "linux,rtas-entry", + &entry, sizeof(entry)); prom_debug("rtas base = 0x%x\n", base); prom_debug("rtas entry = 0x%x\n", entry); @@ -1089,10 +1159,6 @@ static void __init prom_initialize_tce_table(void) if (base < local_alloc_bottom) local_alloc_bottom = base; - /* Save away the TCE table attributes for later use. */ - prom_setprop(node, "linux,tce-base", &base, sizeof(base)); - prom_setprop(node, "linux,tce-size", &minsize, sizeof(minsize)); - /* It seems OF doesn't null-terminate the path :-( */ memset(path, 0, sizeof(path)); /* Call OF to setup the TCE hardware */ @@ -1101,6 +1167,10 @@ static void __init prom_initialize_tce_table(void) prom_printf("package-to-path failed\n"); } + /* Save away the TCE table attributes for later use. */ + prom_setprop(node, path, "linux,tce-base", &base, sizeof(base)); + prom_setprop(node, path, "linux,tce-size", &minsize, sizeof(minsize)); + prom_debug("TCE table: %s\n", path); prom_debug("\tnode = 0x%x\n", node); prom_debug("\tbase = 0x%x\n", base); @@ -1342,6 +1412,7 @@ static void __init prom_init_client_services(unsigned long pp) /* * For really old powermacs, we need to map things we claim. * For that, we need the ihandle of the mmu. + * Also, on the longtrail, we need to work around other bugs. */ static void __init prom_find_mmu(void) { @@ -1355,12 +1426,19 @@ static void __init prom_find_mmu(void) if (prom_getprop(oprom, "model", version, sizeof(version)) <= 0) return; version[sizeof(version) - 1] = 0; - prom_printf("OF version is '%s'\n", version); /* XXX might need to add other versions here */ - if (strcmp(version, "Open Firmware, 1.0.5") != 0) + if (strcmp(version, "Open Firmware, 1.0.5") == 0) + of_workarounds = OF_WA_CLAIM; + else if (strncmp(version, "FirmWorks,3.", 12) == 0) { + of_workarounds = OF_WA_CLAIM | OF_WA_LONGTRAIL; + call_prom("interpret", 1, 1, "dev /memory 0 to allow-reclaim"); + } else return; + _prom->memory = call_prom("open", 1, 1, ADDR("/memory")); prom_getprop(_prom->chosen, "mmu", &_prom->mmumap, sizeof(_prom->mmumap)); + if (!IHANDLE_VALID(_prom->memory) || !IHANDLE_VALID(_prom->mmumap)) + of_workarounds &= ~OF_WA_CLAIM; /* hmmm */ } #else #define prom_find_mmu() @@ -1382,16 +1460,17 @@ static void __init prom_init_stdout(void) memset(path, 0, 256); call_prom("instance-to-path", 3, 1, _prom->stdout, path, 255); val = call_prom("instance-to-package", 1, 1, _prom->stdout); - prom_setprop(_prom->chosen, "linux,stdout-package", &val, sizeof(val)); + prom_setprop(_prom->chosen, "/chosen", "linux,stdout-package", + &val, sizeof(val)); prom_printf("OF stdout device is: %s\n", RELOC(of_stdout_device)); - prom_setprop(_prom->chosen, "linux,stdout-path", - RELOC(of_stdout_device), strlen(RELOC(of_stdout_device))+1); + prom_setprop(_prom->chosen, "/chosen", "linux,stdout-path", + path, strlen(path) + 1); /* If it's a display, note it */ memset(type, 0, sizeof(type)); prom_getprop(val, "device_type", type, sizeof(type)); if (strcmp(type, RELOC("display")) == 0) - prom_setprop(val, "linux,boot-display", NULL, 0); + prom_setprop(val, path, "linux,boot-display", NULL, 0); } static void __init prom_close_stdin(void) @@ -1514,7 +1593,7 @@ static void __init prom_check_displays(void) /* Success */ prom_printf("done\n"); - prom_setprop(node, "linux,opened", NULL, 0); + prom_setprop(node, path, "linux,opened", NULL, 0); /* Setup a usable color table when the appropriate * method is available. Should update this to set-colors */ @@ -1884,9 +1963,11 @@ static void __init fixup_device_tree(void) /* interrupt on this revision of u3 is number 0 and level */ interrupts[0] = 0; interrupts[1] = 1; - prom_setprop(i2c, "interrupts", &interrupts, sizeof(interrupts)); + prom_setprop(i2c, "/u3@0,f8000000/i2c@f8001000", "interrupts", + &interrupts, sizeof(interrupts)); parent = (u32)mpic; - prom_setprop(i2c, "interrupt-parent", &parent, sizeof(parent)); + prom_setprop(i2c, "/u3@0,f8000000/i2c@f8001000", "interrupt-parent", + &parent, sizeof(parent)); #endif } @@ -1922,11 +2003,11 @@ static void __init prom_check_initrd(unsigned long r3, unsigned long r4) RELOC(prom_initrd_end) = RELOC(prom_initrd_start) + r4; val = RELOC(prom_initrd_start); - prom_setprop(_prom->chosen, "linux,initrd-start", &val, - sizeof(val)); + prom_setprop(_prom->chosen, "/chosen", "linux,initrd-start", + &val, sizeof(val)); val = RELOC(prom_initrd_end); - prom_setprop(_prom->chosen, "linux,initrd-end", &val, - sizeof(val)); + prom_setprop(_prom->chosen, "/chosen", "linux,initrd-end", + &val, sizeof(val)); reserve_mem(RELOC(prom_initrd_start), RELOC(prom_initrd_end) - RELOC(prom_initrd_start)); @@ -1968,16 +2049,17 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, */ prom_init_client_services(pp); + /* + * See if this OF is old enough that we need to do explicit maps + * and other workarounds + */ + prom_find_mmu(); + /* * Init prom stdout device */ prom_init_stdout(); - /* - * See if this OF is old enough that we need to do explicit maps - */ - prom_find_mmu(); - /* * Check for an initrd */ @@ -1989,7 +2071,7 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, */ RELOC(of_platform) = prom_find_machine_type(); getprop_rval = RELOC(of_platform); - prom_setprop(_prom->chosen, "linux,platform", + prom_setprop(_prom->chosen, "/chosen", "linux,platform", &getprop_rval, sizeof(getprop_rval)); #ifdef CONFIG_PPC_PSERIES @@ -2050,21 +2132,23 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, * Fill in some infos for use by the kernel later on */ if (RELOC(prom_memory_limit)) - prom_setprop(_prom->chosen, "linux,memory-limit", + prom_setprop(_prom->chosen, "/chosen", "linux,memory-limit", &RELOC(prom_memory_limit), sizeof(prom_memory_limit)); #ifdef CONFIG_PPC64 if (RELOC(ppc64_iommu_off)) - prom_setprop(_prom->chosen, "linux,iommu-off", NULL, 0); + prom_setprop(_prom->chosen, "/chosen", "linux,iommu-off", + NULL, 0); if (RELOC(iommu_force_on)) - prom_setprop(_prom->chosen, "linux,iommu-force-on", NULL, 0); + prom_setprop(_prom->chosen, "/chosen", "linux,iommu-force-on", + NULL, 0); if (RELOC(prom_tce_alloc_start)) { - prom_setprop(_prom->chosen, "linux,tce-alloc-start", + prom_setprop(_prom->chosen, "/chosen", "linux,tce-alloc-start", &RELOC(prom_tce_alloc_start), sizeof(prom_tce_alloc_start)); - prom_setprop(_prom->chosen, "linux,tce-alloc-end", + prom_setprop(_prom->chosen, "/chosen", "linux,tce-alloc-end", &RELOC(prom_tce_alloc_end), sizeof(prom_tce_alloc_end)); } diff --git a/arch/powerpc/platforms/chrp/setup.c b/arch/powerpc/platforms/chrp/setup.c index ecd32d5d85f4..4099ddab9205 100644 --- a/arch/powerpc/platforms/chrp/setup.c +++ b/arch/powerpc/platforms/chrp/setup.c @@ -361,7 +361,9 @@ static void __init chrp_find_openpic(void) printk(KERN_INFO "OpenPIC at %lx\n", opaddr); irq_count = NR_IRQS - NUM_ISA_INTERRUPTS - 4; /* leave room for IPIs */ - prom_get_irq_senses(init_senses, NUM_8259_INTERRUPTS, NR_IRQS - 4); + prom_get_irq_senses(init_senses, NUM_ISA_INTERRUPTS, NR_IRQS - 4); + /* i8259 cascade is always positive level */ + init_senses[0] = IRQ_SENSE_LEVEL | IRQ_POLARITY_POSITIVE; iranges = (unsigned int *) get_property(np, "interrupt-ranges", &len); if (iranges == NULL) diff --git a/arch/ppc/boot/include/of1275.h b/arch/ppc/boot/include/of1275.h index 69173df76db0..4ed88acfa73a 100644 --- a/arch/ppc/boot/include/of1275.h +++ b/arch/ppc/boot/include/of1275.h @@ -19,6 +19,9 @@ extern prom_entry of_prom_entry; /* function declarations */ +int call_prom(const char *service, int nargs, int nret, ...); +int call_prom_ret(const char *service, int nargs, int nret, + unsigned int *rets, ...); void * claim(unsigned int virt, unsigned int size, unsigned int align); int map(unsigned int phys, unsigned int virt, unsigned int size); void enter(void); diff --git a/arch/ppc/boot/of1275/Makefile b/arch/ppc/boot/of1275/Makefile index 02e6f235d7cb..0b979c004972 100644 --- a/arch/ppc/boot/of1275/Makefile +++ b/arch/ppc/boot/of1275/Makefile @@ -3,4 +3,4 @@ # lib-y := claim.o enter.o exit.o finddevice.o getprop.o ofinit.o \ - ofstdio.o read.o release.o write.o map.o + ofstdio.o read.o release.o write.o map.o call_prom.o diff --git a/arch/ppc/boot/of1275/call_prom.c b/arch/ppc/boot/of1275/call_prom.c new file mode 100644 index 000000000000..9479a3a2b8c7 --- /dev/null +++ b/arch/ppc/boot/of1275/call_prom.c @@ -0,0 +1,74 @@ +/* + * Copyright (C) 1996-2005 Paul Mackerras. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ + +#include "of1275.h" +#include + +int call_prom(const char *service, int nargs, int nret, ...) +{ + int i; + struct prom_args { + const char *service; + int nargs; + int nret; + unsigned int args[12]; + } args; + va_list list; + + args.service = service; + args.nargs = nargs; + args.nret = nret; + + va_start(list, nret); + for (i = 0; i < nargs; i++) + args.args[i] = va_arg(list, unsigned int); + va_end(list); + + for (i = 0; i < nret; i++) + args.args[nargs+i] = 0; + + if (of_prom_entry(&args) < 0) + return -1; + + return (nret > 0)? args.args[nargs]: 0; +} + +int call_prom_ret(const char *service, int nargs, int nret, + unsigned int *rets, ...) +{ + int i; + struct prom_args { + const char *service; + int nargs; + int nret; + unsigned int args[12]; + } args; + va_list list; + + args.service = service; + args.nargs = nargs; + args.nret = nret; + + va_start(list, rets); + for (i = 0; i < nargs; i++) + args.args[i] = va_arg(list, unsigned int); + va_end(list); + + for (i = 0; i < nret; i++) + args.args[nargs+i] = 0; + + if (of_prom_entry(&args) < 0) + return -1; + + if (rets != (void *) 0) + for (i = 1; i < nret; ++i) + rets[i-1] = args.args[nargs+i]; + + return (nret > 0)? args.args[nargs]: 0; +} diff --git a/arch/ppc/boot/of1275/claim.c b/arch/ppc/boot/of1275/claim.c index 13169a5c4339..1ed3aeeff8ae 100644 --- a/arch/ppc/boot/of1275/claim.c +++ b/arch/ppc/boot/of1275/claim.c @@ -9,27 +9,84 @@ */ #include "of1275.h" +#include "nonstdio.h" -void * -claim(unsigned int virt, unsigned int size, unsigned int align) +/* + * Older OF's require that when claiming a specific range of addresses, + * we claim the physical space in the /memory node and the virtual + * space in the chosen mmu node, and then do a map operation to + * map virtual to physical. + */ +static int need_map = -1; +static ihandle chosen_mmu; +static phandle memory; + +/* returns true if s2 is a prefix of s1 */ +static int string_match(const char *s1, const char *s2) { - struct prom_args { - char *service; - int nargs; - int nret; - unsigned int virt; - unsigned int size; - unsigned int align; - void *ret; - } args; - - args.service = "claim"; - args.nargs = 3; - args.nret = 1; - args.virt = virt; - args.size = size; - args.align = align; - args.ret = (void *) 0; - (*of_prom_entry)(&args); - return args.ret; + for (; *s2; ++s2) + if (*s1++ != *s2) + return 0; + return 1; +} + +static int check_of_version(void) +{ + phandle oprom, chosen; + char version[64]; + + oprom = finddevice("/openprom"); + if (oprom == OF_INVALID_HANDLE) + return 0; + if (getprop(oprom, "model", version, sizeof(version)) <= 0) + return 0; + version[sizeof(version)-1] = 0; + printf("OF version = '%s'\n", version); + if (!string_match(version, "Open Firmware, 1.") + && !string_match(version, "FirmWorks,3.")) + return 0; + chosen = finddevice("/chosen"); + if (chosen == OF_INVALID_HANDLE) { + chosen = finddevice("/chosen@0"); + if (chosen == OF_INVALID_HANDLE) { + printf("no chosen\n"); + return 0; + } + } + if (getprop(chosen, "mmu", &chosen_mmu, sizeof(chosen_mmu)) <= 0) { + printf("no mmu\n"); + return 0; + } + memory = (ihandle) call_prom("open", 1, 1, "/memory"); + if (memory == OF_INVALID_HANDLE) { + memory = (ihandle) call_prom("open", 1, 1, "/memory@0"); + if (memory == OF_INVALID_HANDLE) { + printf("no memory node\n"); + return 0; + } + } + printf("old OF detected\n"); + return 1; +} + +void *claim(unsigned int virt, unsigned int size, unsigned int align) +{ + int ret; + unsigned int result; + + if (need_map < 0) + need_map = check_of_version(); + if (align || !need_map) + return (void *) call_prom("claim", 3, 1, virt, size, align); + + ret = call_prom_ret("call-method", 5, 2, &result, "claim", memory, + align, size, virt); + if (ret != 0 || result == -1) + return (void *) -1; + ret = call_prom_ret("call-method", 5, 2, &result, "claim", chosen_mmu, + align, size, virt); + /* 0x12 == coherent + read/write */ + ret = call_prom("call-method", 6, 1, "map", chosen_mmu, + 0x12, size, virt, virt); + return virt; } diff --git a/arch/ppc/boot/of1275/finddevice.c b/arch/ppc/boot/of1275/finddevice.c index 2c0f7cbb793e..0dcb1201b772 100644 --- a/arch/ppc/boot/of1275/finddevice.c +++ b/arch/ppc/boot/of1275/finddevice.c @@ -10,22 +10,7 @@ #include "of1275.h" -phandle -finddevice(const char *name) +phandle finddevice(const char *name) { - struct prom_args { - char *service; - int nargs; - int nret; - const char *devspec; - phandle device; - } args; - - args.service = "finddevice"; - args.nargs = 1; - args.nret = 1; - args.devspec = name; - args.device = OF_INVALID_HANDLE; - (*of_prom_entry)(&args); - return args.device; + return (phandle) call_prom("finddevice", 1, 1, name); } From 00557b59c69ce284e5a61bcfcdbcc3dc867cb2da Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 12:01:41 +1100 Subject: [PATCH 057/121] powerpc: Fix find_next_bit on 32-bit We had a "64" that didn't get changed to BITS_PER_LONG, resulting in find_next_bit not working correctly. Signed-off-by: Paul Mackerras --- arch/powerpc/lib/bitops.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/lib/bitops.c b/arch/powerpc/lib/bitops.c index b67ce3004ebf..f68ad71a0187 100644 --- a/arch/powerpc/lib/bitops.c +++ b/arch/powerpc/lib/bitops.c @@ -41,7 +41,7 @@ unsigned long find_next_bit(const unsigned long *addr, unsigned long size, tmp = *p; found_first: - tmp &= (~0UL >> (64 - size)); + tmp &= (~0UL >> (BITS_PER_LONG - size)); if (tmp == 0UL) /* Are any bits set? */ return result + size; /* Nope. */ found_middle: From bfa83a9e03cf8d501c6272999843470afecb32ed Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Thu, 10 Nov 2005 02:25:51 +0100 Subject: [PATCH 058/121] [NETLINK]: Type-safe netlink messages/attributes interface Introduces a new type-safe interface for netlink message and attributes handling. The interface is fully binary compatible with the old interface towards userspace. Besides type safety, this interface features attribute validation capabilities, simplified message contstruction, and documentation. The resulting netlink code should be smaller, less error prone and easier to understand. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- include/linux/netlink.h | 24 +- include/net/netlink.h | 877 ++++++++++++++++++++++++++++++++++++++++ net/netlink/Makefile | 2 +- net/netlink/attr.c | 328 +++++++++++++++ 4 files changed, 1229 insertions(+), 2 deletions(-) create mode 100644 include/net/netlink.h create mode 100644 net/netlink/attr.c diff --git a/include/linux/netlink.h b/include/linux/netlink.h index ba25ca874c20..6a2ccf78a356 100644 --- a/include/linux/netlink.h +++ b/include/linux/netlink.h @@ -71,7 +71,8 @@ struct nlmsghdr #define NLMSG_ALIGNTO 4 #define NLMSG_ALIGN(len) ( ((len)+NLMSG_ALIGNTO-1) & ~(NLMSG_ALIGNTO-1) ) -#define NLMSG_LENGTH(len) ((len)+NLMSG_ALIGN(sizeof(struct nlmsghdr))) +#define NLMSG_HDRLEN ((int) NLMSG_ALIGN(sizeof(struct nlmsghdr))) +#define NLMSG_LENGTH(len) ((len)+NLMSG_ALIGN(NLMSG_HDRLEN)) #define NLMSG_SPACE(len) NLMSG_ALIGN(NLMSG_LENGTH(len)) #define NLMSG_DATA(nlh) ((void*)(((char*)nlh) + NLMSG_LENGTH(0))) #define NLMSG_NEXT(nlh,len) ((len) -= NLMSG_ALIGN((nlh)->nlmsg_len), \ @@ -86,6 +87,8 @@ struct nlmsghdr #define NLMSG_DONE 0x3 /* End of a dump */ #define NLMSG_OVERRUN 0x4 /* Data lost */ +#define NLMSG_MIN_TYPE 0x10 /* < 0x10: reserved control messages */ + struct nlmsgerr { int error; @@ -108,6 +111,25 @@ enum { NETLINK_CONNECTED, }; +/* + * <------- NLA_HDRLEN ------> <-- NLA_ALIGN(payload)--> + * +---------------------+- - -+- - - - - - - - - -+- - -+ + * | Header | Pad | Payload | Pad | + * | (struct nlattr) | ing | | ing | + * +---------------------+- - -+- - - - - - - - - -+- - -+ + * <-------------- nlattr->nla_len --------------> + */ + +struct nlattr +{ + __u16 nla_len; + __u16 nla_type; +}; + +#define NLA_ALIGNTO 4 +#define NLA_ALIGN(len) (((len) + NLA_ALIGNTO - 1) & ~(NLA_ALIGNTO - 1)) +#define NLA_HDRLEN ((int) NLA_ALIGN(sizeof(struct nlattr))) + #ifdef __KERNEL__ #include diff --git a/include/net/netlink.h b/include/net/netlink.h new file mode 100644 index 000000000000..c99e22db9632 --- /dev/null +++ b/include/net/netlink.h @@ -0,0 +1,877 @@ +#ifndef __NET_NETLINK_H +#define __NET_NETLINK_H + +#include +#include + +/* ======================================================================== + * Netlink Messages and Attributes Interface (As Seen On TV) + * ------------------------------------------------------------------------ + * Messages Interface + * ------------------------------------------------------------------------ + * + * Message Format: + * <--- nlmsg_total_size(payload) ---> + * <-- nlmsg_msg_size(payload) -> + * +----------+- - -+-------------+- - -+-------- - - + * | nlmsghdr | Pad | Payload | Pad | nlmsghdr + * +----------+- - -+-------------+- - -+-------- - - + * nlmsg_data(nlh)---^ ^ + * nlmsg_next(nlh)-----------------------+ + * + * Payload Format: + * <---------------------- nlmsg_len(nlh) ---------------------> + * <------ hdrlen ------> <- nlmsg_attrlen(nlh, hdrlen) -> + * +----------------------+- - -+--------------------------------+ + * | Family Header | Pad | Attributes | + * +----------------------+- - -+--------------------------------+ + * nlmsg_attrdata(nlh, hdrlen)---^ + * + * Data Structures: + * struct nlmsghdr netlink message header + * + * Message Construction: + * nlmsg_new() create a new netlink message + * nlmsg_put() add a netlink message to an skb + * nlmsg_put_answer() callback based nlmsg_put() + * nlmsg_end() finanlize netlink message + * nlmsg_cancel() cancel message construction + * nlmsg_free() free a netlink message + * + * Message Sending: + * nlmsg_multicast() multicast message to several groups + * nlmsg_unicast() unicast a message to a single socket + * + * Message Length Calculations: + * nlmsg_msg_size(payload) length of message w/o padding + * nlmsg_total_size(payload) length of message w/ padding + * nlmsg_padlen(payload) length of padding at tail + * + * Message Payload Access: + * nlmsg_data(nlh) head of message payload + * nlmsg_len(nlh) length of message payload + * nlmsg_attrdata(nlh, hdrlen) head of attributes data + * nlmsg_attrlen(nlh, hdrlen) length of attributes data + * + * Message Parsing: + * nlmsg_ok(nlh, remaining) does nlh fit into remaining bytes? + * nlmsg_next(nlh, remaining) get next netlink message + * nlmsg_parse() parse attributes of a message + * nlmsg_find_attr() find an attribute in a message + * nlmsg_for_each_msg() loop over all messages + * nlmsg_validate() validate netlink message incl. attrs + * nlmsg_for_each_attr() loop over all attributes + * + * ------------------------------------------------------------------------ + * Attributes Interface + * ------------------------------------------------------------------------ + * + * Attribute Format: + * <------- nla_total_size(payload) -------> + * <---- nla_attr_size(payload) -----> + * +----------+- - -+- - - - - - - - - +- - -+-------- - - + * | Header | Pad | Payload | Pad | Header + * +----------+- - -+- - - - - - - - - +- - -+-------- - - + * <- nla_len(nla) -> ^ + * nla_data(nla)----^ | + * nla_next(nla)-----------------------------' + * + * Data Structures: + * struct nlattr netlink attribtue header + * + * Attribute Construction: + * nla_reserve(skb, type, len) reserve skb tailroom for an attribute + * nla_put(skb, type, len, data) add attribute to skb + * + * Attribute Construction for Basic Types: + * nla_put_u8(skb, type, value) add u8 attribute to skb + * nla_put_u16(skb, type, value) add u16 attribute to skb + * nla_put_u32(skb, type, value) add u32 attribute to skb + * nla_put_u64(skb, type, value) add u64 attribute to skb + * nla_put_string(skb, type, str) add string attribute to skb + * nla_put_flag(skb, type) add flag attribute to skb + * nla_put_msecs(skb, type, jiffies) add msecs attribute to skb + * + * Exceptions Based Attribute Construction: + * NLA_PUT(skb, type, len, data) add attribute to skb + * NLA_PUT_U8(skb, type, value) add u8 attribute to skb + * NLA_PUT_U16(skb, type, value) add u16 attribute to skb + * NLA_PUT_U32(skb, type, value) add u32 attribute to skb + * NLA_PUT_U64(skb, type, value) add u64 attribute to skb + * NLA_PUT_STRING(skb, type, str) add string attribute to skb + * NLA_PUT_FLAG(skb, type) add flag attribute to skb + * NLA_PUT_MSECS(skb, type, jiffies) add msecs attribute to skb + * + * The meaning of these functions is equal to their lower case + * variants but they jump to the label nla_put_failure in case + * of a failure. + * + * Nested Attributes Construction: + * nla_nest_start(skb, type) start a nested attribute + * nla_nest_end(skb, nla) finalize a nested attribute + * nla_nest_cancel(skb, nla) cancel nested attribute construction + * + * Attribute Length Calculations: + * nla_attr_size(payload) length of attribute w/o padding + * nla_total_size(payload) length of attribute w/ padding + * nla_padlen(payload) length of padding + * + * Attribute Payload Access: + * nla_data(nla) head of attribute payload + * nla_len(nla) length of attribute payload + * + * Attribute Payload Access for Basic Types: + * nla_get_u8(nla) get payload for a u8 attribute + * nla_get_u16(nla) get payload for a u16 attribute + * nla_get_u32(nla) get payload for a u32 attribute + * nla_get_u64(nla) get payload for a u64 attribute + * nla_get_flag(nla) return 1 if flag is true + * nla_get_msecs(nla) get payload for a msecs attribute + * + * Attribute Misc: + * nla_memcpy(dest, nla, count) copy attribute into memory + * nla_memcmp(nla, data, size) compare attribute with memory area + * nla_strlcpy(dst, nla, size) copy attribute to a sized string + * nla_strcmp(nla, str) compare attribute with string + * + * Attribute Parsing: + * nla_ok(nla, remaining) does nla fit into remaining bytes? + * nla_next(nla, remaining) get next netlink attribute + * nla_validate() validate a stream of attributes + * nla_find() find attribute in stream of attributes + * nla_parse() parse and validate stream of attrs + * nla_parse_nested() parse nested attribuets + * nla_for_each_attr() loop over all attributes + *========================================================================= + */ + + /** + * Standard attribute types to specify validation policy + */ +enum { + NLA_UNSPEC, + NLA_U8, + NLA_U16, + NLA_U32, + NLA_U64, + NLA_STRING, + NLA_FLAG, + NLA_MSECS, + NLA_NESTED, + __NLA_TYPE_MAX, +}; + +#define NLA_TYPE_MAX (__NLA_TYPE_MAX - 1) + +/** + * struct nla_policy - attribute validation policy + * @type: Type of attribute or NLA_UNSPEC + * @minlen: Minimal length of payload required to be available + * + * Policies are defined as arrays of this struct, the array must be + * accessible by attribute type up to the highest identifier to be expected. + * + * Example: + * static struct nla_policy my_policy[ATTR_MAX+1] __read_mostly = { + * [ATTR_FOO] = { .type = NLA_U16 }, + * [ATTR_BAR] = { .type = NLA_STRING }, + * [ATTR_BAZ] = { .minlen = sizeof(struct mystruct) }, + * }; + */ +struct nla_policy { + u16 type; + u16 minlen; +}; + +extern int nla_validate(struct nlattr *head, int len, int maxtype, + struct nla_policy *policy); +extern int nla_parse(struct nlattr *tb[], int maxtype, + struct nlattr *head, int len, + struct nla_policy *policy); +extern struct nlattr * nla_find(struct nlattr *head, int len, int attrtype); +extern size_t nla_strlcpy(char *dst, const struct nlattr *nla, + size_t dstsize); +extern int nla_memcpy(void *dest, struct nlattr *src, int count); +extern int nla_memcmp(const struct nlattr *nla, const void *data, + size_t size); +extern int nla_strcmp(const struct nlattr *nla, const char *str); +extern struct nlattr * __nla_reserve(struct sk_buff *skb, int attrtype, + int attrlen); +extern struct nlattr * nla_reserve(struct sk_buff *skb, int attrtype, + int attrlen); +extern void __nla_put(struct sk_buff *skb, int attrtype, + int attrlen, const void *data); +extern int nla_put(struct sk_buff *skb, int attrtype, + int attrlen, const void *data); + +/************************************************************************** + * Netlink Messages + **************************************************************************/ + +/** + * nlmsg_msg_size - length of netlink message not including padding + * @payload: length of message payload + */ +static inline int nlmsg_msg_size(int payload) +{ + return NLMSG_HDRLEN + payload; +} + +/** + * nlmsg_total_size - length of netlink message including padding + * @payload: length of message payload + */ +static inline int nlmsg_total_size(int payload) +{ + return NLMSG_ALIGN(nlmsg_msg_size(payload)); +} + +/** + * nlmsg_padlen - length of padding at the message's tail + * @payload: length of message payload + */ +static inline int nlmsg_padlen(int payload) +{ + return nlmsg_total_size(payload) - nlmsg_msg_size(payload); +} + +/** + * nlmsg_data - head of message payload + * @nlh: netlink messsage header + */ +static inline void *nlmsg_data(const struct nlmsghdr *nlh) +{ + return (unsigned char *) nlh + NLMSG_HDRLEN; +} + +/** + * nlmsg_len - length of message payload + * @nlh: netlink message header + */ +static inline int nlmsg_len(const struct nlmsghdr *nlh) +{ + return nlh->nlmsg_len - NLMSG_HDRLEN; +} + +/** + * nlmsg_attrdata - head of attributes data + * @nlh: netlink message header + * @hdrlen: length of family specific header + */ +static inline struct nlattr *nlmsg_attrdata(const struct nlmsghdr *nlh, + int hdrlen) +{ + unsigned char *data = nlmsg_data(nlh); + return (struct nlattr *) (data + NLMSG_ALIGN(hdrlen)); +} + +/** + * nlmsg_attrlen - length of attributes data + * @nlh: netlink message header + * @hdrlen: length of family specific header + */ +static inline int nlmsg_attrlen(const struct nlmsghdr *nlh, int hdrlen) +{ + return nlmsg_len(nlh) - NLMSG_ALIGN(hdrlen); +} + +/** + * nlmsg_ok - check if the netlink message fits into the remaining bytes + * @nlh: netlink message header + * @remaining: number of bytes remaining in message stream + */ +static inline int nlmsg_ok(const struct nlmsghdr *nlh, int remaining) +{ + return (remaining >= sizeof(struct nlmsghdr) && + nlh->nlmsg_len >= sizeof(struct nlmsghdr) && + nlh->nlmsg_len <= remaining); +} + +/** + * nlmsg_next - next netlink message in message stream + * @nlh: netlink message header + * @remaining: number of bytes remaining in message stream + * + * Returns the next netlink message in the message stream and + * decrements remaining by the size of the current message. + */ +static inline struct nlmsghdr *nlmsg_next(struct nlmsghdr *nlh, int *remaining) +{ + int totlen = NLMSG_ALIGN(nlh->nlmsg_len); + + *remaining -= totlen; + + return (struct nlmsghdr *) ((unsigned char *) nlh + totlen); +} + +/** + * nlmsg_parse - parse attributes of a netlink message + * @nlh: netlink message header + * @hdrlen: length of family specific header + * @tb: destination array with maxtype+1 elements + * @maxtype: maximum attribute type to be expected + * @policy: validation policy + * + * See nla_parse() + */ +static inline int nlmsg_parse(struct nlmsghdr *nlh, int hdrlen, + struct nlattr *tb[], int maxtype, + struct nla_policy *policy) +{ + if (nlh->nlmsg_len < nlmsg_msg_size(hdrlen)) + return -EINVAL; + + return nla_parse(tb, maxtype, nlmsg_attrdata(nlh, hdrlen), + nlmsg_attrlen(nlh, hdrlen), policy); +} + +/** + * nlmsg_find_attr - find a specific attribute in a netlink message + * @nlh: netlink message header + * @hdrlen: length of familiy specific header + * @attrtype: type of attribute to look for + * + * Returns the first attribute which matches the specified type. + */ +static inline struct nlattr *nlmsg_find_attr(struct nlmsghdr *nlh, + int hdrlen, int attrtype) +{ + return nla_find(nlmsg_attrdata(nlh, hdrlen), + nlmsg_attrlen(nlh, hdrlen), attrtype); +} + +/** + * nlmsg_validate - validate a netlink message including attributes + * @nlh: netlinket message header + * @hdrlen: length of familiy specific header + * @maxtype: maximum attribute type to be expected + * @policy: validation policy + */ +static inline int nlmsg_validate(struct nlmsghdr *nlh, int hdrlen, int maxtype, + struct nla_policy *policy) +{ + if (nlh->nlmsg_len < nlmsg_msg_size(hdrlen)) + return -EINVAL; + + return nla_validate(nlmsg_attrdata(nlh, hdrlen), + nlmsg_attrlen(nlh, hdrlen), maxtype, policy); +} + +/** + * nlmsg_for_each_attr - iterate over a stream of attributes + * @pos: loop counter, set to current attribute + * @nlh: netlink message header + * @hdrlen: length of familiy specific header + * @rem: initialized to len, holds bytes currently remaining in stream + */ +#define nlmsg_for_each_attr(pos, nlh, hdrlen, rem) \ + nla_for_each_attr(pos, nlmsg_attrdata(nlh, hdrlen), \ + nlmsg_attrlen(nlh, hdrlen), rem) + +#if 0 +/* FIXME: Enable once all users have been converted */ + +/** + * __nlmsg_put - Add a new netlink message to an skb + * @skb: socket buffer to store message in + * @pid: netlink process id + * @seq: sequence number of message + * @type: message type + * @payload: length of message payload + * @flags: message flags + * + * The caller is responsible to ensure that the skb provides enough + * tailroom for both the netlink header and payload. + */ +static inline struct nlmsghdr *__nlmsg_put(struct sk_buff *skb, u32 pid, + u32 seq, int type, int payload, + int flags) +{ + struct nlmsghdr *nlh; + + nlh = (struct nlmsghdr *) skb_put(skb, nlmsg_total_size(payload)); + nlh->nlmsg_type = type; + nlh->nlmsg_len = nlmsg_msg_size(payload); + nlh->nlmsg_flags = flags; + nlh->nlmsg_pid = pid; + nlh->nlmsg_seq = seq; + + memset((unsigned char *) nlmsg_data(nlh) + payload, 0, + nlmsg_padlen(payload)); + + return nlh; +} +#endif + +/** + * nlmsg_put - Add a new netlink message to an skb + * @skb: socket buffer to store message in + * @pid: netlink process id + * @seq: sequence number of message + * @type: message type + * @payload: length of message payload + * @flags: message flags + * + * Returns NULL if the tailroom of the skb is insufficient to store + * the message header and payload. + */ +static inline struct nlmsghdr *nlmsg_put(struct sk_buff *skb, u32 pid, u32 seq, + int type, int payload, int flags) +{ + if (unlikely(skb_tailroom(skb) < nlmsg_total_size(payload))) + return NULL; + + return __nlmsg_put(skb, pid, seq, type, payload, flags); +} + +/** + * nlmsg_put_answer - Add a new callback based netlink message to an skb + * @skb: socket buffer to store message in + * @cb: netlink callback + * @type: message type + * @payload: length of message payload + * @flags: message flags + * + * Returns NULL if the tailroom of the skb is insufficient to store + * the message header and payload. + */ +static inline struct nlmsghdr *nlmsg_put_answer(struct sk_buff *skb, + struct netlink_callback *cb, + int type, int payload, + int flags) +{ + return nlmsg_put(skb, NETLINK_CB(cb->skb).pid, cb->nlh->nlmsg_seq, + type, payload, flags); +} + +/** + * nlmsg_new - Allocate a new netlink message + * @size: maximum size of message + * + * Use NLMSG_GOODSIZE if size isn't know and you need a good default size. + */ +static inline struct sk_buff *nlmsg_new(int size) +{ + return alloc_skb(NLMSG_GOODSIZE, GFP_KERNEL); +} + +/** + * nlmsg_end - Finalize a netlink message + * @skb: socket buffer the message is stored in + * @nlh: netlink message header + * + * Corrects the netlink message header to include the appeneded + * attributes. Only necessary if attributes have been added to + * the message. + * + * Returns the total data length of the skb. + */ +static inline int nlmsg_end(struct sk_buff *skb, struct nlmsghdr *nlh) +{ + nlh->nlmsg_len = skb->tail - (unsigned char *) nlh; + + return skb->len; +} + +/** + * nlmsg_cancel - Cancel construction of a netlink message + * @skb: socket buffer the message is stored in + * @nlh: netlink message header + * + * Removes the complete netlink message including all + * attributes from the socket buffer again. Returns -1. + */ +static inline int nlmsg_cancel(struct sk_buff *skb, struct nlmsghdr *nlh) +{ + skb_trim(skb, (unsigned char *) nlh - skb->data); + + return -1; +} + +/** + * nlmsg_free - free a netlink message + * @skb: socket buffer of netlink message + */ +static inline void nlmsg_free(struct sk_buff *skb) +{ + kfree_skb(skb); +} + +/** + * nlmsg_multicast - multicast a netlink message + * @sk: netlink socket to spread messages to + * @skb: netlink message as socket buffer + * @pid: own netlink pid to avoid sending to yourself + * @group: multicast group id + */ +static inline int nlmsg_multicast(struct sock *sk, struct sk_buff *skb, + u32 pid, unsigned int group) +{ + int err; + + NETLINK_CB(skb).dst_group = group; + + err = netlink_broadcast(sk, skb, pid, group, GFP_KERNEL); + if (err > 0) + err = 0; + + return err; +} + +/** + * nlmsg_unicast - unicast a netlink message + * @sk: netlink socket to spread message to + * @skb: netlink message as socket buffer + * @pid: netlink pid of the destination socket + */ +static inline int nlmsg_unicast(struct sock *sk, struct sk_buff *skb, u32 pid) +{ + int err; + + err = netlink_unicast(sk, skb, pid, MSG_DONTWAIT); + if (err > 0) + err = 0; + + return err; +} + +/** + * nlmsg_for_each_msg - iterate over a stream of messages + * @pos: loop counter, set to current message + * @head: head of message stream + * @len: length of message stream + * @rem: initialized to len, holds bytes currently remaining in stream + */ +#define nlmsg_for_each_msg(pos, head, len, rem) \ + for (pos = head, rem = len; \ + nlmsg_ok(pos, rem); \ + pos = nlmsg_next(pos, &(rem))) + +/************************************************************************** + * Netlink Attributes + **************************************************************************/ + +/** + * nla_attr_size - length of attribute not including padding + * @payload: length of payload + */ +static inline int nla_attr_size(int payload) +{ + return NLA_HDRLEN + payload; +} + +/** + * nla_total_size - total length of attribute including padding + * @payload: length of payload + */ +static inline int nla_total_size(int payload) +{ + return NLA_ALIGN(nla_attr_size(payload)); +} + +/** + * nla_padlen - length of padding at the tail of attribute + * @payload: length of payload + */ +static inline int nla_padlen(int payload) +{ + return nla_total_size(payload) - nla_attr_size(payload); +} + +/** + * nla_data - head of payload + * @nla: netlink attribute + */ +static inline void *nla_data(const struct nlattr *nla) +{ + return (char *) nla + NLA_HDRLEN; +} + +/** + * nla_len - length of payload + * @nla: netlink attribute + */ +static inline int nla_len(const struct nlattr *nla) +{ + return nla->nla_len - NLA_HDRLEN; +} + +/** + * nla_ok - check if the netlink attribute fits into the remaining bytes + * @nla: netlink attribute + * @remaining: number of bytes remaining in attribute stream + */ +static inline int nla_ok(const struct nlattr *nla, int remaining) +{ + return remaining >= sizeof(*nla) && + nla->nla_len >= sizeof(*nla) && + nla->nla_len <= remaining; +} + +/** + * nla_next - next netlink attribte in attribute stream + * @nla: netlink attribute + * @remaining: number of bytes remaining in attribute stream + * + * Returns the next netlink attribute in the attribute stream and + * decrements remaining by the size of the current attribute. + */ +static inline struct nlattr *nla_next(const struct nlattr *nla, int *remaining) +{ + int totlen = NLA_ALIGN(nla->nla_len); + + *remaining -= totlen; + return (struct nlattr *) ((char *) nla + totlen); +} + +/** + * nla_parse_nested - parse nested attributes + * @tb: destination array with maxtype+1 elements + * @maxtype: maximum attribute type to be expected + * @nla: attribute containing the nested attributes + * @policy: validation policy + * + * See nla_parse() + */ +static inline int nla_parse_nested(struct nlattr *tb[], int maxtype, + struct nlattr *nla, + struct nla_policy *policy) +{ + return nla_parse(tb, maxtype, nla_data(nla), nla_len(nla), policy); +} +/** + * nla_put_u8 - Add a u16 netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + * @value: numeric value + */ +static inline int nla_put_u8(struct sk_buff *skb, int attrtype, u8 value) +{ + return nla_put(skb, attrtype, sizeof(u8), &value); +} + +/** + * nla_put_u16 - Add a u16 netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + * @value: numeric value + */ +static inline int nla_put_u16(struct sk_buff *skb, int attrtype, u16 value) +{ + return nla_put(skb, attrtype, sizeof(u16), &value); +} + +/** + * nla_put_u32 - Add a u32 netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + * @value: numeric value + */ +static inline int nla_put_u32(struct sk_buff *skb, int attrtype, u32 value) +{ + return nla_put(skb, attrtype, sizeof(u32), &value); +} + +/** + * nla_put_64 - Add a u64 netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + * @value: numeric value + */ +static inline int nla_put_u64(struct sk_buff *skb, int attrtype, u64 value) +{ + return nla_put(skb, attrtype, sizeof(u64), &value); +} + +/** + * nla_put_string - Add a string netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + * @str: NUL terminated string + */ +static inline int nla_put_string(struct sk_buff *skb, int attrtype, + const char *str) +{ + return nla_put(skb, attrtype, strlen(str) + 1, str); +} + +/** + * nla_put_flag - Add a flag netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + */ +static inline int nla_put_flag(struct sk_buff *skb, int attrtype) +{ + return nla_put(skb, attrtype, 0, NULL); +} + +/** + * nla_put_msecs - Add a msecs netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + * @jiffies: number of msecs in jiffies + */ +static inline int nla_put_msecs(struct sk_buff *skb, int attrtype, + unsigned long jiffies) +{ + u64 tmp = jiffies_to_msecs(jiffies); + return nla_put(skb, attrtype, sizeof(u64), &tmp); +} + +#define NLA_PUT(skb, attrtype, attrlen, data) \ + do { \ + if (nla_put(skb, attrtype, attrlen, data) < 0) \ + goto nla_put_failure; \ + } while(0) + +#define NLA_PUT_TYPE(skb, type, attrtype, value) \ + do { \ + type __tmp = value; \ + NLA_PUT(skb, attrtype, sizeof(type), &__tmp); \ + } while(0) + +#define NLA_PUT_U8(skb, attrtype, value) \ + NLA_PUT_TYPE(skb, u8, attrtype, value) + +#define NLA_PUT_U16(skb, attrtype, value) \ + NLA_PUT_TYPE(skb, u16, attrtype, value) + +#define NLA_PUT_U32(skb, attrtype, value) \ + NLA_PUT_TYPE(skb, u32, attrtype, value) + +#define NLA_PUT_U64(skb, attrtype, value) \ + NLA_PUT_TYPE(skb, u64, attrtype, value) + +#define NLA_PUT_STRING(skb, attrtype, value) \ + NLA_PUT(skb, attrtype, strlen(value) + 1, value) + +#define NLA_PUT_FLAG(skb, attrtype, value) \ + NLA_PUT(skb, attrtype, 0, NULL) + +#define NLA_PUT_MSECS(skb, attrtype, jiffies) \ + NLA_PUT_U64(skb, attrtype, jiffies_to_msecs(jiffies)) + +/** + * nla_get_u32 - return payload of u32 attribute + * @nla: u32 netlink attribute + */ +static inline u32 nla_get_u32(struct nlattr *nla) +{ + return *(u32 *) nla_data(nla); +} + +/** + * nla_get_u16 - return payload of u16 attribute + * @nla: u16 netlink attribute + */ +static inline u16 nla_get_u16(struct nlattr *nla) +{ + return *(u16 *) nla_data(nla); +} + +/** + * nla_get_u8 - return payload of u8 attribute + * @nla: u8 netlink attribute + */ +static inline u8 nla_get_u8(struct nlattr *nla) +{ + return *(u8 *) nla_data(nla); +} + +/** + * nla_get_u64 - return payload of u64 attribute + * @nla: u64 netlink attribute + */ +static inline u64 nla_get_u64(struct nlattr *nla) +{ + u64 tmp; + + nla_memcpy(&tmp, nla, sizeof(tmp)); + + return tmp; +} + +/** + * nla_get_flag - return payload of flag attribute + * @nla: flag netlink attribute + */ +static inline int nla_get_flag(struct nlattr *nla) +{ + return !!nla; +} + +/** + * nla_get_msecs - return payload of msecs attribute + * @nla: msecs netlink attribute + * + * Returns the number of milliseconds in jiffies. + */ +static inline unsigned long nla_get_msecs(struct nlattr *nla) +{ + u64 msecs = nla_get_u64(nla); + + return msecs_to_jiffies((unsigned long) msecs); +} + +/** + * nla_nest_start - Start a new level of nested attributes + * @skb: socket buffer to add attributes to + * @attrtype: attribute type of container + * + * Returns the container attribute + */ +static inline struct nlattr *nla_nest_start(struct sk_buff *skb, int attrtype) +{ + struct nlattr *start = (struct nlattr *) skb->tail; + + if (nla_put(skb, attrtype, 0, NULL) < 0) + return NULL; + + return start; +} + +/** + * nla_nest_end - Finalize nesting of attributes + * @skb: socket buffer the attribtues are stored in + * @start: container attribute + * + * Corrects the container attribute header to include the all + * appeneded attributes. + * + * Returns the total data length of the skb. + */ +static inline int nla_nest_end(struct sk_buff *skb, struct nlattr *start) +{ + start->nla_len = skb->tail - (unsigned char *) start; + return skb->len; +} + +/** + * nla_nest_cancel - Cancel nesting of attributes + * @skb: socket buffer the message is stored in + * @start: container attribute + * + * Removes the container attribute and including all nested + * attributes. Returns -1. + */ +static inline int nla_nest_cancel(struct sk_buff *skb, struct nlattr *start) +{ + if (start) + skb_trim(skb, (unsigned char *) start - skb->data); + + return -1; +} + +/** + * nla_for_each_attr - iterate over a stream of attributes + * @pos: loop counter, set to current attribute + * @head: head of attribute stream + * @len: length of attribute stream + * @rem: initialized to len, holds bytes currently remaining in stream + */ +#define nla_for_each_attr(pos, head, len, rem) \ + for (pos = head, rem = len; \ + nla_ok(pos, rem); \ + pos = nla_next(pos, &(rem))) + +#endif diff --git a/net/netlink/Makefile b/net/netlink/Makefile index 39d9c2dcd03c..68179015bd39 100644 --- a/net/netlink/Makefile +++ b/net/netlink/Makefile @@ -2,4 +2,4 @@ # Makefile for the netlink driver. # -obj-y := af_netlink.o +obj-y := af_netlink.o attr.o diff --git a/net/netlink/attr.c b/net/netlink/attr.c new file mode 100644 index 000000000000..fffef4ab276f --- /dev/null +++ b/net/netlink/attr.c @@ -0,0 +1,328 @@ +/* + * NETLINK Netlink attributes + * + * Authors: Thomas Graf + * Alexey Kuznetsov + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +static u16 nla_attr_minlen[NLA_TYPE_MAX+1] __read_mostly = { + [NLA_U8] = sizeof(u8), + [NLA_U16] = sizeof(u16), + [NLA_U32] = sizeof(u32), + [NLA_U64] = sizeof(u64), + [NLA_STRING] = 1, + [NLA_NESTED] = NLA_HDRLEN, +}; + +static int validate_nla(struct nlattr *nla, int maxtype, + struct nla_policy *policy) +{ + struct nla_policy *pt; + int minlen = 0; + + if (nla->nla_type <= 0 || nla->nla_type > maxtype) + return 0; + + pt = &policy[nla->nla_type]; + + BUG_ON(pt->type > NLA_TYPE_MAX); + + if (pt->minlen) + minlen = pt->minlen; + else if (pt->type != NLA_UNSPEC) + minlen = nla_attr_minlen[pt->type]; + + if (pt->type == NLA_FLAG && nla_len(nla) > 0) + return -ERANGE; + + if (nla_len(nla) < minlen) + return -ERANGE; + + return 0; +} + +/** + * nla_validate - Validate a stream of attributes + * @head: head of attribute stream + * @len: length of attribute stream + * @maxtype: maximum attribute type to be expected + * @policy: validation policy + * + * Validates all attributes in the specified attribute stream against the + * specified policy. Attributes with a type exceeding maxtype will be + * ignored. See documenation of struct nla_policy for more details. + * + * Returns 0 on success or a negative error code. + */ +int nla_validate(struct nlattr *head, int len, int maxtype, + struct nla_policy *policy) +{ + struct nlattr *nla; + int rem, err; + + nla_for_each_attr(nla, head, len, rem) { + err = validate_nla(nla, maxtype, policy); + if (err < 0) + goto errout; + } + + err = 0; +errout: + return err; +} + +/** + * nla_parse - Parse a stream of attributes into a tb buffer + * @tb: destination array with maxtype+1 elements + * @maxtype: maximum attribute type to be expected + * @head: head of attribute stream + * @len: length of attribute stream + * + * Parses a stream of attributes and stores a pointer to each attribute in + * the tb array accessable via the attribute type. Attributes with a type + * exceeding maxtype will be silently ignored for backwards compatibility + * reasons. policy may be set to NULL if no validation is required. + * + * Returns 0 on success or a negative error code. + */ +int nla_parse(struct nlattr *tb[], int maxtype, struct nlattr *head, int len, + struct nla_policy *policy) +{ + struct nlattr *nla; + int rem, err; + + memset(tb, 0, sizeof(struct nlattr *) * (maxtype + 1)); + + nla_for_each_attr(nla, head, len, rem) { + u16 type = nla->nla_type; + + if (type > 0 && type <= maxtype) { + if (policy) { + err = validate_nla(nla, maxtype, policy); + if (err < 0) + goto errout; + } + + tb[type] = nla; + } + } + + if (unlikely(rem > 0)) + printk(KERN_WARNING "netlink: %d bytes leftover after parsing " + "attributes.\n", rem); + + err = 0; +errout: + return err; +} + +/** + * nla_find - Find a specific attribute in a stream of attributes + * @head: head of attribute stream + * @len: length of attribute stream + * @attrtype: type of attribute to look for + * + * Returns the first attribute in the stream matching the specified type. + */ +struct nlattr *nla_find(struct nlattr *head, int len, int attrtype) +{ + struct nlattr *nla; + int rem; + + nla_for_each_attr(nla, head, len, rem) + if (nla->nla_type == attrtype) + return nla; + + return NULL; +} + +/** + * nla_strlcpy - Copy string attribute payload into a sized buffer + * @dst: where to copy the string to + * @src: attribute to copy the string from + * @dstsize: size of destination buffer + * + * Copies at most dstsize - 1 bytes into the destination buffer. + * The result is always a valid NUL-terminated string. Unlike + * strlcpy the destination buffer is always padded out. + * + * Returns the length of the source buffer. + */ +size_t nla_strlcpy(char *dst, const struct nlattr *nla, size_t dstsize) +{ + size_t srclen = nla_len(nla); + char *src = nla_data(nla); + + if (srclen > 0 && src[srclen - 1] == '\0') + srclen--; + + if (dstsize > 0) { + size_t len = (srclen >= dstsize) ? dstsize - 1 : srclen; + + memset(dst, 0, dstsize); + memcpy(dst, src, len); + } + + return srclen; +} + +/** + * nla_memcpy - Copy a netlink attribute into another memory area + * @dest: where to copy to memcpy + * @src: netlink attribute to copy from + * @count: size of the destination area + * + * Note: The number of bytes copied is limited by the length of + * attribute's payload. memcpy + * + * Returns the number of bytes copied. + */ +int nla_memcpy(void *dest, struct nlattr *src, int count) +{ + int minlen = min_t(int, count, nla_len(src)); + + memcpy(dest, nla_data(src), minlen); + + return minlen; +} + +/** + * nla_memcmp - Compare an attribute with sized memory area + * @nla: netlink attribute + * @data: memory area + * @size: size of memory area + */ +int nla_memcmp(const struct nlattr *nla, const void *data, + size_t size) +{ + int d = nla_len(nla) - size; + + if (d == 0) + d = memcmp(nla_data(nla), data, size); + + return d; +} + +/** + * nla_strcmp - Compare a string attribute against a string + * @nla: netlink string attribute + * @str: another string + */ +int nla_strcmp(const struct nlattr *nla, const char *str) +{ + int len = strlen(str) + 1; + int d = nla_len(nla) - len; + + if (d == 0) + d = memcmp(nla_data(nla), str, len); + + return d; +} + +/** + * __nla_reserve - reserve room for attribute on the skb + * @skb: socket buffer to reserve room on + * @attrtype: attribute type + * @attrlen: length of attribute payload + * + * Adds a netlink attribute header to a socket buffer and reserves + * room for the payload but does not copy it. + * + * The caller is responsible to ensure that the skb provides enough + * tailroom for the attribute header and payload. + */ +struct nlattr *__nla_reserve(struct sk_buff *skb, int attrtype, int attrlen) +{ + struct nlattr *nla; + + nla = (struct nlattr *) skb_put(skb, nla_total_size(attrlen)); + nla->nla_type = attrtype; + nla->nla_len = nla_attr_size(attrlen); + + memset((unsigned char *) nla + nla->nla_len, 0, nla_padlen(attrlen)); + + return nla; +} + +/** + * nla_reserve - reserve room for attribute on the skb + * @skb: socket buffer to reserve room on + * @attrtype: attribute type + * @attrlen: length of attribute payload + * + * Adds a netlink attribute header to a socket buffer and reserves + * room for the payload but does not copy it. + * + * Returns NULL if the tailroom of the skb is insufficient to store + * the attribute header and payload. + */ +struct nlattr *nla_reserve(struct sk_buff *skb, int attrtype, int attrlen) +{ + if (unlikely(skb_tailroom(skb) < nla_total_size(attrlen))) + return NULL; + + return __nla_reserve(skb, attrtype, attrlen); +} + +/** + * __nla_put - Add a netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + * @attrlen: length of attribute payload + * @data: head of attribute payload + * + * The caller is responsible to ensure that the skb provides enough + * tailroom for the attribute header and payload. + */ +void __nla_put(struct sk_buff *skb, int attrtype, int attrlen, + const void *data) +{ + struct nlattr *nla; + + nla = __nla_reserve(skb, attrtype, attrlen); + memcpy(nla_data(nla), data, attrlen); +} + + +/** + * nla_put - Add a netlink attribute to a socket buffer + * @skb: socket buffer to add attribute to + * @attrtype: attribute type + * @attrlen: length of attribute payload + * @data: head of attribute payload + * + * Returns -1 if the tailroom of the skb is insufficient to store + * the attribute header and payload. + */ +int nla_put(struct sk_buff *skb, int attrtype, int attrlen, const void *data) +{ + if (unlikely(skb_tailroom(skb) < nla_total_size(attrlen))) + return -1; + + __nla_put(skb, attrtype, attrlen, data); + return 0; +} + + +EXPORT_SYMBOL(nla_validate); +EXPORT_SYMBOL(nla_parse); +EXPORT_SYMBOL(nla_find); +EXPORT_SYMBOL(nla_strlcpy); +EXPORT_SYMBOL(__nla_reserve); +EXPORT_SYMBOL(nla_reserve); +EXPORT_SYMBOL(__nla_put); +EXPORT_SYMBOL(nla_put); +EXPORT_SYMBOL(nla_memcpy); +EXPORT_SYMBOL(nla_memcmp); +EXPORT_SYMBOL(nla_strcmp); From a8f74b228826eef1cbe04a05647d61e896f5fd63 Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Thu, 10 Nov 2005 02:25:52 +0100 Subject: [PATCH 059/121] [NETLINK]: Make netlink_callback->done() optional Most netlink families make no use of the done() callback, making it optional gets rid of all unnecessary dummy implementations. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- net/core/rtnetlink.c | 8 +------- net/ipv4/inet_diag.c | 9 +-------- net/ipv6/route.c | 2 +- net/netlink/af_netlink.c | 6 ++++-- net/xfrm/xfrm_user.c | 8 +------- 5 files changed, 8 insertions(+), 25 deletions(-) diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 9bed7569ce3f..193fd8637f9f 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -462,11 +462,6 @@ void rtmsg_ifinfo(int type, struct net_device *dev, unsigned change) netlink_broadcast(rtnl, skb, 0, RTNLGRP_LINK, GFP_KERNEL); } -static int rtnetlink_done(struct netlink_callback *cb) -{ - return 0; -} - /* Protected by RTNL sempahore. */ static struct rtattr **rta_buf; static int rtattr_max; @@ -533,8 +528,7 @@ rtnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *errp) goto err_inval; if ((*errp = netlink_dump_start(rtnl, skb, nlh, - link->dumpit, - rtnetlink_done)) != 0) { + link->dumpit, NULL)) != 0) { return -1; } rlen = NLMSG_ALIGN(nlh->nlmsg_len); diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c index 71f3c7350c6e..39061ed53cfd 100644 --- a/net/ipv4/inet_diag.c +++ b/net/ipv4/inet_diag.c @@ -724,12 +724,6 @@ done: return skb->len; } -static int inet_diag_dump_done(struct netlink_callback *cb) -{ - return 0; -} - - static __inline__ int inet_diag_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) { @@ -760,8 +754,7 @@ inet_diag_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh) goto err_inval; } return netlink_dump_start(idiagnl, skb, nlh, - inet_diag_dump, - inet_diag_dump_done); + inet_diag_dump, NULL); } else { return inet_diag_get_exact(skb, nlh); } diff --git a/net/ipv6/route.c b/net/ipv6/route.c index 227e99ed510c..f7f42c3e96cb 100644 --- a/net/ipv6/route.c +++ b/net/ipv6/route.c @@ -1710,7 +1710,7 @@ static void fib6_dump_end(struct netlink_callback *cb) static int fib6_dump_done(struct netlink_callback *cb) { fib6_dump_end(cb); - return cb->done(cb); + return cb->done ? cb->done(cb) : 0; } int inet6_dump_fib(struct sk_buff *skb, struct netlink_callback *cb) diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 5ca283537bc6..f3fb7e575816 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -427,7 +427,8 @@ static int netlink_release(struct socket *sock) spin_lock(&nlk->cb_lock); if (nlk->cb) { - nlk->cb->done(nlk->cb); + if (nlk->cb->done) + nlk->cb->done(nlk->cb); netlink_destroy_callback(nlk->cb); nlk->cb = NULL; } @@ -1322,7 +1323,8 @@ static int netlink_dump(struct sock *sk) skb_queue_tail(&sk->sk_receive_queue, skb); sk->sk_data_ready(sk, skb->len); - cb->done(cb); + if (cb->done) + cb->done(cb); nlk->cb = NULL; spin_unlock(&nlk->cb_lock); diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index c35336a0f71b..6996b7e85665 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -948,11 +948,6 @@ static struct xfrm_link { [XFRM_MSG_FLUSHPOLICY - XFRM_MSG_BASE] = { .doit = xfrm_flush_policy }, }; -static int xfrm_done(struct netlink_callback *cb) -{ - return 0; -} - static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *errp) { struct rtattr *xfrma[XFRMA_MAX]; @@ -990,8 +985,7 @@ static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *err goto err_einval; if ((*errp = netlink_dump_start(xfrm_nl, skb, nlh, - link->dump, - xfrm_done)) != 0) { + link->dump, NULL)) != 0) { return -1; } rlen = NLMSG_ALIGN(nlh->nlmsg_len); From 82ace47a7256fd39d370a6442e0649f75961b831 Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Thu, 10 Nov 2005 02:25:53 +0100 Subject: [PATCH 060/121] [NETLINK]: Generic netlink receive queue processor Introduces netlink_run_queue() to handle the receive queue of a netlink socket in a generic way. Processes as much as there was in the queue upon entry and invokes a callback function for each netlink message found. The callback function may refuse a message by returning a negative error code but setting the error pointer to 0 in which case netlink_run_queue() will return with a qlen != 0. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- include/net/netlink.h | 6 +++ net/netlink/af_netlink.c | 91 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 97 insertions(+) diff --git a/include/net/netlink.h b/include/net/netlink.h index c99e22db9632..640c26a90cf1 100644 --- a/include/net/netlink.h +++ b/include/net/netlink.h @@ -183,6 +183,12 @@ struct nla_policy { u16 minlen; }; +extern void netlink_run_queue(struct sock *sk, unsigned int *qlen, + int (*cb)(struct sk_buff *, + struct nlmsghdr *, int *)); +extern void netlink_queue_skip(struct nlmsghdr *nlh, + struct sk_buff *skb); + extern int nla_validate(struct nlattr *head, int len, int maxtype, struct nla_policy *policy); extern int nla_parse(struct nlattr *tb[], int maxtype, diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index f3fb7e575816..8c38ee6d255e 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -58,6 +58,7 @@ #include #include +#include #define Nprintk(a...) #define NLGRPSZ(x) (ALIGN(x, sizeof(unsigned long) * 8) / 8) @@ -1411,6 +1412,94 @@ void netlink_ack(struct sk_buff *in_skb, struct nlmsghdr *nlh, int err) netlink_unicast(in_skb->sk, skb, NETLINK_CB(in_skb).pid, MSG_DONTWAIT); } +static int netlink_rcv_skb(struct sk_buff *skb, int (*cb)(struct sk_buff *, + struct nlmsghdr *, int *)) +{ + unsigned int total_len; + struct nlmsghdr *nlh; + int err; + + while (skb->len >= nlmsg_total_size(0)) { + nlh = (struct nlmsghdr *) skb->data; + + if (skb->len < nlh->nlmsg_len) + return 0; + + total_len = min(NLMSG_ALIGN(nlh->nlmsg_len), skb->len); + + if (cb(skb, nlh, &err) < 0) { + /* Not an error, but we have to interrupt processing + * here. Note: that in this case we do not pull + * message from skb, it will be processed later. + */ + if (err == 0) + return -1; + netlink_ack(skb, nlh, err); + } else if (nlh->nlmsg_flags & NLM_F_ACK) + netlink_ack(skb, nlh, 0); + + skb_pull(skb, total_len); + } + + return 0; +} + +/** + * nelink_run_queue - Process netlink receive queue. + * @sk: Netlink socket containing the queue + * @qlen: Place to store queue length upon entry + * @cb: Callback function invoked for each netlink message found + * + * Processes as much as there was in the queue upon entry and invokes + * a callback function for each netlink message found. The callback + * function may refuse a message by returning a negative error code + * but setting the error pointer to 0 in which case this function + * returns with a qlen != 0. + * + * qlen must be initialized to 0 before the initial entry, afterwards + * the function may be called repeatedly until qlen reaches 0. + */ +void netlink_run_queue(struct sock *sk, unsigned int *qlen, + int (*cb)(struct sk_buff *, struct nlmsghdr *, int *)) +{ + struct sk_buff *skb; + + if (!*qlen || *qlen > skb_queue_len(&sk->sk_receive_queue)) + *qlen = skb_queue_len(&sk->sk_receive_queue); + + for (; *qlen; (*qlen)--) { + skb = skb_dequeue(&sk->sk_receive_queue); + if (netlink_rcv_skb(skb, cb)) { + if (skb->len) + skb_queue_head(&sk->sk_receive_queue, skb); + else { + kfree_skb(skb); + (*qlen)--; + } + break; + } + + kfree_skb(skb); + } +} + +/** + * netlink_queue_skip - Skip netlink message while processing queue. + * @nlh: Netlink message to be skipped + * @skb: Socket buffer containing the netlink messages. + * + * Pulls the given netlink message off the socket buffer so the next + * call to netlink_queue_run() will not reconsider the message. + */ +void netlink_queue_skip(struct nlmsghdr *nlh, struct sk_buff *skb) +{ + int msglen = NLMSG_ALIGN(nlh->nlmsg_len); + + if (msglen > skb->len) + msglen = skb->len; + + skb_pull(skb, msglen); +} #ifdef CONFIG_PROC_FS struct nl_seq_iter { @@ -1659,6 +1748,8 @@ out: core_initcall(netlink_proto_init); EXPORT_SYMBOL(netlink_ack); +EXPORT_SYMBOL(netlink_run_queue); +EXPORT_SYMBOL(netlink_queue_skip); EXPORT_SYMBOL(netlink_broadcast); EXPORT_SYMBOL(netlink_dump_start); EXPORT_SYMBOL(netlink_kernel_create); From 88fc2c84312d095545c08a9f871ad1888a688cf6 Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Thu, 10 Nov 2005 02:25:54 +0100 Subject: [PATCH 061/121] [XFRM]: Use generic netlink receive queue processor Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- net/xfrm/xfrm_user.c | 61 ++++---------------------------------------- 1 file changed, 5 insertions(+), 56 deletions(-) diff --git a/net/xfrm/xfrm_user.c b/net/xfrm/xfrm_user.c index 6996b7e85665..0cdd9a07e043 100644 --- a/net/xfrm/xfrm_user.c +++ b/net/xfrm/xfrm_user.c @@ -18,7 +18,6 @@ #include #include #include -#include #include #include #include @@ -26,6 +25,7 @@ #include #include #include +#include #include static struct sock *xfrm_nl; @@ -979,8 +979,6 @@ static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *err if ((type == (XFRM_MSG_GETSA - XFRM_MSG_BASE) || type == (XFRM_MSG_GETPOLICY - XFRM_MSG_BASE)) && (nlh->nlmsg_flags & NLM_F_DUMP)) { - u32 rlen; - if (link->dump == NULL) goto err_einval; @@ -988,10 +986,8 @@ static int xfrm_user_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *err link->dump, NULL)) != 0) { return -1; } - rlen = NLMSG_ALIGN(nlh->nlmsg_len); - if (rlen > skb->len) - rlen = skb->len; - skb_pull(skb, rlen); + + netlink_queue_skip(nlh, skb); return -1; } @@ -1026,60 +1022,13 @@ err_einval: return -1; } -static int xfrm_user_rcv_skb(struct sk_buff *skb) -{ - int err; - struct nlmsghdr *nlh; - - while (skb->len >= NLMSG_SPACE(0)) { - u32 rlen; - - nlh = (struct nlmsghdr *) skb->data; - if (nlh->nlmsg_len < sizeof(*nlh) || - skb->len < nlh->nlmsg_len) - return 0; - rlen = NLMSG_ALIGN(nlh->nlmsg_len); - if (rlen > skb->len) - rlen = skb->len; - if (xfrm_user_rcv_msg(skb, nlh, &err) < 0) { - if (err == 0) - return -1; - netlink_ack(skb, nlh, err); - } else if (nlh->nlmsg_flags & NLM_F_ACK) - netlink_ack(skb, nlh, 0); - skb_pull(skb, rlen); - } - - return 0; -} - static void xfrm_netlink_rcv(struct sock *sk, int len) { - unsigned int qlen = skb_queue_len(&sk->sk_receive_queue); + unsigned int qlen = 0; do { - struct sk_buff *skb; - down(&xfrm_cfg_sem); - - if (qlen > skb_queue_len(&sk->sk_receive_queue)) - qlen = skb_queue_len(&sk->sk_receive_queue); - - for (; qlen; qlen--) { - skb = skb_dequeue(&sk->sk_receive_queue); - if (xfrm_user_rcv_skb(skb)) { - if (skb->len) - skb_queue_head(&sk->sk_receive_queue, - skb); - else { - kfree_skb(skb); - qlen--; - } - break; - } - kfree_skb(skb); - } - + netlink_run_queue(sk, &qlen, &xfrm_user_rcv_msg); up(&xfrm_cfg_sem); } while (qlen); From 9ac4a16983ea4edf719c390a1a234d956947688d Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Thu, 10 Nov 2005 02:25:55 +0100 Subject: [PATCH 062/121] [RTNETLINK]: Use generic netlink receive queue processor Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- net/core/rtnetlink.c | 75 +++----------------------------------------- 1 file changed, 5 insertions(+), 70 deletions(-) diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 193fd8637f9f..8700379685e0 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -49,6 +49,7 @@ #include #include #include +#include DECLARE_MUTEX(rtnl_sem); @@ -519,8 +520,6 @@ rtnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *errp) } if (kind == 2 && nlh->nlmsg_flags&NLM_F_DUMP) { - u32 rlen; - if (link->dumpit == NULL) link = &(rtnetlink_links[PF_UNSPEC][type]); @@ -531,10 +530,8 @@ rtnetlink_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, int *errp) link->dumpit, NULL)) != 0) { return -1; } - rlen = NLMSG_ALIGN(nlh->nlmsg_len); - if (rlen > skb->len) - rlen = skb->len; - skb_pull(skb, rlen); + + netlink_queue_skip(nlh, skb); return -1; } @@ -573,75 +570,13 @@ err_inval: return -1; } -/* - * Process one packet of messages. - * Malformed skbs with wrong lengths of messages are discarded silently. - */ - -static inline int rtnetlink_rcv_skb(struct sk_buff *skb) -{ - int err; - struct nlmsghdr * nlh; - - while (skb->len >= NLMSG_SPACE(0)) { - u32 rlen; - - nlh = (struct nlmsghdr *)skb->data; - if (nlh->nlmsg_len < sizeof(*nlh) || skb->len < nlh->nlmsg_len) - return 0; - rlen = NLMSG_ALIGN(nlh->nlmsg_len); - if (rlen > skb->len) - rlen = skb->len; - if (rtnetlink_rcv_msg(skb, nlh, &err)) { - /* Not error, but we must interrupt processing here: - * Note, that in this case we do not pull message - * from skb, it will be processed later. - */ - if (err == 0) - return -1; - netlink_ack(skb, nlh, err); - } else if (nlh->nlmsg_flags&NLM_F_ACK) - netlink_ack(skb, nlh, 0); - skb_pull(skb, rlen); - } - - return 0; -} - -/* - * rtnetlink input queue processing routine: - * - process as much as there was in the queue upon entry. - * - feed skbs to rtnetlink_rcv_skb, until it refuse a message, - * that will occur, when a dump started. - */ - static void rtnetlink_rcv(struct sock *sk, int len) { - unsigned int qlen = skb_queue_len(&sk->sk_receive_queue); + unsigned int qlen = 0; do { - struct sk_buff *skb; - rtnl_lock(); - - if (qlen > skb_queue_len(&sk->sk_receive_queue)) - qlen = skb_queue_len(&sk->sk_receive_queue); - - for (; qlen; qlen--) { - skb = skb_dequeue(&sk->sk_receive_queue); - if (rtnetlink_rcv_skb(skb)) { - if (skb->len) - skb_queue_head(&sk->sk_receive_queue, - skb); - else { - kfree_skb(skb); - qlen--; - } - break; - } - kfree_skb(skb); - } - + netlink_run_queue(sk, &qlen, &rtnetlink_rcv_msg); up(&rtnl_sem); netdev_run_todo(); From 482a8524f85a7d8c40c6fb5d072e85bc2fef327f Mon Sep 17 00:00:00 2001 From: Thomas Graf Date: Thu, 10 Nov 2005 02:25:56 +0100 Subject: [PATCH 063/121] [NETLINK]: Generic netlink family The generic netlink family builds on top of netlink and provides simplifies access for the less demanding netlink users. It solves the problem of protocol numbers running out by introducing a so called controller taking care of id management and name resolving. Generic netlink modules register themself after filling out their id card (struct genl_family), after successful registration the modules are able to register callbacks to command numbers by filling out a struct genl_ops and calling genl_register_op(). The registered callbacks are invoked with attributes parsed making life of simple modules a lot easier. Although generic netlink modules can request static identifiers, it is recommended to use GENL_ID_GENERATE and to let the controller assign a unique identifier to the module. Userspace applications will then ask the controller and lookup the idenfier by the module name. Due to the current multicast implementation of netlink, the number of generic netlink modules is restricted to 1024 to avoid wasting memory for the per socket multiacst subscription bitmask. Signed-off-by: Thomas Graf Signed-off-by: David S. Miller --- include/linux/genetlink.h | 51 ++++ include/net/genetlink.h | 154 ++++++++++ net/netlink/Makefile | 2 +- net/netlink/genetlink.c | 579 ++++++++++++++++++++++++++++++++++++++ 4 files changed, 785 insertions(+), 1 deletion(-) create mode 100644 include/linux/genetlink.h create mode 100644 include/net/genetlink.h create mode 100644 net/netlink/genetlink.c diff --git a/include/linux/genetlink.h b/include/linux/genetlink.h new file mode 100644 index 000000000000..84f12a41dc01 --- /dev/null +++ b/include/linux/genetlink.h @@ -0,0 +1,51 @@ +#ifndef __LINUX_GENERIC_NETLINK_H +#define __LINUX_GENERIC_NETLINK_H + +#include + +#define GENL_NAMSIZ 16 /* length of family name */ + +#define GENL_MIN_ID NLMSG_MIN_TYPE +#define GENL_MAX_ID 1023 + +struct genlmsghdr { + __u8 cmd; + __u8 version; + __u16 reserved; +}; + +#define GENL_HDRLEN NLMSG_ALIGN(sizeof(struct genlmsghdr)) + +/* + * List of reserved static generic netlink identifiers: + */ +#define GENL_ID_GENERATE 0 +#define GENL_ID_CTRL NLMSG_MIN_TYPE + +/************************************************************************** + * Controller + **************************************************************************/ + +enum { + CTRL_CMD_UNSPEC, + CTRL_CMD_NEWFAMILY, + CTRL_CMD_DELFAMILY, + CTRL_CMD_GETFAMILY, + CTRL_CMD_NEWOPS, + CTRL_CMD_DELOPS, + CTRL_CMD_GETOPS, + __CTRL_CMD_MAX, +}; + +#define CTRL_CMD_MAX (__CTRL_CMD_MAX - 1) + +enum { + CTRL_ATTR_UNSPEC, + CTRL_ATTR_FAMILY_ID, + CTRL_ATTR_FAMILY_NAME, + __CTRL_ATTR_MAX, +}; + +#define CTRL_ATTR_MAX (__CTRL_ATTR_MAX - 1) + +#endif /* __LINUX_GENERIC_NETLINK_H */ diff --git a/include/net/genetlink.h b/include/net/genetlink.h new file mode 100644 index 000000000000..52d8b1a73d52 --- /dev/null +++ b/include/net/genetlink.h @@ -0,0 +1,154 @@ +#ifndef __NET_GENERIC_NETLINK_H +#define __NET_GENERIC_NETLINK_H + +#include +#include + +/** + * struct genl_family - generic netlink family + * @id: protocol family idenfitier + * @hdrsize: length of user specific header in bytes + * @name: name of family + * @version: protocol version + * @maxattr: maximum number of attributes supported + * @attrbuf: buffer to store parsed attributes + * @ops_list: list of all assigned operations + * @family_list: family list + */ +struct genl_family +{ + unsigned int id; + unsigned int hdrsize; + char name[GENL_NAMSIZ]; + unsigned int version; + unsigned int maxattr; + struct module * owner; + struct nlattr ** attrbuf; /* private */ + struct list_head ops_list; /* private */ + struct list_head family_list; /* private */ +}; + +#define GENL_ADMIN_PERM 0x01 + +/** + * struct genl_info - receiving information + * @snd_seq: sending sequence number + * @snd_pid: netlink pid of sender + * @nlhdr: netlink message header + * @genlhdr: generic netlink message header + * @userhdr: user specific header + * @attrs: netlink attributes + */ +struct genl_info +{ + u32 snd_seq; + u32 snd_pid; + struct nlmsghdr * nlhdr; + struct genlmsghdr * genlhdr; + void * userhdr; + struct nlattr ** attrs; +}; + +/** + * struct genl_ops - generic netlink operations + * @cmd: command identifier + * @flags: flags + * @policy: attribute validation policy + * @doit: standard command callback + * @dumpit: callback for dumpers + * @ops_list: operations list + */ +struct genl_ops +{ + unsigned int cmd; + unsigned int flags; + struct nla_policy *policy; + int (*doit)(struct sk_buff *skb, + struct genl_info *info); + int (*dumpit)(struct sk_buff *skb, + struct netlink_callback *cb); + struct list_head ops_list; +}; + +extern int genl_register_family(struct genl_family *family); +extern int genl_unregister_family(struct genl_family *family); +extern int genl_register_ops(struct genl_family *, struct genl_ops *ops); +extern int genl_unregister_ops(struct genl_family *, struct genl_ops *ops); + +extern struct sock *genl_sock; + +/** + * genlmsg_put - Add generic netlink header to netlink message + * @skb: socket buffer holding the message + * @pid: netlink pid the message is addressed to + * @seq: sequence number (usually the one of the sender) + * @type: netlink message type + * @hdrlen: length of the user specific header + * @flags netlink message flags + * @cmd: generic netlink command + * @version: version + * + * Returns pointer to user specific header + */ +static inline void *genlmsg_put(struct sk_buff *skb, u32 pid, u32 seq, + int type, int hdrlen, int flags, + u8 cmd, u8 version) +{ + struct nlmsghdr *nlh; + struct genlmsghdr *hdr; + + nlh = nlmsg_put(skb, pid, seq, type, GENL_HDRLEN + hdrlen, flags); + if (nlh == NULL) + return NULL; + + hdr = nlmsg_data(nlh); + hdr->cmd = cmd; + hdr->version = version; + hdr->reserved = 0; + + return (char *) hdr + GENL_HDRLEN; +} + +/** + * genlmsg_end - Finalize a generic netlink message + * @skb: socket buffer the message is stored in + * @hdr: user specific header + */ +static inline int genlmsg_end(struct sk_buff *skb, void *hdr) +{ + return nlmsg_end(skb, hdr - GENL_HDRLEN - NLMSG_HDRLEN); +} + +/** + * genlmsg_cancel - Cancel construction of a generic netlink message + * @skb: socket buffer the message is stored in + * @hdr: generic netlink message header + */ +static inline int genlmsg_cancel(struct sk_buff *skb, void *hdr) +{ + return nlmsg_cancel(skb, hdr - GENL_HDRLEN - NLMSG_HDRLEN); +} + +/** + * genlmsg_multicast - multicast a netlink message + * @skb: netlink message as socket buffer + * @pid: own netlink pid to avoid sending to yourself + * @group: multicast group id + */ +static inline int genlmsg_multicast(struct sk_buff *skb, u32 pid, + unsigned int group) +{ + return nlmsg_multicast(genl_sock, skb, pid, group); +} + +/** + * genlmsg_unicast - unicast a netlink message + * @skb: netlink message as socket buffer + * @pid: netlink pid of the destination socket + */ +static inline int genlmsg_unicast(struct sk_buff *skb, u32 pid) +{ + return nlmsg_unicast(genl_sock, skb, pid); +} + +#endif /* __NET_GENERIC_NETLINK_H */ diff --git a/net/netlink/Makefile b/net/netlink/Makefile index 68179015bd39..e3589c2de49e 100644 --- a/net/netlink/Makefile +++ b/net/netlink/Makefile @@ -2,4 +2,4 @@ # Makefile for the netlink driver. # -obj-y := af_netlink.o attr.o +obj-y := af_netlink.o attr.o genetlink.o diff --git a/net/netlink/genetlink.c b/net/netlink/genetlink.c new file mode 100644 index 000000000000..287cfcc56951 --- /dev/null +++ b/net/netlink/genetlink.c @@ -0,0 +1,579 @@ +/* + * NETLINK Generic Netlink Family + * + * Authors: Jamal Hadi Salim + * Thomas Graf + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +struct sock *genl_sock = NULL; + +static DECLARE_MUTEX(genl_sem); /* serialization of message processing */ + +static void genl_lock(void) +{ + down(&genl_sem); +} + +static int genl_trylock(void) +{ + return down_trylock(&genl_sem); +} + +static void genl_unlock(void) +{ + up(&genl_sem); + + if (genl_sock && genl_sock->sk_receive_queue.qlen) + genl_sock->sk_data_ready(genl_sock, 0); +} + +#define GENL_FAM_TAB_SIZE 16 +#define GENL_FAM_TAB_MASK (GENL_FAM_TAB_SIZE - 1) + +static struct list_head family_ht[GENL_FAM_TAB_SIZE]; + +static int genl_ctrl_event(int event, void *data); + +static inline unsigned int genl_family_hash(unsigned int id) +{ + return id & GENL_FAM_TAB_MASK; +} + +static inline struct list_head *genl_family_chain(unsigned int id) +{ + return &family_ht[genl_family_hash(id)]; +} + +static struct genl_family *genl_family_find_byid(unsigned int id) +{ + struct genl_family *f; + + list_for_each_entry(f, genl_family_chain(id), family_list) + if (f->id == id) + return f; + + return NULL; +} + +static struct genl_family *genl_family_find_byname(char *name) +{ + struct genl_family *f; + int i; + + for (i = 0; i < GENL_FAM_TAB_SIZE; i++) + list_for_each_entry(f, genl_family_chain(i), family_list) + if (strcmp(f->name, name) == 0) + return f; + + return NULL; +} + +static struct genl_ops *genl_get_cmd(u8 cmd, struct genl_family *family) +{ + struct genl_ops *ops; + + list_for_each_entry(ops, &family->ops_list, ops_list) + if (ops->cmd == cmd) + return ops; + + return NULL; +} + +/* Of course we are going to have problems once we hit + * 2^16 alive types, but that can only happen by year 2K +*/ +static inline u16 genl_generate_id(void) +{ + static u16 id_gen_idx; + int overflowed = 0; + + do { + if (id_gen_idx == 0) + id_gen_idx = GENL_MIN_ID; + + if (++id_gen_idx > GENL_MAX_ID) { + if (!overflowed) { + overflowed = 1; + id_gen_idx = 0; + continue; + } else + return 0; + } + + } while (genl_family_find_byid(id_gen_idx)); + + return id_gen_idx; +} + +/** + * genl_register_ops - register generic netlink operations + * @family: generic netlink family + * @ops: operations to be registered + * + * Registers the specified operations and assigns them to the specified + * family. Either a doit or dumpit callback must be specified or the + * operation will fail. Only one operation structure per command + * identifier may be registered. + * + * See include/net/genetlink.h for more documenation on the operations + * structure. + * + * Returns 0 on success or a negative error code. + */ +int genl_register_ops(struct genl_family *family, struct genl_ops *ops) +{ + int err = -EINVAL; + + if (ops->dumpit == NULL && ops->doit == NULL) + goto errout; + + if (genl_get_cmd(ops->cmd, family)) { + err = -EEXIST; + goto errout; + } + + genl_lock(); + list_add_tail(&ops->ops_list, &family->ops_list); + genl_unlock(); + + genl_ctrl_event(CTRL_CMD_NEWOPS, ops); + err = 0; +errout: + return err; +} + +/** + * genl_unregister_ops - unregister generic netlink operations + * @family: generic netlink family + * @ops: operations to be unregistered + * + * Unregisters the specified operations and unassigns them from the + * specified family. The operation blocks until the current message + * processing has finished and doesn't start again until the + * unregister process has finished. + * + * Note: It is not necessary to unregister all operations before + * unregistering the family, unregistering the family will cause + * all assigned operations to be unregistered automatically. + * + * Returns 0 on success or a negative error code. + */ +int genl_unregister_ops(struct genl_family *family, struct genl_ops *ops) +{ + struct genl_ops *rc; + + genl_lock(); + list_for_each_entry(rc, &family->ops_list, ops_list) { + if (rc == ops) { + list_del(&ops->ops_list); + genl_unlock(); + genl_ctrl_event(CTRL_CMD_DELOPS, ops); + return 0; + } + } + genl_unlock(); + + return -ENOENT; +} + +/** + * genl_register_family - register a generic netlink family + * @family: generic netlink family + * + * Registers the specified family after validating it first. Only one + * family may be registered with the same family name or identifier. + * The family id may equal GENL_ID_GENERATE causing an unique id to + * be automatically generated and assigned. + * + * Return 0 on success or a negative error code. + */ +int genl_register_family(struct genl_family *family) +{ + int err = -EINVAL; + + if (family->id && family->id < GENL_MIN_ID) + goto errout; + + if (family->id > GENL_MAX_ID) + goto errout; + + INIT_LIST_HEAD(&family->ops_list); + + genl_lock(); + + if (genl_family_find_byname(family->name)) { + err = -EEXIST; + goto errout_locked; + } + + if (genl_family_find_byid(family->id)) { + err = -EEXIST; + goto errout_locked; + } + + if (!try_module_get(family->owner)) { + err = -EBUSY; + goto errout_locked; + } + + if (family->id == GENL_ID_GENERATE) { + u16 newid = genl_generate_id(); + + if (!newid) { + err = -ENOMEM; + goto errout_locked; + } + + family->id = newid; + } + + if (family->maxattr) { + family->attrbuf = kmalloc((family->maxattr+1) * + sizeof(struct nlattr *), GFP_KERNEL); + if (family->attrbuf == NULL) { + err = -ENOMEM; + goto errout; + } + } else + family->attrbuf = NULL; + + list_add_tail(&family->family_list, genl_family_chain(family->id)); + genl_unlock(); + + genl_ctrl_event(CTRL_CMD_NEWFAMILY, family); + + return 0; + +errout_locked: + genl_unlock(); +errout: + return err; +} + +/** + * genl_unregister_family - unregister generic netlink family + * @family: generic netlink family + * + * Unregisters the specified family. + * + * Returns 0 on success or a negative error code. + */ +int genl_unregister_family(struct genl_family *family) +{ + struct genl_family *rc; + + genl_lock(); + + list_for_each_entry(rc, genl_family_chain(family->id), family_list) { + if (family->id != rc->id || strcmp(rc->name, family->name)) + continue; + + list_del(&rc->family_list); + INIT_LIST_HEAD(&family->ops_list); + genl_unlock(); + + module_put(family->owner); + kfree(family->attrbuf); + genl_ctrl_event(CTRL_CMD_DELFAMILY, family); + return 0; + } + + genl_unlock(); + + return -ENOENT; +} + +static inline int genl_rcv_msg(struct sk_buff *skb, struct nlmsghdr *nlh, + int *errp) +{ + struct genl_ops *ops; + struct genl_family *family; + struct genl_info info; + struct genlmsghdr *hdr = nlmsg_data(nlh); + int hdrlen, err = -EINVAL; + + if (!(nlh->nlmsg_flags & NLM_F_REQUEST)) + goto ignore; + + if (nlh->nlmsg_type < NLMSG_MIN_TYPE) + goto ignore; + + family = genl_family_find_byid(nlh->nlmsg_type); + if (family == NULL) { + err = -ENOENT; + goto errout; + } + + hdrlen = GENL_HDRLEN + family->hdrsize; + if (nlh->nlmsg_len < nlmsg_msg_size(hdrlen)) + goto errout; + + ops = genl_get_cmd(hdr->cmd, family); + if (ops == NULL) { + err = -EOPNOTSUPP; + goto errout; + } + + if ((ops->flags & GENL_ADMIN_PERM) && security_netlink_recv(skb)) { + err = -EPERM; + goto errout; + } + + if (nlh->nlmsg_flags & NLM_F_DUMP) { + if (ops->dumpit == NULL) { + err = -EOPNOTSUPP; + goto errout; + } + + *errp = err = netlink_dump_start(genl_sock, skb, nlh, + ops->dumpit, NULL); + if (err == 0) + skb_pull(skb, min(NLMSG_ALIGN(nlh->nlmsg_len), + skb->len)); + return -1; + } + + if (ops->doit == NULL) { + err = -EOPNOTSUPP; + goto errout; + } + + if (family->attrbuf) { + err = nlmsg_parse(nlh, hdrlen, family->attrbuf, family->maxattr, + ops->policy); + if (err < 0) + goto errout; + } + + info.snd_seq = nlh->nlmsg_seq; + info.snd_pid = NETLINK_CB(skb).pid; + info.nlhdr = nlh; + info.genlhdr = nlmsg_data(nlh); + info.userhdr = nlmsg_data(nlh) + GENL_HDRLEN; + info.attrs = family->attrbuf; + + *errp = err = ops->doit(skb, &info); + return err; + +ignore: + return 0; + +errout: + *errp = err; + return -1; +} + +static void genl_rcv(struct sock *sk, int len) +{ + unsigned int qlen = 0; + + do { + if (genl_trylock()) + return; + netlink_run_queue(sk, &qlen, &genl_rcv_msg); + genl_unlock(); + } while (qlen && genl_sock && genl_sock->sk_receive_queue.qlen); +} + +/************************************************************************** + * Controller + **************************************************************************/ + +static int ctrl_fill_info(struct genl_family *family, u32 pid, u32 seq, + u32 flags, struct sk_buff *skb, u8 cmd) +{ + void *hdr; + + hdr = genlmsg_put(skb, pid, seq, GENL_ID_CTRL, 0, flags, cmd, + family->version); + if (hdr == NULL) + return -1; + + NLA_PUT_STRING(skb, CTRL_ATTR_FAMILY_NAME, family->name); + NLA_PUT_U16(skb, CTRL_ATTR_FAMILY_ID, family->id); + + return genlmsg_end(skb, hdr); + +nla_put_failure: + return genlmsg_cancel(skb, hdr); +} + +static int ctrl_dumpfamily(struct sk_buff *skb, struct netlink_callback *cb) +{ + + int i, n = 0; + struct genl_family *rt; + int chains_to_skip = cb->args[0]; + int fams_to_skip = cb->args[1]; + + for (i = 0; i < GENL_FAM_TAB_SIZE; i++) { + if (i < chains_to_skip) + continue; + n = 0; + list_for_each_entry(rt, genl_family_chain(i), family_list) { + if (++n < fams_to_skip) + continue; + if (ctrl_fill_info(rt, NETLINK_CB(cb->skb).pid, + cb->nlh->nlmsg_seq, NLM_F_MULTI, + skb, CTRL_CMD_NEWFAMILY) < 0) + goto errout; + } + + fams_to_skip = 0; + } + +errout: + cb->args[0] = i; + cb->args[1] = n; + + return skb->len; +} + +static struct sk_buff *ctrl_build_msg(struct genl_family *family, u32 pid, + int seq, int cmd) +{ + struct sk_buff *skb; + int err; + + skb = nlmsg_new(NLMSG_GOODSIZE); + if (skb == NULL) + return ERR_PTR(-ENOBUFS); + + err = ctrl_fill_info(family, pid, seq, 0, skb, cmd); + if (err < 0) { + nlmsg_free(skb); + return ERR_PTR(err); + } + + return skb; +} + +static struct nla_policy ctrl_policy[CTRL_ATTR_MAX+1] __read_mostly = { + [CTRL_ATTR_FAMILY_ID] = { .type = NLA_U16 }, + [CTRL_ATTR_FAMILY_NAME] = { .type = NLA_STRING }, +}; + +static int ctrl_getfamily(struct sk_buff *skb, struct genl_info *info) +{ + struct sk_buff *msg; + struct genl_family *res = NULL; + int err = -EINVAL; + + if (info->attrs[CTRL_ATTR_FAMILY_ID]) { + u16 id = nla_get_u16(info->attrs[CTRL_ATTR_FAMILY_ID]); + res = genl_family_find_byid(id); + } + + if (info->attrs[CTRL_ATTR_FAMILY_NAME]) { + char name[GENL_NAMSIZ]; + + if (nla_strlcpy(name, info->attrs[CTRL_ATTR_FAMILY_NAME], + GENL_NAMSIZ) >= GENL_NAMSIZ) + goto errout; + + res = genl_family_find_byname(name); + } + + if (res == NULL) { + err = -ENOENT; + goto errout; + } + + msg = ctrl_build_msg(res, info->snd_pid, info->snd_seq, + CTRL_CMD_NEWFAMILY); + if (IS_ERR(msg)) { + err = PTR_ERR(msg); + goto errout; + } + + err = genlmsg_unicast(msg, info->snd_pid); +errout: + return err; +} + +static int genl_ctrl_event(int event, void *data) +{ + struct sk_buff *msg; + + if (genl_sock == NULL) + return 0; + + switch (event) { + case CTRL_CMD_NEWFAMILY: + case CTRL_CMD_DELFAMILY: + msg = ctrl_build_msg(data, 0, 0, event); + if (IS_ERR(msg)) + return PTR_ERR(msg); + + genlmsg_multicast(msg, 0, GENL_ID_CTRL); + break; + } + + return 0; +} + +static struct genl_ops genl_ctrl_ops = { + .cmd = CTRL_CMD_GETFAMILY, + .doit = ctrl_getfamily, + .dumpit = ctrl_dumpfamily, + .policy = ctrl_policy, +}; + +static struct genl_family genl_ctrl = { + .id = GENL_ID_CTRL, + .name = "nlctrl", + .version = 0x1, + .maxattr = CTRL_ATTR_MAX, + .owner = THIS_MODULE, +}; + +static int __init genl_init(void) +{ + int i, err; + + for (i = 0; i < GENL_FAM_TAB_SIZE; i++) + INIT_LIST_HEAD(&family_ht[i]); + + err = genl_register_family(&genl_ctrl); + if (err < 0) + goto errout; + + err = genl_register_ops(&genl_ctrl, &genl_ctrl_ops); + if (err < 0) + goto errout_register; + + netlink_set_nonroot(NETLINK_GENERIC, NL_NONROOT_RECV); + genl_sock = netlink_kernel_create(NETLINK_GENERIC, GENL_MAX_ID, + genl_rcv, THIS_MODULE); + if (genl_sock == NULL) { + panic("GENL: Cannot initialize generic netlink\n"); + return -ENOMEM; + } + + return 0; + +errout_register: + genl_unregister_family(&genl_ctrl); +errout: + panic("GENL: Cannot register controller: %d\n", err); + return err; +} + +subsys_initcall(genl_init); + +EXPORT_SYMBOL(genl_sock); +EXPORT_SYMBOL(genl_register_ops); +EXPORT_SYMBOL(genl_unregister_ops); +EXPORT_SYMBOL(genl_register_family); +EXPORT_SYMBOL(genl_unregister_family); From 26ef5c09576496dfd08d2b36ec1d08a6f917a0eb Mon Sep 17 00:00:00 2001 From: David Gibson Date: Thu, 10 Nov 2005 11:50:16 +1100 Subject: [PATCH 064/121] [PATCH] powerpc: Merge cacheflush.h and cache.h The ppc32 and ppc64 versions of cacheflush.h were almost identical. The two versions of cache.h are fairly similar, except for a bunch of register definitions in the ppc32 version which probably belong better elsewhere. This patch, therefore, merges both headers. Notable points: - there are several functions in cacheflush.h which exist only on ppc32 or only on ppc64. These are handled by #ifdef for now, but these should probably be consolidated, along with the actual code behind them later. - Confusingly, both ppc32 and ppc64 have a flush_dcache_range(), but they're subtly different: it uses dcbf on ppc32 and dcbst on ppc64, ppc64 has a flush_inval_dcache_range() which uses dcbf. These too should be merged and consolidated later. - Also flush_dcache_range() was defined in cacheflush.h on ppc64, and in cache.h on ppc32. In the merged version it's in cacheflush.h - On ppc32 flush_icache_range() is a normal function from misc.S. On ppc64, it was wrapper, testing a feature bit before calling __flush_icache_range() which does the actual flush. This patch takes the ppc64 approach, which amounts to no change on ppc32, since CPU_FTR_COHERENT_ICACHE will never be set there, but does mean renaming flush_icache_range() to __flush_icache_range() in arch/ppc/kernel/misc.S and arch/powerpc/kernel/misc_32.S - The PReP register info from asm-ppc/cache.h has moved to arch/ppc/platforms/prep_setup.c - The 8xx register info from asm-ppc/cache.h has moved to a new asm-powerpc/reg_8xx.h, included from reg.h - flush_dcache_all() was defined on ppc32 (only), but was never called (although it was exported). Thus this patch removes it from cacheflush.h and from ARCH=powerpc (misc_32.S) entirely. It's left in ARCH=ppc for now, with the prototype moved to ppc_ksyms.c. Built for Walnut (ARCH=ppc), 32-bit multiplatform (pmac, CHRP and PReP ARCH=ppc, pmac and CHRP ARCH=powerpc). Built and booted on POWER5 LPAR (ARCH=powerpc and ARCH=ppc64). Built for 32-bit powermac (ARCH=ppc and ARCH=powerpc). Built and booted on POWER5 LPAR (ARCH=powerpc and ARCH=ppc64). Built and booted on G5 (ARCH=powerpc) Signed-off-by: David Gibson Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/misc_32.S | 23 +------ arch/ppc/kernel/misc.S | 4 +- arch/ppc/kernel/ppc_ksyms.c | 1 + arch/ppc/platforms/prep_setup.c | 9 +++ include/asm-powerpc/cache.h | 40 +++++++++++ .../{asm-ppc64 => asm-powerpc}/cacheflush.h | 66 ++++++++++++------- include/asm-powerpc/reg.h | 6 +- .../cache.h => asm-powerpc/reg_8xx.h} | 50 ++------------ include/asm-ppc/cacheflush.h | 49 -------------- include/asm-ppc64/cache.h | 36 ---------- 10 files changed, 105 insertions(+), 179 deletions(-) create mode 100644 include/asm-powerpc/cache.h rename include/{asm-ppc64 => asm-powerpc}/cacheflush.h (56%) rename include/{asm-ppc/cache.h => asm-powerpc/reg_8xx.h} (50%) delete mode 100644 include/asm-ppc/cacheflush.h delete mode 100644 include/asm-ppc64/cache.h diff --git a/arch/powerpc/kernel/misc_32.S b/arch/powerpc/kernel/misc_32.S index 3bedb532aed9..f6d84a75ed26 100644 --- a/arch/powerpc/kernel/misc_32.S +++ b/arch/powerpc/kernel/misc_32.S @@ -519,7 +519,7 @@ END_FTR_SECTION_IFCLR(CPU_FTR_SPLIT_ID_CACHE) * * flush_icache_range(unsigned long start, unsigned long stop) */ -_GLOBAL(flush_icache_range) +_GLOBAL(__flush_icache_range) BEGIN_FTR_SECTION blr /* for 601, do nothing */ END_FTR_SECTION_IFCLR(CPU_FTR_SPLIT_ID_CACHE) @@ -607,27 +607,6 @@ _GLOBAL(invalidate_dcache_range) sync /* wait for dcbi's to get to ram */ blr -#ifdef CONFIG_NOT_COHERENT_CACHE -/* - * 40x cores have 8K or 16K dcache and 32 byte line size. - * 44x has a 32K dcache and 32 byte line size. - * 8xx has 1, 2, 4, 8K variants. - * For now, cover the worst case of the 44x. - * Must be called with external interrupts disabled. - */ -#define CACHE_NWAYS 64 -#define CACHE_NLINES 16 - -_GLOBAL(flush_dcache_all) - li r4, (2 * CACHE_NWAYS * CACHE_NLINES) - mtctr r4 - lis r5, KERNELBASE@h -1: lwz r3, 0(r5) /* Load one word from every line */ - addi r5, r5, L1_CACHE_BYTES - bdnz 1b - blr -#endif /* CONFIG_NOT_COHERENT_CACHE */ - /* * Flush a particular page from the data cache to RAM. * Note: this is necessary because the instruction cache does *not* diff --git a/arch/ppc/kernel/misc.S b/arch/ppc/kernel/misc.S index ae6af29938a1..5e61124581d0 100644 --- a/arch/ppc/kernel/misc.S +++ b/arch/ppc/kernel/misc.S @@ -497,9 +497,9 @@ END_FTR_SECTION_IFCLR(CPU_FTR_SPLIT_ID_CACHE) * and invalidate the corresponding instruction cache blocks. * This is a no-op on the 601. * - * flush_icache_range(unsigned long start, unsigned long stop) + * __flush_icache_range(unsigned long start, unsigned long stop) */ -_GLOBAL(flush_icache_range) +_GLOBAL(__flush_icache_range) BEGIN_FTR_SECTION blr /* for 601, do nothing */ END_FTR_SECTION_IFCLR(CPU_FTR_SPLIT_ID_CACHE) diff --git a/arch/ppc/kernel/ppc_ksyms.c b/arch/ppc/kernel/ppc_ksyms.c index 6550de73a855..307077f14936 100644 --- a/arch/ppc/kernel/ppc_ksyms.c +++ b/arch/ppc/kernel/ppc_ksyms.c @@ -175,6 +175,7 @@ EXPORT_SYMBOL(pci_bus_to_phys); #endif /* CONFIG_PCI */ #ifdef CONFIG_NOT_COHERENT_CACHE +extern void flush_dcache_all(void); EXPORT_SYMBOL(flush_dcache_all); #endif diff --git a/arch/ppc/platforms/prep_setup.c b/arch/ppc/platforms/prep_setup.c index 067d7d53b81e..4415748071dc 100644 --- a/arch/ppc/platforms/prep_setup.c +++ b/arch/ppc/platforms/prep_setup.c @@ -61,6 +61,15 @@ #include #include +/* prep registers for L2 */ +#define CACHECRBA 0x80000823 /* Cache configuration register address */ +#define L2CACHE_MASK 0x03 /* Mask for 2 L2 Cache bits */ +#define L2CACHE_512KB 0x00 /* 512KB */ +#define L2CACHE_256KB 0x01 /* 256KB */ +#define L2CACHE_1MB 0x02 /* 1MB */ +#define L2CACHE_NONE 0x03 /* NONE */ +#define L2CACHE_PARITY 0x08 /* Mask for L2 Cache Parity Protected bit */ + TODC_ALLOC(); unsigned char ucSystemType; diff --git a/include/asm-powerpc/cache.h b/include/asm-powerpc/cache.h new file mode 100644 index 000000000000..26ce502e76e8 --- /dev/null +++ b/include/asm-powerpc/cache.h @@ -0,0 +1,40 @@ +#ifndef _ASM_POWERPC_CACHE_H +#define _ASM_POWERPC_CACHE_H + +#ifdef __KERNEL__ + +#include + +/* bytes per L1 cache line */ +#if defined(CONFIG_8xx) || defined(CONFIG_403GCX) +#define L1_CACHE_SHIFT 4 +#define MAX_COPY_PREFETCH 1 +#elif defined(CONFIG_PPC32) +#define L1_CACHE_SHIFT 5 +#define MAX_COPY_PREFETCH 4 +#else /* CONFIG_PPC64 */ +#define L1_CACHE_SHIFT 7 +#endif + +#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) + +#define SMP_CACHE_BYTES L1_CACHE_BYTES +#define L1_CACHE_SHIFT_MAX 7 /* largest L1 which this arch supports */ + +#if defined(__powerpc64__) && !defined(__ASSEMBLY__) +struct ppc64_caches { + u32 dsize; /* L1 d-cache size */ + u32 dline_size; /* L1 d-cache line size */ + u32 log_dline_size; + u32 dlines_per_page; + u32 isize; /* L1 i-cache size */ + u32 iline_size; /* L1 i-cache line size */ + u32 log_iline_size; + u32 ilines_per_page; +}; + +extern struct ppc64_caches ppc64_caches; +#endif /* __powerpc64__ && ! __ASSEMBLY__ */ + +#endif /* __KERNEL__ */ +#endif /* _ASM_POWERPC_CACHE_H */ diff --git a/include/asm-ppc64/cacheflush.h b/include/asm-powerpc/cacheflush.h similarity index 56% rename from include/asm-ppc64/cacheflush.h rename to include/asm-powerpc/cacheflush.h index ffbc08be8e52..8a740c88d93d 100644 --- a/include/asm-ppc64/cacheflush.h +++ b/include/asm-powerpc/cacheflush.h @@ -1,13 +1,20 @@ -#ifndef _PPC64_CACHEFLUSH_H -#define _PPC64_CACHEFLUSH_H +/* + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License + * as published by the Free Software Foundation; either version + * 2 of the License, or (at your option) any later version. + */ +#ifndef _ASM_POWERPC_CACHEFLUSH_H +#define _ASM_POWERPC_CACHEFLUSH_H + +#ifdef __KERNEL__ #include #include /* - * No cache flushing is required when address mappings are - * changed, because the caches on PowerPCs are physically - * addressed. + * No cache flushing is required when address mappings are changed, + * because the caches on PowerPCs are physically addressed. */ #define flush_cache_all() do { } while (0) #define flush_cache_mm(mm) do { } while (0) @@ -22,27 +29,40 @@ extern void flush_dcache_page(struct page *page); #define flush_dcache_mmap_unlock(mapping) do { } while (0) extern void __flush_icache_range(unsigned long, unsigned long); -extern void flush_icache_user_range(struct vm_area_struct *vma, - struct page *page, unsigned long addr, - int len); - -extern void flush_dcache_range(unsigned long start, unsigned long stop); -extern void flush_dcache_phys_range(unsigned long start, unsigned long stop); -extern void flush_inval_dcache_range(unsigned long start, unsigned long stop); - -#define copy_to_user_page(vma, page, vaddr, dst, src, len) \ -do { memcpy(dst, src, len); \ - flush_icache_user_range(vma, page, vaddr, len); \ -} while (0) -#define copy_from_user_page(vma, page, vaddr, dst, src, len) \ - memcpy(dst, src, len) - -extern void __flush_dcache_icache(void *page_va); - static inline void flush_icache_range(unsigned long start, unsigned long stop) { if (!cpu_has_feature(CPU_FTR_COHERENT_ICACHE)) __flush_icache_range(start, stop); } -#endif /* _PPC64_CACHEFLUSH_H */ +extern void flush_icache_user_range(struct vm_area_struct *vma, + struct page *page, unsigned long addr, + int len); +extern void __flush_dcache_icache(void *page_va); +extern void flush_dcache_icache_page(struct page *page); +#if defined(CONFIG_PPC32) && !defined(CONFIG_BOOKE) +extern void __flush_dcache_icache_phys(unsigned long physaddr); +#endif /* CONFIG_PPC32 && !CONFIG_BOOKE */ + +extern void flush_dcache_range(unsigned long start, unsigned long stop); +#ifdef CONFIG_PPC32 +extern void clean_dcache_range(unsigned long start, unsigned long stop); +extern void invalidate_dcache_range(unsigned long start, unsigned long stop); +#endif /* CONFIG_PPC32 */ +#ifdef CONFIG_PPC64 +extern void flush_inval_dcache_range(unsigned long start, unsigned long stop); +extern void flush_dcache_phys_range(unsigned long start, unsigned long stop); +#endif + +#define copy_to_user_page(vma, page, vaddr, dst, src, len) \ + do { \ + memcpy(dst, src, len); \ + flush_icache_user_range(vma, page, vaddr, len); \ + } while (0) +#define copy_from_user_page(vma, page, vaddr, dst, src, len) \ + memcpy(dst, src, len) + + +#endif /* __KERNEL__ */ + +#endif /* _ASM_POWERPC_CACHEFLUSH_H */ diff --git a/include/asm-powerpc/reg.h b/include/asm-powerpc/reg.h index 489cf4c99c21..ef121f4f0bab 100644 --- a/include/asm-powerpc/reg.h +++ b/include/asm-powerpc/reg.h @@ -16,7 +16,11 @@ /* Pickup Book E specific registers. */ #if defined(CONFIG_BOOKE) || defined(CONFIG_40x) #include -#endif +#endif /* CONFIG_BOOKE || CONFIG_40x */ + +#ifdef CONFIG_8xx +#include +#endif /* CONFIG_8xx */ #define MSR_SF_LG 63 /* Enable 64 bit mode */ #define MSR_ISF_LG 61 /* Interrupt 64b mode valid on 630 */ diff --git a/include/asm-ppc/cache.h b/include/asm-powerpc/reg_8xx.h similarity index 50% rename from include/asm-ppc/cache.h rename to include/asm-powerpc/reg_8xx.h index 7a157d0f4b5f..e8ea346b21d3 100644 --- a/include/asm-ppc/cache.h +++ b/include/asm-powerpc/reg_8xx.h @@ -1,49 +1,9 @@ /* - * include/asm-ppc/cache.h + * Contains register definitions common to PowerPC 8xx CPUs. Notice */ -#ifdef __KERNEL__ -#ifndef __ARCH_PPC_CACHE_H -#define __ARCH_PPC_CACHE_H +#ifndef _ASM_POWERPC_REG_8xx_H +#define _ASM_POWERPC_REG_8xx_H -#include - -/* bytes per L1 cache line */ -#if defined(CONFIG_8xx) || defined(CONFIG_403GCX) -#define L1_CACHE_SHIFT 4 -#define MAX_COPY_PREFETCH 1 -#elif defined(CONFIG_PPC64BRIDGE) -#define L1_CACHE_SHIFT 7 -#define MAX_COPY_PREFETCH 1 -#else -#define L1_CACHE_SHIFT 5 -#define MAX_COPY_PREFETCH 4 -#endif - -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) - -#define SMP_CACHE_BYTES L1_CACHE_BYTES -#define L1_CACHE_SHIFT_MAX 7 /* largest L1 which this arch supports */ - -#define L1_CACHE_ALIGN(x) (((x)+(L1_CACHE_BYTES-1))&~(L1_CACHE_BYTES-1)) -#define L1_CACHE_PAGES 8 - -#ifndef __ASSEMBLY__ -extern void clean_dcache_range(unsigned long start, unsigned long stop); -extern void flush_dcache_range(unsigned long start, unsigned long stop); -extern void invalidate_dcache_range(unsigned long start, unsigned long stop); -extern void flush_dcache_all(void); -#endif /* __ASSEMBLY__ */ - -/* prep registers for L2 */ -#define CACHECRBA 0x80000823 /* Cache configuration register address */ -#define L2CACHE_MASK 0x03 /* Mask for 2 L2 Cache bits */ -#define L2CACHE_512KB 0x00 /* 512KB */ -#define L2CACHE_256KB 0x01 /* 256KB */ -#define L2CACHE_1MB 0x02 /* 1MB */ -#define L2CACHE_NONE 0x03 /* NONE */ -#define L2CACHE_PARITY 0x08 /* Mask for L2 Cache Parity Protected bit */ - -#ifdef CONFIG_8xx /* Cache control on the MPC8xx is provided through some additional * special purpose registers. */ @@ -78,7 +38,5 @@ extern void flush_dcache_all(void); #define DC_DFWT 0x40000000 /* Data cache is forced write through */ #define DC_LES 0x20000000 /* Caches are little endian mode */ -#endif /* CONFIG_8xx */ -#endif -#endif /* __KERNEL__ */ +#endif /* _ASM_POWERPC_REG_8xx_H */ diff --git a/include/asm-ppc/cacheflush.h b/include/asm-ppc/cacheflush.h deleted file mode 100644 index 6a243efb3317..000000000000 --- a/include/asm-ppc/cacheflush.h +++ /dev/null @@ -1,49 +0,0 @@ -/* - * include/asm-ppc/cacheflush.h - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - */ -#ifdef __KERNEL__ -#ifndef _PPC_CACHEFLUSH_H -#define _PPC_CACHEFLUSH_H - -#include - -/* - * No cache flushing is required when address mappings are - * changed, because the caches on PowerPCs are physically - * addressed. -- paulus - * Also, when SMP we use the coherency (M) bit of the - * BATs and PTEs. -- Cort - */ -#define flush_cache_all() do { } while (0) -#define flush_cache_mm(mm) do { } while (0) -#define flush_cache_range(vma, a, b) do { } while (0) -#define flush_cache_page(vma, p, pfn) do { } while (0) -#define flush_icache_page(vma, page) do { } while (0) -#define flush_cache_vmap(start, end) do { } while (0) -#define flush_cache_vunmap(start, end) do { } while (0) - -extern void flush_dcache_page(struct page *page); -#define flush_dcache_mmap_lock(mapping) do { } while (0) -#define flush_dcache_mmap_unlock(mapping) do { } while (0) - -extern void flush_icache_range(unsigned long, unsigned long); -extern void flush_icache_user_range(struct vm_area_struct *vma, - struct page *page, unsigned long addr, int len); - -#define copy_to_user_page(vma, page, vaddr, dst, src, len) \ -do { memcpy(dst, src, len); \ - flush_icache_user_range(vma, page, vaddr, len); \ -} while (0) -#define copy_from_user_page(vma, page, vaddr, dst, src, len) \ - memcpy(dst, src, len) - -extern void __flush_dcache_icache(void *page_va); -extern void __flush_dcache_icache_phys(unsigned long physaddr); -extern void flush_dcache_icache_page(struct page *page); -#endif /* _PPC_CACHEFLUSH_H */ -#endif /* __KERNEL__ */ diff --git a/include/asm-ppc64/cache.h b/include/asm-ppc64/cache.h deleted file mode 100644 index 92140a7efbd1..000000000000 --- a/include/asm-ppc64/cache.h +++ /dev/null @@ -1,36 +0,0 @@ -/* - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License - * as published by the Free Software Foundation; either version - * 2 of the License, or (at your option) any later version. - */ -#ifndef __ARCH_PPC64_CACHE_H -#define __ARCH_PPC64_CACHE_H - -#include - -/* bytes per L1 cache line */ -#define L1_CACHE_SHIFT 7 -#define L1_CACHE_BYTES (1 << L1_CACHE_SHIFT) - -#define SMP_CACHE_BYTES L1_CACHE_BYTES -#define L1_CACHE_SHIFT_MAX 7 /* largest L1 which this arch supports */ - -#ifndef __ASSEMBLY__ - -struct ppc64_caches { - u32 dsize; /* L1 d-cache size */ - u32 dline_size; /* L1 d-cache line size */ - u32 log_dline_size; - u32 dlines_per_page; - u32 isize; /* L1 i-cache size */ - u32 iline_size; /* L1 i-cache line size */ - u32 log_iline_size; - u32 ilines_per_page; -}; - -extern struct ppc64_caches ppc64_caches; - -#endif - -#endif From f6d3577da14e877b79517c883d1139ee6ad7da45 Mon Sep 17 00:00:00 2001 From: David Gibson Date: Thu, 10 Nov 2005 12:42:17 +1100 Subject: [PATCH 065/121] [PATCH] powerpc: Move scanlog.c to platforms/pseries scanlog.c is only compiled on pSeries. Thus, this patch moves it to platforms/pseries. Built and booted on pSeries LPAR (ARCH=powerpc and ARCH=ppc64). Built for iSeries (ARCH=powerpc). Signed-off-by: David Gibson Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/pseries/Makefile | 1 + arch/{ppc64/kernel => powerpc/platforms/pseries}/scanlog.c | 0 arch/ppc64/kernel/Makefile | 1 - 3 files changed, 1 insertion(+), 1 deletion(-) rename arch/{ppc64/kernel => powerpc/platforms/pseries}/scanlog.c (100%) diff --git a/arch/powerpc/platforms/pseries/Makefile b/arch/powerpc/platforms/pseries/Makefile index 27515476ad6c..e7ca5b1f591e 100644 --- a/arch/powerpc/platforms/pseries/Makefile +++ b/arch/powerpc/platforms/pseries/Makefile @@ -3,4 +3,5 @@ obj-y := pci.o lpar.o hvCall.o nvram.o reconfig.o \ obj-$(CONFIG_SMP) += smp.o obj-$(CONFIG_IBMVIO) += vio.o obj-$(CONFIG_XICS) += xics.o +obj-$(CONFIG_SCANLOG) += scanlog.o obj-$(CONFIG_EEH) += eeh.o eeh_event.o diff --git a/arch/ppc64/kernel/scanlog.c b/arch/powerpc/platforms/pseries/scanlog.c similarity index 100% rename from arch/ppc64/kernel/scanlog.c rename to arch/powerpc/platforms/pseries/scanlog.c diff --git a/arch/ppc64/kernel/Makefile b/arch/ppc64/kernel/Makefile index 1f71f23cc26d..048ba910f028 100644 --- a/arch/ppc64/kernel/Makefile +++ b/arch/ppc64/kernel/Makefile @@ -36,7 +36,6 @@ ifneq ($(CONFIG_PPC_MERGE),y) obj-$(CONFIG_MODULES) += ppc_ksyms.o endif obj-$(CONFIG_PPC_RTAS) += rtas_pci.o -obj-$(CONFIG_SCANLOG) += scanlog.o obj-$(CONFIG_LPARCFG) += lparcfg.o obj-$(CONFIG_HVC_CONSOLE) += hvconsole.o ifneq ($(CONFIG_PPC_MERGE),y) From 3ddfbcf19b15ccd25a0b4b2dc2e38000e08de739 Mon Sep 17 00:00:00 2001 From: David Gibson Date: Thu, 10 Nov 2005 12:56:55 +1100 Subject: [PATCH 066/121] [PATCH] powerpc: Consolidate asm compatibility macros This patch consolidates macros used to generate assembly for compatibility across different CPUs or configs. A new header, asm-powerpc/asm-compat.h contains the main compatibility macros. It uses some preprocessor magic to make the macros suitable both for use in .S files, and in inline asm in .c files. Headers (bitops.h, uaccess.h, atomic.h, bug.h) which had their own such compatibility macros are changed to use asm-compat.h. ppc_asm.h is now for use in .S files *only*, and a #error enforces that. As such, we're a lot more careless about namespace pollution here than in asm-compat.h. While we're at it, this patch adds a call to the PPC405_ERR77 macro in futex.h which should have had it already, but didn't. Built and booted on pSeries, Maple and iSeries (ARCH=powerpc). Built for 32-bit powermac (ARCH=powerpc) and Walnut (ARCH=ppc). Signed-off-by: David Gibson Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/fpu.S | 24 ++-- arch/powerpc/platforms/iseries/misc.S | 1 + arch/powerpc/xmon/setjmp.S | 176 +++++++++++++------------- arch/ppc/boot/openfirmware/Makefile | 3 +- include/asm-powerpc/asm-compat.h | 55 ++++++++ include/asm-powerpc/atomic.h | 10 +- include/asm-powerpc/bitops.h | 41 +++--- include/asm-powerpc/bug.h | 19 +-- include/asm-powerpc/cputable.h | 2 +- include/asm-powerpc/futex.h | 5 +- include/asm-powerpc/ppc_asm.h | 39 +----- include/asm-powerpc/system.h | 1 - include/asm-powerpc/uaccess.h | 40 +++--- include/asm-ppc64/mmu.h | 2 +- include/asm-ppc64/page.h | 2 +- 15 files changed, 210 insertions(+), 210 deletions(-) create mode 100644 include/asm-powerpc/asm-compat.h diff --git a/arch/powerpc/kernel/fpu.S b/arch/powerpc/kernel/fpu.S index 4d6001fa1cf2..b780b42c95fc 100644 --- a/arch/powerpc/kernel/fpu.S +++ b/arch/powerpc/kernel/fpu.S @@ -41,20 +41,20 @@ _GLOBAL(load_up_fpu) #ifndef CONFIG_SMP LOADBASE(r3, last_task_used_math) toreal(r3) - LDL r4,OFF(last_task_used_math)(r3) - CMPI 0,r4,0 + PPC_LL r4,OFF(last_task_used_math)(r3) + PPC_LCMPI 0,r4,0 beq 1f toreal(r4) addi r4,r4,THREAD /* want last_task_used_math->thread */ SAVE_32FPRS(0, r4) mffs fr0 stfd fr0,THREAD_FPSCR(r4) - LDL r5,PT_REGS(r4) + PPC_LL r5,PT_REGS(r4) toreal(r5) - LDL r4,_MSR-STACK_FRAME_OVERHEAD(r5) + PPC_LL r4,_MSR-STACK_FRAME_OVERHEAD(r5) li r10,MSR_FP|MSR_FE0|MSR_FE1 andc r4,r4,r10 /* disable FP for previous task */ - STL r4,_MSR-STACK_FRAME_OVERHEAD(r5) + PPC_STL r4,_MSR-STACK_FRAME_OVERHEAD(r5) 1: #endif /* CONFIG_SMP */ /* enable use of FP after return */ @@ -77,7 +77,7 @@ _GLOBAL(load_up_fpu) #ifndef CONFIG_SMP subi r4,r5,THREAD fromreal(r4) - STL r4,OFF(last_task_used_math)(r3) + PPC_STL r4,OFF(last_task_used_math)(r3) #endif /* CONFIG_SMP */ /* restore registers and return */ /* we haven't used ctr or xer or lr */ @@ -97,24 +97,24 @@ _GLOBAL(giveup_fpu) MTMSRD(r5) /* enable use of fpu now */ SYNC_601 isync - CMPI 0,r3,0 + PPC_LCMPI 0,r3,0 beqlr- /* if no previous owner, done */ addi r3,r3,THREAD /* want THREAD of task */ - LDL r5,PT_REGS(r3) - CMPI 0,r5,0 + PPC_LL r5,PT_REGS(r3) + PPC_LCMPI 0,r5,0 SAVE_32FPRS(0, r3) mffs fr0 stfd fr0,THREAD_FPSCR(r3) beq 1f - LDL r4,_MSR-STACK_FRAME_OVERHEAD(r5) + PPC_LL r4,_MSR-STACK_FRAME_OVERHEAD(r5) li r3,MSR_FP|MSR_FE0|MSR_FE1 andc r4,r4,r3 /* disable FP for previous task */ - STL r4,_MSR-STACK_FRAME_OVERHEAD(r5) + PPC_STL r4,_MSR-STACK_FRAME_OVERHEAD(r5) 1: #ifndef CONFIG_SMP li r5,0 LOADBASE(r4,last_task_used_math) - STL r5,OFF(last_task_used_math)(r4) + PPC_STL r5,OFF(last_task_used_math)(r4) #endif /* CONFIG_SMP */ blr diff --git a/arch/powerpc/platforms/iseries/misc.S b/arch/powerpc/platforms/iseries/misc.S index 09f14522e176..dfe7aa1ba098 100644 --- a/arch/powerpc/platforms/iseries/misc.S +++ b/arch/powerpc/platforms/iseries/misc.S @@ -15,6 +15,7 @@ #include #include +#include .text diff --git a/arch/powerpc/xmon/setjmp.S b/arch/powerpc/xmon/setjmp.S index f8e40dfd2bff..96a91f10e2ec 100644 --- a/arch/powerpc/xmon/setjmp.S +++ b/arch/powerpc/xmon/setjmp.S @@ -14,61 +14,61 @@ _GLOBAL(xmon_setjmp) mflr r0 - STL r0,0(r3) - STL r1,SZL(r3) - STL r2,2*SZL(r3) + PPC_STL r0,0(r3) + PPC_STL r1,SZL(r3) + PPC_STL r2,2*SZL(r3) mfcr r0 - STL r0,3*SZL(r3) - STL r13,4*SZL(r3) - STL r14,5*SZL(r3) - STL r15,6*SZL(r3) - STL r16,7*SZL(r3) - STL r17,8*SZL(r3) - STL r18,9*SZL(r3) - STL r19,10*SZL(r3) - STL r20,11*SZL(r3) - STL r21,12*SZL(r3) - STL r22,13*SZL(r3) - STL r23,14*SZL(r3) - STL r24,15*SZL(r3) - STL r25,16*SZL(r3) - STL r26,17*SZL(r3) - STL r27,18*SZL(r3) - STL r28,19*SZL(r3) - STL r29,20*SZL(r3) - STL r30,21*SZL(r3) - STL r31,22*SZL(r3) + PPC_STL r0,3*SZL(r3) + PPC_STL r13,4*SZL(r3) + PPC_STL r14,5*SZL(r3) + PPC_STL r15,6*SZL(r3) + PPC_STL r16,7*SZL(r3) + PPC_STL r17,8*SZL(r3) + PPC_STL r18,9*SZL(r3) + PPC_STL r19,10*SZL(r3) + PPC_STL r20,11*SZL(r3) + PPC_STL r21,12*SZL(r3) + PPC_STL r22,13*SZL(r3) + PPC_STL r23,14*SZL(r3) + PPC_STL r24,15*SZL(r3) + PPC_STL r25,16*SZL(r3) + PPC_STL r26,17*SZL(r3) + PPC_STL r27,18*SZL(r3) + PPC_STL r28,19*SZL(r3) + PPC_STL r29,20*SZL(r3) + PPC_STL r30,21*SZL(r3) + PPC_STL r31,22*SZL(r3) li r3,0 blr _GLOBAL(xmon_longjmp) - CMPI r4,0 + PPC_LCMPI r4,0 bne 1f li r4,1 -1: LDL r13,4*SZL(r3) - LDL r14,5*SZL(r3) - LDL r15,6*SZL(r3) - LDL r16,7*SZL(r3) - LDL r17,8*SZL(r3) - LDL r18,9*SZL(r3) - LDL r19,10*SZL(r3) - LDL r20,11*SZL(r3) - LDL r21,12*SZL(r3) - LDL r22,13*SZL(r3) - LDL r23,14*SZL(r3) - LDL r24,15*SZL(r3) - LDL r25,16*SZL(r3) - LDL r26,17*SZL(r3) - LDL r27,18*SZL(r3) - LDL r28,19*SZL(r3) - LDL r29,20*SZL(r3) - LDL r30,21*SZL(r3) - LDL r31,22*SZL(r3) - LDL r0,3*SZL(r3) +1: PPC_LL r13,4*SZL(r3) + PPC_LL r14,5*SZL(r3) + PPC_LL r15,6*SZL(r3) + PPC_LL r16,7*SZL(r3) + PPC_LL r17,8*SZL(r3) + PPC_LL r18,9*SZL(r3) + PPC_LL r19,10*SZL(r3) + PPC_LL r20,11*SZL(r3) + PPC_LL r21,12*SZL(r3) + PPC_LL r22,13*SZL(r3) + PPC_LL r23,14*SZL(r3) + PPC_LL r24,15*SZL(r3) + PPC_LL r25,16*SZL(r3) + PPC_LL r26,17*SZL(r3) + PPC_LL r27,18*SZL(r3) + PPC_LL r28,19*SZL(r3) + PPC_LL r29,20*SZL(r3) + PPC_LL r30,21*SZL(r3) + PPC_LL r31,22*SZL(r3) + PPC_LL r0,3*SZL(r3) mtcrf 0x38,r0 - LDL r0,0(r3) - LDL r1,SZL(r3) - LDL r2,2*SZL(r3) + PPC_LL r0,0(r3) + PPC_LL r1,SZL(r3) + PPC_LL r2,2*SZL(r3) mtlr r0 mr r3,r4 blr @@ -84,52 +84,52 @@ _GLOBAL(xmon_longjmp) * different ABIs, though). */ _GLOBAL(xmon_save_regs) - STL r0,0*SZL(r3) - STL r2,2*SZL(r3) - STL r3,3*SZL(r3) - STL r4,4*SZL(r3) - STL r5,5*SZL(r3) - STL r6,6*SZL(r3) - STL r7,7*SZL(r3) - STL r8,8*SZL(r3) - STL r9,9*SZL(r3) - STL r10,10*SZL(r3) - STL r11,11*SZL(r3) - STL r12,12*SZL(r3) - STL r13,13*SZL(r3) - STL r14,14*SZL(r3) - STL r15,15*SZL(r3) - STL r16,16*SZL(r3) - STL r17,17*SZL(r3) - STL r18,18*SZL(r3) - STL r19,19*SZL(r3) - STL r20,20*SZL(r3) - STL r21,21*SZL(r3) - STL r22,22*SZL(r3) - STL r23,23*SZL(r3) - STL r24,24*SZL(r3) - STL r25,25*SZL(r3) - STL r26,26*SZL(r3) - STL r27,27*SZL(r3) - STL r28,28*SZL(r3) - STL r29,29*SZL(r3) - STL r30,30*SZL(r3) - STL r31,31*SZL(r3) + PPC_STL r0,0*SZL(r3) + PPC_STL r2,2*SZL(r3) + PPC_STL r3,3*SZL(r3) + PPC_STL r4,4*SZL(r3) + PPC_STL r5,5*SZL(r3) + PPC_STL r6,6*SZL(r3) + PPC_STL r7,7*SZL(r3) + PPC_STL r8,8*SZL(r3) + PPC_STL r9,9*SZL(r3) + PPC_STL r10,10*SZL(r3) + PPC_STL r11,11*SZL(r3) + PPC_STL r12,12*SZL(r3) + PPC_STL r13,13*SZL(r3) + PPC_STL r14,14*SZL(r3) + PPC_STL r15,15*SZL(r3) + PPC_STL r16,16*SZL(r3) + PPC_STL r17,17*SZL(r3) + PPC_STL r18,18*SZL(r3) + PPC_STL r19,19*SZL(r3) + PPC_STL r20,20*SZL(r3) + PPC_STL r21,21*SZL(r3) + PPC_STL r22,22*SZL(r3) + PPC_STL r23,23*SZL(r3) + PPC_STL r24,24*SZL(r3) + PPC_STL r25,25*SZL(r3) + PPC_STL r26,26*SZL(r3) + PPC_STL r27,27*SZL(r3) + PPC_STL r28,28*SZL(r3) + PPC_STL r29,29*SZL(r3) + PPC_STL r30,30*SZL(r3) + PPC_STL r31,31*SZL(r3) /* go up one stack frame for SP */ - LDL r4,0(r1) - STL r4,1*SZL(r3) + PPC_LL r4,0(r1) + PPC_STL r4,1*SZL(r3) /* get caller's LR */ - LDL r0,LRSAVE(r4) - STL r0,_NIP-STACK_FRAME_OVERHEAD(r3) - STL r0,_LINK-STACK_FRAME_OVERHEAD(r3) + PPC_LL r0,LRSAVE(r4) + PPC_STL r0,_NIP-STACK_FRAME_OVERHEAD(r3) + PPC_STL r0,_LINK-STACK_FRAME_OVERHEAD(r3) mfmsr r0 - STL r0,_MSR-STACK_FRAME_OVERHEAD(r3) + PPC_STL r0,_MSR-STACK_FRAME_OVERHEAD(r3) mfctr r0 - STL r0,_CTR-STACK_FRAME_OVERHEAD(r3) + PPC_STL r0,_CTR-STACK_FRAME_OVERHEAD(r3) mfxer r0 - STL r0,_XER-STACK_FRAME_OVERHEAD(r3) + PPC_STL r0,_XER-STACK_FRAME_OVERHEAD(r3) mfcr r0 - STL r0,_CCR-STACK_FRAME_OVERHEAD(r3) + PPC_STL r0,_CCR-STACK_FRAME_OVERHEAD(r3) li r0,0 - STL r0,_TRAP-STACK_FRAME_OVERHEAD(r3) + PPC_STL r0,_TRAP-STACK_FRAME_OVERHEAD(r3) blr diff --git a/arch/ppc/boot/openfirmware/Makefile b/arch/ppc/boot/openfirmware/Makefile index 03415238fabf..83a6433459ce 100644 --- a/arch/ppc/boot/openfirmware/Makefile +++ b/arch/ppc/boot/openfirmware/Makefile @@ -80,8 +80,7 @@ $(obj)/note: $(utils)/mknote FORCE $(call if_changed,mknote) -$(obj)/coffcrt0.o: EXTRA_AFLAGS := -traditional -DXCOFF -$(obj)/crt0.o: EXTRA_AFLAGS := -traditional +$(obj)/coffcrt0.o: EXTRA_AFLAGS := -DXCOFF targets += coffcrt0.o crt0.o $(obj)/coffcrt0.o $(obj)/crt0.o: $(common)/crt0.S FORCE $(call if_changed_dep,as_o_S) diff --git a/include/asm-powerpc/asm-compat.h b/include/asm-powerpc/asm-compat.h new file mode 100644 index 000000000000..8b133efc9f79 --- /dev/null +++ b/include/asm-powerpc/asm-compat.h @@ -0,0 +1,55 @@ +#ifndef _ASM_POWERPC_ASM_COMPAT_H +#define _ASM_POWERPC_ASM_COMPAT_H + +#include +#include + +#ifdef __ASSEMBLY__ +# define stringify_in_c(...) __VA_ARGS__ +# define ASM_CONST(x) x +#else +/* This version of stringify will deal with commas... */ +# define __stringify_in_c(...) #__VA_ARGS__ +# define stringify_in_c(...) __stringify_in_c(__VA_ARGS__) " " +# define __ASM_CONST(x) x##UL +# define ASM_CONST(x) __ASM_CONST(x) +#endif + +#ifdef __powerpc64__ + +/* operations for longs and pointers */ +#define PPC_LL stringify_in_c(ld) +#define PPC_STL stringify_in_c(std) +#define PPC_LCMPI stringify_in_c(cmpdi) +#define PPC_LONG stringify_in_c(.llong) +#define PPC_TLNEI stringify_in_c(tdnei) +#define PPC_LLARX stringify_in_c(ldarx) +#define PPC_STLCX stringify_in_c(stdcx.) +#define PPC_CNTLZL stringify_in_c(cntlzd) + +#else /* 32-bit */ + +/* operations for longs and pointers */ +#define PPC_LL stringify_in_c(lwz) +#define PPC_STL stringify_in_c(stw) +#define PPC_LCMPI stringify_in_c(cmpwi) +#define PPC_LONG stringify_in_c(.long) +#define PPC_TLNEI stringify_in_c(twnei) +#define PPC_LLARX stringify_in_c(lwarx) +#define PPC_STLCX stringify_in_c(stwcx.) +#define PPC_CNTLZL stringify_in_c(cntlzw) + +#endif + +#ifdef CONFIG_IBM405_ERR77 +/* Erratum #77 on the 405 means we need a sync or dcbt before every + * stwcx. The old ATOMIC_SYNC_FIX covered some but not all of this. + */ +#define PPC405_ERR77(ra,rb) stringify_in_c(dcbt ra, rb;) +#define PPC405_ERR77_SYNC stringify_in_c(sync;) +#else +#define PPC405_ERR77(ra,rb) +#define PPC405_ERR77_SYNC +#endif + +#endif /* _ASM_POWERPC_ASM_COMPAT_H */ diff --git a/include/asm-powerpc/atomic.h b/include/asm-powerpc/atomic.h index ed4b345ed75d..c5b12fd2b46b 100644 --- a/include/asm-powerpc/atomic.h +++ b/include/asm-powerpc/atomic.h @@ -9,21 +9,13 @@ typedef struct { volatile int counter; } atomic_t; #ifdef __KERNEL__ #include +#include #define ATOMIC_INIT(i) { (i) } #define atomic_read(v) ((v)->counter) #define atomic_set(v,i) (((v)->counter) = (i)) -/* Erratum #77 on the 405 means we need a sync or dcbt before every stwcx. - * The old ATOMIC_SYNC_FIX covered some but not all of this. - */ -#ifdef CONFIG_IBM405_ERR77 -#define PPC405_ERR77(ra,rb) "dcbt " #ra "," #rb ";" -#else -#define PPC405_ERR77(ra,rb) -#endif - static __inline__ void atomic_add(int a, atomic_t *v) { int t; diff --git a/include/asm-powerpc/bitops.h b/include/asm-powerpc/bitops.h index dc25c53704d5..5727229b0444 100644 --- a/include/asm-powerpc/bitops.h +++ b/include/asm-powerpc/bitops.h @@ -40,6 +40,7 @@ #include #include +#include #include /* @@ -52,16 +53,6 @@ #define BITOP_WORD(nr) ((nr) / BITS_PER_LONG) #define BITOP_LE_SWIZZLE ((BITS_PER_LONG-1) & ~0x7) -#ifdef CONFIG_PPC64 -#define LARXL "ldarx" -#define STCXL "stdcx." -#define CNTLZL "cntlzd" -#else -#define LARXL "lwarx" -#define STCXL "stwcx." -#define CNTLZL "cntlzw" -#endif - static __inline__ void set_bit(int nr, volatile unsigned long *addr) { unsigned long old; @@ -69,10 +60,10 @@ static __inline__ void set_bit(int nr, volatile unsigned long *addr) unsigned long *p = ((unsigned long *)addr) + BITOP_WORD(nr); __asm__ __volatile__( -"1:" LARXL " %0,0,%3 # set_bit\n" +"1:" PPC_LLARX "%0,0,%3 # set_bit\n" "or %0,%0,%2\n" PPC405_ERR77(0,%3) - STCXL " %0,0,%3\n" + PPC_STLCX "%0,0,%3\n" "bne- 1b" : "=&r"(old), "=m"(*p) : "r"(mask), "r"(p), "m"(*p) @@ -86,10 +77,10 @@ static __inline__ void clear_bit(int nr, volatile unsigned long *addr) unsigned long *p = ((unsigned long *)addr) + BITOP_WORD(nr); __asm__ __volatile__( -"1:" LARXL " %0,0,%3 # set_bit\n" +"1:" PPC_LLARX "%0,0,%3 # clear_bit\n" "andc %0,%0,%2\n" PPC405_ERR77(0,%3) - STCXL " %0,0,%3\n" + PPC_STLCX "%0,0,%3\n" "bne- 1b" : "=&r"(old), "=m"(*p) : "r"(mask), "r"(p), "m"(*p) @@ -103,10 +94,10 @@ static __inline__ void change_bit(int nr, volatile unsigned long *addr) unsigned long *p = ((unsigned long *)addr) + BITOP_WORD(nr); __asm__ __volatile__( -"1:" LARXL " %0,0,%3 # set_bit\n" +"1:" PPC_LLARX "%0,0,%3 # change_bit\n" "xor %0,%0,%2\n" PPC405_ERR77(0,%3) - STCXL " %0,0,%3\n" + PPC_STLCX "%0,0,%3\n" "bne- 1b" : "=&r"(old), "=m"(*p) : "r"(mask), "r"(p), "m"(*p) @@ -122,10 +113,10 @@ static __inline__ int test_and_set_bit(unsigned long nr, __asm__ __volatile__( EIEIO_ON_SMP -"1:" LARXL " %0,0,%3 # test_and_set_bit\n" +"1:" PPC_LLARX "%0,0,%3 # test_and_set_bit\n" "or %1,%0,%2 \n" PPC405_ERR77(0,%3) - STCXL " %1,0,%3 \n" + PPC_STLCX "%1,0,%3 \n" "bne- 1b" ISYNC_ON_SMP : "=&r" (old), "=&r" (t) @@ -144,10 +135,10 @@ static __inline__ int test_and_clear_bit(unsigned long nr, __asm__ __volatile__( EIEIO_ON_SMP -"1:" LARXL " %0,0,%3 # test_and_clear_bit\n" +"1:" PPC_LLARX "%0,0,%3 # test_and_clear_bit\n" "andc %1,%0,%2 \n" PPC405_ERR77(0,%3) - STCXL " %1,0,%3 \n" + PPC_STLCX "%1,0,%3 \n" "bne- 1b" ISYNC_ON_SMP : "=&r" (old), "=&r" (t) @@ -166,10 +157,10 @@ static __inline__ int test_and_change_bit(unsigned long nr, __asm__ __volatile__( EIEIO_ON_SMP -"1:" LARXL " %0,0,%3 # test_and_change_bit\n" +"1:" PPC_LLARX "%0,0,%3 # test_and_change_bit\n" "xor %1,%0,%2 \n" PPC405_ERR77(0,%3) - STCXL " %1,0,%3 \n" + PPC_STLCX "%1,0,%3 \n" "bne- 1b" ISYNC_ON_SMP : "=&r" (old), "=&r" (t) @@ -184,9 +175,9 @@ static __inline__ void set_bits(unsigned long mask, unsigned long *addr) unsigned long old; __asm__ __volatile__( -"1:" LARXL " %0,0,%3 # set_bit\n" +"1:" PPC_LLARX "%0,0,%3 # set_bits\n" "or %0,%0,%2\n" - STCXL " %0,0,%3\n" + PPC_STLCX "%0,0,%3\n" "bne- 1b" : "=&r" (old), "=m" (*addr) : "r" (mask), "r" (addr), "m" (*addr) @@ -268,7 +259,7 @@ static __inline__ int __ilog2(unsigned long x) { int lz; - asm (CNTLZL " %0,%1" : "=r" (lz) : "r" (x)); + asm (PPC_CNTLZL "%0,%1" : "=r" (lz) : "r" (x)); return BITS_PER_LONG - 1 - lz; } diff --git a/include/asm-powerpc/bug.h b/include/asm-powerpc/bug.h index d625ee55f957..b001ecb3cd99 100644 --- a/include/asm-powerpc/bug.h +++ b/include/asm-powerpc/bug.h @@ -1,6 +1,7 @@ #ifndef _ASM_POWERPC_BUG_H #define _ASM_POWERPC_BUG_H +#include /* * Define an illegal instr to trap on the bug. * We don't use 0 because that marks the end of a function @@ -11,14 +12,6 @@ #ifndef __ASSEMBLY__ -#ifdef __powerpc64__ -#define BUG_TABLE_ENTRY ".llong" -#define BUG_TRAP_OP "tdnei" -#else -#define BUG_TABLE_ENTRY ".long" -#define BUG_TRAP_OP "twnei" -#endif /* __powerpc64__ */ - struct bug_entry { unsigned long bug_addr; long line; @@ -40,16 +33,16 @@ struct bug_entry *find_bug(unsigned long bugaddr); __asm__ __volatile__( \ "1: twi 31,0,0\n" \ ".section __bug_table,\"a\"\n" \ - "\t"BUG_TABLE_ENTRY" 1b,%0,%1,%2\n" \ + "\t"PPC_LONG" 1b,%0,%1,%2\n" \ ".previous" \ : : "i" (__LINE__), "i" (__FILE__), "i" (__FUNCTION__)); \ } while (0) #define BUG_ON(x) do { \ __asm__ __volatile__( \ - "1: "BUG_TRAP_OP" %0,0\n" \ + "1: "PPC_TLNEI" %0,0\n" \ ".section __bug_table,\"a\"\n" \ - "\t"BUG_TABLE_ENTRY" 1b,%1,%2,%3\n" \ + "\t"PPC_LONG" 1b,%1,%2,%3\n" \ ".previous" \ : : "r" ((long)(x)), "i" (__LINE__), \ "i" (__FILE__), "i" (__FUNCTION__)); \ @@ -57,9 +50,9 @@ struct bug_entry *find_bug(unsigned long bugaddr); #define WARN_ON(x) do { \ __asm__ __volatile__( \ - "1: "BUG_TRAP_OP" %0,0\n" \ + "1: "PPC_TLNEI" %0,0\n" \ ".section __bug_table,\"a\"\n" \ - "\t"BUG_TABLE_ENTRY" 1b,%1,%2,%3\n" \ + "\t"PPC_LONG" 1b,%1,%2,%3\n" \ ".previous" \ : : "r" ((long)(x)), \ "i" (__LINE__ + BUG_WARNING_TRAP), \ diff --git a/include/asm-powerpc/cputable.h b/include/asm-powerpc/cputable.h index 79a0556a0ab8..f89fd883e892 100644 --- a/include/asm-powerpc/cputable.h +++ b/include/asm-powerpc/cputable.h @@ -2,7 +2,7 @@ #define __ASM_POWERPC_CPUTABLE_H #include -#include /* for ASM_CONST */ +#include #define PPC_FEATURE_32 0x80000000 #define PPC_FEATURE_64 0x40000000 diff --git a/include/asm-powerpc/futex.h b/include/asm-powerpc/futex.h index 37c94e52ab6d..f0319d50b129 100644 --- a/include/asm-powerpc/futex.h +++ b/include/asm-powerpc/futex.h @@ -7,13 +7,14 @@ #include #include #include -#include +#include #define __futex_atomic_op(insn, ret, oldval, uaddr, oparg) \ __asm__ __volatile ( \ SYNC_ON_SMP \ "1: lwarx %0,0,%2\n" \ insn \ + PPC405_ERR77(0, %2) \ "2: stwcx. %1,0,%2\n" \ "bne- 1b\n" \ "li %1,0\n" \ @@ -23,7 +24,7 @@ ".previous\n" \ ".section __ex_table,\"a\"\n" \ ".align 3\n" \ - DATAL " 1b,4b,2b,4b\n" \ + PPC_LONG "1b,4b,2b,4b\n" \ ".previous" \ : "=&r" (oldval), "=&r" (ret) \ : "b" (uaddr), "i" (-EFAULT), "1" (oparg) \ diff --git a/include/asm-powerpc/ppc_asm.h b/include/asm-powerpc/ppc_asm.h index c534ca41224b..c27baa0563fe 100644 --- a/include/asm-powerpc/ppc_asm.h +++ b/include/asm-powerpc/ppc_asm.h @@ -6,8 +6,13 @@ #include #include +#include -#ifdef __ASSEMBLY__ +#ifndef __ASSEMBLY__ +#error __FILE__ should only be used in assembler files +#else + +#define SZL (BITS_PER_LONG/8) /* * Macros for storing registers into and loading registers from @@ -184,12 +189,6 @@ n: oris reg,reg,(label)@h; \ ori reg,reg,(label)@l; -/* operations for longs and pointers */ -#define LDL ld -#define STL std -#define CMPI cmpdi -#define SZL 8 - /* offsets for stack frame layout */ #define LRSAVE 16 @@ -203,12 +202,6 @@ n: #define OFF(name) name@l -/* operations for longs and pointers */ -#define LDL lwz -#define STL stw -#define CMPI cmpwi -#define SZL 4 - /* offsets for stack frame layout */ #define LRSAVE 4 @@ -266,15 +259,6 @@ END_FTR_SECTION_IFCLR(CPU_FTR_601) #endif -#ifdef CONFIG_IBM405_ERR77 -#define PPC405_ERR77(ra,rb) dcbt ra, rb; -#define PPC405_ERR77_SYNC sync; -#else -#define PPC405_ERR77(ra,rb) -#define PPC405_ERR77_SYNC -#endif - - #ifdef CONFIG_IBM440EP_ERR42 #define PPC440EP_ERR42 isync #else @@ -502,17 +486,6 @@ END_FTR_SECTION_IFCLR(CPU_FTR_601) #define N_SLINE 68 #define N_SO 100 -#define ASM_CONST(x) x -#else - #define __ASM_CONST(x) x##UL - #define ASM_CONST(x) __ASM_CONST(x) - -#ifdef CONFIG_PPC64 -#define DATAL ".llong" -#else -#define DATAL ".long" -#endif - #endif /* __ASSEMBLY__ */ #endif /* _ASM_POWERPC_PPC_ASM_H */ diff --git a/include/asm-powerpc/system.h b/include/asm-powerpc/system.h index 3536a5cd7a2d..f0cce5a30235 100644 --- a/include/asm-powerpc/system.h +++ b/include/asm-powerpc/system.h @@ -8,7 +8,6 @@ #include #include -#include #include /* diff --git a/include/asm-powerpc/uaccess.h b/include/asm-powerpc/uaccess.h index 33af730f0d19..3872e924cdd6 100644 --- a/include/asm-powerpc/uaccess.h +++ b/include/asm-powerpc/uaccess.h @@ -120,14 +120,6 @@ struct exception_table_entry { extern long __put_user_bad(void); -#ifdef __powerpc64__ -#define __EX_TABLE_ALIGN "3" -#define __EX_TABLE_TYPE "llong" -#else -#define __EX_TABLE_ALIGN "2" -#define __EX_TABLE_TYPE "long" -#endif - /* * We don't tell gcc that we are accessing memory, but this is OK * because we do not write to any memory gcc knows about, so there @@ -142,11 +134,12 @@ extern long __put_user_bad(void); " b 2b\n" \ ".previous\n" \ ".section __ex_table,\"a\"\n" \ - " .align " __EX_TABLE_ALIGN "\n" \ - " ."__EX_TABLE_TYPE" 1b,3b\n" \ + " .balign %5\n" \ + PPC_LONG "1b,3b\n" \ ".previous" \ : "=r" (err) \ - : "r" (x), "b" (addr), "i" (-EFAULT), "0" (err)) + : "r" (x), "b" (addr), "i" (-EFAULT), "0" (err),\ + "i"(sizeof(unsigned long))) #ifdef __powerpc64__ #define __put_user_asm2(x, ptr, retval) \ @@ -162,12 +155,13 @@ extern long __put_user_bad(void); " b 3b\n" \ ".previous\n" \ ".section __ex_table,\"a\"\n" \ - " .align " __EX_TABLE_ALIGN "\n" \ - " ." __EX_TABLE_TYPE " 1b,4b\n" \ - " ." __EX_TABLE_TYPE " 2b,4b\n" \ + " .balign %5\n" \ + PPC_LONG "1b,4b\n" \ + PPC_LONG "2b,4b\n" \ ".previous" \ : "=r" (err) \ - : "r" (x), "b" (addr), "i" (-EFAULT), "0" (err)) + : "r" (x), "b" (addr), "i" (-EFAULT), "0" (err),\ + "i"(sizeof(unsigned long))) #endif /* __powerpc64__ */ #define __put_user_size(x, ptr, size, retval) \ @@ -213,11 +207,12 @@ extern long __get_user_bad(void); " b 2b\n" \ ".previous\n" \ ".section __ex_table,\"a\"\n" \ - " .align "__EX_TABLE_ALIGN "\n" \ - " ." __EX_TABLE_TYPE " 1b,3b\n" \ + " .balign %5\n" \ + PPC_LONG "1b,3b\n" \ ".previous" \ : "=r" (err), "=r" (x) \ - : "b" (addr), "i" (-EFAULT), "0" (err)) + : "b" (addr), "i" (-EFAULT), "0" (err), \ + "i"(sizeof(unsigned long))) #ifdef __powerpc64__ #define __get_user_asm2(x, addr, err) \ @@ -235,12 +230,13 @@ extern long __get_user_bad(void); " b 3b\n" \ ".previous\n" \ ".section __ex_table,\"a\"\n" \ - " .align " __EX_TABLE_ALIGN "\n" \ - " ." __EX_TABLE_TYPE " 1b,4b\n" \ - " ." __EX_TABLE_TYPE " 2b,4b\n" \ + " .balign %5\n" \ + PPC_LONG "1b,4b\n" \ + PPC_LONG "2b,4b\n" \ ".previous" \ : "=r" (err), "=&r" (x) \ - : "b" (addr), "i" (-EFAULT), "0" (err)) + : "b" (addr), "i" (-EFAULT), "0" (err), \ + "i"(sizeof(unsigned long))) #endif /* __powerpc64__ */ #define __get_user_size(x, ptr, size, retval) \ diff --git a/include/asm-ppc64/mmu.h b/include/asm-ppc64/mmu.h index 4c18a5cb69f5..d50997bace1c 100644 --- a/include/asm-ppc64/mmu.h +++ b/include/asm-ppc64/mmu.h @@ -14,7 +14,7 @@ #define _PPC64_MMU_H_ #include -#include /* for ASM_CONST */ +#include #include /* diff --git a/include/asm-ppc64/page.h b/include/asm-ppc64/page.h index 82ce187e5be8..e32f1187aa29 100644 --- a/include/asm-ppc64/page.h +++ b/include/asm-ppc64/page.h @@ -11,7 +11,7 @@ */ #include -#include /* for ASM_CONST */ +#include /* * We support either 4k or 64k software page size. When using 64k pages From 799d6046d3fb557006e6d7c9767fdb96479b0e0a Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 13:37:51 +1100 Subject: [PATCH 067/121] [PATCH] powerpc: merge code values for identifying platforms This patch merges platform codes. systemcfg->platform is no longer used, systemcfg use in general is deprecated as much as possible (and renamed _systemcfg before it gets completely moved elsewhere in a future patch), _machine is now used on ppc64 along as ppc32. Platform codes aren't gone yet but we are getting a step closer. A bunch of asm code in head[_64].S is also turned into C code. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/asm-offsets.c | 1 - arch/powerpc/kernel/head_64.S | 89 ++----------------- arch/powerpc/kernel/paca.c | 5 +- arch/powerpc/kernel/ppc_ksyms.c | 3 - arch/powerpc/kernel/prom.c | 16 +--- arch/powerpc/kernel/prom_init.c | 8 +- arch/powerpc/kernel/rtas-proc.c | 2 +- arch/powerpc/kernel/rtas.c | 5 +- arch/powerpc/kernel/setup-common.c | 12 ++- arch/powerpc/kernel/setup_32.c | 2 - arch/powerpc/kernel/setup_64.c | 65 +++++++++----- arch/powerpc/kernel/smp.c | 5 +- arch/powerpc/kernel/sys_ppc32.c | 1 - arch/powerpc/kernel/time.c | 29 +++--- arch/powerpc/kernel/traps.c | 2 +- arch/powerpc/mm/hash_utils_64.c | 29 ++++-- arch/powerpc/mm/stab.c | 21 ++++- arch/powerpc/oprofile/op_model_power4.c | 3 +- arch/powerpc/platforms/iseries/setup.c | 23 +++-- arch/powerpc/platforms/pseries/eeh.c | 3 +- arch/powerpc/platforms/pseries/iommu.c | 3 +- arch/powerpc/platforms/pseries/pci.c | 2 +- arch/powerpc/platforms/pseries/reconfig.c | 2 +- arch/powerpc/platforms/pseries/rtasd.c | 8 +- arch/powerpc/platforms/pseries/setup.c | 8 +- arch/powerpc/platforms/pseries/smp.c | 5 +- arch/powerpc/platforms/pseries/xics.c | 6 +- arch/ppc/kernel/ppc_ksyms.c | 3 - arch/ppc/kernel/setup.c | 1 + arch/ppc64/kernel/asm-offsets.c | 1 - arch/ppc64/kernel/head.S | 82 +---------------- arch/ppc64/kernel/idle.c | 1 - arch/ppc64/kernel/lparcfg.c | 3 +- arch/ppc64/kernel/nvram.c | 5 +- arch/ppc64/kernel/pci.c | 5 +- arch/ppc64/kernel/proc_ppc64.c | 4 +- arch/ppc64/kernel/prom.c | 9 +- arch/ppc64/kernel/prom_init.c | 3 +- arch/ppc64/kernel/vdso.c | 5 +- include/asm-powerpc/firmware.h | 6 ++ include/asm-powerpc/processor.h | 69 +++++++------- include/asm-powerpc/reg.h | 1 + .../{asm-ppc64 => asm-powerpc}/systemcfg.h | 6 +- include/asm-ppc64/mmu.h | 3 + 44 files changed, 234 insertions(+), 331 deletions(-) rename include/{asm-ppc64 => asm-powerpc}/systemcfg.h (96%) diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c index b75757251994..8793102711a8 100644 --- a/arch/powerpc/kernel/asm-offsets.c +++ b/arch/powerpc/kernel/asm-offsets.c @@ -106,7 +106,6 @@ int main(void) DEFINE(ICACHEL1LINESIZE, offsetof(struct ppc64_caches, iline_size)); DEFINE(ICACHEL1LOGLINESIZE, offsetof(struct ppc64_caches, log_iline_size)); DEFINE(ICACHEL1LINESPERPAGE, offsetof(struct ppc64_caches, ilines_per_page)); - DEFINE(PLATFORM, offsetof(struct systemcfg, platform)); DEFINE(PLATFORM_LPAR, PLATFORM_LPAR); /* paca */ diff --git a/arch/powerpc/kernel/head_64.S b/arch/powerpc/kernel/head_64.S index 16ab40daa738..8a8bf79ef044 100644 --- a/arch/powerpc/kernel/head_64.S +++ b/arch/powerpc/kernel/head_64.S @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include @@ -1697,25 +1696,14 @@ _GLOBAL(pmac_secondary_start) * SPRG3 = paca virtual address */ _GLOBAL(__secondary_start) + /* Set thread priority to MEDIUM */ + HMT_MEDIUM - HMT_MEDIUM /* Set thread priority to MEDIUM */ - + /* Load TOC */ ld r2,PACATOC(r13) - li r6,0 - stb r6,PACAPROCENABLED(r13) -#ifndef CONFIG_PPC_ISERIES - /* Initialize the page table pointer register. */ - LOADADDR(r6,_SDR1) - ld r6,0(r6) /* get the value of _SDR1 */ - mtspr SPRN_SDR1,r6 /* set the htab location */ -#endif - /* Initialize the first segment table (or SLB) entry */ - ld r3,PACASTABVIRT(r13) /* get addr of segment table */ -BEGIN_FTR_SECTION - bl .stab_initialize -END_FTR_SECTION_IFCLR(CPU_FTR_SLB) - bl .slb_initialize + /* Do early setup for that CPU (stab, slb, hash table pointer) */ + bl .early_setup_secondary /* Initialize the kernel stack. Just a repeat for iSeries. */ LOADADDR(r3,current_set) @@ -1724,37 +1712,7 @@ END_FTR_SECTION_IFCLR(CPU_FTR_SLB) addi r1,r1,THREAD_SIZE-STACK_FRAME_OVERHEAD std r1,PACAKSAVE(r13) - ld r3,PACASTABREAL(r13) /* get raddr of segment table */ - ori r4,r3,1 /* turn on valid bit */ - -#ifdef CONFIG_PPC_ISERIES - li r0,-1 /* hypervisor call */ - li r3,1 - sldi r3,r3,63 /* 0x8000000000000000 */ - ori r3,r3,4 /* 0x8000000000000004 */ - sc /* HvCall_setASR */ -#else - /* set the ASR */ - ld r3,systemcfg@got(r2) /* r3 = ptr to systemcfg */ - ld r3,0(r3) - lwz r3,PLATFORM(r3) /* r3 = platform flags */ - andi. r3,r3,PLATFORM_LPAR /* Test if bit 0 is set (LPAR bit) */ - beq 98f /* branch if result is 0 */ - mfspr r3,SPRN_PVR - srwi r3,r3,16 - cmpwi r3,0x37 /* SStar */ - beq 97f - cmpwi r3,0x36 /* IStar */ - beq 97f - cmpwi r3,0x34 /* Pulsar */ - bne 98f -97: li r3,H_SET_ASR /* hcall = H_SET_ASR */ - HVSC /* Invoking hcall */ - b 99f -98: /* !(rpa hypervisor) || !(star) */ - mtasr r4 /* set the stab location */ -99: -#endif + /* Clear backchain so we get nice backtraces */ li r7,0 mtlr r7 @@ -1777,6 +1735,7 @@ _GLOBAL(start_secondary_prolog) li r3,0 std r3,0(r1) /* Zero the stack frame pointer */ bl .start_secondary + b . #endif /* @@ -1896,40 +1855,6 @@ _STATIC(start_here_multiplatform) mr r3,r31 bl .early_setup - /* set the ASR */ - ld r3,PACASTABREAL(r13) - ori r4,r3,1 /* turn on valid bit */ - ld r3,systemcfg@got(r2) /* r3 = ptr to systemcfg */ - ld r3,0(r3) - lwz r3,PLATFORM(r3) /* r3 = platform flags */ - andi. r3,r3,PLATFORM_LPAR /* Test if bit 0 is set (LPAR bit) */ - beq 98f /* branch if result is 0 */ - mfspr r3,SPRN_PVR - srwi r3,r3,16 - cmpwi r3,0x37 /* SStar */ - beq 97f - cmpwi r3,0x36 /* IStar */ - beq 97f - cmpwi r3,0x34 /* Pulsar */ - bne 98f -97: li r3,H_SET_ASR /* hcall = H_SET_ASR */ - HVSC /* Invoking hcall */ - b 99f -98: /* !(rpa hypervisor) || !(star) */ - mtasr r4 /* set the stab location */ -99: - /* Set SDR1 (hash table pointer) */ - ld r3,systemcfg@got(r2) /* r3 = ptr to systemcfg */ - ld r3,0(r3) - lwz r3,PLATFORM(r3) /* r3 = platform flags */ - /* Test if bit 0 is set (LPAR bit) */ - andi. r3,r3,PLATFORM_LPAR - bne 98f /* branch if result is !0 */ - LOADADDR(r6,_SDR1) /* Only if NOT LPAR */ - add r6,r6,r26 - ld r6,0(r6) /* get the value of _SDR1 */ - mtspr SPRN_SDR1,r6 /* set the htab location */ -98: LOADADDR(r3,.start_here_common) SET_REG_TO_CONST(r4, MSR_KERNEL) mtspr SPRN_SRR0,r3 diff --git a/arch/powerpc/kernel/paca.c b/arch/powerpc/kernel/paca.c index 179948eb0580..3cf2517c5f91 100644 --- a/arch/powerpc/kernel/paca.c +++ b/arch/powerpc/kernel/paca.c @@ -15,7 +15,7 @@ #include #include #include - +#include #include #include #include @@ -24,8 +24,7 @@ static union { struct systemcfg data; u8 page[PAGE_SIZE]; } systemcfg_store __attribute__((__section__(".data.page.aligned"))); -struct systemcfg *systemcfg = &systemcfg_store.data; -EXPORT_SYMBOL(systemcfg); +struct systemcfg *_systemcfg = &systemcfg_store.data; /* This symbol is provided by the linker - let it fill in the paca diff --git a/arch/powerpc/kernel/ppc_ksyms.c b/arch/powerpc/kernel/ppc_ksyms.c index 5d9fd0369aad..5dcf4ba05ee8 100644 --- a/arch/powerpc/kernel/ppc_ksyms.c +++ b/arch/powerpc/kernel/ppc_ksyms.c @@ -188,9 +188,6 @@ EXPORT_SYMBOL(adb_try_handler_change); EXPORT_SYMBOL(cuda_request); EXPORT_SYMBOL(cuda_poll); #endif /* CONFIG_ADB_CUDA */ -#if defined(CONFIG_PPC_MULTIPLATFORM) && defined(CONFIG_PPC32) -EXPORT_SYMBOL(_machine); -#endif #ifdef CONFIG_PPC_PMAC EXPORT_SYMBOL(sys_ctrler); #endif diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c index 5af39f866735..1bf3642cb859 100644 --- a/arch/powerpc/kernel/prom.c +++ b/arch/powerpc/kernel/prom.c @@ -48,9 +48,6 @@ #include #include #include -#ifdef CONFIG_PPC64 -#include -#endif #ifdef DEBUG #define DBG(fmt...) printk(KERN_ERR fmt) @@ -391,7 +388,7 @@ static int __devinit finish_node_interrupts(struct device_node *np, #ifdef CONFIG_PPC64 /* We offset irq numbers for the u3 MPIC by 128 in PowerMac */ - if (systemcfg->platform == PLATFORM_POWERMAC && ic && ic->parent) { + if (_machine == PLATFORM_POWERMAC && ic && ic->parent) { char *name = get_property(ic->parent, "name", NULL); if (name && !strcmp(name, "u3")) np->intrs[intrcount].line += 128; @@ -1161,13 +1158,9 @@ static int __init early_init_dt_scan_chosen(unsigned long node, prop = (u32 *)of_get_flat_dt_prop(node, "linux,platform", NULL); if (prop == NULL) return 0; -#ifdef CONFIG_PPC64 - systemcfg->platform = *prop; -#else #ifdef CONFIG_PPC_MULTIPLATFORM _machine = *prop; #endif -#endif #ifdef CONFIG_PPC64 /* check if iommu is forced on or off */ @@ -1346,9 +1339,6 @@ void __init early_init_devtree(void *params) of_scan_flat_dt(early_init_dt_scan_memory, NULL); lmb_enforce_memory_limit(memory_limit); lmb_analyze(); -#ifdef CONFIG_PPC64 - systemcfg->physicalMemorySize = lmb_phys_mem_size(); -#endif lmb_reserve(0, __pa(klimit)); DBG("Phys. mem: %lx\n", lmb_phys_mem_size()); @@ -1915,7 +1905,7 @@ static int of_finish_dynamic_node(struct device_node *node, /* We don't support that function on PowerMac, at least * not yet */ - if (systemcfg->platform == PLATFORM_POWERMAC) + if (_machine == PLATFORM_POWERMAC) return -ENODEV; /* fix up new node's linux_phandle field */ @@ -1999,9 +1989,11 @@ int prom_add_property(struct device_node* np, struct property* prop) *next = prop; write_unlock(&devtree_lock); +#ifdef CONFIG_PROC_DEVICETREE /* try to add to proc as well if it was initialized */ if (np->pde) proc_device_tree_add_prop(np->pde, prop); +#endif /* CONFIG_PROC_DEVICETREE */ return 0; } diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c index 09db1bb9ec91..4ce0105c308e 100644 --- a/arch/powerpc/kernel/prom_init.c +++ b/arch/powerpc/kernel/prom_init.c @@ -117,11 +117,6 @@ int of_workarounds; #define prom_debug(x...) #endif -#ifdef CONFIG_PPC32 -#define PLATFORM_POWERMAC _MACH_Pmac -#define PLATFORM_CHRP _MACH_chrp -#endif - typedef u32 prom_arg_t; @@ -2078,7 +2073,8 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, /* * On pSeries, inform the firmware about our capabilities */ - if (RELOC(of_platform) & PLATFORM_PSERIES) + if (RELOC(of_platform) == PLATFORM_PSERIES || + RELOC(of_platform) == PLATFORM_PSERIES_LPAR) prom_send_capabilities(); #endif diff --git a/arch/powerpc/kernel/rtas-proc.c b/arch/powerpc/kernel/rtas-proc.c index 5bdd5b079d96..ae1a36449ccd 100644 --- a/arch/powerpc/kernel/rtas-proc.c +++ b/arch/powerpc/kernel/rtas-proc.c @@ -259,7 +259,7 @@ static int __init proc_rtas_init(void) { struct proc_dir_entry *entry; - if (!(systemcfg->platform & PLATFORM_PSERIES)) + if (_machine != PLATFORM_PSERIES && _machine != PLATFORM_PSERIES_LPAR) return 1; rtas_node = of_find_node_by_name(NULL, "rtas"); diff --git a/arch/powerpc/kernel/rtas.c b/arch/powerpc/kernel/rtas.c index 9d4e07f6f1ec..4283fa33f784 100644 --- a/arch/powerpc/kernel/rtas.c +++ b/arch/powerpc/kernel/rtas.c @@ -29,9 +29,6 @@ #include #include #include -#ifdef CONFIG_PPC64 -#include -#endif struct rtas_t rtas = { .lock = SPIN_LOCK_UNLOCKED @@ -671,7 +668,7 @@ void __init rtas_initialize(void) * the stop-self token if any */ #ifdef CONFIG_PPC64 - if (systemcfg->platform == PLATFORM_PSERIES_LPAR) + if (_machine == PLATFORM_PSERIES_LPAR) rtas_region = min(lmb.rmo_size, RTAS_INSTANTIATE_MAX); #endif rtas_rmo_buf = lmb_alloc_base(RTAS_RMOBUF_MAX, PAGE_SIZE, rtas_region); diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c index 17c46a2e356d..06e4ef215627 100644 --- a/arch/powerpc/kernel/setup-common.c +++ b/arch/powerpc/kernel/setup-common.c @@ -33,6 +33,7 @@ #include #include #include +#include #include #include #include @@ -63,6 +64,11 @@ #define DBG(fmt...) #endif +#ifdef CONFIG_PPC_MULTIPLATFORM +int _machine = 0; +EXPORT_SYMBOL(_machine); +#endif + /* * This still seems to be needed... -- paulus */ @@ -513,8 +519,8 @@ void __init smp_setup_cpu_maps(void) * On pSeries LPAR, we need to know how many cpus * could possibly be added to this partition. */ - if (systemcfg->platform == PLATFORM_PSERIES_LPAR && - (dn = of_find_node_by_path("/rtas"))) { + if (_machine == PLATFORM_PSERIES_LPAR && + (dn = of_find_node_by_path("/rtas"))) { int num_addr_cell, num_size_cell, maxcpus; unsigned int *ireg; @@ -558,7 +564,7 @@ void __init smp_setup_cpu_maps(void) cpu_set(cpu ^ 0x1, cpu_sibling_map[cpu]); } - systemcfg->processorCount = num_present_cpus(); + _systemcfg->processorCount = num_present_cpus(); #endif /* CONFIG_PPC64 */ } #endif /* CONFIG_SMP */ diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c index 55f485f78c00..f73d7681b2e9 100644 --- a/arch/powerpc/kernel/setup_32.c +++ b/arch/powerpc/kernel/setup_32.c @@ -72,8 +72,6 @@ unsigned int DMA_MODE_WRITE; int have_of = 1; #ifdef CONFIG_PPC_MULTIPLATFORM -int _machine = 0; - extern void prep_init(void); extern void pmac_init(void); extern void chrp_init(void); diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index fe39aac4f24d..be607b877a55 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -100,10 +100,9 @@ extern void udbg_init_maple_realmode(void); extern unsigned long klimit; extern void mm_init_ppc64(void); -extern void stab_initialize(unsigned long stab); -extern void htab_initialize(void); extern void early_init_devtree(void *flat_dt); extern void unflatten_device_tree(void); +extern void check_for_initrd(void); int have_of = 1; int boot_cpuid = 0; @@ -256,11 +255,10 @@ void __init early_setup(unsigned long dt_ptr) * Iterate all ppc_md structures until we find the proper * one for the current machine type */ - DBG("Probing machine type for platform %x...\n", - systemcfg->platform); + DBG("Probing machine type for platform %x...\n", _machine); for (mach = machines; *mach; mach++) { - if ((*mach)->probe(systemcfg->platform)) + if ((*mach)->probe(_machine)) break; } /* What can we do if we didn't find ? */ @@ -292,6 +290,28 @@ void __init early_setup(unsigned long dt_ptr) DBG(" <- early_setup()\n"); } +#ifdef CONFIG_SMP +void early_setup_secondary(void) +{ + struct paca_struct *lpaca = get_paca(); + + /* Mark enabled in PACA */ + lpaca->proc_enabled = 0; + + /* Initialize hash table for that CPU */ + htab_initialize_secondary(); + + /* Initialize STAB/SLB. We use a virtual address as it works + * in real mode on pSeries and we want a virutal address on + * iSeries anyway + */ + if (cpu_has_feature(CPU_FTR_SLB)) + slb_initialize(); + else + stab_initialize(lpaca->stab_addr); +} + +#endif /* CONFIG_SMP */ #if defined(CONFIG_SMP) || defined(CONFIG_KEXEC) void smp_release_cpus(void) @@ -317,7 +337,8 @@ void smp_release_cpus(void) #endif /* CONFIG_SMP || CONFIG_KEXEC */ /* - * Initialize some remaining members of the ppc64_caches and systemcfg structures + * Initialize some remaining members of the ppc64_caches and systemcfg + * structures * (at least until we get rid of them completely). This is mostly some * cache informations about the CPU that will be used by cache flush * routines and/or provided to userland @@ -342,7 +363,7 @@ static void __init initialize_cache_info(void) const char *dc, *ic; /* Then read cache informations */ - if (systemcfg->platform == PLATFORM_POWERMAC) { + if (_machine == PLATFORM_POWERMAC) { dc = "d-cache-block-size"; ic = "i-cache-block-size"; } else { @@ -362,8 +383,8 @@ static void __init initialize_cache_info(void) DBG("Argh, can't find dcache properties ! " "sizep: %p, lsizep: %p\n", sizep, lsizep); - systemcfg->dcache_size = ppc64_caches.dsize = size; - systemcfg->dcache_line_size = + _systemcfg->dcache_size = ppc64_caches.dsize = size; + _systemcfg->dcache_line_size = ppc64_caches.dline_size = lsize; ppc64_caches.log_dline_size = __ilog2(lsize); ppc64_caches.dlines_per_page = PAGE_SIZE / lsize; @@ -380,8 +401,8 @@ static void __init initialize_cache_info(void) DBG("Argh, can't find icache properties ! " "sizep: %p, lsizep: %p\n", sizep, lsizep); - systemcfg->icache_size = ppc64_caches.isize = size; - systemcfg->icache_line_size = + _systemcfg->icache_size = ppc64_caches.isize = size; + _systemcfg->icache_line_size = ppc64_caches.iline_size = lsize; ppc64_caches.log_iline_size = __ilog2(lsize); ppc64_caches.ilines_per_page = PAGE_SIZE / lsize; @@ -389,10 +410,12 @@ static void __init initialize_cache_info(void) } /* Add an eye catcher and the systemcfg layout version number */ - strcpy(systemcfg->eye_catcher, "SYSTEMCFG:PPC64"); - systemcfg->version.major = SYSTEMCFG_MAJOR; - systemcfg->version.minor = SYSTEMCFG_MINOR; - systemcfg->processor = mfspr(SPRN_PVR); + strcpy(_systemcfg->eye_catcher, "SYSTEMCFG:PPC64"); + _systemcfg->version.major = SYSTEMCFG_MAJOR; + _systemcfg->version.minor = SYSTEMCFG_MINOR; + _systemcfg->processor = mfspr(SPRN_PVR); + _systemcfg->platform = _machine; + _systemcfg->physicalMemorySize = lmb_phys_mem_size(); DBG(" <- initialize_cache_info()\n"); } @@ -481,10 +504,10 @@ void __init setup_system(void) printk("-----------------------------------------------------\n"); printk("ppc64_pft_size = 0x%lx\n", ppc64_pft_size); printk("ppc64_interrupt_controller = 0x%ld\n", ppc64_interrupt_controller); - printk("systemcfg = 0x%p\n", systemcfg); - printk("systemcfg->platform = 0x%x\n", systemcfg->platform); - printk("systemcfg->processorCount = 0x%lx\n", systemcfg->processorCount); - printk("systemcfg->physicalMemorySize = 0x%lx\n", systemcfg->physicalMemorySize); + printk("systemcfg = 0x%p\n", _systemcfg); + printk("systemcfg->platform = 0x%x\n", _systemcfg->platform); + printk("systemcfg->processorCount = 0x%lx\n", _systemcfg->processorCount); + printk("systemcfg->physicalMemorySize = 0x%lx\n", _systemcfg->physicalMemorySize); printk("ppc64_caches.dcache_line_size = 0x%x\n", ppc64_caches.dline_size); printk("ppc64_caches.icache_line_size = 0x%x\n", @@ -566,12 +589,12 @@ void __init setup_syscall_map(void) for (i = 0; i < __NR_syscalls; i++) { if (sys_call_table[i*2] != sys_ni_syscall) { count64++; - systemcfg->syscall_map_64[i >> 5] |= + _systemcfg->syscall_map_64[i >> 5] |= 0x80000000UL >> (i & 0x1f); } if (sys_call_table[i*2+1] != sys_ni_syscall) { count32++; - systemcfg->syscall_map_32[i >> 5] |= + _systemcfg->syscall_map_32[i >> 5] |= 0x80000000UL >> (i & 0x1f); } } diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c index 5c330c3366e4..7fd530898bd1 100644 --- a/arch/powerpc/kernel/smp.c +++ b/arch/powerpc/kernel/smp.c @@ -44,6 +44,7 @@ #include #include #include +#include #ifdef CONFIG_PPC64 #include #endif @@ -368,7 +369,9 @@ int generic_cpu_disable(void) if (cpu == boot_cpuid) return -EBUSY; - systemcfg->processorCount--; +#ifdef CONFIG_PPC64 + _systemcfg->processorCount--; +#endif cpu_clear(cpu, cpu_online_map); fixup_irqs(cpu_online_map); return 0; diff --git a/arch/powerpc/kernel/sys_ppc32.c b/arch/powerpc/kernel/sys_ppc32.c index a8210ed5c686..9c921d1c4084 100644 --- a/arch/powerpc/kernel/sys_ppc32.c +++ b/arch/powerpc/kernel/sys_ppc32.c @@ -52,7 +52,6 @@ #include #include #include -#include #include /* readdir & getdents */ diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c index a6282b625b44..386006b3d616 100644 --- a/arch/powerpc/kernel/time.c +++ b/arch/powerpc/kernel/time.c @@ -271,13 +271,13 @@ static inline void update_gtod(u64 new_tb_stamp, u64 new_stamp_xsec, * tb_to_xs and stamp_xsec values are consistent. If not, then it * loops back and reads them again until this criteria is met. */ - ++(systemcfg->tb_update_count); + ++(_systemcfg->tb_update_count); smp_wmb(); - systemcfg->tb_orig_stamp = new_tb_stamp; - systemcfg->stamp_xsec = new_stamp_xsec; - systemcfg->tb_to_xs = new_tb_to_xs; + _systemcfg->tb_orig_stamp = new_tb_stamp; + _systemcfg->stamp_xsec = new_stamp_xsec; + _systemcfg->tb_to_xs = new_tb_to_xs; smp_wmb(); - ++(systemcfg->tb_update_count); + ++(_systemcfg->tb_update_count); #endif } @@ -357,8 +357,9 @@ static void iSeries_tb_recal(void) do_gtod.tb_ticks_per_sec = tb_ticks_per_sec; tb_to_xs = divres.result_low; do_gtod.varp->tb_to_xs = tb_to_xs; - systemcfg->tb_ticks_per_sec = tb_ticks_per_sec; - systemcfg->tb_to_xs = tb_to_xs; + _systemcfg->tb_ticks_per_sec = + tb_ticks_per_sec; + _systemcfg->tb_to_xs = tb_to_xs; } else { printk( "Titan recalibrate: FAILED (difference > 4 percent)\n" @@ -559,8 +560,8 @@ int do_settimeofday(struct timespec *tv) update_gtod(tb_last_jiffy, new_xsec, do_gtod.varp->tb_to_xs); #ifdef CONFIG_PPC64 - systemcfg->tz_minuteswest = sys_tz.tz_minuteswest; - systemcfg->tz_dsttime = sys_tz.tz_dsttime; + _systemcfg->tz_minuteswest = sys_tz.tz_minuteswest; + _systemcfg->tz_dsttime = sys_tz.tz_dsttime; #endif write_sequnlock_irqrestore(&xtime_lock, flags); @@ -711,11 +712,11 @@ void __init time_init(void) do_gtod.varp->tb_to_xs = tb_to_xs; do_gtod.tb_to_us = tb_to_us; #ifdef CONFIG_PPC64 - systemcfg->tb_orig_stamp = tb_last_jiffy; - systemcfg->tb_update_count = 0; - systemcfg->tb_ticks_per_sec = tb_ticks_per_sec; - systemcfg->stamp_xsec = xtime.tv_sec * XSEC_PER_SEC; - systemcfg->tb_to_xs = tb_to_xs; + _systemcfg->tb_orig_stamp = tb_last_jiffy; + _systemcfg->tb_update_count = 0; + _systemcfg->tb_ticks_per_sec = tb_ticks_per_sec; + _systemcfg->stamp_xsec = xtime.tv_sec * XSEC_PER_SEC; + _systemcfg->tb_to_xs = tb_to_xs; #endif time_freq = 0; diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c index 0578f8387603..2020bb7648fb 100644 --- a/arch/powerpc/kernel/traps.c +++ b/arch/powerpc/kernel/traps.c @@ -129,7 +129,7 @@ int die(const char *str, struct pt_regs *regs, long err) nl = 1; #endif #ifdef CONFIG_PPC64 - switch (systemcfg->platform) { + switch (_machine) { case PLATFORM_PSERIES: printk("PSERIES "); nl = 1; diff --git a/arch/powerpc/mm/hash_utils_64.c b/arch/powerpc/mm/hash_utils_64.c index c8c9c2df97e9..706e8a63ced9 100644 --- a/arch/powerpc/mm/hash_utils_64.c +++ b/arch/powerpc/mm/hash_utils_64.c @@ -84,10 +84,11 @@ extern unsigned long dart_tablebase; #endif /* CONFIG_U3_DART */ +static unsigned long _SDR1; +struct mmu_psize_def mmu_psize_defs[MMU_PAGE_COUNT]; + hpte_t *htab_address; unsigned long htab_hash_mask; -unsigned long _SDR1; -struct mmu_psize_def mmu_psize_defs[MMU_PAGE_COUNT]; int mmu_linear_psize = MMU_PAGE_4K; int mmu_virtual_psize = MMU_PAGE_4K; #ifdef CONFIG_HUGETLB_PAGE @@ -165,7 +166,7 @@ int htab_bolt_mapping(unsigned long vstart, unsigned long vend, * normal insert callback here. */ #ifdef CONFIG_PPC_ISERIES - if (systemcfg->platform == PLATFORM_ISERIES_LPAR) + if (_machine == PLATFORM_ISERIES_LPAR) ret = iSeries_hpte_insert(hpteg, va, virt_to_abs(paddr), tmp_mode, @@ -174,7 +175,7 @@ int htab_bolt_mapping(unsigned long vstart, unsigned long vend, else #endif #ifdef CONFIG_PPC_PSERIES - if (systemcfg->platform & PLATFORM_LPAR) + if (_machine & PLATFORM_LPAR) ret = pSeries_lpar_hpte_insert(hpteg, va, virt_to_abs(paddr), tmp_mode, @@ -293,7 +294,7 @@ static void __init htab_init_page_sizes(void) * Not in the device-tree, let's fallback on known size * list for 16M capable GP & GR */ - if ((systemcfg->platform != PLATFORM_ISERIES_LPAR) && + if ((_machine != PLATFORM_ISERIES_LPAR) && cpu_has_feature(CPU_FTR_16M_PAGE)) memcpy(mmu_psize_defs, mmu_psize_defaults_gp, sizeof(mmu_psize_defaults_gp)); @@ -364,7 +365,7 @@ static int __init htab_dt_scan_pftsize(unsigned long node, static unsigned long __init htab_get_table_size(void) { - unsigned long rnd_mem_size, pteg_count; + unsigned long mem_size, rnd_mem_size, pteg_count; /* If hash size isn't already provided by the platform, we try to * retreive it from the device-tree. If it's not there neither, we @@ -376,8 +377,9 @@ static unsigned long __init htab_get_table_size(void) return 1UL << ppc64_pft_size; /* round mem_size up to next power of 2 */ - rnd_mem_size = 1UL << __ilog2(systemcfg->physicalMemorySize); - if (rnd_mem_size < systemcfg->physicalMemorySize) + mem_size = lmb_phys_mem_size(); + rnd_mem_size = 1UL << __ilog2(mem_size); + if (rnd_mem_size < mem_size) rnd_mem_size <<= 1; /* # pages / 2 */ @@ -419,7 +421,7 @@ void __init htab_initialize(void) htab_hash_mask = pteg_count - 1; - if (systemcfg->platform & PLATFORM_LPAR) { + if (platform_is_lpar()) { /* Using a hypervisor which owns the htab */ htab_address = NULL; _SDR1 = 0; @@ -440,6 +442,9 @@ void __init htab_initialize(void) /* Initialize the HPT with no entries */ memset((void *)table, 0, htab_size_bytes); + + /* Set SDR1 */ + mtspr(SPRN_SDR1, _SDR1); } mode_rw = _PAGE_ACCESSED | _PAGE_DIRTY | _PAGE_COHERENT | PP_RWXX; @@ -509,6 +514,12 @@ void __init htab_initialize(void) #undef KB #undef MB +void __init htab_initialize_secondary(void) +{ + if (!platform_is_lpar()) + mtspr(SPRN_SDR1, _SDR1); +} + /* * Called by asm hashtable.S for doing lazy icache flush */ diff --git a/arch/powerpc/mm/stab.c b/arch/powerpc/mm/stab.c index fa325dbf98fc..cfbb4e1f966b 100644 --- a/arch/powerpc/mm/stab.c +++ b/arch/powerpc/mm/stab.c @@ -20,6 +20,7 @@ #include #include #include +#include struct stab_entry { unsigned long esid_data; @@ -256,7 +257,7 @@ void stabs_alloc(void) paca[cpu].stab_addr = newstab; paca[cpu].stab_real = virt_to_abs(newstab); - printk(KERN_DEBUG "Segment table for CPU %d at 0x%lx " + printk(KERN_INFO "Segment table for CPU %d at 0x%lx " "virtual, 0x%lx absolute\n", cpu, paca[cpu].stab_addr, paca[cpu].stab_real); } @@ -270,10 +271,28 @@ void stabs_alloc(void) void stab_initialize(unsigned long stab) { unsigned long vsid = get_kernel_vsid(KERNELBASE); + unsigned long stabreal; asm volatile("isync; slbia; isync":::"memory"); make_ste(stab, GET_ESID(KERNELBASE), vsid); /* Order update */ asm volatile("sync":::"memory"); + + /* Set ASR */ + stabreal = get_paca()->stab_real | 0x1ul; + +#ifdef CONFIG_PPC_ISERIES + if (firmware_has_feature(FW_FEATURE_ISERIES)) { + HvCall1(HvCallBaseSetASR, stabreal); + return; + } +#endif /* CONFIG_PPC_ISERIES */ +#ifdef CONFIG_PPC_PSERIES + if (platform_is_lpar()) { + plpar_hcall_norets(H_SET_ASR, stabreal); + return; + } +#endif + mtspr(SPRN_ASR, stabreal); } diff --git a/arch/powerpc/oprofile/op_model_power4.c b/arch/powerpc/oprofile/op_model_power4.c index c4ee5478427b..e3a024e324b6 100644 --- a/arch/powerpc/oprofile/op_model_power4.c +++ b/arch/powerpc/oprofile/op_model_power4.c @@ -233,8 +233,7 @@ static unsigned long get_pc(struct pt_regs *regs) mmcra = mfspr(SPRN_MMCRA); /* Were we in the hypervisor? */ - if ((systemcfg->platform == PLATFORM_PSERIES_LPAR) && - (mmcra & MMCRA_SIHV)) + if (platform_is_lpar() && (mmcra & MMCRA_SIHV)) /* function descriptor madness */ return *((unsigned long *)hypervisor_bucket); diff --git a/arch/powerpc/platforms/iseries/setup.c b/arch/powerpc/platforms/iseries/setup.c index 7f8f0cda6a74..623c39aa0436 100644 --- a/arch/powerpc/platforms/iseries/setup.c +++ b/arch/powerpc/platforms/iseries/setup.c @@ -39,6 +39,7 @@ #include #include #include +#include #include #include @@ -71,7 +72,7 @@ extern void hvlog(char *fmt, ...); #endif /* Function Prototypes */ -static void build_iSeries_Memory_Map(void); +static unsigned long build_iSeries_Memory_Map(void); static void iseries_shared_idle(void); static void iseries_dedicated_idle(void); #ifdef CONFIG_PCI @@ -403,9 +404,11 @@ void mschunks_alloc(unsigned long num_chunks) * a table used to translate Linux's physical addresses to these * absolute addresses. Absolute addresses are needed when * communicating with the hypervisor (e.g. to build HPT entries) + * + * Returns the physical memory size */ -static void __init build_iSeries_Memory_Map(void) +static unsigned long __init build_iSeries_Memory_Map(void) { u32 loadAreaFirstChunk, loadAreaLastChunk, loadAreaSize; u32 nextPhysChunk; @@ -538,7 +541,7 @@ static void __init build_iSeries_Memory_Map(void) * which should be equal to * nextPhysChunk */ - systemcfg->physicalMemorySize = chunk_to_addr(nextPhysChunk); + return chunk_to_addr(nextPhysChunk); } /* @@ -564,8 +567,8 @@ static void __init iSeries_setup_arch(void) printk("Max physical processors = %d\n", itVpdAreas.xSlicMaxPhysicalProcs); - systemcfg->processor = xIoHriProcessorVpd[procIx].xPVR; - printk("Processor version = %x\n", systemcfg->processor); + _systemcfg->processor = xIoHriProcessorVpd[procIx].xPVR; + printk("Processor version = %x\n", _systemcfg->processor); } static void iSeries_show_cpuinfo(struct seq_file *m) @@ -929,7 +932,7 @@ void dt_cpus(struct iseries_flat_dt *dt) dt_end_node(dt); } -void build_flat_dt(struct iseries_flat_dt *dt) +void build_flat_dt(struct iseries_flat_dt *dt, unsigned long phys_mem_size) { u64 tmp[2]; @@ -945,7 +948,7 @@ void build_flat_dt(struct iseries_flat_dt *dt) dt_prop_str(dt, "name", "memory"); dt_prop_str(dt, "device_type", "memory"); tmp[0] = 0; - tmp[1] = systemcfg->physicalMemorySize; + tmp[1] = phys_mem_size; dt_prop_u64_list(dt, "reg", tmp, 2); dt_end_node(dt); @@ -965,13 +968,15 @@ void build_flat_dt(struct iseries_flat_dt *dt) void * __init iSeries_early_setup(void) { + unsigned long phys_mem_size; + iSeries_fixup_klimit(); /* * Initialize the table which translate Linux physical addresses to * AS/400 absolute addresses */ - build_iSeries_Memory_Map(); + phys_mem_size = build_iSeries_Memory_Map(); iSeries_get_cmdline(); @@ -981,7 +986,7 @@ void * __init iSeries_early_setup(void) /* Parse early parameters, in particular mem=x */ parse_early_param(); - build_flat_dt(&iseries_dt); + build_flat_dt(&iseries_dt, phys_mem_size); return (void *) __pa(&iseries_dt); } diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c index b760836bb9d1..a33ba833109c 100644 --- a/arch/powerpc/platforms/pseries/eeh.c +++ b/arch/powerpc/platforms/pseries/eeh.c @@ -32,7 +32,6 @@ #include #include #include -#include #undef DEBUG @@ -1186,7 +1185,7 @@ static int __init eeh_init_proc(void) { struct proc_dir_entry *e; - if (systemcfg->platform & PLATFORM_PSERIES) { + if (platform_is_pseries()) { e = create_proc_entry("ppc64/eeh", 0, NULL); if (e) e->proc_fops = &proc_eeh_operations; diff --git a/arch/powerpc/platforms/pseries/iommu.c b/arch/powerpc/platforms/pseries/iommu.c index fcc50bfd43fd..97ba5214417f 100644 --- a/arch/powerpc/platforms/pseries/iommu.c +++ b/arch/powerpc/platforms/pseries/iommu.c @@ -42,7 +42,6 @@ #include #include #include -#include #include #include #include @@ -582,7 +581,7 @@ void iommu_init_early_pSeries(void) return; } - if (systemcfg->platform & PLATFORM_LPAR) { + if (platform_is_lpar()) { if (firmware_has_feature(FW_FEATURE_MULTITCE)) { ppc_md.tce_build = tce_buildmulti_pSeriesLP; ppc_md.tce_free = tce_freemulti_pSeriesLP; diff --git a/arch/powerpc/platforms/pseries/pci.c b/arch/powerpc/platforms/pseries/pci.c index c198656a3bb5..6b0772f35692 100644 --- a/arch/powerpc/platforms/pseries/pci.c +++ b/arch/powerpc/platforms/pseries/pci.c @@ -123,7 +123,7 @@ static void fixup_winbond_82c105(struct pci_dev* dev) int i; unsigned int reg; - if (!(systemcfg->platform & PLATFORM_PSERIES)) + if (!platform_is_pseries()) return; printk("Using INTC for W82c105 IDE controller.\n"); diff --git a/arch/powerpc/platforms/pseries/reconfig.c b/arch/powerpc/platforms/pseries/reconfig.c index d7d400339458..d8864164dbe8 100644 --- a/arch/powerpc/platforms/pseries/reconfig.c +++ b/arch/powerpc/platforms/pseries/reconfig.c @@ -408,7 +408,7 @@ static int proc_ppc64_create_ofdt(void) { struct proc_dir_entry *ent; - if (!(systemcfg->platform & PLATFORM_PSERIES)) + if (!platform_is_pseries()) return 0; ent = create_proc_entry("ppc64/ofdt", S_IWUSR, NULL); diff --git a/arch/powerpc/platforms/pseries/rtasd.c b/arch/powerpc/platforms/pseries/rtasd.c index e26b0420b6dd..00cf331a1dc4 100644 --- a/arch/powerpc/platforms/pseries/rtasd.c +++ b/arch/powerpc/platforms/pseries/rtasd.c @@ -482,10 +482,12 @@ static int __init rtas_init(void) { struct proc_dir_entry *entry; - /* No RTAS, only warn if we are on a pSeries box */ + if (!platform_is_pseries()) + return 0; + + /* No RTAS */ if (rtas_token("event-scan") == RTAS_UNKNOWN_SERVICE) { - if (systemcfg->platform & PLATFORM_PSERIES) - printk(KERN_INFO "rtasd: no event-scan on system\n"); + printk(KERN_INFO "rtasd: no event-scan on system\n"); return 1; } diff --git a/arch/powerpc/platforms/pseries/setup.c b/arch/powerpc/platforms/pseries/setup.c index a093a0d4dd69..e94247c28d42 100644 --- a/arch/powerpc/platforms/pseries/setup.c +++ b/arch/powerpc/platforms/pseries/setup.c @@ -249,7 +249,7 @@ static void __init pSeries_setup_arch(void) ppc_md.idle_loop = default_idle; } - if (systemcfg->platform & PLATFORM_LPAR) + if (platform_is_lpar()) ppc_md.enable_pmcs = pseries_lpar_enable_pmcs; else ppc_md.enable_pmcs = power4_enable_pmcs; @@ -378,7 +378,7 @@ static void __init pSeries_init_early(void) fw_feature_init(); - if (systemcfg->platform & PLATFORM_LPAR) + if (platform_is_lpar()) hpte_init_lpar(); else { hpte_init_native(); @@ -388,7 +388,7 @@ static void __init pSeries_init_early(void) generic_find_legacy_serial_ports(&physport, &default_speed); - if (systemcfg->platform & PLATFORM_LPAR) + if (platform_is_lpar()) find_udbg_vterm(); else if (physport) { /* Map the uart for udbg. */ @@ -592,7 +592,7 @@ static void pseries_shared_idle(void) static int pSeries_pci_probe_mode(struct pci_bus *bus) { - if (systemcfg->platform & PLATFORM_LPAR) + if (platform_is_lpar()) return PCI_PROBE_DEVTREE; return PCI_PROBE_NORMAL; } diff --git a/arch/powerpc/platforms/pseries/smp.c b/arch/powerpc/platforms/pseries/smp.c index 7a243e8ccd7e..3ba794ca3288 100644 --- a/arch/powerpc/platforms/pseries/smp.c +++ b/arch/powerpc/platforms/pseries/smp.c @@ -46,6 +46,7 @@ #include #include #include +#include #include "plpar_wrappers.h" @@ -96,7 +97,7 @@ int pSeries_cpu_disable(void) int cpu = smp_processor_id(); cpu_clear(cpu, cpu_online_map); - systemcfg->processorCount--; + _systemcfg->processorCount--; /*fix boot_cpuid here*/ if (cpu == boot_cpuid) @@ -441,7 +442,7 @@ void __init smp_init_pSeries(void) smp_ops->cpu_die = pSeries_cpu_die; /* Processors can be added/removed only on LPAR */ - if (systemcfg->platform == PLATFORM_PSERIES_LPAR) + if (platform_is_lpar()) pSeries_reconfig_notifier_register(&pSeries_smp_nb); #endif diff --git a/arch/powerpc/platforms/pseries/xics.c b/arch/powerpc/platforms/pseries/xics.c index 405c4f3229b3..72ac18067ece 100644 --- a/arch/powerpc/platforms/pseries/xics.c +++ b/arch/powerpc/platforms/pseries/xics.c @@ -545,7 +545,9 @@ nextnode: of_node_put(np); } - if (systemcfg->platform == PLATFORM_PSERIES) { + if (platform_is_lpar()) + ops = &pSeriesLP_ops; + else { #ifdef CONFIG_SMP for_each_cpu(i) { int hard_id; @@ -561,8 +563,6 @@ nextnode: #else xics_per_cpu[0] = ioremap(intr_base, intr_size); #endif /* CONFIG_SMP */ - } else if (systemcfg->platform == PLATFORM_PSERIES_LPAR) { - ops = &pSeriesLP_ops; } xics_8259_pic.enable = i8259_pic.enable; diff --git a/arch/ppc/kernel/ppc_ksyms.c b/arch/ppc/kernel/ppc_ksyms.c index 307077f14936..66073f775193 100644 --- a/arch/ppc/kernel/ppc_ksyms.c +++ b/arch/ppc/kernel/ppc_ksyms.c @@ -217,9 +217,6 @@ EXPORT_SYMBOL(adb_try_handler_change); EXPORT_SYMBOL(cuda_request); EXPORT_SYMBOL(cuda_poll); #endif /* CONFIG_ADB_CUDA */ -#ifdef CONFIG_PPC_MULTIPLATFORM -EXPORT_SYMBOL(_machine); -#endif #ifdef CONFIG_PPC_PMAC EXPORT_SYMBOL(sys_ctrler); EXPORT_SYMBOL(pmac_newworld); diff --git a/arch/ppc/kernel/setup.c b/arch/ppc/kernel/setup.c index 6bcb85d2b7fd..dc55e1abc45b 100644 --- a/arch/ppc/kernel/setup.c +++ b/arch/ppc/kernel/setup.c @@ -76,6 +76,7 @@ unsigned int DMA_MODE_WRITE; #ifdef CONFIG_PPC_MULTIPLATFORM int _machine = 0; +EXPORT_SYMBOL(_machine); extern void prep_init(unsigned long r3, unsigned long r4, unsigned long r5, unsigned long r6, unsigned long r7); diff --git a/arch/ppc64/kernel/asm-offsets.c b/arch/ppc64/kernel/asm-offsets.c index bce9065da6cb..84ab5c18ef52 100644 --- a/arch/ppc64/kernel/asm-offsets.c +++ b/arch/ppc64/kernel/asm-offsets.c @@ -74,7 +74,6 @@ int main(void) DEFINE(ICACHEL1LINESIZE, offsetof(struct ppc64_caches, iline_size)); DEFINE(ICACHEL1LOGLINESIZE, offsetof(struct ppc64_caches, log_iline_size)); DEFINE(ICACHEL1LINESPERPAGE, offsetof(struct ppc64_caches, ilines_per_page)); - DEFINE(PLATFORM, offsetof(struct systemcfg, platform)); DEFINE(PLATFORM_LPAR, PLATFORM_LPAR); /* paca */ diff --git a/arch/ppc64/kernel/head.S b/arch/ppc64/kernel/head.S index 9e8050ea1225..1c869ea72d28 100644 --- a/arch/ppc64/kernel/head.S +++ b/arch/ppc64/kernel/head.S @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include @@ -1701,21 +1700,9 @@ _GLOBAL(__secondary_start) HMT_MEDIUM /* Set thread priority to MEDIUM */ ld r2,PACATOC(r13) - li r6,0 - stb r6,PACAPROCENABLED(r13) -#ifndef CONFIG_PPC_ISERIES - /* Initialize the page table pointer register. */ - LOADADDR(r6,_SDR1) - ld r6,0(r6) /* get the value of _SDR1 */ - mtspr SPRN_SDR1,r6 /* set the htab location */ -#endif - /* Initialize the first segment table (or SLB) entry */ - ld r3,PACASTABVIRT(r13) /* get addr of segment table */ -BEGIN_FTR_SECTION - bl .stab_initialize -END_FTR_SECTION_IFCLR(CPU_FTR_SLB) - bl .slb_initialize + /* Do early setup for that CPU */ + bl .early_setup_secondary /* Initialize the kernel stack. Just a repeat for iSeries. */ LOADADDR(r3,current_set) @@ -1724,37 +1711,6 @@ END_FTR_SECTION_IFCLR(CPU_FTR_SLB) addi r1,r1,THREAD_SIZE-STACK_FRAME_OVERHEAD std r1,PACAKSAVE(r13) - ld r3,PACASTABREAL(r13) /* get raddr of segment table */ - ori r4,r3,1 /* turn on valid bit */ - -#ifdef CONFIG_PPC_ISERIES - li r0,-1 /* hypervisor call */ - li r3,1 - sldi r3,r3,63 /* 0x8000000000000000 */ - ori r3,r3,4 /* 0x8000000000000004 */ - sc /* HvCall_setASR */ -#else - /* set the ASR */ - ld r3,systemcfg@got(r2) /* r3 = ptr to systemcfg */ - ld r3,0(r3) - lwz r3,PLATFORM(r3) /* r3 = platform flags */ - andi. r3,r3,PLATFORM_LPAR /* Test if bit 0 is set (LPAR bit) */ - beq 98f /* branch if result is 0 */ - mfspr r3,SPRN_PVR - srwi r3,r3,16 - cmpwi r3,0x37 /* SStar */ - beq 97f - cmpwi r3,0x36 /* IStar */ - beq 97f - cmpwi r3,0x34 /* Pulsar */ - bne 98f -97: li r3,H_SET_ASR /* hcall = H_SET_ASR */ - HVSC /* Invoking hcall */ - b 99f -98: /* !(rpa hypervisor) || !(star) */ - mtasr r4 /* set the stab location */ -99: -#endif li r7,0 mtlr r7 @@ -1896,40 +1852,6 @@ _STATIC(start_here_multiplatform) mr r3,r31 bl .early_setup - /* set the ASR */ - ld r3,PACASTABREAL(r13) - ori r4,r3,1 /* turn on valid bit */ - ld r3,systemcfg@got(r2) /* r3 = ptr to systemcfg */ - ld r3,0(r3) - lwz r3,PLATFORM(r3) /* r3 = platform flags */ - andi. r3,r3,PLATFORM_LPAR /* Test if bit 0 is set (LPAR bit) */ - beq 98f /* branch if result is 0 */ - mfspr r3,SPRN_PVR - srwi r3,r3,16 - cmpwi r3,0x37 /* SStar */ - beq 97f - cmpwi r3,0x36 /* IStar */ - beq 97f - cmpwi r3,0x34 /* Pulsar */ - bne 98f -97: li r3,H_SET_ASR /* hcall = H_SET_ASR */ - HVSC /* Invoking hcall */ - b 99f -98: /* !(rpa hypervisor) || !(star) */ - mtasr r4 /* set the stab location */ -99: - /* Set SDR1 (hash table pointer) */ - ld r3,systemcfg@got(r2) /* r3 = ptr to systemcfg */ - ld r3,0(r3) - lwz r3,PLATFORM(r3) /* r3 = platform flags */ - /* Test if bit 0 is set (LPAR bit) */ - andi. r3,r3,PLATFORM_LPAR - bne 98f /* branch if result is !0 */ - LOADADDR(r6,_SDR1) /* Only if NOT LPAR */ - sub r6,r6,r26 - ld r6,0(r6) /* get the value of _SDR1 */ - mtspr SPRN_SDR1,r6 /* set the htab location */ -98: LOADADDR(r3,.start_here_common) SET_REG_TO_CONST(r4, MSR_KERNEL) mtspr SPRN_SRR0,r3 diff --git a/arch/ppc64/kernel/idle.c b/arch/ppc64/kernel/idle.c index 715bc0e71e0f..b879d3057ef8 100644 --- a/arch/ppc64/kernel/idle.c +++ b/arch/ppc64/kernel/idle.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include diff --git a/arch/ppc64/kernel/lparcfg.c b/arch/ppc64/kernel/lparcfg.c index 3e7b2f28ec83..a80ed307515c 100644 --- a/arch/ppc64/kernel/lparcfg.c +++ b/arch/ppc64/kernel/lparcfg.c @@ -35,6 +35,7 @@ #include #include #include +#include #define MODULE_VERS "1.6" #define MODULE_NAME "lparcfg" @@ -371,7 +372,7 @@ static int lparcfg_data(struct seq_file *m, void *v) lrdrp = (int *)get_property(rtas_node, "ibm,lrdr-capacity", NULL); if (lrdrp == NULL) { - partition_potential_processors = systemcfg->processorCount; + partition_potential_processors = _systemcfg->processorCount; } else { partition_potential_processors = *(lrdrp + 4); } diff --git a/arch/ppc64/kernel/nvram.c b/arch/ppc64/kernel/nvram.c index 4fb1a9f5060d..c0fcd29918ce 100644 --- a/arch/ppc64/kernel/nvram.c +++ b/arch/ppc64/kernel/nvram.c @@ -31,7 +31,6 @@ #include #include #include -#include #undef DEBUG_NVRAM @@ -167,7 +166,7 @@ static int dev_nvram_ioctl(struct inode *inode, struct file *file, case IOC_NVRAM_GET_OFFSET: { int part, offset; - if (systemcfg->platform != PLATFORM_POWERMAC) + if (_machine != PLATFORM_POWERMAC) return -EINVAL; if (copy_from_user(&part, (void __user*)arg, sizeof(part)) != 0) return -EFAULT; @@ -450,7 +449,7 @@ static int nvram_setup_partition(void) * in our nvram, as Apple defined partitions use pretty much * all of the space */ - if (systemcfg->platform == PLATFORM_POWERMAC) + if (_machine == PLATFORM_POWERMAC) return -ENOSPC; /* see if we have an OS partition that meets our needs. diff --git a/arch/ppc64/kernel/pci.c b/arch/ppc64/kernel/pci.c index 30247ff74972..66698fdf059e 100644 --- a/arch/ppc64/kernel/pci.c +++ b/arch/ppc64/kernel/pci.c @@ -1277,12 +1277,9 @@ long sys_pciconfig_iobase(long which, unsigned long in_bus, * G5 machines... So when something asks for bus 0 io base * (bus 0 is HT root), we return the AGP one instead. */ -#ifdef CONFIG_PPC_PMAC - if (systemcfg->platform == PLATFORM_POWERMAC && - machine_is_compatible("MacRISC4")) + if (machine_is_compatible("MacRISC4")) if (in_bus == 0) in_bus = 0xf0; -#endif /* CONFIG_PPC_PMAC */ /* That syscall isn't quite compatible with PCI domains, but it's * used on pre-domains setup. We return the first match diff --git a/arch/ppc64/kernel/proc_ppc64.c b/arch/ppc64/kernel/proc_ppc64.c index 24e955ee9487..c893a11ee198 100644 --- a/arch/ppc64/kernel/proc_ppc64.c +++ b/arch/ppc64/kernel/proc_ppc64.c @@ -53,7 +53,7 @@ static int __init proc_ppc64_create(void) if (!root) return 1; - if (!(systemcfg->platform & (PLATFORM_PSERIES | PLATFORM_CELL))) + if (!(platform_is_pseries() || _machine == PLATFORM_CELL)) return 0; if (!proc_mkdir("rtas", root)) @@ -74,7 +74,7 @@ static int __init proc_ppc64_init(void) if (!pde) return 1; pde->nlink = 1; - pde->data = systemcfg; + pde->data = _systemcfg; pde->size = PAGE_SIZE; pde->proc_fops = &page_map_fops; diff --git a/arch/ppc64/kernel/prom.c b/arch/ppc64/kernel/prom.c index 3402fbee62c7..fbad2c360784 100644 --- a/arch/ppc64/kernel/prom.c +++ b/arch/ppc64/kernel/prom.c @@ -318,7 +318,7 @@ static int __devinit finish_node_interrupts(struct device_node *np, } /* We offset irq numbers for the u3 MPIC by 128 in PowerMac */ - if (systemcfg->platform == PLATFORM_POWERMAC && ic && ic->parent) { + if (_machine == PLATFORM_POWERMAC && ic && ic->parent) { char *name = get_property(ic->parent, "name", NULL); if (name && !strcmp(name, "u3")) np->intrs[intrcount].line += 128; @@ -1065,7 +1065,7 @@ static int __init early_init_dt_scan_chosen(unsigned long node, prop = (u32 *)of_get_flat_dt_prop(node, "linux,platform", NULL); if (prop == NULL) return 0; - systemcfg->platform = *prop; + _machine = *prop; /* check if iommu is forced on or off */ if (of_get_flat_dt_prop(node, "linux,iommu-off", NULL) != NULL) @@ -1230,11 +1230,8 @@ void __init early_init_devtree(void *params) of_scan_flat_dt(early_init_dt_scan_memory, NULL); lmb_enforce_memory_limit(memory_limit); lmb_analyze(); - systemcfg->physicalMemorySize = lmb_phys_mem_size(); lmb_reserve(0, __pa(klimit)); - DBG("Phys. mem: %lx\n", systemcfg->physicalMemorySize); - /* Reserve LMB regions used by kernel, initrd, dt, etc... */ early_reserve_mem(); @@ -1753,7 +1750,7 @@ static int of_finish_dynamic_node(struct device_node *node, /* We don't support that function on PowerMac, at least * not yet */ - if (systemcfg->platform == PLATFORM_POWERMAC) + if (_machine == PLATFORM_POWERMAC) return -ENODEV; /* fix up new node's linux_phandle field */ diff --git a/arch/ppc64/kernel/prom_init.c b/arch/ppc64/kernel/prom_init.c index e4c880dab997..6375f40b23db 100644 --- a/arch/ppc64/kernel/prom_init.c +++ b/arch/ppc64/kernel/prom_init.c @@ -1934,7 +1934,8 @@ unsigned long __init prom_init(unsigned long r3, unsigned long r4, unsigned long /* * On pSeries, inform the firmware about our capabilities */ - if (RELOC(of_platform) & PLATFORM_PSERIES) + if (RELOC(of_platform) == PLATFORM_PSERIES || + RELOC(of_platform) == PLATFORM_PSERIES_LPAR) prom_send_capabilities(); /* diff --git a/arch/ppc64/kernel/vdso.c b/arch/ppc64/kernel/vdso.c index 4aacf521e3e4..1bbacac44988 100644 --- a/arch/ppc64/kernel/vdso.c +++ b/arch/ppc64/kernel/vdso.c @@ -34,6 +34,7 @@ #include #include #include +#include #include #undef DEBUG @@ -179,7 +180,7 @@ static struct page * vdso_vma_nopage(struct vm_area_struct * vma, * Last page is systemcfg. */ if ((vma->vm_end - address) <= PAGE_SIZE) - pg = virt_to_page(systemcfg); + pg = virt_to_page(_systemcfg); else pg = virt_to_page(vbase + offset); @@ -604,7 +605,7 @@ void __init vdso_init(void) get_page(pg); } - get_page(virt_to_page(systemcfg)); + get_page(virt_to_page(_systemcfg)); } int in_gate_area_no_task(unsigned long addr) diff --git a/include/asm-powerpc/firmware.h b/include/asm-powerpc/firmware.h index 806c142ae9ea..12fabbcb04f0 100644 --- a/include/asm-powerpc/firmware.h +++ b/include/asm-powerpc/firmware.h @@ -43,6 +43,7 @@ #define FW_FEATURE_ISERIES (1UL<<21) enum { +#ifdef CONFIG_PPC64 FW_FEATURE_PSERIES_POSSIBLE = FW_FEATURE_PFT | FW_FEATURE_TCE | FW_FEATURE_SPRG0 | FW_FEATURE_DABR | FW_FEATURE_COPY | FW_FEATURE_ASR | FW_FEATURE_DEBUG | FW_FEATURE_TERM | @@ -70,6 +71,11 @@ enum { FW_FEATURE_ISERIES_ALWAYS & #endif FW_FEATURE_POSSIBLE, + +#else /* CONFIG_PPC64 */ + FW_FEATURE_POSSIBLE = 0, + FW_FEATURE_ALWAYS = 0, +#endif }; /* This is used to identify firmware features which are available diff --git a/include/asm-powerpc/processor.h b/include/asm-powerpc/processor.h index 1dc4bf7b52b3..82e3f4e6550d 100644 --- a/include/asm-powerpc/processor.h +++ b/include/asm-powerpc/processor.h @@ -17,65 +17,74 @@ #include #include #include -#ifdef CONFIG_PPC64 -#include -#endif -#ifdef CONFIG_PPC32 -/* 32-bit platform types */ -/* We only need to define a new _MACH_xxx for machines which are part of - * a configuration which supports more than one type of different machine. - * This is currently limited to CONFIG_PPC_MULTIPLATFORM and CHRP/PReP/PMac. - * -- Tom +/* We do _not_ want to define new machine types at all, those must die + * in favor of using the device-tree + * -- BenH. */ -#define _MACH_prep 0x00000001 -#define _MACH_Pmac 0x00000002 /* pmac or pmac clone (non-chrp) */ -#define _MACH_chrp 0x00000004 /* chrp machine */ -/* see residual.h for these */ +/* Platforms codes (to be obsoleted) */ +#define PLATFORM_PSERIES 0x0100 +#define PLATFORM_PSERIES_LPAR 0x0101 +#define PLATFORM_ISERIES_LPAR 0x0201 +#define PLATFORM_LPAR 0x0001 +#define PLATFORM_POWERMAC 0x0400 +#define PLATFORM_MAPLE 0x0500 +#define PLATFORM_PREP 0x0600 +#define PLATFORM_CHRP 0x0700 +#define PLATFORM_CELL 0x1000 + +/* Compat platform codes for 32 bits */ +#define _MACH_prep PLATFORM_PREP +#define _MACH_Pmac PLATFORM_POWERMAC +#define _MACH_chrp PLATFORM_CHRP + +/* PREP sub-platform types see residual.h for these */ #define _PREP_Motorola 0x01 /* motorola prep */ #define _PREP_Firm 0x02 /* firmworks prep */ #define _PREP_IBM 0x00 /* ibm prep */ #define _PREP_Bull 0x03 /* bull prep */ -/* these are arbitrary */ +/* CHRP sub-platform types. These are arbitrary */ #define _CHRP_Motorola 0x04 /* motorola chrp, the cobra */ #define _CHRP_IBM 0x05 /* IBM chrp, the longtrail and longtrail 2 */ #define _CHRP_Pegasos 0x06 /* Genesi/bplan's Pegasos and Pegasos2 */ -#ifdef CONFIG_PPC_MULTIPLATFORM +#define platform_is_pseries() (_machine == PLATFORM_PSERIES || \ + _machine == PLATFORM_PSERIES_LPAR) +#define platform_is_lpar() (!!(_machine & PLATFORM_LPAR)) + +#if defined(CONFIG_PPC_MULTIPLATFORM) extern int _machine; +#ifdef CONFIG_PPC32 + /* what kind of prep workstation we are */ extern int _prep_type; extern int _chrp_type; /* * This is used to identify the board type from a given PReP board - * vendor. Board revision is also made available. + * vendor. Board revision is also made available. This will be moved + * elsewhere soon */ extern unsigned char ucSystemType; extern unsigned char ucBoardRev; extern unsigned char ucBoardRevMaj, ucBoardRevMin; + +#endif /* CONFIG_PPC32 */ + +#elif defined(CONFIG_PPC_ISERIES) +/* + * iSeries is soon to become MULTIPLATFORM hopefully ... + */ +#define _machine CONFIG_PPC_ISERIES_LPAR #else #define _machine 0 #endif /* CONFIG_PPC_MULTIPLATFORM */ -#endif /* CONFIG_PPC32 */ -#ifdef CONFIG_PPC64 -/* Platforms supported by PPC64 */ -#define PLATFORM_PSERIES 0x0100 -#define PLATFORM_PSERIES_LPAR 0x0101 -#define PLATFORM_ISERIES_LPAR 0x0201 -#define PLATFORM_LPAR 0x0001 -#define PLATFORM_POWERMAC 0x0400 -#define PLATFORM_MAPLE 0x0500 -#define PLATFORM_CELL 0x1000 -/* Compatibility with drivers coming from PPC32 world */ -#define _machine (systemcfg->platform) -#define _MACH_Pmac PLATFORM_POWERMAC -#endif + /* * Default implementation of macro that returns current diff --git a/include/asm-powerpc/reg.h b/include/asm-powerpc/reg.h index ef121f4f0bab..eb392d038ed7 100644 --- a/include/asm-powerpc/reg.h +++ b/include/asm-powerpc/reg.h @@ -363,6 +363,7 @@ #define SPRN_RPA 0x3D6 /* Required Physical Address Register */ #define SPRN_SDA 0x3BF /* Sampled Data Address Register */ #define SPRN_SDR1 0x019 /* MMU Hash Base Register */ +#define SPRN_ASR 0x118 /* Address Space Register */ #define SPRN_SIA 0x3BB /* Sampled Instruction Address Register */ #define SPRN_SPRG0 0x110 /* Special Purpose Register General 0 */ #define SPRN_SPRG1 0x111 /* Special Purpose Register General 1 */ diff --git a/include/asm-ppc64/systemcfg.h b/include/asm-powerpc/systemcfg.h similarity index 96% rename from include/asm-ppc64/systemcfg.h rename to include/asm-powerpc/systemcfg.h index 9b86b53129aa..36b5cbe466f1 100644 --- a/include/asm-ppc64/systemcfg.h +++ b/include/asm-powerpc/systemcfg.h @@ -1,7 +1,7 @@ #ifndef _SYSTEMCFG_H #define _SYSTEMCFG_H -/* +/* * Copyright (C) 2002 Peter Bergner , IBM * * This program is free software; you can redistribute it and/or @@ -12,7 +12,7 @@ /* Change Activity: * 2002/09/30 : bergner : Created - * End Change Activity + * End Change Activity */ /* @@ -56,7 +56,7 @@ struct systemcfg { }; #ifdef __KERNEL__ -extern struct systemcfg *systemcfg; +extern struct systemcfg *_systemcfg; /* to be renamed */ #endif #endif /* __ASSEMBLY__ */ diff --git a/include/asm-ppc64/mmu.h b/include/asm-ppc64/mmu.h index d50997bace1c..c43d512d5902 100644 --- a/include/asm-ppc64/mmu.h +++ b/include/asm-ppc64/mmu.h @@ -224,6 +224,8 @@ extern int htab_bolt_mapping(unsigned long vstart, unsigned long vend, unsigned long pstart, unsigned long mode, int psize); +extern void htab_initialize(void); +extern void htab_initialize_secondary(void); extern void hpte_init_native(void); extern void hpte_init_lpar(void); extern void hpte_init_iSeries(void); @@ -245,6 +247,7 @@ extern long iSeries_hpte_insert(unsigned long hpte_group, extern void stabs_alloc(void); extern void slb_initialize(void); +extern void stab_initialize(unsigned long stab); #endif /* __ASSEMBLY__ */ From 3db9aaaf87b77eb9669157c723f6cb7e234bcef0 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 14:23:18 +1100 Subject: [PATCH 068/121] powerpc: Fix typo introduced in merging platform codes Signed-off-by: Paul Mackerras --- include/asm-powerpc/processor.h | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/include/asm-powerpc/processor.h b/include/asm-powerpc/processor.h index 82e3f4e6550d..f6f186b56b0f 100644 --- a/include/asm-powerpc/processor.h +++ b/include/asm-powerpc/processor.h @@ -78,14 +78,11 @@ extern unsigned char ucBoardRevMaj, ucBoardRevMin; /* * iSeries is soon to become MULTIPLATFORM hopefully ... */ -#define _machine CONFIG_PPC_ISERIES_LPAR +#define _machine PLATFORM_ISERIES_LPAR #else #define _machine 0 #endif /* CONFIG_PPC_MULTIPLATFORM */ - - - /* * Default implementation of macro that returns current * instruction pointer ("program counter"). From 0a5cab42a1317326d87b0d074df50705a0c3fa77 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 14:23:54 +1100 Subject: [PATCH 069/121] powerpc: Fix compile error in EEH code with gcc4 Gcc 4 doesn't like being told to inline a recursive function... Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/pseries/eeh.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c index a33ba833109c..3a857b22aa1e 100644 --- a/arch/powerpc/platforms/pseries/eeh.c +++ b/arch/powerpc/platforms/pseries/eeh.c @@ -478,7 +478,7 @@ static struct device_node * find_device_pe(struct device_node *dn) * an interrupt context, which is bad. */ -static inline void __eeh_mark_slot (struct device_node *dn) +static void __eeh_mark_slot (struct device_node *dn) { while (dn) { PCI_DN(dn)->eeh_mode |= EEH_MODE_ISOLATED; @@ -489,7 +489,7 @@ static inline void __eeh_mark_slot (struct device_node *dn) } } -static inline void __eeh_clear_slot (struct device_node *dn) +static void __eeh_clear_slot (struct device_node *dn) { while (dn) { PCI_DN(dn)->eeh_mode &= ~EEH_MODE_ISOLATED; From 094fe2e712f38f49bf79ef93306c61b1b993b07b Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 14:26:12 +1100 Subject: [PATCH 070/121] powerpc: Fixes for 32-bit powermac SMP A couple of bugs crept in with the merge of smp.c... Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/smp.c | 8 +++-- arch/powerpc/platforms/powermac/smp.c | 51 ++++++++++++++++----------- 2 files changed, 37 insertions(+), 22 deletions(-) diff --git a/arch/powerpc/kernel/smp.c b/arch/powerpc/kernel/smp.c index 7fd530898bd1..2ffdc863bff3 100644 --- a/arch/powerpc/kernel/smp.c +++ b/arch/powerpc/kernel/smp.c @@ -369,11 +369,11 @@ int generic_cpu_disable(void) if (cpu == boot_cpuid) return -EBUSY; + cpu_clear(cpu, cpu_online_map); #ifdef CONFIG_PPC64 _systemcfg->processorCount--; -#endif - cpu_clear(cpu, cpu_online_map); fixup_irqs(cpu_online_map); +#endif return 0; } @@ -391,9 +391,11 @@ int generic_cpu_enable(unsigned int cpu) while (!cpu_online(cpu)) cpu_relax(); +#ifdef CONFIG_PPC64 fixup_irqs(cpu_online_map); /* counter the irq disable in fixup_irqs */ local_irq_enable(); +#endif return 0; } @@ -422,7 +424,9 @@ void generic_mach_cpu_die(void) while (__get_cpu_var(cpu_state) != CPU_UP_PREPARE) cpu_relax(); +#ifdef CONFIG_PPC64 flush_tlb_pending(); +#endif cpu_set(cpu, cpu_online_map); local_irq_enable(); } diff --git a/arch/powerpc/platforms/powermac/smp.c b/arch/powerpc/platforms/powermac/smp.c index e1f9443cc872..957b09103422 100644 --- a/arch/powerpc/platforms/powermac/smp.c +++ b/arch/powerpc/platforms/powermac/smp.c @@ -305,9 +305,19 @@ static int __init smp_psurge_probe(void) psurge_start = ioremap(PSURGE_START, 4); psurge_pri_intr = ioremap(PSURGE_PRI_INTR, 4); - /* this is not actually strictly necessary -- paulus. */ - for (i = 1; i < ncpus; ++i) - smp_hw_index[i] = i; + /* + * This is necessary because OF doesn't know about the + * secondary cpu(s), and thus there aren't nodes in the + * device tree for them, and smp_setup_cpu_maps hasn't + * set their bits in cpu_possible_map and cpu_present_map. + */ + if (ncpus > NR_CPUS) + ncpus = NR_CPUS; + for (i = 1; i < ncpus ; ++i) { + cpu_set(i, cpu_present_map); + cpu_set(i, cpu_possible_map); + set_hard_smp_processor_id(i, i); + } if (ppc_md.progress) ppc_md.progress("smp_psurge_probe - done", 0x352); @@ -348,6 +358,7 @@ static void __init psurge_dual_sync_tb(int cpu_nr) int t; set_dec(tb_ticks_per_jiffy); + /* XXX fixme */ set_tb(0, 0); last_jiffy_stamp(cpu_nr) = 0; @@ -363,8 +374,6 @@ static void __init psurge_dual_sync_tb(int cpu_nr) /* now interrupt the secondary, starting both TBs */ psurge_set_ipi(1); - - smp_tb_synchronized = 1; } static struct irqaction psurge_irqaction = { @@ -625,9 +634,8 @@ void smp_core99_give_timebase(void) for (t = 100000; t > 0 && sec_tb_reset; --t) udelay(10); if (sec_tb_reset) + /* XXX BUG_ON here? */ printk(KERN_WARNING "Timeout waiting sync(2) on second CPU\n"); - else - smp_tb_synchronized = 1; /* Now, restart the timebase by leaving the GPIO to an open collector */ pmac_call_feature(PMAC_FTR_WRITE_GPIO, NULL, core99_tb_gpio, 0); @@ -810,19 +818,9 @@ static void __devinit smp_core99_setup_cpu(int cpu_nr) } -/* Core99 Macs (dual G4s and G5s) */ -struct smp_ops_t core99_smp_ops = { - .message_pass = smp_mpic_message_pass, - .probe = smp_core99_probe, - .kick_cpu = smp_core99_kick_cpu, - .setup_cpu = smp_core99_setup_cpu, - .give_timebase = smp_core99_give_timebase, - .take_timebase = smp_core99_take_timebase, -}; - #if defined(CONFIG_HOTPLUG_CPU) && defined(CONFIG_PPC32) -int __cpu_disable(void) +int smp_core99_cpu_disable(void) { cpu_clear(smp_processor_id(), cpu_online_map); @@ -846,7 +844,7 @@ void cpu_die(void) low_cpu_die(); } -void __cpu_die(unsigned int cpu) +void smp_core99_cpu_die(unsigned int cpu) { int timeout; @@ -858,8 +856,21 @@ void __cpu_die(unsigned int cpu) } msleep(1); } - cpu_callin_map[cpu] = 0; cpu_dead[cpu] = 0; } #endif + +/* Core99 Macs (dual G4s and G5s) */ +struct smp_ops_t core99_smp_ops = { + .message_pass = smp_mpic_message_pass, + .probe = smp_core99_probe, + .kick_cpu = smp_core99_kick_cpu, + .setup_cpu = smp_core99_setup_cpu, + .give_timebase = smp_core99_give_timebase, + .take_timebase = smp_core99_take_timebase, +#if defined(CONFIG_HOTPLUG_CPU) && defined(CONFIG_PPC32) + .cpu_disable = smp_core99_cpu_disable, + .cpu_die = smp_core99_cpu_die, +#endif +}; From cbe62e2b4a764aa3f9a2b9c9004f8e86a1f0ade8 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 14:28:03 +1100 Subject: [PATCH 071/121] powerpc: Fix SMP time initialization problem We were getting the last_jiffy per-cpu variable set ahead of the current timebase in smp_space_timers on SMP machines. This caused the loop in timer_interrupt to loop virtually forever, since tb_ticks_since assumes that it will never be called with the timebase behind the last_jiffy value. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/time.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c index 386006b3d616..260b6ecd26a9 100644 --- a/arch/powerpc/kernel/time.c +++ b/arch/powerpc/kernel/time.c @@ -484,6 +484,8 @@ void __init smp_space_timers(unsigned int max_cpus) unsigned long offset = tb_ticks_per_jiffy / max_cpus; unsigned long previous_tb = per_cpu(last_jiffy, boot_cpuid); + /* make sure tb > per_cpu(last_jiffy, cpu) for all cpus always */ + previous_tb -= tb_ticks_per_jiffy; for_each_cpu(i) { if (i != boot_cpuid) { previous_tb += offset; From a7ddc5e85351931b67a48afa22788d77763837d8 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 14:29:18 +1100 Subject: [PATCH 072/121] powerpc: Add user CPU features for POWER4, POWER5, POWER5+ and Cell. This is at the request of the glibc folks, who want to use these bits to select libraries optimized for the microarchitecture and new instructions in these processors. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/cputable.c | 19 +++++++++++-------- include/asm-powerpc/cputable.h | 4 ++++ 2 files changed, 15 insertions(+), 8 deletions(-) diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c index cc4e9eb1c13f..1d85cedbbb7b 100644 --- a/arch/powerpc/kernel/cputable.c +++ b/arch/powerpc/kernel/cputable.c @@ -52,6 +52,9 @@ extern void __setup_cpu_ppc970(unsigned long offset, struct cpu_spec* spec); #define COMMON_USER (PPC_FEATURE_32 | PPC_FEATURE_HAS_FPU | \ PPC_FEATURE_HAS_MMU) #define COMMON_USER_PPC64 (COMMON_USER | PPC_FEATURE_64) +#define COMMON_USER_POWER4 (COMMON_USER_PPC64 | PPC_FEATURE_POWER4) +#define COMMON_USER_POWER5 (COMMON_USER_PPC64 | PPC_FEATURE_POWER5) +#define COMMON_USER_POWER5_PLUS (COMMON_USER_PPC64 | PPC_FEATURE_POWER5_PLUS) /* We only set the spe features if the kernel was compiled with @@ -160,7 +163,7 @@ struct cpu_spec cpu_specs[] = { .pvr_value = 0x00350000, .cpu_name = "POWER4 (gp)", .cpu_features = CPU_FTRS_POWER4, - .cpu_user_features = COMMON_USER_PPC64, + .cpu_user_features = COMMON_USER_POWER4, .icache_bsize = 128, .dcache_bsize = 128, .num_pmcs = 8, @@ -175,7 +178,7 @@ struct cpu_spec cpu_specs[] = { .pvr_value = 0x00380000, .cpu_name = "POWER4+ (gq)", .cpu_features = CPU_FTRS_POWER4, - .cpu_user_features = COMMON_USER_PPC64, + .cpu_user_features = COMMON_USER_POWER4, .icache_bsize = 128, .dcache_bsize = 128, .num_pmcs = 8, @@ -190,7 +193,7 @@ struct cpu_spec cpu_specs[] = { .pvr_value = 0x00390000, .cpu_name = "PPC970", .cpu_features = CPU_FTRS_PPC970, - .cpu_user_features = COMMON_USER_PPC64 | + .cpu_user_features = COMMON_USER_POWER4 | PPC_FEATURE_HAS_ALTIVEC_COMP, .icache_bsize = 128, .dcache_bsize = 128, @@ -212,7 +215,7 @@ struct cpu_spec cpu_specs[] = { #else .cpu_features = CPU_FTRS_PPC970, #endif - .cpu_user_features = COMMON_USER_PPC64 | + .cpu_user_features = COMMON_USER_POWER4 | PPC_FEATURE_HAS_ALTIVEC_COMP, .icache_bsize = 128, .dcache_bsize = 128, @@ -230,7 +233,7 @@ struct cpu_spec cpu_specs[] = { .pvr_value = 0x00440000, .cpu_name = "PPC970MP", .cpu_features = CPU_FTRS_PPC970, - .cpu_user_features = COMMON_USER_PPC64 | + .cpu_user_features = COMMON_USER_POWER4 | PPC_FEATURE_HAS_ALTIVEC_COMP, .icache_bsize = 128, .dcache_bsize = 128, @@ -245,7 +248,7 @@ struct cpu_spec cpu_specs[] = { .pvr_value = 0x003a0000, .cpu_name = "POWER5 (gr)", .cpu_features = CPU_FTRS_POWER5, - .cpu_user_features = COMMON_USER_PPC64, + .cpu_user_features = COMMON_USER_POWER5, .icache_bsize = 128, .dcache_bsize = 128, .num_pmcs = 6, @@ -260,7 +263,7 @@ struct cpu_spec cpu_specs[] = { .pvr_value = 0x003b0000, .cpu_name = "POWER5 (gs)", .cpu_features = CPU_FTRS_POWER5, - .cpu_user_features = COMMON_USER_PPC64, + .cpu_user_features = COMMON_USER_POWER5_PLUS, .icache_bsize = 128, .dcache_bsize = 128, .num_pmcs = 6, @@ -276,7 +279,7 @@ struct cpu_spec cpu_specs[] = { .cpu_name = "Cell Broadband Engine", .cpu_features = CPU_FTRS_CELL, .cpu_user_features = COMMON_USER_PPC64 | - PPC_FEATURE_HAS_ALTIVEC_COMP, + PPC_FEATURE_CELL | PPC_FEATURE_HAS_ALTIVEC_COMP, .icache_bsize = 128, .dcache_bsize = 128, .cpu_setup = __setup_cpu_be, diff --git a/include/asm-powerpc/cputable.h b/include/asm-powerpc/cputable.h index f89fd883e892..04e2726002cf 100644 --- a/include/asm-powerpc/cputable.h +++ b/include/asm-powerpc/cputable.h @@ -16,6 +16,10 @@ #define PPC_FEATURE_HAS_EFP_SINGLE 0x00400000 #define PPC_FEATURE_HAS_EFP_DOUBLE 0x00200000 #define PPC_FEATURE_NO_TB 0x00100000 +#define PPC_FEATURE_POWER4 0x00080000 +#define PPC_FEATURE_POWER5 0x00040000 +#define PPC_FEATURE_POWER5_PLUS 0x00020000 +#define PPC_FEATURE_CELL 0x00010000 #ifdef __KERNEL__ #ifndef __ASSEMBLY__ From e1449ed956ae29129bde3e5137dde1d579d585ff Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 14:30:20 +1100 Subject: [PATCH 073/121] powerpc: 32-bit fixes for xmon This makes the memory examine/change command print the address as 8 digits instead of 16, and makes the memory dump command print 4 4-byte values per line instead of 2 8-byte values. Signed-off-by: Paul Mackerras --- arch/powerpc/xmon/xmon.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/arch/powerpc/xmon/xmon.c b/arch/powerpc/xmon/xmon.c index b43a57425ea9..cfcb2a56d662 100644 --- a/arch/powerpc/xmon/xmon.c +++ b/arch/powerpc/xmon/xmon.c @@ -1792,7 +1792,7 @@ memex(void) for(;;){ if (!mnoread) n = mread(adrs, val, size); - printf("%.16x%c", adrs, brev? 'r': ' '); + printf(REG"%c", adrs, brev? 'r': ' '); if (!mnoread) { if (brev) byterev(val, size); @@ -1971,17 +1971,18 @@ prdump(unsigned long adrs, long ndump) nr = mread(adrs, temp, r); adrs += nr; for (m = 0; m < r; ++m) { - if ((m & 7) == 0 && m > 0) - putchar(' '); + if ((m & (sizeof(long) - 1)) == 0 && m > 0) + putchar(' '); if (m < nr) printf("%.2x", temp[m]); else printf("%s", fault_chars[fault_type]); } - if (m <= 8) - printf(" "); - for (; m < 16; ++m) + for (; m < 16; ++m) { + if ((m & (sizeof(long) - 1)) == 0) + putchar(' '); printf(" "); + } printf(" |"); for (m = 0; m < r; ++m) { if (m < nr) { From 940e3318c36394939d805e797d7be39ddaaa7911 Mon Sep 17 00:00:00 2001 From: Trond Myklebust Date: Wed, 9 Nov 2005 21:45:24 -0500 Subject: [PATCH 074/121] [PATCH] SUNRPC: don't reencode when looping in call transmit. If the call to xprt_transmit() fails due to socket buffer space exhaustion, we do not need to re-encode the RPC message when we loop back through call_transmit. Re-encoding can actually end up triggering the WARN_ON() in call_decode() if we re-encode something like a read() request and auth->au_rslack has changed. It can also cause us to increment the RPCSEC_GSS sequence number beyond the limits of the allowed window. Signed-off-by: Trond Myklebust Signed-off-by: Linus Torvalds --- net/sunrpc/clnt.c | 32 +++++++++++++++++++++++++++++--- 1 file changed, 29 insertions(+), 3 deletions(-) diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c index 702ede309b06..61c3abeaccae 100644 --- a/net/sunrpc/clnt.c +++ b/net/sunrpc/clnt.c @@ -55,6 +55,7 @@ static void call_bind(struct rpc_task *task); static void call_bind_status(struct rpc_task *task); static void call_transmit(struct rpc_task *task); static void call_status(struct rpc_task *task); +static void call_transmit_status(struct rpc_task *task); static void call_refresh(struct rpc_task *task); static void call_refreshresult(struct rpc_task *task); static void call_timeout(struct rpc_task *task); @@ -672,6 +673,18 @@ call_allocate(struct rpc_task *task) rpc_exit(task, -ERESTARTSYS); } +static inline int +rpc_task_need_encode(struct rpc_task *task) +{ + return task->tk_rqstp->rq_snd_buf.len == 0; +} + +static inline void +rpc_task_force_reencode(struct rpc_task *task) +{ + task->tk_rqstp->rq_snd_buf.len = 0; +} + /* * 3. Encode arguments of an RPC call */ @@ -867,12 +880,14 @@ call_transmit(struct rpc_task *task) if (task->tk_status != 0) return; /* Encode here so that rpcsec_gss can use correct sequence number. */ - if (task->tk_rqstp->rq_bytes_sent == 0) { + if (rpc_task_need_encode(task)) { + task->tk_rqstp->rq_bytes_sent = 0; call_encode(task); /* Did the encode result in an error condition? */ if (task->tk_status != 0) goto out_nosend; } + task->tk_action = call_transmit_status; xprt_transmit(task); if (task->tk_status < 0) return; @@ -884,6 +899,7 @@ call_transmit(struct rpc_task *task) out_nosend: /* release socket write lock before attempting to handle error */ xprt_abort_transmit(task); + rpc_task_force_reencode(task); } /* @@ -915,7 +931,6 @@ call_status(struct rpc_task *task) break; case -ECONNREFUSED: case -ENOTCONN: - req->rq_bytes_sent = 0; if (clnt->cl_autobind) clnt->cl_port = 0; task->tk_action = call_bind; @@ -937,7 +952,18 @@ call_status(struct rpc_task *task) } /* - * 6a. Handle RPC timeout + * 6a. Handle transmission errors. + */ +static void +call_transmit_status(struct rpc_task *task) +{ + if (task->tk_status != -EAGAIN) + rpc_task_force_reencode(task); + call_status(task); +} + +/* + * 6b. Handle RPC timeout * We do not release the request slot, so we keep using the * same XID for all retransmits. */ From 676e2497f87501fb190a5b06103d396ca9147373 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Thu, 10 Nov 2005 14:16:21 +1100 Subject: [PATCH 075/121] powerpc: remove some warnings when building iSeries Signed-off-by: Stephen Rothwell --- arch/powerpc/kernel/prom.c | 6 +++--- arch/powerpc/platforms/iseries/setup.c | 1 - 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c index 1bf3642cb859..fd3bcb4a9e9f 100644 --- a/arch/powerpc/kernel/prom.c +++ b/arch/powerpc/kernel/prom.c @@ -1084,9 +1084,9 @@ void __init unflatten_device_tree(void) static int __init early_init_dt_scan_cpus(unsigned long node, const char *uname, int depth, void *data) { - char *type = of_get_flat_dt_prop(node, "device_type", NULL); u32 *prop; - unsigned long size = 0; + unsigned long size; + char *type = of_get_flat_dt_prop(node, "device_type", &size); /* We are scanning "cpu" nodes only */ if (type == NULL || strcmp(type, "cpu") != 0) @@ -1112,7 +1112,7 @@ static int __init early_init_dt_scan_cpus(unsigned long node, #ifdef CONFIG_ALTIVEC /* Check if we have a VMX and eventually update CPU features */ - prop = (u32 *)of_get_flat_dt_prop(node, "ibm,vmx", &size); + prop = (u32 *)of_get_flat_dt_prop(node, "ibm,vmx", NULL); if (prop && (*prop) > 0) { cur_cpu_spec->cpu_features |= CPU_FTR_ALTIVEC; cur_cpu_spec->cpu_user_features |= PPC_FEATURE_HAS_ALTIVEC; diff --git a/arch/powerpc/platforms/iseries/setup.c b/arch/powerpc/platforms/iseries/setup.c index 623c39aa0436..19b415f6e5d5 100644 --- a/arch/powerpc/platforms/iseries/setup.c +++ b/arch/powerpc/platforms/iseries/setup.c @@ -705,7 +705,6 @@ static void iseries_shared_idle(void) static void iseries_dedicated_idle(void) { - long oldval; set_thread_flag(TIF_POLLING_NRFLAG); while (1) { From 87655ff26817993932b7d049c4df226fb2c0ac5f Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 10 Nov 2005 14:53:16 +1100 Subject: [PATCH 076/121] [PATCH] powerpc: 64k pages pmd alloc fix This patch makes the kernel use a different kmem cache for PMD pages as they are smaller than PTE pages. Avoids waste of memory. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Paul Mackerras --- arch/powerpc/mm/init_64.c | 18 ++++++++++++++---- include/asm-ppc64/pgalloc.h | 4 ++-- 2 files changed, 16 insertions(+), 6 deletions(-) diff --git a/arch/powerpc/mm/init_64.c b/arch/powerpc/mm/init_64.c index ce974c83d88a..e274cf10205c 100644 --- a/arch/powerpc/mm/init_64.c +++ b/arch/powerpc/mm/init_64.c @@ -20,6 +20,8 @@ * */ +#undef DEBUG + #include #include #include @@ -64,6 +66,12 @@ #include #include +#ifdef DEBUG +#define DBG(fmt...) printk(fmt) +#else +#define DBG(fmt...) +#endif + #if PGTABLE_RANGE > USER_VSID_RANGE #warning Limited user VSID range means pagetable space is wasted #endif @@ -188,14 +196,14 @@ static void zero_ctor(void *addr, kmem_cache_t *cache, unsigned long flags) } #ifdef CONFIG_PPC_64K_PAGES -static const int pgtable_cache_size[2] = { - PTE_TABLE_SIZE, PGD_TABLE_SIZE +static const unsigned int pgtable_cache_size[3] = { + PTE_TABLE_SIZE, PMD_TABLE_SIZE, PGD_TABLE_SIZE }; static const char *pgtable_cache_name[ARRAY_SIZE(pgtable_cache_size)] = { - "pte_pmd_cache", "pgd_cache", + "pte_pmd_cache", "pmd_cache", "pgd_cache", }; #else -static const int pgtable_cache_size[2] = { +static const unsigned int pgtable_cache_size[2] = { PTE_TABLE_SIZE, PMD_TABLE_SIZE }; static const char *pgtable_cache_name[ARRAY_SIZE(pgtable_cache_size)] = { @@ -213,6 +221,8 @@ void pgtable_cache_init(void) int size = pgtable_cache_size[i]; const char *name = pgtable_cache_name[i]; + DBG("Allocating page table cache %s (#%d) " + "for size: %08x...\n", name, i, size); pgtable_cache[i] = kmem_cache_create(name, size, size, SLAB_HWCACHE_ALIGN | diff --git a/include/asm-ppc64/pgalloc.h b/include/asm-ppc64/pgalloc.h index 98da0e4262bd..dcf3622d1946 100644 --- a/include/asm-ppc64/pgalloc.h +++ b/include/asm-ppc64/pgalloc.h @@ -10,8 +10,8 @@ extern kmem_cache_t *pgtable_cache[]; #ifdef CONFIG_PPC_64K_PAGES #define PTE_CACHE_NUM 0 -#define PMD_CACHE_NUM 0 -#define PGD_CACHE_NUM 1 +#define PMD_CACHE_NUM 1 +#define PGD_CACHE_NUM 2 #else #define PTE_CACHE_NUM 0 #define PMD_CACHE_NUM 1 From 47c2ac8cc3cf7c484e131d10cb812b6ab198885a Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 10 Nov 2005 14:54:58 +1100 Subject: [PATCH 077/121] [PATCH] powerpc: 64k pages vs. U3 iommu That DART (U3 iommu) code didn't properly scale the number of entries when using !4k pages. That caused crashes when booting G5s with more than 2Gb of RAM. This fixes it. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Paul Mackerras --- arch/powerpc/sysdev/u3_iommu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/sysdev/u3_iommu.c b/arch/powerpc/sysdev/u3_iommu.c index 543d65909812..f32baf7f4693 100644 --- a/arch/powerpc/sysdev/u3_iommu.c +++ b/arch/powerpc/sysdev/u3_iommu.c @@ -226,7 +226,7 @@ static void iommu_table_u3_setup(void) iommu_table_u3.it_busno = 0; iommu_table_u3.it_offset = 0; /* it_size is in number of entries */ - iommu_table_u3.it_size = dart_tablesize / sizeof(u32); + iommu_table_u3.it_size = (dart_tablesize / sizeof(u32)) >> DART_PAGE_FACTOR; /* Initialize the common IOMMU code */ iommu_table_u3.it_base = (unsigned long)dart_vbase; From 0f34f490290e05ee6c02e20b3811fce6f09318f4 Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 10 Nov 2005 15:04:24 +1100 Subject: [PATCH 078/121] [PATCH] ppc64: fix PCI IO mapping phbs_remap_io(), which maps the PCI IO space into the kernel virtual space, is called too early on powermac, and thus doesn't work. This fixes it by removing the call from all platforms and putting it back into the ppc64 common code where it belongs, after the actual probing of the bus. That means that before that call, only the ISA IO space (if any) is mapped, any PIO access (from quirks for example) will fail. This happens not to be a problem for now, but we'll have to rework that code if it becomes one in the future. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/maple/pci.c | 3 --- arch/powerpc/platforms/powermac/pci.c | 3 --- arch/powerpc/platforms/pseries/pci.c | 1 - arch/ppc64/kernel/pci.c | 3 +++ 4 files changed, 3 insertions(+), 7 deletions(-) diff --git a/arch/powerpc/platforms/maple/pci.c b/arch/powerpc/platforms/maple/pci.c index 340c21caeae2..895aeb3f75d0 100644 --- a/arch/powerpc/platforms/maple/pci.c +++ b/arch/powerpc/platforms/maple/pci.c @@ -380,9 +380,6 @@ void __init maple_pcibios_fixup(void) for_each_pci_dev(dev) pci_read_irq_line(dev); - /* Do the mapping of the IO space */ - phbs_remap_io(); - DBG(" <- maple_pcibios_fixup\n"); } diff --git a/arch/powerpc/platforms/powermac/pci.c b/arch/powerpc/platforms/powermac/pci.c index 8f818d092e2b..dfd41b9781a9 100644 --- a/arch/powerpc/platforms/powermac/pci.c +++ b/arch/powerpc/platforms/powermac/pci.c @@ -918,9 +918,6 @@ void __init pmac_pci_init(void) PCI_DN(np)->busno = 0xf0; } - /* map in PCI I/O space */ - phbs_remap_io(); - /* pmac_check_ht_link(); */ /* Tell pci.c to not use the common resource allocation mechanism */ diff --git a/arch/powerpc/platforms/pseries/pci.c b/arch/powerpc/platforms/pseries/pci.c index 6b0772f35692..999a9620b5ce 100644 --- a/arch/powerpc/platforms/pseries/pci.c +++ b/arch/powerpc/platforms/pseries/pci.c @@ -107,7 +107,6 @@ static void __init pSeries_request_regions(void) void __init pSeries_final_fixup(void) { - phbs_remap_io(); pSeries_request_regions(); pci_addr_cache_build(); diff --git a/arch/ppc64/kernel/pci.c b/arch/ppc64/kernel/pci.c index 66698fdf059e..46929fb24f3c 100644 --- a/arch/ppc64/kernel/pci.c +++ b/arch/ppc64/kernel/pci.c @@ -548,6 +548,9 @@ static int __init pcibios_init(void) if (ppc64_isabridge_dev != NULL) printk("ISA bridge at %s\n", pci_name(ppc64_isabridge_dev)); + /* map in PCI I/O space */ + phbs_remap_io(); + printk("PCI: Probing PCI hardware done\n"); return 0; From d3d2176a0bc696b5365ce0e8f82a29a521d85fd0 Mon Sep 17 00:00:00 2001 From: David Gibson Date: Thu, 10 Nov 2005 15:26:20 +1100 Subject: [PATCH 079/121] [PATCH] powerpc: Move more ppc64 files with no ppc32 equivalent to powerpc This patch moves a bunch more files from arch/ppc64 and include/asm-ppc64 which have no equivalents in ppc32 code into arch/powerpc and include/asm-powerpc. The file affected are: hvcall.h proc_ppc64.c sysfs.c lparcfg.c rtas_pci.c The only changes apart from the move and corresponding Makefile changes are: - #ifndef/#define in includes updated to _ASM_POWERPC_ form - trailing whitespace removed - comments giving full paths removed Built and booted on POWER5 LPAR (ARCH=powerpc and ARCH=ppc64), built for 32-bit powermac (ARCH=powerpc). Signed-off-by: David Gibson Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/Makefile | 8 ++++++-- arch/{ppc64 => powerpc}/kernel/lparcfg.c | 10 +++++----- arch/{ppc64 => powerpc}/kernel/proc_ppc64.c | 8 +++----- arch/{ppc64 => powerpc}/kernel/rtas_pci.c | 0 arch/{ppc64 => powerpc}/kernel/sysfs.c | 2 +- arch/ppc64/kernel/Makefile | 5 +---- include/{asm-ppc64 => asm-powerpc}/hvcall.h | 14 +++++++------- 7 files changed, 23 insertions(+), 24 deletions(-) rename arch/{ppc64 => powerpc}/kernel/lparcfg.c (99%) rename arch/{ppc64 => powerpc}/kernel/proc_ppc64.c (98%) rename arch/{ppc64 => powerpc}/kernel/rtas_pci.c (100%) rename arch/{ppc64 => powerpc}/kernel/sysfs.c (99%) rename include/{asm-ppc64 => asm-powerpc}/hvcall.h (97%) diff --git a/arch/powerpc/kernel/Makefile b/arch/powerpc/kernel/Makefile index 92cfabf929bc..c04bbd320594 100644 --- a/arch/powerpc/kernel/Makefile +++ b/arch/powerpc/kernel/Makefile @@ -16,13 +16,17 @@ obj-y := semaphore.o cputable.o ptrace.o syscalls.o \ obj-$(CONFIG_PPC64) += setup_64.o binfmt_elf32.o sys_ppc32.o \ signal_64.o ptrace32.o systbl.o \ paca.o ioctl32.o cpu_setup_power4.o \ - firmware.o + firmware.o sysfs.o obj-$(CONFIG_ALTIVEC) += vecemu.o vector.o obj-$(CONFIG_POWER4) += idle_power4.o obj-$(CONFIG_PPC_OF) += of_device.o -obj-$(CONFIG_PPC_RTAS) += rtas.o +procfs-$(CONFIG_PPC64) := proc_ppc64.o +obj-$(CONFIG_PROC_FS) += $(procfs-y) +rtaspci-$(CONFIG_PPC64) := rtas_pci.o +obj-$(CONFIG_PPC_RTAS) += rtas.o $(rtaspci-y) obj-$(CONFIG_RTAS_FLASH) += rtas_flash.o obj-$(CONFIG_RTAS_PROC) += rtas-proc.o +obj-$(CONFIG_LPARCFG) += lparcfg.o obj-$(CONFIG_IBMVIO) += vio.o obj-$(CONFIG_GENERIC_TBSYNC) += smp-tbsync.o diff --git a/arch/ppc64/kernel/lparcfg.c b/arch/powerpc/kernel/lparcfg.c similarity index 99% rename from arch/ppc64/kernel/lparcfg.c rename to arch/powerpc/kernel/lparcfg.c index a80ed307515c..5e954fae031f 100644 --- a/arch/ppc64/kernel/lparcfg.c +++ b/arch/powerpc/kernel/lparcfg.c @@ -97,7 +97,7 @@ static unsigned long get_purr(void) #define lparcfg_write NULL -/* +/* * Methods used to fetch LPAR data when running on an iSeries platform. */ static int lparcfg_data(struct seq_file *m, void *v) @@ -169,7 +169,7 @@ static int lparcfg_data(struct seq_file *m, void *v) #endif /* CONFIG_PPC_ISERIES */ #ifdef CONFIG_PPC_PSERIES -/* +/* * Methods used to fetch LPAR data when running on a pSeries platform. */ @@ -178,7 +178,7 @@ static int lparcfg_data(struct seq_file *m, void *v) * entitled_capacity,unallocated_capacity, * aggregation, resource_capability). * - * R4 = Entitled Processor Capacity Percentage. + * R4 = Entitled Processor Capacity Percentage. * R5 = Unallocated Processor Capacity Percentage. * R6 (AABBCCDDEEFFGGHH). * XXXX - reserved (0) @@ -191,7 +191,7 @@ static int lparcfg_data(struct seq_file *m, void *v) * XX - variable processor Capacity Weight * XX - Unallocated Variable Processor Capacity Weight. * XXXX - Active processors in Physical Processor Pool. - * XXXX - Processors active on platform. + * XXXX - Processors active on platform. */ static unsigned int h_get_ppp(unsigned long *entitled, unsigned long *unallocated, @@ -274,7 +274,7 @@ static void parse_system_parameter_string(struct seq_file *m) if (!workbuffer) { printk(KERN_ERR "%s %s kmalloc failure at line %d \n", __FILE__, __FUNCTION__, __LINE__); - kfree(local_buffer); + kfree(local_buffer); return; } #ifdef LPARCFG_DEBUG diff --git a/arch/ppc64/kernel/proc_ppc64.c b/arch/powerpc/kernel/proc_ppc64.c similarity index 98% rename from arch/ppc64/kernel/proc_ppc64.c rename to arch/powerpc/kernel/proc_ppc64.c index c893a11ee198..a1c19502fe8b 100644 --- a/arch/ppc64/kernel/proc_ppc64.c +++ b/arch/powerpc/kernel/proc_ppc64.c @@ -1,18 +1,16 @@ /* - * arch/ppc64/kernel/proc_ppc64.c - * * Copyright (C) 2001 Mike Corrigan & Dave Engebretsen IBM Corporation - * + * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. - * + * * This program is distributed in the hope that it will be useful, * but WITHOUT ANY WARRANTY; without even the implied warranty of * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the * GNU General Public License for more details. - * + * * You should have received a copy of the GNU General Public License * along with this program; if not, write to the Free Software * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA diff --git a/arch/ppc64/kernel/rtas_pci.c b/arch/powerpc/kernel/rtas_pci.c similarity index 100% rename from arch/ppc64/kernel/rtas_pci.c rename to arch/powerpc/kernel/rtas_pci.c diff --git a/arch/ppc64/kernel/sysfs.c b/arch/powerpc/kernel/sysfs.c similarity index 99% rename from arch/ppc64/kernel/sysfs.c rename to arch/powerpc/kernel/sysfs.c index e99ec62c2c52..850af198fb5f 100644 --- a/arch/ppc64/kernel/sysfs.c +++ b/arch/powerpc/kernel/sysfs.c @@ -232,7 +232,7 @@ static void register_cpu_online(unsigned int cpu) sysdev_create_file(s, &attr_pmc7); if (cur_cpu_spec->num_pmcs >= 8) sysdev_create_file(s, &attr_pmc8); - + if (cpu_has_feature(CPU_FTR_SMT)) sysdev_create_file(s, &attr_purr); } diff --git a/arch/ppc64/kernel/Makefile b/arch/ppc64/kernel/Makefile index 048ba910f028..58b19f107656 100644 --- a/arch/ppc64/kernel/Makefile +++ b/arch/ppc64/kernel/Makefile @@ -15,7 +15,7 @@ obj-y += idle.o dma.o \ align.o \ udbg.o \ rtc.o \ - iommu.o sysfs.o vdso.o + iommu.o vdso.o obj-y += vdso32/ vdso64/ pci-obj-$(CONFIG_PPC_MULTIPLATFORM) += pci_dn.o pci_direct_iommu.o @@ -30,13 +30,10 @@ endif obj-$(CONFIG_PPC_PSERIES) += udbg_16550.o obj-$(CONFIG_KEXEC) += machine_kexec.o -obj-$(CONFIG_PROC_FS) += proc_ppc64.o obj-$(CONFIG_MODULES) += module.o ifneq ($(CONFIG_PPC_MERGE),y) obj-$(CONFIG_MODULES) += ppc_ksyms.o endif -obj-$(CONFIG_PPC_RTAS) += rtas_pci.o -obj-$(CONFIG_LPARCFG) += lparcfg.o obj-$(CONFIG_HVC_CONSOLE) += hvconsole.o ifneq ($(CONFIG_PPC_MERGE),y) obj-$(CONFIG_BOOTX_TEXT) += btext.o diff --git a/include/asm-ppc64/hvcall.h b/include/asm-powerpc/hvcall.h similarity index 97% rename from include/asm-ppc64/hvcall.h rename to include/asm-powerpc/hvcall.h index ab7c3cf24888..d36da61dbc53 100644 --- a/include/asm-ppc64/hvcall.h +++ b/include/asm-powerpc/hvcall.h @@ -1,5 +1,5 @@ -#ifndef _PPC64_HVCALL_H -#define _PPC64_HVCALL_H +#ifndef _ASM_POWERPC_HVCALL_H +#define _ASM_POWERPC_HVCALL_H #define HVSC .long 0x44000022 @@ -138,7 +138,7 @@ long plpar_hcall(unsigned long opcode, */ long plpar_hcall_norets(unsigned long opcode, ...); -/* +/* * Special hcall interface for ibmveth support. * Takes 8 input parms. Returns a rc and stores the * R4 return value in *out1. @@ -153,11 +153,11 @@ long plpar_hcall_8arg_2ret(unsigned long opcode, unsigned long arg7, unsigned long arg8, unsigned long *out1); - + /* plpar_hcall_4out() * - * same as plpar_hcall except with 4 output arguments. - * + * same as plpar_hcall except with 4 output arguments. + * */ long plpar_hcall_4out(unsigned long opcode, unsigned long arg1, @@ -170,4 +170,4 @@ long plpar_hcall_4out(unsigned long opcode, unsigned long *out4); #endif /* __ASSEMBLY__ */ -#endif /* _PPC64_HVCALL_H */ +#endif /* _ASM_POWERPC_HVCALL_H */ From 06a98dba0d4b4f2f9b1f35f636beb166d6cbde34 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Thu, 10 Nov 2005 15:51:14 +1100 Subject: [PATCH 080/121] powerpc: implement atomic64_t on ppc64 Signed-off-by: Stephen Rothwell --- include/asm-powerpc/atomic.h | 178 +++++++++++++++++++++++++++++++++++ 1 file changed, 178 insertions(+) diff --git a/include/asm-powerpc/atomic.h b/include/asm-powerpc/atomic.h index c5b12fd2b46b..9c0b372a46e1 100644 --- a/include/asm-powerpc/atomic.h +++ b/include/asm-powerpc/atomic.h @@ -197,5 +197,183 @@ static __inline__ int atomic_dec_if_positive(atomic_t *v) #define smp_mb__before_atomic_inc() smp_mb() #define smp_mb__after_atomic_inc() smp_mb() +#ifdef __powerpc64__ + +typedef struct { volatile long counter; } atomic64_t; + +#define ATOMIC64_INIT(i) { (i) } + +#define atomic64_read(v) ((v)->counter) +#define atomic64_set(v,i) (((v)->counter) = (i)) + +static __inline__ void atomic64_add(long a, atomic64_t *v) +{ + long t; + + __asm__ __volatile__( +"1: ldarx %0,0,%3 # atomic64_add\n\ + add %0,%2,%0\n\ + stdcx. %0,0,%3 \n\ + bne- 1b" + : "=&r" (t), "=m" (v->counter) + : "r" (a), "r" (&v->counter), "m" (v->counter) + : "cc"); +} + +static __inline__ long atomic64_add_return(long a, atomic64_t *v) +{ + long t; + + __asm__ __volatile__( + EIEIO_ON_SMP +"1: ldarx %0,0,%2 # atomic64_add_return\n\ + add %0,%1,%0\n\ + stdcx. %0,0,%2 \n\ + bne- 1b" + ISYNC_ON_SMP + : "=&r" (t) + : "r" (a), "r" (&v->counter) + : "cc", "memory"); + + return t; +} + +#define atomic64_add_negative(a, v) (atomic64_add_return((a), (v)) < 0) + +static __inline__ void atomic64_sub(long a, atomic64_t *v) +{ + long t; + + __asm__ __volatile__( +"1: ldarx %0,0,%3 # atomic64_sub\n\ + subf %0,%2,%0\n\ + stdcx. %0,0,%3 \n\ + bne- 1b" + : "=&r" (t), "=m" (v->counter) + : "r" (a), "r" (&v->counter), "m" (v->counter) + : "cc"); +} + +static __inline__ long atomic64_sub_return(long a, atomic64_t *v) +{ + long t; + + __asm__ __volatile__( + EIEIO_ON_SMP +"1: ldarx %0,0,%2 # atomic64_sub_return\n\ + subf %0,%1,%0\n\ + stdcx. %0,0,%2 \n\ + bne- 1b" + ISYNC_ON_SMP + : "=&r" (t) + : "r" (a), "r" (&v->counter) + : "cc", "memory"); + + return t; +} + +static __inline__ void atomic64_inc(atomic64_t *v) +{ + long t; + + __asm__ __volatile__( +"1: ldarx %0,0,%2 # atomic64_inc\n\ + addic %0,%0,1\n\ + stdcx. %0,0,%2 \n\ + bne- 1b" + : "=&r" (t), "=m" (v->counter) + : "r" (&v->counter), "m" (v->counter) + : "cc"); +} + +static __inline__ long atomic64_inc_return(atomic64_t *v) +{ + long t; + + __asm__ __volatile__( + EIEIO_ON_SMP +"1: ldarx %0,0,%1 # atomic64_inc_return\n\ + addic %0,%0,1\n\ + stdcx. %0,0,%1 \n\ + bne- 1b" + ISYNC_ON_SMP + : "=&r" (t) + : "r" (&v->counter) + : "cc", "memory"); + + return t; +} + +/* + * atomic64_inc_and_test - increment and test + * @v: pointer of type atomic64_t + * + * Atomically increments @v by 1 + * and returns true if the result is zero, or false for all + * other cases. + */ +#define atomic64_inc_and_test(v) (atomic64_inc_return(v) == 0) + +static __inline__ void atomic64_dec(atomic64_t *v) +{ + long t; + + __asm__ __volatile__( +"1: ldarx %0,0,%2 # atomic64_dec\n\ + addic %0,%0,-1\n\ + stdcx. %0,0,%2\n\ + bne- 1b" + : "=&r" (t), "=m" (v->counter) + : "r" (&v->counter), "m" (v->counter) + : "cc"); +} + +static __inline__ long atomic64_dec_return(atomic64_t *v) +{ + long t; + + __asm__ __volatile__( + EIEIO_ON_SMP +"1: ldarx %0,0,%1 # atomic64_dec_return\n\ + addic %0,%0,-1\n\ + stdcx. %0,0,%1\n\ + bne- 1b" + ISYNC_ON_SMP + : "=&r" (t) + : "r" (&v->counter) + : "cc", "memory"); + + return t; +} + +#define atomic64_sub_and_test(a, v) (atomic64_sub_return((a), (v)) == 0) +#define atomic64_dec_and_test(v) (atomic64_dec_return((v)) == 0) + +/* + * Atomically test *v and decrement if it is greater than 0. + * The function returns the old value of *v minus 1. + */ +static __inline__ long atomic64_dec_if_positive(atomic64_t *v) +{ + long t; + + __asm__ __volatile__( + EIEIO_ON_SMP +"1: ldarx %0,0,%1 # atomic64_dec_if_positive\n\ + addic. %0,%0,-1\n\ + blt- 2f\n\ + stdcx. %0,0,%1\n\ + bne- 1b" + ISYNC_ON_SMP + "\n\ +2:" : "=&r" (t) + : "r" (&v->counter) + : "cc", "memory"); + + return t; +} + +#endif /* __powerpc64__ */ + #endif /* __KERNEL__ */ #endif /* _ASM_POWERPC_ATOMIC_H_ */ From 49b09853df1a303876b82a6480efb2f7b45ef041 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 15:53:40 +1100 Subject: [PATCH 081/121] powerpc: Move some extern declarations from C code into headers This also make klimit have the same type on 32-bit as on 64-bit, namely unsigned long, and defines and initializes it in one place. Signed-off-by: Paul Mackerras --- arch/powerpc/kernel/prom.c | 4 ---- arch/powerpc/kernel/setup-common.c | 2 ++ arch/powerpc/kernel/setup_32.c | 3 +-- arch/powerpc/kernel/setup_64.c | 8 -------- arch/powerpc/mm/init_32.c | 3 --- arch/powerpc/mm/init_64.c | 2 -- arch/powerpc/platforms/iseries/setup.c | 3 +-- include/asm-powerpc/system.h | 1 + include/asm-ppc64/mmu.h | 1 + 9 files changed, 6 insertions(+), 21 deletions(-) diff --git a/arch/powerpc/kernel/prom.c b/arch/powerpc/kernel/prom.c index 1bf3642cb859..6391a4a0709a 100644 --- a/arch/powerpc/kernel/prom.c +++ b/arch/powerpc/kernel/prom.c @@ -71,10 +71,6 @@ struct isa_reg_property { typedef int interpret_func(struct device_node *, unsigned long *, int, int, int); -extern struct rtas_t rtas; -extern struct lmb lmb; -extern unsigned long klimit; - static int __initdata dt_root_addr_cells; static int __initdata dt_root_size_cells; diff --git a/arch/powerpc/kernel/setup-common.c b/arch/powerpc/kernel/setup-common.c index 06e4ef215627..bae4bff138f1 100644 --- a/arch/powerpc/kernel/setup-common.c +++ b/arch/powerpc/kernel/setup-common.c @@ -69,6 +69,8 @@ int _machine = 0; EXPORT_SYMBOL(_machine); #endif +unsigned long klimit = (unsigned long) _end; + /* * This still seems to be needed... -- paulus */ diff --git a/arch/powerpc/kernel/setup_32.c b/arch/powerpc/kernel/setup_32.c index f73d7681b2e9..c98cfcc9cd9a 100644 --- a/arch/powerpc/kernel/setup_32.c +++ b/arch/powerpc/kernel/setup_32.c @@ -279,7 +279,6 @@ arch_initcall(ppc_init); /* Warning, IO base is not yet inited */ void __init setup_arch(char **cmdline_p) { - extern char *klimit; extern void do_init_bootmem(void); /* so udelay does something sensible, assume <= 1000 bogomips */ @@ -338,7 +337,7 @@ void __init setup_arch(char **cmdline_p) init_mm.start_code = PAGE_OFFSET; init_mm.end_code = (unsigned long) _etext; init_mm.end_data = (unsigned long) _edata; - init_mm.brk = (unsigned long) klimit; + init_mm.brk = klimit; /* Save unparsed command line copy for /proc/cmdline */ strlcpy(saved_command_line, cmd_line, COMMAND_LINE_SIZE); diff --git a/arch/powerpc/kernel/setup_64.c b/arch/powerpc/kernel/setup_64.c index be607b877a55..6791668213e7 100644 --- a/arch/powerpc/kernel/setup_64.c +++ b/arch/powerpc/kernel/setup_64.c @@ -96,14 +96,6 @@ extern void udbg_init_maple_realmode(void); do { udbg_putc = call_rtas_display_status_delay; } while(0) #endif -/* extern void *stab; */ -extern unsigned long klimit; - -extern void mm_init_ppc64(void); -extern void early_init_devtree(void *flat_dt); -extern void unflatten_device_tree(void); -extern void check_for_initrd(void); - int have_of = 1; int boot_cpuid = 0; int boot_cpuid_phys = 0; diff --git a/arch/powerpc/mm/init_32.c b/arch/powerpc/mm/init_32.c index 4612a79dfb6e..7d4b8b5f0606 100644 --- a/arch/powerpc/mm/init_32.c +++ b/arch/powerpc/mm/init_32.c @@ -84,9 +84,6 @@ void MMU_init(void); /* XXX should be in current.h -- paulus */ extern struct task_struct *current_set[NR_CPUS]; -char *klimit = _end; -struct device_node *memory_node; - extern int init_bootmem_done; /* diff --git a/arch/powerpc/mm/init_64.c b/arch/powerpc/mm/init_64.c index e274cf10205c..1134f70f231d 100644 --- a/arch/powerpc/mm/init_64.c +++ b/arch/powerpc/mm/init_64.c @@ -80,8 +80,6 @@ #warning TASK_SIZE is smaller than it needs to be. #endif -unsigned long klimit = (unsigned long)_end; - /* max amount of RAM to use */ unsigned long __max_memory; diff --git a/arch/powerpc/platforms/iseries/setup.c b/arch/powerpc/platforms/iseries/setup.c index 623c39aa0436..1f338341d8f2 100644 --- a/arch/powerpc/platforms/iseries/setup.c +++ b/arch/powerpc/platforms/iseries/setup.c @@ -40,7 +40,7 @@ #include #include #include - +#include #include #include #include @@ -85,7 +85,6 @@ static void iSeries_pci_final_fixup(void) { } int piranha_simulator; extern int rd_size; /* Defined in drivers/block/rd.c */ -extern unsigned long klimit; extern unsigned long embedded_sysmap_start; extern unsigned long embedded_sysmap_end; diff --git a/include/asm-powerpc/system.h b/include/asm-powerpc/system.h index f0cce5a30235..5341b75c75cb 100644 --- a/include/asm-powerpc/system.h +++ b/include/asm-powerpc/system.h @@ -179,6 +179,7 @@ extern struct task_struct *_switch(struct thread_struct *prev, extern unsigned int rtas_data; extern int mem_init_done; /* set on boot once kmalloc can be called */ extern unsigned long memory_limit; +extern unsigned long klimit; extern int powersave_nap; /* set if nap mode can be used in idle loop */ diff --git a/include/asm-ppc64/mmu.h b/include/asm-ppc64/mmu.h index c43d512d5902..1a7e0afa2dc6 100644 --- a/include/asm-ppc64/mmu.h +++ b/include/asm-ppc64/mmu.h @@ -229,6 +229,7 @@ extern void htab_initialize_secondary(void); extern void hpte_init_native(void); extern void hpte_init_lpar(void); extern void hpte_init_iSeries(void); +extern void mm_init_ppc64(void); extern long pSeries_lpar_hpte_insert(unsigned long hpte_group, unsigned long va, unsigned long prpn, From d9564ad11454581f85b77026f290f4bb24eecf25 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:50:48 -0600 Subject: [PATCH 082/121] [PATCH] ppc64: mark failed devices 17-eeh-slot-marking-bug.patch A device that experiences a PCI outage may be just one deivce out of many that was affected. In order to avoid repeated reports of a failure, the entire tree of affected devices should be marked as failed. This patch marks up the entire tree. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- arch/powerpc/platforms/pseries/eeh.c | 48 ++++++++++++++++++---------- include/asm-powerpc/ppc-pci.h | 7 ++++ 2 files changed, 39 insertions(+), 16 deletions(-) diff --git a/arch/powerpc/platforms/pseries/eeh.c b/arch/powerpc/platforms/pseries/eeh.c index 3a857b22aa1e..79de2310e70b 100644 --- a/arch/powerpc/platforms/pseries/eeh.c +++ b/arch/powerpc/platforms/pseries/eeh.c @@ -478,32 +478,47 @@ static struct device_node * find_device_pe(struct device_node *dn) * an interrupt context, which is bad. */ -static void __eeh_mark_slot (struct device_node *dn) +static void __eeh_mark_slot (struct device_node *dn, int mode_flag) { while (dn) { - PCI_DN(dn)->eeh_mode |= EEH_MODE_ISOLATED; + if (PCI_DN(dn)) { + PCI_DN(dn)->eeh_mode |= mode_flag; - if (dn->child) - __eeh_mark_slot (dn->child); + if (dn->child) + __eeh_mark_slot (dn->child, mode_flag); + } dn = dn->sibling; } } -static void __eeh_clear_slot (struct device_node *dn) +void eeh_mark_slot (struct device_node *dn, int mode_flag) +{ + dn = find_device_pe (dn); + PCI_DN(dn)->eeh_mode |= mode_flag; + __eeh_mark_slot (dn->child, mode_flag); +} + +static void __eeh_clear_slot (struct device_node *dn, int mode_flag) { while (dn) { - PCI_DN(dn)->eeh_mode &= ~EEH_MODE_ISOLATED; - if (dn->child) - __eeh_clear_slot (dn->child); + if (PCI_DN(dn)) { + PCI_DN(dn)->eeh_mode &= ~mode_flag; + PCI_DN(dn)->eeh_check_count = 0; + if (dn->child) + __eeh_clear_slot (dn->child, mode_flag); + } dn = dn->sibling; } } -static inline void eeh_clear_slot (struct device_node *dn) +void eeh_clear_slot (struct device_node *dn, int mode_flag) { unsigned long flags; spin_lock_irqsave(&confirm_error_lock, flags); - __eeh_clear_slot (dn); + dn = find_device_pe (dn); + PCI_DN(dn)->eeh_mode &= ~mode_flag; + PCI_DN(dn)->eeh_check_count = 0; + __eeh_clear_slot (dn->child, mode_flag); spin_unlock_irqrestore(&confirm_error_lock, flags); } @@ -528,7 +543,6 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) int rets[3]; unsigned long flags; struct pci_dn *pdn; - struct device_node *pe_dn; int rc = 0; __get_cpu_var(total_mmio_ffs)++; @@ -630,8 +644,7 @@ int eeh_dn_check_failure(struct device_node *dn, struct pci_dev *dev) /* Avoid repeated reports of this failure, including problems * with other functions on this device, and functions under * bridges. */ - pe_dn = find_device_pe (dn); - __eeh_mark_slot (pe_dn); + eeh_mark_slot (dn, EEH_MODE_ISOLATED); spin_unlock_irqrestore(&confirm_error_lock, flags); eeh_send_failure_event (dn, dev, rets[0], rets[2]); @@ -743,9 +756,6 @@ rtas_pci_slot_reset(struct pci_dn *pdn, int state) rc, state, pdn->node->full_name); return; } - - if (state == 0) - eeh_clear_slot (pdn->node->parent->child); } /** rtas_set_slot_reset -- assert the pci #RST line for 1/4 second @@ -764,6 +774,12 @@ rtas_set_slot_reset(struct pci_dn *pdn) #define PCI_BUS_RST_HOLD_TIME_MSEC 250 msleep (PCI_BUS_RST_HOLD_TIME_MSEC); + + /* We might get hit with another EEH freeze as soon as the + * pci slot reset line is dropped. Make sure we don't miss + * these, and clear the flag now. */ + eeh_clear_slot (pdn->node, EEH_MODE_ISOLATED); + rtas_pci_slot_reset (pdn, 0); /* After a PCI slot has been reset, the PCI Express spec requires diff --git a/include/asm-powerpc/ppc-pci.h b/include/asm-powerpc/ppc-pci.h index d86c47872bea..9896fade98a7 100644 --- a/include/asm-powerpc/ppc-pci.h +++ b/include/asm-powerpc/ppc-pci.h @@ -87,6 +87,13 @@ void rtas_configure_bridge(struct pci_dn *); int rtas_write_config(struct pci_dn *, int where, int size, u32 val); +/** + * mark and clear slots: find "partition endpoint" PE and set or + * clear the flags for each subnode of the PE. + */ +void eeh_mark_slot (struct device_node *dn, int mode_flag); +void eeh_clear_slot (struct device_node *dn, int mode_flag); + #endif #endif /* _ASM_POWERPC_PPC_PCI_H */ From fe98aeab8494cf431ef62e080cbe1dc1b6f5bd49 Mon Sep 17 00:00:00 2001 From: Linas Vepstas Date: Thu, 3 Nov 2005 18:51:17 -0600 Subject: [PATCH 083/121] [PATCH] ppc64: bugfix: crash on PHB add 19-rpaphp-crashing.patch This patch fixes a bug related to dlpar PHB add, after a PHB removal. -- The crash was due to the PHB not having a pci_dn structure yet, when the phb is being added. This code survived testing, of adding and removeig the PHB and all slots underneath it, 17 times so far, as of this writing. Signed-off-by: Linas Vepstas Signed-off-by: Paul Mackerras --- drivers/pci/hotplug/rpadlpar_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pci/hotplug/rpadlpar_core.c b/drivers/pci/hotplug/rpadlpar_core.c index fcb66b9a0e28..e8593a60ee89 100644 --- a/drivers/pci/hotplug/rpadlpar_core.c +++ b/drivers/pci/hotplug/rpadlpar_core.c @@ -306,7 +306,7 @@ static int dlpar_add_phb(char *drc_name, struct device_node *dn) { struct pci_controller *phb; - if (PCI_DN(dn)->phb) { + if (PCI_DN(dn) && PCI_DN(dn)->phb) { /* PHB already exists */ return -EINVAL; } From 97671e4b295fba1b7d3de2984f98bcdaeb197a26 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 10 Nov 2005 16:06:30 +1100 Subject: [PATCH 084/121] ppc64: Add declarations to ppc64 headers as well as powerpc headers For now, we need these declarations that we moved from C code in the asm-ppc64 versions of these headers as well as the asm-powerpc versions. The asm-ppc64 versions will be disappearing shortly. Signed-off-by: Paul Mackerras --- include/asm-ppc64/prom.h | 2 ++ include/asm-ppc64/system.h | 2 ++ 2 files changed, 4 insertions(+) diff --git a/include/asm-ppc64/prom.h b/include/asm-ppc64/prom.h index 76bb0266d67c..ddfe186589fa 100644 --- a/include/asm-ppc64/prom.h +++ b/include/asm-ppc64/prom.h @@ -204,6 +204,8 @@ extern void of_detach_node(const struct device_node *); extern unsigned long prom_init(unsigned long, unsigned long, unsigned long, unsigned long, unsigned long); extern void finish_device_tree(void); +extern void unflatten_device_tree(void); +extern void early_init_devtree(void *); extern int device_is_compatible(struct device_node *device, const char *); extern int machine_is_compatible(const char *compat); extern unsigned char *get_property(struct device_node *node, const char *name, diff --git a/include/asm-ppc64/system.h b/include/asm-ppc64/system.h index 0cdd66c9f4b7..bf9a6aba19c9 100644 --- a/include/asm-ppc64/system.h +++ b/include/asm-ppc64/system.h @@ -149,6 +149,8 @@ struct thread_struct; extern struct task_struct * _switch(struct thread_struct *prev, struct thread_struct *next); +extern unsigned long klimit; + extern int powersave_nap; /* set if nap mode can be used in idle loop */ /* From fe360cdfd26503a6def1dfebc0ef8371c5071340 Mon Sep 17 00:00:00 2001 From: Stephen Rothwell Date: Thu, 10 Nov 2005 16:07:12 +1100 Subject: [PATCH 085/121] powerpc: fix iSeries build Only MULTIPLATFORM has phbs_remap_io. Signed-off-by: Stephen Rothwell --- arch/ppc64/kernel/pci.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/ppc64/kernel/pci.c b/arch/ppc64/kernel/pci.c index 46929fb24f3c..3cef1b8f57f0 100644 --- a/arch/ppc64/kernel/pci.c +++ b/arch/ppc64/kernel/pci.c @@ -548,8 +548,10 @@ static int __init pcibios_init(void) if (ppc64_isabridge_dev != NULL) printk("ISA bridge at %s\n", pci_name(ppc64_isabridge_dev)); +#ifdef CONFIG_PPC_MULTIPLATFORM /* map in PCI I/O space */ phbs_remap_io(); +#endif printk("PCI: Probing PCI hardware done\n"); From 5ab0b374943b3e42a391a3929e91616ef37dda90 Mon Sep 17 00:00:00 2001 From: Ralf Baechle Date: Wed, 2 Nov 2005 18:08:40 +0000 Subject: [PATCH 086/121] [PATCH] Alchemy PCMCIA: Fix config.h inclusion Add rsp. remove the inclusion of as needed. Signed-off-by: Ralf Baechle Signed-off-by: Dominik Brodowski --- drivers/pcmcia/au1000_db1x00.c | 1 + drivers/pcmcia/au1000_generic.h | 2 ++ drivers/pcmcia/au1000_pb1x00.c | 1 + drivers/pcmcia/au1000_xxs1500.c | 1 - 4 files changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/pcmcia/au1000_db1x00.c b/drivers/pcmcia/au1000_db1x00.c index 24cfee1a412c..abc13f28ba3f 100644 --- a/drivers/pcmcia/au1000_db1x00.c +++ b/drivers/pcmcia/au1000_db1x00.c @@ -30,6 +30,7 @@ * */ +#include #include #include #include diff --git a/drivers/pcmcia/au1000_generic.h b/drivers/pcmcia/au1000_generic.h index b0e7908392a7..f2c970b5f4ff 100644 --- a/drivers/pcmcia/au1000_generic.h +++ b/drivers/pcmcia/au1000_generic.h @@ -22,6 +22,8 @@ #define __ASM_AU1000_PCMCIA_H /* include the world */ +#include + #include #include #include diff --git a/drivers/pcmcia/au1000_pb1x00.c b/drivers/pcmcia/au1000_pb1x00.c index 86c0808d6a05..fd5522ede867 100644 --- a/drivers/pcmcia/au1000_pb1x00.c +++ b/drivers/pcmcia/au1000_pb1x00.c @@ -21,6 +21,7 @@ * with this program; if not, write to the Free Software Foundation, Inc., * 59 Temple Place - Suite 330, Boston MA 02111-1307, USA. */ +#include #include #include #include diff --git a/drivers/pcmcia/au1000_xxs1500.c b/drivers/pcmcia/au1000_xxs1500.c index 01a895bc9a47..01874b0bb03b 100644 --- a/drivers/pcmcia/au1000_xxs1500.c +++ b/drivers/pcmcia/au1000_xxs1500.c @@ -27,7 +27,6 @@ */ #include #include -#include #include #include #include From e34631508861237e598e7d72703eae4478761f37 Mon Sep 17 00:00:00 2001 From: Marcelo Tosatti Date: Sat, 5 Nov 2005 13:03:32 -0200 Subject: [PATCH 087/121] [PCMCIA] MPC8xx PCMCIA update Kconfig entry: dependency on 8xx Makefile: fix whitespace breakage m8xx_pcmcia.c: - asm/segment.h is gone - use generic PCMCIA suspend/resume methods Signed-off-by: Marcelo Tosatti Signed-off-by: Dominik Brodowski --- drivers/pcmcia/Kconfig | 2 +- drivers/pcmcia/Makefile | 6 +++--- drivers/pcmcia/m8xx_pcmcia.c | 24 +++--------------------- 3 files changed, 7 insertions(+), 25 deletions(-) diff --git a/drivers/pcmcia/Kconfig b/drivers/pcmcia/Kconfig index ccf20039e909..309eb557f9a3 100644 --- a/drivers/pcmcia/Kconfig +++ b/drivers/pcmcia/Kconfig @@ -156,7 +156,7 @@ config TCIC config PCMCIA_M8XX tristate "MPC8xx PCMCIA support" - depends on PCMCIA && PPC + depends on PCMCIA && PPC && 8xx select PCCARD_NONSTATIC help Say Y here to include support for PowerPC 8xx series PCMCIA diff --git a/drivers/pcmcia/Makefile b/drivers/pcmcia/Makefile index fe37541abbfe..bcecf5133b7e 100644 --- a/drivers/pcmcia/Makefile +++ b/drivers/pcmcia/Makefile @@ -25,7 +25,7 @@ obj-$(CONFIG_PD6729) += pd6729.o obj-$(CONFIG_I82365) += i82365.o obj-$(CONFIG_I82092) += i82092.o obj-$(CONFIG_TCIC) += tcic.o -obj-$(CONFIG_PCMCIA_M8XX) += m8xx_pcmcia.o +obj-$(CONFIG_PCMCIA_M8XX) += m8xx_pcmcia.o obj-$(CONFIG_HD64465_PCMCIA) += hd64465_ss.o obj-$(CONFIG_PCMCIA_SA1100) += sa11xx_core.o sa1100_cs.o obj-$(CONFIG_PCMCIA_SA1111) += sa11xx_core.o sa1111_cs.o @@ -47,10 +47,10 @@ au1x00_ss-$(CONFIG_MIPS_PB1200) += au1000_db1x00.o au1x00_ss-$(CONFIG_MIPS_PB1500) += au1000_pb1x00.o au1x00_ss-$(CONFIG_MIPS_DB1000) += au1000_db1x00.o au1x00_ss-$(CONFIG_MIPS_DB1100) += au1000_db1x00.o -au1x00_ss-$(CONFIG_MIPS_DB1200) += au1000_db1x00.o +au1x00_ss-$(CONFIG_MIPS_DB1200) += au1000_db1x00.o au1x00_ss-$(CONFIG_MIPS_DB1500) += au1000_db1x00.o au1x00_ss-$(CONFIG_MIPS_DB1550) += au1000_db1x00.o -au1x00_ss-$(CONFIG_MIPS_XXS1500) += au1000_xxs1500.o +au1x00_ss-$(CONFIG_MIPS_XXS1500) += au1000_xxs1500.o sa1111_cs-y += sa1111_generic.o sa1111_cs-$(CONFIG_ASSABET_NEPONSET) += sa1100_neponset.o diff --git a/drivers/pcmcia/m8xx_pcmcia.c b/drivers/pcmcia/m8xx_pcmcia.c index f8bed87cf2f1..6d9f71cfcb34 100644 --- a/drivers/pcmcia/m8xx_pcmcia.c +++ b/drivers/pcmcia/m8xx_pcmcia.c @@ -39,7 +39,6 @@ #include #include -#include #include #include @@ -50,6 +49,7 @@ #include #include #include +#include #include #include @@ -546,29 +546,11 @@ static void m8xx_shutdown(void) free_irq(pcmcia_schlvl, NULL); } -/* copied from tcic.c */ - -static int m8xx_drv_suspend(struct device *dev, pm_message_t state, u32 level) -{ - int ret = 0; - if (level == SUSPEND_SAVE_STATE) - ret = pcmcia_socket_dev_suspend(dev, state); - return ret; -} - -static int m8xx_drv_resume(struct device *dev, u32 level) -{ - int ret = 0; - if (level == RESUME_RESTORE_STATE) - ret = pcmcia_socket_dev_resume(dev); - return ret; -} - static struct device_driver m8xx_driver = { .name = "m8xx-pcmcia", .bus = &platform_bus_type, - .suspend = m8xx_drv_suspend, - .resume = m8xx_drv_resume, + .suspend = pcmcia_socket_dev_suspend, + .resume = pcmcia_socket_dev_resume, }; static struct platform_device m8xx_device = { From a2932b35a0efd01b20a3f598d19db052d64935f5 Mon Sep 17 00:00:00 2001 From: Igor Popik Date: Thu, 6 Oct 2005 20:32:58 +0200 Subject: [PATCH 088/121] [PCMCIA] i82365: release all resources if no devices are found The i82365 driver does not release all the resources when the device is not found. This can cause an oops when reading /proc/ioports after module unload. Signed-off-by: Igor Popik Signed-off-by: Dominik Brodowski --- drivers/pcmcia/i82365.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/pcmcia/i82365.c b/drivers/pcmcia/i82365.c index 7ce455d01cc9..4ddd76239b34 100644 --- a/drivers/pcmcia/i82365.c +++ b/drivers/pcmcia/i82365.c @@ -1366,6 +1366,7 @@ static int __init init_i82365(void) if (sockets == 0) { printk("not found.\n"); platform_device_unregister(&i82365_device); + release_region(i365_base, 2); driver_unregister(&i82365_driver); return -ENODEV; } From af36bef0c5bb82f361ebb2f106f11d0f63dac887 Mon Sep 17 00:00:00 2001 From: Russell King Date: Thu, 10 Nov 2005 11:40:53 +0000 Subject: [PATCH 089/121] [ARM] Fix csumpartial corner case Ji-In Park discovered a bug in csumpartial which caused wrong checksums with misaligned buffers. Signed-off-by: Russell King --- arch/arm/lib/csumpartial.S | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/arch/arm/lib/csumpartial.S b/arch/arm/lib/csumpartial.S index cb5e3708f118..fe797cf320bb 100644 --- a/arch/arm/lib/csumpartial.S +++ b/arch/arm/lib/csumpartial.S @@ -39,6 +39,7 @@ td3 .req lr /* we must have at least one byte. */ tst buf, #1 @ odd address? + movne sum, sum, ror #8 ldrneb td0, [buf], #1 subne len, len, #1 adcnes sum, sum, td0, put_byte_1 @@ -103,6 +104,9 @@ ENTRY(csum_partial) cmp len, #8 @ Ensure that we have at least blo .less8 @ 8 bytes to copy. + tst buf, #1 + movne sum, sum, ror #8 + adds sum, sum, #0 @ C = 0 tst buf, #3 @ Test destination alignment blne .not_aligned @ aligh destination, return here From a7918f39bbe59fe76f43743bdb6bb8b0bdefd94a Mon Sep 17 00:00:00 2001 From: Alessandro Zummo Date: Thu, 10 Nov 2005 14:05:04 +0000 Subject: [PATCH 090/121] [ARM] 3140/1: NSLU2 machine support Patch from Alessandro Zummo This patch adds support for the LinkSys NSLU2 running with both big and little-endian kernels. The LinkSys NSLU2 is a cost engineered ARM, XScale 420 based system similar to the the Intel IXDP425 evaluation board. It uses the IXP4XX ARCH. While this patch applies independently of other patches the resultant kernel requires further patches to successfully use onboard devices, including the onboard flash. Since these patches are independent of this one they will be submitted separately. A defconfig is not included here because not all of the required drivers are actually in the kernel. We intend to provide one as soon as the patches will be incorporated in mainstream. This patch is the combined work of nslu2-linux.org Signed-off-by: John Bowler Signed-off-by: Alessandro Zummo Signed-off-by: Deepak Saxena Signed-off-by: Lennert Buytenhek Signed-off-by: Russell King --- arch/arm/mach-ixp4xx/Kconfig | 10 ++ arch/arm/mach-ixp4xx/Makefile | 1 + arch/arm/mach-ixp4xx/nslu2-pci.c | 77 ++++++++++++++ arch/arm/mach-ixp4xx/nslu2-power.c | 92 +++++++++++++++++ arch/arm/mach-ixp4xx/nslu2-setup.c | 134 +++++++++++++++++++++++++ include/asm-arm/arch-ixp4xx/hardware.h | 1 + include/asm-arm/arch-ixp4xx/irqs.h | 7 ++ include/asm-arm/arch-ixp4xx/nslu2.h | 96 ++++++++++++++++++ 8 files changed, 418 insertions(+) create mode 100644 arch/arm/mach-ixp4xx/nslu2-pci.c create mode 100644 arch/arm/mach-ixp4xx/nslu2-power.c create mode 100644 arch/arm/mach-ixp4xx/nslu2-setup.c create mode 100644 include/asm-arm/arch-ixp4xx/nslu2.h diff --git a/arch/arm/mach-ixp4xx/Kconfig b/arch/arm/mach-ixp4xx/Kconfig index 89762a26495c..385285851cb5 100644 --- a/arch/arm/mach-ixp4xx/Kconfig +++ b/arch/arm/mach-ixp4xx/Kconfig @@ -8,6 +8,16 @@ menu "Intel IXP4xx Implementation Options" comment "IXP4xx Platforms" +# This entry is placed on top because otherwise it would have +# been shown as a submenu. +config MACH_NSLU2 + bool + prompt "NSLU2" if !(MACH_IXDP465 || MACH_IXDPG425 || ARCH_IXDP425 || ARCH_ADI_COYOTE || ARCH_AVILA || ARCH_IXCDP1100 || ARCH_PRPMC1100 || MACH_GTWX5715) + help + Say 'Y' here if you want your kernel to support Linksys's + NSLU2 NAS device. For more information on this platform, + see http://www.nslu2-linux.org + config ARCH_AVILA bool "Avila" help diff --git a/arch/arm/mach-ixp4xx/Makefile b/arch/arm/mach-ixp4xx/Makefile index ddecbda4a633..7a15629c18d0 100644 --- a/arch/arm/mach-ixp4xx/Makefile +++ b/arch/arm/mach-ixp4xx/Makefile @@ -8,4 +8,5 @@ obj-$(CONFIG_ARCH_IXDP4XX) += ixdp425-pci.o ixdp425-setup.o obj-$(CONFIG_MACH_IXDPG425) += ixdpg425-pci.o coyote-setup.o obj-$(CONFIG_ARCH_ADI_COYOTE) += coyote-pci.o coyote-setup.o obj-$(CONFIG_MACH_GTWX5715) += gtwx5715-pci.o gtwx5715-setup.o +obj-$(CONFIG_MACH_NSLU2) += nslu2-pci.o nslu2-setup.o nslu2-power.o diff --git a/arch/arm/mach-ixp4xx/nslu2-pci.c b/arch/arm/mach-ixp4xx/nslu2-pci.c new file mode 100644 index 000000000000..a575f2e0b2c8 --- /dev/null +++ b/arch/arm/mach-ixp4xx/nslu2-pci.c @@ -0,0 +1,77 @@ +/* + * arch/arm/mach-ixp4xx/nslu2-pci.c + * + * NSLU2 board-level PCI initialization + * + * based on ixdp425-pci.c: + * Copyright (C) 2002 Intel Corporation. + * Copyright (C) 2003-2004 MontaVista Software, Inc. + * + * Maintainer: http://www.nslu2-linux.org/ + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#include +#include +#include + +#include +#include + +void __init nslu2_pci_preinit(void) +{ + set_irq_type(IRQ_NSLU2_PCI_INTA, IRQT_LOW); + set_irq_type(IRQ_NSLU2_PCI_INTB, IRQT_LOW); + set_irq_type(IRQ_NSLU2_PCI_INTC, IRQT_LOW); + + gpio_line_isr_clear(NSLU2_PCI_INTA_PIN); + gpio_line_isr_clear(NSLU2_PCI_INTB_PIN); + gpio_line_isr_clear(NSLU2_PCI_INTC_PIN); + + /* INTD is not configured as GPIO is used + * for the power input button. + */ + + ixp4xx_pci_preinit(); +} + +static int __init nslu2_map_irq(struct pci_dev *dev, u8 slot, u8 pin) +{ + static int pci_irq_table[NSLU2_PCI_IRQ_LINES] = { + IRQ_NSLU2_PCI_INTA, + IRQ_NSLU2_PCI_INTB, + IRQ_NSLU2_PCI_INTC, + }; + + int irq = -1; + + if (slot >= 1 && slot <= NSLU2_PCI_MAX_DEV && + pin >= 1 && pin <= NSLU2_PCI_IRQ_LINES) { + irq = pci_irq_table[(slot + pin - 2) % NSLU2_PCI_IRQ_LINES]; + } + + return irq; +} + +struct hw_pci __initdata nslu2_pci = { + .nr_controllers = 1, + .preinit = nslu2_pci_preinit, + .swizzle = pci_std_swizzle, + .setup = ixp4xx_setup, + .scan = ixp4xx_scan_bus, + .map_irq = nslu2_map_irq, +}; + +int __init nslu2_pci_init(void) /* monkey see, monkey do */ +{ + if (machine_is_nslu2()) + pci_common_init(&nslu2_pci); + + return 0; +} + +subsys_initcall(nslu2_pci_init); diff --git a/arch/arm/mach-ixp4xx/nslu2-power.c b/arch/arm/mach-ixp4xx/nslu2-power.c new file mode 100644 index 000000000000..18fbc8c0fb30 --- /dev/null +++ b/arch/arm/mach-ixp4xx/nslu2-power.c @@ -0,0 +1,92 @@ +/* + * arch/arm/mach-ixp4xx/nslu2-power.c + * + * NSLU2 Power/Reset driver + * + * Copyright (C) 2005 Tower Technologies + * + * based on nslu2-io.c + * Copyright (C) 2004 Karen Spearel + * + * Author: Alessandro Zummo + * Maintainers: http://www.nslu2-linux.org/ + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#include +#include +#include + +#include + +extern void ctrl_alt_del(void); + +static irqreturn_t nslu2_power_handler(int irq, void *dev_id, struct pt_regs *regs) +{ + /* Signal init to do the ctrlaltdel action, this will bypass init if + * it hasn't started and do a kernel_restart. + */ + ctrl_alt_del(); + + return IRQ_HANDLED; +} + +static irqreturn_t nslu2_reset_handler(int irq, void *dev_id, struct pt_regs *regs) +{ + /* This is the paper-clip reset, it shuts the machine down directly. + */ + machine_power_off(); + + return IRQ_HANDLED; +} + +static int __init nslu2_power_init(void) +{ + if (!(machine_is_nslu2())) + return 0; + + *IXP4XX_GPIO_GPISR = 0x20400000; /* read the 2 irqs to clr */ + + set_irq_type(NSLU2_RB_IRQ, IRQT_LOW); + set_irq_type(NSLU2_PB_IRQ, IRQT_HIGH); + + gpio_line_isr_clear(NSLU2_RB_GPIO); + gpio_line_isr_clear(NSLU2_PB_GPIO); + + if (request_irq(NSLU2_RB_IRQ, &nslu2_reset_handler, + SA_INTERRUPT, "NSLU2 reset button", NULL) < 0) { + + printk(KERN_DEBUG "Reset Button IRQ %d not available\n", + NSLU2_RB_IRQ); + + return -EIO; + } + + if (request_irq(NSLU2_PB_IRQ, &nslu2_power_handler, + SA_INTERRUPT, "NSLU2 power button", NULL) < 0) { + + printk(KERN_DEBUG "Power Button IRQ %d not available\n", + NSLU2_PB_IRQ); + + return -EIO; + } + + return 0; +} + +static void __exit nslu2_power_exit(void) +{ + free_irq(NSLU2_RB_IRQ, NULL); + free_irq(NSLU2_PB_IRQ, NULL); +} + +module_init(nslu2_power_init); +module_exit(nslu2_power_exit); + +MODULE_AUTHOR("Alessandro Zummo "); +MODULE_DESCRIPTION("NSLU2 Power/Reset driver"); +MODULE_LICENSE("GPL"); diff --git a/arch/arm/mach-ixp4xx/nslu2-setup.c b/arch/arm/mach-ixp4xx/nslu2-setup.c new file mode 100644 index 000000000000..289e94cb65c2 --- /dev/null +++ b/arch/arm/mach-ixp4xx/nslu2-setup.c @@ -0,0 +1,134 @@ +/* + * arch/arm/mach-ixp4xx/nslu2-setup.c + * + * NSLU2 board-setup + * + * based ixdp425-setup.c: + * Copyright (C) 2003-2004 MontaVista Software, Inc. + * + * Author: Mark Rakes + * Maintainers: http://www.nslu2-linux.org/ + * + * Fixed missing init_time in MACHINE_START kas11 10/22/04 + * Changed to conform to new style __init ixdp425 kas11 10/22/04 + */ + +#include +#include +#include + +#include +#include +#include + +static struct flash_platform_data nslu2_flash_data = { + .map_name = "cfi_probe", + .width = 2, +}; + +static struct resource nslu2_flash_resource = { + .start = NSLU2_FLASH_BASE, + .end = NSLU2_FLASH_BASE + NSLU2_FLASH_SIZE, + .flags = IORESOURCE_MEM, +}; + +static struct platform_device nslu2_flash = { + .name = "IXP4XX-Flash", + .id = 0, + .dev.platform_data = &nslu2_flash_data, + .num_resources = 1, + .resource = &nslu2_flash_resource, +}; + +static struct ixp4xx_i2c_pins nslu2_i2c_gpio_pins = { + .sda_pin = NSLU2_SDA_PIN, + .scl_pin = NSLU2_SCL_PIN, +}; + +static struct platform_device nslu2_i2c_controller = { + .name = "IXP4XX-I2C", + .id = 0, + .dev.platform_data = &nslu2_i2c_gpio_pins, + .num_resources = 0, +}; + +static struct resource nslu2_uart_resources[] = { + { + .start = IXP4XX_UART1_BASE_PHYS, + .end = IXP4XX_UART1_BASE_PHYS + 0x0fff, + .flags = IORESOURCE_MEM, + }, + { + .start = IXP4XX_UART2_BASE_PHYS, + .end = IXP4XX_UART2_BASE_PHYS + 0x0fff, + .flags = IORESOURCE_MEM, + } +}; + +static struct plat_serial8250_port nslu2_uart_data[] = { + { + .mapbase = IXP4XX_UART1_BASE_PHYS, + .membase = (char *)IXP4XX_UART1_BASE_VIRT + REG_OFFSET, + .irq = IRQ_IXP4XX_UART1, + .flags = UPF_BOOT_AUTOCONF, + .iotype = UPIO_MEM, + .regshift = 2, + .uartclk = IXP4XX_UART_XTAL, + }, + { + .mapbase = IXP4XX_UART2_BASE_PHYS, + .membase = (char *)IXP4XX_UART2_BASE_VIRT + REG_OFFSET, + .irq = IRQ_IXP4XX_UART2, + .flags = UPF_BOOT_AUTOCONF, + .iotype = UPIO_MEM, + .regshift = 2, + .uartclk = IXP4XX_UART_XTAL, + }, + { } +}; + +static struct platform_device nslu2_uart = { + .name = "serial8250", + .id = PLAT8250_DEV_PLATFORM, + .dev.platform_data = nslu2_uart_data, + .num_resources = 2, + .resource = nslu2_uart_resources, +}; + +static struct platform_device *nslu2_devices[] __initdata = { + &nslu2_i2c_controller, + &nslu2_flash, + &nslu2_uart, +}; + +static void nslu2_power_off(void) +{ + /* This causes the box to drop the power and go dead. */ + + /* enable the pwr cntl gpio */ + gpio_line_config(NSLU2_PO_GPIO, IXP4XX_GPIO_OUT); + + /* do the deed */ + gpio_line_set(NSLU2_PO_GPIO, IXP4XX_GPIO_HIGH); +} + +static void __init nslu2_init(void) +{ + ixp4xx_sys_init(); + + pm_power_off = nslu2_power_off; + + platform_add_devices(nslu2_devices, ARRAY_SIZE(nslu2_devices)); +} + +MACHINE_START(NSLU2, "Linksys NSLU2") + /* Maintainer: www.nslu2-linux.org */ + .phys_ram = PHYS_OFFSET, + .phys_io = IXP4XX_PERIPHERAL_BASE_PHYS, + .io_pg_offst = ((IXP4XX_PERIPHERAL_BASE_VIRT) >> 18) & 0xFFFC, + .boot_params = 0x00000100, + .map_io = ixp4xx_map_io, + .init_irq = ixp4xx_init_irq, + .timer = &ixp4xx_timer, + .init_machine = nslu2_init, +MACHINE_END diff --git a/include/asm-arm/arch-ixp4xx/hardware.h b/include/asm-arm/arch-ixp4xx/hardware.h index 55d85eea8c1a..cfb413c845f7 100644 --- a/include/asm-arm/arch-ixp4xx/hardware.h +++ b/include/asm-arm/arch-ixp4xx/hardware.h @@ -44,5 +44,6 @@ extern unsigned int processor_id; #include "ixdp425.h" #include "coyote.h" #include "prpmc1100.h" +#include "nslu2.h" #endif /* _ASM_ARCH_HARDWARE_H */ diff --git a/include/asm-arm/arch-ixp4xx/irqs.h b/include/asm-arm/arch-ixp4xx/irqs.h index ca808281c7f9..2cf4930372bc 100644 --- a/include/asm-arm/arch-ixp4xx/irqs.h +++ b/include/asm-arm/arch-ixp4xx/irqs.h @@ -93,4 +93,11 @@ #define IRQ_COYOTE_PCI_SLOT1 IRQ_IXP4XX_GPIO11 #define IRQ_COYOTE_IDE IRQ_IXP4XX_GPIO5 +/* + * NSLU2 board IRQs + */ +#define IRQ_NSLU2_PCI_INTA IRQ_IXP4XX_GPIO11 +#define IRQ_NSLU2_PCI_INTB IRQ_IXP4XX_GPIO10 +#define IRQ_NSLU2_PCI_INTC IRQ_IXP4XX_GPIO9 + #endif diff --git a/include/asm-arm/arch-ixp4xx/nslu2.h b/include/asm-arm/arch-ixp4xx/nslu2.h new file mode 100644 index 000000000000..b8b347a559c7 --- /dev/null +++ b/include/asm-arm/arch-ixp4xx/nslu2.h @@ -0,0 +1,96 @@ +/* + * include/asm-arm/arch-ixp4xx/nslu2.h + * + * NSLU2 platform specific definitions + * + * Author: Mark Rakes + * Maintainers: http://www.nslu2-linux.org + * + * based on ixdp425.h: + * Copyright 2004 (c) MontaVista, Software, Inc. + * + * This file is licensed under the terms of the GNU General Public + * License version 2. This program is licensed "as is" without any + * warranty of any kind, whether express or implied. + */ + +#ifndef __ASM_ARCH_HARDWARE_H__ +#error "Do not include this directly, instead #include " +#endif + +#define NSLU2_FLASH_BASE IXP4XX_EXP_BUS_CS0_BASE_PHYS +#define NSLU2_FLASH_SIZE IXP4XX_EXP_BUS_CSX_REGION_SIZE + +#define NSLU2_SDA_PIN 7 +#define NSLU2_SCL_PIN 6 + +/* + * NSLU2 PCI IRQs + */ +#define NSLU2_PCI_MAX_DEV 3 +#define NSLU2_PCI_IRQ_LINES 3 + + +/* PCI controller GPIO to IRQ pin mappings */ +#define NSLU2_PCI_INTA_PIN 11 +#define NSLU2_PCI_INTB_PIN 10 +#define NSLU2_PCI_INTC_PIN 9 +#define NSLU2_PCI_INTD_PIN 8 + + +/* NSLU2 Timer */ +#define NSLU2_FREQ 66000000 +#define NSLU2_CLOCK_TICK_RATE (((NSLU2_FREQ / HZ & ~IXP4XX_OST_RELOAD_MASK) + 1) * HZ) +#define NSLU2_CLOCK_TICKS_PER_USEC ((NSLU2_CLOCK_TICK_RATE + USEC_PER_SEC/2) / USEC_PER_SEC) + +/* GPIO */ + +#define NSLU2_GPIO0 0 +#define NSLU2_GPIO1 1 +#define NSLU2_GPIO2 2 +#define NSLU2_GPIO3 3 +#define NSLU2_GPIO4 4 +#define NSLU2_GPIO5 5 +#define NSLU2_GPIO6 6 +#define NSLU2_GPIO7 7 +#define NSLU2_GPIO8 8 +#define NSLU2_GPIO9 9 +#define NSLU2_GPIO10 10 +#define NSLU2_GPIO11 11 +#define NSLU2_GPIO12 12 +#define NSLU2_GPIO13 13 +#define NSLU2_GPIO14 14 +#define NSLU2_GPIO15 15 + +/* Buttons */ + +#define NSLU2_PB_GPIO NSLU2_GPIO5 +#define NSLU2_PO_GPIO NSLU2_GPIO8 /* power off */ +#define NSLU2_RB_GPIO NSLU2_GPIO12 + +#define NSLU2_PB_IRQ IRQ_IXP4XX_GPIO5 +#define NSLU2_RB_IRQ IRQ_IXP4XX_GPIO12 + +#define NSLU2_PB_BM (1L << NSLU2_PB_GPIO) +#define NSLU2_PO_BM (1L << NSLU2_PO_GPIO) +#define NSLU2_RB_BM (1L << NSLU2_RB_GPIO) + +/* Buzzer */ + +#define NSLU2_GPIO_BUZZ 4 +#define NSLU2_BZ_BM (1L << NSLU2_GPIO_BUZZ) +/* LEDs */ + +#define NSLU2_LED_RED NSLU2_GPIO0 +#define NSLU2_LED_GRN NSLU2_GPIO1 + +#define NSLU2_LED_RED_BM (1L << NSLU2_LED_RED) +#define NSLU2_LED_GRN_BM (1L << NSLU2_LED_GRN) + +#define NSLU2_LED_DISK1 NSLU2_GPIO2 +#define NSLU2_LED_DISK2 NSLU2_GPIO3 + +#define NSLU2_LED_DISK1_BM (1L << NSLU2_GPIO2) +#define NSLU2_LED_DISK2_BM (1L << NSLU2_GPIO3) + + From 3179a019391f0f8081245fd564a5f1be308ba64f Mon Sep 17 00:00:00 2001 From: Tony Lindgren Date: Thu, 10 Nov 2005 14:26:48 +0000 Subject: [PATCH 091/121] [ARM] 3141/1: OMAP 1/5: Update omap1 specific files Patch from Tony Lindgren This patch syncs the mainline kernel with linux-omap tree. The highlights of the patch are: - Omap1 serial pport and framebuffer init updates by Imre Deak - Add support for omap310 processor and Palm Tungsten E PDA by Laurent Gonzales, Romain Goyet, et al. Omap310 and omap1510 processors are now handled as omap15xx. - Omap1 specific changes to shared omap clock framework by Tony Lindgren - Omap1 specific changes to shared omap pin mux framework by Tony Lindgren - Other misc fixes, such as update memory timings for smc91x, omap1 specific device initialization etc. Signed-off-by: Tony Lindgren Signed-off-by: Russell King --- arch/arm/mach-omap1/Kconfig | 32 +- arch/arm/mach-omap1/Makefile | 5 +- arch/arm/mach-omap1/board-generic.c | 33 +- arch/arm/mach-omap1/board-h2.c | 15 +- arch/arm/mach-omap1/board-h3.c | 19 +- arch/arm/mach-omap1/board-innovator.c | 42 +- arch/arm/mach-omap1/board-netstar.c | 15 +- arch/arm/mach-omap1/board-osk.c | 17 +- arch/arm/mach-omap1/board-palmte.c | 87 +++ arch/arm/mach-omap1/board-perseus2.c | 23 +- arch/arm/mach-omap1/board-voiceblue.c | 9 +- arch/arm/mach-omap1/clock.c | 792 ++++++++++++++++++++++++++ arch/arm/mach-omap1/clock.h | 768 +++++++++++++++++++++++++ arch/arm/mach-omap1/devices.c | 221 +------ arch/arm/mach-omap1/id.c | 9 + arch/arm/mach-omap1/io.c | 13 +- arch/arm/mach-omap1/irq.c | 23 +- arch/arm/mach-omap1/leds-h2p2-debug.c | 43 +- arch/arm/mach-omap1/leds.c | 1 - arch/arm/mach-omap1/mux.c | 289 ++++++++++ arch/arm/mach-omap1/serial.c | 9 +- arch/arm/mach-omap1/time.c | 4 +- arch/arm/mm/Kconfig | 4 +- 23 files changed, 2157 insertions(+), 316 deletions(-) create mode 100644 arch/arm/mach-omap1/board-palmte.c create mode 100644 arch/arm/mach-omap1/clock.c create mode 100644 arch/arm/mach-omap1/clock.h create mode 100644 arch/arm/mach-omap1/mux.c diff --git a/arch/arm/mach-omap1/Kconfig b/arch/arm/mach-omap1/Kconfig index 27fc2e8e5fca..86a0f0d14345 100644 --- a/arch/arm/mach-omap1/Kconfig +++ b/arch/arm/mach-omap1/Kconfig @@ -6,10 +6,10 @@ config ARCH_OMAP730 bool "OMAP730 Based System" select ARCH_OMAP_OTG -config ARCH_OMAP1510 +config ARCH_OMAP15XX depends on ARCH_OMAP1 default y - bool "OMAP1510 Based System" + bool "OMAP15xx Based System" config ARCH_OMAP16XX depends on ARCH_OMAP1 @@ -21,7 +21,7 @@ comment "OMAP Board Type" config MACH_OMAP_INNOVATOR bool "TI Innovator" - depends on ARCH_OMAP1 && (ARCH_OMAP1510 || ARCH_OMAP16XX) + depends on ARCH_OMAP1 && (ARCH_OMAP15XX || ARCH_OMAP16XX) help TI OMAP 1510 or 1610 Innovator board support. Say Y here if you have such a board. @@ -64,20 +64,30 @@ config MACH_OMAP_PERSEUS2 config MACH_VOICEBLUE bool "Voiceblue" - depends on ARCH_OMAP1 && ARCH_OMAP1510 + depends on ARCH_OMAP1 && ARCH_OMAP15XX help Support for Voiceblue GSM/VoIP gateway. Say Y here if you have such a board. config MACH_NETSTAR bool "NetStar" - depends on ARCH_OMAP1 && ARCH_OMAP1510 + depends on ARCH_OMAP1 && ARCH_OMAP15XX help Support for NetStar PBX. Say Y here if you have such a board. +config MACH_OMAP_PALMTE + bool "Palm Tungsten E" + depends on ARCH_OMAP1 && ARCH_OMAP15XX + help + Support for the Palm Tungsten E PDA. Currently only the LCD panel + is supported. To boot the kernel, you'll need a PalmOS compatible + bootloader; check out http://palmtelinux.sourceforge.net for more + informations. + Say Y here if you have such a PDA, say NO otherwise. + config MACH_OMAP_GENERIC bool "Generic OMAP board" - depends on ARCH_OMAP1 && (ARCH_OMAP1510 || ARCH_OMAP16XX) + depends on ARCH_OMAP1 && (ARCH_OMAP15XX || ARCH_OMAP16XX) help Support for generic OMAP-1510, 1610 or 1710 board with no FPGA. Can be used as template for porting Linux to @@ -121,32 +131,32 @@ config OMAP_ARM_182MHZ config OMAP_ARM_168MHZ bool "OMAP ARM 168 MHz CPU" - depends on ARCH_OMAP1 && (ARCH_OMAP1510 || ARCH_OMAP16XX || ARCH_OMAP730) + depends on ARCH_OMAP1 && (ARCH_OMAP15XX || ARCH_OMAP16XX || ARCH_OMAP730) help Enable 168MHz clock for OMAP CPU. If unsure, say N. config OMAP_ARM_150MHZ bool "OMAP ARM 150 MHz CPU" - depends on ARCH_OMAP1 && ARCH_OMAP1510 + depends on ARCH_OMAP1 && ARCH_OMAP15XX help Enable 150MHz clock for OMAP CPU. If unsure, say N. config OMAP_ARM_120MHZ bool "OMAP ARM 120 MHz CPU" - depends on ARCH_OMAP1 && (ARCH_OMAP1510 || ARCH_OMAP16XX || ARCH_OMAP730) + depends on ARCH_OMAP1 && (ARCH_OMAP15XX || ARCH_OMAP16XX || ARCH_OMAP730) help Enable 120MHz clock for OMAP CPU. If unsure, say N. config OMAP_ARM_60MHZ bool "OMAP ARM 60 MHz CPU" - depends on ARCH_OMAP1 && (ARCH_OMAP1510 || ARCH_OMAP16XX || ARCH_OMAP730) + depends on ARCH_OMAP1 && (ARCH_OMAP15XX || ARCH_OMAP16XX || ARCH_OMAP730) default y help Enable 60MHz clock for OMAP CPU. If unsure, say Y. config OMAP_ARM_30MHZ bool "OMAP ARM 30 MHz CPU" - depends on ARCH_OMAP1 && (ARCH_OMAP1510 || ARCH_OMAP16XX || ARCH_OMAP730) + depends on ARCH_OMAP1 && (ARCH_OMAP15XX || ARCH_OMAP16XX || ARCH_OMAP730) help Enable 30MHz clock for OMAP CPU. If unsure, say N. diff --git a/arch/arm/mach-omap1/Makefile b/arch/arm/mach-omap1/Makefile index 181a93deaaee..b0b00156faae 100644 --- a/arch/arm/mach-omap1/Makefile +++ b/arch/arm/mach-omap1/Makefile @@ -3,7 +3,7 @@ # # Common support -obj-y := io.o id.o irq.o time.o serial.o devices.o +obj-y := io.o id.o clock.o irq.o time.o mux.o serial.o devices.o led-y := leds.o # Specific board support @@ -15,8 +15,9 @@ obj-$(CONFIG_MACH_OMAP_OSK) += board-osk.o obj-$(CONFIG_MACH_OMAP_H3) += board-h3.o obj-$(CONFIG_MACH_VOICEBLUE) += board-voiceblue.o obj-$(CONFIG_MACH_NETSTAR) += board-netstar.o +obj-$(CONFIG_MACH_OMAP_PALMTE) += board-palmte.o -ifeq ($(CONFIG_ARCH_OMAP1510),y) +ifeq ($(CONFIG_ARCH_OMAP15XX),y) # Innovator-1510 FPGA obj-$(CONFIG_MACH_OMAP_INNOVATOR) += fpga.o endif diff --git a/arch/arm/mach-omap1/board-generic.c b/arch/arm/mach-omap1/board-generic.c index c209c7172a9a..4b292e93fbe2 100644 --- a/arch/arm/mach-omap1/board-generic.c +++ b/arch/arm/mach-omap1/board-generic.c @@ -15,7 +15,7 @@ #include #include -#include +#include #include #include @@ -28,8 +28,6 @@ #include #include -static int __initdata generic_serial_ports[OMAP_MAX_NR_PORTS] = {1, 1, 1}; - static void __init omap_generic_init_irq(void) { omap_init_irq(); @@ -37,7 +35,7 @@ static void __init omap_generic_init_irq(void) /* assume no Mini-AB port */ -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX static struct omap_usb_config generic1510_usb_config __initdata = { .register_host = 1, .register_dev = 1, @@ -76,21 +74,19 @@ static struct omap_mmc_config generic_mmc_config __initdata = { #endif +static struct omap_uart_config generic_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1) | (1 << 2)), +}; + static struct omap_board_config_kernel generic_config[] = { { OMAP_TAG_USB, NULL }, { OMAP_TAG_MMC, &generic_mmc_config }, + { OMAP_TAG_UART, &generic_uart_config }, }; static void __init omap_generic_init(void) { - const struct omap_uart_config *uart_conf; - - /* - * Make sure the serial ports are muxed on at this point. - * You have to mux them off in device drivers later on - * if not needed. - */ -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { generic_config[0].data = &generic1510_usb_config; } @@ -101,20 +97,9 @@ static void __init omap_generic_init(void) } #endif - uart_conf = omap_get_config(OMAP_TAG_UART, struct omap_uart_config); - if (uart_conf != NULL) { - unsigned int enabled_ports, i; - - enabled_ports = uart_conf->enabled_uarts; - for (i = 0; i < 3; i++) { - if (!(enabled_ports & (1 << i))) - generic_serial_ports[i] = 0; - } - } - omap_board_config = generic_config; omap_board_config_size = ARRAY_SIZE(generic_config); - omap_serial_init(generic_serial_ports); + omap_serial_init(); } static void __init omap_generic_map_io(void) diff --git a/arch/arm/mach-omap1/board-h2.c b/arch/arm/mach-omap1/board-h2.c index 4ee6bd8a50b8..a07e2c9307fa 100644 --- a/arch/arm/mach-omap1/board-h2.c +++ b/arch/arm/mach-omap1/board-h2.c @@ -40,8 +40,6 @@ extern int omap_gpio_init(void); -static int __initdata h2_serial_ports[OMAP_MAX_NR_PORTS] = {1, 1, 1}; - static struct mtd_partition h2_partitions[] = { /* bootloader (U-Boot, etc) in first sector */ { @@ -160,9 +158,20 @@ static struct omap_mmc_config h2_mmc_config __initdata = { }, }; +static struct omap_uart_config h2_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1) | (1 << 2)), +}; + +static struct omap_lcd_config h2_lcd_config __initdata = { + .panel_name = "h2", + .ctrl_name = "internal", +}; + static struct omap_board_config_kernel h2_config[] = { { OMAP_TAG_USB, &h2_usb_config }, { OMAP_TAG_MMC, &h2_mmc_config }, + { OMAP_TAG_UART, &h2_uart_config }, + { OMAP_TAG_LCD, &h2_lcd_config }, }; static void __init h2_init(void) @@ -180,12 +189,12 @@ static void __init h2_init(void) platform_add_devices(h2_devices, ARRAY_SIZE(h2_devices)); omap_board_config = h2_config; omap_board_config_size = ARRAY_SIZE(h2_config); + omap_serial_init(); } static void __init h2_map_io(void) { omap_map_common_io(); - omap_serial_init(h2_serial_ports); } MACHINE_START(OMAP_H2, "TI-H2") diff --git a/arch/arm/mach-omap1/board-h3.c b/arch/arm/mach-omap1/board-h3.c index fc824361430d..668e278433c2 100644 --- a/arch/arm/mach-omap1/board-h3.c +++ b/arch/arm/mach-omap1/board-h3.c @@ -41,8 +41,6 @@ extern int omap_gpio_init(void); -static int __initdata h3_serial_ports[OMAP_MAX_NR_PORTS] = {1, 1, 1}; - static struct mtd_partition h3_partitions[] = { /* bootloader (U-Boot, etc) in first sector */ { @@ -168,9 +166,20 @@ static struct omap_mmc_config h3_mmc_config __initdata = { }, }; +static struct omap_uart_config h3_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1) | (1 << 2)), +}; + +static struct omap_lcd_config h3_lcd_config __initdata = { + .panel_name = "h3", + .ctrl_name = "internal", +}; + static struct omap_board_config_kernel h3_config[] = { - { OMAP_TAG_USB, &h3_usb_config }, - { OMAP_TAG_MMC, &h3_mmc_config }, + { OMAP_TAG_USB, &h3_usb_config }, + { OMAP_TAG_MMC, &h3_mmc_config }, + { OMAP_TAG_UART, &h3_uart_config }, + { OMAP_TAG_LCD, &h3_lcd_config }, }; static void __init h3_init(void) @@ -180,6 +189,7 @@ static void __init h3_init(void) (void) platform_add_devices(devices, ARRAY_SIZE(devices)); omap_board_config = h3_config; omap_board_config_size = ARRAY_SIZE(h3_config); + omap_serial_init(); } static void __init h3_init_smc91x(void) @@ -201,7 +211,6 @@ void h3_init_irq(void) static void __init h3_map_io(void) { omap_map_common_io(); - omap_serial_init(h3_serial_ports); } MACHINE_START(OMAP_H3, "TI OMAP1710 H3 board") diff --git a/arch/arm/mach-omap1/board-innovator.c b/arch/arm/mach-omap1/board-innovator.c index a2eac853b2da..95f1ff36cdcb 100644 --- a/arch/arm/mach-omap1/board-innovator.c +++ b/arch/arm/mach-omap1/board-innovator.c @@ -36,8 +36,6 @@ #include #include -static int __initdata innovator_serial_ports[OMAP_MAX_NR_PORTS] = {1, 1, 1}; - static struct mtd_partition innovator_partitions[] = { /* bootloader (U-Boot, etc) in first sector */ { @@ -99,7 +97,7 @@ static struct platform_device innovator_flash_device = { .resource = &innovator_flash_resource, }; -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX /* Only FPGA needs to be mapped here. All others are done with ioremap */ static struct map_desc innovator1510_io_desc[] __initdata = { @@ -136,7 +134,7 @@ static struct platform_device *innovator1510_devices[] __initdata = { &innovator1510_smc91x_device, }; -#endif /* CONFIG_ARCH_OMAP1510 */ +#endif /* CONFIG_ARCH_OMAP15XX */ #ifdef CONFIG_ARCH_OMAP16XX @@ -185,7 +183,7 @@ void innovator_init_irq(void) { omap_init_irq(); omap_gpio_init(); -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { omap1510_fpga_init_irq(); } @@ -193,7 +191,7 @@ void innovator_init_irq(void) innovator_init_smc91x(); } -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX static struct omap_usb_config innovator1510_usb_config __initdata = { /* for bundled non-standard host and peripheral cables */ .hmc_mode = 4, @@ -205,6 +203,11 @@ static struct omap_usb_config innovator1510_usb_config __initdata = { .register_dev = 1, .pins[0] = 2, }; + +static struct omap_lcd_config innovator1510_lcd_config __initdata = { + .panel_name = "inn1510", + .ctrl_name = "internal", +}; #endif #ifdef CONFIG_ARCH_OMAP16XX @@ -222,6 +225,11 @@ static struct omap_usb_config h2_usb_config __initdata = { .pins[1] = 3, }; + +static struct omap_lcd_config innovator1610_lcd_config __initdata = { + .panel_name = "inn1610", + .ctrl_name = "internal", +}; #endif static struct omap_mmc_config innovator_mmc_config __initdata = { @@ -234,14 +242,20 @@ static struct omap_mmc_config innovator_mmc_config __initdata = { }, }; +static struct omap_uart_config innovator_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1) | (1 << 2)), +}; + static struct omap_board_config_kernel innovator_config[] = { { OMAP_TAG_USB, NULL }, + { OMAP_TAG_LCD, NULL }, { OMAP_TAG_MMC, &innovator_mmc_config }, + { OMAP_TAG_UART, &innovator_uart_config }, }; static void __init innovator_init(void) { -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { platform_add_devices(innovator1510_devices, ARRAY_SIZE(innovator1510_devices)); } @@ -252,23 +266,28 @@ static void __init innovator_init(void) } #endif -#ifdef CONFIG_ARCH_OMAP1510 - if (cpu_is_omap1510()) +#ifdef CONFIG_ARCH_OMAP15XX + if (cpu_is_omap1510()) { innovator_config[0].data = &innovator1510_usb_config; + innovator_config[1].data = &innovator1510_lcd_config; + } #endif #ifdef CONFIG_ARCH_OMAP16XX - if (cpu_is_omap1610()) + if (cpu_is_omap1610()) { innovator_config[0].data = &h2_usb_config; + innovator_config[1].data = &innovator1610_lcd_config; + } #endif omap_board_config = innovator_config; omap_board_config_size = ARRAY_SIZE(innovator_config); + omap_serial_init(); } static void __init innovator_map_io(void) { omap_map_common_io(); -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { iotable_init(innovator1510_io_desc, ARRAY_SIZE(innovator1510_io_desc)); udelay(10); /* Delay needed for FPGA */ @@ -280,7 +299,6 @@ static void __init innovator_map_io(void) fpga_read(OMAP1510_FPGA_BOARD_REV)); } #endif - omap_serial_init(innovator_serial_ports); } MACHINE_START(OMAP_INNOVATOR, "TI-Innovator") diff --git a/arch/arm/mach-omap1/board-netstar.c b/arch/arm/mach-omap1/board-netstar.c index c851c2e4dfcb..0448fa7de8a4 100644 --- a/arch/arm/mach-omap1/board-netstar.c +++ b/arch/arm/mach-omap1/board-netstar.c @@ -55,6 +55,14 @@ static struct platform_device *netstar_devices[] __initdata = { &netstar_smc91x_device, }; +static struct omap_uart_config netstar_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1) | (1 << 2)), +}; + +static struct omap_board_config_kernel netstar_config[] = { + { OMAP_TAG_UART, &netstar_uart_config }, +}; + static void __init netstar_init_irq(void) { omap_init_irq(); @@ -92,14 +100,15 @@ static void __init netstar_init(void) /* Switch off red LED */ omap_writeb(0x00, OMAP_LPG1_PMR); /* Disable clock */ omap_writeb(0x80, OMAP_LPG1_LCR); -} -static int __initdata omap_serial_ports[OMAP_MAX_NR_PORTS] = {1, 1, 1}; + omap_board_config = netstar_config; + omap_board_config_size = ARRAY_SIZE(netstar_config); + omap_serial_init(); +} static void __init netstar_map_io(void) { omap_map_common_io(); - omap_serial_init(omap_serial_ports); } #define MACHINE_PANICED 1 diff --git a/arch/arm/mach-omap1/board-osk.c b/arch/arm/mach-omap1/board-osk.c index a88524e7c315..e990e1bc1669 100644 --- a/arch/arm/mach-omap1/board-osk.c +++ b/arch/arm/mach-omap1/board-osk.c @@ -46,8 +46,6 @@ #include #include -static int __initdata osk_serial_ports[OMAP_MAX_NR_PORTS] = {1, 0, 0}; - static struct mtd_partition osk_partitions[] = { /* bootloader (U-Boot, etc) in first sector */ { @@ -155,7 +153,7 @@ static void __init osk_init_smc91x(void) } /* Check EMIFS wait states to fix errors with SMC_GET_PKT_HDR */ - EMIFS_CCS(1) |= 0x2; + EMIFS_CCS(1) |= 0x3; } static void __init osk_init_cf(void) @@ -193,8 +191,19 @@ static struct omap_usb_config osk_usb_config __initdata = { .pins[0] = 2, }; +static struct omap_uart_config osk_uart_config __initdata = { + .enabled_uarts = (1 << 0), +}; + +static struct omap_lcd_config osk_lcd_config __initdata = { + .panel_name = "osk", + .ctrl_name = "internal", +}; + static struct omap_board_config_kernel osk_config[] = { { OMAP_TAG_USB, &osk_usb_config }, + { OMAP_TAG_UART, &osk_uart_config }, + { OMAP_TAG_LCD, &osk_lcd_config }, }; #ifdef CONFIG_OMAP_OSK_MISTRAL @@ -254,13 +263,13 @@ static void __init osk_init(void) omap_board_config_size = ARRAY_SIZE(osk_config); USB_TRANSCEIVER_CTRL_REG |= (3 << 1); + omap_serial_init(); osk_mistral_init(); } static void __init osk_map_io(void) { omap_map_common_io(); - omap_serial_init(osk_serial_ports); } MACHINE_START(OMAP_OSK, "TI-OSK") diff --git a/arch/arm/mach-omap1/board-palmte.c b/arch/arm/mach-omap1/board-palmte.c new file mode 100644 index 000000000000..540b20d78cca --- /dev/null +++ b/arch/arm/mach-omap1/board-palmte.c @@ -0,0 +1,87 @@ +/* + * linux/arch/arm/mach-omap1/board-palmte.c + * + * Modified from board-generic.c + * + * Support for the Palm Tungsten E PDA. + * + * Original version : Laurent Gonzalez + * + * Maintainters : http://palmtelinux.sf.net + * palmtelinux-developpers@lists.sf.net + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include +#include +#include +#include + +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +static void __init omap_generic_init_irq(void) +{ + omap_init_irq(); +} + +static struct omap_usb_config palmte_usb_config __initdata = { + .register_dev = 1, + .hmc_mode = 0, + .pins[0] = 3, +}; + +static struct omap_mmc_config palmte_mmc_config __initdata = { + .mmc [0] = { + .enabled = 1, + .wire4 = 1, + .wp_pin = OMAP_MPUIO(3), + .power_pin = -1, + .switch_pin = -1, + }, +}; + +static struct omap_lcd_config palmte_lcd_config __initdata = { + .panel_name = "palmte", + .ctrl_name = "internal", +}; + +static struct omap_board_config_kernel palmte_config[] = { + { OMAP_TAG_USB, &palmte_usb_config }, + { OMAP_TAG_MMC, &palmte_mmc_config }, + { OMAP_TAG_LCD, &palmte_lcd_config }, +}; + +static void __init omap_generic_init(void) +{ + omap_board_config = palmte_config; + omap_board_config_size = ARRAY_SIZE(palmte_config); +} + +static void __init omap_generic_map_io(void) +{ + omap_map_common_io(); +} + +MACHINE_START(OMAP_PALMTE, "OMAP310 based Palm Tungsten E") + .phys_ram = 0x10000000, + .phys_io = 0xfff00000, + .io_pg_offst = ((0xfef00000) >> 18) & 0xfffc, + .boot_params = 0x10000100, + .map_io = omap_generic_map_io, + .init_irq = omap_generic_init_irq, + .init_machine = omap_generic_init, + .timer = &omap_timer, +MACHINE_END diff --git a/arch/arm/mach-omap1/board-perseus2.c b/arch/arm/mach-omap1/board-perseus2.c index 354b157acb3a..bd900b7ab33c 100644 --- a/arch/arm/mach-omap1/board-perseus2.c +++ b/arch/arm/mach-omap1/board-perseus2.c @@ -29,6 +29,7 @@ #include #include #include +#include static struct resource smc91x_resources[] = { [0] = { @@ -43,8 +44,6 @@ static struct resource smc91x_resources[] = { }, }; -static int __initdata p2_serial_ports[OMAP_MAX_NR_PORTS] = {1, 1, 0}; - static struct mtd_partition p2_partitions[] = { /* bootloader (U-Boot, etc) in first sector */ { @@ -111,9 +110,27 @@ static struct platform_device *devices[] __initdata = { &smc91x_device, }; +static struct omap_uart_config perseus2_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1)), +}; + +static struct omap_lcd_config perseus2_lcd_config __initdata = { + .panel_name = "p2", + .ctrl_name = "internal", +}; + +static struct omap_board_config_kernel perseus2_config[] = { + { OMAP_TAG_UART, &perseus2_uart_config }, + { OMAP_TAG_LCD, &perseus2_lcd_config }, +}; + static void __init omap_perseus2_init(void) { (void) platform_add_devices(devices, ARRAY_SIZE(devices)); + + omap_board_config = perseus2_config; + omap_board_config_size = ARRAY_SIZE(perseus2_config); + omap_serial_init(); } static void __init perseus2_init_smc91x(void) @@ -131,7 +148,6 @@ void omap_perseus2_init_irq(void) omap_gpio_init(); perseus2_init_smc91x(); } - /* Only FPGA needs to be mapped here. All others are done with ioremap */ static struct map_desc omap_perseus2_io_desc[] __initdata = { { @@ -179,7 +195,6 @@ static void __init omap_perseus2_map_io(void) * It is used as the Ethernet controller interrupt */ omap_writel(omap_readl(OMAP730_IO_CONF_9) & 0x1FFFFFFF, OMAP730_IO_CONF_9); - omap_serial_init(p2_serial_ports); } MACHINE_START(OMAP_PERSEUS2, "OMAP730 Perseus2") diff --git a/arch/arm/mach-omap1/board-voiceblue.c b/arch/arm/mach-omap1/board-voiceblue.c index 3f018b296861..6f9a6220e78a 100644 --- a/arch/arm/mach-omap1/board-voiceblue.c +++ b/arch/arm/mach-omap1/board-voiceblue.c @@ -150,9 +150,14 @@ static struct omap_mmc_config voiceblue_mmc_config __initdata = { }, }; +static struct omap_uart_config voiceblue_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1) | (1 << 2)), +}; + static struct omap_board_config_kernel voiceblue_config[] = { { OMAP_TAG_USB, &voiceblue_usb_config }, { OMAP_TAG_MMC, &voiceblue_mmc_config }, + { OMAP_TAG_UART, &voiceblue_uart_config }, }; static void __init voiceblue_init_irq(void) @@ -191,6 +196,7 @@ static void __init voiceblue_init(void) platform_add_devices(voiceblue_devices, ARRAY_SIZE(voiceblue_devices)); omap_board_config = voiceblue_config; omap_board_config_size = ARRAY_SIZE(voiceblue_config); + omap_serial_init(); /* There is a good chance board is going up, so enable power LED * (it is connected through invertor) */ @@ -198,12 +204,9 @@ static void __init voiceblue_init(void) omap_writeb(0x00, OMAP_LPG1_PMR); /* Disable clock */ } -static int __initdata omap_serial_ports[OMAP_MAX_NR_PORTS] = {1, 1, 1}; - static void __init voiceblue_map_io(void) { omap_map_common_io(); - omap_serial_init(omap_serial_ports); } #define MACHINE_PANICED 1 diff --git a/arch/arm/mach-omap1/clock.c b/arch/arm/mach-omap1/clock.c new file mode 100644 index 000000000000..4277eee44ed5 --- /dev/null +++ b/arch/arm/mach-omap1/clock.c @@ -0,0 +1,792 @@ +/* + * linux/arch/arm/mach-omap1/clock.c + * + * Copyright (C) 2004 - 2005 Nokia corporation + * Written by Tuukka Tikkanen + * + * Modified to use omap shared clock framework by + * Tony Lindgren + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include +#include +#include +#include +#include + +#include +#include + +#include +#include +#include + +#include "clock.h" + +__u32 arm_idlect1_mask; + +/*------------------------------------------------------------------------- + * Omap1 specific clock functions + *-------------------------------------------------------------------------*/ + +static void omap1_watchdog_recalc(struct clk * clk) +{ + clk->rate = clk->parent->rate / 14; +} + +static void omap1_uart_recalc(struct clk * clk) +{ + unsigned int val = omap_readl(clk->enable_reg); + if (val & clk->enable_bit) + clk->rate = 48000000; + else + clk->rate = 12000000; +} + +static int omap1_clk_enable_dsp_domain(struct clk *clk) +{ + int retval; + + retval = omap1_clk_use(&api_ck.clk); + if (!retval) { + retval = omap1_clk_enable(clk); + omap1_clk_unuse(&api_ck.clk); + } + + return retval; +} + +static void omap1_clk_disable_dsp_domain(struct clk *clk) +{ + if (omap1_clk_use(&api_ck.clk) == 0) { + omap1_clk_disable(clk); + omap1_clk_unuse(&api_ck.clk); + } +} + +static int omap1_clk_enable_uart_functional(struct clk *clk) +{ + int ret; + struct uart_clk *uclk; + + ret = omap1_clk_enable(clk); + if (ret == 0) { + /* Set smart idle acknowledgement mode */ + uclk = (struct uart_clk *)clk; + omap_writeb((omap_readb(uclk->sysc_addr) & ~0x10) | 8, + uclk->sysc_addr); + } + + return ret; +} + +static void omap1_clk_disable_uart_functional(struct clk *clk) +{ + struct uart_clk *uclk; + + /* Set force idle acknowledgement mode */ + uclk = (struct uart_clk *)clk; + omap_writeb((omap_readb(uclk->sysc_addr) & ~0x18), uclk->sysc_addr); + + omap1_clk_disable(clk); +} + +static void omap1_clk_allow_idle(struct clk *clk) +{ + struct arm_idlect1_clk * iclk = (struct arm_idlect1_clk *)clk; + + if (!(clk->flags & CLOCK_IDLE_CONTROL)) + return; + + if (iclk->no_idle_count > 0 && !(--iclk->no_idle_count)) + arm_idlect1_mask |= 1 << iclk->idlect_shift; +} + +static void omap1_clk_deny_idle(struct clk *clk) +{ + struct arm_idlect1_clk * iclk = (struct arm_idlect1_clk *)clk; + + if (!(clk->flags & CLOCK_IDLE_CONTROL)) + return; + + if (iclk->no_idle_count++ == 0) + arm_idlect1_mask &= ~(1 << iclk->idlect_shift); +} + +static __u16 verify_ckctl_value(__u16 newval) +{ + /* This function checks for following limitations set + * by the hardware (all conditions must be true): + * DSPMMU_CK == DSP_CK or DSPMMU_CK == DSP_CK/2 + * ARM_CK >= TC_CK + * DSP_CK >= TC_CK + * DSPMMU_CK >= TC_CK + * + * In addition following rules are enforced: + * LCD_CK <= TC_CK + * ARMPER_CK <= TC_CK + * + * However, maximum frequencies are not checked for! + */ + __u8 per_exp; + __u8 lcd_exp; + __u8 arm_exp; + __u8 dsp_exp; + __u8 tc_exp; + __u8 dspmmu_exp; + + per_exp = (newval >> CKCTL_PERDIV_OFFSET) & 3; + lcd_exp = (newval >> CKCTL_LCDDIV_OFFSET) & 3; + arm_exp = (newval >> CKCTL_ARMDIV_OFFSET) & 3; + dsp_exp = (newval >> CKCTL_DSPDIV_OFFSET) & 3; + tc_exp = (newval >> CKCTL_TCDIV_OFFSET) & 3; + dspmmu_exp = (newval >> CKCTL_DSPMMUDIV_OFFSET) & 3; + + if (dspmmu_exp < dsp_exp) + dspmmu_exp = dsp_exp; + if (dspmmu_exp > dsp_exp+1) + dspmmu_exp = dsp_exp+1; + if (tc_exp < arm_exp) + tc_exp = arm_exp; + if (tc_exp < dspmmu_exp) + tc_exp = dspmmu_exp; + if (tc_exp > lcd_exp) + lcd_exp = tc_exp; + if (tc_exp > per_exp) + per_exp = tc_exp; + + newval &= 0xf000; + newval |= per_exp << CKCTL_PERDIV_OFFSET; + newval |= lcd_exp << CKCTL_LCDDIV_OFFSET; + newval |= arm_exp << CKCTL_ARMDIV_OFFSET; + newval |= dsp_exp << CKCTL_DSPDIV_OFFSET; + newval |= tc_exp << CKCTL_TCDIV_OFFSET; + newval |= dspmmu_exp << CKCTL_DSPMMUDIV_OFFSET; + + return newval; +} + +static int calc_dsor_exp(struct clk *clk, unsigned long rate) +{ + /* Note: If target frequency is too low, this function will return 4, + * which is invalid value. Caller must check for this value and act + * accordingly. + * + * Note: This function does not check for following limitations set + * by the hardware (all conditions must be true): + * DSPMMU_CK == DSP_CK or DSPMMU_CK == DSP_CK/2 + * ARM_CK >= TC_CK + * DSP_CK >= TC_CK + * DSPMMU_CK >= TC_CK + */ + unsigned long realrate; + struct clk * parent; + unsigned dsor_exp; + + if (unlikely(!(clk->flags & RATE_CKCTL))) + return -EINVAL; + + parent = clk->parent; + if (unlikely(parent == 0)) + return -EIO; + + realrate = parent->rate; + for (dsor_exp=0; dsor_exp<4; dsor_exp++) { + if (realrate <= rate) + break; + + realrate /= 2; + } + + return dsor_exp; +} + +static void omap1_ckctl_recalc(struct clk * clk) +{ + int dsor; + + /* Calculate divisor encoded as 2-bit exponent */ + dsor = 1 << (3 & (omap_readw(ARM_CKCTL) >> clk->rate_offset)); + + if (unlikely(clk->rate == clk->parent->rate / dsor)) + return; /* No change, quick exit */ + clk->rate = clk->parent->rate / dsor; + + if (unlikely(clk->flags & RATE_PROPAGATES)) + propagate_rate(clk); +} + +static void omap1_ckctl_recalc_dsp_domain(struct clk * clk) +{ + int dsor; + + /* Calculate divisor encoded as 2-bit exponent + * + * The clock control bits are in DSP domain, + * so api_ck is needed for access. + * Note that DSP_CKCTL virt addr = phys addr, so + * we must use __raw_readw() instead of omap_readw(). + */ + omap1_clk_use(&api_ck.clk); + dsor = 1 << (3 & (__raw_readw(DSP_CKCTL) >> clk->rate_offset)); + omap1_clk_unuse(&api_ck.clk); + + if (unlikely(clk->rate == clk->parent->rate / dsor)) + return; /* No change, quick exit */ + clk->rate = clk->parent->rate / dsor; + + if (unlikely(clk->flags & RATE_PROPAGATES)) + propagate_rate(clk); +} + +/* MPU virtual clock functions */ +static int omap1_select_table_rate(struct clk * clk, unsigned long rate) +{ + /* Find the highest supported frequency <= rate and switch to it */ + struct mpu_rate * ptr; + + if (clk != &virtual_ck_mpu) + return -EINVAL; + + for (ptr = rate_table; ptr->rate; ptr++) { + if (ptr->xtal != ck_ref.rate) + continue; + + /* DPLL1 cannot be reprogrammed without risking system crash */ + if (likely(ck_dpll1.rate!=0) && ptr->pll_rate != ck_dpll1.rate) + continue; + + /* Can check only after xtal frequency check */ + if (ptr->rate <= rate) + break; + } + + if (!ptr->rate) + return -EINVAL; + + /* + * In most cases we should not need to reprogram DPLL. + * Reprogramming the DPLL is tricky, it must be done from SRAM. + */ + omap_sram_reprogram_clock(ptr->dpllctl_val, ptr->ckctl_val); + + ck_dpll1.rate = ptr->pll_rate; + propagate_rate(&ck_dpll1); + return 0; +} + +static int omap1_clk_set_rate_dsp_domain(struct clk *clk, unsigned long rate) +{ + int ret = -EINVAL; + int dsor_exp; + __u16 regval; + + if (clk->flags & RATE_CKCTL) { + dsor_exp = calc_dsor_exp(clk, rate); + if (dsor_exp > 3) + dsor_exp = -EINVAL; + if (dsor_exp < 0) + return dsor_exp; + + regval = __raw_readw(DSP_CKCTL); + regval &= ~(3 << clk->rate_offset); + regval |= dsor_exp << clk->rate_offset; + __raw_writew(regval, DSP_CKCTL); + clk->rate = clk->parent->rate / (1 << dsor_exp); + ret = 0; + } + + if (unlikely(ret == 0 && (clk->flags & RATE_PROPAGATES))) + propagate_rate(clk); + + return ret; +} + +static long omap1_round_to_table_rate(struct clk * clk, unsigned long rate) +{ + /* Find the highest supported frequency <= rate */ + struct mpu_rate * ptr; + long highest_rate; + + if (clk != &virtual_ck_mpu) + return -EINVAL; + + highest_rate = -EINVAL; + + for (ptr = rate_table; ptr->rate; ptr++) { + if (ptr->xtal != ck_ref.rate) + continue; + + highest_rate = ptr->rate; + + /* Can check only after xtal frequency check */ + if (ptr->rate <= rate) + break; + } + + return highest_rate; +} + +static unsigned calc_ext_dsor(unsigned long rate) +{ + unsigned dsor; + + /* MCLK and BCLK divisor selection is not linear: + * freq = 96MHz / dsor + * + * RATIO_SEL range: dsor <-> RATIO_SEL + * 0..6: (RATIO_SEL+2) <-> (dsor-2) + * 6..48: (8+(RATIO_SEL-6)*2) <-> ((dsor-8)/2+6) + * Minimum dsor is 2 and maximum is 96. Odd divisors starting from 9 + * can not be used. + */ + for (dsor = 2; dsor < 96; ++dsor) { + if ((dsor & 1) && dsor > 8) + continue; + if (rate >= 96000000 / dsor) + break; + } + return dsor; +} + +/* Only needed on 1510 */ +static int omap1_set_uart_rate(struct clk * clk, unsigned long rate) +{ + unsigned int val; + + val = omap_readl(clk->enable_reg); + if (rate == 12000000) + val &= ~(1 << clk->enable_bit); + else if (rate == 48000000) + val |= (1 << clk->enable_bit); + else + return -EINVAL; + omap_writel(val, clk->enable_reg); + clk->rate = rate; + + return 0; +} + +/* External clock (MCLK & BCLK) functions */ +static int omap1_set_ext_clk_rate(struct clk * clk, unsigned long rate) +{ + unsigned dsor; + __u16 ratio_bits; + + dsor = calc_ext_dsor(rate); + clk->rate = 96000000 / dsor; + if (dsor > 8) + ratio_bits = ((dsor - 8) / 2 + 6) << 2; + else + ratio_bits = (dsor - 2) << 2; + + ratio_bits |= omap_readw(clk->enable_reg) & ~0xfd; + omap_writew(ratio_bits, clk->enable_reg); + + return 0; +} + +static long omap1_round_ext_clk_rate(struct clk * clk, unsigned long rate) +{ + return 96000000 / calc_ext_dsor(rate); +} + +static void omap1_init_ext_clk(struct clk * clk) +{ + unsigned dsor; + __u16 ratio_bits; + + /* Determine current rate and ensure clock is based on 96MHz APLL */ + ratio_bits = omap_readw(clk->enable_reg) & ~1; + omap_writew(ratio_bits, clk->enable_reg); + + ratio_bits = (ratio_bits & 0xfc) >> 2; + if (ratio_bits > 6) + dsor = (ratio_bits - 6) * 2 + 8; + else + dsor = ratio_bits + 2; + + clk-> rate = 96000000 / dsor; +} + +static int omap1_clk_use(struct clk *clk) +{ + int ret = 0; + if (clk->usecount++ == 0) { + if (likely(clk->parent)) { + ret = omap1_clk_use(clk->parent); + + if (unlikely(ret != 0)) { + clk->usecount--; + return ret; + } + + if (clk->flags & CLOCK_NO_IDLE_PARENT) + if (!cpu_is_omap24xx()) + omap1_clk_deny_idle(clk->parent); + } + + ret = clk->enable(clk); + + if (unlikely(ret != 0) && clk->parent) { + omap1_clk_unuse(clk->parent); + clk->usecount--; + } + } + + return ret; +} + +static void omap1_clk_unuse(struct clk *clk) +{ + if (clk->usecount > 0 && !(--clk->usecount)) { + clk->disable(clk); + if (likely(clk->parent)) { + omap1_clk_unuse(clk->parent); + if (clk->flags & CLOCK_NO_IDLE_PARENT) + if (!cpu_is_omap24xx()) + omap1_clk_allow_idle(clk->parent); + } + } +} + +static int omap1_clk_enable(struct clk *clk) +{ + __u16 regval16; + __u32 regval32; + + if (clk->flags & ALWAYS_ENABLED) + return 0; + + if (unlikely(clk->enable_reg == 0)) { + printk(KERN_ERR "clock.c: Enable for %s without enable code\n", + clk->name); + return 0; + } + + if (clk->flags & ENABLE_REG_32BIT) { + if (clk->flags & VIRTUAL_IO_ADDRESS) { + regval32 = __raw_readl(clk->enable_reg); + regval32 |= (1 << clk->enable_bit); + __raw_writel(regval32, clk->enable_reg); + } else { + regval32 = omap_readl(clk->enable_reg); + regval32 |= (1 << clk->enable_bit); + omap_writel(regval32, clk->enable_reg); + } + } else { + if (clk->flags & VIRTUAL_IO_ADDRESS) { + regval16 = __raw_readw(clk->enable_reg); + regval16 |= (1 << clk->enable_bit); + __raw_writew(regval16, clk->enable_reg); + } else { + regval16 = omap_readw(clk->enable_reg); + regval16 |= (1 << clk->enable_bit); + omap_writew(regval16, clk->enable_reg); + } + } + + return 0; +} + +static void omap1_clk_disable(struct clk *clk) +{ + __u16 regval16; + __u32 regval32; + + if (clk->enable_reg == 0) + return; + + if (clk->flags & ENABLE_REG_32BIT) { + if (clk->flags & VIRTUAL_IO_ADDRESS) { + regval32 = __raw_readl(clk->enable_reg); + regval32 &= ~(1 << clk->enable_bit); + __raw_writel(regval32, clk->enable_reg); + } else { + regval32 = omap_readl(clk->enable_reg); + regval32 &= ~(1 << clk->enable_bit); + omap_writel(regval32, clk->enable_reg); + } + } else { + if (clk->flags & VIRTUAL_IO_ADDRESS) { + regval16 = __raw_readw(clk->enable_reg); + regval16 &= ~(1 << clk->enable_bit); + __raw_writew(regval16, clk->enable_reg); + } else { + regval16 = omap_readw(clk->enable_reg); + regval16 &= ~(1 << clk->enable_bit); + omap_writew(regval16, clk->enable_reg); + } + } +} + +static long omap1_clk_round_rate(struct clk *clk, unsigned long rate) +{ + int dsor_exp; + + if (clk->flags & RATE_FIXED) + return clk->rate; + + if (clk->flags & RATE_CKCTL) { + dsor_exp = calc_dsor_exp(clk, rate); + if (dsor_exp < 0) + return dsor_exp; + if (dsor_exp > 3) + dsor_exp = 3; + return clk->parent->rate / (1 << dsor_exp); + } + + if(clk->round_rate != 0) + return clk->round_rate(clk, rate); + + return clk->rate; +} + +static int omap1_clk_set_rate(struct clk *clk, unsigned long rate) +{ + int ret = -EINVAL; + int dsor_exp; + __u16 regval; + + if (clk->set_rate) + ret = clk->set_rate(clk, rate); + else if (clk->flags & RATE_CKCTL) { + dsor_exp = calc_dsor_exp(clk, rate); + if (dsor_exp > 3) + dsor_exp = -EINVAL; + if (dsor_exp < 0) + return dsor_exp; + + regval = omap_readw(ARM_CKCTL); + regval &= ~(3 << clk->rate_offset); + regval |= dsor_exp << clk->rate_offset; + regval = verify_ckctl_value(regval); + omap_writew(regval, ARM_CKCTL); + clk->rate = clk->parent->rate / (1 << dsor_exp); + ret = 0; + } + + if (unlikely(ret == 0 && (clk->flags & RATE_PROPAGATES))) + propagate_rate(clk); + + return ret; +} + +/*------------------------------------------------------------------------- + * Omap1 clock reset and init functions + *-------------------------------------------------------------------------*/ + +#ifdef CONFIG_OMAP_RESET_CLOCKS +/* + * Resets some clocks that may be left on from bootloader, + * but leaves serial clocks on. See also omap_late_clk_reset(). + */ +static inline void omap1_early_clk_reset(void) +{ + //omap_writel(0x3 << 29, MOD_CONF_CTRL_0); +} + +static int __init omap1_late_clk_reset(void) +{ + /* Turn off all unused clocks */ + struct clk *p; + __u32 regval32; + + /* USB_REQ_EN will be disabled later if necessary (usb_dc_ck) */ + regval32 = omap_readw(SOFT_REQ_REG) & (1 << 4); + omap_writew(regval32, SOFT_REQ_REG); + omap_writew(0, SOFT_REQ_REG2); + + list_for_each_entry(p, &clocks, node) { + if (p->usecount > 0 || (p->flags & ALWAYS_ENABLED) || + p->enable_reg == 0) + continue; + + /* Clocks in the DSP domain need api_ck. Just assume bootloader + * has not enabled any DSP clocks */ + if ((u32)p->enable_reg == DSP_IDLECT2) { + printk(KERN_INFO "Skipping reset check for DSP domain " + "clock \"%s\"\n", p->name); + continue; + } + + /* Is the clock already disabled? */ + if (p->flags & ENABLE_REG_32BIT) { + if (p->flags & VIRTUAL_IO_ADDRESS) + regval32 = __raw_readl(p->enable_reg); + else + regval32 = omap_readl(p->enable_reg); + } else { + if (p->flags & VIRTUAL_IO_ADDRESS) + regval32 = __raw_readw(p->enable_reg); + else + regval32 = omap_readw(p->enable_reg); + } + + if ((regval32 & (1 << p->enable_bit)) == 0) + continue; + + /* FIXME: This clock seems to be necessary but no-one + * has asked for its activation. */ + if (p == &tc2_ck // FIX: pm.c (SRAM), CCP, Camera + || p == &ck_dpll1out.clk // FIX: SoSSI, SSR + || p == &arm_gpio_ck // FIX: GPIO code for 1510 + ) { + printk(KERN_INFO "FIXME: Clock \"%s\" seems unused\n", + p->name); + continue; + } + + printk(KERN_INFO "Disabling unused clock \"%s\"... ", p->name); + p->disable(p); + printk(" done\n"); + } + + return 0; +} +late_initcall(omap1_late_clk_reset); + +#else +#define omap1_early_clk_reset() {} +#endif + +static struct clk_functions omap1_clk_functions = { + .clk_use = omap1_clk_use, + .clk_unuse = omap1_clk_unuse, + .clk_round_rate = omap1_clk_round_rate, + .clk_set_rate = omap1_clk_set_rate, +}; + +int __init omap1_clk_init(void) +{ + struct clk ** clkp; + const struct omap_clock_config *info; + int crystal_type = 0; /* Default 12 MHz */ + + omap1_early_clk_reset(); + clk_init(&omap1_clk_functions); + + /* By default all idlect1 clocks are allowed to idle */ + arm_idlect1_mask = ~0; + + for (clkp = onchip_clks; clkp < onchip_clks+ARRAY_SIZE(onchip_clks); clkp++) { + if (((*clkp)->flags &CLOCK_IN_OMAP1510) && cpu_is_omap1510()) { + clk_register(*clkp); + continue; + } + + if (((*clkp)->flags &CLOCK_IN_OMAP16XX) && cpu_is_omap16xx()) { + clk_register(*clkp); + continue; + } + + if (((*clkp)->flags &CLOCK_IN_OMAP730) && cpu_is_omap730()) { + clk_register(*clkp); + continue; + } + } + + info = omap_get_config(OMAP_TAG_CLOCK, struct omap_clock_config); + if (info != NULL) { + if (!cpu_is_omap1510()) + crystal_type = info->system_clock_type; + } + +#if defined(CONFIG_ARCH_OMAP730) + ck_ref.rate = 13000000; +#elif defined(CONFIG_ARCH_OMAP16XX) + if (crystal_type == 2) + ck_ref.rate = 19200000; +#endif + + printk("Clocks: ARM_SYSST: 0x%04x DPLL_CTL: 0x%04x ARM_CKCTL: 0x%04x\n", + omap_readw(ARM_SYSST), omap_readw(DPLL_CTL), + omap_readw(ARM_CKCTL)); + + /* We want to be in syncronous scalable mode */ + omap_writew(0x1000, ARM_SYSST); + +#ifdef CONFIG_OMAP_CLOCKS_SET_BY_BOOTLOADER + /* Use values set by bootloader. Determine PLL rate and recalculate + * dependent clocks as if kernel had changed PLL or divisors. + */ + { + unsigned pll_ctl_val = omap_readw(DPLL_CTL); + + ck_dpll1.rate = ck_ref.rate; /* Base xtal rate */ + if (pll_ctl_val & 0x10) { + /* PLL enabled, apply multiplier and divisor */ + if (pll_ctl_val & 0xf80) + ck_dpll1.rate *= (pll_ctl_val & 0xf80) >> 7; + ck_dpll1.rate /= ((pll_ctl_val & 0x60) >> 5) + 1; + } else { + /* PLL disabled, apply bypass divisor */ + switch (pll_ctl_val & 0xc) { + case 0: + break; + case 0x4: + ck_dpll1.rate /= 2; + break; + default: + ck_dpll1.rate /= 4; + break; + } + } + } + propagate_rate(&ck_dpll1); +#else + /* Find the highest supported frequency and enable it */ + if (omap1_select_table_rate(&virtual_ck_mpu, ~0)) { + printk(KERN_ERR "System frequencies not set. Check your config.\n"); + /* Guess sane values (60MHz) */ + omap_writew(0x2290, DPLL_CTL); + omap_writew(0x1005, ARM_CKCTL); + ck_dpll1.rate = 60000000; + propagate_rate(&ck_dpll1); + } +#endif + /* Cache rates for clocks connected to ck_ref (not dpll1) */ + propagate_rate(&ck_ref); + printk(KERN_INFO "Clocking rate (xtal/DPLL1/MPU): " + "%ld.%01ld/%ld.%01ld/%ld.%01ld MHz\n", + ck_ref.rate / 1000000, (ck_ref.rate / 100000) % 10, + ck_dpll1.rate / 1000000, (ck_dpll1.rate / 100000) % 10, + arm_ck.rate / 1000000, (arm_ck.rate / 100000) % 10); + +#ifdef CONFIG_MACH_OMAP_PERSEUS2 + /* Select slicer output as OMAP input clock */ + omap_writew(omap_readw(OMAP730_PCC_UPLD_CTRL) & ~0x1, OMAP730_PCC_UPLD_CTRL); +#endif + + /* Turn off DSP and ARM_TIMXO. Make sure ARM_INTHCK is not divided */ + omap_writew(omap_readw(ARM_CKCTL) & 0x0fff, ARM_CKCTL); + + /* Put DSP/MPUI into reset until needed */ + omap_writew(0, ARM_RSTCT1); + omap_writew(1, ARM_RSTCT2); + omap_writew(0x400, ARM_IDLECT1); + + /* + * According to OMAP5910 Erratum SYS_DMA_1, bit DMACK_REQ (bit 8) + * of the ARM_IDLECT2 register must be set to zero. The power-on + * default value of this bit is one. + */ + omap_writew(0x0000, ARM_IDLECT2); /* Turn LCD clock off also */ + + /* + * Only enable those clocks we will need, let the drivers + * enable other clocks as necessary + */ + clk_use(&armper_ck.clk); + clk_use(&armxor_ck.clk); + clk_use(&armtim_ck.clk); /* This should be done by timer code */ + + if (cpu_is_omap1510()) + clk_enable(&arm_gpio_ck); + + return 0; +} + diff --git a/arch/arm/mach-omap1/clock.h b/arch/arm/mach-omap1/clock.h new file mode 100644 index 000000000000..f3bdfb50e01a --- /dev/null +++ b/arch/arm/mach-omap1/clock.h @@ -0,0 +1,768 @@ +/* + * linux/arch/arm/mach-omap1/clock.h + * + * Copyright (C) 2004 - 2005 Nokia corporation + * Written by Tuukka Tikkanen + * Based on clocks.h by Tony Lindgren, Gordon McNutt and RidgeRun, Inc + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __ARCH_ARM_MACH_OMAP1_CLOCK_H +#define __ARCH_ARM_MACH_OMAP1_CLOCK_H + +static int omap1_clk_enable(struct clk * clk); +static void omap1_clk_disable(struct clk * clk); +static void omap1_ckctl_recalc(struct clk * clk); +static void omap1_watchdog_recalc(struct clk * clk); +static void omap1_ckctl_recalc_dsp_domain(struct clk * clk); +static int omap1_clk_enable_dsp_domain(struct clk * clk); +static int omap1_clk_set_rate_dsp_domain(struct clk * clk, unsigned long rate); +static void omap1_clk_disable_dsp_domain(struct clk * clk); +static int omap1_set_uart_rate(struct clk * clk, unsigned long rate); +static void omap1_uart_recalc(struct clk * clk); +static int omap1_clk_enable_uart_functional(struct clk * clk); +static void omap1_clk_disable_uart_functional(struct clk * clk); +static int omap1_set_ext_clk_rate(struct clk * clk, unsigned long rate); +static long omap1_round_ext_clk_rate(struct clk * clk, unsigned long rate); +static void omap1_init_ext_clk(struct clk * clk); +static int omap1_select_table_rate(struct clk * clk, unsigned long rate); +static long omap1_round_to_table_rate(struct clk * clk, unsigned long rate); +static int omap1_clk_use(struct clk *clk); +static void omap1_clk_unuse(struct clk *clk); + +struct mpu_rate { + unsigned long rate; + unsigned long xtal; + unsigned long pll_rate; + __u16 ckctl_val; + __u16 dpllctl_val; +}; + +struct uart_clk { + struct clk clk; + unsigned long sysc_addr; +}; + +/* Provide a method for preventing idling some ARM IDLECT clocks */ +struct arm_idlect1_clk { + struct clk clk; + unsigned long no_idle_count; + __u8 idlect_shift; +}; + +/* ARM_CKCTL bit shifts */ +#define CKCTL_PERDIV_OFFSET 0 +#define CKCTL_LCDDIV_OFFSET 2 +#define CKCTL_ARMDIV_OFFSET 4 +#define CKCTL_DSPDIV_OFFSET 6 +#define CKCTL_TCDIV_OFFSET 8 +#define CKCTL_DSPMMUDIV_OFFSET 10 +/*#define ARM_TIMXO 12*/ +#define EN_DSPCK 13 +/*#define ARM_INTHCK_SEL 14*/ /* Divide-by-2 for mpu inth_ck */ +/* DSP_CKCTL bit shifts */ +#define CKCTL_DSPPERDIV_OFFSET 0 + +/* ARM_IDLECT2 bit shifts */ +#define EN_WDTCK 0 +#define EN_XORPCK 1 +#define EN_PERCK 2 +#define EN_LCDCK 3 +#define EN_LBCK 4 /* Not on 1610/1710 */ +/*#define EN_HSABCK 5*/ +#define EN_APICK 6 +#define EN_TIMCK 7 +#define DMACK_REQ 8 +#define EN_GPIOCK 9 /* Not on 1610/1710 */ +/*#define EN_LBFREECK 10*/ +#define EN_CKOUT_ARM 11 + +/* ARM_IDLECT3 bit shifts */ +#define EN_OCPI_CK 0 +#define EN_TC1_CK 2 +#define EN_TC2_CK 4 + +/* DSP_IDLECT2 bit shifts (0,1,2 are same as for ARM_IDLECT2) */ +#define EN_DSPTIMCK 5 + +/* Various register defines for clock controls scattered around OMAP chip */ +#define USB_MCLK_EN_BIT 4 /* In ULPD_CLKC_CTRL */ +#define USB_HOST_HHC_UHOST_EN 9 /* In MOD_CONF_CTRL_0 */ +#define SWD_ULPD_PLL_CLK_REQ 1 /* In SWD_CLK_DIV_CTRL_SEL */ +#define COM_ULPD_PLL_CLK_REQ 1 /* In COM_CLK_DIV_CTRL_SEL */ +#define SWD_CLK_DIV_CTRL_SEL 0xfffe0874 +#define COM_CLK_DIV_CTRL_SEL 0xfffe0878 +#define SOFT_REQ_REG 0xfffe0834 +#define SOFT_REQ_REG2 0xfffe0880 + +/*------------------------------------------------------------------------- + * Omap1 MPU rate table + *-------------------------------------------------------------------------*/ +static struct mpu_rate rate_table[] = { + /* MPU MHz, xtal MHz, dpll1 MHz, CKCTL, DPLL_CTL + * NOTE: Comment order here is different from bits in CKCTL value: + * armdiv, dspdiv, dspmmu, tcdiv, perdiv, lcddiv + */ +#if defined(CONFIG_OMAP_ARM_216MHZ) + { 216000000, 12000000, 216000000, 0x050d, 0x2910 }, /* 1/1/2/2/2/8 */ +#endif +#if defined(CONFIG_OMAP_ARM_195MHZ) + { 195000000, 13000000, 195000000, 0x050e, 0x2790 }, /* 1/1/2/2/4/8 */ +#endif +#if defined(CONFIG_OMAP_ARM_192MHZ) + { 192000000, 19200000, 192000000, 0x050f, 0x2510 }, /* 1/1/2/2/8/8 */ + { 192000000, 12000000, 192000000, 0x050f, 0x2810 }, /* 1/1/2/2/8/8 */ + { 96000000, 12000000, 192000000, 0x055f, 0x2810 }, /* 2/2/2/2/8/8 */ + { 48000000, 12000000, 192000000, 0x0baf, 0x2810 }, /* 4/4/4/8/8/8 */ + { 24000000, 12000000, 192000000, 0x0fff, 0x2810 }, /* 8/8/8/8/8/8 */ +#endif +#if defined(CONFIG_OMAP_ARM_182MHZ) + { 182000000, 13000000, 182000000, 0x050e, 0x2710 }, /* 1/1/2/2/4/8 */ +#endif +#if defined(CONFIG_OMAP_ARM_168MHZ) + { 168000000, 12000000, 168000000, 0x010f, 0x2710 }, /* 1/1/1/2/8/8 */ +#endif +#if defined(CONFIG_OMAP_ARM_150MHZ) + { 150000000, 12000000, 150000000, 0x010a, 0x2cb0 }, /* 1/1/1/2/4/4 */ +#endif +#if defined(CONFIG_OMAP_ARM_120MHZ) + { 120000000, 12000000, 120000000, 0x010a, 0x2510 }, /* 1/1/1/2/4/4 */ +#endif +#if defined(CONFIG_OMAP_ARM_96MHZ) + { 96000000, 12000000, 96000000, 0x0005, 0x2410 }, /* 1/1/1/1/2/2 */ +#endif +#if defined(CONFIG_OMAP_ARM_60MHZ) + { 60000000, 12000000, 60000000, 0x0005, 0x2290 }, /* 1/1/1/1/2/2 */ +#endif +#if defined(CONFIG_OMAP_ARM_30MHZ) + { 30000000, 12000000, 60000000, 0x0555, 0x2290 }, /* 2/2/2/2/2/2 */ +#endif + { 0, 0, 0, 0, 0 }, +}; + +/*------------------------------------------------------------------------- + * Omap1 clocks + *-------------------------------------------------------------------------*/ + +static struct clk ck_ref = { + .name = "ck_ref", + .rate = 12000000, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + ALWAYS_ENABLED, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk ck_dpll1 = { + .name = "ck_dpll1", + .parent = &ck_ref, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + RATE_PROPAGATES | ALWAYS_ENABLED, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct arm_idlect1_clk ck_dpll1out = { + .clk = { + .name = "ck_dpll1out", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP16XX | CLOCK_IDLE_CONTROL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_CKOUT_ARM, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 12, +}; + +static struct clk arm_ck = { + .name = "arm_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + RATE_CKCTL | RATE_PROPAGATES | ALWAYS_ENABLED, + .rate_offset = CKCTL_ARMDIV_OFFSET, + .recalc = &omap1_ckctl_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct arm_idlect1_clk armper_ck = { + .clk = { + .name = "armper_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + RATE_CKCTL | CLOCK_IDLE_CONTROL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_PERCK, + .rate_offset = CKCTL_PERDIV_OFFSET, + .recalc = &omap1_ckctl_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 2, +}; + +static struct clk arm_gpio_ck = { + .name = "arm_gpio_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP1510, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_GPIOCK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct arm_idlect1_clk armxor_ck = { + .clk = { + .name = "armxor_ck", + .parent = &ck_ref, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + CLOCK_IDLE_CONTROL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_XORPCK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 1, +}; + +static struct arm_idlect1_clk armtim_ck = { + .clk = { + .name = "armtim_ck", + .parent = &ck_ref, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + CLOCK_IDLE_CONTROL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_TIMCK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 9, +}; + +static struct arm_idlect1_clk armwdt_ck = { + .clk = { + .name = "armwdt_ck", + .parent = &ck_ref, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + CLOCK_IDLE_CONTROL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_WDTCK, + .recalc = &omap1_watchdog_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 0, +}; + +static struct clk arminth_ck16xx = { + .name = "arminth_ck", + .parent = &arm_ck, + .flags = CLOCK_IN_OMAP16XX | ALWAYS_ENABLED, + .recalc = &followparent_recalc, + /* Note: On 16xx the frequency can be divided by 2 by programming + * ARM_CKCTL:ARM_INTHCK_SEL(14) to 1 + * + * 1510 version is in TC clocks. + */ + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk dsp_ck = { + .name = "dsp_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + RATE_CKCTL, + .enable_reg = (void __iomem *)ARM_CKCTL, + .enable_bit = EN_DSPCK, + .rate_offset = CKCTL_DSPDIV_OFFSET, + .recalc = &omap1_ckctl_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk dspmmu_ck = { + .name = "dspmmu_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + RATE_CKCTL | ALWAYS_ENABLED, + .rate_offset = CKCTL_DSPMMUDIV_OFFSET, + .recalc = &omap1_ckctl_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk dspper_ck = { + .name = "dspper_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + RATE_CKCTL | VIRTUAL_IO_ADDRESS, + .enable_reg = (void __iomem *)DSP_IDLECT2, + .enable_bit = EN_PERCK, + .rate_offset = CKCTL_PERDIV_OFFSET, + .recalc = &omap1_ckctl_recalc_dsp_domain, + .set_rate = &omap1_clk_set_rate_dsp_domain, + .enable = &omap1_clk_enable_dsp_domain, + .disable = &omap1_clk_disable_dsp_domain, +}; + +static struct clk dspxor_ck = { + .name = "dspxor_ck", + .parent = &ck_ref, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + VIRTUAL_IO_ADDRESS, + .enable_reg = (void __iomem *)DSP_IDLECT2, + .enable_bit = EN_XORPCK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable_dsp_domain, + .disable = &omap1_clk_disable_dsp_domain, +}; + +static struct clk dsptim_ck = { + .name = "dsptim_ck", + .parent = &ck_ref, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + VIRTUAL_IO_ADDRESS, + .enable_reg = (void __iomem *)DSP_IDLECT2, + .enable_bit = EN_DSPTIMCK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable_dsp_domain, + .disable = &omap1_clk_disable_dsp_domain, +}; + +/* Tie ARM_IDLECT1:IDLIF_ARM to this logical clock structure */ +static struct arm_idlect1_clk tc_ck = { + .clk = { + .name = "tc_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + CLOCK_IN_OMAP730 | RATE_CKCTL | + RATE_PROPAGATES | ALWAYS_ENABLED | + CLOCK_IDLE_CONTROL, + .rate_offset = CKCTL_TCDIV_OFFSET, + .recalc = &omap1_ckctl_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 6, +}; + +static struct clk arminth_ck1510 = { + .name = "arminth_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP1510 | ALWAYS_ENABLED, + .recalc = &followparent_recalc, + /* Note: On 1510 the frequency follows TC_CK + * + * 16xx version is in MPU clocks. + */ + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk tipb_ck = { + /* No-idle controlled by "tc_ck" */ + .name = "tibp_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP1510 | ALWAYS_ENABLED, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk l3_ocpi_ck = { + /* No-idle controlled by "tc_ck" */ + .name = "l3_ocpi_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP16XX, + .enable_reg = (void __iomem *)ARM_IDLECT3, + .enable_bit = EN_OCPI_CK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk tc1_ck = { + .name = "tc1_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP16XX, + .enable_reg = (void __iomem *)ARM_IDLECT3, + .enable_bit = EN_TC1_CK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk tc2_ck = { + .name = "tc2_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP16XX, + .enable_reg = (void __iomem *)ARM_IDLECT3, + .enable_bit = EN_TC2_CK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk dma_ck = { + /* No-idle controlled by "tc_ck" */ + .name = "dma_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + ALWAYS_ENABLED, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk dma_lcdfree_ck = { + .name = "dma_lcdfree_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP16XX | ALWAYS_ENABLED, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct arm_idlect1_clk api_ck = { + .clk = { + .name = "api_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + CLOCK_IDLE_CONTROL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_APICK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 8, +}; + +static struct arm_idlect1_clk lb_ck = { + .clk = { + .name = "lb_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IDLE_CONTROL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_LBCK, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 4, +}; + +static struct clk rhea1_ck = { + .name = "rhea1_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP16XX | ALWAYS_ENABLED, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk rhea2_ck = { + .name = "rhea2_ck", + .parent = &tc_ck.clk, + .flags = CLOCK_IN_OMAP16XX | ALWAYS_ENABLED, + .recalc = &followparent_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk lcd_ck_16xx = { + .name = "lcd_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP16XX | CLOCK_IN_OMAP730 | RATE_CKCTL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_LCDCK, + .rate_offset = CKCTL_LCDDIV_OFFSET, + .recalc = &omap1_ckctl_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct arm_idlect1_clk lcd_ck_1510 = { + .clk = { + .name = "lcd_ck", + .parent = &ck_dpll1, + .flags = CLOCK_IN_OMAP1510 | RATE_CKCTL | + CLOCK_IDLE_CONTROL, + .enable_reg = (void __iomem *)ARM_IDLECT2, + .enable_bit = EN_LCDCK, + .rate_offset = CKCTL_LCDDIV_OFFSET, + .recalc = &omap1_ckctl_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, + }, + .idlect_shift = 3, +}; + +static struct clk uart1_1510 = { + .name = "uart1_ck", + /* Direct from ULPD, no real parent */ + .parent = &armper_ck.clk, + .rate = 12000000, + .flags = CLOCK_IN_OMAP1510 | ENABLE_REG_32BIT | + ALWAYS_ENABLED | CLOCK_NO_IDLE_PARENT, + .enable_reg = (void __iomem *)MOD_CONF_CTRL_0, + .enable_bit = 29, /* Chooses between 12MHz and 48MHz */ + .set_rate = &omap1_set_uart_rate, + .recalc = &omap1_uart_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct uart_clk uart1_16xx = { + .clk = { + .name = "uart1_ck", + /* Direct from ULPD, no real parent */ + .parent = &armper_ck.clk, + .rate = 48000000, + .flags = CLOCK_IN_OMAP16XX | RATE_FIXED | + ENABLE_REG_32BIT | CLOCK_NO_IDLE_PARENT, + .enable_reg = (void __iomem *)MOD_CONF_CTRL_0, + .enable_bit = 29, + .enable = &omap1_clk_enable_uart_functional, + .disable = &omap1_clk_disable_uart_functional, + }, + .sysc_addr = 0xfffb0054, +}; + +static struct clk uart2_ck = { + .name = "uart2_ck", + /* Direct from ULPD, no real parent */ + .parent = &armper_ck.clk, + .rate = 12000000, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + ENABLE_REG_32BIT | ALWAYS_ENABLED | + CLOCK_NO_IDLE_PARENT, + .enable_reg = (void __iomem *)MOD_CONF_CTRL_0, + .enable_bit = 30, /* Chooses between 12MHz and 48MHz */ + .set_rate = &omap1_set_uart_rate, + .recalc = &omap1_uart_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk uart3_1510 = { + .name = "uart3_ck", + /* Direct from ULPD, no real parent */ + .parent = &armper_ck.clk, + .rate = 12000000, + .flags = CLOCK_IN_OMAP1510 | ENABLE_REG_32BIT | + ALWAYS_ENABLED | CLOCK_NO_IDLE_PARENT, + .enable_reg = (void __iomem *)MOD_CONF_CTRL_0, + .enable_bit = 31, /* Chooses between 12MHz and 48MHz */ + .set_rate = &omap1_set_uart_rate, + .recalc = &omap1_uart_recalc, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct uart_clk uart3_16xx = { + .clk = { + .name = "uart3_ck", + /* Direct from ULPD, no real parent */ + .parent = &armper_ck.clk, + .rate = 48000000, + .flags = CLOCK_IN_OMAP16XX | RATE_FIXED | + ENABLE_REG_32BIT | CLOCK_NO_IDLE_PARENT, + .enable_reg = (void __iomem *)MOD_CONF_CTRL_0, + .enable_bit = 31, + .enable = &omap1_clk_enable_uart_functional, + .disable = &omap1_clk_disable_uart_functional, + }, + .sysc_addr = 0xfffb9854, +}; + +static struct clk usb_clko = { /* 6 MHz output on W4_USB_CLKO */ + .name = "usb_clko", + /* Direct from ULPD, no parent */ + .rate = 6000000, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + RATE_FIXED | ENABLE_REG_32BIT, + .enable_reg = (void __iomem *)ULPD_CLOCK_CTRL, + .enable_bit = USB_MCLK_EN_BIT, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk usb_hhc_ck1510 = { + .name = "usb_hhc_ck", + /* Direct from ULPD, no parent */ + .rate = 48000000, /* Actually 2 clocks, 12MHz and 48MHz */ + .flags = CLOCK_IN_OMAP1510 | + RATE_FIXED | ENABLE_REG_32BIT, + .enable_reg = (void __iomem *)MOD_CONF_CTRL_0, + .enable_bit = USB_HOST_HHC_UHOST_EN, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk usb_hhc_ck16xx = { + .name = "usb_hhc_ck", + /* Direct from ULPD, no parent */ + .rate = 48000000, + /* OTG_SYSCON_2.OTG_PADEN == 0 (not 1510-compatible) */ + .flags = CLOCK_IN_OMAP16XX | + RATE_FIXED | ENABLE_REG_32BIT, + .enable_reg = (void __iomem *)OTG_BASE + 0x08 /* OTG_SYSCON_2 */, + .enable_bit = 8 /* UHOST_EN */, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk usb_dc_ck = { + .name = "usb_dc_ck", + /* Direct from ULPD, no parent */ + .rate = 48000000, + .flags = CLOCK_IN_OMAP16XX | RATE_FIXED, + .enable_reg = (void __iomem *)SOFT_REQ_REG, + .enable_bit = 4, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk mclk_1510 = { + .name = "mclk", + /* Direct from ULPD, no parent. May be enabled by ext hardware. */ + .rate = 12000000, + .flags = CLOCK_IN_OMAP1510 | RATE_FIXED, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk mclk_16xx = { + .name = "mclk", + /* Direct from ULPD, no parent. May be enabled by ext hardware. */ + .flags = CLOCK_IN_OMAP16XX, + .enable_reg = (void __iomem *)COM_CLK_DIV_CTRL_SEL, + .enable_bit = COM_ULPD_PLL_CLK_REQ, + .set_rate = &omap1_set_ext_clk_rate, + .round_rate = &omap1_round_ext_clk_rate, + .init = &omap1_init_ext_clk, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk bclk_1510 = { + .name = "bclk", + /* Direct from ULPD, no parent. May be enabled by ext hardware. */ + .rate = 12000000, + .flags = CLOCK_IN_OMAP1510 | RATE_FIXED, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk bclk_16xx = { + .name = "bclk", + /* Direct from ULPD, no parent. May be enabled by ext hardware. */ + .flags = CLOCK_IN_OMAP16XX, + .enable_reg = (void __iomem *)SWD_CLK_DIV_CTRL_SEL, + .enable_bit = SWD_ULPD_PLL_CLK_REQ, + .set_rate = &omap1_set_ext_clk_rate, + .round_rate = &omap1_round_ext_clk_rate, + .init = &omap1_init_ext_clk, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk mmc1_ck = { + .name = "mmc1_ck", + /* Functional clock is direct from ULPD, interface clock is ARMPER */ + .parent = &armper_ck.clk, + .rate = 48000000, + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + RATE_FIXED | ENABLE_REG_32BIT | CLOCK_NO_IDLE_PARENT, + .enable_reg = (void __iomem *)MOD_CONF_CTRL_0, + .enable_bit = 23, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk mmc2_ck = { + .name = "mmc2_ck", + /* Functional clock is direct from ULPD, interface clock is ARMPER */ + .parent = &armper_ck.clk, + .rate = 48000000, + .flags = CLOCK_IN_OMAP16XX | + RATE_FIXED | ENABLE_REG_32BIT | CLOCK_NO_IDLE_PARENT, + .enable_reg = (void __iomem *)MOD_CONF_CTRL_0, + .enable_bit = 20, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk virtual_ck_mpu = { + .name = "mpu", + .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | + VIRTUAL_CLOCK | ALWAYS_ENABLED, + .parent = &arm_ck, /* Is smarter alias for */ + .recalc = &followparent_recalc, + .set_rate = &omap1_select_table_rate, + .round_rate = &omap1_round_to_table_rate, + .enable = &omap1_clk_enable, + .disable = &omap1_clk_disable, +}; + +static struct clk * onchip_clks[] = { + /* non-ULPD clocks */ + &ck_ref, + &ck_dpll1, + /* CK_GEN1 clocks */ + &ck_dpll1out.clk, + &arm_ck, + &armper_ck.clk, + &arm_gpio_ck, + &armxor_ck.clk, + &armtim_ck.clk, + &armwdt_ck.clk, + &arminth_ck1510, &arminth_ck16xx, + /* CK_GEN2 clocks */ + &dsp_ck, + &dspmmu_ck, + &dspper_ck, + &dspxor_ck, + &dsptim_ck, + /* CK_GEN3 clocks */ + &tc_ck.clk, + &tipb_ck, + &l3_ocpi_ck, + &tc1_ck, + &tc2_ck, + &dma_ck, + &dma_lcdfree_ck, + &api_ck.clk, + &lb_ck.clk, + &rhea1_ck, + &rhea2_ck, + &lcd_ck_16xx, + &lcd_ck_1510.clk, + /* ULPD clocks */ + &uart1_1510, + &uart1_16xx.clk, + &uart2_ck, + &uart3_1510, + &uart3_16xx.clk, + &usb_clko, + &usb_hhc_ck1510, &usb_hhc_ck16xx, + &usb_dc_ck, + &mclk_1510, &mclk_16xx, + &bclk_1510, &bclk_16xx, + &mmc1_ck, + &mmc2_ck, + /* Virtual clocks */ + &virtual_ck_mpu, +}; + +#endif diff --git a/arch/arm/mach-omap1/devices.c b/arch/arm/mach-omap1/devices.c index 3c5d901efeaa..ecbc47514adc 100644 --- a/arch/arm/mach-omap1/devices.c +++ b/arch/arm/mach-omap1/devices.c @@ -25,56 +25,7 @@ #include #include - -static void omap_nop_release(struct device *dev) -{ - /* Nothing */ -} - -/*-------------------------------------------------------------------------*/ - -#if defined(CONFIG_I2C_OMAP) || defined(CONFIG_I2C_OMAP_MODULE) - -#define OMAP_I2C_BASE 0xfffb3800 - -static struct resource i2c_resources[] = { - { - .start = OMAP_I2C_BASE, - .end = OMAP_I2C_BASE + 0x3f, - .flags = IORESOURCE_MEM, - }, - { - .start = INT_I2C, - .flags = IORESOURCE_IRQ, - }, -}; - -/* DMA not used; works around erratum writing to non-empty i2c fifo */ - -static struct platform_device omap_i2c_device = { - .name = "i2c_omap", - .id = -1, - .dev = { - .release = omap_nop_release, - }, - .num_resources = ARRAY_SIZE(i2c_resources), - .resource = i2c_resources, -}; - -static void omap_init_i2c(void) -{ - /* FIXME define and use a boot tag, in case of boards that - * either don't wire up I2C, or chips that mux it differently... - * it can include clocking and address info, maybe more. - */ - omap_cfg_reg(I2C_SCL); - omap_cfg_reg(I2C_SDA); - - (void) platform_device_register(&omap_i2c_device); -} -#else -static inline void omap_init_i2c(void) {} -#endif +extern void omap_nop_release(struct device *dev); /*-------------------------------------------------------------------------*/ @@ -110,137 +61,6 @@ static inline void omap_init_irda(void) {} /*-------------------------------------------------------------------------*/ -#if defined(CONFIG_MMC_OMAP) || defined(CONFIG_MMC_OMAP_MODULE) - -#define OMAP_MMC1_BASE 0xfffb7800 -#define OMAP_MMC2_BASE 0xfffb7c00 /* omap16xx only */ - -static struct omap_mmc_conf mmc1_conf; - -static u64 mmc1_dmamask = 0xffffffff; - -static struct resource mmc1_resources[] = { - { - .start = IO_ADDRESS(OMAP_MMC1_BASE), - .end = IO_ADDRESS(OMAP_MMC1_BASE) + 0x7f, - .flags = IORESOURCE_MEM, - }, - { - .start = INT_MMC, - .flags = IORESOURCE_IRQ, - }, -}; - -static struct platform_device mmc_omap_device1 = { - .name = "mmci-omap", - .id = 1, - .dev = { - .release = omap_nop_release, - .dma_mask = &mmc1_dmamask, - .platform_data = &mmc1_conf, - }, - .num_resources = ARRAY_SIZE(mmc1_resources), - .resource = mmc1_resources, -}; - -#ifdef CONFIG_ARCH_OMAP16XX - -static struct omap_mmc_conf mmc2_conf; - -static u64 mmc2_dmamask = 0xffffffff; - -static struct resource mmc2_resources[] = { - { - .start = IO_ADDRESS(OMAP_MMC2_BASE), - .end = IO_ADDRESS(OMAP_MMC2_BASE) + 0x7f, - .flags = IORESOURCE_MEM, - }, - { - .start = INT_1610_MMC2, - .flags = IORESOURCE_IRQ, - }, -}; - -static struct platform_device mmc_omap_device2 = { - .name = "mmci-omap", - .id = 2, - .dev = { - .release = omap_nop_release, - .dma_mask = &mmc2_dmamask, - .platform_data = &mmc2_conf, - }, - .num_resources = ARRAY_SIZE(mmc2_resources), - .resource = mmc2_resources, -}; -#endif - -static void __init omap_init_mmc(void) -{ - const struct omap_mmc_config *mmc_conf; - const struct omap_mmc_conf *mmc; - - /* NOTE: assumes MMC was never (wrongly) enabled */ - mmc_conf = omap_get_config(OMAP_TAG_MMC, struct omap_mmc_config); - if (!mmc_conf) - return; - - /* block 1 is always available and has just one pinout option */ - mmc = &mmc_conf->mmc[0]; - if (mmc->enabled) { - omap_cfg_reg(MMC_CMD); - omap_cfg_reg(MMC_CLK); - omap_cfg_reg(MMC_DAT0); - if (cpu_is_omap1710()) { - omap_cfg_reg(M15_1710_MMC_CLKI); - omap_cfg_reg(P19_1710_MMC_CMDDIR); - omap_cfg_reg(P20_1710_MMC_DATDIR0); - } - if (mmc->wire4) { - omap_cfg_reg(MMC_DAT1); - /* NOTE: DAT2 can be on W10 (here) or M15 */ - if (!mmc->nomux) - omap_cfg_reg(MMC_DAT2); - omap_cfg_reg(MMC_DAT3); - } - mmc1_conf = *mmc; - (void) platform_device_register(&mmc_omap_device1); - } - -#ifdef CONFIG_ARCH_OMAP16XX - /* block 2 is on newer chips, and has many pinout options */ - mmc = &mmc_conf->mmc[1]; - if (mmc->enabled) { - if (!mmc->nomux) { - omap_cfg_reg(Y8_1610_MMC2_CMD); - omap_cfg_reg(Y10_1610_MMC2_CLK); - omap_cfg_reg(R18_1610_MMC2_CLKIN); - omap_cfg_reg(W8_1610_MMC2_DAT0); - if (mmc->wire4) { - omap_cfg_reg(V8_1610_MMC2_DAT1); - omap_cfg_reg(W15_1610_MMC2_DAT2); - omap_cfg_reg(R10_1610_MMC2_DAT3); - } - - /* These are needed for the level shifter */ - omap_cfg_reg(V9_1610_MMC2_CMDDIR); - omap_cfg_reg(V5_1610_MMC2_DATDIR0); - omap_cfg_reg(W19_1610_MMC2_DATDIR1); - } - - /* Feedback clock must be set on OMAP-1710 MMC2 */ - if (cpu_is_omap1710()) - omap_writel(omap_readl(MOD_CONF_CTRL_1) | (1 << 24), - MOD_CONF_CTRL_1); - mmc2_conf = *mmc; - (void) platform_device_register(&mmc_omap_device2); - } -#endif - return; -} -#else -static inline void omap_init_mmc(void) {} -#endif - #if defined(CONFIG_OMAP_RTC) || defined(CONFIG_OMAP_RTC) #define OMAP_RTC_BASE 0xfffb4800 @@ -279,38 +99,6 @@ static void omap_init_rtc(void) static inline void omap_init_rtc(void) {} #endif -/*-------------------------------------------------------------------------*/ - -#if defined(CONFIG_OMAP16XX_WATCHDOG) || defined(CONFIG_OMAP16XX_WATCHDOG_MODULE) - -#define OMAP_WDT_BASE 0xfffeb000 - -static struct resource wdt_resources[] = { - { - .start = OMAP_WDT_BASE, - .end = OMAP_WDT_BASE + 0x4f, - .flags = IORESOURCE_MEM, - }, -}; - -static struct platform_device omap_wdt_device = { - .name = "omap1610_wdt", - .id = -1, - .dev = { - .release = omap_nop_release, - }, - .num_resources = ARRAY_SIZE(wdt_resources), - .resource = wdt_resources, -}; - -static void omap_init_wdt(void) -{ - (void) platform_device_register(&omap_wdt_device); -} -#else -static inline void omap_init_wdt(void) {} -#endif - /*-------------------------------------------------------------------------*/ @@ -334,18 +122,15 @@ static inline void omap_init_wdt(void) {} * may be handled by the boot loader, and drivers should expect it will * normally have been done by the time they're probed. */ -static int __init omap_init_devices(void) +static int __init omap1_init_devices(void) { /* please keep these calls, and their implementations above, * in alphabetical order so they're easier to sort through. */ - omap_init_i2c(); omap_init_irda(); - omap_init_mmc(); omap_init_rtc(); - omap_init_wdt(); return 0; } -arch_initcall(omap_init_devices); +arch_initcall(omap1_init_devices); diff --git a/arch/arm/mach-omap1/id.c b/arch/arm/mach-omap1/id.c index 986c3b7e09bb..5c637c048368 100644 --- a/arch/arm/mach-omap1/id.c +++ b/arch/arm/mach-omap1/id.c @@ -18,6 +18,13 @@ #include +#define OMAP_DIE_ID_0 0xfffe1800 +#define OMAP_DIE_ID_1 0xfffe1804 +#define OMAP_PRODUCTION_ID_0 0xfffe2000 +#define OMAP_PRODUCTION_ID_1 0xfffe2004 +#define OMAP32_ID_0 0xfffed400 +#define OMAP32_ID_1 0xfffed404 + struct omap_id { u16 jtag_id; /* Used to determine OMAP type */ u8 die_rev; /* Processor revision */ @@ -27,6 +34,7 @@ struct omap_id { /* Register values to detect the OMAP version */ static struct omap_id omap_ids[] __initdata = { + { .jtag_id = 0xb574, .die_rev = 0x2, .omap_id = 0x03310315, .type = 0x03100000}, { .jtag_id = 0x355f, .die_rev = 0x0, .omap_id = 0x03320000, .type = 0x07300100}, { .jtag_id = 0xb55f, .die_rev = 0x0, .omap_id = 0x03320000, .type = 0x07300300}, { .jtag_id = 0xb470, .die_rev = 0x0, .omap_id = 0x03310100, .type = 0x15100000}, @@ -164,6 +172,7 @@ void __init omap_check_revision(void) case 0x07: system_rev |= 0x07; break; + case 0x03: case 0x15: system_rev |= 0x15; break; diff --git a/arch/arm/mach-omap1/io.c b/arch/arm/mach-omap1/io.c index 79fb86535ebc..a7a19f75b9e1 100644 --- a/arch/arm/mach-omap1/io.c +++ b/arch/arm/mach-omap1/io.c @@ -15,9 +15,10 @@ #include #include +#include #include -extern int clk_init(void); +extern int omap1_clk_init(void); extern void omap_check_revision(void); extern void omap_sram_init(void); @@ -50,7 +51,7 @@ static struct map_desc omap730_io_desc[] __initdata = { }; #endif -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX static struct map_desc omap1510_io_desc[] __initdata = { { .virtual = OMAP1510_DSP_BASE, @@ -98,7 +99,7 @@ static void __init _omap_map_io(void) iotable_init(omap730_io_desc, ARRAY_SIZE(omap730_io_desc)); } #endif -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { iotable_init(omap1510_io_desc, ARRAY_SIZE(omap1510_io_desc)); } @@ -119,7 +120,7 @@ static void __init _omap_map_io(void) /* Must init clocks early to assure that timer interrupt works */ - clk_init(); + omap1_clk_init(); } /* @@ -127,7 +128,9 @@ static void __init _omap_map_io(void) */ void __init omap_map_common_io(void) { - if (!initialized) + if (!initialized) { _omap_map_io(); + omap1_mux_init(); + } } diff --git a/arch/arm/mach-omap1/irq.c b/arch/arm/mach-omap1/irq.c index 192ce6055faa..ed65a7d2e941 100644 --- a/arch/arm/mach-omap1/irq.c +++ b/arch/arm/mach-omap1/irq.c @@ -47,6 +47,7 @@ #include #include #include +#include #include @@ -147,11 +148,15 @@ static struct omap_irq_bank omap730_irq_banks[] = { }; #endif -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX static struct omap_irq_bank omap1510_irq_banks[] = { { .base_reg = OMAP_IH1_BASE, .trigger_map = 0xb3febfff }, { .base_reg = OMAP_IH2_BASE, .trigger_map = 0xffbfffed }, }; +static struct omap_irq_bank omap310_irq_banks[] = { + { .base_reg = OMAP_IH1_BASE, .trigger_map = 0xb3faefc3 }, + { .base_reg = OMAP_IH2_BASE, .trigger_map = 0x65b3c061 }, +}; #endif #if defined(CONFIG_ARCH_OMAP16XX) @@ -181,11 +186,15 @@ void __init omap_init_irq(void) irq_bank_count = ARRAY_SIZE(omap730_irq_banks); } #endif -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { irq_banks = omap1510_irq_banks; irq_bank_count = ARRAY_SIZE(omap1510_irq_banks); } + if (cpu_is_omap310()) { + irq_banks = omap310_irq_banks; + irq_bank_count = ARRAY_SIZE(omap310_irq_banks); + } #endif #if defined(CONFIG_ARCH_OMAP16XX) if (cpu_is_omap16xx()) { @@ -226,9 +235,11 @@ void __init omap_init_irq(void) } /* Unmask level 2 handler */ - if (cpu_is_omap730()) { + + if (cpu_is_omap730()) omap_unmask_irq(INT_730_IH2_IRQ); - } else { - omap_unmask_irq(INT_IH2_IRQ); - } + else if (cpu_is_omap1510()) + omap_unmask_irq(INT_1510_IH2_IRQ); + else if (cpu_is_omap16xx()) + omap_unmask_irq(INT_1610_IH2_IRQ); } diff --git a/arch/arm/mach-omap1/leds-h2p2-debug.c b/arch/arm/mach-omap1/leds-h2p2-debug.c index 399010c14036..650650815915 100644 --- a/arch/arm/mach-omap1/leds-h2p2-debug.c +++ b/arch/arm/mach-omap1/leds-h2p2-debug.c @@ -18,6 +18,7 @@ #include #include #include +#include #include #include @@ -63,14 +64,19 @@ void h2p2_dbg_leds_event(led_event_t evt) case led_stop: case led_halted: /* all leds off during suspend or shutdown */ - omap_set_gpio_dataout(GPIO_TIMER, 0); - omap_set_gpio_dataout(GPIO_IDLE, 0); + + if (! machine_is_omap_perseus2()) { + omap_set_gpio_dataout(GPIO_TIMER, 0); + omap_set_gpio_dataout(GPIO_IDLE, 0); + } + __raw_writew(~0, &fpga->leds); led_state &= ~LED_STATE_ENABLED; if (evt == led_halted) { iounmap(fpga); fpga = NULL; } + goto done; case led_claim: @@ -85,18 +91,37 @@ void h2p2_dbg_leds_event(led_event_t evt) #ifdef CONFIG_LEDS_TIMER case led_timer: led_state ^= LED_TIMER_ON; - omap_set_gpio_dataout(GPIO_TIMER, led_state & LED_TIMER_ON); - goto done; + + if (machine_is_omap_perseus2()) + hw_led_state ^= H2P2_DBG_FPGA_P2_LED_TIMER; + else { + omap_set_gpio_dataout(GPIO_TIMER, led_state & LED_TIMER_ON); + goto done; + } + + break; #endif #ifdef CONFIG_LEDS_CPU case led_idle_start: - omap_set_gpio_dataout(GPIO_IDLE, 1); - goto done; + if (machine_is_omap_perseus2()) + hw_led_state |= H2P2_DBG_FPGA_P2_LED_IDLE; + else { + omap_set_gpio_dataout(GPIO_IDLE, 1); + goto done; + } + + break; case led_idle_end: - omap_set_gpio_dataout(GPIO_IDLE, 0); - goto done; + if (machine_is_omap_perseus2()) + hw_led_state &= ~H2P2_DBG_FPGA_P2_LED_IDLE; + else { + omap_set_gpio_dataout(GPIO_IDLE, 0); + goto done; + } + + break; #endif case led_green_on: @@ -135,7 +160,7 @@ void h2p2_dbg_leds_event(led_event_t evt) /* * Actually burn the LEDs */ - if (led_state & LED_STATE_CLAIMED) + if (led_state & LED_STATE_ENABLED) __raw_writew(~hw_led_state, &fpga->leds); done: diff --git a/arch/arm/mach-omap1/leds.c b/arch/arm/mach-omap1/leds.c index 5c6b1bb6e722..3f9dcac4fd41 100644 --- a/arch/arm/mach-omap1/leds.c +++ b/arch/arm/mach-omap1/leds.c @@ -33,7 +33,6 @@ omap_leds_init(void) if (machine_is_omap_h2() || machine_is_omap_h3() - || machine_is_omap_perseus2() #ifdef CONFIG_OMAP_OSK_MISTRAL || machine_is_omap_osk() #endif diff --git a/arch/arm/mach-omap1/mux.c b/arch/arm/mach-omap1/mux.c new file mode 100644 index 000000000000..d4b8d624e742 --- /dev/null +++ b/arch/arm/mach-omap1/mux.c @@ -0,0 +1,289 @@ +/* + * linux/arch/arm/mach-omap1/mux.c + * + * OMAP1 pin multiplexing configurations + * + * Copyright (C) 2003 - 2005 Nokia Corporation + * + * Written by Tony Lindgren + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + */ +#include +#include +#include +#include +#include +#include + +#include + +#ifdef CONFIG_OMAP_MUX + +#ifdef CONFIG_ARCH_OMAP730 +struct pin_config __initdata_or_module omap730_pins[] = { +MUX_CFG_730("E2_730_KBR0", 12, 21, 0, 0, 20, 1, NA, 0, 0) +MUX_CFG_730("J7_730_KBR1", 12, 25, 0, 0, 24, 1, NA, 0, 0) +MUX_CFG_730("E1_730_KBR2", 12, 29, 0, 0, 28, 1, NA, 0, 0) +MUX_CFG_730("F3_730_KBR3", 13, 1, 0, 0, 0, 1, NA, 0, 0) +MUX_CFG_730("D2_730_KBR4", 13, 5, 0, 0, 4, 1, NA, 0, 0) +MUX_CFG_730("C2_730_KBC0", 13, 9, 0, 0, 8, 1, NA, 0, 0) +MUX_CFG_730("D3_730_KBC1", 13, 13, 0, 0, 12, 1, NA, 0, 0) +MUX_CFG_730("E4_730_KBC2", 13, 17, 0, 0, 16, 1, NA, 0, 0) +MUX_CFG_730("F4_730_KBC3", 13, 21, 0, 0, 20, 1, NA, 0, 0) +MUX_CFG_730("E3_730_KBC4", 13, 25, 0, 0, 24, 1, NA, 0, 0) +}; +#endif + +#if defined(CONFIG_ARCH_OMAP15XX) || defined(CONFIG_ARCH_OMAP16XX) +struct pin_config __initdata_or_module omap1xxx_pins[] = { +/* + * description mux mode mux pull pull pull pu_pd pu dbg + * reg offset mode reg bit ena reg + */ +MUX_CFG("UART1_TX", 9, 21, 1, 2, 3, 0, NA, 0, 0) +MUX_CFG("UART1_RTS", 9, 12, 1, 2, 0, 0, NA, 0, 0) + +/* UART2 (COM_UART_GATING), conflicts with USB2 */ +MUX_CFG("UART2_TX", C, 27, 1, 3, 3, 0, NA, 0, 0) +MUX_CFG("UART2_RX", C, 18, 0, 3, 1, 1, NA, 0, 0) +MUX_CFG("UART2_CTS", C, 21, 0, 3, 1, 1, NA, 0, 0) +MUX_CFG("UART2_RTS", C, 24, 1, 3, 2, 0, NA, 0, 0) + +/* UART3 (GIGA_UART_GATING) */ +MUX_CFG("UART3_TX", 6, 0, 1, 0, 30, 0, NA, 0, 0) +MUX_CFG("UART3_RX", 6, 3, 0, 0, 31, 1, NA, 0, 0) +MUX_CFG("UART3_CTS", 5, 12, 2, 0, 24, 0, NA, 0, 0) +MUX_CFG("UART3_RTS", 5, 15, 2, 0, 25, 0, NA, 0, 0) +MUX_CFG("UART3_CLKREQ", 9, 27, 0, 2, 5, 0, NA, 0, 0) +MUX_CFG("UART3_BCLK", A, 0, 0, 2, 6, 0, NA, 0, 0) +MUX_CFG("Y15_1610_UART3_RTS", A, 0, 1, 2, 6, 0, NA, 0, 0) + +/* PWT & PWL, conflicts with UART3 */ +MUX_CFG("PWT", 6, 0, 2, 0, 30, 0, NA, 0, 0) +MUX_CFG("PWL", 6, 3, 1, 0, 31, 1, NA, 0, 0) + +/* USB internal master generic */ +MUX_CFG("R18_USB_VBUS", 7, 9, 2, 1, 11, 0, NA, 0, 1) +MUX_CFG("R18_1510_USB_GPIO0", 7, 9, 0, 1, 11, 1, NA, 0, 1) +/* works around erratum: W4_USB_PUEN and W4_USB_PUDIS are switched! */ +MUX_CFG("W4_USB_PUEN", D, 3, 3, 3, 5, 1, NA, 0, 1) +MUX_CFG("W4_USB_CLKO", D, 3, 1, 3, 5, 0, NA, 0, 1) +MUX_CFG("W4_USB_HIGHZ", D, 3, 4, 3, 5, 0, 3, 0, 1) +MUX_CFG("W4_GPIO58", D, 3, 7, 3, 5, 0, 3, 0, 1) + +/* USB1 master */ +MUX_CFG("USB1_SUSP", 8, 27, 2, 1, 27, 0, NA, 0, 1) +MUX_CFG("USB1_SE0", 9, 0, 2, 1, 28, 0, NA, 0, 1) +MUX_CFG("W13_1610_USB1_SE0", 9, 0, 4, 1, 28, 0, NA, 0, 1) +MUX_CFG("USB1_TXEN", 9, 3, 2, 1, 29, 0, NA, 0, 1) +MUX_CFG("USB1_TXD", 9, 24, 1, 2, 4, 0, NA, 0, 1) +MUX_CFG("USB1_VP", A, 3, 1, 2, 7, 0, NA, 0, 1) +MUX_CFG("USB1_VM", A, 6, 1, 2, 8, 0, NA, 0, 1) +MUX_CFG("USB1_RCV", A, 9, 1, 2, 9, 0, NA, 0, 1) +MUX_CFG("USB1_SPEED", A, 12, 2, 2, 10, 0, NA, 0, 1) +MUX_CFG("R13_1610_USB1_SPEED", A, 12, 5, 2, 10, 0, NA, 0, 1) +MUX_CFG("R13_1710_USB1_SEO", A, 12, 5, 2, 10, 0, NA, 0, 1) + +/* USB2 master */ +MUX_CFG("USB2_SUSP", B, 3, 1, 2, 17, 0, NA, 0, 1) +MUX_CFG("USB2_VP", B, 6, 1, 2, 18, 0, NA, 0, 1) +MUX_CFG("USB2_TXEN", B, 9, 1, 2, 19, 0, NA, 0, 1) +MUX_CFG("USB2_VM", C, 18, 1, 3, 0, 0, NA, 0, 1) +MUX_CFG("USB2_RCV", C, 21, 1, 3, 1, 0, NA, 0, 1) +MUX_CFG("USB2_SE0", C, 24, 2, 3, 2, 0, NA, 0, 1) +MUX_CFG("USB2_TXD", C, 27, 2, 3, 3, 0, NA, 0, 1) + +/* OMAP-1510 GPIO */ +MUX_CFG("R18_1510_GPIO0", 7, 9, 0, 1, 11, 1, 0, 0, 1) +MUX_CFG("R19_1510_GPIO1", 7, 6, 0, 1, 10, 1, 0, 0, 1) +MUX_CFG("M14_1510_GPIO2", 7, 3, 0, 1, 9, 1, 0, 0, 1) + +/* OMAP1610 GPIO */ +MUX_CFG("P18_1610_GPIO3", 7, 0, 0, 1, 8, 0, NA, 0, 1) +MUX_CFG("Y15_1610_GPIO17", A, 0, 7, 2, 6, 0, NA, 0, 1) + +/* OMAP-1710 GPIO */ +MUX_CFG("R18_1710_GPIO0", 7, 9, 0, 1, 11, 1, 1, 1, 1) +MUX_CFG("V2_1710_GPIO10", F, 27, 1, 4, 3, 1, 4, 1, 1) +MUX_CFG("N21_1710_GPIO14", 6, 9, 0, 1, 1, 1, 1, 1, 1) +MUX_CFG("W15_1710_GPIO40", 9, 27, 7, 2, 5, 1, 2, 1, 1) + +/* MPUIO */ +MUX_CFG("MPUIO2", 7, 18, 0, 1, 14, 1, NA, 0, 1) +MUX_CFG("N15_1610_MPUIO2", 7, 18, 0, 1, 14, 1, 1, 0, 1) +MUX_CFG("MPUIO4", 7, 15, 0, 1, 13, 1, NA, 0, 1) +MUX_CFG("MPUIO5", 7, 12, 0, 1, 12, 1, NA, 0, 1) + +MUX_CFG("T20_1610_MPUIO5", 7, 12, 0, 1, 12, 0, 3, 0, 1) +MUX_CFG("W11_1610_MPUIO6", 10, 15, 2, 3, 8, 0, 3, 0, 1) +MUX_CFG("V10_1610_MPUIO7", A, 24, 2, 2, 14, 0, 2, 0, 1) +MUX_CFG("W11_1610_MPUIO9", 10, 15, 1, 3, 8, 0, 3, 0, 1) +MUX_CFG("V10_1610_MPUIO10", A, 24, 1, 2, 14, 0, 2, 0, 1) +MUX_CFG("W10_1610_MPUIO11", A, 18, 2, 2, 11, 0, 2, 0, 1) +MUX_CFG("E20_1610_MPUIO13", 3, 21, 1, 0, 7, 0, 0, 0, 1) +MUX_CFG("U20_1610_MPUIO14", 9, 6, 6, 0, 30, 0, 0, 0, 1) +MUX_CFG("E19_1610_MPUIO15", 3, 18, 1, 0, 6, 0, 0, 0, 1) + +/* MCBSP2 */ +MUX_CFG("MCBSP2_CLKR", C, 6, 0, 2, 27, 1, NA, 0, 1) +MUX_CFG("MCBSP2_CLKX", C, 9, 0, 2, 29, 1, NA, 0, 1) +MUX_CFG("MCBSP2_DR", C, 0, 0, 2, 26, 1, NA, 0, 1) +MUX_CFG("MCBSP2_DX", C, 15, 0, 2, 31, 1, NA, 0, 1) +MUX_CFG("MCBSP2_FSR", C, 12, 0, 2, 30, 1, NA, 0, 1) +MUX_CFG("MCBSP2_FSX", C, 3, 0, 2, 27, 1, NA, 0, 1) + +/* MCBSP3 NOTE: Mode must 1 for clock */ +MUX_CFG("MCBSP3_CLKX", 9, 3, 1, 1, 29, 0, NA, 0, 1) + +/* Misc ballouts */ +MUX_CFG("BALLOUT_V8_ARMIO3", B, 18, 0, 2, 25, 1, NA, 0, 1) +MUX_CFG("N20_HDQ", 6, 18, 1, 1, 4, 0, 1, 4, 0) + +/* OMAP-1610 MMC2 */ +MUX_CFG("W8_1610_MMC2_DAT0", B, 21, 6, 2, 23, 1, 2, 1, 1) +MUX_CFG("V8_1610_MMC2_DAT1", B, 27, 6, 2, 25, 1, 2, 1, 1) +MUX_CFG("W15_1610_MMC2_DAT2", 9, 12, 6, 2, 5, 1, 2, 1, 1) +MUX_CFG("R10_1610_MMC2_DAT3", B, 18, 6, 2, 22, 1, 2, 1, 1) +MUX_CFG("Y10_1610_MMC2_CLK", B, 3, 6, 2, 17, 0, 2, 0, 1) +MUX_CFG("Y8_1610_MMC2_CMD", B, 24, 6, 2, 24, 1, 2, 1, 1) +MUX_CFG("V9_1610_MMC2_CMDDIR", B, 12, 6, 2, 20, 0, 2, 1, 1) +MUX_CFG("V5_1610_MMC2_DATDIR0", B, 15, 6, 2, 21, 0, 2, 1, 1) +MUX_CFG("W19_1610_MMC2_DATDIR1", 8, 15, 6, 1, 23, 0, 1, 1, 1) +MUX_CFG("R18_1610_MMC2_CLKIN", 7, 9, 6, 1, 11, 0, 1, 11, 1) + +/* OMAP-1610 External Trace Interface */ +MUX_CFG("M19_1610_ETM_PSTAT0", 5, 27, 1, 0, 29, 0, 0, 0, 1) +MUX_CFG("L15_1610_ETM_PSTAT1", 5, 24, 1, 0, 28, 0, 0, 0, 1) +MUX_CFG("L18_1610_ETM_PSTAT2", 5, 21, 1, 0, 27, 0, 0, 0, 1) +MUX_CFG("L19_1610_ETM_D0", 5, 18, 1, 0, 26, 0, 0, 0, 1) +MUX_CFG("J19_1610_ETM_D6", 5, 0, 1, 0, 20, 0, 0, 0, 1) +MUX_CFG("J18_1610_ETM_D7", 5, 27, 1, 0, 19, 0, 0, 0, 1) + +/* OMAP16XX GPIO */ +MUX_CFG("P20_1610_GPIO4", 6, 27, 0, 1, 7, 0, 1, 1, 1) +MUX_CFG("V9_1610_GPIO7", B, 12, 1, 2, 20, 0, 2, 1, 1) +MUX_CFG("W8_1610_GPIO9", B, 21, 0, 2, 23, 0, 2, 1, 1) +MUX_CFG("N20_1610_GPIO11", 6, 18, 0, 1, 4, 0, 1, 1, 1) +MUX_CFG("N19_1610_GPIO13", 6, 12, 0, 1, 2, 0, 1, 1, 1) +MUX_CFG("P10_1610_GPIO22", C, 0, 7, 2, 26, 0, 2, 1, 1) +MUX_CFG("V5_1610_GPIO24", B, 15, 7, 2, 21, 0, 2, 1, 1) +MUX_CFG("AA20_1610_GPIO_41", 9, 9, 7, 1, 31, 0, 1, 1, 1) +MUX_CFG("W19_1610_GPIO48", 8, 15, 7, 1, 23, 1, 1, 0, 1) +MUX_CFG("M7_1610_GPIO62", 10, 0, 0, 4, 24, 0, 4, 0, 1) +MUX_CFG("V14_16XX_GPIO37", 9, 18, 7, 2, 2, 0, 2, 2, 0) +MUX_CFG("R9_16XX_GPIO18", C, 18, 7, 3, 0, 0, 3, 0, 0) +MUX_CFG("L14_16XX_GPIO49", 6, 3, 7, 0, 31, 0, 0, 31, 0) + +/* OMAP-1610 uWire */ +MUX_CFG("V19_1610_UWIRE_SCLK", 8, 6, 0, 1, 20, 0, 1, 1, 1) +MUX_CFG("U18_1610_UWIRE_SDI", 8, 0, 0, 1, 18, 0, 1, 1, 1) +MUX_CFG("W21_1610_UWIRE_SDO", 8, 3, 0, 1, 19, 0, 1, 1, 1) +MUX_CFG("N14_1610_UWIRE_CS0", 8, 9, 1, 1, 21, 0, 1, 1, 1) +MUX_CFG("P15_1610_UWIRE_CS3", 8, 12, 1, 1, 22, 0, 1, 1, 1) +MUX_CFG("N15_1610_UWIRE_CS1", 7, 18, 2, 1, 14, 0, NA, 0, 1) + +/* OMAP-1610 Flash */ +MUX_CFG("L3_1610_FLASH_CS2B_OE",10, 6, 1, NA, 0, 0, NA, 0, 1) +MUX_CFG("M8_1610_FLASH_CS2B_WE",10, 3, 1, NA, 0, 0, NA, 0, 1) + +/* First MMC interface, same on 1510, 1610 and 1710 */ +MUX_CFG("MMC_CMD", A, 27, 0, 2, 15, 1, 2, 1, 1) +MUX_CFG("MMC_DAT1", A, 24, 0, 2, 14, 1, 2, 1, 1) +MUX_CFG("MMC_DAT2", A, 18, 0, 2, 12, 1, 2, 1, 1) +MUX_CFG("MMC_DAT0", B, 0, 0, 2, 16, 1, 2, 1, 1) +MUX_CFG("MMC_CLK", A, 21, 0, NA, 0, 0, NA, 0, 1) +MUX_CFG("MMC_DAT3", 10, 15, 0, 3, 8, 1, 3, 1, 1) +MUX_CFG("M15_1710_MMC_CLKI", 6, 21, 2, 0, 0, 0, NA, 0, 1) +MUX_CFG("P19_1710_MMC_CMDDIR", 6, 24, 6, 0, 0, 0, NA, 0, 1) +MUX_CFG("P20_1710_MMC_DATDIR0", 6, 27, 5, 0, 0, 0, NA, 0, 1) + +/* OMAP-1610 USB0 alternate configuration */ +MUX_CFG("W9_USB0_TXEN", B, 9, 5, 2, 19, 0, 2, 0, 1) +MUX_CFG("AA9_USB0_VP", B, 6, 5, 2, 18, 0, 2, 0, 1) +MUX_CFG("Y5_USB0_RCV", C, 21, 5, 3, 1, 0, 1, 0, 1) +MUX_CFG("R9_USB0_VM", C, 18, 5, 3, 0, 0, 3, 0, 1) +MUX_CFG("V6_USB0_TXD", C, 27, 5, 3, 3, 0, 3, 0, 1) +MUX_CFG("W5_USB0_SE0", C, 24, 5, 3, 2, 0, 3, 0, 1) +MUX_CFG("V9_USB0_SPEED", B, 12, 5, 2, 20, 0, 2, 0, 1) +MUX_CFG("Y10_USB0_SUSP", B, 3, 5, 2, 17, 0, 2, 0, 1) + +/* USB2 interface */ +MUX_CFG("W9_USB2_TXEN", B, 9, 1, NA, 0, 0, NA, 0, 1) +MUX_CFG("AA9_USB2_VP", B, 6, 1, NA, 0, 0, NA, 0, 1) +MUX_CFG("Y5_USB2_RCV", C, 21, 1, NA, 0, 0, NA, 0, 1) +MUX_CFG("R9_USB2_VM", C, 18, 1, NA, 0, 0, NA, 0, 1) +MUX_CFG("V6_USB2_TXD", C, 27, 2, NA, 0, 0, NA, 0, 1) +MUX_CFG("W5_USB2_SE0", C, 24, 2, NA, 0, 0, NA, 0, 1) + +/* 16XX UART */ +MUX_CFG("R13_1610_UART1_TX", A, 12, 6, 2, 10, 0, 2, 10, 1) +MUX_CFG("V14_16XX_UART1_RX", 9, 18, 0, 2, 2, 0, 2, 2, 1) +MUX_CFG("R14_1610_UART1_CTS", 9, 15, 0, 2, 1, 0, 2, 1, 1) +MUX_CFG("AA15_1610_UART1_RTS", 9, 12, 1, 2, 0, 0, 2, 0, 1) +MUX_CFG("R9_16XX_UART2_RX", C, 18, 0, 3, 0, 0, 3, 0, 1) +MUX_CFG("L14_16XX_UART3_RX", 6, 3, 0, 0, 31, 0, 0, 31, 1) + +/* I2C interface */ +MUX_CFG("I2C_SCL", 7, 24, 0, NA, 0, 0, NA, 0, 0) +MUX_CFG("I2C_SDA", 7, 27, 0, NA, 0, 0, NA, 0, 0) + +/* Keypad */ +MUX_CFG("F18_1610_KBC0", 3, 15, 0, 0, 5, 1, 0, 0, 0) +MUX_CFG("D20_1610_KBC1", 3, 12, 0, 0, 4, 1, 0, 0, 0) +MUX_CFG("D19_1610_KBC2", 3, 9, 0, 0, 3, 1, 0, 0, 0) +MUX_CFG("E18_1610_KBC3", 3, 6, 0, 0, 2, 1, 0, 0, 0) +MUX_CFG("C21_1610_KBC4", 3, 3, 0, 0, 1, 1, 0, 0, 0) +MUX_CFG("G18_1610_KBR0", 4, 0, 0, 0, 10, 1, 0, 1, 0) +MUX_CFG("F19_1610_KBR1", 3, 27, 0, 0, 9, 1, 0, 1, 0) +MUX_CFG("H14_1610_KBR2", 3, 24, 0, 0, 8, 1, 0, 1, 0) +MUX_CFG("E20_1610_KBR3", 3, 21, 0, 0, 7, 1, 0, 1, 0) +MUX_CFG("E19_1610_KBR4", 3, 18, 0, 0, 6, 1, 0, 1, 0) +MUX_CFG("N19_1610_KBR5", 6, 12, 1, 1, 2, 1, 1, 1, 0) + +/* Power management */ +MUX_CFG("T20_1610_LOW_PWR", 7, 12, 1, NA, 0, 0, NA, 0, 0) + +/* MCLK Settings */ +MUX_CFG("V5_1710_MCLK_ON", B, 15, 0, NA, 0, 0, NA, 0, 0) +MUX_CFG("V5_1710_MCLK_OFF", B, 15, 6, NA, 0, 0, NA, 0, 0) +MUX_CFG("R10_1610_MCLK_ON", B, 18, 0, NA, 22, 0, NA, 1, 0) +MUX_CFG("R10_1610_MCLK_OFF", B, 18, 6, 2, 22, 1, 2, 1, 1) + +/* CompactFlash controller, conflicts with MMC1 */ +MUX_CFG("P11_1610_CF_CD2", A, 27, 3, 2, 15, 1, 2, 1, 1) +MUX_CFG("R11_1610_CF_IOIS16", B, 0, 3, 2, 16, 1, 2, 1, 1) +MUX_CFG("V10_1610_CF_IREQ", A, 24, 3, 2, 14, 0, 2, 0, 1) +MUX_CFG("W10_1610_CF_RESET", A, 18, 3, 2, 12, 1, 2, 1, 1) +MUX_CFG("W11_1610_CF_CD1", 10, 15, 3, 3, 8, 1, 3, 1, 1) +}; +#endif /* CONFIG_ARCH_OMAP15XX || CONFIG_ARCH_OMAP16XX */ + +int __init omap1_mux_init(void) +{ + +#ifdef CONFIG_ARCH_OMAP730 + omap_mux_register(omap730_pins, ARRAY_SIZE(omap730_pins)); +#endif + +#if defined(CONFIG_ARCH_OMAP15XX) || defined(CONFIG_ARCH_OMAP16XX) + omap_mux_register(omap1xxx_pins, ARRAY_SIZE(omap1xxx_pins)); +#endif + + return 0; +} + +#endif diff --git a/arch/arm/mach-omap1/serial.c b/arch/arm/mach-omap1/serial.c index 40c4f7c40e73..6810cfb84462 100644 --- a/arch/arm/mach-omap1/serial.c +++ b/arch/arm/mach-omap1/serial.c @@ -109,9 +109,10 @@ static struct platform_device serial_device = { * By default UART2 does not work on Innovator-1510 if you have * USB OHCI enabled. To use UART2, you must disable USB2 first. */ -void __init omap_serial_init(int ports[OMAP_MAX_NR_PORTS]) +void __init omap_serial_init(void) { int i; + const struct omap_uart_config *info; if (cpu_is_omap730()) { serial_platform_data[0].regshift = 0; @@ -126,10 +127,14 @@ void __init omap_serial_init(int ports[OMAP_MAX_NR_PORTS]) serial_platform_data[2].uartclk = OMAP1510_BASE_BAUD * 16; } + info = omap_get_config(OMAP_TAG_UART, struct omap_uart_config); + if (info == NULL) + return; + for (i = 0; i < OMAP_MAX_NR_PORTS; i++) { unsigned char reg; - if (ports[i] == 0) { + if (!((1 << i) & info->enabled_uarts)) { serial_platform_data[i].membase = NULL; serial_platform_data[i].mapbase = 0; continue; diff --git a/arch/arm/mach-omap1/time.c b/arch/arm/mach-omap1/time.c index 191a9b1ee9b7..cdbf4d7620c6 100644 --- a/arch/arm/mach-omap1/time.c +++ b/arch/arm/mach-omap1/time.c @@ -226,8 +226,8 @@ unsigned long long sched_clock(void) #ifdef CONFIG_OMAP_32K_TIMER -#ifdef CONFIG_ARCH_OMAP1510 -#error OMAP 32KHz timer does not currently work on 1510! +#ifdef CONFIG_ARCH_OMAP15XX +#error OMAP 32KHz timer does not currently work on 15XX! #endif /* diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index e3c14d6b4328..4e8d19318d22 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -102,8 +102,8 @@ config CPU_ARM922T # ARM925T config CPU_ARM925T bool "Support ARM925T processor" if ARCH_OMAP1 - depends on ARCH_OMAP1510 - default y if ARCH_OMAP1510 + depends on ARCH_OMAP15XX + default y if ARCH_OMAP15XX select CPU_32v4 select CPU_ABRT_EV4T select CPU_CACHE_V4WT From 1a8bfa1eb998af6e650ad26201f7cae9f2a2fdc8 Mon Sep 17 00:00:00 2001 From: Tony Lindgren Date: Thu, 10 Nov 2005 14:26:50 +0000 Subject: [PATCH 092/121] [ARM] 3142/1: OMAP 2/5: Update files common to omap1 and omap2 Patch from Tony Lindgren This patch syncs the mainline kernel with linux-omap tree. The highlights of the patch are: - Serial port and framebuffer init improvments by Imre Deak - Common omap pin mux framework by Tony Lindgren - Common omap clock framework by Tony Lindren Signed-off-by: Tony Lindgren Signed-off-by: Russell King --- arch/arm/plat-omap/Makefile | 2 +- arch/arm/plat-omap/clock.c | 1443 +++++----------------------------- arch/arm/plat-omap/clock.h | 120 --- arch/arm/plat-omap/common.c | 48 +- arch/arm/plat-omap/devices.c | 381 +++++++++ arch/arm/plat-omap/dma.c | 938 ++++++++++++++-------- arch/arm/plat-omap/gpio.c | 40 +- arch/arm/plat-omap/mcbsp.c | 22 +- arch/arm/plat-omap/mux.c | 65 +- arch/arm/plat-omap/pm.c | 105 ++- arch/arm/plat-omap/sleep.S | 139 +++- arch/arm/plat-omap/sram.c | 143 +++- arch/arm/plat-omap/sram.h | 21 - arch/arm/plat-omap/usb.c | 11 +- 14 files changed, 1661 insertions(+), 1817 deletions(-) delete mode 100644 arch/arm/plat-omap/clock.h create mode 100644 arch/arm/plat-omap/devices.c delete mode 100644 arch/arm/plat-omap/sram.h diff --git a/arch/arm/plat-omap/Makefile b/arch/arm/plat-omap/Makefile index 7e144f9cad1c..9ccf1943fc94 100644 --- a/arch/arm/plat-omap/Makefile +++ b/arch/arm/plat-omap/Makefile @@ -3,7 +3,7 @@ # # Common support -obj-y := common.o sram.o sram-fn.o clock.o dma.o mux.o gpio.o mcbsp.o usb.o +obj-y := common.o sram.o sram-fn.o clock.o devices.o dma.o mux.o gpio.o mcbsp.o usb.o obj-m := obj-n := obj- := diff --git a/arch/arm/plat-omap/clock.c b/arch/arm/plat-omap/clock.c index a020fe16428f..7ce39b986e23 100644 --- a/arch/arm/plat-omap/clock.c +++ b/arch/arm/plat-omap/clock.c @@ -1,15 +1,20 @@ /* * linux/arch/arm/plat-omap/clock.c * - * Copyright (C) 2004 Nokia corporation + * Copyright (C) 2004 - 2005 Nokia corporation * Written by Tuukka Tikkanen * + * Modified for omap shared clock framework by Tony Lindgren + * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. */ -#include +#include +#include #include +#include +#include #include #include #include @@ -18,562 +23,20 @@ #include #include #include -#include -#include -#include "clock.h" -#include "sram.h" +#include -static LIST_HEAD(clocks); +LIST_HEAD(clocks); static DECLARE_MUTEX(clocks_sem); -static DEFINE_SPINLOCK(clockfw_lock); -static void propagate_rate(struct clk * clk); -/* UART clock function */ -static int set_uart_rate(struct clk * clk, unsigned long rate); -/* External clock (MCLK & BCLK) functions */ -static int set_ext_clk_rate(struct clk * clk, unsigned long rate); -static long round_ext_clk_rate(struct clk * clk, unsigned long rate); -static void init_ext_clk(struct clk * clk); -/* MPU virtual clock functions */ -static int select_table_rate(struct clk * clk, unsigned long rate); -static long round_to_table_rate(struct clk * clk, unsigned long rate); -void clk_setdpll(__u16, __u16); +DEFINE_SPINLOCK(clockfw_lock); -static struct mpu_rate rate_table[] = { - /* MPU MHz, xtal MHz, dpll1 MHz, CKCTL, DPLL_CTL - * armdiv, dspdiv, dspmmu, tcdiv, perdiv, lcddiv - */ -#if defined(CONFIG_OMAP_ARM_216MHZ) - { 216000000, 12000000, 216000000, 0x050d, 0x2910 }, /* 1/1/2/2/2/8 */ -#endif -#if defined(CONFIG_OMAP_ARM_195MHZ) - { 195000000, 13000000, 195000000, 0x050e, 0x2790 }, /* 1/1/2/2/4/8 */ -#endif -#if defined(CONFIG_OMAP_ARM_192MHZ) - { 192000000, 19200000, 192000000, 0x050f, 0x2510 }, /* 1/1/2/2/8/8 */ - { 192000000, 12000000, 192000000, 0x050f, 0x2810 }, /* 1/1/2/2/8/8 */ - { 96000000, 12000000, 192000000, 0x055f, 0x2810 }, /* 2/2/2/2/8/8 */ - { 48000000, 12000000, 192000000, 0x0baf, 0x2810 }, /* 4/8/4/4/8/8 */ - { 24000000, 12000000, 192000000, 0x0fff, 0x2810 }, /* 8/8/8/8/8/8 */ -#endif -#if defined(CONFIG_OMAP_ARM_182MHZ) - { 182000000, 13000000, 182000000, 0x050e, 0x2710 }, /* 1/1/2/2/4/8 */ -#endif -#if defined(CONFIG_OMAP_ARM_168MHZ) - { 168000000, 12000000, 168000000, 0x010f, 0x2710 }, /* 1/1/1/2/8/8 */ -#endif -#if defined(CONFIG_OMAP_ARM_150MHZ) - { 150000000, 12000000, 150000000, 0x010a, 0x2cb0 }, /* 1/1/1/2/4/4 */ -#endif -#if defined(CONFIG_OMAP_ARM_120MHZ) - { 120000000, 12000000, 120000000, 0x010a, 0x2510 }, /* 1/1/1/2/4/4 */ -#endif -#if defined(CONFIG_OMAP_ARM_96MHZ) - { 96000000, 12000000, 96000000, 0x0005, 0x2410 }, /* 1/1/1/1/2/2 */ -#endif -#if defined(CONFIG_OMAP_ARM_60MHZ) - { 60000000, 12000000, 60000000, 0x0005, 0x2290 }, /* 1/1/1/1/2/2 */ -#endif -#if defined(CONFIG_OMAP_ARM_30MHZ) - { 30000000, 12000000, 60000000, 0x0555, 0x2290 }, /* 2/2/2/2/2/2 */ -#endif - { 0, 0, 0, 0, 0 }, -}; +static struct clk_functions *arch_clock; +/*------------------------------------------------------------------------- + * Standard clock functions defined in asm/hardware/clock.h + *-------------------------------------------------------------------------*/ -static void ckctl_recalc(struct clk * clk); -int __clk_enable(struct clk *clk); -void __clk_disable(struct clk *clk); -void __clk_unuse(struct clk *clk); -int __clk_use(struct clk *clk); - - -static void followparent_recalc(struct clk * clk) -{ - clk->rate = clk->parent->rate; -} - - -static void watchdog_recalc(struct clk * clk) -{ - clk->rate = clk->parent->rate / 14; -} - -static void uart_recalc(struct clk * clk) -{ - unsigned int val = omap_readl(clk->enable_reg); - if (val & clk->enable_bit) - clk->rate = 48000000; - else - clk->rate = 12000000; -} - -static struct clk ck_ref = { - .name = "ck_ref", - .rate = 12000000, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - ALWAYS_ENABLED, -}; - -static struct clk ck_dpll1 = { - .name = "ck_dpll1", - .parent = &ck_ref, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - RATE_PROPAGATES | ALWAYS_ENABLED, -}; - -static struct clk ck_dpll1out = { - .name = "ck_dpll1out", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP16XX, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_CKOUT_ARM, - .recalc = &followparent_recalc, -}; - -static struct clk arm_ck = { - .name = "arm_ck", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - RATE_CKCTL | RATE_PROPAGATES | ALWAYS_ENABLED, - .rate_offset = CKCTL_ARMDIV_OFFSET, - .recalc = &ckctl_recalc, -}; - -static struct clk armper_ck = { - .name = "armper_ck", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - RATE_CKCTL, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_PERCK, - .rate_offset = CKCTL_PERDIV_OFFSET, - .recalc = &ckctl_recalc, -}; - -static struct clk arm_gpio_ck = { - .name = "arm_gpio_ck", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP1510, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_GPIOCK, - .recalc = &followparent_recalc, -}; - -static struct clk armxor_ck = { - .name = "armxor_ck", - .parent = &ck_ref, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_XORPCK, - .recalc = &followparent_recalc, -}; - -static struct clk armtim_ck = { - .name = "armtim_ck", - .parent = &ck_ref, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_TIMCK, - .recalc = &followparent_recalc, -}; - -static struct clk armwdt_ck = { - .name = "armwdt_ck", - .parent = &ck_ref, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_WDTCK, - .recalc = &watchdog_recalc, -}; - -static struct clk arminth_ck16xx = { - .name = "arminth_ck", - .parent = &arm_ck, - .flags = CLOCK_IN_OMAP16XX | ALWAYS_ENABLED, - .recalc = &followparent_recalc, - /* Note: On 16xx the frequency can be divided by 2 by programming - * ARM_CKCTL:ARM_INTHCK_SEL(14) to 1 - * - * 1510 version is in TC clocks. - */ -}; - -static struct clk dsp_ck = { - .name = "dsp_ck", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - RATE_CKCTL, - .enable_reg = ARM_CKCTL, - .enable_bit = EN_DSPCK, - .rate_offset = CKCTL_DSPDIV_OFFSET, - .recalc = &ckctl_recalc, -}; - -static struct clk dspmmu_ck = { - .name = "dspmmu_ck", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - RATE_CKCTL | ALWAYS_ENABLED, - .rate_offset = CKCTL_DSPMMUDIV_OFFSET, - .recalc = &ckctl_recalc, -}; - -static struct clk dspper_ck = { - .name = "dspper_ck", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - RATE_CKCTL | DSP_DOMAIN_CLOCK | VIRTUAL_IO_ADDRESS, - .enable_reg = DSP_IDLECT2, - .enable_bit = EN_PERCK, - .rate_offset = CKCTL_PERDIV_OFFSET, - .recalc = &followparent_recalc, - //.recalc = &ckctl_recalc, -}; - -static struct clk dspxor_ck = { - .name = "dspxor_ck", - .parent = &ck_ref, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - DSP_DOMAIN_CLOCK | VIRTUAL_IO_ADDRESS, - .enable_reg = DSP_IDLECT2, - .enable_bit = EN_XORPCK, - .recalc = &followparent_recalc, -}; - -static struct clk dsptim_ck = { - .name = "dsptim_ck", - .parent = &ck_ref, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - DSP_DOMAIN_CLOCK | VIRTUAL_IO_ADDRESS, - .enable_reg = DSP_IDLECT2, - .enable_bit = EN_DSPTIMCK, - .recalc = &followparent_recalc, -}; - -static struct clk tc_ck = { - .name = "tc_ck", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | CLOCK_IN_OMAP730 | - RATE_CKCTL | RATE_PROPAGATES | ALWAYS_ENABLED, - .rate_offset = CKCTL_TCDIV_OFFSET, - .recalc = &ckctl_recalc, -}; - -static struct clk arminth_ck1510 = { - .name = "arminth_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP1510 | ALWAYS_ENABLED, - .recalc = &followparent_recalc, - /* Note: On 1510 the frequency follows TC_CK - * - * 16xx version is in MPU clocks. - */ -}; - -static struct clk tipb_ck = { - .name = "tibp_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP1510 | ALWAYS_ENABLED, - .recalc = &followparent_recalc, -}; - -static struct clk l3_ocpi_ck = { - .name = "l3_ocpi_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP16XX, - .enable_reg = ARM_IDLECT3, - .enable_bit = EN_OCPI_CK, - .recalc = &followparent_recalc, -}; - -static struct clk tc1_ck = { - .name = "tc1_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP16XX, - .enable_reg = ARM_IDLECT3, - .enable_bit = EN_TC1_CK, - .recalc = &followparent_recalc, -}; - -static struct clk tc2_ck = { - .name = "tc2_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP16XX, - .enable_reg = ARM_IDLECT3, - .enable_bit = EN_TC2_CK, - .recalc = &followparent_recalc, -}; - -static struct clk dma_ck = { - .name = "dma_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - ALWAYS_ENABLED, - .recalc = &followparent_recalc, -}; - -static struct clk dma_lcdfree_ck = { - .name = "dma_lcdfree_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP16XX | ALWAYS_ENABLED, - .recalc = &followparent_recalc, -}; - -static struct clk api_ck = { - .name = "api_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_APICK, - .recalc = &followparent_recalc, -}; - -static struct clk lb_ck = { - .name = "lb_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP1510, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_LBCK, - .recalc = &followparent_recalc, -}; - -static struct clk rhea1_ck = { - .name = "rhea1_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP16XX | ALWAYS_ENABLED, - .recalc = &followparent_recalc, -}; - -static struct clk rhea2_ck = { - .name = "rhea2_ck", - .parent = &tc_ck, - .flags = CLOCK_IN_OMAP16XX | ALWAYS_ENABLED, - .recalc = &followparent_recalc, -}; - -static struct clk lcd_ck = { - .name = "lcd_ck", - .parent = &ck_dpll1, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | CLOCK_IN_OMAP730 | - RATE_CKCTL, - .enable_reg = ARM_IDLECT2, - .enable_bit = EN_LCDCK, - .rate_offset = CKCTL_LCDDIV_OFFSET, - .recalc = &ckctl_recalc, -}; - -static struct clk uart1_1510 = { - .name = "uart1_ck", - /* Direct from ULPD, no parent */ - .rate = 12000000, - .flags = CLOCK_IN_OMAP1510 | ENABLE_REG_32BIT | ALWAYS_ENABLED, - .enable_reg = MOD_CONF_CTRL_0, - .enable_bit = 29, /* Chooses between 12MHz and 48MHz */ - .set_rate = &set_uart_rate, - .recalc = &uart_recalc, -}; - -static struct clk uart1_16xx = { - .name = "uart1_ck", - /* Direct from ULPD, no parent */ - .rate = 48000000, - .flags = CLOCK_IN_OMAP16XX | RATE_FIXED | ENABLE_REG_32BIT, - .enable_reg = MOD_CONF_CTRL_0, - .enable_bit = 29, -}; - -static struct clk uart2_ck = { - .name = "uart2_ck", - /* Direct from ULPD, no parent */ - .rate = 12000000, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | ENABLE_REG_32BIT | - ALWAYS_ENABLED, - .enable_reg = MOD_CONF_CTRL_0, - .enable_bit = 30, /* Chooses between 12MHz and 48MHz */ - .set_rate = &set_uart_rate, - .recalc = &uart_recalc, -}; - -static struct clk uart3_1510 = { - .name = "uart3_ck", - /* Direct from ULPD, no parent */ - .rate = 12000000, - .flags = CLOCK_IN_OMAP1510 | ENABLE_REG_32BIT | ALWAYS_ENABLED, - .enable_reg = MOD_CONF_CTRL_0, - .enable_bit = 31, /* Chooses between 12MHz and 48MHz */ - .set_rate = &set_uart_rate, - .recalc = &uart_recalc, -}; - -static struct clk uart3_16xx = { - .name = "uart3_ck", - /* Direct from ULPD, no parent */ - .rate = 48000000, - .flags = CLOCK_IN_OMAP16XX | RATE_FIXED | ENABLE_REG_32BIT, - .enable_reg = MOD_CONF_CTRL_0, - .enable_bit = 31, -}; - -static struct clk usb_clko = { /* 6 MHz output on W4_USB_CLKO */ - .name = "usb_clko", - /* Direct from ULPD, no parent */ - .rate = 6000000, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - RATE_FIXED | ENABLE_REG_32BIT, - .enable_reg = ULPD_CLOCK_CTRL, - .enable_bit = USB_MCLK_EN_BIT, -}; - -static struct clk usb_hhc_ck1510 = { - .name = "usb_hhc_ck", - /* Direct from ULPD, no parent */ - .rate = 48000000, /* Actually 2 clocks, 12MHz and 48MHz */ - .flags = CLOCK_IN_OMAP1510 | - RATE_FIXED | ENABLE_REG_32BIT, - .enable_reg = MOD_CONF_CTRL_0, - .enable_bit = USB_HOST_HHC_UHOST_EN, -}; - -static struct clk usb_hhc_ck16xx = { - .name = "usb_hhc_ck", - /* Direct from ULPD, no parent */ - .rate = 48000000, - /* OTG_SYSCON_2.OTG_PADEN == 0 (not 1510-compatible) */ - .flags = CLOCK_IN_OMAP16XX | - RATE_FIXED | ENABLE_REG_32BIT, - .enable_reg = OTG_BASE + 0x08 /* OTG_SYSCON_2 */, - .enable_bit = 8 /* UHOST_EN */, -}; - -static struct clk usb_dc_ck = { - .name = "usb_dc_ck", - /* Direct from ULPD, no parent */ - .rate = 48000000, - .flags = CLOCK_IN_OMAP16XX | RATE_FIXED, - .enable_reg = SOFT_REQ_REG, - .enable_bit = 4, -}; - -static struct clk mclk_1510 = { - .name = "mclk", - /* Direct from ULPD, no parent. May be enabled by ext hardware. */ - .rate = 12000000, - .flags = CLOCK_IN_OMAP1510 | RATE_FIXED, -}; - -static struct clk mclk_16xx = { - .name = "mclk", - /* Direct from ULPD, no parent. May be enabled by ext hardware. */ - .flags = CLOCK_IN_OMAP16XX, - .enable_reg = COM_CLK_DIV_CTRL_SEL, - .enable_bit = COM_ULPD_PLL_CLK_REQ, - .set_rate = &set_ext_clk_rate, - .round_rate = &round_ext_clk_rate, - .init = &init_ext_clk, -}; - -static struct clk bclk_1510 = { - .name = "bclk", - /* Direct from ULPD, no parent. May be enabled by ext hardware. */ - .rate = 12000000, - .flags = CLOCK_IN_OMAP1510 | RATE_FIXED, -}; - -static struct clk bclk_16xx = { - .name = "bclk", - /* Direct from ULPD, no parent. May be enabled by ext hardware. */ - .flags = CLOCK_IN_OMAP16XX, - .enable_reg = SWD_CLK_DIV_CTRL_SEL, - .enable_bit = SWD_ULPD_PLL_CLK_REQ, - .set_rate = &set_ext_clk_rate, - .round_rate = &round_ext_clk_rate, - .init = &init_ext_clk, -}; - -static struct clk mmc1_ck = { - .name = "mmc1_ck", - /* Functional clock is direct from ULPD, interface clock is ARMPER */ - .parent = &armper_ck, - .rate = 48000000, - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - RATE_FIXED | ENABLE_REG_32BIT, - .enable_reg = MOD_CONF_CTRL_0, - .enable_bit = 23, -}; - -static struct clk mmc2_ck = { - .name = "mmc2_ck", - /* Functional clock is direct from ULPD, interface clock is ARMPER */ - .parent = &armper_ck, - .rate = 48000000, - .flags = CLOCK_IN_OMAP16XX | - RATE_FIXED | ENABLE_REG_32BIT, - .enable_reg = MOD_CONF_CTRL_0, - .enable_bit = 20, -}; - -static struct clk virtual_ck_mpu = { - .name = "mpu", - .flags = CLOCK_IN_OMAP1510 | CLOCK_IN_OMAP16XX | - VIRTUAL_CLOCK | ALWAYS_ENABLED, - .parent = &arm_ck, /* Is smarter alias for */ - .recalc = &followparent_recalc, - .set_rate = &select_table_rate, - .round_rate = &round_to_table_rate, -}; - - -static struct clk * onchip_clks[] = { - /* non-ULPD clocks */ - &ck_ref, - &ck_dpll1, - /* CK_GEN1 clocks */ - &ck_dpll1out, - &arm_ck, - &armper_ck, - &arm_gpio_ck, - &armxor_ck, - &armtim_ck, - &armwdt_ck, - &arminth_ck1510, &arminth_ck16xx, - /* CK_GEN2 clocks */ - &dsp_ck, - &dspmmu_ck, - &dspper_ck, - &dspxor_ck, - &dsptim_ck, - /* CK_GEN3 clocks */ - &tc_ck, - &tipb_ck, - &l3_ocpi_ck, - &tc1_ck, - &tc2_ck, - &dma_ck, - &dma_lcdfree_ck, - &api_ck, - &lb_ck, - &rhea1_ck, - &rhea2_ck, - &lcd_ck, - /* ULPD clocks */ - &uart1_1510, - &uart1_16xx, - &uart2_ck, - &uart3_1510, - &uart3_16xx, - &usb_clko, - &usb_hhc_ck1510, &usb_hhc_ck16xx, - &usb_dc_ck, - &mclk_1510, &mclk_16xx, - &bclk_1510, &bclk_16xx, - &mmc1_ck, - &mmc2_ck, - /* Virtual clocks */ - &virtual_ck_mpu, -}; - -struct clk *clk_get(struct device *dev, const char *id) +struct clk * clk_get(struct device *dev, const char *id) { struct clk *p, *clk = ERR_PTR(-ENOENT); @@ -590,6 +53,89 @@ struct clk *clk_get(struct device *dev, const char *id) } EXPORT_SYMBOL(clk_get); +int clk_enable(struct clk *clk) +{ + unsigned long flags; + int ret = 0; + + spin_lock_irqsave(&clockfw_lock, flags); + if (clk->enable) + ret = clk->enable(clk); + else if (arch_clock->clk_enable) + ret = arch_clock->clk_enable(clk); + else + printk(KERN_ERR "Could not enable clock %s\n", clk->name); + spin_unlock_irqrestore(&clockfw_lock, flags); + + return ret; +} +EXPORT_SYMBOL(clk_enable); + +void clk_disable(struct clk *clk) +{ + unsigned long flags; + + spin_lock_irqsave(&clockfw_lock, flags); + if (clk->disable) + clk->disable(clk); + else if (arch_clock->clk_disable) + arch_clock->clk_disable(clk); + else + printk(KERN_ERR "Could not disable clock %s\n", clk->name); + spin_unlock_irqrestore(&clockfw_lock, flags); +} +EXPORT_SYMBOL(clk_disable); + +int clk_use(struct clk *clk) +{ + unsigned long flags; + int ret = 0; + + spin_lock_irqsave(&clockfw_lock, flags); + if (arch_clock->clk_use) + ret = arch_clock->clk_use(clk); + spin_unlock_irqrestore(&clockfw_lock, flags); + + return ret; +} +EXPORT_SYMBOL(clk_use); + +void clk_unuse(struct clk *clk) +{ + unsigned long flags; + + spin_lock_irqsave(&clockfw_lock, flags); + if (arch_clock->clk_unuse) + arch_clock->clk_unuse(clk); + spin_unlock_irqrestore(&clockfw_lock, flags); +} +EXPORT_SYMBOL(clk_unuse); + +int clk_get_usecount(struct clk *clk) +{ + unsigned long flags; + int ret = 0; + + spin_lock_irqsave(&clockfw_lock, flags); + ret = clk->usecount; + spin_unlock_irqrestore(&clockfw_lock, flags); + + return ret; +} +EXPORT_SYMBOL(clk_get_usecount); + +unsigned long clk_get_rate(struct clk *clk) +{ + unsigned long flags; + unsigned long ret = 0; + + spin_lock_irqsave(&clockfw_lock, flags); + ret = clk->rate; + spin_unlock_irqrestore(&clockfw_lock, flags); + + return ret; +} +EXPORT_SYMBOL(clk_get_rate); void clk_put(struct clk *clk) { @@ -598,526 +144,109 @@ void clk_put(struct clk *clk) } EXPORT_SYMBOL(clk_put); - -int __clk_enable(struct clk *clk) -{ - __u16 regval16; - __u32 regval32; - - if (clk->flags & ALWAYS_ENABLED) - return 0; - - if (unlikely(clk->enable_reg == 0)) { - printk(KERN_ERR "clock.c: Enable for %s without enable code\n", - clk->name); - return 0; - } - - if (clk->flags & DSP_DOMAIN_CLOCK) { - __clk_use(&api_ck); - } - - if (clk->flags & ENABLE_REG_32BIT) { - if (clk->flags & VIRTUAL_IO_ADDRESS) { - regval32 = __raw_readl(clk->enable_reg); - regval32 |= (1 << clk->enable_bit); - __raw_writel(regval32, clk->enable_reg); - } else { - regval32 = omap_readl(clk->enable_reg); - regval32 |= (1 << clk->enable_bit); - omap_writel(regval32, clk->enable_reg); - } - } else { - if (clk->flags & VIRTUAL_IO_ADDRESS) { - regval16 = __raw_readw(clk->enable_reg); - regval16 |= (1 << clk->enable_bit); - __raw_writew(regval16, clk->enable_reg); - } else { - regval16 = omap_readw(clk->enable_reg); - regval16 |= (1 << clk->enable_bit); - omap_writew(regval16, clk->enable_reg); - } - } - - if (clk->flags & DSP_DOMAIN_CLOCK) { - __clk_unuse(&api_ck); - } - - return 0; -} - - -void __clk_disable(struct clk *clk) -{ - __u16 regval16; - __u32 regval32; - - if (clk->enable_reg == 0) - return; - - if (clk->flags & DSP_DOMAIN_CLOCK) { - __clk_use(&api_ck); - } - - if (clk->flags & ENABLE_REG_32BIT) { - if (clk->flags & VIRTUAL_IO_ADDRESS) { - regval32 = __raw_readl(clk->enable_reg); - regval32 &= ~(1 << clk->enable_bit); - __raw_writel(regval32, clk->enable_reg); - } else { - regval32 = omap_readl(clk->enable_reg); - regval32 &= ~(1 << clk->enable_bit); - omap_writel(regval32, clk->enable_reg); - } - } else { - if (clk->flags & VIRTUAL_IO_ADDRESS) { - regval16 = __raw_readw(clk->enable_reg); - regval16 &= ~(1 << clk->enable_bit); - __raw_writew(regval16, clk->enable_reg); - } else { - regval16 = omap_readw(clk->enable_reg); - regval16 &= ~(1 << clk->enable_bit); - omap_writew(regval16, clk->enable_reg); - } - } - - if (clk->flags & DSP_DOMAIN_CLOCK) { - __clk_unuse(&api_ck); - } -} - - -void __clk_unuse(struct clk *clk) -{ - if (clk->usecount > 0 && !(--clk->usecount)) { - __clk_disable(clk); - if (likely(clk->parent)) - __clk_unuse(clk->parent); - } -} - - -int __clk_use(struct clk *clk) -{ - int ret = 0; - if (clk->usecount++ == 0) { - if (likely(clk->parent)) - ret = __clk_use(clk->parent); - - if (unlikely(ret != 0)) { - clk->usecount--; - return ret; - } - - ret = __clk_enable(clk); - - if (unlikely(ret != 0) && clk->parent) { - __clk_unuse(clk->parent); - clk->usecount--; - } - } - - return ret; -} - - -int clk_enable(struct clk *clk) -{ - unsigned long flags; - int ret; - - spin_lock_irqsave(&clockfw_lock, flags); - ret = __clk_enable(clk); - spin_unlock_irqrestore(&clockfw_lock, flags); - return ret; -} -EXPORT_SYMBOL(clk_enable); - - -void clk_disable(struct clk *clk) -{ - unsigned long flags; - - spin_lock_irqsave(&clockfw_lock, flags); - __clk_disable(clk); - spin_unlock_irqrestore(&clockfw_lock, flags); -} -EXPORT_SYMBOL(clk_disable); - - -int clk_use(struct clk *clk) -{ - unsigned long flags; - int ret = 0; - - spin_lock_irqsave(&clockfw_lock, flags); - ret = __clk_use(clk); - spin_unlock_irqrestore(&clockfw_lock, flags); - return ret; -} -EXPORT_SYMBOL(clk_use); - - -void clk_unuse(struct clk *clk) -{ - unsigned long flags; - - spin_lock_irqsave(&clockfw_lock, flags); - __clk_unuse(clk); - spin_unlock_irqrestore(&clockfw_lock, flags); -} -EXPORT_SYMBOL(clk_unuse); - - -int clk_get_usecount(struct clk *clk) -{ - return clk->usecount; -} -EXPORT_SYMBOL(clk_get_usecount); - - -unsigned long clk_get_rate(struct clk *clk) -{ - return clk->rate; -} -EXPORT_SYMBOL(clk_get_rate); - - -static __u16 verify_ckctl_value(__u16 newval) -{ - /* This function checks for following limitations set - * by the hardware (all conditions must be true): - * DSPMMU_CK == DSP_CK or DSPMMU_CK == DSP_CK/2 - * ARM_CK >= TC_CK - * DSP_CK >= TC_CK - * DSPMMU_CK >= TC_CK - * - * In addition following rules are enforced: - * LCD_CK <= TC_CK - * ARMPER_CK <= TC_CK - * - * However, maximum frequencies are not checked for! - */ - __u8 per_exp; - __u8 lcd_exp; - __u8 arm_exp; - __u8 dsp_exp; - __u8 tc_exp; - __u8 dspmmu_exp; - - per_exp = (newval >> CKCTL_PERDIV_OFFSET) & 3; - lcd_exp = (newval >> CKCTL_LCDDIV_OFFSET) & 3; - arm_exp = (newval >> CKCTL_ARMDIV_OFFSET) & 3; - dsp_exp = (newval >> CKCTL_DSPDIV_OFFSET) & 3; - tc_exp = (newval >> CKCTL_TCDIV_OFFSET) & 3; - dspmmu_exp = (newval >> CKCTL_DSPMMUDIV_OFFSET) & 3; - - if (dspmmu_exp < dsp_exp) - dspmmu_exp = dsp_exp; - if (dspmmu_exp > dsp_exp+1) - dspmmu_exp = dsp_exp+1; - if (tc_exp < arm_exp) - tc_exp = arm_exp; - if (tc_exp < dspmmu_exp) - tc_exp = dspmmu_exp; - if (tc_exp > lcd_exp) - lcd_exp = tc_exp; - if (tc_exp > per_exp) - per_exp = tc_exp; - - newval &= 0xf000; - newval |= per_exp << CKCTL_PERDIV_OFFSET; - newval |= lcd_exp << CKCTL_LCDDIV_OFFSET; - newval |= arm_exp << CKCTL_ARMDIV_OFFSET; - newval |= dsp_exp << CKCTL_DSPDIV_OFFSET; - newval |= tc_exp << CKCTL_TCDIV_OFFSET; - newval |= dspmmu_exp << CKCTL_DSPMMUDIV_OFFSET; - - return newval; -} - - -static int calc_dsor_exp(struct clk *clk, unsigned long rate) -{ - /* Note: If target frequency is too low, this function will return 4, - * which is invalid value. Caller must check for this value and act - * accordingly. - * - * Note: This function does not check for following limitations set - * by the hardware (all conditions must be true): - * DSPMMU_CK == DSP_CK or DSPMMU_CK == DSP_CK/2 - * ARM_CK >= TC_CK - * DSP_CK >= TC_CK - * DSPMMU_CK >= TC_CK - */ - unsigned long realrate; - struct clk * parent; - unsigned dsor_exp; - - if (unlikely(!(clk->flags & RATE_CKCTL))) - return -EINVAL; - - parent = clk->parent; - if (unlikely(parent == 0)) - return -EIO; - - realrate = parent->rate; - for (dsor_exp=0; dsor_exp<4; dsor_exp++) { - if (realrate <= rate) - break; - - realrate /= 2; - } - - return dsor_exp; -} - - -static void ckctl_recalc(struct clk * clk) -{ - int dsor; - - /* Calculate divisor encoded as 2-bit exponent */ - if (clk->flags & DSP_DOMAIN_CLOCK) { - /* The clock control bits are in DSP domain, - * so api_ck is needed for access. - * Note that DSP_CKCTL virt addr = phys addr, so - * we must use __raw_readw() instead of omap_readw(). - */ - __clk_use(&api_ck); - dsor = 1 << (3 & (__raw_readw(DSP_CKCTL) >> clk->rate_offset)); - __clk_unuse(&api_ck); - } else { - dsor = 1 << (3 & (omap_readw(ARM_CKCTL) >> clk->rate_offset)); - } - if (unlikely(clk->rate == clk->parent->rate / dsor)) - return; /* No change, quick exit */ - clk->rate = clk->parent->rate / dsor; - - if (unlikely(clk->flags & RATE_PROPAGATES)) - propagate_rate(clk); -} - +/*------------------------------------------------------------------------- + * Optional clock functions defined in asm/hardware/clock.h + *-------------------------------------------------------------------------*/ long clk_round_rate(struct clk *clk, unsigned long rate) { - int dsor_exp; + unsigned long flags; + long ret = 0; - if (clk->flags & RATE_FIXED) - return clk->rate; + spin_lock_irqsave(&clockfw_lock, flags); + if (arch_clock->clk_round_rate) + ret = arch_clock->clk_round_rate(clk, rate); + spin_unlock_irqrestore(&clockfw_lock, flags); - if (clk->flags & RATE_CKCTL) { - dsor_exp = calc_dsor_exp(clk, rate); - if (dsor_exp < 0) - return dsor_exp; - if (dsor_exp > 3) - dsor_exp = 3; - return clk->parent->rate / (1 << dsor_exp); - } - - if(clk->round_rate != 0) - return clk->round_rate(clk, rate); - - return clk->rate; + return ret; } EXPORT_SYMBOL(clk_round_rate); - -static void propagate_rate(struct clk * clk) -{ - struct clk ** clkp; - - for (clkp = onchip_clks; clkp < onchip_clks+ARRAY_SIZE(onchip_clks); clkp++) { - if (likely((*clkp)->parent != clk)) continue; - if (likely((*clkp)->recalc)) - (*clkp)->recalc(*clkp); - } -} - - -static int select_table_rate(struct clk * clk, unsigned long rate) -{ - /* Find the highest supported frequency <= rate and switch to it */ - struct mpu_rate * ptr; - - if (clk != &virtual_ck_mpu) - return -EINVAL; - - for (ptr = rate_table; ptr->rate; ptr++) { - if (ptr->xtal != ck_ref.rate) - continue; - - /* DPLL1 cannot be reprogrammed without risking system crash */ - if (likely(ck_dpll1.rate!=0) && ptr->pll_rate != ck_dpll1.rate) - continue; - - /* Can check only after xtal frequency check */ - if (ptr->rate <= rate) - break; - } - - if (!ptr->rate) - return -EINVAL; - - /* - * In most cases we should not need to reprogram DPLL. - * Reprogramming the DPLL is tricky, it must be done from SRAM. - */ - omap_sram_reprogram_clock(ptr->dpllctl_val, ptr->ckctl_val); - - ck_dpll1.rate = ptr->pll_rate; - propagate_rate(&ck_dpll1); - return 0; -} - - -static long round_to_table_rate(struct clk * clk, unsigned long rate) -{ - /* Find the highest supported frequency <= rate */ - struct mpu_rate * ptr; - long highest_rate; - - if (clk != &virtual_ck_mpu) - return -EINVAL; - - highest_rate = -EINVAL; - - for (ptr = rate_table; ptr->rate; ptr++) { - if (ptr->xtal != ck_ref.rate) - continue; - - highest_rate = ptr->rate; - - /* Can check only after xtal frequency check */ - if (ptr->rate <= rate) - break; - } - - return highest_rate; -} - - int clk_set_rate(struct clk *clk, unsigned long rate) { - int ret = -EINVAL; - int dsor_exp; - __u16 regval; - unsigned long flags; + unsigned long flags; + int ret = 0; - if (clk->flags & RATE_CKCTL) { - dsor_exp = calc_dsor_exp(clk, rate); - if (dsor_exp > 3) - dsor_exp = -EINVAL; - if (dsor_exp < 0) - return dsor_exp; - - spin_lock_irqsave(&clockfw_lock, flags); - regval = omap_readw(ARM_CKCTL); - regval &= ~(3 << clk->rate_offset); - regval |= dsor_exp << clk->rate_offset; - regval = verify_ckctl_value(regval); - omap_writew(regval, ARM_CKCTL); - clk->rate = clk->parent->rate / (1 << dsor_exp); - spin_unlock_irqrestore(&clockfw_lock, flags); - ret = 0; - } else if(clk->set_rate != 0) { - spin_lock_irqsave(&clockfw_lock, flags); - ret = clk->set_rate(clk, rate); - spin_unlock_irqrestore(&clockfw_lock, flags); - } - - if (unlikely(ret == 0 && (clk->flags & RATE_PROPAGATES))) - propagate_rate(clk); + spin_lock_irqsave(&clockfw_lock, flags); + if (arch_clock->clk_set_rate) + ret = arch_clock->clk_set_rate(clk, rate); + spin_unlock_irqrestore(&clockfw_lock, flags); return ret; } EXPORT_SYMBOL(clk_set_rate); - -static unsigned calc_ext_dsor(unsigned long rate) +int clk_set_parent(struct clk *clk, struct clk *parent) { - unsigned dsor; + unsigned long flags; + int ret = 0; - /* MCLK and BCLK divisor selection is not linear: - * freq = 96MHz / dsor - * - * RATIO_SEL range: dsor <-> RATIO_SEL - * 0..6: (RATIO_SEL+2) <-> (dsor-2) - * 6..48: (8+(RATIO_SEL-6)*2) <-> ((dsor-8)/2+6) - * Minimum dsor is 2 and maximum is 96. Odd divisors starting from 9 - * can not be used. - */ - for (dsor = 2; dsor < 96; ++dsor) { - if ((dsor & 1) && dsor > 8) - continue; - if (rate >= 96000000 / dsor) - break; + spin_lock_irqsave(&clockfw_lock, flags); + if (arch_clock->clk_set_parent) + ret = arch_clock->clk_set_parent(clk, parent); + spin_unlock_irqrestore(&clockfw_lock, flags); + + return ret; +} +EXPORT_SYMBOL(clk_set_parent); + +struct clk *clk_get_parent(struct clk *clk) +{ + unsigned long flags; + struct clk * ret = NULL; + + spin_lock_irqsave(&clockfw_lock, flags); + if (arch_clock->clk_get_parent) + ret = arch_clock->clk_get_parent(clk); + spin_unlock_irqrestore(&clockfw_lock, flags); + + return ret; +} +EXPORT_SYMBOL(clk_get_parent); + +/*------------------------------------------------------------------------- + * OMAP specific clock functions shared between omap1 and omap2 + *-------------------------------------------------------------------------*/ + +unsigned int __initdata mpurate; + +/* + * By default we use the rate set by the bootloader. + * You can override this with mpurate= cmdline option. + */ +static int __init omap_clk_setup(char *str) +{ + get_option(&str, &mpurate); + + if (!mpurate) + return 1; + + if (mpurate < 1000) + mpurate *= 1000000; + + return 1; +} +__setup("mpurate=", omap_clk_setup); + +/* Used for clocks that always have same value as the parent clock */ +void followparent_recalc(struct clk *clk) +{ + clk->rate = clk->parent->rate; +} + +/* Propagate rate to children */ +void propagate_rate(struct clk * tclk) +{ + struct clk *clkp; + + list_for_each_entry(clkp, &clocks, node) { + if (likely(clkp->parent != tclk)) + continue; + if (likely((u32)clkp->recalc)) + clkp->recalc(clkp); } - return dsor; } -/* Only needed on 1510 */ -static int set_uart_rate(struct clk * clk, unsigned long rate) -{ - unsigned int val; - - val = omap_readl(clk->enable_reg); - if (rate == 12000000) - val &= ~(1 << clk->enable_bit); - else if (rate == 48000000) - val |= (1 << clk->enable_bit); - else - return -EINVAL; - omap_writel(val, clk->enable_reg); - clk->rate = rate; - - return 0; -} - -static int set_ext_clk_rate(struct clk * clk, unsigned long rate) -{ - unsigned dsor; - __u16 ratio_bits; - - dsor = calc_ext_dsor(rate); - clk->rate = 96000000 / dsor; - if (dsor > 8) - ratio_bits = ((dsor - 8) / 2 + 6) << 2; - else - ratio_bits = (dsor - 2) << 2; - - ratio_bits |= omap_readw(clk->enable_reg) & ~0xfd; - omap_writew(ratio_bits, clk->enable_reg); - - return 0; -} - - -static long round_ext_clk_rate(struct clk * clk, unsigned long rate) -{ - return 96000000 / calc_ext_dsor(rate); -} - - -static void init_ext_clk(struct clk * clk) -{ - unsigned dsor; - __u16 ratio_bits; - - /* Determine current rate and ensure clock is based on 96MHz APLL */ - ratio_bits = omap_readw(clk->enable_reg) & ~1; - omap_writew(ratio_bits, clk->enable_reg); - - ratio_bits = (ratio_bits & 0xfc) >> 2; - if (ratio_bits > 6) - dsor = (ratio_bits - 6) * 2 + 8; - else - dsor = ratio_bits + 2; - - clk-> rate = 96000000 / dsor; -} - - int clk_register(struct clk *clk) { down(&clocks_sem); @@ -1125,6 +254,7 @@ int clk_register(struct clk *clk) if (clk->init) clk->init(clk); up(&clocks_sem); + return 0; } EXPORT_SYMBOL(clk_register); @@ -1137,203 +267,38 @@ void clk_unregister(struct clk *clk) } EXPORT_SYMBOL(clk_unregister); -#ifdef CONFIG_OMAP_RESET_CLOCKS -/* - * Resets some clocks that may be left on from bootloader, - * but leaves serial clocks on. See also omap_late_clk_reset(). - */ -static inline void omap_early_clk_reset(void) +void clk_deny_idle(struct clk *clk) { - //omap_writel(0x3 << 29, MOD_CONF_CTRL_0); + unsigned long flags; + + spin_lock_irqsave(&clockfw_lock, flags); + if (arch_clock->clk_deny_idle) + arch_clock->clk_deny_idle(clk); + spin_unlock_irqrestore(&clockfw_lock, flags); } -#else -#define omap_early_clk_reset() {} -#endif +EXPORT_SYMBOL(clk_deny_idle); -int __init clk_init(void) +void clk_allow_idle(struct clk *clk) { - struct clk ** clkp; - const struct omap_clock_config *info; - int crystal_type = 0; /* Default 12 MHz */ + unsigned long flags; - omap_early_clk_reset(); + spin_lock_irqsave(&clockfw_lock, flags); + if (arch_clock->clk_allow_idle) + arch_clock->clk_allow_idle(clk); + spin_unlock_irqrestore(&clockfw_lock, flags); +} +EXPORT_SYMBOL(clk_allow_idle); - for (clkp = onchip_clks; clkp < onchip_clks+ARRAY_SIZE(onchip_clks); clkp++) { - if (((*clkp)->flags &CLOCK_IN_OMAP1510) && cpu_is_omap1510()) { - clk_register(*clkp); - continue; - } +/*-------------------------------------------------------------------------*/ - if (((*clkp)->flags &CLOCK_IN_OMAP16XX) && cpu_is_omap16xx()) { - clk_register(*clkp); - continue; - } - - if (((*clkp)->flags &CLOCK_IN_OMAP730) && cpu_is_omap730()) { - clk_register(*clkp); - continue; - } +int __init clk_init(struct clk_functions * custom_clocks) +{ + if (!custom_clocks) { + printk(KERN_ERR "No custom clock functions registered\n"); + BUG(); } - info = omap_get_config(OMAP_TAG_CLOCK, struct omap_clock_config); - if (info != NULL) { - if (!cpu_is_omap1510()) - crystal_type = info->system_clock_type; - } - -#if defined(CONFIG_ARCH_OMAP730) - ck_ref.rate = 13000000; -#elif defined(CONFIG_ARCH_OMAP16XX) - if (crystal_type == 2) - ck_ref.rate = 19200000; -#endif - - printk("Clocks: ARM_SYSST: 0x%04x DPLL_CTL: 0x%04x ARM_CKCTL: 0x%04x\n", - omap_readw(ARM_SYSST), omap_readw(DPLL_CTL), - omap_readw(ARM_CKCTL)); - - /* We want to be in syncronous scalable mode */ - omap_writew(0x1000, ARM_SYSST); - -#ifdef CONFIG_OMAP_CLOCKS_SET_BY_BOOTLOADER - /* Use values set by bootloader. Determine PLL rate and recalculate - * dependent clocks as if kernel had changed PLL or divisors. - */ - { - unsigned pll_ctl_val = omap_readw(DPLL_CTL); - - ck_dpll1.rate = ck_ref.rate; /* Base xtal rate */ - if (pll_ctl_val & 0x10) { - /* PLL enabled, apply multiplier and divisor */ - if (pll_ctl_val & 0xf80) - ck_dpll1.rate *= (pll_ctl_val & 0xf80) >> 7; - ck_dpll1.rate /= ((pll_ctl_val & 0x60) >> 5) + 1; - } else { - /* PLL disabled, apply bypass divisor */ - switch (pll_ctl_val & 0xc) { - case 0: - break; - case 0x4: - ck_dpll1.rate /= 2; - break; - default: - ck_dpll1.rate /= 4; - break; - } - } - } - propagate_rate(&ck_dpll1); -#else - /* Find the highest supported frequency and enable it */ - if (select_table_rate(&virtual_ck_mpu, ~0)) { - printk(KERN_ERR "System frequencies not set. Check your config.\n"); - /* Guess sane values (60MHz) */ - omap_writew(0x2290, DPLL_CTL); - omap_writew(0x1005, ARM_CKCTL); - ck_dpll1.rate = 60000000; - propagate_rate(&ck_dpll1); - } -#endif - /* Cache rates for clocks connected to ck_ref (not dpll1) */ - propagate_rate(&ck_ref); - printk(KERN_INFO "Clocking rate (xtal/DPLL1/MPU): " - "%ld.%01ld/%ld.%01ld/%ld.%01ld MHz\n", - ck_ref.rate / 1000000, (ck_ref.rate / 100000) % 10, - ck_dpll1.rate / 1000000, (ck_dpll1.rate / 100000) % 10, - arm_ck.rate / 1000000, (arm_ck.rate / 100000) % 10); - -#ifdef CONFIG_MACH_OMAP_PERSEUS2 - /* Select slicer output as OMAP input clock */ - omap_writew(omap_readw(OMAP730_PCC_UPLD_CTRL) & ~0x1, OMAP730_PCC_UPLD_CTRL); -#endif - - /* Turn off DSP and ARM_TIMXO. Make sure ARM_INTHCK is not divided */ - omap_writew(omap_readw(ARM_CKCTL) & 0x0fff, ARM_CKCTL); - - /* Put DSP/MPUI into reset until needed */ - omap_writew(0, ARM_RSTCT1); - omap_writew(1, ARM_RSTCT2); - omap_writew(0x400, ARM_IDLECT1); - - /* - * According to OMAP5910 Erratum SYS_DMA_1, bit DMACK_REQ (bit 8) - * of the ARM_IDLECT2 register must be set to zero. The power-on - * default value of this bit is one. - */ - omap_writew(0x0000, ARM_IDLECT2); /* Turn LCD clock off also */ - - /* - * Only enable those clocks we will need, let the drivers - * enable other clocks as necessary - */ - clk_use(&armper_ck); - clk_use(&armxor_ck); - clk_use(&armtim_ck); - - if (cpu_is_omap1510()) - clk_enable(&arm_gpio_ck); + arch_clock = custom_clocks; return 0; } - - -#ifdef CONFIG_OMAP_RESET_CLOCKS - -static int __init omap_late_clk_reset(void) -{ - /* Turn off all unused clocks */ - struct clk *p; - __u32 regval32; - - /* USB_REQ_EN will be disabled later if necessary (usb_dc_ck) */ - regval32 = omap_readw(SOFT_REQ_REG) & (1 << 4); - omap_writew(regval32, SOFT_REQ_REG); - omap_writew(0, SOFT_REQ_REG2); - - list_for_each_entry(p, &clocks, node) { - if (p->usecount > 0 || (p->flags & ALWAYS_ENABLED) || - p->enable_reg == 0) - continue; - - /* Assume no DSP clocks have been activated by bootloader */ - if (p->flags & DSP_DOMAIN_CLOCK) - continue; - - /* Is the clock already disabled? */ - if (p->flags & ENABLE_REG_32BIT) { - if (p->flags & VIRTUAL_IO_ADDRESS) - regval32 = __raw_readl(p->enable_reg); - else - regval32 = omap_readl(p->enable_reg); - } else { - if (p->flags & VIRTUAL_IO_ADDRESS) - regval32 = __raw_readw(p->enable_reg); - else - regval32 = omap_readw(p->enable_reg); - } - - if ((regval32 & (1 << p->enable_bit)) == 0) - continue; - - /* FIXME: This clock seems to be necessary but no-one - * has asked for its activation. */ - if (p == &tc2_ck // FIX: pm.c (SRAM), CCP, Camera - || p == &ck_dpll1out // FIX: SoSSI, SSR - || p == &arm_gpio_ck // FIX: GPIO code for 1510 - ) { - printk(KERN_INFO "FIXME: Clock \"%s\" seems unused\n", - p->name); - continue; - } - - printk(KERN_INFO "Disabling unused clock \"%s\"... ", p->name); - __clk_disable(p); - printk(" done\n"); - } - - return 0; -} - -late_initcall(omap_late_clk_reset); - -#endif diff --git a/arch/arm/plat-omap/clock.h b/arch/arm/plat-omap/clock.h deleted file mode 100644 index a89e1e8c2519..000000000000 --- a/arch/arm/plat-omap/clock.h +++ /dev/null @@ -1,120 +0,0 @@ -/* - * linux/arch/arm/plat-omap/clock.h - * - * Copyright (C) 2004 Nokia corporation - * Written by Tuukka Tikkanen - * Based on clocks.h by Tony Lindgren, Gordon McNutt and RidgeRun, Inc - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ - -#ifndef __ARCH_ARM_OMAP_CLOCK_H -#define __ARCH_ARM_OMAP_CLOCK_H - -struct module; - -struct clk { - struct list_head node; - struct module *owner; - const char *name; - struct clk *parent; - unsigned long rate; - __s8 usecount; - __u16 flags; - __u32 enable_reg; - __u8 enable_bit; - __u8 rate_offset; - void (*recalc)(struct clk *); - int (*set_rate)(struct clk *, unsigned long); - long (*round_rate)(struct clk *, unsigned long); - void (*init)(struct clk *); -}; - - -struct mpu_rate { - unsigned long rate; - unsigned long xtal; - unsigned long pll_rate; - __u16 ckctl_val; - __u16 dpllctl_val; -}; - - -/* Clock flags */ -#define RATE_CKCTL 1 -#define RATE_FIXED 2 -#define RATE_PROPAGATES 4 -#define VIRTUAL_CLOCK 8 -#define ALWAYS_ENABLED 16 -#define ENABLE_REG_32BIT 32 -#define CLOCK_IN_OMAP16XX 64 -#define CLOCK_IN_OMAP1510 128 -#define CLOCK_IN_OMAP730 256 -#define DSP_DOMAIN_CLOCK 512 -#define VIRTUAL_IO_ADDRESS 1024 - -/* ARM_CKCTL bit shifts */ -#define CKCTL_PERDIV_OFFSET 0 -#define CKCTL_LCDDIV_OFFSET 2 -#define CKCTL_ARMDIV_OFFSET 4 -#define CKCTL_DSPDIV_OFFSET 6 -#define CKCTL_TCDIV_OFFSET 8 -#define CKCTL_DSPMMUDIV_OFFSET 10 -/*#define ARM_TIMXO 12*/ -#define EN_DSPCK 13 -/*#define ARM_INTHCK_SEL 14*/ /* Divide-by-2 for mpu inth_ck */ -/* DSP_CKCTL bit shifts */ -#define CKCTL_DSPPERDIV_OFFSET 0 - -/* ARM_IDLECT1 bit shifts */ -/*#define IDLWDT_ARM 0*/ -/*#define IDLXORP_ARM 1*/ -/*#define IDLPER_ARM 2*/ -/*#define IDLLCD_ARM 3*/ -/*#define IDLLB_ARM 4*/ -/*#define IDLHSAB_ARM 5*/ -/*#define IDLIF_ARM 6*/ -/*#define IDLDPLL_ARM 7*/ -/*#define IDLAPI_ARM 8*/ -/*#define IDLTIM_ARM 9*/ -/*#define SETARM_IDLE 11*/ - -/* ARM_IDLECT2 bit shifts */ -#define EN_WDTCK 0 -#define EN_XORPCK 1 -#define EN_PERCK 2 -#define EN_LCDCK 3 -#define EN_LBCK 4 /* Not on 1610/1710 */ -/*#define EN_HSABCK 5*/ -#define EN_APICK 6 -#define EN_TIMCK 7 -#define DMACK_REQ 8 -#define EN_GPIOCK 9 /* Not on 1610/1710 */ -/*#define EN_LBFREECK 10*/ -#define EN_CKOUT_ARM 11 - -/* ARM_IDLECT3 bit shifts */ -#define EN_OCPI_CK 0 -#define EN_TC1_CK 2 -#define EN_TC2_CK 4 - -/* DSP_IDLECT2 bit shifts (0,1,2 are same as for ARM_IDLECT2) */ -#define EN_DSPTIMCK 5 - -/* Various register defines for clock controls scattered around OMAP chip */ -#define USB_MCLK_EN_BIT 4 /* In ULPD_CLKC_CTRL */ -#define USB_HOST_HHC_UHOST_EN 9 /* In MOD_CONF_CTRL_0 */ -#define SWD_ULPD_PLL_CLK_REQ 1 /* In SWD_CLK_DIV_CTRL_SEL */ -#define COM_ULPD_PLL_CLK_REQ 1 /* In COM_CLK_DIV_CTRL_SEL */ -#define SWD_CLK_DIV_CTRL_SEL 0xfffe0874 -#define COM_CLK_DIV_CTRL_SEL 0xfffe0878 -#define SOFT_REQ_REG 0xfffe0834 -#define SOFT_REQ_REG2 0xfffe0880 - -int clk_register(struct clk *clk); -void clk_unregister(struct clk *clk); -int clk_init(void); - -#endif diff --git a/arch/arm/plat-omap/common.c b/arch/arm/plat-omap/common.c index 02bcc6c1cd1b..ccdb452630cf 100644 --- a/arch/arm/plat-omap/common.c +++ b/arch/arm/plat-omap/common.c @@ -31,7 +31,7 @@ #include #include -#include "clock.h" +#include #define NO_LENGTH_CHECK 0xffffffff @@ -117,19 +117,43 @@ EXPORT_SYMBOL(omap_get_var_config); static int __init omap_add_serial_console(void) { - const struct omap_serial_console_config *info; + const struct omap_serial_console_config *con_info; + const struct omap_uart_config *uart_info; + static char speed[11], *opt = NULL; + int line, i, uart_idx; - info = omap_get_config(OMAP_TAG_SERIAL_CONSOLE, - struct omap_serial_console_config); - if (info != NULL && info->console_uart) { - static char speed[11], *opt = NULL; + uart_info = omap_get_config(OMAP_TAG_UART, struct omap_uart_config); + con_info = omap_get_config(OMAP_TAG_SERIAL_CONSOLE, + struct omap_serial_console_config); + if (uart_info == NULL || con_info == NULL) + return 0; - if (info->console_speed) { - snprintf(speed, sizeof(speed), "%u", info->console_speed); - opt = speed; - } - return add_preferred_console("ttyS", info->console_uart - 1, opt); + if (con_info->console_uart == 0) + return 0; + + if (con_info->console_speed) { + snprintf(speed, sizeof(speed), "%u", con_info->console_speed); + opt = speed; } - return 0; + + uart_idx = con_info->console_uart - 1; + if (uart_idx >= OMAP_MAX_NR_PORTS) { + printk(KERN_INFO "Console: external UART#%d. " + "Not adding it as console this time.\n", + uart_idx + 1); + return 0; + } + if (!(uart_info->enabled_uarts & (1 << uart_idx))) { + printk(KERN_ERR "Console: Selected UART#%d is " + "not enabled for this platform\n", + uart_idx + 1); + return -1; + } + line = 0; + for (i = 0; i < uart_idx; i++) { + if (uart_info->enabled_uarts & (1 << i)) + line++; + } + return add_preferred_console("ttyS", line, opt); } console_initcall(omap_add_serial_console); diff --git a/arch/arm/plat-omap/devices.c b/arch/arm/plat-omap/devices.c new file mode 100644 index 000000000000..9dcce904b608 --- /dev/null +++ b/arch/arm/plat-omap/devices.c @@ -0,0 +1,381 @@ +/* + * linux/arch/arm/plat-omap/devices.c + * + * Common platform device setup/initialization for OMAP1 and OMAP2 + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +#include +#include +#include +#include + + +void omap_nop_release(struct device *dev) +{ + /* Nothing */ +} + +/*-------------------------------------------------------------------------*/ + +#if defined(CONFIG_I2C_OMAP) || defined(CONFIG_I2C_OMAP_MODULE) + +#define OMAP1_I2C_BASE 0xfffb3800 +#define OMAP2_I2C_BASE1 0x48070000 +#define OMAP_I2C_SIZE 0x3f +#define OMAP1_I2C_INT INT_I2C +#define OMAP2_I2C_INT1 56 + +static struct resource i2c_resources1[] = { + { + .start = 0, + .end = 0, + .flags = IORESOURCE_MEM, + }, + { + .start = 0, + .flags = IORESOURCE_IRQ, + }, +}; + +/* DMA not used; works around erratum writing to non-empty i2c fifo */ + +static struct platform_device omap_i2c_device1 = { + .name = "i2c_omap", + .id = 1, + .dev = { + .release = omap_nop_release, + }, + .num_resources = ARRAY_SIZE(i2c_resources1), + .resource = i2c_resources1, +}; + +/* See also arch/arm/mach-omap2/devices.c for second I2C on 24xx */ +static void omap_init_i2c(void) +{ + if (cpu_is_omap24xx()) { + i2c_resources1[0].start = OMAP2_I2C_BASE1; + i2c_resources1[0].end = OMAP2_I2C_BASE1 + OMAP_I2C_SIZE; + i2c_resources1[1].start = OMAP2_I2C_INT1; + } else { + i2c_resources1[0].start = OMAP1_I2C_BASE; + i2c_resources1[0].end = OMAP1_I2C_BASE + OMAP_I2C_SIZE; + i2c_resources1[1].start = OMAP1_I2C_INT; + } + + /* FIXME define and use a boot tag, in case of boards that + * either don't wire up I2C, or chips that mux it differently... + * it can include clocking and address info, maybe more. + */ + if (cpu_is_omap24xx()) { + omap_cfg_reg(M19_24XX_I2C1_SCL); + omap_cfg_reg(L15_24XX_I2C1_SDA); + } else { + omap_cfg_reg(I2C_SCL); + omap_cfg_reg(I2C_SDA); + } + + (void) platform_device_register(&omap_i2c_device1); +} + +#else +static inline void omap_init_i2c(void) {} +#endif + +/*-------------------------------------------------------------------------*/ + +#if defined(CONFIG_MMC_OMAP) || defined(CONFIG_MMC_OMAP_MODULE) + +#ifdef CONFIG_ARCH_OMAP24XX +#define OMAP_MMC1_BASE 0x4809c000 +#define OMAP_MMC1_INT 83 +#else +#define OMAP_MMC1_BASE 0xfffb7800 +#define OMAP_MMC1_INT INT_MMC +#endif +#define OMAP_MMC2_BASE 0xfffb7c00 /* omap16xx only */ + +static struct omap_mmc_conf mmc1_conf; + +static u64 mmc1_dmamask = 0xffffffff; + +static struct resource mmc1_resources[] = { + { + .start = IO_ADDRESS(OMAP_MMC1_BASE), + .end = IO_ADDRESS(OMAP_MMC1_BASE) + 0x7f, + .flags = IORESOURCE_MEM, + }, + { + .start = OMAP_MMC1_INT, + .flags = IORESOURCE_IRQ, + }, +}; + +static struct platform_device mmc_omap_device1 = { + .name = "mmci-omap", + .id = 1, + .dev = { + .release = omap_nop_release, + .dma_mask = &mmc1_dmamask, + .platform_data = &mmc1_conf, + }, + .num_resources = ARRAY_SIZE(mmc1_resources), + .resource = mmc1_resources, +}; + +#ifdef CONFIG_ARCH_OMAP16XX + +static struct omap_mmc_conf mmc2_conf; + +static u64 mmc2_dmamask = 0xffffffff; + +static struct resource mmc2_resources[] = { + { + .start = IO_ADDRESS(OMAP_MMC2_BASE), + .end = IO_ADDRESS(OMAP_MMC2_BASE) + 0x7f, + .flags = IORESOURCE_MEM, + }, + { + .start = INT_1610_MMC2, + .flags = IORESOURCE_IRQ, + }, +}; + +static struct platform_device mmc_omap_device2 = { + .name = "mmci-omap", + .id = 2, + .dev = { + .release = omap_nop_release, + .dma_mask = &mmc2_dmamask, + .platform_data = &mmc2_conf, + }, + .num_resources = ARRAY_SIZE(mmc2_resources), + .resource = mmc2_resources, +}; +#endif + +static void __init omap_init_mmc(void) +{ + const struct omap_mmc_config *mmc_conf; + const struct omap_mmc_conf *mmc; + + /* NOTE: assumes MMC was never (wrongly) enabled */ + mmc_conf = omap_get_config(OMAP_TAG_MMC, struct omap_mmc_config); + if (!mmc_conf) + return; + + /* block 1 is always available and has just one pinout option */ + mmc = &mmc_conf->mmc[0]; + if (mmc->enabled) { + if (!cpu_is_omap24xx()) { + omap_cfg_reg(MMC_CMD); + omap_cfg_reg(MMC_CLK); + omap_cfg_reg(MMC_DAT0); + if (cpu_is_omap1710()) { + omap_cfg_reg(M15_1710_MMC_CLKI); + omap_cfg_reg(P19_1710_MMC_CMDDIR); + omap_cfg_reg(P20_1710_MMC_DATDIR0); + } + } + if (mmc->wire4) { + if (!cpu_is_omap24xx()) { + omap_cfg_reg(MMC_DAT1); + /* NOTE: DAT2 can be on W10 (here) or M15 */ + if (!mmc->nomux) + omap_cfg_reg(MMC_DAT2); + omap_cfg_reg(MMC_DAT3); + } + } + mmc1_conf = *mmc; + (void) platform_device_register(&mmc_omap_device1); + } + +#ifdef CONFIG_ARCH_OMAP16XX + /* block 2 is on newer chips, and has many pinout options */ + mmc = &mmc_conf->mmc[1]; + if (mmc->enabled) { + if (!mmc->nomux) { + omap_cfg_reg(Y8_1610_MMC2_CMD); + omap_cfg_reg(Y10_1610_MMC2_CLK); + omap_cfg_reg(R18_1610_MMC2_CLKIN); + omap_cfg_reg(W8_1610_MMC2_DAT0); + if (mmc->wire4) { + omap_cfg_reg(V8_1610_MMC2_DAT1); + omap_cfg_reg(W15_1610_MMC2_DAT2); + omap_cfg_reg(R10_1610_MMC2_DAT3); + } + + /* These are needed for the level shifter */ + omap_cfg_reg(V9_1610_MMC2_CMDDIR); + omap_cfg_reg(V5_1610_MMC2_DATDIR0); + omap_cfg_reg(W19_1610_MMC2_DATDIR1); + } + + /* Feedback clock must be set on OMAP-1710 MMC2 */ + if (cpu_is_omap1710()) + omap_writel(omap_readl(MOD_CONF_CTRL_1) | (1 << 24), + MOD_CONF_CTRL_1); + mmc2_conf = *mmc; + (void) platform_device_register(&mmc_omap_device2); + } +#endif + return; +} +#else +static inline void omap_init_mmc(void) {} +#endif + +#if defined(CONFIG_OMAP_WATCHDOG) || defined(CONFIG_OMAP_WATCHDOG_MODULE) + +#ifdef CONFIG_ARCH_OMAP24XX +#define OMAP_WDT_BASE 0x48022000 +#else +#define OMAP_WDT_BASE 0xfffeb000 +#endif + +static struct resource wdt_resources[] = { + { + .start = OMAP_WDT_BASE, + .end = OMAP_WDT_BASE + 0x4f, + .flags = IORESOURCE_MEM, + }, +}; + +static struct platform_device omap_wdt_device = { + .name = "omap_wdt", + .id = -1, + .dev = { + .release = omap_nop_release, + }, + .num_resources = ARRAY_SIZE(wdt_resources), + .resource = wdt_resources, +}; + +static void omap_init_wdt(void) +{ + (void) platform_device_register(&omap_wdt_device); +} +#else +static inline void omap_init_wdt(void) {} +#endif + +/*-------------------------------------------------------------------------*/ + +#if defined(CONFIG_OMAP_RNG) || defined(CONFIG_OMAP_RNG_MODULE) + +#ifdef CONFIG_ARCH_OMAP24XX +#define OMAP_RNG_BASE 0x480A0000 +#else +#define OMAP_RNG_BASE 0xfffe5000 +#endif + +static struct resource rng_resources[] = { + { + .start = OMAP_RNG_BASE, + .end = OMAP_RNG_BASE + 0x4f, + .flags = IORESOURCE_MEM, + }, +}; + +static struct platform_device omap_rng_device = { + .name = "omap_rng", + .id = -1, + .dev = { + .release = omap_nop_release, + }, + .num_resources = ARRAY_SIZE(rng_resources), + .resource = rng_resources, +}; + +static void omap_init_rng(void) +{ + (void) platform_device_register(&omap_rng_device); +} +#else +static inline void omap_init_rng(void) {} +#endif + +#if defined(CONFIG_FB_OMAP) || defined(CONFIG_FB_OMAP_MODULE) + +static struct omap_lcd_config omap_fb_conf; + +static u64 omap_fb_dma_mask = ~(u32)0; + +static struct platform_device omap_fb_device = { + .name = "omapfb", + .id = -1, + .dev = { + .release = omap_nop_release, + .dma_mask = &omap_fb_dma_mask, + .coherent_dma_mask = ~(u32)0, + .platform_data = &omap_fb_conf, + }, + .num_resources = 0, +}; + +static inline void omap_init_fb(void) +{ + const struct omap_lcd_config *conf; + + conf = omap_get_config(OMAP_TAG_LCD, struct omap_lcd_config); + if (conf != NULL) + omap_fb_conf = *conf; + platform_device_register(&omap_fb_device); +} + +#else + +static inline void omap_init_fb(void) {} + +#endif + +/* + * This gets called after board-specific INIT_MACHINE, and initializes most + * on-chip peripherals accessible on this board (except for few like USB): + * + * (a) Does any "standard config" pin muxing needed. Board-specific + * code will have muxed GPIO pins and done "nonstandard" setup; + * that code could live in the boot loader. + * (b) Populating board-specific platform_data with the data drivers + * rely on to handle wiring variations. + * (c) Creating platform devices as meaningful on this board and + * with this kernel configuration. + * + * Claiming GPIOs, and setting their direction and initial values, is the + * responsibility of the device drivers. So is responding to probe(). + * + * Board-specific knowlege like creating devices or pin setup is to be + * kept out of drivers as much as possible. In particular, pin setup + * may be handled by the boot loader, and drivers should expect it will + * normally have been done by the time they're probed. + */ +static int __init omap_init_devices(void) +{ + /* please keep these calls, and their implementations above, + * in alphabetical order so they're easier to sort through. + */ + omap_init_fb(); + omap_init_i2c(); + omap_init_mmc(); + omap_init_wdt(); + omap_init_rng(); + + return 0; +} +arch_initcall(omap_init_devices); + diff --git a/arch/arm/plat-omap/dma.c b/arch/arm/plat-omap/dma.c index da7b65145658..f5cc21ad0956 100644 --- a/arch/arm/plat-omap/dma.c +++ b/arch/arm/plat-omap/dma.c @@ -6,6 +6,8 @@ * DMA channel linking for 1610 by Samuel Ortiz * Graphics DMA and LCD DMA graphics tranformations * by Imre Deak + * OMAP2 support Copyright (C) 2004-2005 Texas Instruments, Inc. + * Merged to support both OMAP1 and OMAP2 by Tony Lindgren * Some functions based on earlier dma-omap.c Copyright (C) 2001 RidgeRun, Inc. * * Support functions for the OMAP internal DMA channels. @@ -31,8 +33,15 @@ #include -#define OMAP_DMA_ACTIVE 0x01 +#define DEBUG_PRINTS +#undef DEBUG_PRINTS +#ifdef DEBUG_PRINTS +#define debug_printk(x) printk x +#else +#define debug_printk(x) +#endif +#define OMAP_DMA_ACTIVE 0x01 #define OMAP_DMA_CCR_EN (1 << 7) #define OMAP_FUNC_MUX_ARM_BASE (0xfffe1000 + 0xec) @@ -55,7 +64,7 @@ static int dma_chan_count; static spinlock_t dma_chan_lock; static struct omap_dma_lch dma_chan[OMAP_LOGICAL_DMA_CH_COUNT]; -const static u8 dma_irq[OMAP_LOGICAL_DMA_CH_COUNT] = { +const static u8 omap1_dma_irq[OMAP_LOGICAL_DMA_CH_COUNT] = { INT_DMA_CH0_6, INT_DMA_CH1_7, INT_DMA_CH2_8, INT_DMA_CH3, INT_DMA_CH4, INT_DMA_CH5, INT_1610_DMA_CH6, INT_1610_DMA_CH7, INT_1610_DMA_CH8, INT_1610_DMA_CH9, INT_1610_DMA_CH10, @@ -63,6 +72,20 @@ const static u8 dma_irq[OMAP_LOGICAL_DMA_CH_COUNT] = { INT_1610_DMA_CH14, INT_1610_DMA_CH15, INT_DMA_LCD }; +#define REVISIT_24XX() printk(KERN_ERR "FIXME: no %s on 24xx\n", \ + __FUNCTION__); + +#ifdef CONFIG_ARCH_OMAP15XX +/* Returns 1 if the DMA module is in OMAP1510-compatible mode, 0 otherwise */ +int omap_dma_in_1510_mode(void) +{ + return enable_1510_mode; +} +#else +#define omap_dma_in_1510_mode() 0 +#endif + +#ifdef CONFIG_ARCH_OMAP1 static inline int get_gdma_dev(int req) { u32 reg = OMAP_FUNC_MUX_ARM_BASE + ((req - 1) / 5) * 4; @@ -82,6 +105,9 @@ static inline void set_gdma_dev(int req, int dev) l |= (dev - 1) << shift; omap_writel(l, reg); } +#else +#define set_gdma_dev(req, dev) do {} while (0) +#endif static void clear_lch_regs(int lch) { @@ -121,38 +147,62 @@ void omap_set_dma_priority(int dst_port, int priority) } void omap_set_dma_transfer_params(int lch, int data_type, int elem_count, - int frame_count, int sync_mode) + int frame_count, int sync_mode, + int dma_trigger, int src_or_dst_synch) { - u16 w; + OMAP_DMA_CSDP_REG(lch) &= ~0x03; + OMAP_DMA_CSDP_REG(lch) |= data_type; - w = omap_readw(OMAP_DMA_CSDP(lch)); - w &= ~0x03; - w |= data_type; - omap_writew(w, OMAP_DMA_CSDP(lch)); + if (cpu_class_is_omap1()) { + OMAP_DMA_CCR_REG(lch) &= ~(1 << 5); + if (sync_mode == OMAP_DMA_SYNC_FRAME) + OMAP_DMA_CCR_REG(lch) |= 1 << 5; - w = omap_readw(OMAP_DMA_CCR(lch)); - w &= ~(1 << 5); - if (sync_mode == OMAP_DMA_SYNC_FRAME) - w |= 1 << 5; - omap_writew(w, OMAP_DMA_CCR(lch)); + OMAP1_DMA_CCR2_REG(lch) &= ~(1 << 2); + if (sync_mode == OMAP_DMA_SYNC_BLOCK) + OMAP1_DMA_CCR2_REG(lch) |= 1 << 2; + } - w = omap_readw(OMAP_DMA_CCR2(lch)); - w &= ~(1 << 2); - if (sync_mode == OMAP_DMA_SYNC_BLOCK) - w |= 1 << 2; - omap_writew(w, OMAP_DMA_CCR2(lch)); + if (cpu_is_omap24xx() && dma_trigger) { + u32 val = OMAP_DMA_CCR_REG(lch); - omap_writew(elem_count, OMAP_DMA_CEN(lch)); - omap_writew(frame_count, OMAP_DMA_CFN(lch)); + if (dma_trigger > 63) + val |= 1 << 20; + if (dma_trigger > 31) + val |= 1 << 19; + val |= (dma_trigger & 0x1f); + + if (sync_mode & OMAP_DMA_SYNC_FRAME) + val |= 1 << 5; + + if (sync_mode & OMAP_DMA_SYNC_BLOCK) + val |= 1 << 18; + + if (src_or_dst_synch) + val |= 1 << 24; /* source synch */ + else + val &= ~(1 << 24); /* dest synch */ + + OMAP_DMA_CCR_REG(lch) = val; + } + + OMAP_DMA_CEN_REG(lch) = elem_count; + OMAP_DMA_CFN_REG(lch) = frame_count; } + void omap_set_dma_color_mode(int lch, enum omap_dma_color_mode mode, u32 color) { u16 w; BUG_ON(omap_dma_in_1510_mode()); - w = omap_readw(OMAP_DMA_CCR2(lch)) & ~0x03; + if (cpu_is_omap24xx()) { + REVISIT_24XX(); + return; + } + + w = OMAP1_DMA_CCR2_REG(lch) & ~0x03; switch (mode) { case OMAP_DMA_CONSTANT_FILL: w |= 0x01; @@ -165,63 +215,84 @@ void omap_set_dma_color_mode(int lch, enum omap_dma_color_mode mode, u32 color) default: BUG(); } - omap_writew(w, OMAP_DMA_CCR2(lch)); + OMAP1_DMA_CCR2_REG(lch) = w; - w = omap_readw(OMAP_DMA_LCH_CTRL(lch)) & ~0x0f; + w = OMAP1_DMA_LCH_CTRL_REG(lch) & ~0x0f; /* Default is channel type 2D */ if (mode) { - omap_writew((u16)color, OMAP_DMA_COLOR_L(lch)); - omap_writew((u16)(color >> 16), OMAP_DMA_COLOR_U(lch)); + OMAP1_DMA_COLOR_L_REG(lch) = (u16)color; + OMAP1_DMA_COLOR_U_REG(lch) = (u16)(color >> 16); w |= 1; /* Channel type G */ } - omap_writew(w, OMAP_DMA_LCH_CTRL(lch)); + OMAP1_DMA_LCH_CTRL_REG(lch) = w; } - +/* Note that src_port is only for omap1 */ void omap_set_dma_src_params(int lch, int src_port, int src_amode, - unsigned long src_start) + unsigned long src_start, + int src_ei, int src_fi) { - u16 w; + if (cpu_class_is_omap1()) { + OMAP_DMA_CSDP_REG(lch) &= ~(0x1f << 2); + OMAP_DMA_CSDP_REG(lch) |= src_port << 2; + } - w = omap_readw(OMAP_DMA_CSDP(lch)); - w &= ~(0x1f << 2); - w |= src_port << 2; - omap_writew(w, OMAP_DMA_CSDP(lch)); + OMAP_DMA_CCR_REG(lch) &= ~(0x03 << 12); + OMAP_DMA_CCR_REG(lch) |= src_amode << 12; - w = omap_readw(OMAP_DMA_CCR(lch)); - w &= ~(0x03 << 12); - w |= src_amode << 12; - omap_writew(w, OMAP_DMA_CCR(lch)); + if (cpu_class_is_omap1()) { + OMAP1_DMA_CSSA_U_REG(lch) = src_start >> 16; + OMAP1_DMA_CSSA_L_REG(lch) = src_start; + } - omap_writew(src_start >> 16, OMAP_DMA_CSSA_U(lch)); - omap_writew(src_start, OMAP_DMA_CSSA_L(lch)); + if (cpu_is_omap24xx()) + OMAP2_DMA_CSSA_REG(lch) = src_start; + + OMAP_DMA_CSEI_REG(lch) = src_ei; + OMAP_DMA_CSFI_REG(lch) = src_fi; +} + +void omap_set_dma_params(int lch, struct omap_dma_channel_params * params) +{ + omap_set_dma_transfer_params(lch, params->data_type, + params->elem_count, params->frame_count, + params->sync_mode, params->trigger, + params->src_or_dst_synch); + omap_set_dma_src_params(lch, params->src_port, + params->src_amode, params->src_start, + params->src_ei, params->src_fi); + + omap_set_dma_dest_params(lch, params->dst_port, + params->dst_amode, params->dst_start, + params->dst_ei, params->dst_fi); } void omap_set_dma_src_index(int lch, int eidx, int fidx) { - omap_writew(eidx, OMAP_DMA_CSEI(lch)); - omap_writew(fidx, OMAP_DMA_CSFI(lch)); + if (cpu_is_omap24xx()) { + REVISIT_24XX(); + return; + } + OMAP_DMA_CSEI_REG(lch) = eidx; + OMAP_DMA_CSFI_REG(lch) = fidx; } void omap_set_dma_src_data_pack(int lch, int enable) { - u16 w; - - w = omap_readw(OMAP_DMA_CSDP(lch)) & ~(1 << 6); - w |= enable ? (1 << 6) : 0; - omap_writew(w, OMAP_DMA_CSDP(lch)); + OMAP_DMA_CSDP_REG(lch) &= ~(1 << 6); + if (enable) + OMAP_DMA_CSDP_REG(lch) |= (1 << 6); } void omap_set_dma_src_burst_mode(int lch, enum omap_dma_burst_mode burst_mode) { - u16 w; + OMAP_DMA_CSDP_REG(lch) &= ~(0x03 << 7); - w = omap_readw(OMAP_DMA_CSDP(lch)) & ~(0x03 << 7); switch (burst_mode) { case OMAP_DMA_DATA_BURST_DIS: break; case OMAP_DMA_DATA_BURST_4: - w |= (0x01 << 7); + OMAP_DMA_CSDP_REG(lch) |= (0x02 << 7); break; case OMAP_DMA_DATA_BURST_8: /* not supported by current hardware @@ -231,175 +302,87 @@ void omap_set_dma_src_burst_mode(int lch, enum omap_dma_burst_mode burst_mode) default: BUG(); } - omap_writew(w, OMAP_DMA_CSDP(lch)); } +/* Note that dest_port is only for OMAP1 */ void omap_set_dma_dest_params(int lch, int dest_port, int dest_amode, - unsigned long dest_start) + unsigned long dest_start, + int dst_ei, int dst_fi) { - u16 w; + if (cpu_class_is_omap1()) { + OMAP_DMA_CSDP_REG(lch) &= ~(0x1f << 9); + OMAP_DMA_CSDP_REG(lch) |= dest_port << 9; + } - w = omap_readw(OMAP_DMA_CSDP(lch)); - w &= ~(0x1f << 9); - w |= dest_port << 9; - omap_writew(w, OMAP_DMA_CSDP(lch)); + OMAP_DMA_CCR_REG(lch) &= ~(0x03 << 14); + OMAP_DMA_CCR_REG(lch) |= dest_amode << 14; - w = omap_readw(OMAP_DMA_CCR(lch)); - w &= ~(0x03 << 14); - w |= dest_amode << 14; - omap_writew(w, OMAP_DMA_CCR(lch)); + if (cpu_class_is_omap1()) { + OMAP1_DMA_CDSA_U_REG(lch) = dest_start >> 16; + OMAP1_DMA_CDSA_L_REG(lch) = dest_start; + } - omap_writew(dest_start >> 16, OMAP_DMA_CDSA_U(lch)); - omap_writew(dest_start, OMAP_DMA_CDSA_L(lch)); + if (cpu_is_omap24xx()) + OMAP2_DMA_CDSA_REG(lch) = dest_start; + + OMAP_DMA_CDEI_REG(lch) = dst_ei; + OMAP_DMA_CDFI_REG(lch) = dst_fi; } void omap_set_dma_dest_index(int lch, int eidx, int fidx) { - omap_writew(eidx, OMAP_DMA_CDEI(lch)); - omap_writew(fidx, OMAP_DMA_CDFI(lch)); + if (cpu_is_omap24xx()) { + REVISIT_24XX(); + return; + } + OMAP_DMA_CDEI_REG(lch) = eidx; + OMAP_DMA_CDFI_REG(lch) = fidx; } void omap_set_dma_dest_data_pack(int lch, int enable) { - u16 w; - - w = omap_readw(OMAP_DMA_CSDP(lch)) & ~(1 << 13); - w |= enable ? (1 << 13) : 0; - omap_writew(w, OMAP_DMA_CSDP(lch)); + OMAP_DMA_CSDP_REG(lch) &= ~(1 << 13); + if (enable) + OMAP_DMA_CSDP_REG(lch) |= 1 << 13; } void omap_set_dma_dest_burst_mode(int lch, enum omap_dma_burst_mode burst_mode) { - u16 w; + OMAP_DMA_CSDP_REG(lch) &= ~(0x03 << 14); - w = omap_readw(OMAP_DMA_CSDP(lch)) & ~(0x03 << 14); switch (burst_mode) { case OMAP_DMA_DATA_BURST_DIS: break; case OMAP_DMA_DATA_BURST_4: - w |= (0x01 << 14); + OMAP_DMA_CSDP_REG(lch) |= (0x02 << 14); break; case OMAP_DMA_DATA_BURST_8: - w |= (0x03 << 14); + OMAP_DMA_CSDP_REG(lch) |= (0x03 << 14); break; default: printk(KERN_ERR "Invalid DMA burst mode\n"); BUG(); return; } - omap_writew(w, OMAP_DMA_CSDP(lch)); } -static inline void init_intr(int lch) +static inline void omap_enable_channel_irq(int lch) { - u16 w; + u32 status; /* Read CSR to make sure it's cleared. */ - w = omap_readw(OMAP_DMA_CSR(lch)); + status = OMAP_DMA_CSR_REG(lch); + /* Enable some nice interrupts. */ - omap_writew(dma_chan[lch].enabled_irqs, OMAP_DMA_CICR(lch)); + OMAP_DMA_CICR_REG(lch) = dma_chan[lch].enabled_irqs; + dma_chan[lch].flags |= OMAP_DMA_ACTIVE; } -static inline void enable_lnk(int lch) +static void omap_disable_channel_irq(int lch) { - u16 w; - - /* Clear the STOP_LNK bits */ - w = omap_readw(OMAP_DMA_CLNK_CTRL(lch)); - w &= ~(1 << 14); - omap_writew(w, OMAP_DMA_CLNK_CTRL(lch)); - - /* And set the ENABLE_LNK bits */ - if (dma_chan[lch].next_lch != -1) - omap_writew(dma_chan[lch].next_lch | (1 << 15), - OMAP_DMA_CLNK_CTRL(lch)); -} - -static inline void disable_lnk(int lch) -{ - u16 w; - - /* Disable interrupts */ - omap_writew(0, OMAP_DMA_CICR(lch)); - - /* Set the STOP_LNK bit */ - w = omap_readw(OMAP_DMA_CLNK_CTRL(lch)); - w |= (1 << 14); - w = omap_writew(w, OMAP_DMA_CLNK_CTRL(lch)); - - dma_chan[lch].flags &= ~OMAP_DMA_ACTIVE; -} - -void omap_start_dma(int lch) -{ - u16 w; - - if (!omap_dma_in_1510_mode() && dma_chan[lch].next_lch != -1) { - int next_lch, cur_lch; - char dma_chan_link_map[OMAP_LOGICAL_DMA_CH_COUNT]; - - dma_chan_link_map[lch] = 1; - /* Set the link register of the first channel */ - enable_lnk(lch); - - memset(dma_chan_link_map, 0, sizeof(dma_chan_link_map)); - cur_lch = dma_chan[lch].next_lch; - do { - next_lch = dma_chan[cur_lch].next_lch; - - /* The loop case: we've been here already */ - if (dma_chan_link_map[cur_lch]) - break; - /* Mark the current channel */ - dma_chan_link_map[cur_lch] = 1; - - enable_lnk(cur_lch); - init_intr(cur_lch); - - cur_lch = next_lch; - } while (next_lch != -1); - } - - init_intr(lch); - - w = omap_readw(OMAP_DMA_CCR(lch)); - w |= OMAP_DMA_CCR_EN; - omap_writew(w, OMAP_DMA_CCR(lch)); - dma_chan[lch].flags |= OMAP_DMA_ACTIVE; -} - -void omap_stop_dma(int lch) -{ - u16 w; - - if (!omap_dma_in_1510_mode() && dma_chan[lch].next_lch != -1) { - int next_lch, cur_lch = lch; - char dma_chan_link_map[OMAP_LOGICAL_DMA_CH_COUNT]; - - memset(dma_chan_link_map, 0, sizeof(dma_chan_link_map)); - do { - /* The loop case: we've been here already */ - if (dma_chan_link_map[cur_lch]) - break; - /* Mark the current channel */ - dma_chan_link_map[cur_lch] = 1; - - disable_lnk(cur_lch); - - next_lch = dma_chan[cur_lch].next_lch; - cur_lch = next_lch; - } while (next_lch != -1); - - return; - } - /* Disable all interrupts on the channel */ - omap_writew(0, OMAP_DMA_CICR(lch)); - - w = omap_readw(OMAP_DMA_CCR(lch)); - w &= ~OMAP_DMA_CCR_EN; - omap_writew(w, OMAP_DMA_CCR(lch)); - dma_chan[lch].flags &= ~OMAP_DMA_ACTIVE; + if (cpu_is_omap24xx()) + OMAP_DMA_CICR_REG(lch) = 0; } void omap_enable_dma_irq(int lch, u16 bits) @@ -412,55 +395,45 @@ void omap_disable_dma_irq(int lch, u16 bits) dma_chan[lch].enabled_irqs &= ~bits; } -static int dma_handle_ch(int ch) +static inline void enable_lnk(int lch) { - u16 csr; + if (cpu_class_is_omap1()) + OMAP_DMA_CLNK_CTRL_REG(lch) &= ~(1 << 14); - if (enable_1510_mode && ch >= 6) { - csr = dma_chan[ch].saved_csr; - dma_chan[ch].saved_csr = 0; - } else - csr = omap_readw(OMAP_DMA_CSR(ch)); - if (enable_1510_mode && ch <= 2 && (csr >> 7) != 0) { - dma_chan[ch + 6].saved_csr = csr >> 7; - csr &= 0x7f; - } - if ((csr & 0x3f) == 0) - return 0; - if (unlikely(dma_chan[ch].dev_id == -1)) { - printk(KERN_WARNING "Spurious interrupt from DMA channel %d (CSR %04x)\n", - ch, csr); - return 0; - } - if (unlikely(csr & OMAP_DMA_TOUT_IRQ)) - printk(KERN_WARNING "DMA timeout with device %d\n", dma_chan[ch].dev_id); - if (unlikely(csr & OMAP_DMA_DROP_IRQ)) - printk(KERN_WARNING "DMA synchronization event drop occurred with device %d\n", - dma_chan[ch].dev_id); - if (likely(csr & OMAP_DMA_BLOCK_IRQ)) - dma_chan[ch].flags &= ~OMAP_DMA_ACTIVE; - if (likely(dma_chan[ch].callback != NULL)) - dma_chan[ch].callback(ch, csr, dma_chan[ch].data); - return 1; + /* Set the ENABLE_LNK bits */ + if (dma_chan[lch].next_lch != -1) + OMAP_DMA_CLNK_CTRL_REG(lch) = + dma_chan[lch].next_lch | (1 << 15); } -static irqreturn_t dma_irq_handler(int irq, void *dev_id, struct pt_regs *regs) +static inline void disable_lnk(int lch) { - int ch = ((int) dev_id) - 1; - int handled = 0; - - for (;;) { - int handled_now = 0; - - handled_now += dma_handle_ch(ch); - if (enable_1510_mode && dma_chan[ch + 6].saved_csr) - handled_now += dma_handle_ch(ch + 6); - if (!handled_now) - break; - handled += handled_now; + /* Disable interrupts */ + if (cpu_class_is_omap1()) { + OMAP_DMA_CICR_REG(lch) = 0; + /* Set the STOP_LNK bit */ + OMAP_DMA_CLNK_CTRL_REG(lch) |= 1 << 14; } - return handled ? IRQ_HANDLED : IRQ_NONE; + if (cpu_is_omap24xx()) { + omap_disable_channel_irq(lch); + /* Clear the ENABLE_LNK bit */ + OMAP_DMA_CLNK_CTRL_REG(lch) &= ~(1 << 15); + } + + dma_chan[lch].flags &= ~OMAP_DMA_ACTIVE; +} + +static inline void omap2_enable_irq_lch(int lch) +{ + u32 val; + + if (!cpu_is_omap24xx()) + return; + + val = omap_readl(OMAP_DMA4_IRQENABLE_L0); + val |= 1 << lch; + omap_writel(val, OMAP_DMA4_IRQENABLE_L0); } int omap_request_dma(int dev_id, const char *dev_name, @@ -485,14 +458,23 @@ int omap_request_dma(int dev_id, const char *dev_name, } chan = dma_chan + free_ch; chan->dev_id = dev_id; - clear_lch_regs(free_ch); + + if (cpu_class_is_omap1()) + clear_lch_regs(free_ch); + + if (cpu_is_omap24xx()) + omap_clear_dma(free_ch); + spin_unlock_irqrestore(&dma_chan_lock, flags); - chan->dev_id = dev_id; chan->dev_name = dev_name; chan->callback = callback; chan->data = data; - chan->enabled_irqs = OMAP_DMA_TOUT_IRQ | OMAP_DMA_DROP_IRQ | OMAP_DMA_BLOCK_IRQ; + chan->enabled_irqs = OMAP_DMA_TOUT_IRQ | OMAP_DMA_DROP_IRQ | + OMAP_DMA_BLOCK_IRQ; + + if (cpu_is_omap24xx()) + chan->enabled_irqs |= OMAP2_DMA_TRANS_ERR_IRQ; if (cpu_is_omap16xx()) { /* If the sync device is set, configure it dynamically. */ @@ -502,37 +484,242 @@ int omap_request_dma(int dev_id, const char *dev_name, } /* Disable the 1510 compatibility mode and set the sync device * id. */ - omap_writew(dev_id | (1 << 10), OMAP_DMA_CCR(free_ch)); - } else { - omap_writew(dev_id, OMAP_DMA_CCR(free_ch)); + OMAP_DMA_CCR_REG(free_ch) = dev_id | (1 << 10); + } else if (cpu_is_omap730() || cpu_is_omap15xx()) { + OMAP_DMA_CCR_REG(free_ch) = dev_id; } + + if (cpu_is_omap24xx()) { + omap2_enable_irq_lch(free_ch); + + omap_enable_channel_irq(free_ch); + /* Clear the CSR register and IRQ status register */ + OMAP_DMA_CSR_REG(free_ch) = 0x0; + omap_writel(~0x0, OMAP_DMA4_IRQSTATUS_L0); + } + *dma_ch_out = free_ch; return 0; } -void omap_free_dma(int ch) +void omap_free_dma(int lch) { unsigned long flags; spin_lock_irqsave(&dma_chan_lock, flags); - if (dma_chan[ch].dev_id == -1) { - printk("omap_dma: trying to free nonallocated DMA channel %d\n", ch); + if (dma_chan[lch].dev_id == -1) { + printk("omap_dma: trying to free nonallocated DMA channel %d\n", + lch); spin_unlock_irqrestore(&dma_chan_lock, flags); return; } - dma_chan[ch].dev_id = -1; + dma_chan[lch].dev_id = -1; + dma_chan[lch].next_lch = -1; + dma_chan[lch].callback = NULL; spin_unlock_irqrestore(&dma_chan_lock, flags); - /* Disable all DMA interrupts for the channel. */ - omap_writew(0, OMAP_DMA_CICR(ch)); - /* Make sure the DMA transfer is stopped. */ - omap_writew(0, OMAP_DMA_CCR(ch)); + if (cpu_class_is_omap1()) { + /* Disable all DMA interrupts for the channel. */ + OMAP_DMA_CICR_REG(lch) = 0; + /* Make sure the DMA transfer is stopped. */ + OMAP_DMA_CCR_REG(lch) = 0; + } + + if (cpu_is_omap24xx()) { + u32 val; + /* Disable interrupts */ + val = omap_readl(OMAP_DMA4_IRQENABLE_L0); + val &= ~(1 << lch); + omap_writel(val, OMAP_DMA4_IRQENABLE_L0); + + /* Clear the CSR register and IRQ status register */ + OMAP_DMA_CSR_REG(lch) = 0x0; + + val = omap_readl(OMAP_DMA4_IRQSTATUS_L0); + val |= 1 << lch; + omap_writel(val, OMAP_DMA4_IRQSTATUS_L0); + + /* Disable all DMA interrupts for the channel. */ + OMAP_DMA_CICR_REG(lch) = 0; + + /* Make sure the DMA transfer is stopped. */ + OMAP_DMA_CCR_REG(lch) = 0; + omap_clear_dma(lch); + } } -int omap_dma_in_1510_mode(void) +/* + * Clears any DMA state so the DMA engine is ready to restart with new buffers + * through omap_start_dma(). Any buffers in flight are discarded. + */ +void omap_clear_dma(int lch) { - return enable_1510_mode; + unsigned long flags; + + local_irq_save(flags); + + if (cpu_class_is_omap1()) { + int status; + OMAP_DMA_CCR_REG(lch) &= ~OMAP_DMA_CCR_EN; + + /* Clear pending interrupts */ + status = OMAP_DMA_CSR_REG(lch); + } + + if (cpu_is_omap24xx()) { + int i; + u32 lch_base = OMAP24XX_DMA_BASE + lch * 0x60 + 0x80; + for (i = 0; i < 0x44; i += 4) + omap_writel(0, lch_base + i); + } + + local_irq_restore(flags); +} + +void omap_start_dma(int lch) +{ + if (!omap_dma_in_1510_mode() && dma_chan[lch].next_lch != -1) { + int next_lch, cur_lch; + char dma_chan_link_map[OMAP_LOGICAL_DMA_CH_COUNT]; + + dma_chan_link_map[lch] = 1; + /* Set the link register of the first channel */ + enable_lnk(lch); + + memset(dma_chan_link_map, 0, sizeof(dma_chan_link_map)); + cur_lch = dma_chan[lch].next_lch; + do { + next_lch = dma_chan[cur_lch].next_lch; + + /* The loop case: we've been here already */ + if (dma_chan_link_map[cur_lch]) + break; + /* Mark the current channel */ + dma_chan_link_map[cur_lch] = 1; + + enable_lnk(cur_lch); + omap_enable_channel_irq(cur_lch); + + cur_lch = next_lch; + } while (next_lch != -1); + } else if (cpu_is_omap24xx()) { + /* Errata: Need to write lch even if not using chaining */ + OMAP_DMA_CLNK_CTRL_REG(lch) = lch; + } + + omap_enable_channel_irq(lch); + + /* Errata: On ES2.0 BUFFERING disable must be set. + * This will always fail on ES1.0 */ + if (cpu_is_omap24xx()) { + OMAP_DMA_CCR_REG(lch) |= OMAP_DMA_CCR_EN; + } + + OMAP_DMA_CCR_REG(lch) |= OMAP_DMA_CCR_EN; + + dma_chan[lch].flags |= OMAP_DMA_ACTIVE; +} + +void omap_stop_dma(int lch) +{ + if (!omap_dma_in_1510_mode() && dma_chan[lch].next_lch != -1) { + int next_lch, cur_lch = lch; + char dma_chan_link_map[OMAP_LOGICAL_DMA_CH_COUNT]; + + memset(dma_chan_link_map, 0, sizeof(dma_chan_link_map)); + do { + /* The loop case: we've been here already */ + if (dma_chan_link_map[cur_lch]) + break; + /* Mark the current channel */ + dma_chan_link_map[cur_lch] = 1; + + disable_lnk(cur_lch); + + next_lch = dma_chan[cur_lch].next_lch; + cur_lch = next_lch; + } while (next_lch != -1); + + return; + } + + /* Disable all interrupts on the channel */ + if (cpu_class_is_omap1()) + OMAP_DMA_CICR_REG(lch) = 0; + + OMAP_DMA_CCR_REG(lch) &= ~OMAP_DMA_CCR_EN; + dma_chan[lch].flags &= ~OMAP_DMA_ACTIVE; +} + +/* + * Returns current physical source address for the given DMA channel. + * If the channel is running the caller must disable interrupts prior calling + * this function and process the returned value before re-enabling interrupt to + * prevent races with the interrupt handler. Note that in continuous mode there + * is a chance for CSSA_L register overflow inbetween the two reads resulting + * in incorrect return value. + */ +dma_addr_t omap_get_dma_src_pos(int lch) +{ + dma_addr_t offset; + + if (cpu_class_is_omap1()) + offset = (dma_addr_t) (OMAP1_DMA_CSSA_L_REG(lch) | + (OMAP1_DMA_CSSA_U_REG(lch) << 16)); + + if (cpu_is_omap24xx()) + offset = OMAP_DMA_CSAC_REG(lch); + + return offset; +} + +/* + * Returns current physical destination address for the given DMA channel. + * If the channel is running the caller must disable interrupts prior calling + * this function and process the returned value before re-enabling interrupt to + * prevent races with the interrupt handler. Note that in continuous mode there + * is a chance for CDSA_L register overflow inbetween the two reads resulting + * in incorrect return value. + */ +dma_addr_t omap_get_dma_dst_pos(int lch) +{ + dma_addr_t offset; + + if (cpu_class_is_omap1()) + offset = (dma_addr_t) (OMAP1_DMA_CDSA_L_REG(lch) | + (OMAP1_DMA_CDSA_U_REG(lch) << 16)); + + if (cpu_is_omap24xx()) + offset = OMAP2_DMA_CDSA_REG(lch); + + return offset; +} + +/* + * Returns current source transfer counting for the given DMA channel. + * Can be used to monitor the progress of a transfer inside a block. + * It must be called with disabled interrupts. + */ +int omap_get_dma_src_addr_counter(int lch) +{ + return (dma_addr_t) OMAP_DMA_CSAC_REG(lch); +} + +int omap_dma_running(void) +{ + int lch; + + /* Check if LCD DMA is running */ + if (cpu_is_omap16xx()) + if (omap_readw(OMAP1610_DMA_LCD_CCR) & OMAP_DMA_CCR_EN) + return 1; + + for (lch = 0; lch < dma_chan_count; lch++) + if (OMAP_DMA_CCR_REG(lch) & OMAP_DMA_CCR_EN) + return 1; + + return 0; } /* @@ -550,7 +737,8 @@ void omap_dma_link_lch (int lch_head, int lch_queue) if ((dma_chan[lch_head].dev_id == -1) || (dma_chan[lch_queue].dev_id == -1)) { - printk(KERN_ERR "omap_dma: trying to link non requested channels\n"); + printk(KERN_ERR "omap_dma: trying to link " + "non requested channels\n"); dump_stack(); } @@ -570,20 +758,149 @@ void omap_dma_unlink_lch (int lch_head, int lch_queue) if (dma_chan[lch_head].next_lch != lch_queue || dma_chan[lch_head].next_lch == -1) { - printk(KERN_ERR "omap_dma: trying to unlink non linked channels\n"); + printk(KERN_ERR "omap_dma: trying to unlink " + "non linked channels\n"); dump_stack(); } if ((dma_chan[lch_head].flags & OMAP_DMA_ACTIVE) || (dma_chan[lch_head].flags & OMAP_DMA_ACTIVE)) { - printk(KERN_ERR "omap_dma: You need to stop the DMA channels before unlinking\n"); + printk(KERN_ERR "omap_dma: You need to stop the DMA channels " + "before unlinking\n"); dump_stack(); } dma_chan[lch_head].next_lch = -1; } +/*----------------------------------------------------------------------------*/ + +#ifdef CONFIG_ARCH_OMAP1 + +static int omap1_dma_handle_ch(int ch) +{ + u16 csr; + + if (enable_1510_mode && ch >= 6) { + csr = dma_chan[ch].saved_csr; + dma_chan[ch].saved_csr = 0; + } else + csr = OMAP_DMA_CSR_REG(ch); + if (enable_1510_mode && ch <= 2 && (csr >> 7) != 0) { + dma_chan[ch + 6].saved_csr = csr >> 7; + csr &= 0x7f; + } + if ((csr & 0x3f) == 0) + return 0; + if (unlikely(dma_chan[ch].dev_id == -1)) { + printk(KERN_WARNING "Spurious interrupt from DMA channel " + "%d (CSR %04x)\n", ch, csr); + return 0; + } + if (unlikely(csr & OMAP_DMA_TOUT_IRQ)) + printk(KERN_WARNING "DMA timeout with device %d\n", + dma_chan[ch].dev_id); + if (unlikely(csr & OMAP_DMA_DROP_IRQ)) + printk(KERN_WARNING "DMA synchronization event drop occurred " + "with device %d\n", dma_chan[ch].dev_id); + if (likely(csr & OMAP_DMA_BLOCK_IRQ)) + dma_chan[ch].flags &= ~OMAP_DMA_ACTIVE; + if (likely(dma_chan[ch].callback != NULL)) + dma_chan[ch].callback(ch, csr, dma_chan[ch].data); + return 1; +} + +static irqreturn_t omap1_dma_irq_handler(int irq, void *dev_id, + struct pt_regs *regs) +{ + int ch = ((int) dev_id) - 1; + int handled = 0; + + for (;;) { + int handled_now = 0; + + handled_now += omap1_dma_handle_ch(ch); + if (enable_1510_mode && dma_chan[ch + 6].saved_csr) + handled_now += omap1_dma_handle_ch(ch + 6); + if (!handled_now) + break; + handled += handled_now; + } + + return handled ? IRQ_HANDLED : IRQ_NONE; +} + +#else +#define omap1_dma_irq_handler NULL +#endif + +#ifdef CONFIG_ARCH_OMAP2 + +static int omap2_dma_handle_ch(int ch) +{ + u32 status = OMAP_DMA_CSR_REG(ch); + u32 val; + + if (!status) + return 0; + if (unlikely(dma_chan[ch].dev_id == -1)) + return 0; + /* REVISIT: According to 24xx TRM, there's no TOUT_IE */ + if (unlikely(status & OMAP_DMA_TOUT_IRQ)) + printk(KERN_INFO "DMA timeout with device %d\n", + dma_chan[ch].dev_id); + if (unlikely(status & OMAP_DMA_DROP_IRQ)) + printk(KERN_INFO + "DMA synchronization event drop occurred with device " + "%d\n", dma_chan[ch].dev_id); + + if (unlikely(status & OMAP2_DMA_TRANS_ERR_IRQ)) + printk(KERN_INFO "DMA transaction error with device %d\n", + dma_chan[ch].dev_id); + + OMAP_DMA_CSR_REG(ch) = 0x20; + + val = omap_readl(OMAP_DMA4_IRQSTATUS_L0); + /* ch in this function is from 0-31 while in register it is 1-32 */ + val = 1 << (ch); + omap_writel(val, OMAP_DMA4_IRQSTATUS_L0); + + if (likely(dma_chan[ch].callback != NULL)) + dma_chan[ch].callback(ch, status, dma_chan[ch].data); + + return 0; +} + +/* STATUS register count is from 1-32 while our is 0-31 */ +static irqreturn_t omap2_dma_irq_handler(int irq, void *dev_id, + struct pt_regs *regs) +{ + u32 val; + int i; + + val = omap_readl(OMAP_DMA4_IRQSTATUS_L0); + + for (i = 1; i <= OMAP_LOGICAL_DMA_CH_COUNT; i++) { + int active = val & (1 << (i - 1)); + if (active) + omap2_dma_handle_ch(i - 1); + } + + return IRQ_HANDLED; +} + +static struct irqaction omap24xx_dma_irq = { + .name = "DMA", + .handler = omap2_dma_irq_handler, + .flags = SA_INTERRUPT +}; + +#else +static struct irqaction omap24xx_dma_irq; +#endif + +/*----------------------------------------------------------------------------*/ static struct lcd_dma_info { spinlock_t lock; @@ -795,7 +1112,7 @@ static void set_b1_regs(void) /* Always set the source port as SDRAM for now*/ w &= ~(0x03 << 6); if (lcd_dma.callback != NULL) - w |= 1 << 1; /* Block interrupt enable */ + w |= 1 << 1; /* Block interrupt enable */ else w &= ~(1 << 1); omap_writew(w, OMAP1610_DMA_LCD_CTRL); @@ -814,7 +1131,8 @@ static void set_b1_regs(void) omap_writew(fi, OMAP1610_DMA_LCD_SRC_FI_B1_L); } -static irqreturn_t lcd_dma_irq_handler(int irq, void *dev_id, struct pt_regs *regs) +static irqreturn_t lcd_dma_irq_handler(int irq, void *dev_id, + struct pt_regs *regs) { u16 w; @@ -870,7 +1188,8 @@ void omap_free_lcd_dma(void) return; } if (!enable_1510_mode) - omap_writew(omap_readw(OMAP1610_DMA_LCD_CCR) & ~1, OMAP1610_DMA_LCD_CCR); + omap_writew(omap_readw(OMAP1610_DMA_LCD_CCR) & ~1, + OMAP1610_DMA_LCD_CCR); lcd_dma.reserved = 0; spin_unlock(&lcd_dma.lock); } @@ -939,93 +1258,24 @@ void omap_stop_lcd_dma(void) omap_writew(w, OMAP1610_DMA_LCD_CTRL); } -/* - * Clears any DMA state so the DMA engine is ready to restart with new buffers - * through omap_start_dma(). Any buffers in flight are discarded. - */ -void omap_clear_dma(int lch) -{ - unsigned long flags; - int status; - - local_irq_save(flags); - omap_writew(omap_readw(OMAP_DMA_CCR(lch)) & ~OMAP_DMA_CCR_EN, - OMAP_DMA_CCR(lch)); - status = OMAP_DMA_CSR(lch); /* clear pending interrupts */ - local_irq_restore(flags); -} - -/* - * Returns current physical source address for the given DMA channel. - * If the channel is running the caller must disable interrupts prior calling - * this function and process the returned value before re-enabling interrupt to - * prevent races with the interrupt handler. Note that in continuous mode there - * is a chance for CSSA_L register overflow inbetween the two reads resulting - * in incorrect return value. - */ -dma_addr_t omap_get_dma_src_pos(int lch) -{ - return (dma_addr_t) (omap_readw(OMAP_DMA_CSSA_L(lch)) | - (omap_readw(OMAP_DMA_CSSA_U(lch)) << 16)); -} - -/* - * Returns current physical destination address for the given DMA channel. - * If the channel is running the caller must disable interrupts prior calling - * this function and process the returned value before re-enabling interrupt to - * prevent races with the interrupt handler. Note that in continuous mode there - * is a chance for CDSA_L register overflow inbetween the two reads resulting - * in incorrect return value. - */ -dma_addr_t omap_get_dma_dst_pos(int lch) -{ - return (dma_addr_t) (omap_readw(OMAP_DMA_CDSA_L(lch)) | - (omap_readw(OMAP_DMA_CDSA_U(lch)) << 16)); -} - -/* - * Returns current source transfer counting for the given DMA channel. - * Can be used to monitor the progress of a transfer inside a block. - * It must be called with disabled interrupts. - */ -int omap_get_dma_src_addr_counter(int lch) -{ - return (dma_addr_t) omap_readw(OMAP_DMA_CSAC(lch)); -} - -int omap_dma_running(void) -{ - int lch; - - /* Check if LCD DMA is running */ - if (cpu_is_omap16xx()) - if (omap_readw(OMAP1610_DMA_LCD_CCR) & OMAP_DMA_CCR_EN) - return 1; - - for (lch = 0; lch < dma_chan_count; lch++) { - u16 w; - - w = omap_readw(OMAP_DMA_CCR(lch)); - if (w & OMAP_DMA_CCR_EN) - return 1; - } - return 0; -} +/*----------------------------------------------------------------------------*/ static int __init omap_init_dma(void) { int ch, r; - if (cpu_is_omap1510()) { - printk(KERN_INFO "DMA support for OMAP1510 initialized\n"); + if (cpu_is_omap15xx()) { + printk(KERN_INFO "DMA support for OMAP15xx initialized\n"); dma_chan_count = 9; enable_1510_mode = 1; } else if (cpu_is_omap16xx() || cpu_is_omap730()) { printk(KERN_INFO "OMAP DMA hardware version %d\n", omap_readw(OMAP_DMA_HW_ID)); printk(KERN_INFO "DMA capabilities: %08x:%08x:%04x:%04x:%04x\n", - (omap_readw(OMAP_DMA_CAPS_0_U) << 16) | omap_readw(OMAP_DMA_CAPS_0_L), - (omap_readw(OMAP_DMA_CAPS_1_U) << 16) | omap_readw(OMAP_DMA_CAPS_1_L), + (omap_readw(OMAP_DMA_CAPS_0_U) << 16) | + omap_readw(OMAP_DMA_CAPS_0_L), + (omap_readw(OMAP_DMA_CAPS_1_U) << 16) | + omap_readw(OMAP_DMA_CAPS_1_L), omap_readw(OMAP_DMA_CAPS_2), omap_readw(OMAP_DMA_CAPS_3), omap_readw(OMAP_DMA_CAPS_4)); if (!enable_1510_mode) { @@ -1038,6 +1288,11 @@ static int __init omap_init_dma(void) dma_chan_count = 16; } else dma_chan_count = 9; + } else if (cpu_is_omap24xx()) { + u8 revision = omap_readb(OMAP_DMA4_REVISION); + printk(KERN_INFO "OMAP DMA hardware revision %d.%d\n", + revision >> 4, revision & 0xf); + dma_chan_count = OMAP_LOGICAL_DMA_CH_COUNT; } else { dma_chan_count = 0; return 0; @@ -1049,41 +1304,56 @@ static int __init omap_init_dma(void) memset(&dma_chan, 0, sizeof(dma_chan)); for (ch = 0; ch < dma_chan_count; ch++) { + omap_clear_dma(ch); dma_chan[ch].dev_id = -1; dma_chan[ch].next_lch = -1; if (ch >= 6 && enable_1510_mode) continue; - /* request_irq() doesn't like dev_id (ie. ch) being zero, - * so we have to kludge around this. */ - r = request_irq(dma_irq[ch], dma_irq_handler, 0, "DMA", - (void *) (ch + 1)); + if (cpu_class_is_omap1()) { + /* request_irq() doesn't like dev_id (ie. ch) being + * zero, so we have to kludge around this. */ + r = request_irq(omap1_dma_irq[ch], + omap1_dma_irq_handler, 0, "DMA", + (void *) (ch + 1)); + if (r != 0) { + int i; + + printk(KERN_ERR "unable to request IRQ %d " + "for DMA (error %d)\n", + omap1_dma_irq[ch], r); + for (i = 0; i < ch; i++) + free_irq(omap1_dma_irq[i], + (void *) (i + 1)); + return r; + } + } + } + + if (cpu_is_omap24xx()) + setup_irq(INT_24XX_SDMA_IRQ0, &omap24xx_dma_irq); + + /* FIXME: Update LCD DMA to work on 24xx */ + if (cpu_class_is_omap1()) { + r = request_irq(INT_DMA_LCD, lcd_dma_irq_handler, 0, + "LCD DMA", NULL); if (r != 0) { int i; - printk(KERN_ERR "unable to request IRQ %d for DMA (error %d)\n", - dma_irq[ch], r); - for (i = 0; i < ch; i++) - free_irq(dma_irq[i], (void *) (i + 1)); + printk(KERN_ERR "unable to request IRQ for LCD DMA " + "(error %d)\n", r); + for (i = 0; i < dma_chan_count; i++) + free_irq(omap1_dma_irq[i], (void *) (i + 1)); return r; } } - r = request_irq(INT_DMA_LCD, lcd_dma_irq_handler, 0, "LCD DMA", NULL); - if (r != 0) { - int i; - printk(KERN_ERR "unable to request IRQ for LCD DMA (error %d)\n", r); - for (i = 0; i < dma_chan_count; i++) - free_irq(dma_irq[i], (void *) (i + 1)); - return r; - } return 0; } arch_initcall(omap_init_dma); - EXPORT_SYMBOL(omap_get_dma_src_pos); EXPORT_SYMBOL(omap_get_dma_dst_pos); EXPORT_SYMBOL(omap_get_dma_src_addr_counter); @@ -1109,6 +1379,8 @@ EXPORT_SYMBOL(omap_set_dma_dest_index); EXPORT_SYMBOL(omap_set_dma_dest_data_pack); EXPORT_SYMBOL(omap_set_dma_dest_burst_mode); +EXPORT_SYMBOL(omap_set_dma_params); + EXPORT_SYMBOL(omap_dma_link_lch); EXPORT_SYMBOL(omap_dma_unlink_lch); diff --git a/arch/arm/plat-omap/gpio.c b/arch/arm/plat-omap/gpio.c index 55059a24ad41..76f721d85137 100644 --- a/arch/arm/plat-omap/gpio.c +++ b/arch/arm/plat-omap/gpio.c @@ -140,7 +140,7 @@ static struct gpio_bank gpio_bank_1610[5] = { }; #endif -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX static struct gpio_bank gpio_bank_1510[2] = { { OMAP_MPUIO_BASE, INT_MPUIO, IH_MPUIO_BASE, METHOD_MPUIO }, { OMAP1510_GPIO_BASE, INT_GPIO_BANK1, IH_GPIO_BASE, METHOD_GPIO_1510 } @@ -173,7 +173,7 @@ static int gpio_bank_count; static inline struct gpio_bank *get_gpio_bank(int gpio) { -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { if (OMAP_GPIO_IS_MPUIO(gpio)) return &gpio_bank[0]; @@ -222,7 +222,7 @@ static inline int gpio_valid(int gpio) return -1; return 0; } -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510() && gpio < 16) return 0; #endif @@ -654,7 +654,7 @@ int omap_request_gpio(int gpio) /* Set trigger to none. You need to enable the trigger after request_irq */ _set_gpio_triggering(bank, get_gpio_index(gpio), IRQT_NOEDGE); -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (bank->method == METHOD_GPIO_1510) { void __iomem *reg; @@ -739,7 +739,7 @@ static void gpio_irq_handler(unsigned int irq, struct irqdesc *desc, bank = (struct gpio_bank *) desc->data; if (bank->method == METHOD_MPUIO) isr_reg = bank->base + OMAP_MPUIO_GPIO_INT; -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (bank->method == METHOD_GPIO_1510) isr_reg = bank->base + OMAP1510_GPIO_INT_STATUS; #endif @@ -774,7 +774,7 @@ static void gpio_irq_handler(unsigned int irq, struct irqdesc *desc, d = irq_desc + gpio_irq; desc_handle_irq(gpio_irq, d, regs); } - } + } } static void gpio_ack_irq(unsigned int irq) @@ -837,8 +837,9 @@ static struct irqchip mpuio_irq_chip = { .unmask = mpuio_unmask_irq }; -static int initialized = 0; -static struct clk * gpio_ck = NULL; +static int initialized; +static struct clk * gpio_ick; +static struct clk * gpio_fck; static int __init _omap_gpio_init(void) { @@ -848,14 +849,26 @@ static int __init _omap_gpio_init(void) initialized = 1; if (cpu_is_omap1510()) { - gpio_ck = clk_get(NULL, "arm_gpio_ck"); - if (IS_ERR(gpio_ck)) + gpio_ick = clk_get(NULL, "arm_gpio_ck"); + if (IS_ERR(gpio_ick)) printk("Could not get arm_gpio_ck\n"); else - clk_use(gpio_ck); + clk_use(gpio_ick); + } + if (cpu_is_omap24xx()) { + gpio_ick = clk_get(NULL, "gpios_ick"); + if (IS_ERR(gpio_ick)) + printk("Could not get gpios_ick\n"); + else + clk_use(gpio_ick); + gpio_fck = clk_get(NULL, "gpios_fck"); + if (IS_ERR(gpio_ick)) + printk("Could not get gpios_fck\n"); + else + clk_use(gpio_fck); } -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { printk(KERN_INFO "OMAP1510 GPIO hardware\n"); gpio_bank_count = 2; @@ -901,7 +914,7 @@ static int __init _omap_gpio_init(void) if (bank->method == METHOD_MPUIO) { omap_writew(0xFFFF, OMAP_MPUIO_BASE + OMAP_MPUIO_GPIO_MASKIT); } -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (bank->method == METHOD_GPIO_1510) { __raw_writew(0xffff, bank->base + OMAP1510_GPIO_INT_MASK); __raw_writew(0x0000, bank->base + OMAP1510_GPIO_INT_STATUS); @@ -1038,6 +1051,7 @@ static struct sys_device omap_gpio_device = { /* * This may get called early from board specific init + * for boards that have interrupts routed via FPGA. */ int omap_gpio_init(void) { diff --git a/arch/arm/plat-omap/mcbsp.c b/arch/arm/plat-omap/mcbsp.c index 9c9b7df3faf6..ea9475c86656 100644 --- a/arch/arm/plat-omap/mcbsp.c +++ b/arch/arm/plat-omap/mcbsp.c @@ -491,17 +491,20 @@ int omap_mcbsp_xmit_buffer(unsigned int id, dma_addr_t buffer, unsigned int leng omap_set_dma_transfer_params(mcbsp[id].dma_tx_lch, OMAP_DMA_DATA_TYPE_S16, length >> 1, 1, - OMAP_DMA_SYNC_ELEMENT); + OMAP_DMA_SYNC_ELEMENT, + 0, 0); omap_set_dma_dest_params(mcbsp[id].dma_tx_lch, OMAP_DMA_PORT_TIPB, OMAP_DMA_AMODE_CONSTANT, - mcbsp[id].io_base + OMAP_MCBSP_REG_DXR1); + mcbsp[id].io_base + OMAP_MCBSP_REG_DXR1, + 0, 0); omap_set_dma_src_params(mcbsp[id].dma_tx_lch, OMAP_DMA_PORT_EMIFF, OMAP_DMA_AMODE_POST_INC, - buffer); + buffer, + 0, 0); omap_start_dma(mcbsp[id].dma_tx_lch); wait_for_completion(&(mcbsp[id].tx_dma_completion)); @@ -531,17 +534,20 @@ int omap_mcbsp_recv_buffer(unsigned int id, dma_addr_t buffer, unsigned int leng omap_set_dma_transfer_params(mcbsp[id].dma_rx_lch, OMAP_DMA_DATA_TYPE_S16, length >> 1, 1, - OMAP_DMA_SYNC_ELEMENT); + OMAP_DMA_SYNC_ELEMENT, + 0, 0); omap_set_dma_src_params(mcbsp[id].dma_rx_lch, OMAP_DMA_PORT_TIPB, OMAP_DMA_AMODE_CONSTANT, - mcbsp[id].io_base + OMAP_MCBSP_REG_DRR1); + mcbsp[id].io_base + OMAP_MCBSP_REG_DRR1, + 0, 0); omap_set_dma_dest_params(mcbsp[id].dma_rx_lch, OMAP_DMA_PORT_EMIFF, OMAP_DMA_AMODE_POST_INC, - buffer); + buffer, + 0, 0); omap_start_dma(mcbsp[id].dma_rx_lch); wait_for_completion(&(mcbsp[id].rx_dma_completion)); @@ -643,7 +649,7 @@ static const struct omap_mcbsp_info mcbsp_730[] = { }; #endif -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX static const struct omap_mcbsp_info mcbsp_1510[] = { [0] = { .virt_base = OMAP1510_MCBSP1_BASE, .dma_rx_sync = OMAP_DMA_MCBSP1_RX, @@ -712,7 +718,7 @@ static int __init omap_mcbsp_init(void) mcbsp_count = ARRAY_SIZE(mcbsp_730); } #endif -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX if (cpu_is_omap1510()) { mcbsp_info = mcbsp_1510; mcbsp_count = ARRAY_SIZE(mcbsp_1510); diff --git a/arch/arm/plat-omap/mux.c b/arch/arm/plat-omap/mux.c index 64482040f89e..8c1c016aa689 100644 --- a/arch/arm/plat-omap/mux.c +++ b/arch/arm/plat-omap/mux.c @@ -3,7 +3,7 @@ * * Utility to set the Omap MUX and PULL_DWN registers from a table in mux.h * - * Copyright (C) 2003 Nokia Corporation + * Copyright (C) 2003 - 2005 Nokia Corporation * * Written by Tony Lindgren * @@ -25,38 +25,74 @@ #include #include #include +#include #include #include #include - -#define __MUX_C__ #include #ifdef CONFIG_OMAP_MUX +#define OMAP24XX_L4_BASE 0x48000000 +#define OMAP24XX_PULL_ENA (1 << 3) +#define OMAP24XX_PULL_UP (1 << 4) + +static struct pin_config * pin_table; +static unsigned long pin_table_sz; + +extern struct pin_config * omap730_pins; +extern struct pin_config * omap1xxx_pins; +extern struct pin_config * omap24xx_pins; + +int __init omap_mux_register(struct pin_config * pins, unsigned long size) +{ + pin_table = pins; + pin_table_sz = size; + + return 0; +} + /* * Sets the Omap MUX and PULL_DWN registers based on the table */ -int __init_or_module -omap_cfg_reg(const reg_cfg_t reg_cfg) +int __init_or_module omap_cfg_reg(const unsigned long index) { static DEFINE_SPINLOCK(mux_spin_lock); unsigned long flags; - reg_cfg_set *cfg; + struct pin_config *cfg; unsigned int reg_orig = 0, reg = 0, pu_pd_orig = 0, pu_pd = 0, pull_orig = 0, pull = 0; unsigned int mask, warn = 0; - if (cpu_is_omap7xx()) - return 0; + if (!pin_table) + BUG(); - if (reg_cfg > ARRAY_SIZE(reg_cfg_table)) { - printk(KERN_ERR "MUX: reg_cfg %d\n", reg_cfg); - return -EINVAL; + if (index >= pin_table_sz) { + printk(KERN_ERR "Invalid pin mux index: %lu (%lu)\n", + index, pin_table_sz); + dump_stack(); + return -ENODEV; } - cfg = (reg_cfg_set *)®_cfg_table[reg_cfg]; + cfg = (struct pin_config *)&pin_table[index]; + if (cpu_is_omap24xx()) { + u8 reg = 0; + + reg |= cfg->mask & 0x7; + if (cfg->pull_val) + reg |= OMAP24XX_PULL_ENA; + if(cfg->pu_pd_val) + reg |= OMAP24XX_PULL_UP; +#ifdef CONFIG_OMAP_MUX_DEBUG + printk("Muxing %s (0x%08x): 0x%02x -> 0x%02x\n", + cfg->name, OMAP24XX_L4_BASE + cfg->mux_reg, + omap_readb(OMAP24XX_L4_BASE + cfg->mux_reg), reg); +#endif + omap_writeb(reg, OMAP24XX_L4_BASE + cfg->mux_reg); + + return 0; + } /* Check the mux register in question */ if (cfg->mux_reg) { @@ -157,7 +193,8 @@ omap_cfg_reg(const reg_cfg_t reg_cfg) return 0; #endif } - EXPORT_SYMBOL(omap_cfg_reg); - +#else +#define omap_mux_init() do {} while(0) +#define omap_cfg_reg(x) do {} while(0) #endif /* CONFIG_OMAP_MUX */ diff --git a/arch/arm/plat-omap/pm.c b/arch/arm/plat-omap/pm.c index e15c6c1ddec9..966cca031ca7 100644 --- a/arch/arm/plat-omap/pm.c +++ b/arch/arm/plat-omap/pm.c @@ -54,11 +54,12 @@ #include #include -#include "clock.h" -#include "sram.h" +#include +#include static unsigned int arm_sleep_save[ARM_SLEEP_SAVE_SIZE]; static unsigned short ulpd_sleep_save[ULPD_SLEEP_SAVE_SIZE]; +static unsigned int mpui730_sleep_save[MPUI730_SLEEP_SAVE_SIZE]; static unsigned int mpui1510_sleep_save[MPUI1510_SLEEP_SAVE_SIZE]; static unsigned int mpui1610_sleep_save[MPUI1610_SLEEP_SAVE_SIZE]; @@ -120,8 +121,8 @@ void omap_pm_idle(void) */ static void omap_pm_wakeup_setup(void) { - u32 level1_wake = OMAP_IRQ_BIT(INT_IH2_IRQ); - u32 level2_wake = OMAP_IRQ_BIT(INT_UART2) | OMAP_IRQ_BIT(INT_KEYBOARD); + u32 level1_wake = 0; + u32 level2_wake = OMAP_IRQ_BIT(INT_UART2); /* * Turn off all interrupts except GPIO bank 1, L1-2nd level cascade, @@ -129,19 +130,29 @@ static void omap_pm_wakeup_setup(void) * drivers must still separately call omap_set_gpio_wakeup() to * wake up to a GPIO interrupt. */ - if (cpu_is_omap1510() || cpu_is_omap16xx()) - level1_wake |= OMAP_IRQ_BIT(INT_GPIO_BANK1); - else if (cpu_is_omap730()) - level1_wake |= OMAP_IRQ_BIT(INT_730_GPIO_BANK1); + if (cpu_is_omap730()) + level1_wake = OMAP_IRQ_BIT(INT_730_GPIO_BANK1) | + OMAP_IRQ_BIT(INT_730_IH2_IRQ); + else if (cpu_is_omap1510()) + level1_wake = OMAP_IRQ_BIT(INT_GPIO_BANK1) | + OMAP_IRQ_BIT(INT_1510_IH2_IRQ); + else if (cpu_is_omap16xx()) + level1_wake = OMAP_IRQ_BIT(INT_GPIO_BANK1) | + OMAP_IRQ_BIT(INT_1610_IH2_IRQ); omap_writel(~level1_wake, OMAP_IH1_MIR); - if (cpu_is_omap1510()) - omap_writel(~level2_wake, OMAP_IH2_MIR); - - /* INT_1610_WAKE_UP_REQ is needed for GPIO wakeup... */ - if (cpu_is_omap16xx()) { + if (cpu_is_omap730()) { omap_writel(~level2_wake, OMAP_IH2_0_MIR); + omap_writel(~(OMAP_IRQ_BIT(INT_730_WAKE_UP_REQ) | OMAP_IRQ_BIT(INT_730_MPUIO_KEYPAD)), OMAP_IH2_1_MIR); + } else if (cpu_is_omap1510()) { + level2_wake |= OMAP_IRQ_BIT(INT_KEYBOARD); + omap_writel(~level2_wake, OMAP_IH2_MIR); + } else if (cpu_is_omap16xx()) { + level2_wake |= OMAP_IRQ_BIT(INT_KEYBOARD); + omap_writel(~level2_wake, OMAP_IH2_0_MIR); + + /* INT_1610_WAKE_UP_REQ is needed for GPIO wakeup... */ omap_writel(~OMAP_IRQ_BIT(INT_1610_WAKE_UP_REQ), OMAP_IH2_1_MIR); omap_writel(~0x0, OMAP_IH2_2_MIR); omap_writel(~0x0, OMAP_IH2_3_MIR); @@ -185,7 +196,17 @@ void omap_pm_suspend(void) * Save interrupt, MPUI, ARM and UPLD control registers. */ - if (cpu_is_omap1510()) { + if (cpu_is_omap730()) { + MPUI730_SAVE(OMAP_IH1_MIR); + MPUI730_SAVE(OMAP_IH2_0_MIR); + MPUI730_SAVE(OMAP_IH2_1_MIR); + MPUI730_SAVE(MPUI_CTRL); + MPUI730_SAVE(MPUI_DSP_BOOT_CONFIG); + MPUI730_SAVE(MPUI_DSP_API_CONFIG); + MPUI730_SAVE(EMIFS_CONFIG); + MPUI730_SAVE(EMIFF_SDRAM_CONFIG); + + } else if (cpu_is_omap1510()) { MPUI1510_SAVE(OMAP_IH1_MIR); MPUI1510_SAVE(OMAP_IH2_MIR); MPUI1510_SAVE(MPUI_CTRL); @@ -280,7 +301,13 @@ void omap_pm_suspend(void) ULPD_RESTORE(ULPD_CLOCK_CTRL); ULPD_RESTORE(ULPD_STATUS_REQ); - if (cpu_is_omap1510()) { + if (cpu_is_omap730()) { + MPUI730_RESTORE(EMIFS_CONFIG); + MPUI730_RESTORE(EMIFF_SDRAM_CONFIG); + MPUI730_RESTORE(OMAP_IH1_MIR); + MPUI730_RESTORE(OMAP_IH2_0_MIR); + MPUI730_RESTORE(OMAP_IH2_1_MIR); + } else if (cpu_is_omap1510()) { MPUI1510_RESTORE(MPUI_CTRL); MPUI1510_RESTORE(MPUI_DSP_BOOT_CONFIG); MPUI1510_RESTORE(MPUI_DSP_API_CONFIG); @@ -355,7 +382,14 @@ static int omap_pm_read_proc( ULPD_SAVE(ULPD_DPLL_CTRL); ULPD_SAVE(ULPD_POWER_CTRL); - if (cpu_is_omap1510()) { + if (cpu_is_omap730()) { + MPUI730_SAVE(MPUI_CTRL); + MPUI730_SAVE(MPUI_DSP_STATUS); + MPUI730_SAVE(MPUI_DSP_BOOT_CONFIG); + MPUI730_SAVE(MPUI_DSP_API_CONFIG); + MPUI730_SAVE(EMIFF_SDRAM_CONFIG); + MPUI730_SAVE(EMIFS_CONFIG); + } else if (cpu_is_omap1510()) { MPUI1510_SAVE(MPUI_CTRL); MPUI1510_SAVE(MPUI_DSP_STATUS); MPUI1510_SAVE(MPUI_DSP_BOOT_CONFIG); @@ -404,7 +438,21 @@ static int omap_pm_read_proc( ULPD_SHOW(ULPD_STATUS_REQ), ULPD_SHOW(ULPD_POWER_CTRL)); - if (cpu_is_omap1510()) { + if (cpu_is_omap730()) { + my_buffer_offset += sprintf(my_base + my_buffer_offset, + "MPUI730_CTRL_REG 0x%-8x \n" + "MPUI730_DSP_STATUS_REG: 0x%-8x \n" + "MPUI730_DSP_BOOT_CONFIG_REG: 0x%-8x \n" + "MPUI730_DSP_API_CONFIG_REG: 0x%-8x \n" + "MPUI730_SDRAM_CONFIG_REG: 0x%-8x \n" + "MPUI730_EMIFS_CONFIG_REG: 0x%-8x \n", + MPUI730_SHOW(MPUI_CTRL), + MPUI730_SHOW(MPUI_DSP_STATUS), + MPUI730_SHOW(MPUI_DSP_BOOT_CONFIG), + MPUI730_SHOW(MPUI_DSP_API_CONFIG), + MPUI730_SHOW(EMIFF_SDRAM_CONFIG), + MPUI730_SHOW(EMIFS_CONFIG)); + } else if (cpu_is_omap1510()) { my_buffer_offset += sprintf(my_base + my_buffer_offset, "MPUI1510_CTRL_REG 0x%-8x \n" "MPUI1510_DSP_STATUS_REG: 0x%-8x \n" @@ -553,7 +601,12 @@ static int __init omap_pm_init(void) * These routines need to be in SRAM as that's the only * memory the MPU can see when it wakes up. */ - if (cpu_is_omap1510()) { + if (cpu_is_omap730()) { + omap_sram_idle = omap_sram_push(omap730_idle_loop_suspend, + omap730_idle_loop_suspend_sz); + omap_sram_suspend = omap_sram_push(omap730_cpu_suspend, + omap730_cpu_suspend_sz); + } else if (cpu_is_omap1510()) { omap_sram_idle = omap_sram_push(omap1510_idle_loop_suspend, omap1510_idle_loop_suspend_sz); omap_sram_suspend = omap_sram_push(omap1510_cpu_suspend, @@ -572,7 +625,11 @@ static int __init omap_pm_init(void) pm_idle = omap_pm_idle; - setup_irq(INT_1610_WAKE_UP_REQ, &omap_wakeup_irq); + if (cpu_is_omap730()) + setup_irq(INT_730_WAKE_UP_REQ, &omap_wakeup_irq); + else if (cpu_is_omap16xx()) + setup_irq(INT_1610_WAKE_UP_REQ, &omap_wakeup_irq); + #if 0 /* --- BEGIN BOARD-DEPENDENT CODE --- */ /* Sleepx mask direction */ @@ -591,7 +648,9 @@ static int __init omap_pm_init(void) omap_writew(ULPD_POWER_CTRL_REG_VAL, ULPD_POWER_CTRL); /* Configure IDLECT3 */ - if (cpu_is_omap16xx()) + if (cpu_is_omap730()) + omap_writel(OMAP730_IDLECT3_VAL, OMAP730_IDLECT3); + else if (cpu_is_omap16xx()) omap_writel(OMAP1610_IDLECT3_VAL, OMAP1610_IDLECT3); pm_set_ops(&omap_pm_ops); @@ -600,8 +659,10 @@ static int __init omap_pm_init(void) omap_pm_init_proc(); #endif - /* configure LOW_PWR pin */ - omap_cfg_reg(T20_1610_LOW_PWR); + if (cpu_is_omap16xx()) { + /* configure LOW_PWR pin */ + omap_cfg_reg(T20_1610_LOW_PWR); + } return 0; } diff --git a/arch/arm/plat-omap/sleep.S b/arch/arm/plat-omap/sleep.S index 9f745836f6aa..4cd7d292f854 100644 --- a/arch/arm/plat-omap/sleep.S +++ b/arch/arm/plat-omap/sleep.S @@ -1,7 +1,7 @@ /* * linux/arch/arm/plat-omap/sleep.S * - * Low-level OMAP1510/1610 sleep/wakeUp support + * Low-level OMAP730/1510/1610 sleep/wakeUp support * * Initial SA1110 code: * Copyright (c) 2001 Cliff Brake @@ -52,7 +52,57 @@ * processor specific functions here. */ -#ifdef CONFIG_ARCH_OMAP1510 +#if defined(CONFIG_ARCH_OMAP730) +ENTRY(omap730_idle_loop_suspend) + + stmfd sp!, {r0 - r12, lr} @ save registers on stack + + @ load base address of ARM_IDLECT1 and ARM_IDLECT2 + mov r4, #CLKGEN_REG_ASM_BASE & 0xff000000 + orr r4, r4, #CLKGEN_REG_ASM_BASE & 0x00ff0000 + orr r4, r4, #CLKGEN_REG_ASM_BASE & 0x0000ff00 + + @ turn off clock domains + @ get ARM_IDLECT2 into r2 + ldrh r2, [r4, #ARM_IDLECT2_ASM_OFFSET & 0xff] + mov r5, #OMAP730_IDLECT2_SLEEP_VAL & 0xff + orr r5, r5, #OMAP730_IDLECT2_SLEEP_VAL & 0xff00 + strh r5, [r4, #ARM_IDLECT2_ASM_OFFSET & 0xff] + + @ request ARM idle + @ get ARM_IDLECT1 into r1 + ldrh r1, [r4, #ARM_IDLECT1_ASM_OFFSET & 0xff] + orr r3, r1, #OMAP730_IDLE_LOOP_REQUEST & 0xffff + strh r3, [r4, #ARM_IDLECT1_ASM_OFFSET & 0xff] + + mov r5, #IDLE_WAIT_CYCLES & 0xff + orr r5, r5, #IDLE_WAIT_CYCLES & 0xff00 +l_730: subs r5, r5, #1 + bne l_730 +/* + * Let's wait for the next clock tick to wake us up. + */ + mov r0, #0 + mcr p15, 0, r0, c7, c0, 4 @ wait for interrupt +/* + * omap730_idle_loop_suspend()'s resume point. + * + * It will just start executing here, so we'll restore stuff from the + * stack, reset the ARM_IDLECT1 and ARM_IDLECT2. + */ + + @ restore ARM_IDLECT1 and ARM_IDLECT2 and return + @ r1 has ARM_IDLECT1 and r2 still has ARM_IDLECT2 + strh r2, [r4, #ARM_IDLECT2_ASM_OFFSET & 0xff] + strh r1, [r4, #ARM_IDLECT1_ASM_OFFSET & 0xff] + + ldmfd sp!, {r0 - r12, pc} @ restore regs and return + +ENTRY(omap730_idle_loop_suspend_sz) + .word . - omap730_idle_loop_suspend +#endif /* CONFIG_ARCH_OMAP730 */ + +#ifdef CONFIG_ARCH_OMAP15XX ENTRY(omap1510_idle_loop_suspend) stmfd sp!, {r0 - r12, lr} @ save registers on stack @@ -100,7 +150,7 @@ l_1510: subs r5, r5, #1 ENTRY(omap1510_idle_loop_suspend_sz) .word . - omap1510_idle_loop_suspend -#endif /* CONFIG_ARCH_OMAP1510 */ +#endif /* CONFIG_ARCH_OMAP15XX */ #if defined(CONFIG_ARCH_OMAP16XX) ENTRY(omap1610_idle_loop_suspend) @@ -169,7 +219,86 @@ ENTRY(omap1610_idle_loop_suspend_sz) * */ -#ifdef CONFIG_ARCH_OMAP1510 +#if defined(CONFIG_ARCH_OMAP730) +ENTRY(omap730_cpu_suspend) + + @ save registers on stack + stmfd sp!, {r0 - r12, lr} + + @ Drain write cache + mov r4, #0 + mcr p15, 0, r0, c7, c10, 4 + nop + + @ load base address of Traffic Controller + mov r6, #TCMIF_ASM_BASE & 0xff000000 + orr r6, r6, #TCMIF_ASM_BASE & 0x00ff0000 + orr r6, r6, #TCMIF_ASM_BASE & 0x0000ff00 + + @ prepare to put SDRAM into self-refresh manually + ldr r7, [r6, #EMIFF_SDRAM_CONFIG_ASM_OFFSET & 0xff] + orr r9, r7, #SELF_REFRESH_MODE & 0xff000000 + orr r9, r9, #SELF_REFRESH_MODE & 0x000000ff + str r9, [r6, #EMIFF_SDRAM_CONFIG_ASM_OFFSET & 0xff] + + @ prepare to put EMIFS to Sleep + ldr r8, [r6, #EMIFS_CONFIG_ASM_OFFSET & 0xff] + orr r9, r8, #IDLE_EMIFS_REQUEST & 0xff + str r9, [r6, #EMIFS_CONFIG_ASM_OFFSET & 0xff] + + @ load base address of ARM_IDLECT1 and ARM_IDLECT2 + mov r4, #CLKGEN_REG_ASM_BASE & 0xff000000 + orr r4, r4, #CLKGEN_REG_ASM_BASE & 0x00ff0000 + orr r4, r4, #CLKGEN_REG_ASM_BASE & 0x0000ff00 + + @ turn off clock domains + @ do not disable PERCK (0x04) + mov r5, #OMAP730_IDLECT2_SLEEP_VAL & 0xff + orr r5, r5, #OMAP730_IDLECT2_SLEEP_VAL & 0xff00 + strh r5, [r4, #ARM_IDLECT2_ASM_OFFSET & 0xff] + + @ request ARM idle + mov r3, #OMAP730_IDLECT1_SLEEP_VAL & 0xff + orr r3, r3, #OMAP730_IDLECT1_SLEEP_VAL & 0xff00 + strh r3, [r4, #ARM_IDLECT1_ASM_OFFSET & 0xff] + + @ disable instruction cache + mrc p15, 0, r9, c1, c0, 0 + bic r2, r9, #0x1000 + mcr p15, 0, r2, c1, c0, 0 + nop + +/* + * Let's wait for the next wake up event to wake us up. r0 can't be + * used here because r0 holds ARM_IDLECT1 + */ + mov r2, #0 + mcr p15, 0, r2, c7, c0, 4 @ wait for interrupt +/* + * omap730_cpu_suspend()'s resume point. + * + * It will just start executing here, so we'll restore stuff from the + * stack. + */ + @ re-enable Icache + mcr p15, 0, r9, c1, c0, 0 + + @ reset the ARM_IDLECT1 and ARM_IDLECT2. + strh r1, [r4, #ARM_IDLECT2_ASM_OFFSET & 0xff] + strh r0, [r4, #ARM_IDLECT1_ASM_OFFSET & 0xff] + + @ Restore EMIFF controls + str r7, [r6, #EMIFF_SDRAM_CONFIG_ASM_OFFSET & 0xff] + str r8, [r6, #EMIFS_CONFIG_ASM_OFFSET & 0xff] + + @ restore regs and return + ldmfd sp!, {r0 - r12, pc} + +ENTRY(omap730_cpu_suspend_sz) + .word . - omap730_cpu_suspend +#endif /* CONFIG_ARCH_OMAP730 */ + +#ifdef CONFIG_ARCH_OMAP15XX ENTRY(omap1510_cpu_suspend) @ save registers on stack @@ -241,7 +370,7 @@ l_1510_2: ENTRY(omap1510_cpu_suspend_sz) .word . - omap1510_cpu_suspend -#endif /* CONFIG_ARCH_OMAP1510 */ +#endif /* CONFIG_ARCH_OMAP15XX */ #if defined(CONFIG_ARCH_OMAP16XX) ENTRY(omap1610_cpu_suspend) diff --git a/arch/arm/plat-omap/sram.c b/arch/arm/plat-omap/sram.c index 7ad69f14a3e7..792f66375830 100644 --- a/arch/arm/plat-omap/sram.c +++ b/arch/arm/plat-omap/sram.c @@ -20,10 +20,13 @@ #include #include -#include "sram.h" +#include + +#define OMAP1_SRAM_PA 0x20000000 +#define OMAP1_SRAM_VA 0xd0000000 +#define OMAP2_SRAM_PA 0x40200000 +#define OMAP2_SRAM_VA 0xd0000000 -#define OMAP1_SRAM_BASE 0xd0000000 -#define OMAP1_SRAM_START 0x20000000 #define SRAM_BOOTLOADER_SZ 0x80 static unsigned long omap_sram_base; @@ -31,37 +34,40 @@ static unsigned long omap_sram_size; static unsigned long omap_sram_ceil; /* - * The amount of SRAM depends on the core type: - * 730 = 200K, 1510 = 512K, 5912 = 256K, 1610 = 16K, 1710 = 16K + * The amount of SRAM depends on the core type. * Note that we cannot try to test for SRAM here because writes * to secure SRAM will hang the system. Also the SRAM is not * yet mapped at this point. */ void __init omap_detect_sram(void) { - omap_sram_base = OMAP1_SRAM_BASE; + if (!cpu_is_omap24xx()) + omap_sram_base = OMAP1_SRAM_VA; + else + omap_sram_base = OMAP2_SRAM_VA; if (cpu_is_omap730()) - omap_sram_size = 0x32000; - else if (cpu_is_omap1510()) - omap_sram_size = 0x80000; + omap_sram_size = 0x32000; /* 200K */ + else if (cpu_is_omap15xx()) + omap_sram_size = 0x30000; /* 192K */ else if (cpu_is_omap1610() || cpu_is_omap1621() || cpu_is_omap1710()) - omap_sram_size = 0x4000; + omap_sram_size = 0x4000; /* 16K */ else if (cpu_is_omap1611()) - omap_sram_size = 0x3e800; + omap_sram_size = 0x3e800; /* 250K */ + else if (cpu_is_omap2420()) + omap_sram_size = 0xa0014; /* 640K */ else { printk(KERN_ERR "Could not detect SRAM size\n"); omap_sram_size = 0x4000; } - printk(KERN_INFO "SRAM size: 0x%lx\n", omap_sram_size); omap_sram_ceil = omap_sram_base + omap_sram_size; } static struct map_desc omap_sram_io_desc[] __initdata = { { /* .length gets filled in at runtime */ - .virtual = OMAP1_SRAM_BASE, - .pfn = __phys_to_pfn(OMAP1_SRAM_START), + .virtual = OMAP1_SRAM_VA, + .pfn = __phys_to_pfn(OMAP1_SRAM_PA), .type = MT_DEVICE } }; @@ -76,10 +82,19 @@ void __init omap_map_sram(void) if (omap_sram_size == 0) return; + if (cpu_is_omap24xx()) { + omap_sram_io_desc[0].virtual = OMAP2_SRAM_VA; + omap_sram_io_desc[0].pfn = __phys_to_pfn(OMAP2_SRAM_PA); + } + omap_sram_io_desc[0].length = (omap_sram_size + PAGE_SIZE-1)/PAGE_SIZE; omap_sram_io_desc[0].length *= PAGE_SIZE; iotable_init(omap_sram_io_desc, ARRAY_SIZE(omap_sram_io_desc)); + printk(KERN_INFO "SRAM: Mapped pa 0x%08lx to va 0x%08lx size: 0x%lx\n", + omap_sram_io_desc[0].pfn, omap_sram_io_desc[0].virtual, + omap_sram_io_desc[0].length); + /* * Looks like we need to preserve some bootloader code at the * beginning of SRAM for jumping to flash for reboot to work... @@ -88,16 +103,6 @@ void __init omap_map_sram(void) omap_sram_size - SRAM_BOOTLOADER_SZ); } -static void (*_omap_sram_reprogram_clock)(u32 dpllctl, u32 ckctl) = NULL; - -void omap_sram_reprogram_clock(u32 dpllctl, u32 ckctl) -{ - if (_omap_sram_reprogram_clock == NULL) - panic("Cannot use SRAM"); - - return _omap_sram_reprogram_clock(dpllctl, ckctl); -} - void * omap_sram_push(void * start, unsigned long size) { if (size > (omap_sram_ceil - (omap_sram_base + SRAM_BOOTLOADER_SZ))) { @@ -111,10 +116,94 @@ void * omap_sram_push(void * start, unsigned long size) return (void *)omap_sram_ceil; } -void __init omap_sram_init(void) +static void omap_sram_error(void) +{ + panic("Uninitialized SRAM function\n"); +} + +#ifdef CONFIG_ARCH_OMAP1 + +static void (*_omap_sram_reprogram_clock)(u32 dpllctl, u32 ckctl); + +void omap_sram_reprogram_clock(u32 dpllctl, u32 ckctl) +{ + if (!_omap_sram_reprogram_clock) + omap_sram_error(); + + return _omap_sram_reprogram_clock(dpllctl, ckctl); +} + +int __init omap1_sram_init(void) +{ + _omap_sram_reprogram_clock = omap_sram_push(sram_reprogram_clock, + sram_reprogram_clock_sz); + + return 0; +} + +#else +#define omap1_sram_init() do {} while (0) +#endif + +#ifdef CONFIG_ARCH_OMAP2 + +static void (*_omap2_sram_ddr_init)(u32 *slow_dll_ctrl, u32 fast_dll_ctrl, + u32 base_cs, u32 force_unlock); + +void omap2_sram_ddr_init(u32 *slow_dll_ctrl, u32 fast_dll_ctrl, + u32 base_cs, u32 force_unlock) +{ + if (!_omap2_sram_ddr_init) + omap_sram_error(); + + return _omap2_sram_ddr_init(slow_dll_ctrl, fast_dll_ctrl, + base_cs, force_unlock); +} + +static void (*_omap2_sram_reprogram_sdrc)(u32 perf_level, u32 dll_val, + u32 mem_type); + +void omap2_sram_reprogram_sdrc(u32 perf_level, u32 dll_val, u32 mem_type) +{ + if (!_omap2_sram_reprogram_sdrc) + omap_sram_error(); + + return _omap2_sram_reprogram_sdrc(perf_level, dll_val, mem_type); +} + +static u32 (*_omap2_set_prcm)(u32 dpll_ctrl_val, u32 sdrc_rfr_val, int bypass); + +u32 omap2_set_prcm(u32 dpll_ctrl_val, u32 sdrc_rfr_val, int bypass) +{ + if (!_omap2_set_prcm) + omap_sram_error(); + + return _omap2_set_prcm(dpll_ctrl_val, sdrc_rfr_val, bypass); +} + +int __init omap2_sram_init(void) +{ + _omap2_sram_ddr_init = omap_sram_push(sram_ddr_init, sram_ddr_init_sz); + + _omap2_sram_reprogram_sdrc = omap_sram_push(sram_reprogram_sdrc, + sram_reprogram_sdrc_sz); + _omap2_set_prcm = omap_sram_push(sram_set_prcm, sram_set_prcm_sz); + + return 0; +} +#else +#define omap2_sram_init() do {} while (0) +#endif + +int __init omap_sram_init(void) { omap_detect_sram(); omap_map_sram(); - _omap_sram_reprogram_clock = omap_sram_push(sram_reprogram_clock, - sram_reprogram_clock_sz); + + if (!cpu_is_omap24xx()) + omap1_sram_init(); + else + omap2_sram_init(); + + return 0; } diff --git a/arch/arm/plat-omap/sram.h b/arch/arm/plat-omap/sram.h deleted file mode 100644 index 71984efa6ae8..000000000000 --- a/arch/arm/plat-omap/sram.h +++ /dev/null @@ -1,21 +0,0 @@ -/* - * linux/arch/arm/plat-omap/sram.h - * - * Interface for functions that need to be run in internal SRAM - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - */ - -#ifndef __ARCH_ARM_OMAP_SRAM_H -#define __ARCH_ARM_OMAP_SRAM_H - -extern void * omap_sram_push(void * start, unsigned long size); -extern void omap_sram_reprogram_clock(u32 dpllctl, u32 ckctl); - -/* Do not use these */ -extern void sram_reprogram_clock(u32 ckctl, u32 dpllctl); -extern unsigned long sram_reprogram_clock_sz; - -#endif diff --git a/arch/arm/plat-omap/usb.c b/arch/arm/plat-omap/usb.c index 205e2d0b826d..00afc7a8c2ab 100644 --- a/arch/arm/plat-omap/usb.c +++ b/arch/arm/plat-omap/usb.c @@ -91,6 +91,8 @@ EXPORT_SYMBOL(otg_set_transceiver); /*-------------------------------------------------------------------------*/ +#if defined(CONFIG_ARCH_OMAP_OTG) || defined(CONFIG_ARCH_OMAP15XX) + static u32 __init omap_usb0_init(unsigned nwires, unsigned is_device) { u32 syscon1 = 0; @@ -271,6 +273,8 @@ static u32 __init omap_usb2_init(unsigned nwires, unsigned alt_pingroup) return syscon1 << 24; } +#endif + /*-------------------------------------------------------------------------*/ #if defined(CONFIG_USB_GADGET_OMAP) || \ @@ -494,7 +498,7 @@ static inline void omap_otg_init(struct omap_usb_config *config) {} /*-------------------------------------------------------------------------*/ -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX #define ULPD_DPLL_CTRL_REG __REG16(ULPD_DPLL_CTRL) #define DPLL_IOB (1 << 13) @@ -507,7 +511,6 @@ static inline void omap_otg_init(struct omap_usb_config *config) {} static void __init omap_1510_usb_init(struct omap_usb_config *config) { - int status; unsigned int val; omap_usb0_init(config->pins[0], is_usb0_device(config)); @@ -539,6 +542,8 @@ static void __init omap_1510_usb_init(struct omap_usb_config *config) #ifdef CONFIG_USB_GADGET_OMAP if (config->register_dev) { + int status; + udc_device.dev.platform_data = config; status = platform_device_register(&udc_device); if (status) @@ -549,6 +554,8 @@ static void __init omap_1510_usb_init(struct omap_usb_config *config) #if defined(CONFIG_USB_OHCI_HCD) || defined(CONFIG_USB_OHCI_HCD_MODULE) if (config->register_host) { + int status; + ohci_device.dev.platform_data = config; status = platform_device_register(&ohci_device); if (status) From 1dbae815a724303b46ab4663b5fc23c13e9d9690 Mon Sep 17 00:00:00 2001 From: Tony Lindgren Date: Thu, 10 Nov 2005 14:26:51 +0000 Subject: [PATCH 093/121] [ARM] 3145/1: OMAP 3a/5: Add support for omap24xx Patch from Tony Lindgren This patch adds support for omap24xx series of processors. The files live in arch/arm/mach-omap2, and share common files with omap15xx and omap16xx processors in arch/arm/plat-omap. Omap24xx support was originally added for 2.6.9 by TI. This code was then improved and integrated to share common code with omap15xx and omap16xx processors by various omap developers, such as Paul Mundt, Juha Yrjola, Imre Deak, Tony Lindgren, Richard Woodruff, Nishant Menon, Komal Shah et al. Signed-off-by: Tony Lindgren Signed-off-by: Russell King --- arch/arm/Kconfig | 2 + arch/arm/Makefile | 1 + arch/arm/mach-omap2/Kconfig | 22 ++ arch/arm/mach-omap2/Makefile | 13 + arch/arm/mach-omap2/Makefile.boot | 3 + arch/arm/mach-omap2/board-generic.c | 80 ++++++ arch/arm/mach-omap2/board-h4.c | 197 +++++++++++++ arch/arm/mach-omap2/devices.c | 89 ++++++ arch/arm/mach-omap2/id.c | 124 ++++++++ arch/arm/mach-omap2/io.c | 53 ++++ arch/arm/mach-omap2/irq.c | 149 ++++++++++ arch/arm/mach-omap2/mux.c | 65 +++++ arch/arm/mach-omap2/prcm.h | 419 ++++++++++++++++++++++++++++ arch/arm/mach-omap2/serial.c | 180 ++++++++++++ arch/arm/mach-omap2/sram-fn.S | 333 ++++++++++++++++++++++ arch/arm/mach-omap2/timer-gp.c | 126 +++++++++ arch/arm/mm/Kconfig | 2 +- 17 files changed, 1857 insertions(+), 1 deletion(-) create mode 100644 arch/arm/mach-omap2/Kconfig create mode 100644 arch/arm/mach-omap2/Makefile create mode 100644 arch/arm/mach-omap2/Makefile.boot create mode 100644 arch/arm/mach-omap2/board-generic.c create mode 100644 arch/arm/mach-omap2/board-h4.c create mode 100644 arch/arm/mach-omap2/devices.c create mode 100644 arch/arm/mach-omap2/id.c create mode 100644 arch/arm/mach-omap2/io.c create mode 100644 arch/arm/mach-omap2/irq.c create mode 100644 arch/arm/mach-omap2/mux.c create mode 100644 arch/arm/mach-omap2/prcm.h create mode 100644 arch/arm/mach-omap2/serial.c create mode 100644 arch/arm/mach-omap2/sram-fn.S create mode 100644 arch/arm/mach-omap2/timer-gp.c diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index ec77721507cb..3df7cbd924a1 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -239,6 +239,8 @@ source "arch/arm/plat-omap/Kconfig" source "arch/arm/mach-omap1/Kconfig" +source "arch/arm/mach-omap2/Kconfig" + source "arch/arm/mach-s3c2410/Kconfig" source "arch/arm/mach-lh7a40x/Kconfig" diff --git a/arch/arm/Makefile b/arch/arm/Makefile index 114cda7f1b73..81bd2193fe6d 100644 --- a/arch/arm/Makefile +++ b/arch/arm/Makefile @@ -93,6 +93,7 @@ textaddr-$(CONFIG_ARCH_FORTUNET) := 0xc0008000 machine-$(CONFIG_ARCH_IXP4XX) := ixp4xx machine-$(CONFIG_ARCH_IXP2000) := ixp2000 machine-$(CONFIG_ARCH_OMAP1) := omap1 + machine-$(CONFIG_ARCH_OMAP2) := omap2 incdir-$(CONFIG_ARCH_OMAP) := omap machine-$(CONFIG_ARCH_S3C2410) := s3c2410 machine-$(CONFIG_ARCH_LH7A40X) := lh7a40x diff --git a/arch/arm/mach-omap2/Kconfig b/arch/arm/mach-omap2/Kconfig new file mode 100644 index 000000000000..578880943cf2 --- /dev/null +++ b/arch/arm/mach-omap2/Kconfig @@ -0,0 +1,22 @@ +comment "OMAP Core Type" + depends on ARCH_OMAP2 + +config ARCH_OMAP24XX + bool "OMAP24xx Based System" + depends on ARCH_OMAP2 + +config ARCH_OMAP2420 + bool "OMAP2420 support" + depends on ARCH_OMAP24XX + +comment "OMAP Board Type" + depends on ARCH_OMAP2 + +config MACH_OMAP_GENERIC + bool "Generic OMAP board" + depends on ARCH_OMAP2 && ARCH_OMAP24XX + +config MACH_OMAP_H4 + bool "OMAP 2420 H4 board" + depends on ARCH_OMAP2 && ARCH_OMAP24XX + diff --git a/arch/arm/mach-omap2/Makefile b/arch/arm/mach-omap2/Makefile new file mode 100644 index 000000000000..42041166435c --- /dev/null +++ b/arch/arm/mach-omap2/Makefile @@ -0,0 +1,13 @@ +# +# Makefile for the linux kernel. +# + +# Common support +obj-y := irq.o id.o io.o sram-fn.o clock.o mux.o devices.o serial.o + +obj-$(CONFIG_OMAP_MPU_TIMER) += timer-gp.o + +# Specific board support +obj-$(CONFIG_MACH_OMAP_GENERIC) += board-generic.o +obj-$(CONFIG_MACH_OMAP_H4) += board-h4.o + diff --git a/arch/arm/mach-omap2/Makefile.boot b/arch/arm/mach-omap2/Makefile.boot new file mode 100644 index 000000000000..565aff7f37a9 --- /dev/null +++ b/arch/arm/mach-omap2/Makefile.boot @@ -0,0 +1,3 @@ + zreladdr-y := 0x80008000 +params_phys-y := 0x80000100 +initrd_phys-y := 0x80800000 diff --git a/arch/arm/mach-omap2/board-generic.c b/arch/arm/mach-omap2/board-generic.c new file mode 100644 index 000000000000..c602e7a3d93e --- /dev/null +++ b/arch/arm/mach-omap2/board-generic.c @@ -0,0 +1,80 @@ +/* + * linux/arch/arm/mach-omap/omap2/board-generic.c + * + * Copyright (C) 2005 Nokia Corporation + * Author: Paul Mundt + * + * Modified from mach-omap/omap1/board-generic.c + * + * Code for generic OMAP2 board. Should work on many OMAP2 systems where + * the bootloader passes the board-specific data to the kernel. + * Do not put any board specific code to this file; create a new machine + * type if you need custom low-level initializations. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include +#include +#include + +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +static void __init omap_generic_init_irq(void) +{ + omap_init_irq(); +} + +static struct omap_uart_config generic_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1) | (1 << 2)), +}; + +static struct omap_mmc_config generic_mmc_config __initdata = { + .mmc [0] = { + .enabled = 0, + .wire4 = 0, + .wp_pin = -1, + .power_pin = -1, + .switch_pin = -1, + }, +}; + +static struct omap_board_config_kernel generic_config[] = { + { OMAP_TAG_UART, &generic_uart_config }, + { OMAP_TAG_MMC, &generic_mmc_config }, +}; + +static void __init omap_generic_init(void) +{ + omap_board_config = generic_config; + omap_board_config_size = ARRAY_SIZE(generic_config); + omap_serial_init(); +} + +static void __init omap_generic_map_io(void) +{ + omap_map_common_io(); +} + +MACHINE_START(OMAP_GENERIC, "Generic OMAP24xx") + /* Maintainer: Paul Mundt */ + .phys_ram = 0x80000000, + .phys_io = 0x48000000, + .io_pg_offst = ((0xd8000000) >> 18) & 0xfffc, + .boot_params = 0x80000100, + .map_io = omap_generic_map_io, + .init_irq = omap_generic_init_irq, + .init_machine = omap_generic_init, + .timer = &omap_timer, +MACHINE_END diff --git a/arch/arm/mach-omap2/board-h4.c b/arch/arm/mach-omap2/board-h4.c new file mode 100644 index 000000000000..f2554469a76a --- /dev/null +++ b/arch/arm/mach-omap2/board-h4.c @@ -0,0 +1,197 @@ +/* + * linux/arch/arm/mach-omap/omap2/board-h4.c + * + * Copyright (C) 2005 Nokia Corporation + * Author: Paul Mundt + * + * Modified from mach-omap/omap1/board-generic.c + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include +#include + +#include +#include + +static struct mtd_partition h4_partitions[] = { + /* bootloader (U-Boot, etc) in first sector */ + { + .name = "bootloader", + .offset = 0, + .size = SZ_128K, + .mask_flags = MTD_WRITEABLE, /* force read-only */ + }, + /* bootloader params in the next sector */ + { + .name = "params", + .offset = MTDPART_OFS_APPEND, + .size = SZ_128K, + .mask_flags = 0, + }, + /* kernel */ + { + .name = "kernel", + .offset = MTDPART_OFS_APPEND, + .size = SZ_2M, + .mask_flags = 0 + }, + /* file system */ + { + .name = "filesystem", + .offset = MTDPART_OFS_APPEND, + .size = MTDPART_SIZ_FULL, + .mask_flags = 0 + } +}; + +static struct flash_platform_data h4_flash_data = { + .map_name = "cfi_probe", + .width = 2, + .parts = h4_partitions, + .nr_parts = ARRAY_SIZE(h4_partitions), +}; + +static struct resource h4_flash_resource = { + .start = H4_CS0_BASE, + .end = H4_CS0_BASE + SZ_64M - 1, + .flags = IORESOURCE_MEM, +}; + +static struct platform_device h4_flash_device = { + .name = "omapflash", + .id = 0, + .dev = { + .platform_data = &h4_flash_data, + }, + .num_resources = 1, + .resource = &h4_flash_resource, +}; + +static struct resource h4_smc91x_resources[] = { + [0] = { + .start = OMAP24XX_ETHR_START, /* Physical */ + .end = OMAP24XX_ETHR_START + 0xf, + .flags = IORESOURCE_MEM, + }, + [1] = { + .start = OMAP_GPIO_IRQ(OMAP24XX_ETHR_GPIO_IRQ), + .end = OMAP_GPIO_IRQ(OMAP24XX_ETHR_GPIO_IRQ), + .flags = IORESOURCE_IRQ, + }, +}; + +static struct platform_device h4_smc91x_device = { + .name = "smc91x", + .id = -1, + .num_resources = ARRAY_SIZE(h4_smc91x_resources), + .resource = h4_smc91x_resources, +}; + +static struct platform_device *h4_devices[] __initdata = { + &h4_smc91x_device, + &h4_flash_device, +}; + +static inline void __init h4_init_smc91x(void) +{ + /* Make sure CS1 timings are correct */ + GPMC_CONFIG1_1 = 0x00011200; + GPMC_CONFIG2_1 = 0x001f1f01; + GPMC_CONFIG3_1 = 0x00080803; + GPMC_CONFIG4_1 = 0x1c091c09; + GPMC_CONFIG5_1 = 0x041f1f1f; + GPMC_CONFIG6_1 = 0x000004c4; + GPMC_CONFIG7_1 = 0x00000f40 | (0x08000000 >> 24); + udelay(100); + + omap_cfg_reg(M15_24XX_GPIO92); + if (omap_request_gpio(OMAP24XX_ETHR_GPIO_IRQ) < 0) { + printk(KERN_ERR "Failed to request GPIO%d for smc91x IRQ\n", + OMAP24XX_ETHR_GPIO_IRQ); + return; + } + omap_set_gpio_direction(OMAP24XX_ETHR_GPIO_IRQ, 1); +} + +static void __init omap_h4_init_irq(void) +{ + omap_init_irq(); + omap_gpio_init(); + h4_init_smc91x(); +} + +static struct omap_uart_config h4_uart_config __initdata = { + .enabled_uarts = ((1 << 0) | (1 << 1) | (1 << 2)), +}; + +static struct omap_mmc_config h4_mmc_config __initdata = { + .mmc [0] = { + .enabled = 1, + .wire4 = 1, + .wp_pin = -1, + .power_pin = -1, + .switch_pin = -1, + }, +}; + +static struct omap_lcd_config h4_lcd_config __initdata = { + .panel_name = "h4", + .ctrl_name = "internal", +}; + +static struct omap_board_config_kernel h4_config[] = { + { OMAP_TAG_UART, &h4_uart_config }, + { OMAP_TAG_MMC, &h4_mmc_config }, + { OMAP_TAG_LCD, &h4_lcd_config }, +}; + +static void __init omap_h4_init(void) +{ + /* + * Make sure the serial ports are muxed on at this point. + * You have to mux them off in device drivers later on + * if not needed. + */ + platform_add_devices(h4_devices, ARRAY_SIZE(h4_devices)); + omap_board_config = h4_config; + omap_board_config_size = ARRAY_SIZE(h4_config); + omap_serial_init(); +} + +static void __init omap_h4_map_io(void) +{ + omap_map_common_io(); +} + +MACHINE_START(OMAP_H4, "OMAP2420 H4 board") + /* Maintainer: Paul Mundt */ + .phys_ram = 0x80000000, + .phys_io = 0x48000000, + .io_pg_offst = ((0xd8000000) >> 18) & 0xfffc, + .boot_params = 0x80000100, + .map_io = omap_h4_map_io, + .init_irq = omap_h4_init_irq, + .init_machine = omap_h4_init, + .timer = &omap_timer, +MACHINE_END diff --git a/arch/arm/mach-omap2/devices.c b/arch/arm/mach-omap2/devices.c new file mode 100644 index 000000000000..7181edb89352 --- /dev/null +++ b/arch/arm/mach-omap2/devices.c @@ -0,0 +1,89 @@ +/* + * linux/arch/arm/mach-omap2/devices.c + * + * OMAP2 platform device setup/initialization + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + */ + +#include +#include +#include +#include +#include + +#include +#include +#include +#include + +#include +#include +#include +#include + +extern void omap_nop_release(struct device *dev); + +/*-------------------------------------------------------------------------*/ + +#if defined(CONFIG_I2C_OMAP) || defined(CONFIG_I2C_OMAP_MODULE) + +#define OMAP2_I2C_BASE2 0x48072000 +#define OMAP2_I2C_INT2 57 + +static struct resource i2c_resources2[] = { + { + .start = OMAP2_I2C_BASE2, + .end = OMAP2_I2C_BASE2 + 0x3f, + .flags = IORESOURCE_MEM, + }, + { + .start = OMAP2_I2C_INT2, + .flags = IORESOURCE_IRQ, + }, +}; + +static struct platform_device omap_i2c_device2 = { + .name = "i2c_omap", + .id = 2, + .dev = { + .release = omap_nop_release, + }, + .num_resources = ARRAY_SIZE(i2c_resources2), + .resource = i2c_resources2, +}; + +/* See also arch/arm/plat-omap/devices.c for first I2C on 24xx */ +static void omap_init_i2c(void) +{ + /* REVISIT: Second I2C not in use on H4? */ + if (machine_is_omap_h4()) + return; + + omap_cfg_reg(J15_24XX_I2C2_SCL); + omap_cfg_reg(H19_24XX_I2C2_SDA); + (void) platform_device_register(&omap_i2c_device2); +} + +#else + +static void omap_init_i2c(void) {} + +#endif + +/*-------------------------------------------------------------------------*/ + +static int __init omap2_init_devices(void) +{ + /* please keep these calls, and their implementations above, + * in alphabetical order so they're easier to sort through. + */ + omap_init_i2c(); + + return 0; +} +arch_initcall(omap2_init_devices); + diff --git a/arch/arm/mach-omap2/id.c b/arch/arm/mach-omap2/id.c new file mode 100644 index 000000000000..76187300f2b6 --- /dev/null +++ b/arch/arm/mach-omap2/id.c @@ -0,0 +1,124 @@ +/* + * linux/arch/arm/mach-omap2/id.c + * + * OMAP2 CPU identification code + * + * Copyright (C) 2005 Nokia Corporation + * Written by Tony Lindgren + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include +#include +#include +#include + +#include + +#define OMAP24XX_TAP_BASE io_p2v(0x48014000) + +#define OMAP_TAP_IDCODE 0x0204 +#define OMAP_TAP_PROD_ID 0x0208 + +#define OMAP_TAP_DIE_ID_0 0x0218 +#define OMAP_TAP_DIE_ID_1 0x021C +#define OMAP_TAP_DIE_ID_2 0x0220 +#define OMAP_TAP_DIE_ID_3 0x0224 + +/* system_rev fields for OMAP2 processors: + * CPU id bits [31:16], + * CPU device type [15:12], (unprg,normal,POP) + * CPU revision [11:08] + * CPU class bits [07:00] + */ + +struct omap_id { + u16 hawkeye; /* Silicon type (Hawkeye id) */ + u8 dev; /* Device type from production_id reg */ + u32 type; /* combined type id copied to system_rev */ +}; + +/* Register values to detect the OMAP version */ +static struct omap_id omap_ids[] __initdata = { + { .hawkeye = 0xb5d9, .dev = 0x0, .type = 0x24200000 }, + { .hawkeye = 0xb5d9, .dev = 0x1, .type = 0x24201000 }, + { .hawkeye = 0xb5d9, .dev = 0x2, .type = 0x24202000 }, + { .hawkeye = 0xb5d9, .dev = 0x4, .type = 0x24220000 }, + { .hawkeye = 0xb5d9, .dev = 0x8, .type = 0x24230000 }, + { .hawkeye = 0xb68a, .dev = 0x0, .type = 0x24300000 }, +}; + +static u32 __init read_tap_reg(int reg) +{ + return __raw_readl(OMAP24XX_TAP_BASE + reg); +} + +void __init omap2_check_revision(void) +{ + int i, j; + u32 idcode; + u32 prod_id; + u16 hawkeye; + u8 dev_type; + u8 rev; + + idcode = read_tap_reg(OMAP_TAP_IDCODE); + prod_id = read_tap_reg(OMAP_TAP_PROD_ID); + hawkeye = (idcode >> 12) & 0xffff; + rev = (idcode >> 28) & 0x0f; + dev_type = (prod_id >> 16) & 0x0f; + +#ifdef DEBUG + printk(KERN_DEBUG "OMAP_TAP_IDCODE 0x%08x REV %i HAWKEYE 0x%04x MANF %03x\n", + idcode, rev, hawkeye, (idcode >> 1) & 0x7ff); + printk(KERN_DEBUG "OMAP_TAP_DIE_ID_0: 0x%08x\n", + read_tap_reg(OMAP_TAP_DIE_ID_0)); + printk(KERN_DEBUG "OMAP_TAP_DIE_ID_1: 0x%08x DEV_REV: %i\n", + read_tap_reg(OMAP_TAP_DIE_ID_1), + (read_tap_reg(OMAP_TAP_DIE_ID_1) >> 28) & 0xf); + printk(KERN_DEBUG "OMAP_TAP_DIE_ID_2: 0x%08x\n", + read_tap_reg(OMAP_TAP_DIE_ID_2)); + printk(KERN_DEBUG "OMAP_TAP_DIE_ID_3: 0x%08x\n", + read_tap_reg(OMAP_TAP_DIE_ID_3)); + printk(KERN_DEBUG "OMAP_TAP_PROD_ID_0: 0x%08x DEV_TYPE: %i\n", + prod_id, dev_type); +#endif + + /* Check hawkeye ids */ + for (i = 0; i < ARRAY_SIZE(omap_ids); i++) { + if (hawkeye == omap_ids[i].hawkeye) + break; + } + + if (i == ARRAY_SIZE(omap_ids)) { + printk(KERN_ERR "Unknown OMAP CPU id\n"); + return; + } + + for (j = i; j < ARRAY_SIZE(omap_ids); j++) { + if (dev_type == omap_ids[j].dev) + break; + } + + if (j == ARRAY_SIZE(omap_ids)) { + printk(KERN_ERR "Unknown OMAP device type. " + "Handling it as OMAP%04x\n", + omap_ids[i].type >> 16); + j = i; + } + system_rev = omap_ids[j].type; + + system_rev |= rev << 8; + + /* Add the cpu class info (24xx) */ + system_rev |= 0x24; + + pr_info("OMAP%04x", system_rev >> 16); + if ((system_rev >> 8) & 0x0f) + printk("%x", (system_rev >> 8) & 0x0f); + printk("\n"); +} + diff --git a/arch/arm/mach-omap2/io.c b/arch/arm/mach-omap2/io.c new file mode 100644 index 000000000000..8ea67bf196a5 --- /dev/null +++ b/arch/arm/mach-omap2/io.c @@ -0,0 +1,53 @@ +/* + * linux/arch/arm/mach-omap2/io.c + * + * OMAP2 I/O mapping code + * + * Copyright (C) 2005 Nokia Corporation + * Author: Juha Yrjölä + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#include +#include +#include +#include + +#include +#include +#include + +extern void omap_sram_init(void); +extern int omap2_clk_init(void); +extern void omap2_check_revision(void); + +/* + * The machine specific code may provide the extra mapping besides the + * default mapping provided here. + */ +static struct map_desc omap2_io_desc[] __initdata = { + { + .virtual = L3_24XX_VIRT, + .pfn = __phys_to_pfn(L3_24XX_PHYS), + .length = L3_24XX_SIZE, + .type = MT_DEVICE + }, + { + .virtual = L4_24XX_VIRT, + .pfn = __phys_to_pfn(L4_24XX_PHYS), + .length = L4_24XX_SIZE, + .type = MT_DEVICE + } +}; + +void __init omap_map_common_io(void) +{ + iotable_init(omap2_io_desc, ARRAY_SIZE(omap2_io_desc)); + omap2_check_revision(); + omap_sram_init(); + omap2_mux_init(); + omap2_clk_init(); +} diff --git a/arch/arm/mach-omap2/irq.c b/arch/arm/mach-omap2/irq.c new file mode 100644 index 000000000000..d7baff675cfe --- /dev/null +++ b/arch/arm/mach-omap2/irq.c @@ -0,0 +1,149 @@ +/* + * linux/arch/arm/mach-omap/omap2/irq.c + * + * Interrupt handler for OMAP2 boards. + * + * Copyright (C) 2005 Nokia Corporation + * Author: Paul Mundt + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + */ +#include +#include +#include +#include +#include +#include +#include +#include + +#define INTC_REVISION 0x0000 +#define INTC_SYSCONFIG 0x0010 +#define INTC_SYSSTATUS 0x0014 +#define INTC_CONTROL 0x0048 +#define INTC_MIR_CLEAR0 0x0088 +#define INTC_MIR_SET0 0x008c + +/* + * OMAP2 has a number of different interrupt controllers, each interrupt + * controller is identified as its own "bank". Register definitions are + * fairly consistent for each bank, but not all registers are implemented + * for each bank.. when in doubt, consult the TRM. + */ +static struct omap_irq_bank { + unsigned long base_reg; + unsigned int nr_irqs; +} __attribute__ ((aligned(4))) irq_banks[] = { + { + /* MPU INTC */ + .base_reg = OMAP24XX_IC_BASE, + .nr_irqs = 96, + }, { + /* XXX: DSP INTC */ + +#if 0 + /* + * Commented out for now until we fix the IVA clocking + */ +#ifdef CONFIG_ARCH_OMAP2420 + }, { + /* IVA INTC (2420 only) */ + .base_reg = OMAP24XX_IVA_INTC_BASE, + .nr_irqs = 16, /* Actually 32, but only 16 are used */ +#endif +#endif + } +}; + +/* XXX: FIQ and additional INTC support (only MPU at the moment) */ +static void omap_ack_irq(unsigned int irq) +{ + omap_writel(0x1, irq_banks[0].base_reg + INTC_CONTROL); +} + +static void omap_mask_irq(unsigned int irq) +{ + int offset = (irq >> 5) << 5; + + if (irq >= 64) { + irq %= 64; + } else if (irq >= 32) { + irq %= 32; + } + + omap_writel(1 << irq, irq_banks[0].base_reg + INTC_MIR_SET0 + offset); +} + +static void omap_unmask_irq(unsigned int irq) +{ + int offset = (irq >> 5) << 5; + + if (irq >= 64) { + irq %= 64; + } else if (irq >= 32) { + irq %= 32; + } + + omap_writel(1 << irq, irq_banks[0].base_reg + INTC_MIR_CLEAR0 + offset); +} + +static void omap_mask_ack_irq(unsigned int irq) +{ + omap_mask_irq(irq); + omap_ack_irq(irq); +} + +static struct irqchip omap_irq_chip = { + .ack = omap_mask_ack_irq, + .mask = omap_mask_irq, + .unmask = omap_unmask_irq, +}; + +static void __init omap_irq_bank_init_one(struct omap_irq_bank *bank) +{ + unsigned long tmp; + + tmp = omap_readl(bank->base_reg + INTC_REVISION) & 0xff; + printk(KERN_INFO "IRQ: Found an INTC at 0x%08lx " + "(revision %ld.%ld) with %d interrupts\n", + bank->base_reg, tmp >> 4, tmp & 0xf, bank->nr_irqs); + + tmp = omap_readl(bank->base_reg + INTC_SYSCONFIG); + tmp |= 1 << 1; /* soft reset */ + omap_writel(tmp, bank->base_reg + INTC_SYSCONFIG); + + while (!(omap_readl(bank->base_reg + INTC_SYSSTATUS) & 0x1)) + /* Wait for reset to complete */; +} + +void __init omap_init_irq(void) +{ + unsigned long nr_irqs = 0; + unsigned int nr_banks = 0; + int i; + + for (i = 0; i < ARRAY_SIZE(irq_banks); i++) { + struct omap_irq_bank *bank = irq_banks + i; + + /* XXX */ + if (!bank->base_reg) + continue; + + omap_irq_bank_init_one(bank); + + nr_irqs += bank->nr_irqs; + nr_banks++; + } + + printk(KERN_INFO "Total of %ld interrupts on %d active controller%s\n", + nr_irqs, nr_banks, nr_banks > 1 ? "s" : ""); + + for (i = 0; i < nr_irqs; i++) { + set_irq_chip(i, &omap_irq_chip); + set_irq_handler(i, do_level_IRQ); + set_irq_flags(i, IRQF_VALID); + } +} + diff --git a/arch/arm/mach-omap2/mux.c b/arch/arm/mach-omap2/mux.c new file mode 100644 index 000000000000..ea4654815dd1 --- /dev/null +++ b/arch/arm/mach-omap2/mux.c @@ -0,0 +1,65 @@ +/* + * linux/arch/arm/mach-omap2/mux.c + * + * OMAP1 pin multiplexing configurations + * + * Copyright (C) 2003 - 2005 Nokia Corporation + * + * Written by Tony Lindgren + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + */ +#include +#include +#include +#include +#include +#include + +#include + +#ifdef CONFIG_OMAP_MUX + +/* NOTE: See mux.h for the enumeration */ + +struct pin_config __initdata_or_module omap24xx_pins[] = { +/* + * description mux mux pull pull debug + * offset mode ena type + */ + +/* 24xx I2C */ +MUX_CFG_24XX("M19_24XX_I2C1_SCL", 0x111, 0, 0, 0, 1) +MUX_CFG_24XX("L15_24XX_I2C1_SDA", 0x112, 0, 0, 0, 1) +MUX_CFG_24XX("J15_24XX_I2C2_SCL", 0x113, 0, 0, 0, 1) +MUX_CFG_24XX("H19_24XX_I2C2_SDA", 0x114, 0, 0, 0, 1) + +/* Menelaus interrupt */ +MUX_CFG_24XX("W19_24XX_SYS_NIRQ", 0x12c, 0, 1, 1, 1) + +/* 24xx GPIO */ +MUX_CFG_24XX("Y20_24XX_GPIO60", 0x12c, 3, 0, 0, 1) +MUX_CFG_24XX("M15_24XX_GPIO92", 0x10a, 3, 0, 0, 1) + +}; + +int __init omap2_mux_init(void) +{ + omap_mux_register(omap24xx_pins, ARRAY_SIZE(omap24xx_pins)); + return 0; +} + +#endif diff --git a/arch/arm/mach-omap2/prcm.h b/arch/arm/mach-omap2/prcm.h new file mode 100644 index 000000000000..2eb89b936c83 --- /dev/null +++ b/arch/arm/mach-omap2/prcm.h @@ -0,0 +1,419 @@ +/* + * prcm.h - Access definations for use in OMAP24XX clock and power management + * + * Copyright (C) 2005 Texas Instruments, Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef __ASM_ARM_ARCH_DPM_PRCM_H +#define __ASM_ARM_ARCH_DPM_PRCM_H + +/* SET_PERFORMANCE_LEVEL PARAMETERS */ +#define PRCM_HALF_SPEED 1 +#define PRCM_FULL_SPEED 2 + +#ifndef __ASSEMBLER__ + +#define PRCM_REG32(offset) __REG32(OMAP24XX_PRCM_BASE + (offset)) + +#define PRCM_REVISION PRCM_REG32(0x000) +#define PRCM_SYSCONFIG PRCM_REG32(0x010) +#define PRCM_IRQSTATUS_MPU PRCM_REG32(0x018) +#define PRCM_IRQENABLE_MPU PRCM_REG32(0x01C) +#define PRCM_VOLTCTRL PRCM_REG32(0x050) +#define PRCM_VOLTST PRCM_REG32(0x054) +#define PRCM_CLKSRC_CTRL PRCM_REG32(0x060) +#define PRCM_CLKOUT_CTRL PRCM_REG32(0x070) +#define PRCM_CLKEMUL_CTRL PRCM_REG32(0x078) +#define PRCM_CLKCFG_CTRL PRCM_REG32(0x080) +#define PRCM_CLKCFG_STATUS PRCM_REG32(0x084) +#define PRCM_VOLTSETUP PRCM_REG32(0x090) +#define PRCM_CLKSSETUP PRCM_REG32(0x094) +#define PRCM_POLCTRL PRCM_REG32(0x098) + +/* GENERAL PURPOSE */ +#define GENERAL_PURPOSE1 PRCM_REG32(0x0B0) +#define GENERAL_PURPOSE2 PRCM_REG32(0x0B4) +#define GENERAL_PURPOSE3 PRCM_REG32(0x0B8) +#define GENERAL_PURPOSE4 PRCM_REG32(0x0BC) +#define GENERAL_PURPOSE5 PRCM_REG32(0x0C0) +#define GENERAL_PURPOSE6 PRCM_REG32(0x0C4) +#define GENERAL_PURPOSE7 PRCM_REG32(0x0C8) +#define GENERAL_PURPOSE8 PRCM_REG32(0x0CC) +#define GENERAL_PURPOSE9 PRCM_REG32(0x0D0) +#define GENERAL_PURPOSE10 PRCM_REG32(0x0D4) +#define GENERAL_PURPOSE11 PRCM_REG32(0x0D8) +#define GENERAL_PURPOSE12 PRCM_REG32(0x0DC) +#define GENERAL_PURPOSE13 PRCM_REG32(0x0E0) +#define GENERAL_PURPOSE14 PRCM_REG32(0x0E4) +#define GENERAL_PURPOSE15 PRCM_REG32(0x0E8) +#define GENERAL_PURPOSE16 PRCM_REG32(0x0EC) +#define GENERAL_PURPOSE17 PRCM_REG32(0x0F0) +#define GENERAL_PURPOSE18 PRCM_REG32(0x0F4) +#define GENERAL_PURPOSE19 PRCM_REG32(0x0F8) +#define GENERAL_PURPOSE20 PRCM_REG32(0x0FC) + +/* MPU */ +#define CM_CLKSEL_MPU PRCM_REG32(0x140) +#define CM_CLKSTCTRL_MPU PRCM_REG32(0x148) +#define RM_RSTST_MPU PRCM_REG32(0x158) +#define PM_WKDEP_MPU PRCM_REG32(0x1C8) +#define PM_EVGENCTRL_MPU PRCM_REG32(0x1D4) +#define PM_EVEGENONTIM_MPU PRCM_REG32(0x1D8) +#define PM_EVEGENOFFTIM_MPU PRCM_REG32(0x1DC) +#define PM_PWSTCTRL_MPU PRCM_REG32(0x1E0) +#define PM_PWSTST_MPU PRCM_REG32(0x1E4) + +/* CORE */ +#define CM_FCLKEN1_CORE PRCM_REG32(0x200) +#define CM_FCLKEN2_CORE PRCM_REG32(0x204) +#define CM_FCLKEN3_CORE PRCM_REG32(0x208) +#define CM_ICLKEN1_CORE PRCM_REG32(0x210) +#define CM_ICLKEN2_CORE PRCM_REG32(0x214) +#define CM_ICLKEN3_CORE PRCM_REG32(0x218) +#define CM_ICLKEN4_CORE PRCM_REG32(0x21C) +#define CM_IDLEST1_CORE PRCM_REG32(0x220) +#define CM_IDLEST2_CORE PRCM_REG32(0x224) +#define CM_IDLEST3_CORE PRCM_REG32(0x228) +#define CM_IDLEST4_CORE PRCM_REG32(0x22C) +#define CM_AUTOIDLE1_CORE PRCM_REG32(0x230) +#define CM_AUTOIDLE2_CORE PRCM_REG32(0x234) +#define CM_AUTOIDLE3_CORE PRCM_REG32(0x238) +#define CM_AUTOIDLE4_CORE PRCM_REG32(0x23C) +#define CM_CLKSEL1_CORE PRCM_REG32(0x240) +#define CM_CLKSEL2_CORE PRCM_REG32(0x244) +#define CM_CLKSTCTRL_CORE PRCM_REG32(0x248) +#define PM_WKEN1_CORE PRCM_REG32(0x2A0) +#define PM_WKEN2_CORE PRCM_REG32(0x2A4) +#define PM_WKST1_CORE PRCM_REG32(0x2B0) +#define PM_WKST2_CORE PRCM_REG32(0x2B4) +#define PM_WKDEP_CORE PRCM_REG32(0x2C8) +#define PM_PWSTCTRL_CORE PRCM_REG32(0x2E0) +#define PM_PWSTST_CORE PRCM_REG32(0x2E4) + +/* GFX */ +#define CM_FCLKEN_GFX PRCM_REG32(0x300) +#define CM_ICLKEN_GFX PRCM_REG32(0x310) +#define CM_IDLEST_GFX PRCM_REG32(0x320) +#define CM_CLKSEL_GFX PRCM_REG32(0x340) +#define CM_CLKSTCTRL_GFX PRCM_REG32(0x348) +#define RM_RSTCTRL_GFX PRCM_REG32(0x350) +#define RM_RSTST_GFX PRCM_REG32(0x358) +#define PM_WKDEP_GFX PRCM_REG32(0x3C8) +#define PM_PWSTCTRL_GFX PRCM_REG32(0x3E0) +#define PM_PWSTST_GFX PRCM_REG32(0x3E4) + +/* WAKE-UP */ +#define CM_FCLKEN_WKUP PRCM_REG32(0x400) +#define CM_ICLKEN_WKUP PRCM_REG32(0x410) +#define CM_IDLEST_WKUP PRCM_REG32(0x420) +#define CM_AUTOIDLE_WKUP PRCM_REG32(0x430) +#define CM_CLKSEL_WKUP PRCM_REG32(0x440) +#define RM_RSTCTRL_WKUP PRCM_REG32(0x450) +#define RM_RSTTIME_WKUP PRCM_REG32(0x454) +#define RM_RSTST_WKUP PRCM_REG32(0x458) +#define PM_WKEN_WKUP PRCM_REG32(0x4A0) +#define PM_WKST_WKUP PRCM_REG32(0x4B0) + +/* CLOCKS */ +#define CM_CLKEN_PLL PRCM_REG32(0x500) +#define CM_IDLEST_CKGEN PRCM_REG32(0x520) +#define CM_AUTOIDLE_PLL PRCM_REG32(0x530) +#define CM_CLKSEL1_PLL PRCM_REG32(0x540) +#define CM_CLKSEL2_PLL PRCM_REG32(0x544) + +/* DSP */ +#define CM_FCLKEN_DSP PRCM_REG32(0x800) +#define CM_ICLKEN_DSP PRCM_REG32(0x810) +#define CM_IDLEST_DSP PRCM_REG32(0x820) +#define CM_AUTOIDLE_DSP PRCM_REG32(0x830) +#define CM_CLKSEL_DSP PRCM_REG32(0x840) +#define CM_CLKSTCTRL_DSP PRCM_REG32(0x848) +#define RM_RSTCTRL_DSP PRCM_REG32(0x850) +#define RM_RSTST_DSP PRCM_REG32(0x858) +#define PM_WKEN_DSP PRCM_REG32(0x8A0) +#define PM_WKDEP_DSP PRCM_REG32(0x8C8) +#define PM_PWSTCTRL_DSP PRCM_REG32(0x8E0) +#define PM_PWSTST_DSP PRCM_REG32(0x8E4) +#define PRCM_IRQSTATUS_DSP PRCM_REG32(0x8F0) +#define PRCM_IRQENABLE_DSP PRCM_REG32(0x8F4) + +/* IVA */ +#define PRCM_IRQSTATUS_IVA PRCM_REG32(0x8F8) +#define PRCM_IRQENABLE_IVA PRCM_REG32(0x8FC) + +/* Modem on 2430 */ +#define CM_FCLKEN_MDM PRCM_REG32(0xC00) +#define CM_ICLKEN_MDM PRCM_REG32(0xC10) +#define CM_IDLEST_MDM PRCM_REG32(0xC20) +#define CM_CLKSEL_MDM PRCM_REG32(0xC40) + +/* FIXME: Move to header for 2430 */ +#define DISP_BASE (OMAP24XX_L4_IO_BASE+0x50000) +#define DISP_REG32(offset) __REG32(DISP_BASE + (offset)) + +#define GPMC_BASE (OMAP24XX_GPMC_BASE) +#define GPMC_REG32(offset) __REG32(GPMC_BASE + (offset)) + +#define GPT1_BASE (OMAP24XX_GPT1) +#define GPT1_REG32(offset) __REG32(GPT1_BASE + (offset)) + +/* Misc sysconfig */ +#define DISPC_SYSCONFIG DISP_REG32(0x410) +#define SPI_BASE (OMAP24XX_L4_IO_BASE+0x98000) +#define MCSPI1_SYSCONFIG __REG32(SPI_BASE + 0x10) +#define MCSPI2_SYSCONFIG __REG32(SPI_BASE+0x2000 + 0x10) + +//#define DSP_MMU_SYSCONFIG 0x5A000010 +#define CAMERA_MMU_SYSCONFIG __REG32(DISP_BASE+0x2C10) +//#define IVA_MMU_SYSCONFIG 0x5D000010 +//#define DSP_DMA_SYSCONFIG 0x00FCC02C +#define CAMERA_DMA_SYSCONFIG __REG32(DISP_BASE+0x282C) +#define SYSTEM_DMA_SYSCONFIG __REG32(DISP_BASE+0x602C) +#define GPMC_SYSCONFIG GPMC_REG32(0x010) +#define MAILBOXES_SYSCONFIG __REG32(OMAP24XX_L4_IO_BASE+0x94010) +#define UART1_SYSCONFIG __REG32(OMAP24XX_L4_IO_BASE+0x6A054) +#define UART2_SYSCONFIG __REG32(OMAP24XX_L4_IO_BASE+0x6C054) +#define UART3_SYSCONFIG __REG32(OMAP24XX_L4_IO_BASE+0x6E054) +//#define IVA_SYSCONFIG 0x5C060010 +#define SDRC_SYSCONFIG __REG32(OMAP24XX_SDRC_BASE+0x10) +#define SMS_SYSCONFIG __REG32(OMAP24XX_SMS_BASE+0x10) +#define SSI_SYSCONFIG __REG32(DISP_BASE+0x8010) +//#define VLYNQ_SYSCONFIG 0x67FFFE10 + +/* rkw - good cannidates for PM_ to start what nm was trying */ +#define OMAP24XX_GPT2 (OMAP24XX_L4_IO_BASE+0x2A000) +#define OMAP24XX_GPT3 (OMAP24XX_L4_IO_BASE+0x78000) +#define OMAP24XX_GPT4 (OMAP24XX_L4_IO_BASE+0x7A000) +#define OMAP24XX_GPT5 (OMAP24XX_L4_IO_BASE+0x7C000) +#define OMAP24XX_GPT6 (OMAP24XX_L4_IO_BASE+0x7E000) +#define OMAP24XX_GPT7 (OMAP24XX_L4_IO_BASE+0x80000) +#define OMAP24XX_GPT8 (OMAP24XX_L4_IO_BASE+0x82000) +#define OMAP24XX_GPT9 (OMAP24XX_L4_IO_BASE+0x84000) +#define OMAP24XX_GPT10 (OMAP24XX_L4_IO_BASE+0x86000) +#define OMAP24XX_GPT11 (OMAP24XX_L4_IO_BASE+0x88000) +#define OMAP24XX_GPT12 (OMAP24XX_L4_IO_BASE+0x8A000) + +#define GPTIMER1_SYSCONFIG GPT1_REG32(0x010) +#define GPTIMER2_SYSCONFIG __REG32(OMAP24XX_GPT2 + 0x10) +#define GPTIMER3_SYSCONFIG __REG32(OMAP24XX_GPT3 + 0x10) +#define GPTIMER4_SYSCONFIG __REG32(OMAP24XX_GPT4 + 0x10) +#define GPTIMER5_SYSCONFIG __REG32(OMAP24XX_GPT5 + 0x10) +#define GPTIMER6_SYSCONFIG __REG32(OMAP24XX_GPT6 + 0x10) +#define GPTIMER7_SYSCONFIG __REG32(OMAP24XX_GPT7 + 0x10) +#define GPTIMER8_SYSCONFIG __REG32(OMAP24XX_GPT8 + 0x10) +#define GPTIMER9_SYSCONFIG __REG32(OMAP24XX_GPT9 + 0x10) +#define GPTIMER10_SYSCONFIG __REG32(OMAP24XX_GPT10 + 0x10) +#define GPTIMER11_SYSCONFIG __REG32(OMAP24XX_GPT11 + 0x10) +#define GPTIMER12_SYSCONFIG __REG32(OMAP24XX_GPT12 + 0x10) + +#define GPIOX_BASE(X) (OMAP24XX_GPIO_BASE+(0x2000*((X)-1))) + +#define GPIO1_SYSCONFIG __REG32((GPIOX_BASE(1)+0x10)) +#define GPIO2_SYSCONFIG __REG32((GPIOX_BASE(2)+0x10)) +#define GPIO3_SYSCONFIG __REG32((GPIOX_BASE(3)+0x10)) +#define GPIO4_SYSCONFIG __REG32((GPIOX_BASE(4)+0x10)) + +/* GP TIMER 1 */ +#define GPTIMER1_TISTAT GPT1_REG32(0x014) +#define GPTIMER1_TISR GPT1_REG32(0x018) +#define GPTIMER1_TIER GPT1_REG32(0x01C) +#define GPTIMER1_TWER GPT1_REG32(0x020) +#define GPTIMER1_TCLR GPT1_REG32(0x024) +#define GPTIMER1_TCRR GPT1_REG32(0x028) +#define GPTIMER1_TLDR GPT1_REG32(0x02C) +#define GPTIMER1_TTGR GPT1_REG32(0x030) +#define GPTIMER1_TWPS GPT1_REG32(0x034) +#define GPTIMER1_TMAR GPT1_REG32(0x038) +#define GPTIMER1_TCAR1 GPT1_REG32(0x03C) +#define GPTIMER1_TSICR GPT1_REG32(0x040) +#define GPTIMER1_TCAR2 GPT1_REG32(0x044) + +/* rkw -- base fix up please... */ +#define GPTIMER3_TISR __REG32(OMAP24XX_L4_IO_BASE+0x78018) + +/* SDRC */ +#define SDRC_DLLA_CTRL __REG32(OMAP24XX_SDRC_BASE+0x060) +#define SDRC_DLLA_STATUS __REG32(OMAP24XX_SDRC_BASE+0x064) +#define SDRC_DLLB_CTRL __REG32(OMAP24XX_SDRC_BASE+0x068) +#define SDRC_DLLB_STATUS __REG32(OMAP24XX_SDRC_BASE+0x06C) +#define SDRC_POWER __REG32(OMAP24XX_SDRC_BASE+0x070) +#define SDRC_MR_0 __REG32(OMAP24XX_SDRC_BASE+0x084) + +/* GPIO 1 */ +#define GPIO1_BASE GPIOX_BASE(1) +#define GPIO1_REG32(offset) __REG32(GPIO1_BASE + (offset)) +#define GPIO1_IRQENABLE1 GPIO1_REG32(0x01C) +#define GPIO1_IRQSTATUS1 GPIO1_REG32(0x018) +#define GPIO1_IRQENABLE2 GPIO1_REG32(0x02C) +#define GPIO1_IRQSTATUS2 GPIO1_REG32(0x028) +#define GPIO1_WAKEUPENABLE GPIO1_REG32(0x020) +#define GPIO1_RISINGDETECT GPIO1_REG32(0x048) +#define GPIO1_DATAIN GPIO1_REG32(0x038) +#define GPIO1_OE GPIO1_REG32(0x034) +#define GPIO1_DATAOUT GPIO1_REG32(0x03C) + +/* GPIO2 */ +#define GPIO2_BASE GPIOX_BASE(2) +#define GPIO2_REG32(offset) __REG32(GPIO2_BASE + (offset)) +#define GPIO2_IRQENABLE1 GPIO2_REG32(0x01C) +#define GPIO2_IRQSTATUS1 GPIO2_REG32(0x018) +#define GPIO2_IRQENABLE2 GPIO2_REG32(0x02C) +#define GPIO2_IRQSTATUS2 GPIO2_REG32(0x028) +#define GPIO2_WAKEUPENABLE GPIO2_REG32(0x020) +#define GPIO2_RISINGDETECT GPIO2_REG32(0x048) +#define GPIO2_DATAIN GPIO2_REG32(0x038) +#define GPIO2_OE GPIO2_REG32(0x034) +#define GPIO2_DATAOUT GPIO2_REG32(0x03C) + +/* GPIO 3 */ +#define GPIO3_BASE GPIOX_BASE(3) +#define GPIO3_REG32(offset) __REG32(GPIO3_BASE + (offset)) +#define GPIO3_IRQENABLE1 GPIO3_REG32(0x01C) +#define GPIO3_IRQSTATUS1 GPIO3_REG32(0x018) +#define GPIO3_IRQENABLE2 GPIO3_REG32(0x02C) +#define GPIO3_IRQSTATUS2 GPIO3_REG32(0x028) +#define GPIO3_WAKEUPENABLE GPIO3_REG32(0x020) +#define GPIO3_RISINGDETECT GPIO3_REG32(0x048) +#define GPIO3_FALLINGDETECT GPIO3_REG32(0x04C) +#define GPIO3_DATAIN GPIO3_REG32(0x038) +#define GPIO3_OE GPIO3_REG32(0x034) +#define GPIO3_DATAOUT GPIO3_REG32(0x03C) +#define GPIO3_DEBOUNCENABLE GPIO3_REG32(0x050) +#define GPIO3_DEBOUNCINGTIME GPIO3_REG32(0x054) + +/* GPIO 4 */ +#define GPIO4_BASE GPIOX_BASE(4) +#define GPIO4_REG32(offset) __REG32(GPIO4_BASE + (offset)) +#define GPIO4_IRQENABLE1 GPIO4_REG32(0x01C) +#define GPIO4_IRQSTATUS1 GPIO4_REG32(0x018) +#define GPIO4_IRQENABLE2 GPIO4_REG32(0x02C) +#define GPIO4_IRQSTATUS2 GPIO4_REG32(0x028) +#define GPIO4_WAKEUPENABLE GPIO4_REG32(0x020) +#define GPIO4_RISINGDETECT GPIO4_REG32(0x048) +#define GPIO4_FALLINGDETECT GPIO4_REG32(0x04C) +#define GPIO4_DATAIN GPIO4_REG32(0x038) +#define GPIO4_OE GPIO4_REG32(0x034) +#define GPIO4_DATAOUT GPIO4_REG32(0x03C) +#define GPIO4_DEBOUNCENABLE GPIO4_REG32(0x050) +#define GPIO4_DEBOUNCINGTIME GPIO4_REG32(0x054) + + +/* IO CONFIG */ +#define CONTROL_BASE (OMAP24XX_CTRL_BASE) +#define CONTROL_REG32(offset) __REG32(CONTROL_BASE + (offset)) + +#define CONTROL_PADCONF_SPI1_NCS2 CONTROL_REG32(0x104) +#define CONTROL_PADCONF_SYS_XTALOUT CONTROL_REG32(0x134) +#define CONTROL_PADCONF_UART1_RX CONTROL_REG32(0x0C8) +#define CONTROL_PADCONF_MCBSP1_DX CONTROL_REG32(0x10C) +#define CONTROL_PADCONF_GPMC_NCS4 CONTROL_REG32(0x090) +#define CONTROL_PADCONF_DSS_D5 CONTROL_REG32(0x0B8) +#define CONTROL_PADCONF_DSS_D9 CONTROL_REG32(0x0BC) +#define CONTROL_PADCONF_DSS_D13 CONTROL_REG32(0x0C0) +#define CONTROL_PADCONF_DSS_VSYNC CONTROL_REG32(0x0CC) + +/* CONTROL */ +#define CONTROL_DEVCONF CONTROL_REG32(0x274) + +/* INTERRUPT CONTROLLER */ +#define INTC_BASE (OMAP24XX_L4_IO_BASE+0xfe000) +#define INTC_REG32(offset) __REG32(INTC_BASE + (offset)) + +#define INTC1_U_BASE INTC_REG32(0x000) +#define INTC_MIR0 INTC_REG32(0x084) +#define INTC_MIR_SET0 INTC_REG32(0x08C) +#define INTC_MIR_CLEAR0 INTC_REG32(0x088) +#define INTC_ISR_CLEAR0 INTC_REG32(0x094) +#define INTC_MIR1 INTC_REG32(0x0A4) +#define INTC_MIR_SET1 INTC_REG32(0x0AC) +#define INTC_MIR_CLEAR1 INTC_REG32(0x0A8) +#define INTC_ISR_CLEAR1 INTC_REG32(0x0B4) +#define INTC_MIR2 INTC_REG32(0x0C4) +#define INTC_MIR_SET2 INTC_REG32(0x0CC) +#define INTC_MIR_CLEAR2 INTC_REG32(0x0C8) +#define INTC_ISR_CLEAR2 INTC_REG32(0x0D4) +#define INTC_SIR_IRQ INTC_REG32(0x040) +#define INTC_CONTROL INTC_REG32(0x048) +#define INTC_ILR11 INTC_REG32(0x12C) +#define INTC_ILR32 INTC_REG32(0x180) +#define INTC_ILR37 INTC_REG32(0x194) +#define INTC_SYSCONFIG INTC_REG32(0x010) + +/* RAM FIREWALL */ +#define RAMFW_BASE (0x68005000) +#define RAMFW_REG32(offset) __REG32(RAMFW_BASE + (offset)) + +#define RAMFW_REQINFOPERM0 RAMFW_REG32(0x048) +#define RAMFW_READPERM0 RAMFW_REG32(0x050) +#define RAMFW_WRITEPERM0 RAMFW_REG32(0x058) + +/* GPMC CS1 FPGA ON USER INTERFACE MODULE */ +//#define DEBUG_BOARD_LED_REGISTER 0x04000014 + +/* GPMC CS0 */ +#define GPMC_CONFIG1_0 GPMC_REG32(0x060) +#define GPMC_CONFIG2_0 GPMC_REG32(0x064) +#define GPMC_CONFIG3_0 GPMC_REG32(0x068) +#define GPMC_CONFIG4_0 GPMC_REG32(0x06C) +#define GPMC_CONFIG5_0 GPMC_REG32(0x070) +#define GPMC_CONFIG6_0 GPMC_REG32(0x074) +#define GPMC_CONFIG7_0 GPMC_REG32(0x078) + +/* DSS */ +#define DSS_CONTROL DISP_REG32(0x040) +#define DISPC_CONTROL DISP_REG32(0x440) +#define DISPC_SYSSTATUS DISP_REG32(0x414) +#define DISPC_IRQSTATUS DISP_REG32(0x418) +#define DISPC_IRQENABLE DISP_REG32(0x41C) +#define DISPC_CONFIG DISP_REG32(0x444) +#define DISPC_DEFAULT_COLOR0 DISP_REG32(0x44C) +#define DISPC_DEFAULT_COLOR1 DISP_REG32(0x450) +#define DISPC_TRANS_COLOR0 DISP_REG32(0x454) +#define DISPC_TRANS_COLOR1 DISP_REG32(0x458) +#define DISPC_LINE_NUMBER DISP_REG32(0x460) +#define DISPC_TIMING_H DISP_REG32(0x464) +#define DISPC_TIMING_V DISP_REG32(0x468) +#define DISPC_POL_FREQ DISP_REG32(0x46C) +#define DISPC_DIVISOR DISP_REG32(0x470) +#define DISPC_SIZE_DIG DISP_REG32(0x478) +#define DISPC_SIZE_LCD DISP_REG32(0x47C) +#define DISPC_GFX_BA0 DISP_REG32(0x480) +#define DISPC_GFX_BA1 DISP_REG32(0x484) +#define DISPC_GFX_POSITION DISP_REG32(0x488) +#define DISPC_GFX_SIZE DISP_REG32(0x48C) +#define DISPC_GFX_ATTRIBUTES DISP_REG32(0x4A0) +#define DISPC_GFX_FIFO_THRESHOLD DISP_REG32(0x4A4) +#define DISPC_GFX_ROW_INC DISP_REG32(0x4AC) +#define DISPC_GFX_PIXEL_INC DISP_REG32(0x4B0) +#define DISPC_GFX_WINDOW_SKIP DISP_REG32(0x4B4) +#define DISPC_GFX_TABLE_BA DISP_REG32(0x4B8) +#define DISPC_DATA_CYCLE1 DISP_REG32(0x5D4) +#define DISPC_DATA_CYCLE2 DISP_REG32(0x5D8) +#define DISPC_DATA_CYCLE3 DISP_REG32(0x5DC) + +/* Wake up define for board */ +#define GPIO97 (1 << 1) +#define GPIO88 (1 << 24) + +#endif /* __ASSEMBLER__ */ + +#endif + + + + + diff --git a/arch/arm/mach-omap2/serial.c b/arch/arm/mach-omap2/serial.c new file mode 100644 index 000000000000..f4df04fe1dd8 --- /dev/null +++ b/arch/arm/mach-omap2/serial.c @@ -0,0 +1,180 @@ +/* + * arch/arm/mach-omap/omap2/serial.c + * + * OMAP2 serial support. + * + * Copyright (C) 2005 Nokia Corporation + * Author: Paul Mundt + * + * Based off of arch/arm/mach-omap/omap1/serial.c + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + */ +#include +#include +#include +#include + +#include +#include + +#include +#include + +static struct clk * uart1_ick = NULL; +static struct clk * uart1_fck = NULL; +static struct clk * uart2_ick = NULL; +static struct clk * uart2_fck = NULL; +static struct clk * uart3_ick = NULL; +static struct clk * uart3_fck = NULL; + +static struct plat_serial8250_port serial_platform_data[] = { + { + .membase = (char *)IO_ADDRESS(OMAP_UART1_BASE), + .mapbase = (unsigned long)OMAP_UART1_BASE, + .irq = 72, + .flags = UPF_BOOT_AUTOCONF, + .iotype = UPIO_MEM, + .regshift = 2, + .uartclk = OMAP16XX_BASE_BAUD * 16, + }, { + .membase = (char *)IO_ADDRESS(OMAP_UART2_BASE), + .mapbase = (unsigned long)OMAP_UART2_BASE, + .irq = 73, + .flags = UPF_BOOT_AUTOCONF, + .iotype = UPIO_MEM, + .regshift = 2, + .uartclk = OMAP16XX_BASE_BAUD * 16, + }, { + .membase = (char *)IO_ADDRESS(OMAP_UART3_BASE), + .mapbase = (unsigned long)OMAP_UART3_BASE, + .irq = 74, + .flags = UPF_BOOT_AUTOCONF, + .iotype = UPIO_MEM, + .regshift = 2, + .uartclk = OMAP16XX_BASE_BAUD * 16, + }, { + .flags = 0 + } +}; + +static inline unsigned int serial_read_reg(struct plat_serial8250_port *up, + int offset) +{ + offset <<= up->regshift; + return (unsigned int)__raw_readb(up->membase + offset); +} + +static inline void serial_write_reg(struct plat_serial8250_port *p, int offset, + int value) +{ + offset <<= p->regshift; + __raw_writeb(value, (unsigned long)(p->membase + offset)); +} + +/* + * Internal UARTs need to be initialized for the 8250 autoconfig to work + * properly. Note that the TX watermark initialization may not be needed + * once the 8250.c watermark handling code is merged. + */ +static inline void __init omap_serial_reset(struct plat_serial8250_port *p) +{ + serial_write_reg(p, UART_OMAP_MDR1, 0x07); + serial_write_reg(p, UART_OMAP_SCR, 0x08); + serial_write_reg(p, UART_OMAP_MDR1, 0x00); + serial_write_reg(p, UART_OMAP_SYSC, 0x01); +} + +void __init omap_serial_init() +{ + int i; + const struct omap_uart_config *info; + + /* + * Make sure the serial ports are muxed on at this point. + * You have to mux them off in device drivers later on + * if not needed. + */ + + info = omap_get_config(OMAP_TAG_UART, + struct omap_uart_config); + + if (info == NULL) + return; + + for (i = 0; i < OMAP_MAX_NR_PORTS; i++) { + struct plat_serial8250_port *p = serial_platform_data + i; + + if (!(info->enabled_uarts & (1 << i))) { + p->membase = 0; + p->mapbase = 0; + continue; + } + + switch (i) { + case 0: + uart1_ick = clk_get(NULL, "uart1_ick"); + if (IS_ERR(uart1_ick)) + printk("Could not get uart1_ick\n"); + else { + clk_use(uart1_ick); + } + + uart1_fck = clk_get(NULL, "uart1_fck"); + if (IS_ERR(uart1_fck)) + printk("Could not get uart1_fck\n"); + else { + clk_use(uart1_fck); + } + break; + case 1: + uart2_ick = clk_get(NULL, "uart2_ick"); + if (IS_ERR(uart2_ick)) + printk("Could not get uart2_ick\n"); + else { + clk_use(uart2_ick); + } + + uart2_fck = clk_get(NULL, "uart2_fck"); + if (IS_ERR(uart2_fck)) + printk("Could not get uart2_fck\n"); + else { + clk_use(uart2_fck); + } + break; + case 2: + uart3_ick = clk_get(NULL, "uart3_ick"); + if (IS_ERR(uart3_ick)) + printk("Could not get uart3_ick\n"); + else { + clk_use(uart3_ick); + } + + uart3_fck = clk_get(NULL, "uart3_fck"); + if (IS_ERR(uart3_fck)) + printk("Could not get uart3_fck\n"); + else { + clk_use(uart3_fck); + } + break; + } + + omap_serial_reset(p); + } +} + +static struct platform_device serial_device = { + .name = "serial8250", + .id = 0, + .dev = { + .platform_data = serial_platform_data, + }, +}; + +static int __init omap_init(void) +{ + return platform_device_register(&serial_device); +} +arch_initcall(omap_init); diff --git a/arch/arm/mach-omap2/sram-fn.S b/arch/arm/mach-omap2/sram-fn.S new file mode 100644 index 000000000000..2a869e203342 --- /dev/null +++ b/arch/arm/mach-omap2/sram-fn.S @@ -0,0 +1,333 @@ +/* + * linux/arch/arm/mach-omap1/sram.S + * + * Omap2 specific functions that need to be run in internal SRAM + * + * (C) Copyright 2004 + * Texas Instruments, + * Richard Woodruff + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation; either version 2 of + * the License, or (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR /PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, + * MA 02111-1307 USA + */ +#include +#include +#include +#include +#include + +#include + +#define TIMER_32KSYNCT_CR_V IO_ADDRESS(OMAP24XX_32KSYNCT_BASE + 0x010) + +#define CM_CLKSEL2_PLL_V IO_ADDRESS(OMAP24XX_PRCM_BASE + 0x544) +#define PRCM_VOLTCTRL_V IO_ADDRESS(OMAP24XX_PRCM_BASE + 0x050) +#define PRCM_CLKCFG_CTRL_V IO_ADDRESS(OMAP24XX_PRCM_BASE + 0x080) +#define CM_CLKEN_PLL_V IO_ADDRESS(OMAP24XX_PRCM_BASE + 0x500) +#define CM_IDLEST_CKGEN_V IO_ADDRESS(OMAP24XX_PRCM_BASE + 0x520) +#define CM_CLKSEL1_PLL_V IO_ADDRESS(OMAP24XX_PRCM_BASE + 0x540) + +#define SDRC_DLLA_CTRL_V IO_ADDRESS(OMAP24XX_SDRC_BASE + 0x060) +#define SDRC_RFR_CTRL_V IO_ADDRESS(OMAP24XX_SDRC_BASE + 0x0a4) + + .text + +ENTRY(sram_ddr_init) + stmfd sp!, {r0 - r12, lr} @ save registers on stack + + mov r12, r2 @ capture CS1 vs CS0 + mov r8, r3 @ capture force parameter + + /* frequency shift down */ + ldr r2, cm_clksel2_pll @ get address of dpllout reg + mov r3, #0x1 @ value for 1x operation + str r3, [r2] @ go to L1-freq operation + + /* voltage shift down */ + mov r9, #0x1 @ set up for L1 voltage call + bl voltage_shift @ go drop voltage + + /* dll lock mode */ + ldr r11, sdrc_dlla_ctrl @ addr of dlla ctrl + ldr r10, [r11] @ get current val + cmp r12, #0x1 @ cs1 base (2422 es2.05/1) + addeq r11, r11, #0x8 @ if cs1 base, move to DLLB + mvn r9, #0x4 @ mask to get clear bit2 + and r10, r10, r9 @ clear bit2 for lock mode. + orr r10, r10, #0x8 @ make sure DLL on (es2 bit pos) + orr r10, r10, #0x2 @ 90 degree phase for all below 133Mhz + str r10, [r11] @ commit to DLLA_CTRL + bl i_dll_wait @ wait for dll to lock + + /* get dll value */ + add r11, r11, #0x4 @ get addr of status reg + ldr r10, [r11] @ get locked value + + /* voltage shift up */ + mov r9, #0x0 @ shift back to L0-voltage + bl voltage_shift @ go raise voltage + + /* frequency shift up */ + mov r3, #0x2 @ value for 2x operation + str r3, [r2] @ go to L0-freq operation + + /* reset entry mode for dllctrl */ + sub r11, r11, #0x4 @ move from status to ctrl + cmp r12, #0x1 @ normalize if cs1 based + subeq r11, r11, #0x8 @ possibly back to DLLA + cmp r8, #0x1 @ if forced unlock exit + orreq r1, r1, #0x4 @ make sure exit with unlocked value + str r1, [r11] @ restore DLLA_CTRL high value + add r11, r11, #0x8 @ move to DLLB_CTRL addr + str r1, [r11] @ set value DLLB_CTRL + bl i_dll_wait @ wait for possible lock + + /* set up for return, DDR should be good */ + str r10, [r0] @ write dll_status and return counter + ldmfd sp!, {r0 - r12, pc} @ restore regs and return + + /* ensure the DLL has relocked */ +i_dll_wait: + mov r4, #0x800 @ delay DLL relock, min 0x400 L3 clocks +i_dll_delay: + subs r4, r4, #0x1 + bne i_dll_delay + mov pc, lr + + /* + * shift up or down voltage, use R9 as input to tell level. + * wait for it to finish, use 32k sync counter, 1tick=31uS. + */ +voltage_shift: + ldr r4, prcm_voltctrl @ get addr of volt ctrl. + ldr r5, [r4] @ get value. + ldr r6, prcm_mask_val @ get value of mask + and r5, r5, r6 @ apply mask to clear bits + orr r5, r5, r9 @ bulld value for L0/L1-volt operation. + str r5, [r4] @ set up for change. + mov r3, #0x4000 @ get val for force + orr r5, r5, r3 @ build value for force + str r5, [r4] @ Force transition to L1 + + ldr r3, timer_32ksynct_cr @ get addr of counter + ldr r5, [r3] @ get value + add r5, r5, #0x3 @ give it at most 93uS +volt_delay: + ldr r7, [r3] @ get timer value + cmp r5, r7 @ time up? + bhi volt_delay @ not yet->branch + mov pc, lr @ back to caller. + +/* relative load constants */ +cm_clksel2_pll: + .word CM_CLKSEL2_PLL_V +sdrc_dlla_ctrl: + .word SDRC_DLLA_CTRL_V +prcm_voltctrl: + .word PRCM_VOLTCTRL_V +prcm_mask_val: + .word 0xFFFF3FFC +timer_32ksynct_cr: + .word TIMER_32KSYNCT_CR_V +ENTRY(sram_ddr_init_sz) + .word . - sram_ddr_init + +/* + * Reprograms memory timings. + * r0 = [PRCM_FULL | PRCM_HALF] r1 = SDRC_DLLA_CTRL value r2 = [DDR | SDR] + * PRCM_FULL = 2, PRCM_HALF = 1, DDR = 1, SDR = 0 + */ +ENTRY(sram_reprogram_sdrc) + stmfd sp!, {r0 - r10, lr} @ save registers on stack + mov r3, #0x0 @ clear for mrc call + mcr p15, 0, r3, c7, c10, 4 @ memory barrier, finish ARM SDR/DDR + nop + nop + ldr r6, ddr_sdrc_rfr_ctrl @ get addr of refresh reg + ldr r5, [r6] @ get value + mov r5, r5, lsr #8 @ isolate rfr field and drop burst + + cmp r0, #0x1 @ going to half speed? + movne r9, #0x0 @ if up set flag up for pre up, hi volt + + blne voltage_shift_c @ adjust voltage + + cmp r0, #0x1 @ going to half speed (post branch link) + moveq r5, r5, lsr #1 @ divide by 2 if to half + movne r5, r5, lsl #1 @ mult by 2 if to full + mov r5, r5, lsl #8 @ put rfr field back into place + add r5, r5, #0x1 @ turn on burst of 1 + ldr r4, ddr_cm_clksel2_pll @ get address of out reg + ldr r3, [r4] @ get curr value + orr r3, r3, #0x3 + bic r3, r3, #0x3 @ clear lower bits + orr r3, r3, r0 @ new state value + str r3, [r4] @ set new state (pll/x, x=1 or 2) + nop + nop + + moveq r9, #0x1 @ if speed down, post down, drop volt + bleq voltage_shift_c + + mcr p15, 0, r3, c7, c10, 4 @ memory barrier + str r5, [r6] @ set new RFR_1 value + add r6, r6, #0x30 @ get RFR_2 addr + str r5, [r6] @ set RFR_2 + nop + cmp r2, #0x1 @ (SDR or DDR) do we need to adjust DLL + bne freq_out @ leave if SDR, no DLL function + + /* With DDR, we need to take care of the DLL for the frequency change */ + ldr r2, ddr_sdrc_dlla_ctrl @ addr of dlla ctrl + str r1, [r2] @ write out new SDRC_DLLA_CTRL + add r2, r2, #0x8 @ addr to SDRC_DLLB_CTRL + str r1, [r2] @ commit to SDRC_DLLB_CTRL + mov r1, #0x2000 @ wait DLL relock, min 0x400 L3 clocks +dll_wait: + subs r1, r1, #0x1 + bne dll_wait +freq_out: + ldmfd sp!, {r0 - r10, pc} @ restore regs and return + + /* + * shift up or down voltage, use R9 as input to tell level. + * wait for it to finish, use 32k sync counter, 1tick=31uS. + */ +voltage_shift_c: + ldr r10, ddr_prcm_voltctrl @ get addr of volt ctrl + ldr r8, [r10] @ get value + ldr r7, ddr_prcm_mask_val @ get value of mask + and r8, r8, r7 @ apply mask to clear bits + orr r8, r8, r9 @ bulld value for L0/L1-volt operation. + str r8, [r10] @ set up for change. + mov r7, #0x4000 @ get val for force + orr r8, r8, r7 @ build value for force + str r8, [r10] @ Force transition to L1 + + ldr r10, ddr_timer_32ksynct @ get addr of counter + ldr r8, [r10] @ get value + add r8, r8, #0x2 @ give it at most 62uS (min 31+) +volt_delay_c: + ldr r7, [r10] @ get timer value + cmp r8, r7 @ time up? + bhi volt_delay_c @ not yet->branch + mov pc, lr @ back to caller + +ddr_cm_clksel2_pll: + .word CM_CLKSEL2_PLL_V +ddr_sdrc_dlla_ctrl: + .word SDRC_DLLA_CTRL_V +ddr_sdrc_rfr_ctrl: + .word SDRC_RFR_CTRL_V +ddr_prcm_voltctrl: + .word PRCM_VOLTCTRL_V +ddr_prcm_mask_val: + .word 0xFFFF3FFC +ddr_timer_32ksynct: + .word TIMER_32KSYNCT_CR_V + +ENTRY(sram_reprogram_sdrc_sz) + .word . - sram_reprogram_sdrc + +/* + * Set dividers and pll. Also recalculate DLL value for DDR and unlock mode. + */ +ENTRY(sram_set_prcm) + stmfd sp!, {r0-r12, lr} @ regs to stack + adr r4, pbegin @ addr of preload start + adr r8, pend @ addr of preload end + mcrr p15, 1, r8, r4, c12 @ preload into icache +pbegin: + /* move into fast relock bypass */ + ldr r8, pll_ctl @ get addr + ldr r5, [r8] @ get val + mvn r6, #0x3 @ clear mask + and r5, r5, r6 @ clear field + orr r7, r5, #0x2 @ fast relock val + str r7, [r8] @ go to fast relock + ldr r4, pll_stat @ addr of stat +block: + /* wait for bypass */ + ldr r8, [r4] @ stat value + and r8, r8, #0x3 @ mask for stat + cmp r8, #0x1 @ there yet + bne block @ loop if not + + /* set new dpll dividers _after_ in bypass */ + ldr r4, pll_div @ get addr + str r0, [r4] @ set dpll ctrl val + + ldr r4, set_config @ get addr + mov r8, #1 @ valid cfg msk + str r8, [r4] @ make dividers take + + mov r4, #100 @ dead spin a bit +wait_a_bit: + subs r4, r4, #1 @ dec loop + bne wait_a_bit @ delay done? + + /* check if staying in bypass */ + cmp r2, #0x1 @ stay in bypass? + beq pend @ jump over dpll relock + + /* relock DPLL with new vals */ + ldr r5, pll_stat @ get addr + ldr r4, pll_ctl @ get addr + orr r8, r7, #0x3 @ val for lock dpll + str r8, [r4] @ set val + mov r0, #1000 @ dead spin a bit +wait_more: + subs r0, r0, #1 @ dec loop + bne wait_more @ delay done? +wait_lock: + ldr r8, [r5] @ get lock val + and r8, r8, #3 @ isolate field + cmp r8, #2 @ locked? + bne wait_lock @ wait if not +pend: + /* update memory timings & briefly lock dll */ + ldr r4, sdrc_rfr @ get addr + str r1, [r4] @ update refresh timing + ldr r11, dlla_ctrl @ get addr of DLLA ctrl + ldr r10, [r11] @ get current val + mvn r9, #0x4 @ mask to get clear bit2 + and r10, r10, r9 @ clear bit2 for lock mode + orr r10, r10, #0x8 @ make sure DLL on (es2 bit pos) + str r10, [r11] @ commit to DLLA_CTRL + add r11, r11, #0x8 @ move to dllb + str r10, [r11] @ hit DLLB also + + mov r4, #0x800 @ relock time (min 0x400 L3 clocks) +wait_dll_lock: + subs r4, r4, #0x1 + bne wait_dll_lock + nop + ldmfd sp!, {r0-r12, pc} @ restore regs and return + +set_config: + .word PRCM_CLKCFG_CTRL_V +pll_ctl: + .word CM_CLKEN_PLL_V +pll_stat: + .word CM_IDLEST_CKGEN_V +pll_div: + .word CM_CLKSEL1_PLL_V +sdrc_rfr: + .word SDRC_RFR_CTRL_V +dlla_ctrl: + .word SDRC_DLLA_CTRL_V + +ENTRY(sram_set_prcm_sz) + .word . - sram_set_prcm diff --git a/arch/arm/mach-omap2/timer-gp.c b/arch/arm/mach-omap2/timer-gp.c new file mode 100644 index 000000000000..9ec11443200f --- /dev/null +++ b/arch/arm/mach-omap2/timer-gp.c @@ -0,0 +1,126 @@ +/* + * linux/arch/arm/mach-omap2/timer-gp.c + * + * OMAP2 GP timer support. + * + * Copyright (C) 2005 Nokia Corporation + * Author: Paul Mundt + * Juha Yrjölä + * + * Some parts based off of TI's 24xx code: + * + * Copyright (C) 2004 Texas Instruments, Inc. + * + * Roughly modelled after the OMAP1 MPU timer code. + * + * This file is subject to the terms and conditions of the GNU General Public + * License. See the file "COPYING" in the main directory of this archive + * for more details. + */ +#include +#include +#include +#include +#include +#include +#include +#include + +#define OMAP2_GP_TIMER1_BASE 0x48028000 +#define OMAP2_GP_TIMER2_BASE 0x4802a000 +#define OMAP2_GP_TIMER3_BASE 0x48078000 +#define OMAP2_GP_TIMER4_BASE 0x4807a000 + +#define GP_TIMER_TIDR 0x00 +#define GP_TIMER_TISR 0x18 +#define GP_TIMER_TIER 0x1c +#define GP_TIMER_TCLR 0x24 +#define GP_TIMER_TCRR 0x28 +#define GP_TIMER_TLDR 0x2c +#define GP_TIMER_TSICR 0x40 + +#define OS_TIMER_NR 1 /* GP timer 2 */ + +static unsigned long timer_base[] = { + IO_ADDRESS(OMAP2_GP_TIMER1_BASE), + IO_ADDRESS(OMAP2_GP_TIMER2_BASE), + IO_ADDRESS(OMAP2_GP_TIMER3_BASE), + IO_ADDRESS(OMAP2_GP_TIMER4_BASE), +}; + +static inline unsigned int timer_read_reg(int nr, unsigned int reg) +{ + return __raw_readl(timer_base[nr] + reg); +} + +static inline void timer_write_reg(int nr, unsigned int reg, unsigned int val) +{ + __raw_writel(val, timer_base[nr] + reg); +} + +/* Note that we always enable the clock prescale divider bit */ +static inline void omap2_gp_timer_start(int nr, unsigned long load_val) +{ + unsigned int tmp; + + tmp = 0xffffffff - load_val; + + timer_write_reg(nr, GP_TIMER_TLDR, tmp); + timer_write_reg(nr, GP_TIMER_TCRR, tmp); + timer_write_reg(nr, GP_TIMER_TIER, 1 << 1); + timer_write_reg(nr, GP_TIMER_TCLR, (1 << 5) | (1 << 1) | 1); +} + +static irqreturn_t omap2_gp_timer_interrupt(int irq, void *dev_id, + struct pt_regs *regs) +{ + write_seqlock(&xtime_lock); + + timer_write_reg(OS_TIMER_NR, GP_TIMER_TISR, 1 << 1); + timer_tick(regs); + + write_sequnlock(&xtime_lock); + + return IRQ_HANDLED; +} + +static struct irqaction omap2_gp_timer_irq = { + .name = "gp timer", + .flags = SA_INTERRUPT, + .handler = omap2_gp_timer_interrupt, +}; + +static void __init omap2_gp_timer_init(void) +{ + struct clk * sys_ck; + u32 tick_period = 120000; + u32 l; + + /* Reset clock and prescale value */ + timer_write_reg(OS_TIMER_NR, GP_TIMER_TCLR, 0); + + sys_ck = clk_get(NULL, "sys_ck"); + if (IS_ERR(sys_ck)) + printk(KERN_ERR "Could not get sys_ck\n"); + else { + clk_use(sys_ck); + tick_period = clk_get_rate(sys_ck) / 100; + clk_put(sys_ck); + } + + tick_period /= 2; /* Minimum prescale divider is 2 */ + tick_period -= 1; + + l = timer_read_reg(OS_TIMER_NR, GP_TIMER_TIDR); + printk(KERN_INFO "OMAP2 GP timer (HW version %d.%d)\n", + (l >> 4) & 0x0f, l & 0x0f); + + setup_irq(38, &omap2_gp_timer_irq); + + omap2_gp_timer_start(OS_TIMER_NR, tick_period); +} + +struct sys_timer omap_timer = { + .init = omap2_gp_timer_init, +}; + diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index 4e8d19318d22..e84fdde6edf8 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -242,7 +242,7 @@ config CPU_XSCALE # ARMv6 config CPU_V6 bool "Support ARM V6 processor" - depends on ARCH_INTEGRATOR || MACH_REALVIEW_EB + depends on ARCH_INTEGRATOR || MACH_REALVIEW_EB || ARCH_OMAP2 select CPU_32v6 select CPU_ABRT_EV6 select CPU_CACHE_V6 From 046d6b28ebce92aab2865aa3eb3a0d8ede57e17e Mon Sep 17 00:00:00 2001 From: Tony Lindgren Date: Thu, 10 Nov 2005 14:26:52 +0000 Subject: [PATCH 094/121] [ARM] 3146/1: OMAP 3b/5: Add omap24xx clock framework Patch from Tony Lindgren This patch adds omap24xx specific clock code by Richard Woodruff, Nishant Menon, Tony Lindgren et al. Signed-off-by: Tony Lindgren Signed-off-by: Russell King --- arch/arm/mach-omap2/clock.c | 1129 +++++++++++++++++++ arch/arm/mach-omap2/clock.h | 2103 +++++++++++++++++++++++++++++++++++ 2 files changed, 3232 insertions(+) create mode 100644 arch/arm/mach-omap2/clock.c create mode 100644 arch/arm/mach-omap2/clock.h diff --git a/arch/arm/mach-omap2/clock.c b/arch/arm/mach-omap2/clock.c new file mode 100644 index 000000000000..85818d9f2635 --- /dev/null +++ b/arch/arm/mach-omap2/clock.c @@ -0,0 +1,1129 @@ +/* + * linux/arch/arm/mach-omap2/clock.c + * + * Copyright (C) 2005 Texas Instruments Inc. + * Richard Woodruff + * Created for OMAP2. + * + * Cleaned up and modified to use omap shared clock framework by + * Tony Lindgren + * + * Based on omap1 clock.c, Copyright (C) 2004 - 2005 Nokia corporation + * Written by Tuukka Tikkanen + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include +#include +#include +#include +#include +#include +#include + +#include + +#include +#include +#include +#include + +#include "clock.h" + +//#define DOWN_VARIABLE_DPLL 1 /* Experimental */ + +static struct prcm_config *curr_prcm_set; +static struct memory_timings mem_timings; +static u32 curr_perf_level = PRCM_FULL_SPEED; + +/*------------------------------------------------------------------------- + * Omap2 specific clock functions + *-------------------------------------------------------------------------*/ + +/* Recalculate SYST_CLK */ +static void omap2_sys_clk_recalc(struct clk * clk) +{ + u32 div = PRCM_CLKSRC_CTRL; + div &= (1 << 7) | (1 << 6); /* Test if ext clk divided by 1 or 2 */ + div >>= clk->rate_offset; + clk->rate = (clk->parent->rate / div); + propagate_rate(clk); +} + +static u32 omap2_get_dpll_rate(struct clk * tclk) +{ + int dpll_clk, dpll_mult, dpll_div, amult; + + dpll_mult = (CM_CLKSEL1_PLL >> 12) & 0x03ff; /* 10 bits */ + dpll_div = (CM_CLKSEL1_PLL >> 8) & 0x0f; /* 4 bits */ + dpll_clk = (tclk->parent->rate * dpll_mult) / (dpll_div + 1); + amult = CM_CLKSEL2_PLL & 0x3; + dpll_clk *= amult; + + return dpll_clk; +} + +static void omap2_followparent_recalc(struct clk *clk) +{ + followparent_recalc(clk); +} + +static void omap2_propagate_rate(struct clk * clk) +{ + if (!(clk->flags & RATE_FIXED)) + clk->rate = clk->parent->rate; + + propagate_rate(clk); +} + +/* Enable an APLL if off */ +static void omap2_clk_fixed_enable(struct clk *clk) +{ + u32 cval, i=0; + + if (clk->enable_bit == 0xff) /* Parent will do it */ + return; + + cval = CM_CLKEN_PLL; + + if ((cval & (0x3 << clk->enable_bit)) == (0x3 << clk->enable_bit)) + return; + + cval &= ~(0x3 << clk->enable_bit); + cval |= (0x3 << clk->enable_bit); + CM_CLKEN_PLL = cval; + + if (clk == &apll96_ck) + cval = (1 << 8); + else if (clk == &apll54_ck) + cval = (1 << 6); + + while (!CM_IDLEST_CKGEN & cval) { /* Wait for lock */ + ++i; + udelay(1); + if (i == 100000) + break; + } +} + +/* Enables clock without considering parent dependencies or use count + * REVISIT: Maybe change this to use clk->enable like on omap1? + */ +static int omap2_clk_enable(struct clk * clk) +{ + u32 regval32; + + if (clk->flags & ALWAYS_ENABLED) + return 0; + + if (unlikely(clk->enable_reg == 0)) { + printk(KERN_ERR "clock.c: Enable for %s without enable code\n", + clk->name); + return 0; + } + + if (clk->enable_reg == (void __iomem *)&CM_CLKEN_PLL) { + omap2_clk_fixed_enable(clk); + return 0; + } + + regval32 = __raw_readl(clk->enable_reg); + regval32 |= (1 << clk->enable_bit); + __raw_writel(regval32, clk->enable_reg); + + return 0; +} + +/* Stop APLL */ +static void omap2_clk_fixed_disable(struct clk *clk) +{ + u32 cval; + + if(clk->enable_bit == 0xff) /* let parent off do it */ + return; + + cval = CM_CLKEN_PLL; + cval &= ~(0x3 << clk->enable_bit); + CM_CLKEN_PLL = cval; +} + +/* Disables clock without considering parent dependencies or use count */ +static void omap2_clk_disable(struct clk *clk) +{ + u32 regval32; + + if (clk->enable_reg == 0) + return; + + if (clk->enable_reg == (void __iomem *)&CM_CLKEN_PLL) { + omap2_clk_fixed_disable(clk); + return; + } + + regval32 = __raw_readl(clk->enable_reg); + regval32 &= ~(1 << clk->enable_bit); + __raw_writel(regval32, clk->enable_reg); +} + +static int omap2_clk_use(struct clk *clk) +{ + int ret = 0; + + if (clk->usecount++ == 0) { + if (likely((u32)clk->parent)) + ret = omap2_clk_use(clk->parent); + + if (unlikely(ret != 0)) { + clk->usecount--; + return ret; + } + + ret = omap2_clk_enable(clk); + + if (unlikely(ret != 0) && clk->parent) { + omap2_clk_unuse(clk->parent); + clk->usecount--; + } + } + + return ret; +} + +static void omap2_clk_unuse(struct clk *clk) +{ + if (clk->usecount > 0 && !(--clk->usecount)) { + omap2_clk_disable(clk); + if (likely((u32)clk->parent)) + omap2_clk_unuse(clk->parent); + } +} + +/* + * Uses the current prcm set to tell if a rate is valid. + * You can go slower, but not faster within a given rate set. + */ +static u32 omap2_dpll_round_rate(unsigned long target_rate) +{ + u32 high, low; + + if ((CM_CLKSEL2_PLL & 0x3) == 1) { /* DPLL clockout */ + high = curr_prcm_set->dpll_speed * 2; + low = curr_prcm_set->dpll_speed; + } else { /* DPLL clockout x 2 */ + high = curr_prcm_set->dpll_speed; + low = curr_prcm_set->dpll_speed / 2; + } + +#ifdef DOWN_VARIABLE_DPLL + if (target_rate > high) + return high; + else + return target_rate; +#else + if (target_rate > low) + return high; + else + return low; +#endif + +} + +/* + * Used for clocks that are part of CLKSEL_xyz governed clocks. + * REVISIT: Maybe change to use clk->enable() functions like on omap1? + */ +static void omap2_clksel_recalc(struct clk * clk) +{ + u32 fixed = 0, div = 0; + + if (clk == &dpll_ck) { + clk->rate = omap2_get_dpll_rate(clk); + fixed = 1; + div = 0; + } + + if (clk == &iva1_mpu_int_ifck) { + div = 2; + fixed = 1; + } + + if ((clk == &dss1_fck) && ((CM_CLKSEL1_CORE & (0x1f << 8)) == 0)) { + clk->rate = sys_ck.rate; + return; + } + + if (!fixed) { + div = omap2_clksel_get_divisor(clk); + if (div == 0) + return; + } + + if (div != 0) { + if (unlikely(clk->rate == clk->parent->rate / div)) + return; + clk->rate = clk->parent->rate / div; + } + + if (unlikely(clk->flags & RATE_PROPAGATES)) + propagate_rate(clk); +} + +/* + * Finds best divider value in an array based on the source and target + * rates. The divider array must be sorted with smallest divider first. + */ +static inline u32 omap2_divider_from_table(u32 size, u32 *div_array, + u32 src_rate, u32 tgt_rate) +{ + int i, test_rate; + + if (div_array == NULL) + return ~1; + + for (i=0; i < size; i++) { + test_rate = src_rate / *div_array; + if (test_rate <= tgt_rate) + return *div_array; + ++div_array; + } + + return ~0; /* No acceptable divider */ +} + +/* + * Find divisor for the given clock and target rate. + * + * Note that this will not work for clocks which are part of CONFIG_PARTICIPANT, + * they are only settable as part of virtual_prcm set. + */ +static u32 omap2_clksel_round_rate(struct clk *tclk, u32 target_rate, + u32 *new_div) +{ + u32 gfx_div[] = {2, 3, 4}; + u32 sysclkout_div[] = {1, 2, 4, 8, 16}; + u32 dss1_div[] = {1, 2, 3, 4, 5, 6, 8, 9, 12, 16}; + u32 vylnq_div[] = {1, 2, 3, 4, 6, 8, 9, 12, 16, 18}; + u32 best_div = ~0, asize = 0; + u32 *div_array = NULL; + + switch (tclk->flags & SRC_RATE_SEL_MASK) { + case CM_GFX_SEL1: + asize = 3; + div_array = gfx_div; + break; + case CM_PLL_SEL1: + return omap2_dpll_round_rate(target_rate); + case CM_SYSCLKOUT_SEL1: + asize = 5; + div_array = sysclkout_div; + break; + case CM_CORE_SEL1: + if(tclk == &dss1_fck){ + if(tclk->parent == &core_ck){ + asize = 10; + div_array = dss1_div; + } else { + *new_div = 0; /* fixed clk */ + return(tclk->parent->rate); + } + } else if((tclk == &vlynq_fck) && cpu_is_omap2420()){ + if(tclk->parent == &core_ck){ + asize = 10; + div_array = vylnq_div; + } else { + *new_div = 0; /* fixed clk */ + return(tclk->parent->rate); + } + } + break; + } + + best_div = omap2_divider_from_table(asize, div_array, + tclk->parent->rate, target_rate); + if (best_div == ~0){ + *new_div = 1; + return best_div; /* signal error */ + } + + *new_div = best_div; + return (tclk->parent->rate / best_div); +} + +/* Given a clock and a rate apply a clock specific rounding function */ +static long omap2_clk_round_rate(struct clk *clk, unsigned long rate) +{ + u32 new_div = 0; + int valid_rate; + + if (clk->flags & RATE_FIXED) + return clk->rate; + + if (clk->flags & RATE_CKCTL) { + valid_rate = omap2_clksel_round_rate(clk, rate, &new_div); + return valid_rate; + } + + if (clk->round_rate != 0) + return clk->round_rate(clk, rate); + + return clk->rate; +} + +/* + * Check the DLL lock state, and return tue if running in unlock mode. + * This is needed to compenste for the shifted DLL value in unlock mode. + */ +static u32 omap2_dll_force_needed(void) +{ + u32 dll_state = SDRC_DLLA_CTRL; /* dlla and dllb are a set */ + + if ((dll_state & (1 << 2)) == (1 << 2)) + return 1; + else + return 0; +} + +static void omap2_init_memory_params(u32 force_lock_to_unlock_mode) +{ + unsigned long dll_cnt; + u32 fast_dll = 0; + + mem_timings.m_type = !((SDRC_MR_0 & 0x3) == 0x1); /* DDR = 1, SDR = 0 */ + + /* 2422 es2.05 and beyond has a single SIP DDR instead of 2 like others. + * In the case of 2422, its ok to use CS1 instead of CS0. + */ + +#if 0 /* FIXME: Enable after 24xx cpu detection works */ + ctype = get_cpu_type(); + if (cpu_is_omap2422()) + mem_timings.base_cs = 1; + else +#endif + mem_timings.base_cs = 0; + + if (mem_timings.m_type != M_DDR) + return; + + /* With DDR we need to determine the low frequency DLL value */ + if (((mem_timings.fast_dll_ctrl & (1 << 2)) == M_LOCK_CTRL)) + mem_timings.dll_mode = M_UNLOCK; + else + mem_timings.dll_mode = M_LOCK; + + if (mem_timings.base_cs == 0) { + fast_dll = SDRC_DLLA_CTRL; + dll_cnt = SDRC_DLLA_STATUS & 0xff00; + } else { + fast_dll = SDRC_DLLB_CTRL; + dll_cnt = SDRC_DLLB_STATUS & 0xff00; + } + if (force_lock_to_unlock_mode) { + fast_dll &= ~0xff00; + fast_dll |= dll_cnt; /* Current lock mode */ + } + mem_timings.fast_dll_ctrl = fast_dll; + + /* No disruptions, DDR will be offline & C-ABI not followed */ + omap2_sram_ddr_init(&mem_timings.slow_dll_ctrl, + mem_timings.fast_dll_ctrl, + mem_timings.base_cs, + force_lock_to_unlock_mode); + mem_timings.slow_dll_ctrl &= 0xff00; /* Keep lock value */ + + /* Turn status into unlock ctrl */ + mem_timings.slow_dll_ctrl |= + ((mem_timings.fast_dll_ctrl & 0xF) | (1 << 2)); + + /* 90 degree phase for anything below 133Mhz */ + mem_timings.slow_dll_ctrl |= (1 << 1); +} + +static u32 omap2_reprogram_sdrc(u32 level, u32 force) +{ + u32 prev = curr_perf_level, flags; + + if ((curr_perf_level == level) && !force) + return prev; + + if (level == PRCM_HALF_SPEED) { + local_irq_save(flags); + PRCM_VOLTSETUP = 0xffff; + omap2_sram_reprogram_sdrc(PRCM_HALF_SPEED, + mem_timings.slow_dll_ctrl, + mem_timings.m_type); + curr_perf_level = PRCM_HALF_SPEED; + local_irq_restore(flags); + } + if (level == PRCM_FULL_SPEED) { + local_irq_save(flags); + PRCM_VOLTSETUP = 0xffff; + omap2_sram_reprogram_sdrc(PRCM_FULL_SPEED, + mem_timings.fast_dll_ctrl, + mem_timings.m_type); + curr_perf_level = PRCM_FULL_SPEED; + local_irq_restore(flags); + } + + return prev; +} + +static int omap2_reprogram_dpll(struct clk * clk, unsigned long rate) +{ + u32 flags, cur_rate, low, mult, div, valid_rate, done_rate; + u32 bypass = 0; + struct prcm_config tmpset; + int ret = -EINVAL; + + local_irq_save(flags); + cur_rate = omap2_get_dpll_rate(&dpll_ck); + mult = CM_CLKSEL2_PLL & 0x3; + + if ((rate == (cur_rate / 2)) && (mult == 2)) { + omap2_reprogram_sdrc(PRCM_HALF_SPEED, 1); + } else if ((rate == (cur_rate * 2)) && (mult == 1)) { + omap2_reprogram_sdrc(PRCM_FULL_SPEED, 1); + } else if (rate != cur_rate) { + valid_rate = omap2_dpll_round_rate(rate); + if (valid_rate != rate) + goto dpll_exit; + + if ((CM_CLKSEL2_PLL & 0x3) == 1) + low = curr_prcm_set->dpll_speed; + else + low = curr_prcm_set->dpll_speed / 2; + + tmpset.cm_clksel1_pll = CM_CLKSEL1_PLL; + tmpset.cm_clksel1_pll &= ~(0x3FFF << 8); + div = ((curr_prcm_set->xtal_speed / 1000000) - 1); + tmpset.cm_clksel2_pll = CM_CLKSEL2_PLL; + tmpset.cm_clksel2_pll &= ~0x3; + if (rate > low) { + tmpset.cm_clksel2_pll |= 0x2; + mult = ((rate / 2) / 1000000); + done_rate = PRCM_FULL_SPEED; + } else { + tmpset.cm_clksel2_pll |= 0x1; + mult = (rate / 1000000); + done_rate = PRCM_HALF_SPEED; + } + tmpset.cm_clksel1_pll |= ((div << 8) | (mult << 12)); + + /* Worst case */ + tmpset.base_sdrc_rfr = V24XX_SDRC_RFR_CTRL_BYPASS; + + if (rate == curr_prcm_set->xtal_speed) /* If asking for 1-1 */ + bypass = 1; + + omap2_reprogram_sdrc(PRCM_FULL_SPEED, 1); /* For init_mem */ + + /* Force dll lock mode */ + omap2_set_prcm(tmpset.cm_clksel1_pll, tmpset.base_sdrc_rfr, + bypass); + + /* Errata: ret dll entry state */ + omap2_init_memory_params(omap2_dll_force_needed()); + omap2_reprogram_sdrc(done_rate, 0); + } + omap2_clksel_recalc(&dpll_ck); + ret = 0; + +dpll_exit: + local_irq_restore(flags); + return(ret); +} + +/* Just return the MPU speed */ +static void omap2_mpu_recalc(struct clk * clk) +{ + clk->rate = curr_prcm_set->mpu_speed; +} + +/* + * Look for a rate equal or less than the target rate given a configuration set. + * + * What's not entirely clear is "which" field represents the key field. + * Some might argue L3-DDR, others ARM, others IVA. This code is simple and + * just uses the ARM rates. + */ +static long omap2_round_to_table_rate(struct clk * clk, unsigned long rate) +{ + struct prcm_config * ptr; + long highest_rate; + + if (clk != &virt_prcm_set) + return -EINVAL; + + highest_rate = -EINVAL; + + for (ptr = rate_table; ptr->mpu_speed; ptr++) { + if (ptr->xtal_speed != sys_ck.rate) + continue; + + highest_rate = ptr->mpu_speed; + + /* Can check only after xtal frequency check */ + if (ptr->mpu_speed <= rate) + break; + } + return highest_rate; +} + +/* + * omap2_convert_field_to_div() - turn field value into integer divider + */ +static u32 omap2_clksel_to_divisor(u32 div_sel, u32 field_val) +{ + u32 i; + u32 clkout_array[] = {1, 2, 4, 8, 16}; + + if ((div_sel & SRC_RATE_SEL_MASK) == CM_SYSCLKOUT_SEL1) { + for (i = 0; i < 5; i++) { + if (field_val == i) + return clkout_array[i]; + } + return ~0; + } else + return field_val; +} + +/* + * Returns the CLKSEL divider register value + * REVISIT: This should be cleaned up to work nicely with void __iomem * + */ +static u32 omap2_get_clksel(u32 *div_sel, u32 *field_mask, + struct clk *clk) +{ + int ret = ~0; + u32 reg_val, div_off; + u32 div_addr = 0; + u32 mask = ~0; + + div_off = clk->rate_offset; + + switch ((*div_sel & SRC_RATE_SEL_MASK)) { + case CM_MPU_SEL1: + div_addr = (u32)&CM_CLKSEL_MPU; + mask = 0x1f; + break; + case CM_DSP_SEL1: + div_addr = (u32)&CM_CLKSEL_DSP; + if (cpu_is_omap2420()) { + if ((div_off == 0) || (div_off == 8)) + mask = 0x1f; + else if (div_off == 5) + mask = 0x3; + } else if (cpu_is_omap2430()) { + if (div_off == 0) + mask = 0x1f; + else if (div_off == 5) + mask = 0x3; + } + break; + case CM_GFX_SEL1: + div_addr = (u32)&CM_CLKSEL_GFX; + if (div_off == 0) + mask = 0x7; + break; + case CM_MODEM_SEL1: + div_addr = (u32)&CM_CLKSEL_MDM; + if (div_off == 0) + mask = 0xf; + break; + case CM_SYSCLKOUT_SEL1: + div_addr = (u32)&PRCM_CLKOUT_CTRL; + if ((div_off == 3) || (div_off = 11)) + mask= 0x3; + break; + case CM_CORE_SEL1: + div_addr = (u32)&CM_CLKSEL1_CORE; + switch (div_off) { + case 0: /* l3 */ + case 8: /* dss1 */ + case 15: /* vylnc-2420 */ + case 20: /* ssi */ + mask = 0x1f; break; + case 5: /* l4 */ + mask = 0x3; break; + case 13: /* dss2 */ + mask = 0x1; break; + case 25: /* usb */ + mask = 0xf; break; + } + } + + *field_mask = mask; + + if (unlikely(mask == ~0)) + div_addr = 0; + + *div_sel = div_addr; + + if (unlikely(div_addr == 0)) + return ret; + + /* Isolate field */ + reg_val = __raw_readl((void __iomem *)div_addr) & (mask << div_off); + + /* Normalize back to divider value */ + reg_val >>= div_off; + + return reg_val; +} + +/* + * Return divider to be applied to parent clock. + * Return 0 on error. + */ +static u32 omap2_clksel_get_divisor(struct clk *clk) +{ + int ret = 0; + u32 div, div_sel, div_off, field_mask, field_val; + + /* isolate control register */ + div_sel = (SRC_RATE_SEL_MASK & clk->flags); + + div_off = clk->rate_offset; + field_val = omap2_get_clksel(&div_sel, &field_mask, clk); + if (div_sel == 0) + return ret; + + div_sel = (SRC_RATE_SEL_MASK & clk->flags); + div = omap2_clksel_to_divisor(div_sel, field_val); + + return div; +} + +/* Set the clock rate for a clock source */ +static int omap2_clk_set_rate(struct clk *clk, unsigned long rate) + +{ + int ret = -EINVAL; + void __iomem * reg; + u32 div_sel, div_off, field_mask, field_val, reg_val, validrate; + u32 new_div = 0; + + if (!(clk->flags & CONFIG_PARTICIPANT) && (clk->flags & RATE_CKCTL)) { + if (clk == &dpll_ck) + return omap2_reprogram_dpll(clk, rate); + + /* Isolate control register */ + div_sel = (SRC_RATE_SEL_MASK & clk->flags); + div_off = clk->src_offset; + + validrate = omap2_clksel_round_rate(clk, rate, &new_div); + if(validrate != rate) + return(ret); + + field_val = omap2_get_clksel(&div_sel, &field_mask, clk); + if (div_sel == 0) + return ret; + + if(clk->flags & CM_SYSCLKOUT_SEL1){ + switch(new_div){ + case 16: field_val = 4; break; + case 8: field_val = 3; break; + case 4: field_val = 2; break; + case 2: field_val = 1; break; + case 1: field_val = 0; break; + } + } + else + field_val = new_div; + + reg = (void __iomem *)div_sel; + + reg_val = __raw_readl(reg); + reg_val &= ~(field_mask << div_off); + reg_val |= (field_val << div_off); + + __raw_writel(reg_val, reg); + clk->rate = clk->parent->rate / field_val; + + if (clk->flags & DELAYED_APP) + __raw_writel(0x1, (void __iomem *)&PRCM_CLKCFG_CTRL); + ret = 0; + } else if (clk->set_rate != 0) + ret = clk->set_rate(clk, rate); + + if (unlikely(ret == 0 && (clk->flags & RATE_PROPAGATES))) + propagate_rate(clk); + + return ret; +} + +/* Converts encoded control register address into a full address */ +static u32 omap2_get_src_field(u32 *type_to_addr, u32 reg_offset, + struct clk *src_clk, u32 *field_mask) +{ + u32 val = ~0, src_reg_addr = 0, mask = 0; + + /* Find target control register.*/ + switch ((*type_to_addr & SRC_RATE_SEL_MASK)) { + case CM_CORE_SEL1: + src_reg_addr = (u32)&CM_CLKSEL1_CORE; + if (reg_offset == 13) { /* DSS2_fclk */ + mask = 0x1; + if (src_clk == &sys_ck) + val = 0; + if (src_clk == &func_48m_ck) + val = 1; + } else if (reg_offset == 8) { /* DSS1_fclk */ + mask = 0x1f; + if (src_clk == &sys_ck) + val = 0; + else if (src_clk == &core_ck) /* divided clock */ + val = 0x10; /* rate needs fixing */ + } else if ((reg_offset == 15) && cpu_is_omap2420()){ /*vlnyq*/ + mask = 0x1F; + if(src_clk == &func_96m_ck) + val = 0; + else if (src_clk == &core_ck) + val = 0x10; + } + break; + case CM_CORE_SEL2: + src_reg_addr = (u32)&CM_CLKSEL2_CORE; + mask = 0x3; + if (src_clk == &func_32k_ck) + val = 0x0; + if (src_clk == &sys_ck) + val = 0x1; + if (src_clk == &alt_ck) + val = 0x2; + break; + case CM_WKUP_SEL1: + src_reg_addr = (u32)&CM_CLKSEL2_CORE; + mask = 0x3; + if (src_clk == &func_32k_ck) + val = 0x0; + if (src_clk == &sys_ck) + val = 0x1; + if (src_clk == &alt_ck) + val = 0x2; + break; + case CM_PLL_SEL1: + src_reg_addr = (u32)&CM_CLKSEL1_PLL; + mask = 0x1; + if (reg_offset == 0x3) { + if (src_clk == &apll96_ck) + val = 0; + if (src_clk == &alt_ck) + val = 1; + } + else if (reg_offset == 0x5) { + if (src_clk == &apll54_ck) + val = 0; + if (src_clk == &alt_ck) + val = 1; + } + break; + case CM_PLL_SEL2: + src_reg_addr = (u32)&CM_CLKSEL2_PLL; + mask = 0x3; + if (src_clk == &func_32k_ck) + val = 0x0; + if (src_clk == &dpll_ck) + val = 0x2; + break; + case CM_SYSCLKOUT_SEL1: + src_reg_addr = (u32)&PRCM_CLKOUT_CTRL; + mask = 0x3; + if (src_clk == &dpll_ck) + val = 0; + if (src_clk == &sys_ck) + val = 1; + if (src_clk == &func_54m_ck) + val = 2; + if (src_clk == &func_96m_ck) + val = 3; + break; + } + + if (val == ~0) /* Catch errors in offset */ + *type_to_addr = 0; + else + *type_to_addr = src_reg_addr; + *field_mask = mask; + + return val; +} + +static int omap2_clk_set_parent(struct clk *clk, struct clk *new_parent) +{ + void __iomem * reg; + u32 src_sel, src_off, field_val, field_mask, reg_val, rate; + int ret = -EINVAL; + + if (unlikely(clk->flags & CONFIG_PARTICIPANT)) + return ret; + + if (clk->flags & SRC_SEL_MASK) { /* On-chip SEL collection */ + src_sel = (SRC_RATE_SEL_MASK & clk->flags); + src_off = clk->src_offset; + + if (src_sel == 0) + goto set_parent_error; + + field_val = omap2_get_src_field(&src_sel, src_off, new_parent, + &field_mask); + + reg = (void __iomem *)src_sel; + + if (clk->usecount > 0) + omap2_clk_disable(clk); + + /* Set new source value (previous dividers if any in effect) */ + reg_val = __raw_readl(reg) & ~(field_mask << src_off); + reg_val |= (field_val << src_off); + __raw_writel(reg_val, reg); + + if (clk->flags & DELAYED_APP) + __raw_writel(0x1, (void __iomem *)&PRCM_CLKCFG_CTRL); + + if (clk->usecount > 0) + omap2_clk_enable(clk); + + clk->parent = new_parent; + + /* SRC_RATE_SEL_MASK clocks follow their parents rates.*/ + if ((new_parent == &core_ck) && (clk == &dss1_fck)) + clk->rate = new_parent->rate / 0x10; + else + clk->rate = new_parent->rate; + + if (unlikely(clk->flags & RATE_PROPAGATES)) + propagate_rate(clk); + + return 0; + } else { + clk->parent = new_parent; + rate = new_parent->rate; + omap2_clk_set_rate(clk, rate); + ret = 0; + } + + set_parent_error: + return ret; +} + +/* Sets basic clocks based on the specified rate */ +static int omap2_select_table_rate(struct clk * clk, unsigned long rate) +{ + u32 flags, cur_rate, done_rate, bypass = 0; + u8 cpu_mask = 0; + struct prcm_config *prcm; + unsigned long found_speed = 0; + + if (clk != &virt_prcm_set) + return -EINVAL; + + /* FIXME: Change cpu_is_omap2420() to cpu_is_omap242x() */ + if (cpu_is_omap2420()) + cpu_mask = RATE_IN_242X; + else if (cpu_is_omap2430()) + cpu_mask = RATE_IN_243X; + + for (prcm = rate_table; prcm->mpu_speed; prcm++) { + if (!(prcm->flags & cpu_mask)) + continue; + + if (prcm->xtal_speed != sys_ck.rate) + continue; + + if (prcm->mpu_speed <= rate) { + found_speed = prcm->mpu_speed; + break; + } + } + + if (!found_speed) { + printk(KERN_INFO "Could not set MPU rate to %luMHz\n", + rate / 1000000); + return -EINVAL; + } + + curr_prcm_set = prcm; + cur_rate = omap2_get_dpll_rate(&dpll_ck); + + if (prcm->dpll_speed == cur_rate / 2) { + omap2_reprogram_sdrc(PRCM_HALF_SPEED, 1); + } else if (prcm->dpll_speed == cur_rate * 2) { + omap2_reprogram_sdrc(PRCM_FULL_SPEED, 1); + } else if (prcm->dpll_speed != cur_rate) { + local_irq_save(flags); + + if (prcm->dpll_speed == prcm->xtal_speed) + bypass = 1; + + if ((prcm->cm_clksel2_pll & 0x3) == 2) + done_rate = PRCM_FULL_SPEED; + else + done_rate = PRCM_HALF_SPEED; + + /* MPU divider */ + CM_CLKSEL_MPU = prcm->cm_clksel_mpu; + + /* dsp + iva1 div(2420), iva2.1(2430) */ + CM_CLKSEL_DSP = prcm->cm_clksel_dsp; + + CM_CLKSEL_GFX = prcm->cm_clksel_gfx; + + /* Major subsystem dividers */ + CM_CLKSEL1_CORE = prcm->cm_clksel1_core; + if (cpu_is_omap2430()) + CM_CLKSEL_MDM = prcm->cm_clksel_mdm; + + /* x2 to enter init_mem */ + omap2_reprogram_sdrc(PRCM_FULL_SPEED, 1); + + omap2_set_prcm(prcm->cm_clksel1_pll, prcm->base_sdrc_rfr, + bypass); + + omap2_init_memory_params(omap2_dll_force_needed()); + omap2_reprogram_sdrc(done_rate, 0); + + local_irq_restore(flags); + } + omap2_clksel_recalc(&dpll_ck); + + return 0; +} + +/*------------------------------------------------------------------------- + * Omap2 clock reset and init functions + *-------------------------------------------------------------------------*/ + +static struct clk_functions omap2_clk_functions = { + .clk_enable = omap2_clk_enable, + .clk_disable = omap2_clk_disable, + .clk_use = omap2_clk_use, + .clk_unuse = omap2_clk_unuse, + .clk_round_rate = omap2_clk_round_rate, + .clk_set_rate = omap2_clk_set_rate, + .clk_set_parent = omap2_clk_set_parent, +}; + +static void __init omap2_get_crystal_rate(struct clk *osc, struct clk *sys) +{ + u32 div, aplls, sclk = 13000000; + + aplls = CM_CLKSEL1_PLL; + aplls &= ((1 << 23) | (1 << 24) | (1 << 25)); + aplls >>= 23; /* Isolate field, 0,2,3 */ + + if (aplls == 0) + sclk = 19200000; + else if (aplls == 2) + sclk = 13000000; + else if (aplls == 3) + sclk = 12000000; + + div = PRCM_CLKSRC_CTRL; + div &= ((1 << 7) | (1 << 6)); + div >>= sys->rate_offset; + + osc->rate = sclk * div; + sys->rate = sclk; +} + +#ifdef CONFIG_OMAP_RESET_CLOCKS +static void __init omap2_disable_unused_clocks(void) +{ + struct clk *ck; + u32 regval32; + + list_for_each_entry(ck, &clocks, node) { + if (ck->usecount > 0 || (ck->flags & ALWAYS_ENABLED) || + ck->enable_reg == 0) + continue; + + regval32 = __raw_readl(ck->enable_reg); + if ((regval32 & (1 << ck->enable_bit)) == 0) + continue; + + printk(KERN_INFO "Disabling unused clock \"%s\"\n", ck->name); + omap2_clk_disable(ck); + } +} +late_initcall(omap2_disable_unused_clocks); +#endif + +/* + * Switch the MPU rate if specified on cmdline. + * We cannot do this early until cmdline is parsed. + */ +static int __init omap2_clk_arch_init(void) +{ + if (!mpurate) + return -EINVAL; + + if (omap2_select_table_rate(&virt_prcm_set, mpurate)) + printk(KERN_ERR "Could not find matching MPU rate\n"); + + propagate_rate(&osc_ck); /* update main root fast */ + propagate_rate(&func_32k_ck); /* update main root slow */ + + printk(KERN_INFO "Switched to new clocking rate (Crystal/DPLL/MPU): " + "%ld.%01ld/%ld/%ld MHz\n", + (sys_ck.rate / 1000000), (sys_ck.rate / 100000) % 10, + (dpll_ck.rate / 1000000), (mpu_ck.rate / 1000000)) ; + + return 0; +} +arch_initcall(omap2_clk_arch_init); + +int __init omap2_clk_init(void) +{ + struct prcm_config *prcm; + struct clk ** clkp; + u32 clkrate; + + clk_init(&omap2_clk_functions); + omap2_get_crystal_rate(&osc_ck, &sys_ck); + + for (clkp = onchip_clks; clkp < onchip_clks + ARRAY_SIZE(onchip_clks); + clkp++) { + + if ((*clkp)->flags & CLOCK_IN_OMAP242X && cpu_is_omap2420()) { + clk_register(*clkp); + continue; + } + + if ((*clkp)->flags & CLOCK_IN_OMAP243X && cpu_is_omap2430()) { + clk_register(*clkp); + continue; + } + } + + /* Check the MPU rate set by bootloader */ + clkrate = omap2_get_dpll_rate(&dpll_ck); + for (prcm = rate_table; prcm->mpu_speed; prcm++) { + if (prcm->xtal_speed != sys_ck.rate) + continue; + if (prcm->dpll_speed <= clkrate) + break; + } + curr_prcm_set = prcm; + + propagate_rate(&osc_ck); /* update main root fast */ + propagate_rate(&func_32k_ck); /* update main root slow */ + + printk(KERN_INFO "Clocking rate (Crystal/DPLL/MPU): " + "%ld.%01ld/%ld/%ld MHz\n", + (sys_ck.rate / 1000000), (sys_ck.rate / 100000) % 10, + (dpll_ck.rate / 1000000), (mpu_ck.rate / 1000000)) ; + + /* + * Only enable those clocks we will need, let the drivers + * enable other clocks as necessary + */ + clk_use(&sync_32k_ick); + clk_use(&omapctrl_ick); + if (cpu_is_omap2430()) + clk_use(&sdrc_ick); + + return 0; +} diff --git a/arch/arm/mach-omap2/clock.h b/arch/arm/mach-omap2/clock.h new file mode 100644 index 000000000000..4aeab5591bd3 --- /dev/null +++ b/arch/arm/mach-omap2/clock.h @@ -0,0 +1,2103 @@ +/* + * linux/arch/arm/mach-omap24xx/clock.h + * + * Copyright (C) 2005 Texas Instruments Inc. + * Richard Woodruff + * Created for OMAP2. + * + * Copyright (C) 2004 Nokia corporation + * Written by Tuukka Tikkanen + * Based on clocks.h by Tony Lindgren, Gordon McNutt and RidgeRun, Inc + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __ARCH_ARM_MACH_OMAP2_CLOCK_H +#define __ARCH_ARM_MACH_OMAP2_CLOCK_H + +static void omap2_sys_clk_recalc(struct clk * clk); +static void omap2_clksel_recalc(struct clk * clk); +static void omap2_followparent_recalc(struct clk * clk); +static void omap2_propagate_rate(struct clk * clk); +static void omap2_mpu_recalc(struct clk * clk); +static int omap2_select_table_rate(struct clk * clk, unsigned long rate); +static long omap2_round_to_table_rate(struct clk * clk, unsigned long rate); +static void omap2_clk_unuse(struct clk *clk); +static void omap2_sys_clk_recalc(struct clk * clk); +static u32 omap2_clksel_to_divisor(u32 div_sel, u32 field_val); +static u32 omap2_clksel_get_divisor(struct clk *clk); + + +#define RATE_IN_242X (1 << 0) +#define RATE_IN_243X (1 << 1) + +/* Memory timings */ +#define M_DDR 1 +#define M_LOCK_CTRL (1 << 2) +#define M_UNLOCK 0 +#define M_LOCK 1 + +struct memory_timings { + u32 m_type; /* ddr = 1, sdr = 0 */ + u32 dll_mode; /* use lock mode = 1, unlock mode = 0 */ + u32 slow_dll_ctrl; /* unlock mode, dll value for slow speed */ + u32 fast_dll_ctrl; /* unlock mode, dll value for fast speed */ + u32 base_cs; /* base chip select to use for calculations */ +}; + +/* Key dividers which make up a PRCM set. Ratio's for a PRCM are mandated. + * xtal_speed, dpll_speed, mpu_speed, CM_CLKSEL_MPU,CM_CLKSEL_DSP + * CM_CLKSEL_GFX, CM_CLKSEL1_CORE, CM_CLKSEL1_PLL CM_CLKSEL2_PLL, CM_CLKSEL_MDM + */ +struct prcm_config { + unsigned long xtal_speed; /* crystal rate */ + unsigned long dpll_speed; /* dpll: out*xtal*M/(N-1)table_recalc */ + unsigned long mpu_speed; /* speed of MPU */ + unsigned long cm_clksel_mpu; /* mpu divider */ + unsigned long cm_clksel_dsp; /* dsp+iva1 div(2420), iva2.1(2430) */ + unsigned long cm_clksel_gfx; /* gfx dividers */ + unsigned long cm_clksel1_core; /* major subsystem dividers */ + unsigned long cm_clksel1_pll; /* m,n */ + unsigned long cm_clksel2_pll; /* dpllx1 or x2 out */ + unsigned long cm_clksel_mdm; /* modem dividers 2430 only */ + unsigned long base_sdrc_rfr; /* base refresh timing for a set */ + unsigned char flags; +}; + +/* Mask for clksel which support parent settign in set_rate */ +#define SRC_SEL_MASK (CM_CORE_SEL1 | CM_CORE_SEL2 | CM_WKUP_SEL1 | \ + CM_PLL_SEL1 | CM_PLL_SEL2 | CM_SYSCLKOUT_SEL1) + +/* Mask for clksel regs which support rate operations */ +#define SRC_RATE_SEL_MASK (CM_MPU_SEL1 | CM_DSP_SEL1 | CM_GFX_SEL1 | \ + CM_MODEM_SEL1 | CM_CORE_SEL1 | CM_CORE_SEL2 | \ + CM_WKUP_SEL1 | CM_PLL_SEL1 | CM_PLL_SEL2 | \ + CM_SYSCLKOUT_SEL1) + +/* + * The OMAP2 processor can be run at several discrete 'PRCM configurations'. + * These configurations are characterized by voltage and speed for clocks. + * The device is only validated for certain combinations. One way to express + * these combinations is via the 'ratio's' which the clocks operate with + * respect to each other. These ratio sets are for a given voltage/DPLL + * setting. All configurations can be described by a DPLL setting and a ratio + * There are 3 ratio sets for the 2430 and X ratio sets for 2420. + * + * 2430 differs from 2420 in that there are no more phase synchronizers used. + * They both have a slightly different clock domain setup. 2420(iva1,dsp) vs + * 2430 (iva2.1, NOdsp, mdm) + */ + +/* Core fields for cm_clksel, not ratio governed */ +#define RX_CLKSEL_DSS1 (0x10 << 8) +#define RX_CLKSEL_DSS2 (0x0 << 13) +#define RX_CLKSEL_SSI (0x5 << 20) + +/*------------------------------------------------------------------------- + * Voltage/DPLL ratios + *-------------------------------------------------------------------------*/ + +/* 2430 Ratio's, 2430-Ratio Config 1 */ +#define R1_CLKSEL_L3 (4 << 0) +#define R1_CLKSEL_L4 (2 << 5) +#define R1_CLKSEL_USB (4 << 25) +#define R1_CM_CLKSEL1_CORE_VAL R1_CLKSEL_USB | RX_CLKSEL_SSI | \ + RX_CLKSEL_DSS2 | RX_CLKSEL_DSS1 | \ + R1_CLKSEL_L4 | R1_CLKSEL_L3 +#define R1_CLKSEL_MPU (2 << 0) +#define R1_CM_CLKSEL_MPU_VAL R1_CLKSEL_MPU +#define R1_CLKSEL_DSP (2 << 0) +#define R1_CLKSEL_DSP_IF (2 << 5) +#define R1_CM_CLKSEL_DSP_VAL R1_CLKSEL_DSP | R1_CLKSEL_DSP_IF +#define R1_CLKSEL_GFX (2 << 0) +#define R1_CM_CLKSEL_GFX_VAL R1_CLKSEL_GFX +#define R1_CLKSEL_MDM (4 << 0) +#define R1_CM_CLKSEL_MDM_VAL R1_CLKSEL_MDM + +/* 2430-Ratio Config 2 */ +#define R2_CLKSEL_L3 (6 << 0) +#define R2_CLKSEL_L4 (2 << 5) +#define R2_CLKSEL_USB (2 << 25) +#define R2_CM_CLKSEL1_CORE_VAL R2_CLKSEL_USB | RX_CLKSEL_SSI | \ + RX_CLKSEL_DSS2 | RX_CLKSEL_DSS1 | \ + R2_CLKSEL_L4 | R2_CLKSEL_L3 +#define R2_CLKSEL_MPU (2 << 0) +#define R2_CM_CLKSEL_MPU_VAL R2_CLKSEL_MPU +#define R2_CLKSEL_DSP (2 << 0) +#define R2_CLKSEL_DSP_IF (3 << 5) +#define R2_CM_CLKSEL_DSP_VAL R2_CLKSEL_DSP | R2_CLKSEL_DSP_IF +#define R2_CLKSEL_GFX (2 << 0) +#define R2_CM_CLKSEL_GFX_VAL R2_CLKSEL_GFX +#define R2_CLKSEL_MDM (6 << 0) +#define R2_CM_CLKSEL_MDM_VAL R2_CLKSEL_MDM + +/* 2430-Ratio Bootm (BYPASS) */ +#define RB_CLKSEL_L3 (1 << 0) +#define RB_CLKSEL_L4 (1 << 5) +#define RB_CLKSEL_USB (1 << 25) +#define RB_CM_CLKSEL1_CORE_VAL RB_CLKSEL_USB | RX_CLKSEL_SSI | \ + RX_CLKSEL_DSS2 | RX_CLKSEL_DSS1 | \ + RB_CLKSEL_L4 | RB_CLKSEL_L3 +#define RB_CLKSEL_MPU (1 << 0) +#define RB_CM_CLKSEL_MPU_VAL RB_CLKSEL_MPU +#define RB_CLKSEL_DSP (1 << 0) +#define RB_CLKSEL_DSP_IF (1 << 5) +#define RB_CM_CLKSEL_DSP_VAL RB_CLKSEL_DSP | RB_CLKSEL_DSP_IF +#define RB_CLKSEL_GFX (1 << 0) +#define RB_CM_CLKSEL_GFX_VAL RB_CLKSEL_GFX +#define RB_CLKSEL_MDM (1 << 0) +#define RB_CM_CLKSEL_MDM_VAL RB_CLKSEL_MDM + +/* 2420 Ratio Equivalents */ +#define RXX_CLKSEL_VLYNQ (0x12 << 15) +#define RXX_CLKSEL_SSI (0x8 << 20) + +/* 2420-PRCM III 532MHz core */ +#define RIII_CLKSEL_L3 (4 << 0) /* 133MHz */ +#define RIII_CLKSEL_L4 (2 << 5) /* 66.5MHz */ +#define RIII_CLKSEL_USB (4 << 25) /* 33.25MHz */ +#define RIII_CM_CLKSEL1_CORE_VAL RIII_CLKSEL_USB | RXX_CLKSEL_SSI | \ + RXX_CLKSEL_VLYNQ | RX_CLKSEL_DSS2 | \ + RX_CLKSEL_DSS1 | RIII_CLKSEL_L4 | \ + RIII_CLKSEL_L3 +#define RIII_CLKSEL_MPU (2 << 0) /* 266MHz */ +#define RIII_CM_CLKSEL_MPU_VAL RIII_CLKSEL_MPU +#define RIII_CLKSEL_DSP (3 << 0) /* c5x - 177.3MHz */ +#define RIII_CLKSEL_DSP_IF (2 << 5) /* c5x - 88.67MHz */ +#define RIII_SYNC_DSP (1 << 7) /* Enable sync */ +#define RIII_CLKSEL_IVA (6 << 8) /* iva1 - 88.67MHz */ +#define RIII_SYNC_IVA (1 << 13) /* Enable sync */ +#define RIII_CM_CLKSEL_DSP_VAL RIII_SYNC_IVA | RIII_CLKSEL_IVA | \ + RIII_SYNC_DSP | RIII_CLKSEL_DSP_IF | \ + RIII_CLKSEL_DSP +#define RIII_CLKSEL_GFX (2 << 0) /* 66.5MHz */ +#define RIII_CM_CLKSEL_GFX_VAL RIII_CLKSEL_GFX + +/* 2420-PRCM II 600MHz core */ +#define RII_CLKSEL_L3 (6 << 0) /* 100MHz */ +#define RII_CLKSEL_L4 (2 << 5) /* 50MHz */ +#define RII_CLKSEL_USB (2 << 25) /* 50MHz */ +#define RII_CM_CLKSEL1_CORE_VAL RII_CLKSEL_USB | \ + RXX_CLKSEL_SSI | RXX_CLKSEL_VLYNQ | \ + RX_CLKSEL_DSS2 | RX_CLKSEL_DSS1 | \ + RII_CLKSEL_L4 | RII_CLKSEL_L3 +#define RII_CLKSEL_MPU (2 << 0) /* 300MHz */ +#define RII_CM_CLKSEL_MPU_VAL RII_CLKSEL_MPU +#define RII_CLKSEL_DSP (3 << 0) /* c5x - 200MHz */ +#define RII_CLKSEL_DSP_IF (2 << 5) /* c5x - 100MHz */ +#define RII_SYNC_DSP (0 << 7) /* Bypass sync */ +#define RII_CLKSEL_IVA (6 << 8) /* iva1 - 200MHz */ +#define RII_SYNC_IVA (0 << 13) /* Bypass sync */ +#define RII_CM_CLKSEL_DSP_VAL RII_SYNC_IVA | RII_CLKSEL_IVA | \ + RII_SYNC_DSP | RII_CLKSEL_DSP_IF | \ + RII_CLKSEL_DSP +#define RII_CLKSEL_GFX (2 << 0) /* 50MHz */ +#define RII_CM_CLKSEL_GFX_VAL RII_CLKSEL_GFX + +/* 2420-PRCM VII (boot) */ +#define RVII_CLKSEL_L3 (1 << 0) +#define RVII_CLKSEL_L4 (1 << 5) +#define RVII_CLKSEL_DSS1 (1 << 8) +#define RVII_CLKSEL_DSS2 (0 << 13) +#define RVII_CLKSEL_VLYNQ (1 << 15) +#define RVII_CLKSEL_SSI (1 << 20) +#define RVII_CLKSEL_USB (1 << 25) + +#define RVII_CM_CLKSEL1_CORE_VAL RVII_CLKSEL_USB | RVII_CLKSEL_SSI | \ + RVII_CLKSEL_VLYNQ | RVII_CLKSEL_DSS2 | \ + RVII_CLKSEL_DSS1 | RVII_CLKSEL_L4 | RVII_CLKSEL_L3 + +#define RVII_CLKSEL_MPU (1 << 0) /* all divide by 1 */ +#define RVII_CM_CLKSEL_MPU_VAL RVII_CLKSEL_MPU + +#define RVII_CLKSEL_DSP (1 << 0) +#define RVII_CLKSEL_DSP_IF (1 << 5) +#define RVII_SYNC_DSP (0 << 7) +#define RVII_CLKSEL_IVA (1 << 8) +#define RVII_SYNC_IVA (0 << 13) +#define RVII_CM_CLKSEL_DSP_VAL RVII_SYNC_IVA | RVII_CLKSEL_IVA | RVII_SYNC_DSP | \ + RVII_CLKSEL_DSP_IF | RVII_CLKSEL_DSP + +#define RVII_CLKSEL_GFX (1 << 0) +#define RVII_CM_CLKSEL_GFX_VAL RVII_CLKSEL_GFX + +/*------------------------------------------------------------------------- + * 2430 Target modes: Along with each configuration the CPU has several + * modes which goes along with them. Modes mainly are the addition of + * describe DPLL combinations to go along with a ratio. + *-------------------------------------------------------------------------*/ + +/* Hardware governed */ +#define MX_48M_SRC (0 << 3) +#define MX_54M_SRC (0 << 5) +#define MX_APLLS_CLIKIN_12 (3 << 23) +#define MX_APLLS_CLIKIN_13 (2 << 23) +#define MX_APLLS_CLIKIN_19_2 (0 << 23) + +/* + * 2430 - standalone, 2*ref*M/(n+1), M/N is for exactness not relock speed + * #2 (ratio1) baseport-target + * #5a (ratio1) baseport-target, target DPLL = 266*2 = 532MHz + */ +#define M5A_DPLL_MULT_12 (133 << 12) +#define M5A_DPLL_DIV_12 (5 << 8) +#define M5A_CM_CLKSEL1_PLL_12_VAL MX_48M_SRC | MX_54M_SRC | \ + M5A_DPLL_DIV_12 | M5A_DPLL_MULT_12 | \ + MX_APLLS_CLIKIN_12 +#define M5A_DPLL_MULT_13 (266 << 12) +#define M5A_DPLL_DIV_13 (12 << 8) +#define M5A_CM_CLKSEL1_PLL_13_VAL MX_48M_SRC | MX_54M_SRC | \ + M5A_DPLL_DIV_13 | M5A_DPLL_MULT_13 | \ + MX_APLLS_CLIKIN_13 +#define M5A_DPLL_MULT_19 (180 << 12) +#define M5A_DPLL_DIV_19 (12 << 8) +#define M5A_CM_CLKSEL1_PLL_19_VAL MX_48M_SRC | MX_54M_SRC | \ + M5A_DPLL_DIV_19 | M5A_DPLL_MULT_19 | \ + MX_APLLS_CLIKIN_19_2 +/* #5b (ratio1) target DPLL = 200*2 = 400MHz */ +#define M5B_DPLL_MULT_12 (50 << 12) +#define M5B_DPLL_DIV_12 (2 << 8) +#define M5B_CM_CLKSEL1_PLL_12_VAL MX_48M_SRC | MX_54M_SRC | \ + M5B_DPLL_DIV_12 | M5B_DPLL_MULT_12 | \ + MX_APLLS_CLIKIN_12 +#define M5B_DPLL_MULT_13 (200 << 12) +#define M5B_DPLL_DIV_13 (12 << 8) + +#define M5B_CM_CLKSEL1_PLL_13_VAL MX_48M_SRC | MX_54M_SRC | \ + M5B_DPLL_DIV_13 | M5B_DPLL_MULT_13 | \ + MX_APLLS_CLIKIN_13 +#define M5B_DPLL_MULT_19 (125 << 12) +#define M5B_DPLL_DIV_19 (31 << 8) +#define M5B_CM_CLKSEL1_PLL_19_VAL MX_48M_SRC | MX_54M_SRC | \ + M5B_DPLL_DIV_19 | M5B_DPLL_MULT_19 | \ + MX_APLLS_CLIKIN_19_2 +/* + * #4 (ratio2) + * #3 (ratio2) baseport-target, target DPLL = 330*2 = 660MHz + */ +#define M3_DPLL_MULT_12 (55 << 12) +#define M3_DPLL_DIV_12 (1 << 8) +#define M3_CM_CLKSEL1_PLL_12_VAL MX_48M_SRC | MX_54M_SRC | \ + M3_DPLL_DIV_12 | M3_DPLL_MULT_12 | \ + MX_APLLS_CLIKIN_12 +#define M3_DPLL_MULT_13 (330 << 12) +#define M3_DPLL_DIV_13 (12 << 8) +#define M3_CM_CLKSEL1_PLL_13_VAL MX_48M_SRC | MX_54M_SRC | \ + M3_DPLL_DIV_13 | M3_DPLL_MULT_13 | \ + MX_APLLS_CLIKIN_13 +#define M3_DPLL_MULT_19 (275 << 12) +#define M3_DPLL_DIV_19 (15 << 8) +#define M3_CM_CLKSEL1_PLL_19_VAL MX_48M_SRC | MX_54M_SRC | \ + M3_DPLL_DIV_19 | M3_DPLL_MULT_19 | \ + MX_APLLS_CLIKIN_19_2 +/* boot (boot) */ +#define MB_DPLL_MULT (1 << 12) +#define MB_DPLL_DIV (0 << 8) +#define MB_CM_CLKSEL1_PLL_12_VAL MX_48M_SRC | MX_54M_SRC | MB_DPLL_DIV |\ + MB_DPLL_MULT | MX_APLLS_CLIKIN_12 + +#define MB_CM_CLKSEL1_PLL_13_VAL MX_48M_SRC | MX_54M_SRC | MB_DPLL_DIV |\ + MB_DPLL_MULT | MX_APLLS_CLIKIN_13 + +#define MB_CM_CLKSEL1_PLL_19_VAL MX_48M_SRC | MX_54M_SRC | MB_DPLL_DIV |\ + MB_DPLL_MULT | MX_APLLS_CLIKIN_19 + +/* + * 2430 - chassis (sedna) + * 165 (ratio1) same as above #2 + * 150 (ratio1) + * 133 (ratio2) same as above #4 + * 110 (ratio2) same as above #3 + * 104 (ratio2) + * boot (boot) + */ + +/* + * 2420 Equivalent - mode registers + * PRCM II , target DPLL = 2*300MHz = 600MHz + */ +#define MII_DPLL_MULT_12 (50 << 12) +#define MII_DPLL_DIV_12 (1 << 8) +#define MII_CM_CLKSEL1_PLL_12_VAL MX_48M_SRC | MX_54M_SRC | \ + MII_DPLL_DIV_12 | MII_DPLL_MULT_12 | \ + MX_APLLS_CLIKIN_12 +#define MII_DPLL_MULT_13 (300 << 12) +#define MII_DPLL_DIV_13 (12 << 8) +#define MII_CM_CLKSEL1_PLL_13_VAL MX_48M_SRC | MX_54M_SRC | \ + MII_DPLL_DIV_13 | MII_DPLL_MULT_13 | \ + MX_APLLS_CLIKIN_13 + +/* PRCM III target DPLL = 2*266 = 532MHz*/ +#define MIII_DPLL_MULT_12 (133 << 12) +#define MIII_DPLL_DIV_12 (5 << 8) +#define MIII_CM_CLKSEL1_PLL_12_VAL MX_48M_SRC | MX_54M_SRC | \ + MIII_DPLL_DIV_12 | MIII_DPLL_MULT_12 | \ + MX_APLLS_CLIKIN_12 +#define MIII_DPLL_MULT_13 (266 << 12) +#define MIII_DPLL_DIV_13 (12 << 8) +#define MIII_CM_CLKSEL1_PLL_13_VAL MX_48M_SRC | MX_54M_SRC | \ + MIII_DPLL_DIV_13 | MIII_DPLL_MULT_13 | \ + MX_APLLS_CLIKIN_13 + +/* PRCM VII (boot bypass) */ +#define MVII_CM_CLKSEL1_PLL_12_VAL MB_CM_CLKSEL1_PLL_12_VAL +#define MVII_CM_CLKSEL1_PLL_13_VAL MB_CM_CLKSEL1_PLL_13_VAL + +/* High and low operation value */ +#define MX_CLKSEL2_PLL_2x_VAL (2 << 0) +#define MX_CLKSEL2_PLL_1x_VAL (1 << 0) + +/* + * These represent optimal values for common parts, it won't work for all. + * As long as you scale down, most parameters are still work, they just + * become sub-optimal. The RFR value goes in the oppisite direction. If you + * don't adjust it down as your clock period increases the refresh interval + * will not be met. Setting all parameters for complete worst case may work, + * but may cut memory performance by 2x. Due to errata the DLLs need to be + * unlocked and their value needs run time calibration. A dynamic call is + * need for that as no single right value exists acorss production samples. + * + * Only the FULL speed values are given. Current code is such that rate + * changes must be made at DPLLoutx2. The actual value adjustment for low + * frequency operation will be handled by omap_set_performance() + * + * By having the boot loader boot up in the fastest L4 speed available likely + * will result in something which you can switch between. + */ +#define V24XX_SDRC_RFR_CTRL_133MHz (0x0003de00 | 1) +#define V24XX_SDRC_RFR_CTRL_100MHz (0x0002da01 | 1) +#define V24XX_SDRC_RFR_CTRL_110MHz (0x0002da01 | 1) /* Need to calc */ +#define V24XX_SDRC_RFR_CTRL_BYPASS (0x00005000 | 1) /* Need to calc */ + +/* MPU speed defines */ +#define S12M 12000000 +#define S13M 13000000 +#define S19M 19200000 +#define S26M 26000000 +#define S100M 100000000 +#define S133M 133000000 +#define S150M 150000000 +#define S165M 165000000 +#define S200M 200000000 +#define S266M 266000000 +#define S300M 300000000 +#define S330M 330000000 +#define S400M 400000000 +#define S532M 532000000 +#define S600M 600000000 +#define S660M 660000000 + +/*------------------------------------------------------------------------- + * Key dividers which make up a PRCM set. Ratio's for a PRCM are mandated. + * xtal_speed, dpll_speed, mpu_speed, CM_CLKSEL_MPU, + * CM_CLKSEL_DSP, CM_CLKSEL_GFX, CM_CLKSEL1_CORE, CM_CLKSEL1_PLL, + * CM_CLKSEL2_PLL, CM_CLKSEL_MDM + * + * Filling in table based on H4 boards and 2430-SDPs variants available. + * There are quite a few more rates combinations which could be defined. + * + * When multiple values are defiend the start up will try and choose the + * fastest one. If a 'fast' value is defined, then automatically, the /2 + * one should be included as it can be used. Generally having more that + * one fast set does not make sense, as static timings need to be changed + * to change the set. The exception is the bypass setting which is + * availble for low power bypass. + * + * Note: This table needs to be sorted, fastest to slowest. + *-------------------------------------------------------------------------*/ +static struct prcm_config rate_table[] = { + /* PRCM II - FAST */ + {S12M, S600M, S300M, RII_CM_CLKSEL_MPU_VAL, /* 300MHz ARM */ + RII_CM_CLKSEL_DSP_VAL, RII_CM_CLKSEL_GFX_VAL, + RII_CM_CLKSEL1_CORE_VAL, MII_CM_CLKSEL1_PLL_12_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_100MHz, + RATE_IN_242X}, + + {S13M, S600M, S300M, RII_CM_CLKSEL_MPU_VAL, /* 300MHz ARM */ + RII_CM_CLKSEL_DSP_VAL, RII_CM_CLKSEL_GFX_VAL, + RII_CM_CLKSEL1_CORE_VAL, MII_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_100MHz, + RATE_IN_242X}, + + /* PRCM III - FAST */ + {S12M, S532M, S266M, RIII_CM_CLKSEL_MPU_VAL, /* 266MHz ARM */ + RIII_CM_CLKSEL_DSP_VAL, RIII_CM_CLKSEL_GFX_VAL, + RIII_CM_CLKSEL1_CORE_VAL, MIII_CM_CLKSEL1_PLL_12_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_133MHz, + RATE_IN_242X}, + + {S13M, S532M, S266M, RIII_CM_CLKSEL_MPU_VAL, /* 266MHz ARM */ + RIII_CM_CLKSEL_DSP_VAL, RIII_CM_CLKSEL_GFX_VAL, + RIII_CM_CLKSEL1_CORE_VAL, MIII_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_133MHz, + RATE_IN_242X}, + + /* PRCM II - SLOW */ + {S12M, S300M, S150M, RII_CM_CLKSEL_MPU_VAL, /* 150MHz ARM */ + RII_CM_CLKSEL_DSP_VAL, RII_CM_CLKSEL_GFX_VAL, + RII_CM_CLKSEL1_CORE_VAL, MII_CM_CLKSEL1_PLL_12_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_100MHz, + RATE_IN_242X}, + + {S13M, S300M, S150M, RII_CM_CLKSEL_MPU_VAL, /* 150MHz ARM */ + RII_CM_CLKSEL_DSP_VAL, RII_CM_CLKSEL_GFX_VAL, + RII_CM_CLKSEL1_CORE_VAL, MII_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_100MHz, + RATE_IN_242X}, + + /* PRCM III - SLOW */ + {S12M, S266M, S133M, RIII_CM_CLKSEL_MPU_VAL, /* 133MHz ARM */ + RIII_CM_CLKSEL_DSP_VAL, RIII_CM_CLKSEL_GFX_VAL, + RIII_CM_CLKSEL1_CORE_VAL, MIII_CM_CLKSEL1_PLL_12_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_133MHz, + RATE_IN_242X}, + + {S13M, S266M, S133M, RIII_CM_CLKSEL_MPU_VAL, /* 133MHz ARM */ + RIII_CM_CLKSEL_DSP_VAL, RIII_CM_CLKSEL_GFX_VAL, + RIII_CM_CLKSEL1_CORE_VAL, MIII_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_133MHz, + RATE_IN_242X}, + + /* PRCM-VII (boot-bypass) */ + {S12M, S12M, S12M, RVII_CM_CLKSEL_MPU_VAL, /* 12MHz ARM*/ + RVII_CM_CLKSEL_DSP_VAL, RVII_CM_CLKSEL_GFX_VAL, + RVII_CM_CLKSEL1_CORE_VAL, MVII_CM_CLKSEL1_PLL_12_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_BYPASS, + RATE_IN_242X}, + + /* PRCM-VII (boot-bypass) */ + {S13M, S13M, S13M, RVII_CM_CLKSEL_MPU_VAL, /* 13MHz ARM */ + RVII_CM_CLKSEL_DSP_VAL, RVII_CM_CLKSEL_GFX_VAL, + RVII_CM_CLKSEL1_CORE_VAL, MVII_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, 0, V24XX_SDRC_RFR_CTRL_BYPASS, + RATE_IN_242X}, + + /* PRCM #3 - ratio2 (ES2) - FAST */ + {S13M, S660M, S330M, R2_CM_CLKSEL_MPU_VAL, /* 330MHz ARM */ + R2_CM_CLKSEL_DSP_VAL, R2_CM_CLKSEL_GFX_VAL, + R2_CM_CLKSEL1_CORE_VAL, M3_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, R2_CM_CLKSEL_MDM_VAL, + V24XX_SDRC_RFR_CTRL_110MHz, + RATE_IN_243X}, + + /* PRCM #5a - ratio1 - FAST */ + {S13M, S532M, S266M, R1_CM_CLKSEL_MPU_VAL, /* 266MHz ARM */ + R1_CM_CLKSEL_DSP_VAL, R1_CM_CLKSEL_GFX_VAL, + R1_CM_CLKSEL1_CORE_VAL, M5A_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, R1_CM_CLKSEL_MDM_VAL, + V24XX_SDRC_RFR_CTRL_133MHz, + RATE_IN_243X}, + + /* PRCM #5b - ratio1 - FAST */ + {S13M, S400M, S200M, R1_CM_CLKSEL_MPU_VAL, /* 200MHz ARM */ + R1_CM_CLKSEL_DSP_VAL, R1_CM_CLKSEL_GFX_VAL, + R1_CM_CLKSEL1_CORE_VAL, M5B_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, R1_CM_CLKSEL_MDM_VAL, + V24XX_SDRC_RFR_CTRL_100MHz, + RATE_IN_243X}, + + /* PRCM #3 - ratio2 (ES2) - SLOW */ + {S13M, S330M, S165M, R2_CM_CLKSEL_MPU_VAL, /* 165MHz ARM */ + R2_CM_CLKSEL_DSP_VAL, R2_CM_CLKSEL_GFX_VAL, + R2_CM_CLKSEL1_CORE_VAL, M3_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_1x_VAL, R2_CM_CLKSEL_MDM_VAL, + V24XX_SDRC_RFR_CTRL_110MHz, + RATE_IN_243X}, + + /* PRCM #5a - ratio1 - SLOW */ + {S13M, S266M, S133M, R1_CM_CLKSEL_MPU_VAL, /* 133MHz ARM */ + R1_CM_CLKSEL_DSP_VAL, R1_CM_CLKSEL_GFX_VAL, + R1_CM_CLKSEL1_CORE_VAL, M5A_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_1x_VAL, R1_CM_CLKSEL_MDM_VAL, + V24XX_SDRC_RFR_CTRL_133MHz, + RATE_IN_243X}, + + /* PRCM #5b - ratio1 - SLOW*/ + {S13M, S200M, S100M, R1_CM_CLKSEL_MPU_VAL, /* 100MHz ARM */ + R1_CM_CLKSEL_DSP_VAL, R1_CM_CLKSEL_GFX_VAL, + R1_CM_CLKSEL1_CORE_VAL, M5B_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_1x_VAL, R1_CM_CLKSEL_MDM_VAL, + V24XX_SDRC_RFR_CTRL_100MHz, + RATE_IN_243X}, + + /* PRCM-boot/bypass */ + {S13M, S13M, S13M, RB_CM_CLKSEL_MPU_VAL, /* 13Mhz */ + RB_CM_CLKSEL_DSP_VAL, RB_CM_CLKSEL_GFX_VAL, + RB_CM_CLKSEL1_CORE_VAL, MB_CM_CLKSEL1_PLL_13_VAL, + MX_CLKSEL2_PLL_2x_VAL, RB_CM_CLKSEL_MDM_VAL, + V24XX_SDRC_RFR_CTRL_BYPASS, + RATE_IN_243X}, + + /* PRCM-boot/bypass */ + {S12M, S12M, S12M, RB_CM_CLKSEL_MPU_VAL, /* 12Mhz */ + RB_CM_CLKSEL_DSP_VAL, RB_CM_CLKSEL_GFX_VAL, + RB_CM_CLKSEL1_CORE_VAL, MB_CM_CLKSEL1_PLL_12_VAL, + MX_CLKSEL2_PLL_2x_VAL, RB_CM_CLKSEL_MDM_VAL, + V24XX_SDRC_RFR_CTRL_BYPASS, + RATE_IN_243X}, + + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}, +}; + +/*------------------------------------------------------------------------- + * 24xx clock tree. + * + * NOTE:In many cases here we are assigning a 'default' parent. In many + * cases the parent is selectable. The get/set parent calls will also + * switch sources. + * + * Many some clocks say always_enabled, but they can be auto idled for + * power savings. They will always be available upon clock request. + * + * Several sources are given initial rates which may be wrong, this will + * be fixed up in the init func. + * + * Things are broadly separated below by clock domains. It is + * noteworthy that most periferals have dependencies on multiple clock + * domains. Many get their interface clocks from the L4 domain, but get + * functional clocks from fixed sources or other core domain derived + * clocks. + *-------------------------------------------------------------------------*/ + +/* Base external input clocks */ +static struct clk func_32k_ck = { + .name = "func_32k_ck", + .rate = 32000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | ALWAYS_ENABLED, +}; + +/* Typical 12/13MHz in standalone mode, will be 26Mhz in chassis mode */ +static struct clk osc_ck = { /* (*12, *13, 19.2, *26, 38.4)MHz */ + .name = "osc_ck", + .rate = 26000000, /* fixed up in clock init */ + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | ALWAYS_ENABLED | RATE_PROPAGATES, +}; + +/* With out modem likely 12MHz, with modem likely 13MHz */ +static struct clk sys_ck = { /* (*12, *13, 19.2, 26, 38.4)MHz */ + .name = "sys_ck", /* ~ ref_clk also */ + .parent = &osc_ck, + .rate = 13000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | ALWAYS_ENABLED | RATE_PROPAGATES, + .rate_offset = 6, /* sysclkdiv 1 or 2, already handled or no boot */ + .recalc = &omap2_sys_clk_recalc, +}; + +static struct clk alt_ck = { /* Typical 54M or 48M, may not exist */ + .name = "alt_ck", + .rate = 54000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | ALWAYS_ENABLED | RATE_PROPAGATES, + .recalc = &omap2_propagate_rate, +}; + +/* + * Analog domain root source clocks + */ + +/* dpll_ck, is broken out in to special cases through clksel */ +static struct clk dpll_ck = { + .name = "dpll_ck", + .parent = &sys_ck, /* Can be func_32k also */ + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_PROPAGATES | RATE_CKCTL | CM_PLL_SEL1, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk apll96_ck = { + .name = "apll96_ck", + .parent = &sys_ck, + .rate = 96000000, + .flags = CLOCK_IN_OMAP242X |CLOCK_IN_OMAP243X | + RATE_FIXED | RATE_PROPAGATES, + .enable_reg = (void __iomem *)&CM_CLKEN_PLL, + .enable_bit = 0x2, + .recalc = &omap2_propagate_rate, +}; + +static struct clk apll54_ck = { + .name = "apll54_ck", + .parent = &sys_ck, + .rate = 54000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | RATE_PROPAGATES, + .enable_reg = (void __iomem *)&CM_CLKEN_PLL, + .enable_bit = 0x6, + .recalc = &omap2_propagate_rate, +}; + +/* + * PRCM digital base sources + */ +static struct clk func_54m_ck = { + .name = "func_54m_ck", + .parent = &apll54_ck, /* can also be alt_clk */ + .rate = 54000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | CM_PLL_SEL1 | RATE_PROPAGATES, + .src_offset = 5, + .enable_reg = (void __iomem *)&CM_CLKEN_PLL, + .enable_bit = 0xff, + .recalc = &omap2_propagate_rate, +}; + +static struct clk core_ck = { + .name = "core_ck", + .parent = &dpll_ck, /* can also be 32k */ + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + ALWAYS_ENABLED | RATE_PROPAGATES, + .recalc = &omap2_propagate_rate, +}; + +static struct clk sleep_ck = { /* sys_clk or 32k */ + .name = "sleep_ck", + .parent = &func_32k_ck, + .rate = 32000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .recalc = &omap2_propagate_rate, +}; + +static struct clk func_96m_ck = { + .name = "func_96m_ck", + .parent = &apll96_ck, + .rate = 96000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | RATE_PROPAGATES, + .enable_reg = (void __iomem *)&CM_CLKEN_PLL, + .enable_bit = 0xff, + .recalc = &omap2_propagate_rate, +}; + +static struct clk func_48m_ck = { + .name = "func_48m_ck", + .parent = &apll96_ck, /* 96M or Alt */ + .rate = 48000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | CM_PLL_SEL1 | RATE_PROPAGATES, + .src_offset = 3, + .enable_reg = (void __iomem *)&CM_CLKEN_PLL, + .enable_bit = 0xff, + .recalc = &omap2_propagate_rate, +}; + +static struct clk func_12m_ck = { + .name = "func_12m_ck", + .parent = &func_48m_ck, + .rate = 12000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | RATE_PROPAGATES, + .recalc = &omap2_propagate_rate, + .enable_reg = (void __iomem *)&CM_CLKEN_PLL, + .enable_bit = 0xff, +}; + +/* Secure timer, only available in secure mode */ +static struct clk wdt1_osc_ck = { + .name = "ck_wdt1_osc", + .parent = &osc_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk sys_clkout = { + .name = "sys_clkout", + .parent = &func_54m_ck, + .rate = 54000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_SYSCLKOUT_SEL1 | RATE_CKCTL, + .src_offset = 0, + .enable_reg = (void __iomem *)&PRCM_CLKOUT_CTRL, + .enable_bit = 7, + .rate_offset = 3, + .recalc = &omap2_clksel_recalc, +}; + +/* In 2430, new in 2420 ES2 */ +static struct clk sys_clkout2 = { + .name = "sys_clkout2", + .parent = &func_54m_ck, + .rate = 54000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_SYSCLKOUT_SEL1 | RATE_CKCTL, + .src_offset = 8, + .enable_reg = (void __iomem *)&PRCM_CLKOUT_CTRL, + .enable_bit = 15, + .rate_offset = 11, + .recalc = &omap2_clksel_recalc, +}; + +/* + * MPU clock domain + * Clocks: + * MPU_FCLK, MPU_ICLK + * INT_M_FCLK, INT_M_I_CLK + * + * - Individual clocks are hardware managed. + * - Base divider comes from: CM_CLKSEL_MPU + * + */ +static struct clk mpu_ck = { /* Control cpu */ + .name = "mpu_ck", + .parent = &core_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | RATE_CKCTL | + ALWAYS_ENABLED | CM_MPU_SEL1 | DELAYED_APP | + CONFIG_PARTICIPANT | RATE_PROPAGATES, + .rate_offset = 0, /* bits 0-4 */ + .recalc = &omap2_clksel_recalc, +}; + +/* + * DSP (2430-IVA2.1) (2420-UMA+IVA1) clock domain + * Clocks: + * 2430: IVA2.1_FCLK, IVA2.1_ICLK + * 2420: UMA_FCLK, UMA_ICLK, IVA_MPU, IVA_COP + */ +static struct clk iva2_1_fck = { + .name = "iva2_1_fck", + .parent = &core_ck, + .flags = CLOCK_IN_OMAP243X | RATE_CKCTL | CM_DSP_SEL1 | + DELAYED_APP | RATE_PROPAGATES | + CONFIG_PARTICIPANT, + .rate_offset = 0, + .enable_reg = (void __iomem *)&CM_FCLKEN_DSP, + .enable_bit = 0, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk iva2_1_ick = { + .name = "iva2_1_ick", + .parent = &iva2_1_fck, + .flags = CLOCK_IN_OMAP243X | RATE_CKCTL | CM_DSP_SEL1 | + DELAYED_APP | CONFIG_PARTICIPANT, + .rate_offset = 5, + .recalc = &omap2_clksel_recalc, +}; + +/* + * Won't be too specific here. The core clock comes into this block + * it is divided then tee'ed. One branch goes directly to xyz enable + * controls. The other branch gets further divided by 2 then possibly + * routed into a synchronizer and out of clocks abc. + */ +static struct clk dsp_fck = { + .name = "dsp_fck", + .parent = &core_ck, + .flags = CLOCK_IN_OMAP242X | RATE_CKCTL | CM_DSP_SEL1 | + DELAYED_APP | CONFIG_PARTICIPANT | RATE_PROPAGATES, + .rate_offset = 0, + .enable_reg = (void __iomem *)&CM_FCLKEN_DSP, + .enable_bit = 0, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk dsp_ick = { + .name = "dsp_ick", /* apparently ipi and isp */ + .parent = &dsp_fck, + .flags = CLOCK_IN_OMAP242X | RATE_CKCTL | CM_DSP_SEL1 | + DELAYED_APP | CONFIG_PARTICIPANT, + .rate_offset = 5, + .enable_reg = (void __iomem *)&CM_ICLKEN_DSP, + .enable_bit = 1, /* for ipi */ + .recalc = &omap2_clksel_recalc, +}; + +static struct clk iva1_ifck = { + .name = "iva1_ifck", + .parent = &core_ck, + .flags = CLOCK_IN_OMAP242X | CM_DSP_SEL1 | RATE_CKCTL | + CONFIG_PARTICIPANT | RATE_PROPAGATES | DELAYED_APP, + .rate_offset= 8, + .enable_reg = (void __iomem *)&CM_FCLKEN_DSP, + .enable_bit = 10, + .recalc = &omap2_clksel_recalc, +}; + +/* IVA1 mpu/int/i/f clocks are /2 of parent */ +static struct clk iva1_mpu_int_ifck = { + .name = "iva1_mpu_int_ifck", + .parent = &iva1_ifck, + .flags = CLOCK_IN_OMAP242X | RATE_CKCTL | CM_DSP_SEL1, + .enable_reg = (void __iomem *)&CM_FCLKEN_DSP, + .enable_bit = 8, + .recalc = &omap2_clksel_recalc, +}; + +/* + * L3 clock domain + * L3 clocks are used for both interface and functional clocks to + * multiple entities. Some of these clocks are completely managed + * by hardware, and some others allow software control. Hardware + * managed ones general are based on directly CLK_REQ signals and + * various auto idle settings. The functional spec sets many of these + * as 'tie-high' for their enables. + * + * I-CLOCKS: + * L3-Interconnect, SMS, GPMC, SDRC, OCM_RAM, OCM_ROM, SDMA + * CAM, HS-USB. + * F-CLOCK + * SSI. + * + * GPMC memories and SDRC have timing and clock sensitive registers which + * may very well need notification when the clock changes. Currently for low + * operating points, these are taken care of in sleep.S. + */ +static struct clk core_l3_ck = { /* Used for ick and fck, interconnect */ + .name = "core_l3_ck", + .parent = &core_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL | ALWAYS_ENABLED | CM_CORE_SEL1 | + DELAYED_APP | CONFIG_PARTICIPANT | + RATE_PROPAGATES, + .rate_offset = 0, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk usb_l4_ick = { /* FS-USB interface clock */ + .name = "usb_l4_ick", + .parent = &core_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL | CM_CORE_SEL1 | DELAYED_APP | + CONFIG_PARTICIPANT, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 0, + .rate_offset = 25, + .recalc = &omap2_clksel_recalc, +}; + +/* + * SSI is in L3 management domain, its direct parent is core not l3, + * many core power domain entities are grouped into the L3 clock + * domain. + * SSI_SSR_FCLK, SSI_SST_FCLK, SSI_L4_CLIK + * + * ssr = core/1/2/3/4/5, sst = 1/2 ssr. + */ +static struct clk ssi_ssr_sst_fck = { + .name = "ssi_fck", + .parent = &core_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL | CM_CORE_SEL1 | DELAYED_APP, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, /* bit 1 */ + .enable_bit = 1, + .rate_offset = 20, + .recalc = &omap2_clksel_recalc, +}; + +/* + * GFX clock domain + * Clocks: + * GFX_FCLK, GFX_ICLK + * GFX_CG1(2d), GFX_CG2(3d) + * + * GFX_FCLK runs from L3, and is divided by (1,2,3,4) + * The 2d and 3d clocks run at a hardware determined + * divided value of fclk. + * + */ +static struct clk gfx_3d_fck = { + .name = "gfx_3d_fck", + .parent = &core_l3_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL | CM_GFX_SEL1, + .enable_reg = (void __iomem *)&CM_FCLKEN_GFX, + .enable_bit = 2, + .rate_offset= 0, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk gfx_2d_fck = { + .name = "gfx_2d_fck", + .parent = &core_l3_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL | CM_GFX_SEL1, + .enable_reg = (void __iomem *)&CM_FCLKEN_GFX, + .enable_bit = 1, + .rate_offset= 0, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk gfx_ick = { + .name = "gfx_ick", /* From l3 */ + .parent = &core_l3_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL, + .enable_reg = (void __iomem *)&CM_ICLKEN_GFX, /* bit 0 */ + .enable_bit = 0, + .recalc = &omap2_followparent_recalc, +}; + +/* + * Modem clock domain (2430) + * CLOCKS: + * MDM_OSC_CLK + * MDM_ICLK + */ +static struct clk mdm_ick = { /* used both as a ick and fck */ + .name = "mdm_ick", + .parent = &core_ck, + .flags = CLOCK_IN_OMAP243X | RATE_CKCTL | CM_MODEM_SEL1 | + DELAYED_APP | CONFIG_PARTICIPANT, + .rate_offset = 0, + .enable_reg = (void __iomem *)&CM_ICLKEN_MDM, + .enable_bit = 0, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk mdm_osc_ck = { + .name = "mdm_osc_ck", + .rate = 26000000, + .parent = &osc_ck, + .flags = CLOCK_IN_OMAP243X | RATE_FIXED, + .enable_reg = (void __iomem *)&CM_FCLKEN_MDM, + .enable_bit = 1, + .recalc = &omap2_followparent_recalc, +}; + +/* + * L4 clock management domain + * + * This domain contains lots of interface clocks from the L4 interface, some + * functional clocks. Fixed APLL functional source clocks are managed in + * this domain. + */ +static struct clk l4_ck = { /* used both as an ick and fck */ + .name = "l4_ck", + .parent = &core_l3_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL | ALWAYS_ENABLED | CM_CORE_SEL1 | + DELAYED_APP | RATE_PROPAGATES, + .rate_offset = 5, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk ssi_l4_ick = { + .name = "ssi_l4_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | RATE_CKCTL, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, /* bit 1 */ + .enable_bit = 1, + .recalc = &omap2_followparent_recalc, +}; + +/* + * DSS clock domain + * CLOCKs: + * DSS_L4_ICLK, DSS_L3_ICLK, + * DSS_CLK1, DSS_CLK2, DSS_54MHz_CLK + * + * DSS is both initiator and target. + */ +static struct clk dss_ick = { /* Enables both L3,L4 ICLK's */ + .name = "dss_ick", + .parent = &l4_ck, /* really both l3 and l4 */ + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | RATE_CKCTL, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 0, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk dss1_fck = { + .name = "dss1_fck", + .parent = &core_ck, /* Core or sys */ + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL | CM_CORE_SEL1 | DELAYED_APP, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 0, + .rate_offset = 8, + .src_offset = 8, + .recalc = &omap2_clksel_recalc, +}; + +static struct clk dss2_fck = { /* Alt clk used in power management */ + .name = "dss2_fck", + .parent = &sys_ck, /* fixed at sys_ck or 48MHz */ + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_CKCTL | CM_CORE_SEL1 | RATE_FIXED, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 1, + .src_offset = 13, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk dss_54m_fck = { /* Alt clk used in power management */ + .name = "dss_54m_fck", /* 54m tv clk */ + .parent = &func_54m_ck, + .rate = 54000000, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + RATE_FIXED | RATE_PROPAGATES, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 2, + .recalc = &omap2_propagate_rate, +}; + +/* + * CORE power domain ICLK & FCLK defines. + * Many of the these can have more than one possible parent. Entries + * here will likely have an L4 interface parent, and may have multiple + * functional clock parents. + */ +static struct clk gpt1_ick = { + .name = "gpt1_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN_WKUP, /* Bit4 */ + .enable_bit = 0, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt1_fck = { + .name = "gpt1_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_WKUP_SEL1, + .enable_reg = (void __iomem *)&CM_FCLKEN_WKUP, + .enable_bit = 0, + .src_offset = 0, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt2_ick = { + .name = "gpt2_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* bit4 */ + .enable_bit = 0, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt2_fck = { + .name = "gpt2_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 4, + .src_offset = 2, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt3_ick = { + .name = "gpt3_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* Bit5 */ + .enable_bit = 5, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt3_fck = { + .name = "gpt3_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 5, + .src_offset = 4, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt4_ick = { + .name = "gpt4_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* Bit6 */ + .enable_bit = 6, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt4_fck = { + .name = "gpt4_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 6, + .src_offset = 6, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt5_ick = { + .name = "gpt5_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* Bit7 */ + .enable_bit = 7, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt5_fck = { + .name = "gpt5_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 7, + .src_offset = 8, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt6_ick = { + .name = "gpt6_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_bit = 8, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* bit8 */ + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt6_fck = { + .name = "gpt6_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 8, + .src_offset = 10, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt7_ick = { + .name = "gpt7_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* bit9 */ + .enable_bit = 9, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt7_fck = { + .name = "gpt7_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 9, + .src_offset = 12, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt8_ick = { + .name = "gpt8_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* bit10 */ + .enable_bit = 10, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt8_fck = { + .name = "gpt8_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 10, + .src_offset = 14, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt9_ick = { + .name = "gpt9_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 11, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt9_fck = { + .name = "gpt9_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 11, + .src_offset = 16, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt10_ick = { + .name = "gpt10_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 12, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt10_fck = { + .name = "gpt10_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 12, + .src_offset = 18, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt11_ick = { + .name = "gpt11_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 13, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt11_fck = { + .name = "gpt11_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 13, + .src_offset = 20, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt12_ick = { + .name = "gpt12_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* bit14 */ + .enable_bit = 14, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpt12_fck = { + .name = "gpt12_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + CM_CORE_SEL2, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 14, + .src_offset = 22, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp1_ick = { + .name = "mcbsp1_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_bit = 15, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, /* bit16 */ + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp1_fck = { + .name = "mcbsp1_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_bit = 15, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp2_ick = { + .name = "mcbsp2_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_bit = 16, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp2_fck = { + .name = "mcbsp2_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_bit = 16, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp3_ick = { + .name = "mcbsp3_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 3, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp3_fck = { + .name = "mcbsp3_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 3, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp4_ick = { + .name = "mcbsp4_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 4, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp4_fck = { + .name = "mcbsp4_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 4, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp5_ick = { + .name = "mcbsp5_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 5, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcbsp5_fck = { + .name = "mcbsp5_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 5, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcspi1_ick = { + .name = "mcspi1_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 17, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcspi1_fck = { + .name = "mcspi1_fck", + .parent = &func_48m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 17, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcspi2_ick = { + .name = "mcspi2_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 18, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcspi2_fck = { + .name = "mcspi2_fck", + .parent = &func_48m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 18, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcspi3_ick = { + .name = "mcspi3_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 9, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mcspi3_fck = { + .name = "mcspi3_fck", + .parent = &func_48m_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 9, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk uart1_ick = { + .name = "uart1_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 21, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk uart1_fck = { + .name = "uart1_fck", + .parent = &func_48m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 21, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk uart2_ick = { + .name = "uart2_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 22, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk uart2_fck = { + .name = "uart2_fck", + .parent = &func_48m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 22, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk uart3_ick = { + .name = "uart3_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 2, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk uart3_fck = { + .name = "uart3_fck", + .parent = &func_48m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 2, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpios_ick = { + .name = "gpios_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN_WKUP, + .enable_bit = 2, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpios_fck = { + .name = "gpios_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN_WKUP, + .enable_bit = 2, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mpu_wdt_ick = { + .name = "mpu_wdt_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN_WKUP, + .enable_bit = 3, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mpu_wdt_fck = { + .name = "mpu_wdt_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN_WKUP, + .enable_bit = 3, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk sync_32k_ick = { + .name = "sync_32k_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN_WKUP, + .enable_bit = 1, + .recalc = &omap2_followparent_recalc, +}; +static struct clk wdt1_ick = { + .name = "wdt1_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN_WKUP, + .enable_bit = 4, + .recalc = &omap2_followparent_recalc, +}; +static struct clk omapctrl_ick = { + .name = "omapctrl_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN_WKUP, + .enable_bit = 5, + .recalc = &omap2_followparent_recalc, +}; +static struct clk icr_ick = { + .name = "icr_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN_WKUP, + .enable_bit = 6, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk cam_ick = { + .name = "cam_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 31, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk cam_fck = { + .name = "cam_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 31, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mailboxes_ick = { + .name = "mailboxes_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 30, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk wdt4_ick = { + .name = "wdt4_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 29, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk wdt4_fck = { + .name = "wdt4_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 29, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk wdt3_ick = { + .name = "wdt3_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 28, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk wdt3_fck = { + .name = "wdt3_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 28, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mspro_ick = { + .name = "mspro_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 27, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mspro_fck = { + .name = "mspro_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 27, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mmc_ick = { + .name = "mmc_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 26, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mmc_fck = { + .name = "mmc_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 26, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk fac_ick = { + .name = "fac_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 25, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk fac_fck = { + .name = "fac_fck", + .parent = &func_12m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 25, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk eac_ick = { + .name = "eac_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 24, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk eac_fck = { + .name = "eac_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 24, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk hdq_ick = { + .name = "hdq_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 23, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk hdq_fck = { + .name = "hdq_fck", + .parent = &func_12m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 23, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk i2c2_ick = { + .name = "i2c2_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 20, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk i2c2_fck = { + .name = "i2c2_fck", + .parent = &func_12m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 20, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk i2chs2_fck = { + .name = "i2chs2_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 20, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk i2c1_ick = { + .name = "i2c1_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 19, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk i2c1_fck = { + .name = "i2c1_fck", + .parent = &func_12m_ck, + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 19, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk i2chs1_fck = { + .name = "i2chs1_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 19, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk vlynq_ick = { + .name = "vlynq_ick", + .parent = &core_l3_ck, + .flags = CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN1_CORE, + .enable_bit = 3, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk vlynq_fck = { + .name = "vlynq_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP242X | RATE_CKCTL | CM_CORE_SEL1 | DELAYED_APP, + .enable_reg = (void __iomem *)&CM_FCLKEN1_CORE, + .enable_bit = 3, + .src_offset = 15, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk sdrc_ick = { + .name = "sdrc_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN3_CORE, + .enable_bit = 2, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk des_ick = { + .name = "des_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X | CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN4_CORE, + .enable_bit = 0, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk sha_ick = { + .name = "sha_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X | CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN4_CORE, + .enable_bit = 1, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk rng_ick = { + .name = "rng_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X | CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN4_CORE, + .enable_bit = 2, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk aes_ick = { + .name = "aes_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X | CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN4_CORE, + .enable_bit = 3, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk pka_ick = { + .name = "pka_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X | CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_ICLKEN4_CORE, + .enable_bit = 4, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk usb_fck = { + .name = "usb_fck", + .parent = &func_48m_ck, + .flags = CLOCK_IN_OMAP243X | CLOCK_IN_OMAP242X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 0, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk usbhs_ick = { + .name = "usbhs_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 6, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mmchs1_ick = { + .name = "mmchs1_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 7, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mmchs1_fck = { + .name = "mmchs1_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 7, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mmchs2_ick = { + .name = "mmchs2_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 8, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mmchs2_fck = { + .name = "mmchs2_fck", + .parent = &func_96m_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 8, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpio5_ick = { + .name = "gpio5_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 10, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk gpio5_fck = { + .name = "gpio5_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 10, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mdm_intc_ick = { + .name = "mdm_intc_ick", + .parent = &l4_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_ICLKEN2_CORE, + .enable_bit = 11, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mmchsdb1_fck = { + .name = "mmchsdb1_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 16, + .recalc = &omap2_followparent_recalc, +}; + +static struct clk mmchsdb2_fck = { + .name = "mmchsdb2_fck", + .parent = &func_32k_ck, + .flags = CLOCK_IN_OMAP243X, + .enable_reg = (void __iomem *)&CM_FCLKEN2_CORE, + .enable_bit = 17, + .recalc = &omap2_followparent_recalc, +}; + +/* + * This clock is a composite clock which does entire set changes then + * forces a rebalance. It keys on the MPU speed, but it really could + * be any key speed part of a set in the rate table. + * + * to really change a set, you need memory table sets which get changed + * in sram, pre-notifiers & post notifiers, changing the top set, without + * having low level display recalc's won't work... this is why dpm notifiers + * work, isr's off, walk a list of clocks already _off_ and not messing with + * the bus. + * + * This clock should have no parent. It embodies the entire upper level + * active set. A parent will mess up some of the init also. + */ +static struct clk virt_prcm_set = { + .name = "virt_prcm_set", + .flags = CLOCK_IN_OMAP242X | CLOCK_IN_OMAP243X | + VIRTUAL_CLOCK | ALWAYS_ENABLED | DELAYED_APP, + .parent = &mpu_ck, /* Indexed by mpu speed, no parent */ + .recalc = &omap2_mpu_recalc, /* sets are keyed on mpu rate */ + .set_rate = &omap2_select_table_rate, + .round_rate = &omap2_round_to_table_rate, +}; + +static struct clk *onchip_clks[] = { + /* external root sources */ + &func_32k_ck, + &osc_ck, + &sys_ck, + &alt_ck, + /* internal analog sources */ + &dpll_ck, + &apll96_ck, + &apll54_ck, + /* internal prcm root sources */ + &func_54m_ck, + &core_ck, + &sleep_ck, + &func_96m_ck, + &func_48m_ck, + &func_12m_ck, + &wdt1_osc_ck, + &sys_clkout, + &sys_clkout2, + /* mpu domain clocks */ + &mpu_ck, + /* dsp domain clocks */ + &iva2_1_fck, /* 2430 */ + &iva2_1_ick, + &dsp_ick, /* 2420 */ + &dsp_fck, + &iva1_ifck, + &iva1_mpu_int_ifck, + /* GFX domain clocks */ + &gfx_3d_fck, + &gfx_2d_fck, + &gfx_ick, + /* Modem domain clocks */ + &mdm_ick, + &mdm_osc_ck, + /* DSS domain clocks */ + &dss_ick, + &dss1_fck, + &dss2_fck, + &dss_54m_fck, + /* L3 domain clocks */ + &core_l3_ck, + &ssi_ssr_sst_fck, + &usb_l4_ick, + /* L4 domain clocks */ + &l4_ck, /* used as both core_l4 and wu_l4 */ + &ssi_l4_ick, + /* virtual meta-group clock */ + &virt_prcm_set, + /* general l4 interface ck, multi-parent functional clk */ + &gpt1_ick, + &gpt1_fck, + &gpt2_ick, + &gpt2_fck, + &gpt3_ick, + &gpt3_fck, + &gpt4_ick, + &gpt4_fck, + &gpt5_ick, + &gpt5_fck, + &gpt6_ick, + &gpt6_fck, + &gpt7_ick, + &gpt7_fck, + &gpt8_ick, + &gpt8_fck, + &gpt9_ick, + &gpt9_fck, + &gpt10_ick, + &gpt10_fck, + &gpt11_ick, + &gpt11_fck, + &gpt12_ick, + &gpt12_fck, + &mcbsp1_ick, + &mcbsp1_fck, + &mcbsp2_ick, + &mcbsp2_fck, + &mcbsp3_ick, + &mcbsp3_fck, + &mcbsp4_ick, + &mcbsp4_fck, + &mcbsp5_ick, + &mcbsp5_fck, + &mcspi1_ick, + &mcspi1_fck, + &mcspi2_ick, + &mcspi2_fck, + &mcspi3_ick, + &mcspi3_fck, + &uart1_ick, + &uart1_fck, + &uart2_ick, + &uart2_fck, + &uart3_ick, + &uart3_fck, + &gpios_ick, + &gpios_fck, + &mpu_wdt_ick, + &mpu_wdt_fck, + &sync_32k_ick, + &wdt1_ick, + &omapctrl_ick, + &icr_ick, + &cam_fck, + &cam_ick, + &mailboxes_ick, + &wdt4_ick, + &wdt4_fck, + &wdt3_ick, + &wdt3_fck, + &mspro_ick, + &mspro_fck, + &mmc_ick, + &mmc_fck, + &fac_ick, + &fac_fck, + &eac_ick, + &eac_fck, + &hdq_ick, + &hdq_fck, + &i2c1_ick, + &i2c1_fck, + &i2chs1_fck, + &i2c2_ick, + &i2c2_fck, + &i2chs2_fck, + &vlynq_ick, + &vlynq_fck, + &sdrc_ick, + &des_ick, + &sha_ick, + &rng_ick, + &aes_ick, + &pka_ick, + &usb_fck, + &usbhs_ick, + &mmchs1_ick, + &mmchs1_fck, + &mmchs2_ick, + &mmchs2_fck, + &gpio5_ick, + &gpio5_fck, + &mdm_intc_ick, + &mmchsdb1_fck, + &mmchsdb2_fck, +}; + +#endif From 9ad5897c2659b3c610e0c717e8b3dbfb496d2c74 Mon Sep 17 00:00:00 2001 From: Tony Lindgren Date: Thu, 10 Nov 2005 14:26:53 +0000 Subject: [PATCH 095/121] [ARM] 3143/1: OMAP 4/5: Update omap include files Patch from Tony Lindgren This patch syncs the mainline kernel with linux-omap tree. This patch contains changes to common header files for omap1xxx and omap24xx by various omap developers, and improved cpu detection by Imre Deak Signed-off-by: Tony Lindgren Signed-off-by: Russell King --- include/asm-arm/arch-omap/board-h4.h | 4 + include/asm-arm/arch-omap/board-innovator.h | 4 +- include/asm-arm/arch-omap/clock.h | 91 +++++ include/asm-arm/arch-omap/common.h | 2 +- include/asm-arm/arch-omap/cpu.h | 82 ++-- include/asm-arm/arch-omap/dma.h | 261 +++++++++--- include/asm-arm/arch-omap/entry-macro.S | 14 + include/asm-arm/arch-omap/fpga.h | 4 +- include/asm-arm/arch-omap/gpio.h | 2 +- include/asm-arm/arch-omap/hardware.h | 8 +- include/asm-arm/arch-omap/io.h | 32 +- include/asm-arm/arch-omap/irqs.h | 13 +- include/asm-arm/arch-omap/memory.h | 4 +- include/asm-arm/arch-omap/menelaus.h | 22 + include/asm-arm/arch-omap/mux.h | 329 ++++----------- include/asm-arm/arch-omap/omap1510.h | 6 +- include/asm-arm/arch-omap/omap24xx.h | 17 +- include/asm-arm/arch-omap/omapfb.h | 281 +++++++++++++ include/asm-arm/arch-omap/pm.h | 40 +- include/asm-arm/arch-omap/prcm.h | 429 ++++++++++++++++++++ include/asm-arm/arch-omap/sram.h | 38 ++ include/asm-arm/arch-omap/system.h | 35 +- include/asm-arm/arch-omap/timex.h | 8 + include/asm-arm/arch-omap/uncompress.h | 6 +- 24 files changed, 1371 insertions(+), 361 deletions(-) create mode 100644 include/asm-arm/arch-omap/clock.h create mode 100644 include/asm-arm/arch-omap/menelaus.h create mode 100644 include/asm-arm/arch-omap/omapfb.h create mode 100644 include/asm-arm/arch-omap/prcm.h create mode 100644 include/asm-arm/arch-omap/sram.h diff --git a/include/asm-arm/arch-omap/board-h4.h b/include/asm-arm/arch-omap/board-h4.h index d64ee9211eed..95ba43042a01 100644 --- a/include/asm-arm/arch-omap/board-h4.h +++ b/include/asm-arm/arch-omap/board-h4.h @@ -34,5 +34,9 @@ #define OMAP24XX_ETHR_START 0x08000300 #define OMAP24XX_ETHR_GPIO_IRQ 92 +#define H4_CS0_BASE 0x04000000 + +#define H4_CS0_BASE 0x04000000 + #endif /* __ASM_ARCH_OMAP_H4_H */ diff --git a/include/asm-arm/arch-omap/board-innovator.h b/include/asm-arm/arch-omap/board-innovator.h index 79574e0ed13d..b3cf33441f6e 100644 --- a/include/asm-arm/arch-omap/board-innovator.h +++ b/include/asm-arm/arch-omap/board-innovator.h @@ -26,7 +26,7 @@ #ifndef __ASM_ARCH_OMAP_INNOVATOR_H #define __ASM_ARCH_OMAP_INNOVATOR_H -#if defined (CONFIG_ARCH_OMAP1510) +#if defined (CONFIG_ARCH_OMAP15XX) #ifndef OMAP_SDRAM_DEVICE #define OMAP_SDRAM_DEVICE D256M_1X16_4B @@ -44,7 +44,7 @@ void fpga_write(unsigned char val, int reg); unsigned char fpga_read(int reg); #endif -#endif /* CONFIG_ARCH_OMAP1510 */ +#endif /* CONFIG_ARCH_OMAP15XX */ #if defined (CONFIG_ARCH_OMAP16XX) diff --git a/include/asm-arm/arch-omap/clock.h b/include/asm-arm/arch-omap/clock.h new file mode 100644 index 000000000000..740c297eb11c --- /dev/null +++ b/include/asm-arm/arch-omap/clock.h @@ -0,0 +1,91 @@ +/* + * linux/include/asm-arm/arch-omap/clock.h + * + * Copyright (C) 2004 - 2005 Nokia corporation + * Written by Tuukka Tikkanen + * Based on clocks.h by Tony Lindgren, Gordon McNutt and RidgeRun, Inc + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __ARCH_ARM_OMAP_CLOCK_H +#define __ARCH_ARM_OMAP_CLOCK_H + +struct module; + +struct clk { + struct list_head node; + struct module *owner; + const char *name; + struct clk *parent; + unsigned long rate; + __u32 flags; + void __iomem *enable_reg; + __u8 enable_bit; + __u8 rate_offset; + __u8 src_offset; + __s8 usecount; + void (*recalc)(struct clk *); + int (*set_rate)(struct clk *, unsigned long); + long (*round_rate)(struct clk *, unsigned long); + void (*init)(struct clk *); + int (*enable)(struct clk *); + void (*disable)(struct clk *); +}; + +struct clk_functions { + int (*clk_enable)(struct clk *clk); + void (*clk_disable)(struct clk *clk); + int (*clk_use)(struct clk *clk); + void (*clk_unuse)(struct clk *clk); + long (*clk_round_rate)(struct clk *clk, unsigned long rate); + int (*clk_set_rate)(struct clk *clk, unsigned long rate); + int (*clk_set_parent)(struct clk *clk, struct clk *parent); + struct clk * (*clk_get_parent)(struct clk *clk); + void (*clk_allow_idle)(struct clk *clk); + void (*clk_deny_idle)(struct clk *clk); +}; + +extern unsigned int mpurate; +extern struct list_head clocks; +extern spinlock_t clockfw_lock; + +extern int clk_init(struct clk_functions * custom_clocks); +extern int clk_register(struct clk *clk); +extern void clk_unregister(struct clk *clk); +extern void propagate_rate(struct clk *clk); +extern void followparent_recalc(struct clk * clk); +extern void clk_allow_idle(struct clk *clk); +extern void clk_deny_idle(struct clk *clk); + +/* Clock flags */ +#define RATE_CKCTL (1 << 0) /* Main fixed ratio clocks */ +#define RATE_FIXED (1 << 1) /* Fixed clock rate */ +#define RATE_PROPAGATES (1 << 2) /* Program children too */ +#define VIRTUAL_CLOCK (1 << 3) /* Composite clock from table */ +#define ALWAYS_ENABLED (1 << 4) /* Clock cannot be disabled */ +#define ENABLE_REG_32BIT (1 << 5) /* Use 32-bit access */ +#define VIRTUAL_IO_ADDRESS (1 << 6) /* Clock in virtual address */ +#define CLOCK_IDLE_CONTROL (1 << 7) +#define CLOCK_NO_IDLE_PARENT (1 << 8) +#define DELAYED_APP (1 << 9) /* Delay application of clock */ +#define CONFIG_PARTICIPANT (1 << 10) /* Fundamental clock */ +#define CM_MPU_SEL1 (1 << 11) /* Domain divider/source */ +#define CM_DSP_SEL1 (1 << 12) +#define CM_GFX_SEL1 (1 << 13) +#define CM_MODEM_SEL1 (1 << 14) +#define CM_CORE_SEL1 (1 << 15) /* Sets divider for many */ +#define CM_CORE_SEL2 (1 << 16) /* sets parent for GPT */ +#define CM_WKUP_SEL1 (1 << 17) +#define CM_PLL_SEL1 (1 << 18) +#define CM_PLL_SEL2 (1 << 19) +#define CM_SYSCLKOUT_SEL1 (1 << 20) +#define CLOCK_IN_OMAP730 (1 << 21) +#define CLOCK_IN_OMAP1510 (1 << 22) +#define CLOCK_IN_OMAP16XX (1 << 23) +#define CLOCK_IN_OMAP242X (1 << 24) +#define CLOCK_IN_OMAP243X (1 << 25) + +#endif diff --git a/include/asm-arm/arch-omap/common.h b/include/asm-arm/arch-omap/common.h index 2a676b4f13b5..08d58abd8218 100644 --- a/include/asm-arm/arch-omap/common.h +++ b/include/asm-arm/arch-omap/common.h @@ -31,6 +31,6 @@ struct sys_timer; extern void omap_map_common_io(void); extern struct sys_timer omap_timer; -extern void omap_serial_init(int ports[]); +extern void omap_serial_init(void); #endif /* __ARCH_ARM_MACH_OMAP_COMMON_H */ diff --git a/include/asm-arm/arch-omap/cpu.h b/include/asm-arm/arch-omap/cpu.h index 1119e2b53e72..ec7eb675d922 100644 --- a/include/asm-arm/arch-omap/cpu.h +++ b/include/asm-arm/arch-omap/cpu.h @@ -28,12 +28,7 @@ extern unsigned int system_rev; -#define OMAP_DIE_ID_0 0xfffe1800 -#define OMAP_DIE_ID_1 0xfffe1804 -#define OMAP_PRODUCTION_ID_0 0xfffe2000 -#define OMAP_PRODUCTION_ID_1 0xfffe2004 -#define OMAP32_ID_0 0xfffed400 -#define OMAP32_ID_1 0xfffed404 +#define omap2_cpu_rev() ((system_rev >> 8) & 0x0f) /* * Test if multicore OMAP support is needed @@ -50,7 +45,7 @@ extern unsigned int system_rev; # define OMAP_NAME omap730 # endif #endif -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX # ifdef OMAP_NAME # undef MULTI_OMAP1 # define MULTI_OMAP1 @@ -79,9 +74,11 @@ extern unsigned int system_rev; * Macros to group OMAP into cpu classes. * These can be used in most places. * cpu_is_omap7xx(): True for OMAP730 - * cpu_is_omap15xx(): True for OMAP1510 and OMAP5910 + * cpu_is_omap15xx(): True for OMAP1510, OMAP5910 and OMAP310 * cpu_is_omap16xx(): True for OMAP1610, OMAP5912 and OMAP1710 - * cpu_is_omap24xx(): True for OMAP2420 + * cpu_is_omap24xx(): True for OMAP2420, OMAP2422, OMAP2423, OMAP2430 + * cpu_is_omap242x(): True for OMAP2420, OMAP2422, OMAP2423 + * cpu_is_omap243x(): True for OMAP2430 */ #define GET_OMAP_CLASS (system_rev & 0xff) @@ -91,22 +88,35 @@ static inline int is_omap ##class (void) \ return (GET_OMAP_CLASS == (id)) ? 1 : 0; \ } +#define GET_OMAP_SUBCLASS ((system_rev >> 20) & 0x0fff) + +#define IS_OMAP_SUBCLASS(subclass, id) \ +static inline int is_omap ##subclass (void) \ +{ \ + return (GET_OMAP_SUBCLASS == (id)) ? 1 : 0; \ +} + IS_OMAP_CLASS(7xx, 0x07) IS_OMAP_CLASS(15xx, 0x15) IS_OMAP_CLASS(16xx, 0x16) IS_OMAP_CLASS(24xx, 0x24) +IS_OMAP_SUBCLASS(242x, 0x242) +IS_OMAP_SUBCLASS(243x, 0x243) + #define cpu_is_omap7xx() 0 #define cpu_is_omap15xx() 0 #define cpu_is_omap16xx() 0 #define cpu_is_omap24xx() 0 +#define cpu_is_omap242x() 0 +#define cpu_is_omap243x() 0 #if defined(MULTI_OMAP1) # if defined(CONFIG_ARCH_OMAP730) # undef cpu_is_omap7xx # define cpu_is_omap7xx() is_omap7xx() # endif -# if defined(CONFIG_ARCH_OMAP1510) +# if defined(CONFIG_ARCH_OMAP15XX) # undef cpu_is_omap15xx # define cpu_is_omap15xx() is_omap15xx() # endif @@ -119,7 +129,7 @@ IS_OMAP_CLASS(24xx, 0x24) # undef cpu_is_omap7xx # define cpu_is_omap7xx() 1 # endif -# if defined(CONFIG_ARCH_OMAP1510) +# if defined(CONFIG_ARCH_OMAP15XX) # undef cpu_is_omap15xx # define cpu_is_omap15xx() 1 # endif @@ -129,13 +139,18 @@ IS_OMAP_CLASS(24xx, 0x24) # endif # if defined(CONFIG_ARCH_OMAP24XX) # undef cpu_is_omap24xx +# undef cpu_is_omap242x +# undef cpu_is_omap243x # define cpu_is_omap24xx() 1 +# define cpu_is_omap242x() is_omap242x() +# define cpu_is_omap243x() is_omap243x() # endif #endif /* * Macros to detect individual cpu types. * These are only rarely needed. + * cpu_is_omap330(): True for OMAP330 * cpu_is_omap730(): True for OMAP730 * cpu_is_omap1510(): True for OMAP1510 * cpu_is_omap1610(): True for OMAP1610 @@ -144,6 +159,9 @@ IS_OMAP_CLASS(24xx, 0x24) * cpu_is_omap1621(): True for OMAP1621 * cpu_is_omap1710(): True for OMAP1710 * cpu_is_omap2420(): True for OMAP2420 + * cpu_is_omap2422(): True for OMAP2422 + * cpu_is_omap2423(): True for OMAP2423 + * cpu_is_omap2430(): True for OMAP2430 */ #define GET_OMAP_TYPE ((system_rev >> 16) & 0xffff) @@ -153,6 +171,7 @@ static inline int is_omap ##type (void) \ return (GET_OMAP_TYPE == (id)) ? 1 : 0; \ } +IS_OMAP_TYPE(310, 0x0310) IS_OMAP_TYPE(730, 0x0730) IS_OMAP_TYPE(1510, 0x1510) IS_OMAP_TYPE(1610, 0x1610) @@ -161,7 +180,11 @@ IS_OMAP_TYPE(5912, 0x1611) IS_OMAP_TYPE(1621, 0x1621) IS_OMAP_TYPE(1710, 0x1710) IS_OMAP_TYPE(2420, 0x2420) +IS_OMAP_TYPE(2422, 0x2422) +IS_OMAP_TYPE(2423, 0x2423) +IS_OMAP_TYPE(2430, 0x2430) +#define cpu_is_omap310() 0 #define cpu_is_omap730() 0 #define cpu_is_omap1510() 0 #define cpu_is_omap1610() 0 @@ -170,31 +193,33 @@ IS_OMAP_TYPE(2420, 0x2420) #define cpu_is_omap1621() 0 #define cpu_is_omap1710() 0 #define cpu_is_omap2420() 0 +#define cpu_is_omap2422() 0 +#define cpu_is_omap2423() 0 +#define cpu_is_omap2430() 0 #if defined(MULTI_OMAP1) # if defined(CONFIG_ARCH_OMAP730) # undef cpu_is_omap730 # define cpu_is_omap730() is_omap730() # endif -# if defined(CONFIG_ARCH_OMAP1510) -# undef cpu_is_omap1510 -# define cpu_is_omap1510() is_omap1510() -# endif #else # if defined(CONFIG_ARCH_OMAP730) # undef cpu_is_omap730 # define cpu_is_omap730() 1 # endif -# if defined(CONFIG_ARCH_OMAP1510) -# undef cpu_is_omap1510 -# define cpu_is_omap1510() 1 -# endif #endif /* * Whether we have MULTI_OMAP1 or not, we still need to distinguish - * between 1611B/5912 and 1710. + * between 330 vs. 1510 and 1611B/5912 vs. 1710. */ +#if defined(CONFIG_ARCH_OMAP15XX) +# undef cpu_is_omap310 +# undef cpu_is_omap1510 +# define cpu_is_omap310() is_omap310() +# define cpu_is_omap1510() is_omap1510() +#endif + #if defined(CONFIG_ARCH_OMAP16XX) # undef cpu_is_omap1610 # undef cpu_is_omap1611 @@ -208,9 +233,20 @@ IS_OMAP_TYPE(2420, 0x2420) # define cpu_is_omap1710() is_omap1710() #endif -#if defined(CONFIG_ARCH_OMAP2420) -# undef cpu_is_omap2420 -# define cpu_is_omap2420() 1 +#if defined(CONFIG_ARCH_OMAP24XX) +# undef cpu_is_omap2420 +# undef cpu_is_omap2422 +# undef cpu_is_omap2423 +# undef cpu_is_omap2430 +# define cpu_is_omap2420() is_omap2420() +# define cpu_is_omap2422() is_omap2422() +# define cpu_is_omap2423() is_omap2423() +# define cpu_is_omap2430() is_omap2430() #endif +/* Macros to detect if we have OMAP1 or OMAP2 */ +#define cpu_class_is_omap1() (cpu_is_omap730() || cpu_is_omap15xx() || \ + cpu_is_omap16xx()) +#define cpu_class_is_omap2() cpu_is_omap24xx() + #endif diff --git a/include/asm-arm/arch-omap/dma.h b/include/asm-arm/arch-omap/dma.h index 04ebef5c6e95..ccbcb580a5c1 100644 --- a/include/asm-arm/arch-omap/dma.h +++ b/include/asm-arm/arch-omap/dma.h @@ -22,9 +22,109 @@ #define __ASM_ARCH_DMA_H #define MAX_DMA_ADDRESS 0xffffffff +#define MAX_DMA_CHANNELS 0 + +/* Hardware registers for omap1 */ +#define OMAP_DMA_BASE (0xfffed800) +#define OMAP_DMA_GCR (OMAP_DMA_BASE + 0x400) +#define OMAP_DMA_GSCR (OMAP_DMA_BASE + 0x404) +#define OMAP_DMA_GRST (OMAP_DMA_BASE + 0x408) +#define OMAP_DMA_HW_ID (OMAP_DMA_BASE + 0x442) +#define OMAP_DMA_PCH2_ID (OMAP_DMA_BASE + 0x444) +#define OMAP_DMA_PCH0_ID (OMAP_DMA_BASE + 0x446) +#define OMAP_DMA_PCH1_ID (OMAP_DMA_BASE + 0x448) +#define OMAP_DMA_PCHG_ID (OMAP_DMA_BASE + 0x44a) +#define OMAP_DMA_PCHD_ID (OMAP_DMA_BASE + 0x44c) +#define OMAP_DMA_CAPS_0_U (OMAP_DMA_BASE + 0x44e) +#define OMAP_DMA_CAPS_0_L (OMAP_DMA_BASE + 0x450) +#define OMAP_DMA_CAPS_1_U (OMAP_DMA_BASE + 0x452) +#define OMAP_DMA_CAPS_1_L (OMAP_DMA_BASE + 0x454) +#define OMAP_DMA_CAPS_2 (OMAP_DMA_BASE + 0x456) +#define OMAP_DMA_CAPS_3 (OMAP_DMA_BASE + 0x458) +#define OMAP_DMA_CAPS_4 (OMAP_DMA_BASE + 0x45a) +#define OMAP_DMA_PCH2_SR (OMAP_DMA_BASE + 0x460) +#define OMAP_DMA_PCH0_SR (OMAP_DMA_BASE + 0x480) +#define OMAP_DMA_PCH1_SR (OMAP_DMA_BASE + 0x482) +#define OMAP_DMA_PCHD_SR (OMAP_DMA_BASE + 0x4c0) + +/* Hardware registers for omap2 */ +#define OMAP24XX_DMA_BASE (L4_24XX_BASE + 0x56000) +#define OMAP_DMA4_REVISION (OMAP24XX_DMA_BASE + 0x00) +#define OMAP_DMA4_GCR_REG (OMAP24XX_DMA_BASE + 0x78) +#define OMAP_DMA4_IRQSTATUS_L0 (OMAP24XX_DMA_BASE + 0x08) +#define OMAP_DMA4_IRQSTATUS_L1 (OMAP24XX_DMA_BASE + 0x0c) +#define OMAP_DMA4_IRQSTATUS_L2 (OMAP24XX_DMA_BASE + 0x10) +#define OMAP_DMA4_IRQSTATUS_L3 (OMAP24XX_DMA_BASE + 0x14) +#define OMAP_DMA4_IRQENABLE_L0 (OMAP24XX_DMA_BASE + 0x18) +#define OMAP_DMA4_IRQENABLE_L1 (OMAP24XX_DMA_BASE + 0x1c) +#define OMAP_DMA4_IRQENABLE_L2 (OMAP24XX_DMA_BASE + 0x20) +#define OMAP_DMA4_IRQENABLE_L3 (OMAP24XX_DMA_BASE + 0x24) +#define OMAP_DMA4_SYSSTATUS (OMAP24XX_DMA_BASE + 0x28) +#define OMAP_DMA4_CAPS_0 (OMAP24XX_DMA_BASE + 0x64) +#define OMAP_DMA4_CAPS_2 (OMAP24XX_DMA_BASE + 0x6c) +#define OMAP_DMA4_CAPS_3 (OMAP24XX_DMA_BASE + 0x70) +#define OMAP_DMA4_CAPS_4 (OMAP24XX_DMA_BASE + 0x74) + +#ifdef CONFIG_ARCH_OMAP1 #define OMAP_LOGICAL_DMA_CH_COUNT 17 +/* Common channel specific registers for omap1 */ +#define OMAP_DMA_CSDP_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x00) +#define OMAP_DMA_CCR_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x02) +#define OMAP_DMA_CICR_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x04) +#define OMAP_DMA_CSR_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x06) +#define OMAP_DMA_CEN_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x10) +#define OMAP_DMA_CFN_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x12) +#define OMAP_DMA_CSFI_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x14) +#define OMAP_DMA_CSEI_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x16) +#define OMAP_DMA_CSAC_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x18) +#define OMAP_DMA_CDAC_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x1a) +#define OMAP_DMA_CDEI_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x1c) +#define OMAP_DMA_CDFI_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x1e) +#define OMAP_DMA_CLNK_CTRL_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x28) + +#else + +#define OMAP_LOGICAL_DMA_CH_COUNT 32 /* REVISIT: Is this 32 + 2? */ + +/* Common channel specific registers for omap2 */ +#define OMAP_DMA_CCR_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0x80) +#define OMAP_DMA_CLNK_CTRL_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0x84) +#define OMAP_DMA_CICR_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0x88) +#define OMAP_DMA_CSR_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0x8c) +#define OMAP_DMA_CSDP_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0x90) +#define OMAP_DMA_CEN_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0x94) +#define OMAP_DMA_CFN_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0x98) +#define OMAP_DMA_CSEI_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xa4) +#define OMAP_DMA_CSFI_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xa8) +#define OMAP_DMA_CDEI_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xac) +#define OMAP_DMA_CDFI_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xb0) +#define OMAP_DMA_CSAC_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xb4) +#define OMAP_DMA_CDAC_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xb8) + +#endif + +/* Channel specific registers only on omap1 */ +#define OMAP1_DMA_CSSA_L_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x08) +#define OMAP1_DMA_CSSA_U_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x0a) +#define OMAP1_DMA_CDSA_L_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x0c) +#define OMAP1_DMA_CDSA_U_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x0e) +#define OMAP1_DMA_COLOR_L_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x20) +#define OMAP1_DMA_CCR2_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x24) +#define OMAP1_DMA_COLOR_U_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x22) +#define OMAP1_DMA_LCH_CTRL_REG(n) __REG16(OMAP_DMA_BASE + 0x40 * (n) + 0x2a) + +/* Channel specific registers only on omap2 */ +#define OMAP2_DMA_CSSA_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0x9c) +#define OMAP2_DMA_CDSA_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xa0) +#define OMAP2_DMA_CCEN_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xbc) +#define OMAP2_DMA_CCFN_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xc0) +#define OMAP2_DMA_COLOR_REG(n) __REG32(OMAP24XX_DMA_BASE + 0x60 * (n) + 0xc4) + +/*----------------------------------------------------------------------------*/ + +/* DMA channels for omap1 */ #define OMAP_DMA_NO_DEVICE 0 #define OMAP_DMA_MCSI1_TX 1 #define OMAP_DMA_MCSI1_RX 2 @@ -85,29 +185,72 @@ #define OMAP_DMA_MMC2_RX 55 #define OMAP_DMA_CRYPTO_DES_OUT 56 +/* DMA channels for 24xx */ +#define OMAP24XX_DMA_NO_DEVICE 0 +#define OMAP24XX_DMA_XTI_DMA 1 /* S_DMA_0 */ +#define OMAP24XX_DMA_EXT_NDMA_REQ0 2 /* S_DMA_1 */ +#define OMAP24XX_DMA_EXT_NDMA_REQ1 3 /* S_DMA_2 */ +#define OMAP24XX_DMA_GPMC 4 /* S_DMA_3 */ +#define OMAP24XX_DMA_GFX 5 /* S_DMA_4 */ +#define OMAP24XX_DMA_DSS 6 /* S_DMA_5 */ +#define OMAP24XX_DMA_VLYNQ_TX 7 /* S_DMA_6 */ +#define OMAP24XX_DMA_CWT 8 /* S_DMA_7 */ +#define OMAP24XX_DMA_AES_TX 9 /* S_DMA_8 */ +#define OMAP24XX_DMA_AES_RX 10 /* S_DMA_9 */ +#define OMAP24XX_DMA_DES_TX 11 /* S_DMA_10 */ +#define OMAP24XX_DMA_DES_RX 12 /* S_DMA_11 */ +#define OMAP24XX_DMA_SHA1MD5_RX 13 /* S_DMA_12 */ -#define OMAP_DMA_BASE (0xfffed800) -#define OMAP_DMA_GCR (OMAP_DMA_BASE + 0x400) -#define OMAP_DMA_GSCR (OMAP_DMA_BASE + 0x404) -#define OMAP_DMA_GRST (OMAP_DMA_BASE + 0x408) -#define OMAP_DMA_HW_ID (OMAP_DMA_BASE + 0x442) -#define OMAP_DMA_PCH2_ID (OMAP_DMA_BASE + 0x444) -#define OMAP_DMA_PCH0_ID (OMAP_DMA_BASE + 0x446) -#define OMAP_DMA_PCH1_ID (OMAP_DMA_BASE + 0x448) -#define OMAP_DMA_PCHG_ID (OMAP_DMA_BASE + 0x44a) -#define OMAP_DMA_PCHD_ID (OMAP_DMA_BASE + 0x44c) -#define OMAP_DMA_CAPS_0_U (OMAP_DMA_BASE + 0x44e) -#define OMAP_DMA_CAPS_0_L (OMAP_DMA_BASE + 0x450) -#define OMAP_DMA_CAPS_1_U (OMAP_DMA_BASE + 0x452) -#define OMAP_DMA_CAPS_1_L (OMAP_DMA_BASE + 0x454) -#define OMAP_DMA_CAPS_2 (OMAP_DMA_BASE + 0x456) -#define OMAP_DMA_CAPS_3 (OMAP_DMA_BASE + 0x458) -#define OMAP_DMA_CAPS_4 (OMAP_DMA_BASE + 0x45a) -#define OMAP_DMA_PCH2_SR (OMAP_DMA_BASE + 0x460) -#define OMAP_DMA_PCH0_SR (OMAP_DMA_BASE + 0x480) -#define OMAP_DMA_PCH1_SR (OMAP_DMA_BASE + 0x482) -#define OMAP_DMA_PCHD_SR (OMAP_DMA_BASE + 0x4c0) +#define OMAP24XX_DMA_EAC_AC_RD 17 /* S_DMA_16 */ +#define OMAP24XX_DMA_EAC_AC_WR 18 /* S_DMA_17 */ +#define OMAP24XX_DMA_EAC_MD_UL_RD 19 /* S_DMA_18 */ +#define OMAP24XX_DMA_EAC_MD_UL_WR 20 /* S_DMA_19 */ +#define OMAP24XX_DMA_EAC_MD_DL_RD 21 /* S_DMA_20 */ +#define OMAP24XX_DMA_EAC_MD_DL_WR 22 /* S_DMA_21 */ +#define OMAP24XX_DMA_EAC_BT_UL_RD 23 /* S_DMA_22 */ +#define OMAP24XX_DMA_EAC_BT_UL_WR 24 /* S_DMA_23 */ +#define OMAP24XX_DMA_EAC_BT_DL_RD 25 /* S_DMA_24 */ +#define OMAP24XX_DMA_EAC_BT_DL_WR 26 /* S_DMA_25 */ +#define OMAP24XX_DMA_I2C1_TX 27 /* S_DMA_26 */ +#define OMAP24XX_DMA_I2C1_RX 28 /* S_DMA_27 */ +#define OMAP24XX_DMA_I2C2_TX 29 /* S_DMA_28 */ +#define OMAP24XX_DMA_I2C2_RX 30 /* S_DMA_29 */ +#define OMAP24XX_DMA_MCBSP1_TX 31 /* SDMA_30 */ +#define OMAP24XX_DMA_MCBSP1_RX 32 /* SDMA_31 */ +#define OMAP24XX_DMA_MCBSP2_TX 33 /* SDMA_32 */ +#define OMAP24XX_DMA_MCBSP2_RX 34 /* SDMA_33 */ +#define OMAP24XX_DMA_SPI1_TX0 35 /* SDMA_34 */ +#define OMAP24XX_DMA_SPI1_RX0 36 /* SDMA_35 */ +#define OMAP24XX_DMA_SPI1_TX1 37 /* SDMA_36 */ +#define OMAP24XX_DMA_SPI1_RX1 38 /* SDMA_37 */ +#define OMAP24XX_DMA_SPI1_TX2 39 /* SDMA_38 */ +#define OMAP24XX_DMA_SPI1_RX2 40 /* SDMA_39 */ +#define OMAP24XX_DMA_SPI1_TX3 41 /* SDMA_40 */ +#define OMAP24XX_DMA_SPI1_RX3 42 /* SDMA_41 */ +#define OMAP24XX_DMA_SPI2_TX0 43 /* SDMA_42 */ +#define OMAP24XX_DMA_SPI2_RX0 44 /* SDMA_43 */ +#define OMAP24XX_DMA_SPI2_TX1 45 /* SDMA_44 */ +#define OMAP24XX_DMA_SPI2_RX1 46 /* SDMA_45 */ +#define OMAP24XX_DMA_UART1_TX 49 /* SDMA_48 */ +#define OMAP24XX_DMA_UART1_RX 50 /* SDMA_49 */ +#define OMAP24XX_DMA_UART2_TX 51 /* SDMA_50 */ +#define OMAP24XX_DMA_UART2_RX 52 /* SDMA_51 */ +#define OMAP24XX_DMA_UART3_TX 53 /* SDMA_52 */ +#define OMAP24XX_DMA_UART3_RX 54 /* SDMA_53 */ +#define OMAP24XX_DMA_USB_W2FC_TX0 55 /* SDMA_54 */ +#define OMAP24XX_DMA_USB_W2FC_RX0 56 /* SDMA_55 */ +#define OMAP24XX_DMA_USB_W2FC_TX1 57 /* SDMA_56 */ +#define OMAP24XX_DMA_USB_W2FC_RX1 58 /* SDMA_57 */ +#define OMAP24XX_DMA_USB_W2FC_TX2 59 /* SDMA_58 */ +#define OMAP24XX_DMA_USB_W2FC_RX2 60 /* SDMA_59 */ +#define OMAP24XX_DMA_MMC1_TX 61 /* SDMA_60 */ +#define OMAP24XX_DMA_MMC1_RX 62 /* SDMA_61 */ +#define OMAP24XX_DMA_MS 63 /* SDMA_62 */ + +/*----------------------------------------------------------------------------*/ + +/* Hardware registers for LCD DMA */ #define OMAP1510_DMA_LCD_BASE (0xfffedb00) #define OMAP1510_DMA_LCD_CTRL (OMAP1510_DMA_LCD_BASE + 0x00) #define OMAP1510_DMA_LCD_TOP_F1_L (OMAP1510_DMA_LCD_BASE + 0x02) @@ -116,7 +259,7 @@ #define OMAP1510_DMA_LCD_BOT_F1_U (OMAP1510_DMA_LCD_BASE + 0x08) #define OMAP1610_DMA_LCD_BASE (0xfffee300) -#define OMAP1610_DMA_LCD_CSDP (OMAP1610_DMA_LCD_BASE + 0xc0) +#define OMAP1610_DMA_LCD_CSDP (OMAP1610_DMA_LCD_BASE + 0xc0) #define OMAP1610_DMA_LCD_CCR (OMAP1610_DMA_LCD_BASE + 0xc2) #define OMAP1610_DMA_LCD_CTRL (OMAP1610_DMA_LCD_BASE + 0xc4) #define OMAP1610_DMA_LCD_TOP_B1_L (OMAP1610_DMA_LCD_BASE + 0xc8) @@ -134,37 +277,18 @@ #define OMAP1610_DMA_LCD_LCH_CTRL (OMAP1610_DMA_LCD_BASE + 0xea) #define OMAP1610_DMA_LCD_SRC_FI_B1_U (OMAP1610_DMA_LCD_BASE + 0xf4) - -/* Every LCh has its own set of the registers below */ -#define OMAP_DMA_CSDP(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x00) -#define OMAP_DMA_CCR(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x02) -#define OMAP_DMA_CICR(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x04) -#define OMAP_DMA_CSR(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x06) -#define OMAP_DMA_CSSA_L(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x08) -#define OMAP_DMA_CSSA_U(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x0a) -#define OMAP_DMA_CDSA_L(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x0c) -#define OMAP_DMA_CDSA_U(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x0e) -#define OMAP_DMA_CEN(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x10) -#define OMAP_DMA_CFN(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x12) -#define OMAP_DMA_CSFI(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x14) -#define OMAP_DMA_CSEI(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x16) -#define OMAP_DMA_CSAC(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x18) -#define OMAP_DMA_CDAC(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x1a) -#define OMAP_DMA_CDEI(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x1c) -#define OMAP_DMA_CDFI(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x1e) -#define OMAP_DMA_COLOR_L(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x20) -#define OMAP_DMA_COLOR_U(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x22) -#define OMAP_DMA_CCR2(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x24) -#define OMAP_DMA_CLNK_CTRL(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x28) -#define OMAP_DMA_LCH_CTRL(n) (OMAP_DMA_BASE + 0x40 * (n) + 0x2a) - -#define OMAP_DMA_TOUT_IRQ (1 << 0) +#define OMAP_DMA_TOUT_IRQ (1 << 0) /* Only on omap1 */ #define OMAP_DMA_DROP_IRQ (1 << 1) #define OMAP_DMA_HALF_IRQ (1 << 2) #define OMAP_DMA_FRAME_IRQ (1 << 3) #define OMAP_DMA_LAST_IRQ (1 << 4) #define OMAP_DMA_BLOCK_IRQ (1 << 5) -#define OMAP_DMA_SYNC_IRQ (1 << 6) +#define OMAP1_DMA_SYNC_IRQ (1 << 6) +#define OMAP2_DMA_PKT_IRQ (1 << 7) +#define OMAP2_DMA_TRANS_ERR_IRQ (1 << 8) +#define OMAP2_DMA_SECURE_ERR_IRQ (1 << 9) +#define OMAP2_DMA_SUPERVISOR_ERR_IRQ (1 << 10) +#define OMAP2_DMA_MISALIGNED_ERR_IRQ (1 << 11) #define OMAP_DMA_DATA_TYPE_S8 0x00 #define OMAP_DMA_DATA_TYPE_S16 0x01 @@ -194,6 +318,7 @@ enum { OMAP_LCD_DMA_B2_BOTTOM }; +/* REVISIT: Check if BURST_4 is really 1 (or 2) */ enum omap_dma_burst_mode { OMAP_DMA_DATA_BURST_DIS = 0, OMAP_DMA_DATA_BURST_4, @@ -206,6 +331,31 @@ enum omap_dma_color_mode { OMAP_DMA_TRANSPARENT_COPY }; +struct omap_dma_channel_params { + int data_type; /* data type 8,16,32 */ + int elem_count; /* number of elements in a frame */ + int frame_count; /* number of frames in a element */ + + int src_port; /* Only on OMAP1 REVISIT: Is this needed? */ + int src_amode; /* constant , post increment, indexed , double indexed */ + int src_start; /* source address : physical */ + int src_ei; /* source element index */ + int src_fi; /* source frame index */ + + int dst_port; /* Only on OMAP1 REVISIT: Is this needed? */ + int dst_amode; /* constant , post increment, indexed , double indexed */ + int dst_start; /* source address : physical */ + int dst_ei; /* source element index */ + int dst_fi; /* source frame index */ + + int trigger; /* trigger attached if the channel is synchronized */ + int sync_mode; /* sycn on element, frame , block or packet */ + int src_or_dst_synch; /* source synch(1) or destination synch(0) */ + + int ie; /* interrupt enabled */ +}; + + extern void omap_set_dma_priority(int dst_port, int priority); extern int omap_request_dma(int dev_id, const char *dev_name, void (* callback)(int lch, u16 ch_status, void *data), @@ -217,24 +367,30 @@ extern void omap_start_dma(int lch); extern void omap_stop_dma(int lch); extern void omap_set_dma_transfer_params(int lch, int data_type, int elem_count, int frame_count, - int sync_mode); + int sync_mode, + int dma_trigger, int src_or_dst_synch); extern void omap_set_dma_color_mode(int lch, enum omap_dma_color_mode mode, u32 color); extern void omap_set_dma_src_params(int lch, int src_port, int src_amode, - unsigned long src_start); + unsigned long src_start, + int src_ei, int src_fi); extern void omap_set_dma_src_index(int lch, int eidx, int fidx); extern void omap_set_dma_src_data_pack(int lch, int enable); extern void omap_set_dma_src_burst_mode(int lch, enum omap_dma_burst_mode burst_mode); extern void omap_set_dma_dest_params(int lch, int dest_port, int dest_amode, - unsigned long dest_start); + unsigned long dest_start, + int dst_ei, int dst_fi); extern void omap_set_dma_dest_index(int lch, int eidx, int fidx); extern void omap_set_dma_dest_data_pack(int lch, int enable); extern void omap_set_dma_dest_burst_mode(int lch, enum omap_dma_burst_mode burst_mode); +extern void omap_set_dma_params(int lch, + struct omap_dma_channel_params * params); + extern void omap_dma_link_lch (int lch_head, int lch_queue); extern void omap_dma_unlink_lch (int lch_head, int lch_queue); @@ -244,9 +400,6 @@ extern int omap_get_dma_src_addr_counter(int lch); extern void omap_clear_dma(int lch); extern int omap_dma_running(void); -/* Returns 1 if the DMA module is in OMAP1510-compatible mode, 0 otherwise */ -extern int omap_dma_in_1510_mode(void); - /* LCD DMA functions */ extern int omap_request_lcd_dma(void (* callback)(u16 status, void *data), void *data); diff --git a/include/asm-arm/arch-omap/entry-macro.S b/include/asm-arm/arch-omap/entry-macro.S index 0d29b9c56a95..f8814a84910e 100644 --- a/include/asm-arm/arch-omap/entry-macro.S +++ b/include/asm-arm/arch-omap/entry-macro.S @@ -10,6 +10,20 @@ #if defined(CONFIG_ARCH_OMAP1) +#if defined(CONFIG_ARCH_OMAP730) && \ + (defined(CONFIG_ARCH_OMAP15XX) || defined(CONFIG_ARCH_OMAP16XX)) +#error "FIXME: OMAP730 doesn't support multiple-OMAP" +#elif defined(CONFIG_ARCH_OMAP730) +#define INT_IH2_IRQ INT_730_IH2_IRQ +#elif defined(CONFIG_ARCH_OMAP15XX) +#define INT_IH2_IRQ INT_1510_IH2_IRQ +#elif defined(CONFIG_ARCH_OMAP16XX) +#define INT_IH2_IRQ INT_1610_IH2_IRQ +#else +#warning "IH2 IRQ defaulted" +#define INT_IH2_IRQ INT_1510_IH2_IRQ +#endif + .macro disable_fiq .endm diff --git a/include/asm-arm/arch-omap/fpga.h b/include/asm-arm/arch-omap/fpga.h index 676807dc50e1..6a883e0bdbb8 100644 --- a/include/asm-arm/arch-omap/fpga.h +++ b/include/asm-arm/arch-omap/fpga.h @@ -19,7 +19,7 @@ #ifndef __ASM_ARCH_OMAP_FPGA_H #define __ASM_ARCH_OMAP_FPGA_H -#if defined(CONFIG_MACH_OMAP_INNOVATOR) && defined(CONFIG_ARCH_OMAP1510) +#if defined(CONFIG_MACH_OMAP_INNOVATOR) && defined(CONFIG_ARCH_OMAP15XX) extern void omap1510_fpga_init_irq(void); #else #define omap1510_fpga_init_irq() (0) @@ -77,6 +77,8 @@ struct h2p2_dbg_fpga { #define H2P2_DBG_FPGA_LOAD_METER_SIZE 11 #define H2P2_DBG_FPGA_LOAD_METER_MASK ((1 << H2P2_DBG_FPGA_LOAD_METER_SIZE) - 1) +#define H2P2_DBG_FPGA_P2_LED_TIMER (1 << 0) +#define H2P2_DBG_FPGA_P2_LED_IDLE (1 << 1) /* * --------------------------------------------------------------------------- diff --git a/include/asm-arm/arch-omap/gpio.h b/include/asm-arm/arch-omap/gpio.h index 74cb2b93b700..1b3885741ac1 100644 --- a/include/asm-arm/arch-omap/gpio.h +++ b/include/asm-arm/arch-omap/gpio.h @@ -67,7 +67,7 @@ #define OMAP_GPIO_IRQ(nr) (OMAP_GPIO_IS_MPUIO(nr) ? \ IH_MPUIO_BASE + ((nr) & 0x0f) : \ - IH_GPIO_BASE + ((nr) & 0x3f)) + IH_GPIO_BASE + (nr)) extern int omap_gpio_init(void); /* Call from board init only */ extern int omap_request_gpio(int gpio); diff --git a/include/asm-arm/arch-omap/hardware.h b/include/asm-arm/arch-omap/hardware.h index 60201e1dd6ad..5406b875c422 100644 --- a/include/asm-arm/arch-omap/hardware.h +++ b/include/asm-arm/arch-omap/hardware.h @@ -267,8 +267,6 @@ #define OMAP_LPG2_LCR (OMAP_LPG2_BASE + 0x00) #define OMAP_LPG2_PMR (OMAP_LPG2_BASE + 0x04) -#ifndef __ASSEMBLER__ - /* * --------------------------------------------------------------------------- * Processor specific defines @@ -277,13 +275,11 @@ #include "omap730.h" #include "omap1510.h" - -#ifdef CONFIG_ARCH_OMAP24XX #include "omap24xx.h" -#endif - #include "omap16xx.h" +#ifndef __ASSEMBLER__ + /* * --------------------------------------------------------------------------- * Board specific defines diff --git a/include/asm-arm/arch-omap/io.h b/include/asm-arm/arch-omap/io.h index 3d5bcd545082..f5bcc9a1aed6 100644 --- a/include/asm-arm/arch-omap/io.h +++ b/include/asm-arm/arch-omap/io.h @@ -52,23 +52,33 @@ * ---------------------------------------------------------------------------- */ +#define PCIO_BASE 0 + #if defined(CONFIG_ARCH_OMAP1) + #define IO_PHYS 0xFFFB0000 -#define IO_OFFSET -0x01000000 /* Virtual IO = 0xfefb0000 */ +#define IO_OFFSET 0x01000000 /* Virtual IO = 0xfefb0000 */ #define IO_SIZE 0x40000 +#define IO_VIRT (IO_PHYS - IO_OFFSET) +#define IO_ADDRESS(pa) ((pa) - IO_OFFSET) +#define io_p2v(pa) ((pa) - IO_OFFSET) +#define io_v2p(va) ((va) + IO_OFFSET) #elif defined(CONFIG_ARCH_OMAP2) -#define IO_PHYS 0x48000000 /* L4 peripherals; other stuff has to be mapped * - * manually. */ -#define IO_OFFSET 0x90000000 /* Virtual IO = 0xd8000000 */ -#define IO_SIZE 0x08000000 -#endif -#define IO_VIRT (IO_PHYS + IO_OFFSET) -#define IO_ADDRESS(x) ((x) + IO_OFFSET) -#define PCIO_BASE 0 -#define io_p2v(x) ((x) + IO_OFFSET) -#define io_v2p(x) ((x) - IO_OFFSET) +/* We map both L3 and L4 on OMAP2 */ +#define L3_24XX_PHYS L3_24XX_BASE /* 0x68000000 */ +#define L3_24XX_VIRT 0xf8000000 +#define L3_24XX_SIZE SZ_1M /* 44kB of 128MB used, want 1MB sect */ +#define L4_24XX_PHYS L4_24XX_BASE /* 0x48000000 */ +#define L4_24XX_VIRT 0xd8000000 +#define L4_24XX_SIZE SZ_1M /* 1MB of 128MB used, want 1MB sect */ +#define IO_OFFSET 0x90000000 +#define IO_ADDRESS(pa) ((pa) + IO_OFFSET) /* Works for L3 and L4 */ +#define io_p2v(pa) ((pa) + IO_OFFSET) /* Works for L3 and L4 */ +#define io_v2p(va) ((va) - IO_OFFSET) /* Works for L3 and L4 */ + +#endif #ifndef __ASSEMBLER__ diff --git a/include/asm-arm/arch-omap/irqs.h b/include/asm-arm/arch-omap/irqs.h index 74e108ccac16..9779686bdceb 100644 --- a/include/asm-arm/arch-omap/irqs.h +++ b/include/asm-arm/arch-omap/irqs.h @@ -22,8 +22,8 @@ * are different. */ -#ifndef __ASM_ARCH_OMAP1510_IRQS_H -#define __ASM_ARCH_OMAP1510_IRQS_H +#ifndef __ASM_ARCH_OMAP15XX_IRQS_H +#define __ASM_ARCH_OMAP15XX_IRQS_H /* * IRQ numbers for interrupt handler 1 @@ -31,7 +31,6 @@ * NOTE: See also the OMAP-1510 and 1610 specific IRQ numbers below * */ -#define INT_IH2_IRQ 0 #define INT_CAMERA 1 #define INT_FIQ 3 #define INT_RTDX 6 @@ -60,6 +59,7 @@ /* * OMAP-1510 specific IRQ numbers for interrupt handler 1 */ +#define INT_1510_IH2_IRQ 0 #define INT_1510_RES2 2 #define INT_1510_SPI_TX 4 #define INT_1510_SPI_RX 5 @@ -71,6 +71,7 @@ /* * OMAP-1610 specific IRQ numbers for interrupt handler 1 */ +#define INT_1610_IH2_IRQ 0 #define INT_1610_IH2_FIQ 2 #define INT_1610_McBSP2_TX 4 #define INT_1610_McBSP2_RX 5 @@ -231,6 +232,12 @@ #define INT_730_DMA_CH15 (62 + IH2_BASE) #define INT_730_NAND (63 + IH2_BASE) +#define INT_24XX_SYS_NIRQ 7 +#define INT_24XX_SDMA_IRQ0 12 +#define INT_24XX_SDMA_IRQ1 13 +#define INT_24XX_SDMA_IRQ2 14 +#define INT_24XX_SDMA_IRQ3 15 +#define INT_24XX_DSS_IRQ 25 #define INT_24XX_GPIO_BANK1 29 #define INT_24XX_GPIO_BANK2 30 #define INT_24XX_GPIO_BANK3 31 diff --git a/include/asm-arm/arch-omap/memory.h b/include/asm-arm/arch-omap/memory.h index bf545b6e0a26..df50dd53e1dd 100644 --- a/include/asm-arm/arch-omap/memory.h +++ b/include/asm-arm/arch-omap/memory.h @@ -61,7 +61,7 @@ * Note that the is_lbus_device() test is not very efficient on 1510 * because of the strncmp(). */ -#ifdef CONFIG_ARCH_OMAP1510 +#ifdef CONFIG_ARCH_OMAP15XX /* * OMAP-1510 Local Bus address offset @@ -84,7 +84,7 @@ virt_to_lbus(addr) : \ __virt_to_bus(addr);}) -#endif /* CONFIG_ARCH_OMAP1510 */ +#endif /* CONFIG_ARCH_OMAP15XX */ #endif diff --git a/include/asm-arm/arch-omap/menelaus.h b/include/asm-arm/arch-omap/menelaus.h new file mode 100644 index 000000000000..46be8b8d6346 --- /dev/null +++ b/include/asm-arm/arch-omap/menelaus.h @@ -0,0 +1,22 @@ +/* + * linux/include/asm-arm/arch-omap/menelaus.h + * + * Functions to access Menelaus power management chip + */ + +#ifndef __ASM_ARCH_MENELAUS_H +#define __ASM_ARCH_MENELAUS_H + +extern void menelaus_mmc_register(void (*callback)(u8 card_mask), + unsigned long data); +extern void menelaus_mmc_remove(void); +extern void menelaus_mmc_opendrain(int enable); + +#if defined(CONFIG_ARCH_OMAP24XX) && defined(CONFIG_MENELAUS) +#define omap_has_menelaus() 1 +#else +#define omap_has_menelaus() 0 +#endif + +#endif + diff --git a/include/asm-arm/arch-omap/mux.h b/include/asm-arm/arch-omap/mux.h index 1b1ad4105349..13415a9aab06 100644 --- a/include/asm-arm/arch-omap/mux.h +++ b/include/asm-arm/arch-omap/mux.h @@ -4,7 +4,7 @@ * Table of the Omap register configurations for the FUNC_MUX and * PULL_DWN combinations. * - * Copyright (C) 2003 Nokia Corporation + * Copyright (C) 2003 - 2005 Nokia Corporation * * Written by Tony Lindgren * @@ -58,6 +58,16 @@ .pu_pd_reg = PU_PD_SEL_##reg, \ .pu_pd_val = status, +#define MUX_REG_730(reg, mode_offset, mode) .mux_reg_name = "OMAP730_IO_CONF_"#reg, \ + .mux_reg = OMAP730_IO_CONF_##reg, \ + .mask_offset = mode_offset, \ + .mask = mode, + +#define PULL_REG_730(reg, bit, status) .pull_name = "OMAP730_IO_CONF_"#reg, \ + .pull_reg = OMAP730_IO_CONF_##reg, \ + .pull_bit = bit, \ + .pull_val = status, + #else #define MUX_REG(reg, mode_offset, mode) .mux_reg = FUNC_MUX_CTRL_##reg, \ @@ -71,6 +81,15 @@ #define PU_PD_REG(reg, status) .pu_pd_reg = PU_PD_SEL_##reg, \ .pu_pd_val = status, +#define MUX_REG_730(reg, mode_offset, mode) \ + .mux_reg = OMAP730_IO_CONF_##reg, \ + .mask_offset = mode_offset, \ + .mask = mode, + +#define PULL_REG_730(reg, bit, status) .pull_reg = OMAP730_IO_CONF_##reg, \ + .pull_bit = bit, \ + .pull_val = status, + #endif /* CONFIG_OMAP_MUX_DEBUG */ #define MUX_CFG(desc, mux_reg, mode_offset, mode, \ @@ -84,13 +103,44 @@ PU_PD_REG(pu_pd_reg, pu_pd_status) \ }, + +/* + * OMAP730 has a slightly different config for the pin mux. + * - config regs are the OMAP730_IO_CONF_x regs (see omap730.h) regs and + * not the FUNC_MUX_CTRL_x regs from hardware.h + * - for pull-up/down, only has one enable bit which is is in the same register + * as mux config + */ +#define MUX_CFG_730(desc, mux_reg, mode_offset, mode, \ + pull_reg, pull_bit, pull_status, \ + pu_pd_reg, pu_pd_status, debug_status)\ +{ \ + .name = desc, \ + .debug = debug_status, \ + MUX_REG_730(mux_reg, mode_offset, mode) \ + PULL_REG_730(mux_reg, pull_bit, pull_status) \ + PU_PD_REG(pu_pd_reg, pu_pd_status) \ +}, + +#define MUX_CFG_24XX(desc, reg_offset, mode, \ + pull_en, pull_mode, dbg) \ +{ \ + .name = desc, \ + .debug = dbg, \ + .mux_reg = reg_offset, \ + .mask = mode, \ + .pull_val = pull_en, \ + .pu_pd_val = pull_mode, \ +}, + + #define PULL_DISABLED 0 #define PULL_ENABLED 1 #define PULL_DOWN 0 #define PULL_UP 1 -typedef struct { +struct pin_config { char *name; unsigned char busy; unsigned char debug; @@ -108,13 +158,23 @@ typedef struct { const char *pu_pd_name; const unsigned int pu_pd_reg; const unsigned char pu_pd_val; -} reg_cfg_set; +}; -/* - * Lookup table for FUNC_MUX and PULL_DWN register combinations for each - * device. See also reg_cfg_table below for the register values. - */ -typedef enum { +enum omap730_index { + /* OMAP 730 keyboard */ + E2_730_KBR0, + J7_730_KBR1, + E1_730_KBR2, + F3_730_KBR3, + D2_730_KBR4, + C2_730_KBC0, + D3_730_KBC1, + E4_730_KBC2, + F4_730_KBC3, + E3_730_KBC4, +}; + +enum omap1xxx_index { /* UART1 (BT_UART_GATING)*/ UART1_TX = 0, UART1_RTS, @@ -331,245 +391,34 @@ typedef enum { V10_1610_CF_IREQ, W10_1610_CF_RESET, W11_1610_CF_CD1, -} reg_cfg_t; - -#if defined(__MUX_C__) && defined(CONFIG_OMAP_MUX) - -/* - * Table of various FUNC_MUX and PULL_DWN combinations for each device. - * See also reg_cfg_t above for the lookup table. - */ -static const reg_cfg_set __initdata_or_module -reg_cfg_table[] = { -/* - * description mux mode mux pull pull pull pu_pd pu dbg - * reg offset mode reg bit ena reg - */ -MUX_CFG("UART1_TX", 9, 21, 1, 2, 3, 0, NA, 0, 0) -MUX_CFG("UART1_RTS", 9, 12, 1, 2, 0, 0, NA, 0, 0) - -/* UART2 (COM_UART_GATING), conflicts with USB2 */ -MUX_CFG("UART2_TX", C, 27, 1, 3, 3, 0, NA, 0, 0) -MUX_CFG("UART2_RX", C, 18, 0, 3, 1, 1, NA, 0, 0) -MUX_CFG("UART2_CTS", C, 21, 0, 3, 1, 1, NA, 0, 0) -MUX_CFG("UART2_RTS", C, 24, 1, 3, 2, 0, NA, 0, 0) - -/* UART3 (GIGA_UART_GATING) */ -MUX_CFG("UART3_TX", 6, 0, 1, 0, 30, 0, NA, 0, 0) -MUX_CFG("UART3_RX", 6, 3, 0, 0, 31, 1, NA, 0, 0) -MUX_CFG("UART3_CTS", 5, 12, 2, 0, 24, 0, NA, 0, 0) -MUX_CFG("UART3_RTS", 5, 15, 2, 0, 25, 0, NA, 0, 0) -MUX_CFG("UART3_CLKREQ", 9, 27, 0, 2, 5, 0, NA, 0, 0) -MUX_CFG("UART3_BCLK", A, 0, 0, 2, 6, 0, NA, 0, 0) -MUX_CFG("Y15_1610_UART3_RTS", A, 0, 1, 2, 6, 0, NA, 0, 0) - -/* PWT & PWL, conflicts with UART3 */ -MUX_CFG("PWT", 6, 0, 2, 0, 30, 0, NA, 0, 0) -MUX_CFG("PWL", 6, 3, 1, 0, 31, 1, NA, 0, 0) - -/* USB internal master generic */ -MUX_CFG("R18_USB_VBUS", 7, 9, 2, 1, 11, 0, NA, 0, 1) -MUX_CFG("R18_1510_USB_GPIO0", 7, 9, 0, 1, 11, 1, NA, 0, 1) -/* works around erratum: W4_USB_PUEN and W4_USB_PUDIS are switched! */ -MUX_CFG("W4_USB_PUEN", D, 3, 3, 3, 5, 1, NA, 0, 1) -MUX_CFG("W4_USB_CLKO", D, 3, 1, 3, 5, 0, NA, 0, 1) -MUX_CFG("W4_USB_HIGHZ", D, 3, 4, 3, 5, 0, 3, 0, 1) -MUX_CFG("W4_GPIO58", D, 3, 7, 3, 5, 0, 3, 0, 1) - -/* USB1 master */ -MUX_CFG("USB1_SUSP", 8, 27, 2, 1, 27, 0, NA, 0, 1) -MUX_CFG("USB1_SE0", 9, 0, 2, 1, 28, 0, NA, 0, 1) -MUX_CFG("W13_1610_USB1_SE0", 9, 0, 4, 1, 28, 0, NA, 0, 1) -MUX_CFG("USB1_TXEN", 9, 3, 2, 1, 29, 0, NA, 0, 1) -MUX_CFG("USB1_TXD", 9, 24, 1, 2, 4, 0, NA, 0, 1) -MUX_CFG("USB1_VP", A, 3, 1, 2, 7, 0, NA, 0, 1) -MUX_CFG("USB1_VM", A, 6, 1, 2, 8, 0, NA, 0, 1) -MUX_CFG("USB1_RCV", A, 9, 1, 2, 9, 0, NA, 0, 1) -MUX_CFG("USB1_SPEED", A, 12, 2, 2, 10, 0, NA, 0, 1) -MUX_CFG("R13_1610_USB1_SPEED", A, 12, 5, 2, 10, 0, NA, 0, 1) -MUX_CFG("R13_1710_USB1_SEO", A, 12, 5, 2, 10, 0, NA, 0, 1) - -/* USB2 master */ -MUX_CFG("USB2_SUSP", B, 3, 1, 2, 17, 0, NA, 0, 1) -MUX_CFG("USB2_VP", B, 6, 1, 2, 18, 0, NA, 0, 1) -MUX_CFG("USB2_TXEN", B, 9, 1, 2, 19, 0, NA, 0, 1) -MUX_CFG("USB2_VM", C, 18, 1, 3, 0, 0, NA, 0, 1) -MUX_CFG("USB2_RCV", C, 21, 1, 3, 1, 0, NA, 0, 1) -MUX_CFG("USB2_SE0", C, 24, 2, 3, 2, 0, NA, 0, 1) -MUX_CFG("USB2_TXD", C, 27, 2, 3, 3, 0, NA, 0, 1) - -/* OMAP-1510 GPIO */ -MUX_CFG("R18_1510_GPIO0", 7, 9, 0, 1, 11, 1, 0, 0, 1) -MUX_CFG("R19_1510_GPIO1", 7, 6, 0, 1, 10, 1, 0, 0, 1) -MUX_CFG("M14_1510_GPIO2", 7, 3, 0, 1, 9, 1, 0, 0, 1) - -/* OMAP1610 GPIO */ -MUX_CFG("P18_1610_GPIO3", 7, 0, 0, 1, 8, 0, NA, 0, 1) -MUX_CFG("Y15_1610_GPIO17", A, 0, 7, 2, 6, 0, NA, 0, 1) - -/* OMAP-1710 GPIO */ -MUX_CFG("R18_1710_GPIO0", 7, 9, 0, 1, 11, 1, 1, 1, 1) -MUX_CFG("V2_1710_GPIO10", F, 27, 1, 4, 3, 1, 4, 1, 1) -MUX_CFG("N21_1710_GPIO14", 6, 9, 0, 1, 1, 1, 1, 1, 1) -MUX_CFG("W15_1710_GPIO40", 9, 27, 7, 2, 5, 1, 2, 1, 1) - -/* MPUIO */ -MUX_CFG("MPUIO2", 7, 18, 0, 1, 14, 1, NA, 0, 1) -MUX_CFG("N15_1610_MPUIO2", 7, 18, 0, 1, 14, 1, 1, 0, 1) -MUX_CFG("MPUIO4", 7, 15, 0, 1, 13, 1, NA, 0, 1) -MUX_CFG("MPUIO5", 7, 12, 0, 1, 12, 1, NA, 0, 1) - -MUX_CFG("T20_1610_MPUIO5", 7, 12, 0, 1, 12, 0, 3, 0, 1) -MUX_CFG("W11_1610_MPUIO6", 10, 15, 2, 3, 8, 0, 3, 0, 1) -MUX_CFG("V10_1610_MPUIO7", A, 24, 2, 2, 14, 0, 2, 0, 1) -MUX_CFG("W11_1610_MPUIO9", 10, 15, 1, 3, 8, 0, 3, 0, 1) -MUX_CFG("V10_1610_MPUIO10", A, 24, 1, 2, 14, 0, 2, 0, 1) -MUX_CFG("W10_1610_MPUIO11", A, 18, 2, 2, 11, 0, 2, 0, 1) -MUX_CFG("E20_1610_MPUIO13", 3, 21, 1, 0, 7, 0, 0, 0, 1) -MUX_CFG("U20_1610_MPUIO14", 9, 6, 6, 0, 30, 0, 0, 0, 1) -MUX_CFG("E19_1610_MPUIO15", 3, 18, 1, 0, 6, 0, 0, 0, 1) - -/* MCBSP2 */ -MUX_CFG("MCBSP2_CLKR", C, 6, 0, 2, 27, 1, NA, 0, 1) -MUX_CFG("MCBSP2_CLKX", C, 9, 0, 2, 29, 1, NA, 0, 1) -MUX_CFG("MCBSP2_DR", C, 0, 0, 2, 26, 1, NA, 0, 1) -MUX_CFG("MCBSP2_DX", C, 15, 0, 2, 31, 1, NA, 0, 1) -MUX_CFG("MCBSP2_FSR", C, 12, 0, 2, 30, 1, NA, 0, 1) -MUX_CFG("MCBSP2_FSX", C, 3, 0, 2, 27, 1, NA, 0, 1) - -/* MCBSP3 NOTE: Mode must 1 for clock */ -MUX_CFG("MCBSP3_CLKX", 9, 3, 1, 1, 29, 0, NA, 0, 1) - -/* Misc ballouts */ -MUX_CFG("BALLOUT_V8_ARMIO3", B, 18, 0, 2, 25, 1, NA, 0, 1) -MUX_CFG("N20_HDQ", 6, 18, 1, 1, 4, 0, 1, 4, 0) - -/* OMAP-1610 MMC2 */ -MUX_CFG("W8_1610_MMC2_DAT0", B, 21, 6, 2, 23, 1, 2, 1, 1) -MUX_CFG("V8_1610_MMC2_DAT1", B, 27, 6, 2, 25, 1, 2, 1, 1) -MUX_CFG("W15_1610_MMC2_DAT2", 9, 12, 6, 2, 5, 1, 2, 1, 1) -MUX_CFG("R10_1610_MMC2_DAT3", B, 18, 6, 2, 22, 1, 2, 1, 1) -MUX_CFG("Y10_1610_MMC2_CLK", B, 3, 6, 2, 17, 0, 2, 0, 1) -MUX_CFG("Y8_1610_MMC2_CMD", B, 24, 6, 2, 24, 1, 2, 1, 1) -MUX_CFG("V9_1610_MMC2_CMDDIR", B, 12, 6, 2, 20, 0, 2, 1, 1) -MUX_CFG("V5_1610_MMC2_DATDIR0", B, 15, 6, 2, 21, 0, 2, 1, 1) -MUX_CFG("W19_1610_MMC2_DATDIR1", 8, 15, 6, 1, 23, 0, 1, 1, 1) -MUX_CFG("R18_1610_MMC2_CLKIN", 7, 9, 6, 1, 11, 0, 1, 11, 1) - -/* OMAP-1610 External Trace Interface */ -MUX_CFG("M19_1610_ETM_PSTAT0", 5, 27, 1, 0, 29, 0, 0, 0, 1) -MUX_CFG("L15_1610_ETM_PSTAT1", 5, 24, 1, 0, 28, 0, 0, 0, 1) -MUX_CFG("L18_1610_ETM_PSTAT2", 5, 21, 1, 0, 27, 0, 0, 0, 1) -MUX_CFG("L19_1610_ETM_D0", 5, 18, 1, 0, 26, 0, 0, 0, 1) -MUX_CFG("J19_1610_ETM_D6", 5, 0, 1, 0, 20, 0, 0, 0, 1) -MUX_CFG("J18_1610_ETM_D7", 5, 27, 1, 0, 19, 0, 0, 0, 1) - -/* OMAP16XX GPIO */ -MUX_CFG("P20_1610_GPIO4", 6, 27, 0, 1, 7, 0, 1, 1, 1) -MUX_CFG("V9_1610_GPIO7", B, 12, 1, 2, 20, 0, 2, 1, 1) -MUX_CFG("W8_1610_GPIO9", B, 21, 0, 2, 23, 0, 2, 1, 1) -MUX_CFG("N20_1610_GPIO11", 6, 18, 0, 1, 4, 0, 1, 1, 1) -MUX_CFG("N19_1610_GPIO13", 6, 12, 0, 1, 2, 0, 1, 1, 1) -MUX_CFG("P10_1610_GPIO22", C, 0, 7, 2, 26, 0, 2, 1, 1) -MUX_CFG("V5_1610_GPIO24", B, 15, 7, 2, 21, 0, 2, 1, 1) -MUX_CFG("AA20_1610_GPIO_41", 9, 9, 7, 1, 31, 0, 1, 1, 1) -MUX_CFG("W19_1610_GPIO48", 8, 15, 7, 1, 23, 1, 1, 0, 1) -MUX_CFG("M7_1610_GPIO62", 10, 0, 0, 4, 24, 0, 4, 0, 1) -MUX_CFG("V14_16XX_GPIO37", 9, 18, 7, 2, 2, 0, 2, 2, 0) -MUX_CFG("R9_16XX_GPIO18", C, 18, 7, 3, 0, 0, 3, 0, 0) -MUX_CFG("L14_16XX_GPIO49", 6, 3, 7, 0, 31, 0, 0, 31, 0) - -/* OMAP-1610 uWire */ -MUX_CFG("V19_1610_UWIRE_SCLK", 8, 6, 0, 1, 20, 0, 1, 1, 1) -MUX_CFG("U18_1610_UWIRE_SDI", 8, 0, 0, 1, 18, 0, 1, 1, 1) -MUX_CFG("W21_1610_UWIRE_SDO", 8, 3, 0, 1, 19, 0, 1, 1, 1) -MUX_CFG("N14_1610_UWIRE_CS0", 8, 9, 1, 1, 21, 0, 1, 1, 1) -MUX_CFG("P15_1610_UWIRE_CS3", 8, 12, 1, 1, 22, 0, 1, 1, 1) -MUX_CFG("N15_1610_UWIRE_CS1", 7, 18, 2, 1, 14, 0, NA, 0, 1) - -/* OMAP-1610 Flash */ -MUX_CFG("L3_1610_FLASH_CS2B_OE",10, 6, 1, NA, 0, 0, NA, 0, 1) -MUX_CFG("M8_1610_FLASH_CS2B_WE",10, 3, 1, NA, 0, 0, NA, 0, 1) - -/* First MMC interface, same on 1510, 1610 and 1710 */ -MUX_CFG("MMC_CMD", A, 27, 0, 2, 15, 1, 2, 1, 1) -MUX_CFG("MMC_DAT1", A, 24, 0, 2, 14, 1, 2, 1, 1) -MUX_CFG("MMC_DAT2", A, 18, 0, 2, 12, 1, 2, 1, 1) -MUX_CFG("MMC_DAT0", B, 0, 0, 2, 16, 1, 2, 1, 1) -MUX_CFG("MMC_CLK", A, 21, 0, NA, 0, 0, NA, 0, 1) -MUX_CFG("MMC_DAT3", 10, 15, 0, 3, 8, 1, 3, 1, 1) -MUX_CFG("M15_1710_MMC_CLKI", 6, 21, 2, 0, 0, 0, NA, 0, 1) -MUX_CFG("P19_1710_MMC_CMDDIR", 6, 24, 6, 0, 0, 0, NA, 0, 1) -MUX_CFG("P20_1710_MMC_DATDIR0", 6, 27, 5, 0, 0, 0, NA, 0, 1) - -/* OMAP-1610 USB0 alternate configuration */ -MUX_CFG("W9_USB0_TXEN", B, 9, 5, 2, 19, 0, 2, 0, 1) -MUX_CFG("AA9_USB0_VP", B, 6, 5, 2, 18, 0, 2, 0, 1) -MUX_CFG("Y5_USB0_RCV", C, 21, 5, 3, 1, 0, 1, 0, 1) -MUX_CFG("R9_USB0_VM", C, 18, 5, 3, 0, 0, 3, 0, 1) -MUX_CFG("V6_USB0_TXD", C, 27, 5, 3, 3, 0, 3, 0, 1) -MUX_CFG("W5_USB0_SE0", C, 24, 5, 3, 2, 0, 3, 0, 1) -MUX_CFG("V9_USB0_SPEED", B, 12, 5, 2, 20, 0, 2, 0, 1) -MUX_CFG("Y10_USB0_SUSP", B, 3, 5, 2, 17, 0, 2, 0, 1) - -/* USB2 interface */ -MUX_CFG("W9_USB2_TXEN", B, 9, 1, NA, 0, 0, NA, 0, 1) -MUX_CFG("AA9_USB2_VP", B, 6, 1, NA, 0, 0, NA, 0, 1) -MUX_CFG("Y5_USB2_RCV", C, 21, 1, NA, 0, 0, NA, 0, 1) -MUX_CFG("R9_USB2_VM", C, 18, 1, NA, 0, 0, NA, 0, 1) -MUX_CFG("V6_USB2_TXD", C, 27, 2, NA, 0, 0, NA, 0, 1) -MUX_CFG("W5_USB2_SE0", C, 24, 2, NA, 0, 0, NA, 0, 1) - -/* 16XX UART */ -MUX_CFG("R13_1610_UART1_TX", A, 12, 6, 2, 10, 0, 2, 10, 1) -MUX_CFG("V14_16XX_UART1_RX", 9, 18, 0, 2, 2, 0, 2, 2, 1) -MUX_CFG("R14_1610_UART1_CTS", 9, 15, 0, 2, 1, 0, 2, 1, 1) -MUX_CFG("AA15_1610_UART1_RTS", 9, 12, 1, 2, 0, 0, 2, 0, 1) -MUX_CFG("R9_16XX_UART2_RX", C, 18, 0, 3, 0, 0, 3, 0, 1) -MUX_CFG("L14_16XX_UART3_RX", 6, 3, 0, 0, 31, 0, 0, 31, 1) - -/* I2C interface */ -MUX_CFG("I2C_SCL", 7, 24, 0, NA, 0, 0, NA, 0, 0) -MUX_CFG("I2C_SDA", 7, 27, 0, NA, 0, 0, NA, 0, 0) - -/* Keypad */ -MUX_CFG("F18_1610_KBC0", 3, 15, 0, 0, 5, 1, 0, 0, 0) -MUX_CFG("D20_1610_KBC1", 3, 12, 0, 0, 4, 1, 0, 0, 0) -MUX_CFG("D19_1610_KBC2", 3, 9, 0, 0, 3, 1, 0, 0, 0) -MUX_CFG("E18_1610_KBC3", 3, 6, 0, 0, 2, 1, 0, 0, 0) -MUX_CFG("C21_1610_KBC4", 3, 3, 0, 0, 1, 1, 0, 0, 0) -MUX_CFG("G18_1610_KBR0", 4, 0, 0, 0, 10, 1, 0, 1, 0) -MUX_CFG("F19_1610_KBR1", 3, 27, 0, 0, 9, 1, 0, 1, 0) -MUX_CFG("H14_1610_KBR2", 3, 24, 0, 0, 8, 1, 0, 1, 0) -MUX_CFG("E20_1610_KBR3", 3, 21, 0, 0, 7, 1, 0, 1, 0) -MUX_CFG("E19_1610_KBR4", 3, 18, 0, 0, 6, 1, 0, 1, 0) -MUX_CFG("N19_1610_KBR5", 6, 12, 1, 1, 2, 1, 1, 1, 0) - -/* Power management */ -MUX_CFG("T20_1610_LOW_PWR", 7, 12, 1, NA, 0, 0, NA, 0, 0) - -/* MCLK Settings */ -MUX_CFG("V5_1710_MCLK_ON", B, 15, 0, NA, 0, 0, NA, 0, 0) -MUX_CFG("V5_1710_MCLK_OFF", B, 15, 6, NA, 0, 0, NA, 0, 0) -MUX_CFG("R10_1610_MCLK_ON", B, 18, 0, NA, 22, 0, NA, 1, 0) -MUX_CFG("R10_1610_MCLK_OFF", B, 18, 6, 2, 22, 1, 2, 1, 1) - -/* CompactFlash controller, conflicts with MMC1 */ -MUX_CFG("P11_1610_CF_CD2", A, 27, 3, 2, 15, 1, 2, 1, 1) -MUX_CFG("R11_1610_CF_IOIS16", B, 0, 3, 2, 16, 1, 2, 1, 1) -MUX_CFG("V10_1610_CF_IREQ", A, 24, 3, 2, 14, 0, 2, 0, 1) -MUX_CFG("W10_1610_CF_RESET", A, 18, 3, 2, 12, 1, 2, 1, 1) -MUX_CFG("W11_1610_CF_CD1", 10, 15, 3, 3, 8, 1, 3, 1, 1) }; -#endif /* __MUX_C__ */ +enum omap24xx_index { + /* 24xx I2C */ + M19_24XX_I2C1_SCL, + L15_24XX_I2C1_SDA, + J15_24XX_I2C2_SCL, + H19_24XX_I2C2_SDA, + + /* 24xx Menelaus interrupt */ + W19_24XX_SYS_NIRQ, + + /* 24xx GPIO */ + Y20_24XX_GPIO60, + M15_24XX_GPIO92, +}; #ifdef CONFIG_OMAP_MUX /* setup pin muxing in Linux */ -extern int omap_cfg_reg(reg_cfg_t reg_cfg); +extern int omap1_mux_init(void); +extern int omap2_mux_init(void); +extern int omap_mux_register(struct pin_config * pins, unsigned long size); +extern int omap_cfg_reg(unsigned long reg_cfg); #else /* boot loader does it all (no warnings from CONFIG_OMAP_MUX_WARNINGS) */ -static inline int omap_cfg_reg(reg_cfg_t reg_cfg) { return 0; } +static inline int omap1_mux_init(void) { return 0; } +static inline int omap2_mux_init(void) { return 0; } +static inline int omap_cfg_reg(unsigned long reg_cfg) { return 0; } #endif #endif diff --git a/include/asm-arm/arch-omap/omap1510.h b/include/asm-arm/arch-omap/omap1510.h index f086a3933906..c575d354850f 100644 --- a/include/asm-arm/arch-omap/omap1510.h +++ b/include/asm-arm/arch-omap/omap1510.h @@ -25,8 +25,8 @@ * 675 Mass Ave, Cambridge, MA 02139, USA. */ -#ifndef __ASM_ARCH_OMAP1510_H -#define __ASM_ARCH_OMAP1510_H +#ifndef __ASM_ARCH_OMAP15XX_H +#define __ASM_ARCH_OMAP15XX_H /* * ---------------------------------------------------------------------------- @@ -44,5 +44,5 @@ #define OMAP1510_DSPREG_SIZE SZ_128K #define OMAP1510_DSPREG_START 0xE1000000 -#endif /* __ASM_ARCH_OMAP1510_H */ +#endif /* __ASM_ARCH_OMAP15XX_H */ diff --git a/include/asm-arm/arch-omap/omap24xx.h b/include/asm-arm/arch-omap/omap24xx.h index a9105466a417..6e59805fa654 100644 --- a/include/asm-arm/arch-omap/omap24xx.h +++ b/include/asm-arm/arch-omap/omap24xx.h @@ -1,15 +1,24 @@ #ifndef __ASM_ARCH_OMAP24XX_H #define __ASM_ARCH_OMAP24XX_H -#define OMAP24XX_L4_IO_BASE 0x48000000 +/* + * Please place only base defines here and put the rest in device + * specific headers. Note also that some of these defines are needed + * for omap1 to compile without adding ifdefs. + */ + +#define L4_24XX_BASE 0x48000000 +#define L3_24XX_BASE 0x68000000 /* interrupt controller */ -#define OMAP24XX_IC_BASE (OMAP24XX_L4_IO_BASE + 0xfe000) +#define OMAP24XX_IC_BASE (L4_24XX_BASE + 0xfe000) #define VA_IC_BASE IO_ADDRESS(OMAP24XX_IC_BASE) - #define OMAP24XX_IVA_INTC_BASE 0x40000000 - #define IRQ_SIR_IRQ 0x0040 +#define OMAP24XX_32KSYNCT_BASE (L4_24XX_BASE + 0x4000) +#define OMAP24XX_PRCM_BASE (L4_24XX_BASE + 0x8000) +#define OMAP24XX_SDRC_BASE (L3_24XX_BASE + 0x9000) + #endif /* __ASM_ARCH_OMAP24XX_H */ diff --git a/include/asm-arm/arch-omap/omapfb.h b/include/asm-arm/arch-omap/omapfb.h new file mode 100644 index 000000000000..4ba2622cc142 --- /dev/null +++ b/include/asm-arm/arch-omap/omapfb.h @@ -0,0 +1,281 @@ +/* + * File: include/asm-arm/arch-omap/omapfb.h + * + * Framebuffer driver for TI OMAP boards + * + * Copyright (C) 2004 Nokia Corporation + * Author: Imre Deak + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or (at your + * option) any later version. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + */ + +#ifndef __OMAPFB_H +#define __OMAPFB_H + +/* IOCTL commands. */ + +#define OMAP_IOW(num, dtype) _IOW('O', num, dtype) +#define OMAP_IOR(num, dtype) _IOR('O', num, dtype) +#define OMAP_IOWR(num, dtype) _IOWR('O', num, dtype) +#define OMAP_IO(num) _IO('O', num) + +#define OMAPFB_MIRROR OMAP_IOW(31, int) +#define OMAPFB_SYNC_GFX OMAP_IO(37) +#define OMAPFB_VSYNC OMAP_IO(38) +#define OMAPFB_SET_UPDATE_MODE OMAP_IOW(40, enum omapfb_update_mode) +#define OMAPFB_GET_CAPS OMAP_IOR(42, unsigned long) +#define OMAPFB_GET_UPDATE_MODE OMAP_IOW(43, enum omapfb_update_mode) +#define OMAPFB_LCD_TEST OMAP_IOW(45, int) +#define OMAPFB_CTRL_TEST OMAP_IOW(46, int) +#define OMAPFB_UPDATE_WINDOW OMAP_IOW(47, struct omapfb_update_window) +#define OMAPFB_SETUP_PLANE OMAP_IOW(48, struct omapfb_setup_plane) +#define OMAPFB_ENABLE_PLANE OMAP_IOW(49, struct omapfb_enable_plane) +#define OMAPFB_SET_COLOR_KEY OMAP_IOW(50, struct omapfb_color_key) + +#define OMAPFB_CAPS_GENERIC_MASK 0x00000fff +#define OMAPFB_CAPS_LCDC_MASK 0x00fff000 +#define OMAPFB_CAPS_PANEL_MASK 0xff000000 + +#define OMAPFB_CAPS_MANUAL_UPDATE 0x00001000 +#define OMAPFB_CAPS_SET_BACKLIGHT 0x01000000 + +/* Values from DSP must map to lower 16-bits */ +#define OMAPFB_FORMAT_MASK 0x00ff +#define OMAPFB_FORMAT_FLAG_DOUBLE 0x0100 + +enum omapfb_color_format { + OMAPFB_COLOR_RGB565 = 0, + OMAPFB_COLOR_YUV422, + OMAPFB_COLOR_YUV420, + OMAPFB_COLOR_CLUT_8BPP, + OMAPFB_COLOR_CLUT_4BPP, + OMAPFB_COLOR_CLUT_2BPP, + OMAPFB_COLOR_CLUT_1BPP, +}; + +struct omapfb_update_window { + u32 x, y; + u32 width, height; + u32 format; +}; + +enum omapfb_plane { + OMAPFB_PLANE_GFX = 0, + OMAPFB_PLANE_VID1, + OMAPFB_PLANE_VID2, +}; + +enum omapfb_channel_out { + OMAPFB_CHANNEL_OUT_LCD = 0, + OMAPFB_CHANNEL_OUT_DIGIT, +}; + +struct omapfb_setup_plane { + u8 plane; + u8 channel_out; + u32 offset; + u32 pos_x, pos_y; + u32 width, height; + u32 color_mode; +}; + +struct omapfb_enable_plane { + u8 plane; + u8 enable; +}; + +enum omapfb_color_key_type { + OMAPFB_COLOR_KEY_DISABLED = 0, + OMAPFB_COLOR_KEY_GFX_DST, + OMAPFB_COLOR_KEY_VID_SRC, +}; + +struct omapfb_color_key { + u8 channel_out; + u32 background; + u32 trans_key; + u8 key_type; +}; + +enum omapfb_update_mode { + OMAPFB_UPDATE_DISABLED = 0, + OMAPFB_AUTO_UPDATE, + OMAPFB_MANUAL_UPDATE +}; + +#ifdef __KERNEL__ + +#include +#include +#include + +#define OMAP_LCDC_INV_VSYNC 0x0001 +#define OMAP_LCDC_INV_HSYNC 0x0002 +#define OMAP_LCDC_INV_PIX_CLOCK 0x0004 +#define OMAP_LCDC_INV_OUTPUT_EN 0x0008 +#define OMAP_LCDC_HSVS_RISING_EDGE 0x0010 +#define OMAP_LCDC_HSVS_OPPOSITE 0x0020 + +#define OMAP_LCDC_SIGNAL_MASK 0x003f + +#define OMAP_LCDC_PANEL_TFT 0x0100 + +#ifdef CONFIG_ARCH_OMAP1 +#define OMAPFB_PLANE_NUM 1 +#else +#define OMAPFB_PLANE_NUM 3 +#endif + +struct omapfb_device; + +struct lcd_panel { + const char *name; + int config; /* TFT/STN, signal inversion */ + int bpp; /* Pixel format in fb mem */ + int data_lines; /* Lines on LCD HW interface */ + + int x_res, y_res; + int pixel_clock; /* In kHz */ + int hsw; /* Horizontal synchronization + pulse width */ + int hfp; /* Horizontal front porch */ + int hbp; /* Horizontal back porch */ + int vsw; /* Vertical synchronization + pulse width */ + int vfp; /* Vertical front porch */ + int vbp; /* Vertical back porch */ + int acb; /* ac-bias pin frequency */ + int pcd; /* pixel clock divider. + Obsolete use pixel_clock instead */ + + int (*init) (struct omapfb_device *fbdev); + void (*cleanup) (void); + int (*enable) (void); + void (*disable) (void); + unsigned long (*get_caps) (void); + int (*set_bklight_level)(unsigned int level); + unsigned int (*get_bklight_level)(void); + unsigned int (*get_bklight_max) (void); + int (*run_test) (int test_num); +}; + +struct omapfb_device; + +struct extif_timings { + int cs_on_time; + int cs_off_time; + int we_on_time; + int we_off_time; + int re_on_time; + int re_off_time; + int we_cycle_time; + int re_cycle_time; + int cs_pulse_width; + int access_time; +}; + +struct lcd_ctrl_extif { + int (*init) (void); + void (*cleanup) (void); + void (*set_timings) (const struct extif_timings *timings); + void (*write_command) (u32 cmd); + u32 (*read_data) (void); + void (*write_data) (u32 data); + void (*transfer_area) (int width, int height, + void (callback)(void * data), void *data); +}; + +struct lcd_ctrl { + const char *name; + void *data; + + int (*init) (struct omapfb_device *fbdev, + int ext_mode, int req_vram_size); + void (*cleanup) (void); + void (*get_vram_layout)(unsigned long *size, + void **virt_base, + dma_addr_t *phys_base); + unsigned long (*get_caps) (void); + int (*set_update_mode)(enum omapfb_update_mode mode); + enum omapfb_update_mode (*get_update_mode)(void); + int (*setup_plane) (int plane, int channel_out, + unsigned long offset, + int screen_width, + int pos_x, int pos_y, int width, + int height, int color_mode); + int (*enable_plane) (int plane, int enable); + int (*update_window) (struct omapfb_update_window *win, + void (*callback)(void *), + void *callback_data); + void (*sync) (void); + void (*suspend) (void); + void (*resume) (void); + int (*run_test) (int test_num); + int (*setcolreg) (u_int regno, u16 red, u16 green, + u16 blue, u16 transp, + int update_hw_mem); + int (*set_color_key) (struct omapfb_color_key *ck); + +}; + +enum omapfb_state { + OMAPFB_DISABLED = 0, + OMAPFB_SUSPENDED= 99, + OMAPFB_ACTIVE = 100 +}; + +struct omapfb_device { + int state; + int ext_lcdc; /* Using external + LCD controller */ + struct semaphore rqueue_sema; + + void *vram_virt_base; + dma_addr_t vram_phys_base; + unsigned long vram_size; + + int color_mode; + int palette_size; + int mirror; + u32 pseudo_palette[17]; + + struct lcd_panel *panel; /* LCD panel */ + struct lcd_ctrl *ctrl; /* LCD controller */ + struct lcd_ctrl *int_ctrl; /* internal LCD ctrl */ + struct lcd_ctrl_extif *ext_if; /* LCD ctrl external + interface */ + struct fb_info *fb_info; + + struct device *dev; +}; + +extern struct lcd_panel h3_panel; +extern struct lcd_panel h2_panel; +extern struct lcd_panel p2_panel; +extern struct lcd_panel osk_panel; +extern struct lcd_panel innovator1610_panel; +extern struct lcd_panel innovator1510_panel; + +#ifdef CONFIG_ARCH_OMAP1 +extern struct lcd_ctrl omap1_lcd_ctrl; +#else +extern struct lcd_ctrl omap2_disp_ctrl; +#endif + +extern void omapfb_write_first_pixel(struct omapfb_device *fbdev, u16 pixval); + +#endif /* __KERNEL__ */ + +#endif /* __OMAPFB_H */ diff --git a/include/asm-arm/arch-omap/pm.h b/include/asm-arm/arch-omap/pm.h index fbd742d0c499..7c790425e363 100644 --- a/include/asm-arm/arch-omap/pm.h +++ b/include/asm-arm/arch-omap/pm.h @@ -98,7 +98,14 @@ #define OMAP1610_IDLECT3 0xfffece24 #define OMAP1610_IDLE_LOOP_REQUEST 0x0400 -#if !defined(CONFIG_ARCH_OMAP1510) && \ +#define OMAP730_IDLECT1_SLEEP_VAL 0x16c7 +#define OMAP730_IDLECT2_SLEEP_VAL 0x09c7 +#define OMAP730_IDLECT3_VAL 0x3f +#define OMAP730_IDLECT3 0xfffece24 +#define OMAP730_IDLE_LOOP_REQUEST 0x0C00 + +#if !defined(CONFIG_ARCH_OMAP730) && \ + !defined(CONFIG_ARCH_OMAP15XX) && \ !defined(CONFIG_ARCH_OMAP16XX) && \ !defined(CONFIG_ARCH_OMAP24XX) #error "Power management for this processor not implemented yet" @@ -107,8 +114,10 @@ #ifndef __ASSEMBLER__ extern void omap_pm_idle(void); extern void omap_pm_suspend(void); +extern void omap730_cpu_suspend(unsigned short, unsigned short); extern void omap1510_cpu_suspend(unsigned short, unsigned short); extern void omap1610_cpu_suspend(unsigned short, unsigned short); +extern void omap730_idle_loop_suspend(void); extern void omap1510_idle_loop_suspend(void); extern void omap1610_idle_loop_suspend(void); @@ -118,6 +127,8 @@ extern void omap_serial_wake_trigger(int enable); #define omap_serial_wake_trigger(x) {} #endif /* CONFIG_OMAP_SERIAL_WAKE */ +extern unsigned int omap730_cpu_suspend_sz; +extern unsigned int omap730_idle_loop_suspend_sz; extern unsigned int omap1510_cpu_suspend_sz; extern unsigned int omap1510_idle_loop_suspend_sz; extern unsigned int omap1610_cpu_suspend_sz; @@ -131,6 +142,10 @@ extern unsigned int omap1610_idle_loop_suspend_sz; #define ULPD_RESTORE(x) omap_writew((ulpd_sleep_save[ULPD_SLEEP_SAVE_##x]), (x)) #define ULPD_SHOW(x) ulpd_sleep_save[ULPD_SLEEP_SAVE_##x] +#define MPUI730_SAVE(x) mpui730_sleep_save[MPUI730_SLEEP_SAVE_##x] = omap_readl(x) +#define MPUI730_RESTORE(x) omap_writel((mpui730_sleep_save[MPUI730_SLEEP_SAVE_##x]), (x)) +#define MPUI730_SHOW(x) mpui730_sleep_save[MPUI730_SLEEP_SAVE_##x] + #define MPUI1510_SAVE(x) mpui1510_sleep_save[MPUI1510_SLEEP_SAVE_##x] = omap_readl(x) #define MPUI1510_RESTORE(x) omap_writel((mpui1510_sleep_save[MPUI1510_SLEEP_SAVE_##x]), (x)) #define MPUI1510_SHOW(x) mpui1510_sleep_save[MPUI1510_SLEEP_SAVE_##x] @@ -188,13 +203,34 @@ enum mpui1510_save_state { MPUI1510_SLEEP_SAVE_EMIFS_CONFIG, MPUI1510_SLEEP_SAVE_OMAP_IH1_MIR, MPUI1510_SLEEP_SAVE_OMAP_IH2_MIR, -#if defined(CONFIG_ARCH_OMAP1510) +#if defined(CONFIG_ARCH_OMAP15XX) MPUI1510_SLEEP_SAVE_SIZE #else MPUI1510_SLEEP_SAVE_SIZE = 0 #endif }; +enum mpui730_save_state { + MPUI730_SLEEP_SAVE_START = 0, + /* + * MPUI registers 32 bits + */ + MPUI730_SLEEP_SAVE_MPUI_CTRL, + MPUI730_SLEEP_SAVE_MPUI_DSP_BOOT_CONFIG, + MPUI730_SLEEP_SAVE_MPUI_DSP_API_CONFIG, + MPUI730_SLEEP_SAVE_MPUI_DSP_STATUS, + MPUI730_SLEEP_SAVE_EMIFF_SDRAM_CONFIG, + MPUI730_SLEEP_SAVE_EMIFS_CONFIG, + MPUI730_SLEEP_SAVE_OMAP_IH1_MIR, + MPUI730_SLEEP_SAVE_OMAP_IH2_0_MIR, + MPUI730_SLEEP_SAVE_OMAP_IH2_1_MIR, +#if defined(CONFIG_ARCH_OMAP730) + MPUI730_SLEEP_SAVE_SIZE +#else + MPUI730_SLEEP_SAVE_SIZE = 0 +#endif +}; + enum mpui1610_save_state { MPUI1610_SLEEP_SAVE_START = 0, /* diff --git a/include/asm-arm/arch-omap/prcm.h b/include/asm-arm/arch-omap/prcm.h new file mode 100644 index 000000000000..7b48a5cbb15f --- /dev/null +++ b/include/asm-arm/arch-omap/prcm.h @@ -0,0 +1,429 @@ +/* + * prcm.h - Access definations for use in OMAP24XX clock and power management + * + * Copyright (C) 2005 Texas Instruments, Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#ifndef __ASM_ARM_ARCH_DPM_PRCM_H +#define __ASM_ARM_ARCH_DPM_PRCM_H + +/* SET_PERFORMANCE_LEVEL PARAMETERS */ +#define PRCM_HALF_SPEED 1 +#define PRCM_FULL_SPEED 2 + +#ifndef __ASSEMBLER__ + +#define PRCM_REG32(offset) __REG32(OMAP24XX_PRCM_BASE + (offset)) + +#define PRCM_REVISION PRCM_REG32(0x000) +#define PRCM_SYSCONFIG PRCM_REG32(0x010) +#define PRCM_IRQSTATUS_MPU PRCM_REG32(0x018) +#define PRCM_IRQENABLE_MPU PRCM_REG32(0x01C) +#define PRCM_VOLTCTRL PRCM_REG32(0x050) +#define PRCM_VOLTST PRCM_REG32(0x054) +#define PRCM_CLKSRC_CTRL PRCM_REG32(0x060) +#define PRCM_CLKOUT_CTRL PRCM_REG32(0x070) +#define PRCM_CLKEMUL_CTRL PRCM_REG32(0x078) +#define PRCM_CLKCFG_CTRL PRCM_REG32(0x080) +#define PRCM_CLKCFG_STATUS PRCM_REG32(0x084) +#define PRCM_VOLTSETUP PRCM_REG32(0x090) +#define PRCM_CLKSSETUP PRCM_REG32(0x094) +#define PRCM_POLCTRL PRCM_REG32(0x098) + +/* GENERAL PURPOSE */ +#define GENERAL_PURPOSE1 PRCM_REG32(0x0B0) +#define GENERAL_PURPOSE2 PRCM_REG32(0x0B4) +#define GENERAL_PURPOSE3 PRCM_REG32(0x0B8) +#define GENERAL_PURPOSE4 PRCM_REG32(0x0BC) +#define GENERAL_PURPOSE5 PRCM_REG32(0x0C0) +#define GENERAL_PURPOSE6 PRCM_REG32(0x0C4) +#define GENERAL_PURPOSE7 PRCM_REG32(0x0C8) +#define GENERAL_PURPOSE8 PRCM_REG32(0x0CC) +#define GENERAL_PURPOSE9 PRCM_REG32(0x0D0) +#define GENERAL_PURPOSE10 PRCM_REG32(0x0D4) +#define GENERAL_PURPOSE11 PRCM_REG32(0x0D8) +#define GENERAL_PURPOSE12 PRCM_REG32(0x0DC) +#define GENERAL_PURPOSE13 PRCM_REG32(0x0E0) +#define GENERAL_PURPOSE14 PRCM_REG32(0x0E4) +#define GENERAL_PURPOSE15 PRCM_REG32(0x0E8) +#define GENERAL_PURPOSE16 PRCM_REG32(0x0EC) +#define GENERAL_PURPOSE17 PRCM_REG32(0x0F0) +#define GENERAL_PURPOSE18 PRCM_REG32(0x0F4) +#define GENERAL_PURPOSE19 PRCM_REG32(0x0F8) +#define GENERAL_PURPOSE20 PRCM_REG32(0x0FC) + +/* MPU */ +#define CM_CLKSEL_MPU PRCM_REG32(0x140) +#define CM_CLKSTCTRL_MPU PRCM_REG32(0x148) +#define RM_RSTST_MPU PRCM_REG32(0x158) +#define PM_WKDEP_MPU PRCM_REG32(0x1C8) +#define PM_EVGENCTRL_MPU PRCM_REG32(0x1D4) +#define PM_EVEGENONTIM_MPU PRCM_REG32(0x1D8) +#define PM_EVEGENOFFTIM_MPU PRCM_REG32(0x1DC) +#define PM_PWSTCTRL_MPU PRCM_REG32(0x1E0) +#define PM_PWSTST_MPU PRCM_REG32(0x1E4) + +/* CORE */ +#define CM_FCLKEN1_CORE PRCM_REG32(0x200) +#define CM_FCLKEN2_CORE PRCM_REG32(0x204) +#define CM_FCLKEN3_CORE PRCM_REG32(0x208) +#define CM_ICLKEN1_CORE PRCM_REG32(0x210) +#define CM_ICLKEN2_CORE PRCM_REG32(0x214) +#define CM_ICLKEN3_CORE PRCM_REG32(0x218) +#define CM_ICLKEN4_CORE PRCM_REG32(0x21C) +#define CM_IDLEST1_CORE PRCM_REG32(0x220) +#define CM_IDLEST2_CORE PRCM_REG32(0x224) +#define CM_IDLEST3_CORE PRCM_REG32(0x228) +#define CM_IDLEST4_CORE PRCM_REG32(0x22C) +#define CM_AUTOIDLE1_CORE PRCM_REG32(0x230) +#define CM_AUTOIDLE2_CORE PRCM_REG32(0x234) +#define CM_AUTOIDLE3_CORE PRCM_REG32(0x238) +#define CM_AUTOIDLE4_CORE PRCM_REG32(0x23C) +#define CM_CLKSEL1_CORE PRCM_REG32(0x240) +#define CM_CLKSEL2_CORE PRCM_REG32(0x244) +#define CM_CLKSTCTRL_CORE PRCM_REG32(0x248) +#define PM_WKEN1_CORE PRCM_REG32(0x2A0) +#define PM_WKEN2_CORE PRCM_REG32(0x2A4) +#define PM_WKST1_CORE PRCM_REG32(0x2B0) +#define PM_WKST2_CORE PRCM_REG32(0x2B4) +#define PM_WKDEP_CORE PRCM_REG32(0x2C8) +#define PM_PWSTCTRL_CORE PRCM_REG32(0x2E0) +#define PM_PWSTST_CORE PRCM_REG32(0x2E4) + +/* GFX */ +#define CM_FCLKEN_GFX PRCM_REG32(0x300) +#define CM_ICLKEN_GFX PRCM_REG32(0x310) +#define CM_IDLEST_GFX PRCM_REG32(0x320) +#define CM_CLKSEL_GFX PRCM_REG32(0x340) +#define CM_CLKSTCTRL_GFX PRCM_REG32(0x348) +#define RM_RSTCTRL_GFX PRCM_REG32(0x350) +#define RM_RSTST_GFX PRCM_REG32(0x358) +#define PM_WKDEP_GFX PRCM_REG32(0x3C8) +#define PM_PWSTCTRL_GFX PRCM_REG32(0x3E0) +#define PM_PWSTST_GFX PRCM_REG32(0x3E4) + +/* WAKE-UP */ +#define CM_FCLKEN_WKUP PRCM_REG32(0x400) +#define CM_ICLKEN_WKUP PRCM_REG32(0x410) +#define CM_IDLEST_WKUP PRCM_REG32(0x420) +#define CM_AUTOIDLE_WKUP PRCM_REG32(0x430) +#define CM_CLKSEL_WKUP PRCM_REG32(0x440) +#define RM_RSTCTRL_WKUP PRCM_REG32(0x450) +#define RM_RSTTIME_WKUP PRCM_REG32(0x454) +#define RM_RSTST_WKUP PRCM_REG32(0x458) +#define PM_WKEN_WKUP PRCM_REG32(0x4A0) +#define PM_WKST_WKUP PRCM_REG32(0x4B0) + +/* CLOCKS */ +#define CM_CLKEN_PLL PRCM_REG32(0x500) +#define CM_IDLEST_CKGEN PRCM_REG32(0x520) +#define CM_AUTOIDLE_PLL PRCM_REG32(0x530) +#define CM_CLKSEL1_PLL PRCM_REG32(0x540) +#define CM_CLKSEL2_PLL PRCM_REG32(0x544) + +/* DSP */ +#define CM_FCLKEN_DSP PRCM_REG32(0x800) +#define CM_ICLKEN_DSP PRCM_REG32(0x810) +#define CM_IDLEST_DSP PRCM_REG32(0x820) +#define CM_AUTOIDLE_DSP PRCM_REG32(0x830) +#define CM_CLKSEL_DSP PRCM_REG32(0x840) +#define CM_CLKSTCTRL_DSP PRCM_REG32(0x848) +#define RM_RSTCTRL_DSP PRCM_REG32(0x850) +#define RM_RSTST_DSP PRCM_REG32(0x858) +#define PM_WKEN_DSP PRCM_REG32(0x8A0) +#define PM_WKDEP_DSP PRCM_REG32(0x8C8) +#define PM_PWSTCTRL_DSP PRCM_REG32(0x8E0) +#define PM_PWSTST_DSP PRCM_REG32(0x8E4) +#define PRCM_IRQSTATUS_DSP PRCM_REG32(0x8F0) +#define PRCM_IRQENABLE_DSP PRCM_REG32(0x8F4) + +/* IVA */ +#define PRCM_IRQSTATUS_IVA PRCM_REG32(0x8F8) +#define PRCM_IRQENABLE_IVA PRCM_REG32(0x8FC) + +/* Modem on 2430 */ +#define CM_FCLKEN_MDM PRCM_REG32(0xC00) +#define CM_ICLKEN_MDM PRCM_REG32(0xC10) +#define CM_IDLEST_MDM PRCM_REG32(0xC20) +#define CM_CLKSEL_MDM PRCM_REG32(0xC40) + +/* FIXME: Move to header for 2430 */ +#define DISP_BASE (OMAP24XX_L4_IO_BASE+0x50000) +#define DISP_REG32(offset) __REG32(DISP_BASE + (offset)) + +#define OMAP24XX_GPMC_BASE (L3_24XX_BASE + 0xa000) +#define GPMC_BASE (OMAP24XX_GPMC_BASE) +#define GPMC_REG32(offset) __REG32(GPMC_BASE + (offset)) + +#define GPT1_BASE (OMAP24XX_GPT1) +#define GPT1_REG32(offset) __REG32(GPT1_BASE + (offset)) + +/* Misc sysconfig */ +#define DISPC_SYSCONFIG DISP_REG32(0x410) +#define SPI_BASE (OMAP24XX_L4_IO_BASE+0x98000) +#define MCSPI1_SYSCONFIG __REG32(SPI_BASE + 0x10) +#define MCSPI2_SYSCONFIG __REG32(SPI_BASE+0x2000 + 0x10) + +//#define DSP_MMU_SYSCONFIG 0x5A000010 +#define CAMERA_MMU_SYSCONFIG __REG32(DISP_BASE+0x2C10) +//#define IVA_MMU_SYSCONFIG 0x5D000010 +//#define DSP_DMA_SYSCONFIG 0x00FCC02C +#define CAMERA_DMA_SYSCONFIG __REG32(DISP_BASE+0x282C) +#define SYSTEM_DMA_SYSCONFIG __REG32(DISP_BASE+0x602C) +#define GPMC_SYSCONFIG GPMC_REG32(0x010) +#define MAILBOXES_SYSCONFIG __REG32(OMAP24XX_L4_IO_BASE+0x94010) +#define UART1_SYSCONFIG __REG32(OMAP24XX_L4_IO_BASE+0x6A054) +#define UART2_SYSCONFIG __REG32(OMAP24XX_L4_IO_BASE+0x6C054) +#define UART3_SYSCONFIG __REG32(OMAP24XX_L4_IO_BASE+0x6E054) +//#define IVA_SYSCONFIG 0x5C060010 +#define SDRC_SYSCONFIG __REG32(OMAP24XX_SDRC_BASE+0x10) +#define SMS_SYSCONFIG __REG32(OMAP24XX_SMS_BASE+0x10) +#define SSI_SYSCONFIG __REG32(DISP_BASE+0x8010) +//#define VLYNQ_SYSCONFIG 0x67FFFE10 + +/* rkw - good cannidates for PM_ to start what nm was trying */ +#define OMAP24XX_GPT2 (OMAP24XX_L4_IO_BASE+0x2A000) +#define OMAP24XX_GPT3 (OMAP24XX_L4_IO_BASE+0x78000) +#define OMAP24XX_GPT4 (OMAP24XX_L4_IO_BASE+0x7A000) +#define OMAP24XX_GPT5 (OMAP24XX_L4_IO_BASE+0x7C000) +#define OMAP24XX_GPT6 (OMAP24XX_L4_IO_BASE+0x7E000) +#define OMAP24XX_GPT7 (OMAP24XX_L4_IO_BASE+0x80000) +#define OMAP24XX_GPT8 (OMAP24XX_L4_IO_BASE+0x82000) +#define OMAP24XX_GPT9 (OMAP24XX_L4_IO_BASE+0x84000) +#define OMAP24XX_GPT10 (OMAP24XX_L4_IO_BASE+0x86000) +#define OMAP24XX_GPT11 (OMAP24XX_L4_IO_BASE+0x88000) +#define OMAP24XX_GPT12 (OMAP24XX_L4_IO_BASE+0x8A000) + +#define GPTIMER1_SYSCONFIG GPT1_REG32(0x010) +#define GPTIMER2_SYSCONFIG __REG32(OMAP24XX_GPT2 + 0x10) +#define GPTIMER3_SYSCONFIG __REG32(OMAP24XX_GPT3 + 0x10) +#define GPTIMER4_SYSCONFIG __REG32(OMAP24XX_GPT4 + 0x10) +#define GPTIMER5_SYSCONFIG __REG32(OMAP24XX_GPT5 + 0x10) +#define GPTIMER6_SYSCONFIG __REG32(OMAP24XX_GPT6 + 0x10) +#define GPTIMER7_SYSCONFIG __REG32(OMAP24XX_GPT7 + 0x10) +#define GPTIMER8_SYSCONFIG __REG32(OMAP24XX_GPT8 + 0x10) +#define GPTIMER9_SYSCONFIG __REG32(OMAP24XX_GPT9 + 0x10) +#define GPTIMER10_SYSCONFIG __REG32(OMAP24XX_GPT10 + 0x10) +#define GPTIMER11_SYSCONFIG __REG32(OMAP24XX_GPT11 + 0x10) +#define GPTIMER12_SYSCONFIG __REG32(OMAP24XX_GPT12 + 0x10) + +#define GPIOX_BASE(X) (OMAP24XX_GPIO_BASE+(0x2000*((X)-1))) + +#define GPIO1_SYSCONFIG __REG32((GPIOX_BASE(1)+0x10)) +#define GPIO2_SYSCONFIG __REG32((GPIOX_BASE(2)+0x10)) +#define GPIO3_SYSCONFIG __REG32((GPIOX_BASE(3)+0x10)) +#define GPIO4_SYSCONFIG __REG32((GPIOX_BASE(4)+0x10)) + +/* GP TIMER 1 */ +#define GPTIMER1_TISTAT GPT1_REG32(0x014) +#define GPTIMER1_TISR GPT1_REG32(0x018) +#define GPTIMER1_TIER GPT1_REG32(0x01C) +#define GPTIMER1_TWER GPT1_REG32(0x020) +#define GPTIMER1_TCLR GPT1_REG32(0x024) +#define GPTIMER1_TCRR GPT1_REG32(0x028) +#define GPTIMER1_TLDR GPT1_REG32(0x02C) +#define GPTIMER1_TTGR GPT1_REG32(0x030) +#define GPTIMER1_TWPS GPT1_REG32(0x034) +#define GPTIMER1_TMAR GPT1_REG32(0x038) +#define GPTIMER1_TCAR1 GPT1_REG32(0x03C) +#define GPTIMER1_TSICR GPT1_REG32(0x040) +#define GPTIMER1_TCAR2 GPT1_REG32(0x044) + +/* rkw -- base fix up please... */ +#define GPTIMER3_TISR __REG32(OMAP24XX_L4_IO_BASE+0x78018) + +/* SDRC */ +#define SDRC_DLLA_CTRL __REG32(OMAP24XX_SDRC_BASE+0x060) +#define SDRC_DLLA_STATUS __REG32(OMAP24XX_SDRC_BASE+0x064) +#define SDRC_DLLB_CTRL __REG32(OMAP24XX_SDRC_BASE+0x068) +#define SDRC_DLLB_STATUS __REG32(OMAP24XX_SDRC_BASE+0x06C) +#define SDRC_POWER __REG32(OMAP24XX_SDRC_BASE+0x070) +#define SDRC_MR_0 __REG32(OMAP24XX_SDRC_BASE+0x084) + +/* GPIO 1 */ +#define GPIO1_BASE GPIOX_BASE(1) +#define GPIO1_REG32(offset) __REG32(GPIO1_BASE + (offset)) +#define GPIO1_IRQENABLE1 GPIO1_REG32(0x01C) +#define GPIO1_IRQSTATUS1 GPIO1_REG32(0x018) +#define GPIO1_IRQENABLE2 GPIO1_REG32(0x02C) +#define GPIO1_IRQSTATUS2 GPIO1_REG32(0x028) +#define GPIO1_WAKEUPENABLE GPIO1_REG32(0x020) +#define GPIO1_RISINGDETECT GPIO1_REG32(0x048) +#define GPIO1_DATAIN GPIO1_REG32(0x038) +#define GPIO1_OE GPIO1_REG32(0x034) +#define GPIO1_DATAOUT GPIO1_REG32(0x03C) + +/* GPIO2 */ +#define GPIO2_BASE GPIOX_BASE(2) +#define GPIO2_REG32(offset) __REG32(GPIO2_BASE + (offset)) +#define GPIO2_IRQENABLE1 GPIO2_REG32(0x01C) +#define GPIO2_IRQSTATUS1 GPIO2_REG32(0x018) +#define GPIO2_IRQENABLE2 GPIO2_REG32(0x02C) +#define GPIO2_IRQSTATUS2 GPIO2_REG32(0x028) +#define GPIO2_WAKEUPENABLE GPIO2_REG32(0x020) +#define GPIO2_RISINGDETECT GPIO2_REG32(0x048) +#define GPIO2_DATAIN GPIO2_REG32(0x038) +#define GPIO2_OE GPIO2_REG32(0x034) +#define GPIO2_DATAOUT GPIO2_REG32(0x03C) + +/* GPIO 3 */ +#define GPIO3_BASE GPIOX_BASE(3) +#define GPIO3_REG32(offset) __REG32(GPIO3_BASE + (offset)) +#define GPIO3_IRQENABLE1 GPIO3_REG32(0x01C) +#define GPIO3_IRQSTATUS1 GPIO3_REG32(0x018) +#define GPIO3_IRQENABLE2 GPIO3_REG32(0x02C) +#define GPIO3_IRQSTATUS2 GPIO3_REG32(0x028) +#define GPIO3_WAKEUPENABLE GPIO3_REG32(0x020) +#define GPIO3_RISINGDETECT GPIO3_REG32(0x048) +#define GPIO3_FALLINGDETECT GPIO3_REG32(0x04C) +#define GPIO3_DATAIN GPIO3_REG32(0x038) +#define GPIO3_OE GPIO3_REG32(0x034) +#define GPIO3_DATAOUT GPIO3_REG32(0x03C) +#define GPIO3_DEBOUNCENABLE GPIO3_REG32(0x050) +#define GPIO3_DEBOUNCINGTIME GPIO3_REG32(0x054) + +/* GPIO 4 */ +#define GPIO4_BASE GPIOX_BASE(4) +#define GPIO4_REG32(offset) __REG32(GPIO4_BASE + (offset)) +#define GPIO4_IRQENABLE1 GPIO4_REG32(0x01C) +#define GPIO4_IRQSTATUS1 GPIO4_REG32(0x018) +#define GPIO4_IRQENABLE2 GPIO4_REG32(0x02C) +#define GPIO4_IRQSTATUS2 GPIO4_REG32(0x028) +#define GPIO4_WAKEUPENABLE GPIO4_REG32(0x020) +#define GPIO4_RISINGDETECT GPIO4_REG32(0x048) +#define GPIO4_FALLINGDETECT GPIO4_REG32(0x04C) +#define GPIO4_DATAIN GPIO4_REG32(0x038) +#define GPIO4_OE GPIO4_REG32(0x034) +#define GPIO4_DATAOUT GPIO4_REG32(0x03C) +#define GPIO4_DEBOUNCENABLE GPIO4_REG32(0x050) +#define GPIO4_DEBOUNCINGTIME GPIO4_REG32(0x054) + + +/* IO CONFIG */ +#define CONTROL_BASE (OMAP24XX_CTRL_BASE) +#define CONTROL_REG32(offset) __REG32(CONTROL_BASE + (offset)) + +#define CONTROL_PADCONF_SPI1_NCS2 CONTROL_REG32(0x104) +#define CONTROL_PADCONF_SYS_XTALOUT CONTROL_REG32(0x134) +#define CONTROL_PADCONF_UART1_RX CONTROL_REG32(0x0C8) +#define CONTROL_PADCONF_MCBSP1_DX CONTROL_REG32(0x10C) +#define CONTROL_PADCONF_GPMC_NCS4 CONTROL_REG32(0x090) +#define CONTROL_PADCONF_DSS_D5 CONTROL_REG32(0x0B8) +#define CONTROL_PADCONF_DSS_D9 CONTROL_REG32(0x0BC) +#define CONTROL_PADCONF_DSS_D13 CONTROL_REG32(0x0C0) +#define CONTROL_PADCONF_DSS_VSYNC CONTROL_REG32(0x0CC) + +/* CONTROL */ +#define CONTROL_DEVCONF CONTROL_REG32(0x274) + +/* INTERRUPT CONTROLLER */ +#define INTC_BASE (OMAP24XX_L4_IO_BASE+0xfe000) +#define INTC_REG32(offset) __REG32(INTC_BASE + (offset)) + +#define INTC1_U_BASE INTC_REG32(0x000) +#define INTC_MIR0 INTC_REG32(0x084) +#define INTC_MIR_SET0 INTC_REG32(0x08C) +#define INTC_MIR_CLEAR0 INTC_REG32(0x088) +#define INTC_ISR_CLEAR0 INTC_REG32(0x094) +#define INTC_MIR1 INTC_REG32(0x0A4) +#define INTC_MIR_SET1 INTC_REG32(0x0AC) +#define INTC_MIR_CLEAR1 INTC_REG32(0x0A8) +#define INTC_ISR_CLEAR1 INTC_REG32(0x0B4) +#define INTC_MIR2 INTC_REG32(0x0C4) +#define INTC_MIR_SET2 INTC_REG32(0x0CC) +#define INTC_MIR_CLEAR2 INTC_REG32(0x0C8) +#define INTC_ISR_CLEAR2 INTC_REG32(0x0D4) +#define INTC_SIR_IRQ INTC_REG32(0x040) +#define INTC_CONTROL INTC_REG32(0x048) +#define INTC_ILR11 INTC_REG32(0x12C) +#define INTC_ILR32 INTC_REG32(0x180) +#define INTC_ILR37 INTC_REG32(0x194) +#define INTC_SYSCONFIG INTC_REG32(0x010) + +/* RAM FIREWALL */ +#define RAMFW_BASE (0x68005000) +#define RAMFW_REG32(offset) __REG32(RAMFW_BASE + (offset)) + +#define RAMFW_REQINFOPERM0 RAMFW_REG32(0x048) +#define RAMFW_READPERM0 RAMFW_REG32(0x050) +#define RAMFW_WRITEPERM0 RAMFW_REG32(0x058) + +/* GPMC CS1 FPGA ON USER INTERFACE MODULE */ +//#define DEBUG_BOARD_LED_REGISTER 0x04000014 + +/* GPMC CS0 */ +#define GPMC_CONFIG1_0 GPMC_REG32(0x060) +#define GPMC_CONFIG2_0 GPMC_REG32(0x064) +#define GPMC_CONFIG3_0 GPMC_REG32(0x068) +#define GPMC_CONFIG4_0 GPMC_REG32(0x06C) +#define GPMC_CONFIG5_0 GPMC_REG32(0x070) +#define GPMC_CONFIG6_0 GPMC_REG32(0x074) +#define GPMC_CONFIG7_0 GPMC_REG32(0x078) + +/* GPMC CS1 */ +#define GPMC_CONFIG1_1 GPMC_REG32(0x090) +#define GPMC_CONFIG2_1 GPMC_REG32(0x094) +#define GPMC_CONFIG3_1 GPMC_REG32(0x098) +#define GPMC_CONFIG4_1 GPMC_REG32(0x09C) +#define GPMC_CONFIG5_1 GPMC_REG32(0x0a0) +#define GPMC_CONFIG6_1 GPMC_REG32(0x0a4) +#define GPMC_CONFIG7_1 GPMC_REG32(0x0a8) + +/* DSS */ +#define DSS_CONTROL DISP_REG32(0x040) +#define DISPC_CONTROL DISP_REG32(0x440) +#define DISPC_SYSSTATUS DISP_REG32(0x414) +#define DISPC_IRQSTATUS DISP_REG32(0x418) +#define DISPC_IRQENABLE DISP_REG32(0x41C) +#define DISPC_CONFIG DISP_REG32(0x444) +#define DISPC_DEFAULT_COLOR0 DISP_REG32(0x44C) +#define DISPC_DEFAULT_COLOR1 DISP_REG32(0x450) +#define DISPC_TRANS_COLOR0 DISP_REG32(0x454) +#define DISPC_TRANS_COLOR1 DISP_REG32(0x458) +#define DISPC_LINE_NUMBER DISP_REG32(0x460) +#define DISPC_TIMING_H DISP_REG32(0x464) +#define DISPC_TIMING_V DISP_REG32(0x468) +#define DISPC_POL_FREQ DISP_REG32(0x46C) +#define DISPC_DIVISOR DISP_REG32(0x470) +#define DISPC_SIZE_DIG DISP_REG32(0x478) +#define DISPC_SIZE_LCD DISP_REG32(0x47C) +#define DISPC_GFX_BA0 DISP_REG32(0x480) +#define DISPC_GFX_BA1 DISP_REG32(0x484) +#define DISPC_GFX_POSITION DISP_REG32(0x488) +#define DISPC_GFX_SIZE DISP_REG32(0x48C) +#define DISPC_GFX_ATTRIBUTES DISP_REG32(0x4A0) +#define DISPC_GFX_FIFO_THRESHOLD DISP_REG32(0x4A4) +#define DISPC_GFX_ROW_INC DISP_REG32(0x4AC) +#define DISPC_GFX_PIXEL_INC DISP_REG32(0x4B0) +#define DISPC_GFX_WINDOW_SKIP DISP_REG32(0x4B4) +#define DISPC_GFX_TABLE_BA DISP_REG32(0x4B8) +#define DISPC_DATA_CYCLE1 DISP_REG32(0x5D4) +#define DISPC_DATA_CYCLE2 DISP_REG32(0x5D8) +#define DISPC_DATA_CYCLE3 DISP_REG32(0x5DC) + +/* Wake up define for board */ +#define GPIO97 (1 << 1) +#define GPIO88 (1 << 24) + +#endif /* __ASSEMBLER__ */ + +#endif + + + + + diff --git a/include/asm-arm/arch-omap/sram.h b/include/asm-arm/arch-omap/sram.h new file mode 100644 index 000000000000..e72ccbf0fe06 --- /dev/null +++ b/include/asm-arm/arch-omap/sram.h @@ -0,0 +1,38 @@ +/* + * linux/include/asm-arm/arch-omap/sram.h + * + * Interface for functions that need to be run in internal SRAM + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ + +#ifndef __ARCH_ARM_OMAP_SRAM_H +#define __ARCH_ARM_OMAP_SRAM_H + +extern void * omap_sram_push(void * start, unsigned long size); +extern void omap_sram_reprogram_clock(u32 dpllctl, u32 ckctl); + +extern void omap2_sram_ddr_init(u32 *slow_dll_ctrl, u32 fast_dll_ctrl, + u32 base_cs, u32 force_unlock); +extern void omap2_sram_reprogram_sdrc(u32 perf_level, u32 dll_val, + u32 mem_type); +extern u32 omap2_set_prcm(u32 dpll_ctrl_val, u32 sdrc_rfr_val, int bypass); + + +/* Do not use these */ +extern void sram_reprogram_clock(u32 ckctl, u32 dpllctl); +extern unsigned long sram_reprogram_clock_sz; + +extern void sram_ddr_init(u32 *slow_dll_ctrl, u32 fast_dll_ctrl, + u32 base_cs, u32 force_unlock); +extern unsigned long sram_ddr_init_sz; + +extern u32 sram_set_prcm(u32 dpll_ctrl_val, u32 sdrc_rfr_val, int bypass); +extern unsigned long sram_set_prcm_sz; + +extern void sram_reprogram_sdrc(u32 perf_level, u32 dll_val, u32 mem_type); +extern unsigned long sram_reprogram_sdrc_sz; + +#endif diff --git a/include/asm-arm/arch-omap/system.h b/include/asm-arm/arch-omap/system.h index ff37bc27e603..b43cdd2a3874 100644 --- a/include/asm-arm/arch-omap/system.h +++ b/include/asm-arm/arch-omap/system.h @@ -6,18 +6,21 @@ #define __ASM_ARCH_SYSTEM_H #include #include +#include #include -#include +#include + +#ifndef CONFIG_MACH_VOICEBLUE +#define voiceblue_reset() do {} while (0) +#endif static inline void arch_idle(void) { cpu_do_idle(); } -static inline void arch_reset(char mode) +static inline void omap1_arch_reset(char mode) { - -#ifdef CONFIG_ARCH_OMAP16XX /* * Workaround for 5912/1611b bug mentioned in sprz209d.pdf p. 28 * "Global Software Reset Affects Traffic Controller Frequency". @@ -27,13 +30,31 @@ static inline void arch_reset(char mode) DPLL_CTL); omap_writew(0x8, ARM_RSTCT1); } -#endif -#ifdef CONFIG_MACH_VOICEBLUE + if (machine_is_voiceblue()) voiceblue_reset(); else -#endif omap_writew(1, ARM_RSTCT1); } +static inline void omap2_arch_reset(char mode) +{ + u32 rate; + struct clk *vclk, *sclk; + + vclk = clk_get(NULL, "virt_prcm_set"); + sclk = clk_get(NULL, "sys_ck"); + rate = clk_get_rate(sclk); + clk_set_rate(vclk, rate); /* go to bypass for OMAP limitation */ + RM_RSTCTRL_WKUP |= 2; +} + +static inline void arch_reset(char mode) +{ + if (!cpu_is_omap24xx()) + omap1_arch_reset(mode); + else + omap2_arch_reset(mode); +} + #endif diff --git a/include/asm-arm/arch-omap/timex.h b/include/asm-arm/arch-omap/timex.h index b61ddb491e83..21f2e367185a 100644 --- a/include/asm-arm/arch-omap/timex.h +++ b/include/asm-arm/arch-omap/timex.h @@ -28,6 +28,14 @@ #if !defined(__ASM_ARCH_OMAP_TIMEX_H) #define __ASM_ARCH_OMAP_TIMEX_H +/* + * OMAP 32KHz timer updates time one jiffie at a time from a secondary timer, + * and that's why the CLOCK_TICK_RATE is not 32768. + */ +#ifdef CONFIG_OMAP_32K_TIMER +#define CLOCK_TICK_RATE (CONFIG_OMAP_32K_TIMER_HZ) +#else #define CLOCK_TICK_RATE (HZ * 100000UL) +#endif #endif /* __ASM_ARCH_OMAP_TIMEX_H */ diff --git a/include/asm-arm/arch-omap/uncompress.h b/include/asm-arm/arch-omap/uncompress.h index 3545c86859cc..c718264affbd 100644 --- a/include/asm-arm/arch-omap/uncompress.h +++ b/include/asm-arm/arch-omap/uncompress.h @@ -36,10 +36,14 @@ putstr(const char *s) volatile u8 * uart = 0; int shift = 2; +#ifdef CONFIG_MACH_OMAP_PALMTE + return; +#endif + #ifdef CONFIG_ARCH_OMAP #ifdef CONFIG_OMAP_LL_DEBUG_UART3 uart = (volatile u8 *)(OMAP_UART3_BASE); -#elif CONFIG_OMAP_LL_DEBUG_UART2 +#elif defined(CONFIG_OMAP_LL_DEBUG_UART2) uart = (volatile u8 *)(OMAP_UART2_BASE); #else uart = (volatile u8 *)(OMAP_UART1_BASE); From b1faebb672edecf48d4745fb05eeca3c704b6823 Mon Sep 17 00:00:00 2001 From: Tony Lindgren Date: Thu, 10 Nov 2005 14:26:54 +0000 Subject: [PATCH 096/121] [ARM] 3144/1: OMAP 5/5: Update omap H2 defconfig Patch from Tony Lindgren This patch updates omap H2 defconfig. Signed-off-by: Tony Lindgren Signed-off-by: Russell King --- arch/arm/configs/omap_h2_1610_defconfig | 97 +++++++++++++++++-------- include/asm-arm/arch-omap/board-h4.h | 2 + 2 files changed, 68 insertions(+), 31 deletions(-) diff --git a/arch/arm/configs/omap_h2_1610_defconfig b/arch/arm/configs/omap_h2_1610_defconfig index 4198677cd394..529f0f72e1e9 100644 --- a/arch/arm/configs/omap_h2_1610_defconfig +++ b/arch/arm/configs/omap_h2_1610_defconfig @@ -1,7 +1,7 @@ # # Automatically generated make config: don't edit -# Linux kernel version: 2.6.13 -# Mon Sep 5 18:07:12 2005 +# Linux kernel version: 2.6.14 +# Wed Nov 9 18:53:40 2005 # CONFIG_ARM=y CONFIG_MMU=y @@ -22,6 +22,7 @@ CONFIG_INIT_ENV_ARG_LIMIT=32 # General setup # CONFIG_LOCALVERSION="" +CONFIG_LOCALVERSION_AUTO=y CONFIG_SWAP=y CONFIG_SYSVIPC=y # CONFIG_POSIX_MQUEUE is not set @@ -31,6 +32,7 @@ CONFIG_SYSCTL=y # CONFIG_HOTPLUG is not set CONFIG_KOBJECT_UEVENT=y # CONFIG_IKCONFIG is not set +CONFIG_INITRAMFS_SOURCE="" # CONFIG_EMBEDDED is not set CONFIG_KALLSYMS=y # CONFIG_KALLSYMS_EXTRA_PASS is not set @@ -59,6 +61,23 @@ CONFIG_OBSOLETE_MODPARM=y # CONFIG_MODULE_SRCVERSION_ALL is not set # CONFIG_KMOD is not set +# +# Block layer +# + +# +# IO Schedulers +# +CONFIG_IOSCHED_NOOP=y +CONFIG_IOSCHED_AS=y +CONFIG_IOSCHED_DEADLINE=y +CONFIG_IOSCHED_CFQ=y +CONFIG_DEFAULT_AS=y +# CONFIG_DEFAULT_DEADLINE is not set +# CONFIG_DEFAULT_CFQ is not set +# CONFIG_DEFAULT_NOOP is not set +CONFIG_DEFAULT_IOSCHED="anticipatory" + # # System Type # @@ -81,6 +100,7 @@ CONFIG_OBSOLETE_MODPARM=y # CONFIG_ARCH_LH7A40X is not set CONFIG_ARCH_OMAP=y # CONFIG_ARCH_VERSATILE is not set +# CONFIG_ARCH_REALVIEW is not set # CONFIG_ARCH_IMX is not set # CONFIG_ARCH_H720X is not set # CONFIG_ARCH_AAEC2000 is not set @@ -112,7 +132,7 @@ CONFIG_OMAP_SERIAL_WAKE=y # OMAP Core Type # # CONFIG_ARCH_OMAP730 is not set -# CONFIG_ARCH_OMAP1510 is not set +# CONFIG_ARCH_OMAP15XX is not set CONFIG_ARCH_OMAP16XX=y # @@ -177,6 +197,8 @@ CONFIG_FLATMEM_MANUAL=y # CONFIG_SPARSEMEM_MANUAL is not set CONFIG_FLATMEM=y CONFIG_FLAT_NODE_MEM_MAP=y +# CONFIG_SPARSEMEM_STATIC is not set +CONFIG_SPLIT_PTLOCK_CPUS=4096 # CONFIG_LEDS is not set CONFIG_ALIGNMENT_TRAP=y @@ -258,13 +280,18 @@ CONFIG_IP_PNP_BOOTP=y # CONFIG_INET_ESP is not set # CONFIG_INET_IPCOMP is not set # CONFIG_INET_TUNNEL is not set -CONFIG_IP_TCPDIAG=y -# CONFIG_IP_TCPDIAG_IPV6 is not set +CONFIG_INET_DIAG=y +CONFIG_INET_TCP_DIAG=y # CONFIG_TCP_CONG_ADVANCED is not set CONFIG_TCP_CONG_BIC=y # CONFIG_IPV6 is not set # CONFIG_NETFILTER is not set +# +# DCCP Configuration (EXPERIMENTAL) +# +# CONFIG_IP_DCCP is not set + # # SCTP Configuration (EXPERIMENTAL) # @@ -281,6 +308,10 @@ CONFIG_TCP_CONG_BIC=y # CONFIG_NET_DIVERT is not set # CONFIG_ECONET is not set # CONFIG_WAN_ROUTER is not set + +# +# QoS and/or fair queueing +# # CONFIG_NET_SCHED is not set # CONFIG_NET_CLS_ROUTE is not set @@ -291,6 +322,7 @@ CONFIG_TCP_CONG_BIC=y # CONFIG_HAMRADIO is not set # CONFIG_IRDA is not set # CONFIG_BT is not set +# CONFIG_IEEE80211 is not set # # Device Drivers @@ -328,21 +360,13 @@ CONFIG_BLK_DEV_RAM=y CONFIG_BLK_DEV_RAM_COUNT=16 CONFIG_BLK_DEV_RAM_SIZE=8192 CONFIG_BLK_DEV_INITRD=y -CONFIG_INITRAMFS_SOURCE="" # CONFIG_CDROM_PKTCDVD is not set - -# -# IO Schedulers -# -CONFIG_IOSCHED_NOOP=y -CONFIG_IOSCHED_AS=y -CONFIG_IOSCHED_DEADLINE=y -CONFIG_IOSCHED_CFQ=y CONFIG_ATA_OVER_ETH=m # # SCSI device support # +# CONFIG_RAID_ATTRS is not set CONFIG_SCSI=y CONFIG_SCSI_PROC_FS=y @@ -369,10 +393,12 @@ CONFIG_SCSI_PROC_FS=y # CONFIG_SCSI_SPI_ATTRS is not set # CONFIG_SCSI_FC_ATTRS is not set # CONFIG_SCSI_ISCSI_ATTRS is not set +# CONFIG_SCSI_SAS_ATTRS is not set # # SCSI low-level drivers # +# CONFIG_ISCSI_TCP is not set # CONFIG_SCSI_SATA is not set # CONFIG_SCSI_DEBUG is not set @@ -403,6 +429,11 @@ CONFIG_NETDEVICES=y # CONFIG_EQUALIZER is not set # CONFIG_TUN is not set +# +# PHY device support +# +# CONFIG_PHYLIB is not set + # # Ethernet (10 or 100Mbit) # @@ -439,6 +470,7 @@ CONFIG_PPP=y # CONFIG_PPP_SYNC_TTY is not set # CONFIG_PPP_DEFLATE is not set # CONFIG_PPP_BSDCOMP is not set +# CONFIG_PPP_MPPE is not set # CONFIG_PPPOE is not set CONFIG_SLIP=y CONFIG_SLIP_COMPRESSED=y @@ -541,24 +573,28 @@ CONFIG_WATCHDOG_NOWAYOUT=y # # TPM devices # +# CONFIG_TELCLOCK is not set # # I2C support # # CONFIG_I2C is not set -# CONFIG_I2C_SENSOR is not set -CONFIG_ISP1301_OMAP=y # # Hardware Monitoring support # CONFIG_HWMON=y +# CONFIG_HWMON_VID is not set # CONFIG_HWMON_DEBUG_CHIP is not set # # Misc devices # +# +# Multimedia Capabilities Port drivers +# + # # Multimedia devices # @@ -576,7 +612,6 @@ CONFIG_FB=y # CONFIG_FB_CFB_FILLRECT is not set # CONFIG_FB_CFB_COPYAREA is not set # CONFIG_FB_CFB_IMAGEBLIT is not set -# CONFIG_FB_SOFT_CURSOR is not set # CONFIG_FB_MACMODES is not set CONFIG_FB_MODE_HELPERS=y # CONFIG_FB_TILEBLITTING is not set @@ -589,6 +624,7 @@ CONFIG_FB_MODE_HELPERS=y # CONFIG_VGA_CONSOLE is not set CONFIG_DUMMY_CONSOLE=y CONFIG_FRAMEBUFFER_CONSOLE=y +# CONFIG_FRAMEBUFFER_CONSOLE_ROTATION is not set CONFIG_FONTS=y CONFIG_FONT_8x8=y CONFIG_FONT_8x16=y @@ -600,6 +636,7 @@ CONFIG_FONT_8x16=y # CONFIG_FONT_SUN8x16 is not set # CONFIG_FONT_SUN12x22 is not set # CONFIG_FONT_10x18 is not set +# CONFIG_FONT_RL is not set # # Logo configuration @@ -624,10 +661,10 @@ CONFIG_SOUND=y # Open Sound System # CONFIG_SOUND_PRIME=y +# CONFIG_OBSOLETE_OSS_DRIVER is not set # CONFIG_SOUND_MSNDCLAS is not set # CONFIG_SOUND_MSNDPIN is not set # CONFIG_SOUND_OSS is not set -# CONFIG_SOUND_AD1980 is not set # # USB support @@ -636,23 +673,22 @@ CONFIG_USB_ARCH_HAS_HCD=y CONFIG_USB_ARCH_HAS_OHCI=y # CONFIG_USB is not set +# +# NOTE: USB_STORAGE enables SCSI, and 'SCSI disk support' +# + # # USB Gadget Support # -CONFIG_USB_GADGET=y -# CONFIG_USB_GADGET_DEBUG_FILES is not set -CONFIG_USB_GADGET_SELECTED=y +# CONFIG_USB_GADGET is not set # CONFIG_USB_GADGET_NET2280 is not set # CONFIG_USB_GADGET_PXA2XX is not set # CONFIG_USB_GADGET_GOKU is not set # CONFIG_USB_GADGET_LH7A40X is not set -CONFIG_USB_GADGET_OMAP=y -CONFIG_USB_OMAP=y +# CONFIG_USB_GADGET_OMAP is not set # CONFIG_USB_GADGET_DUMMY_HCD is not set -# CONFIG_USB_GADGET_DUALSPEED is not set # CONFIG_USB_ZERO is not set -CONFIG_USB_ETH=y -CONFIG_USB_ETH_RNDIS=y +# CONFIG_USB_ETH is not set # CONFIG_USB_GADGETFS is not set # CONFIG_USB_FILE_STORAGE is not set # CONFIG_USB_G_SERIAL is not set @@ -673,10 +709,6 @@ CONFIG_EXT2_FS=y # CONFIG_REISERFS_FS is not set # CONFIG_JFS_FS is not set # CONFIG_FS_POSIX_ACL is not set - -# -# XFS support -# # CONFIG_XFS_FS is not set # CONFIG_MINIX_FS is not set CONFIG_ROMFS_FS=y @@ -685,6 +717,7 @@ CONFIG_INOTIFY=y CONFIG_DNOTIFY=y # CONFIG_AUTOFS_FS is not set # CONFIG_AUTOFS4_FS is not set +# CONFIG_FUSE_FS is not set # # CD-ROM/DVD Filesystems @@ -706,10 +739,10 @@ CONFIG_FAT_DEFAULT_CODEPAGE=437 # CONFIG_PROC_FS=y CONFIG_SYSFS=y -# CONFIG_DEVPTS_FS_XATTR is not set # CONFIG_TMPFS is not set # CONFIG_HUGETLB_PAGE is not set CONFIG_RAMFS=y +# CONFIG_RELAYFS_FS is not set # # Miscellaneous filesystems @@ -750,6 +783,7 @@ CONFIG_RPCSEC_GSS_KRB5=y # CONFIG_NCP_FS is not set # CONFIG_CODA_FS is not set # CONFIG_AFS_FS is not set +# CONFIG_9P_FS is not set # # Partition Types @@ -859,6 +893,7 @@ CONFIG_CRYPTO_DES=y # Library routines # # CONFIG_CRC_CCITT is not set +# CONFIG_CRC16 is not set CONFIG_CRC32=y # CONFIG_LIBCRC32C is not set CONFIG_ZLIB_INFLATE=y diff --git a/include/asm-arm/arch-omap/board-h4.h b/include/asm-arm/arch-omap/board-h4.h index 95ba43042a01..33ea29a41654 100644 --- a/include/asm-arm/arch-omap/board-h4.h +++ b/include/asm-arm/arch-omap/board-h4.h @@ -38,5 +38,7 @@ #define H4_CS0_BASE 0x04000000 +#define H4_CS0_BASE 0x04000000 + #endif /* __ASM_ARCH_OMAP_H4_H */ From 078abcf95cdb95c78d786dbc61ae3c22ee70fb61 Mon Sep 17 00:00:00 2001 From: Richard Purdie Date: Thu, 10 Nov 2005 17:42:29 +0000 Subject: [PATCH 097/121] [ARM] 3096/1: Add SharpSL Zaurus power and battery management core driver Patch from Richard Purdie This patch adds a power and battery management core driver which with the addition of the right device files, supports the c7x0 and cxx00 series of Sharp Zaurus handhelds. The driver is complex for several reasons. Battery charging is manually monitored and controlled. When suspended, the device needs to periodically partially resume, check the charging status and then re-suspend. It does without bothering the higher linux layers as a full resume and re-suspend is unnecessary. The code is carefully written to avoid interrupts or calling code outside the module under these circumstances. It also vets the various wake up sources and monitors the device's power situation. Hooks to limit the backlight intensity and to notify the battery monitoring code of backlight events are connected/added as the backlight is one of the biggest users of power on the device. Signed-off-by: Richard Purdie Signed-off-by: Russell King --- arch/arm/mach-pxa/sharpsl.h | 87 +++ arch/arm/mach-pxa/sharpsl_pm.c | 992 +++++++++++++++++++++++++++++ drivers/video/backlight/corgi_bl.c | 6 + include/asm-arm/arch-pxa/sharpsl.h | 8 +- 4 files changed, 1092 insertions(+), 1 deletion(-) create mode 100644 arch/arm/mach-pxa/sharpsl_pm.c diff --git a/arch/arm/mach-pxa/sharpsl.h b/arch/arm/mach-pxa/sharpsl.h index 3977a77aacdd..4879c0f7da72 100644 --- a/arch/arm/mach-pxa/sharpsl.h +++ b/arch/arm/mach-pxa/sharpsl.h @@ -32,3 +32,90 @@ void corgi_put_hsync(void); void spitz_put_hsync(void); void corgi_wait_hsync(void); void spitz_wait_hsync(void); + +/* + * SharpSL Battery/PM Driver + */ + +struct sharpsl_charger_machinfo { + void (*init)(void); + int gpio_acin; + int gpio_batfull; + int gpio_batlock; + int gpio_fatal; + int (*status_acin)(void); + void (*discharge)(int); + void (*discharge1)(int); + void (*charge)(int); + void (*chargeled)(int); + void (*measure_temp)(int); + void (*presuspend)(void); + void (*postsuspend)(void); + unsigned long (*charger_wakeup)(void); + int (*should_wakeup)(unsigned int resume_on_alarm); + int bat_levels; + struct battery_thresh *bat_levels_noac; + struct battery_thresh *bat_levels_acin; + int status_high_acin; + int status_low_acin; + int status_high_noac; + int status_low_noac; +}; + +struct battery_thresh { + int voltage; + int percentage; +}; + +struct battery_stat { + int ac_status; /* APM AC Present/Not Present */ + int mainbat_status; /* APM Main Battery Status */ + int mainbat_percent; /* Main Battery Percentage Charge */ + int mainbat_voltage; /* Main Battery Voltage */ +}; + +struct sharpsl_pm_status { + struct device *dev; + struct timer_list ac_timer; + struct timer_list chrg_full_timer; + + int charge_mode; +#define CHRG_ERROR (-1) +#define CHRG_OFF (0) +#define CHRG_ON (1) +#define CHRG_DONE (2) + + unsigned int flags; +#define SHARPSL_SUSPENDED (1 << 0) /* Device is Suspended */ +#define SHARPSL_ALARM_ACTIVE (1 << 1) /* Alarm is for charging event (not user) */ +#define SHARPSL_BL_LIMIT (1 << 2) /* Backlight Intensity Limited */ +#define SHARPSL_APM_QUEUED (1 << 3) /* APM Event Queued */ +#define SHARPSL_DO_OFFLINE_CHRG (1 << 4) /* Trigger the offline charger */ + + int full_count; + unsigned long charge_start_time; + struct sharpsl_charger_machinfo *machinfo; + struct battery_stat battstat; +}; + +extern struct sharpsl_pm_status sharpsl_pm; +extern struct battery_thresh spitz_battery_levels_acin[]; +extern struct battery_thresh spitz_battery_levels_noac[]; + +#define READ_GPIO_BIT(x) (GPLR(x) & GPIO_bit(x)) + +#define SHARPSL_LED_ERROR 2 +#define SHARPSL_LED_ON 1 +#define SHARPSL_LED_OFF 0 + +#define CHARGE_ON() sharpsl_pm.machinfo->charge(1) +#define CHARGE_OFF() sharpsl_pm.machinfo->charge(0) +#define CHARGE_LED_ON() sharpsl_pm.machinfo->chargeled(SHARPSL_LED_ON) +#define CHARGE_LED_OFF() sharpsl_pm.machinfo->chargeled(SHARPSL_LED_OFF) +#define CHARGE_LED_ERR() sharpsl_pm.machinfo->chargeled(SHARPSL_LED_ERROR) +#define DISCHARGE_ON() sharpsl_pm.machinfo->discharge(1) +#define DISCHARGE_OFF() sharpsl_pm.machinfo->discharge(0) +#define STATUS_AC_IN sharpsl_pm.machinfo->status_acin() +#define STATUS_BATT_LOCKED READ_GPIO_BIT(sharpsl_pm.machinfo->gpio_batlock) +#define STATUS_CHRG_FULL READ_GPIO_BIT(sharpsl_pm.machinfo->gpio_batfull) +#define STATUS_FATAL READ_GPIO_BIT(sharpsl_pm.machinfo->gpio_fatal) diff --git a/arch/arm/mach-pxa/sharpsl_pm.c b/arch/arm/mach-pxa/sharpsl_pm.c new file mode 100644 index 000000000000..6c9e871c53d8 --- /dev/null +++ b/arch/arm/mach-pxa/sharpsl_pm.c @@ -0,0 +1,992 @@ +/* + * Battery and Power Management code for the Sharp SL-C7xx and SL-Cxx00 + * series of PDAs + * + * Copyright (c) 2004-2005 Richard Purdie + * + * Based on code written by Sharp for 2.4 kernels + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + */ + +#undef DEBUG + +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#include +#include +#include +#include "sharpsl.h" + +/* + * Constants + */ +#define SHARPSL_CHARGE_ON_TIME_INTERVAL (msecs_to_jiffies(1*60*1000)) /* 1 min */ +#define SHARPSL_CHARGE_FINISH_TIME (msecs_to_jiffies(10*60*1000)) /* 10 min */ +#define SHARPSL_BATCHK_TIME (msecs_to_jiffies(15*1000)) /* 15 sec */ +#define SHARPSL_BATCHK_TIME_SUSPEND (60*10) /* 10 min */ +#define SHARPSL_WAIT_CO_TIME 15 /* 15 sec */ +#define SHARPSL_WAIT_DISCHARGE_ON 100 /* 100 msec */ +#define SHARPSL_CHECK_BATTERY_WAIT_TIME_TEMP 10 /* 10 msec */ +#define SHARPSL_CHECK_BATTERY_WAIT_TIME_VOLT 10 /* 10 msec */ +#define SHARPSL_CHECK_BATTERY_WAIT_TIME_JKVAD 10 /* 10 msec */ +#define SHARPSL_CHARGE_WAIT_TIME 15 /* 15 msec */ +#define SHARPSL_CHARGE_CO_CHECK_TIME 5 /* 5 msec */ +#define SHARPSL_CHARGE_RETRY_CNT 1 /* eqv. 10 min */ + +#define SHARPSL_CHARGE_ON_VOLT 0x99 /* 2.9V */ +#define SHARPSL_CHARGE_ON_TEMP 0xe0 /* 2.9V */ +#define SHARPSL_CHARGE_ON_JKVAD_HIGH 0x9b /* 6V */ +#define SHARPSL_CHARGE_ON_JKVAD_LOW 0x34 /* 2V */ +#define SHARPSL_FATAL_ACIN_VOLT 182 /* 3.45V */ +#define SHARPSL_FATAL_NOACIN_VOLT 170 /* 3.40V */ + +struct battery_thresh spitz_battery_levels_acin[] = { + { 213, 100}, + { 212, 98}, + { 211, 95}, + { 210, 93}, + { 209, 90}, + { 208, 88}, + { 207, 85}, + { 206, 83}, + { 205, 80}, + { 204, 78}, + { 203, 75}, + { 202, 73}, + { 201, 70}, + { 200, 68}, + { 199, 65}, + { 198, 63}, + { 197, 60}, + { 196, 58}, + { 195, 55}, + { 194, 53}, + { 193, 50}, + { 192, 48}, + { 192, 45}, + { 191, 43}, + { 191, 40}, + { 190, 38}, + { 190, 35}, + { 189, 33}, + { 188, 30}, + { 187, 28}, + { 186, 25}, + { 185, 23}, + { 184, 20}, + { 183, 18}, + { 182, 15}, + { 181, 13}, + { 180, 10}, + { 179, 8}, + { 178, 5}, + { 0, 0}, +}; + +struct battery_thresh spitz_battery_levels_noac[] = { + { 213, 100}, + { 212, 98}, + { 211, 95}, + { 210, 93}, + { 209, 90}, + { 208, 88}, + { 207, 85}, + { 206, 83}, + { 205, 80}, + { 204, 78}, + { 203, 75}, + { 202, 73}, + { 201, 70}, + { 200, 68}, + { 199, 65}, + { 198, 63}, + { 197, 60}, + { 196, 58}, + { 195, 55}, + { 194, 53}, + { 193, 50}, + { 192, 48}, + { 191, 45}, + { 190, 43}, + { 189, 40}, + { 188, 38}, + { 187, 35}, + { 186, 33}, + { 185, 30}, + { 184, 28}, + { 183, 25}, + { 182, 23}, + { 181, 20}, + { 180, 18}, + { 179, 15}, + { 178, 13}, + { 177, 10}, + { 176, 8}, + { 175, 5}, + { 0, 0}, +}; + +/* MAX1111 Commands */ +#define MAXCTRL_PD0 1u << 0 +#define MAXCTRL_PD1 1u << 1 +#define MAXCTRL_SGL 1u << 2 +#define MAXCTRL_UNI 1u << 3 +#define MAXCTRL_SEL_SH 4 +#define MAXCTRL_STR 1u << 7 + +/* MAX1111 Channel Definitions */ +#define BATT_AD 4u +#define BATT_THM 2u +#define JK_VAD 6u + + +/* + * Prototypes + */ +static int sharpsl_read_MainBattery(void); +static int sharpsl_off_charge_battery(void); +static int sharpsl_check_battery(int mode); +static int sharpsl_ac_check(void); +static int sharpsl_fatal_check(void); +static int sharpsl_average_value(int ad); +static void sharpsl_average_clear(void); +static void sharpsl_charge_toggle(void *private_); +static void sharpsl_battery_thread(void *private_); + + +/* + * Variables + */ +struct sharpsl_pm_status sharpsl_pm; +DECLARE_WORK(toggle_charger, sharpsl_charge_toggle, NULL); +DECLARE_WORK(sharpsl_bat, sharpsl_battery_thread, NULL); + + +static int get_percentage(int voltage) +{ + int i = sharpsl_pm.machinfo->bat_levels - 1; + struct battery_thresh *thresh; + + if (sharpsl_pm.charge_mode == CHRG_ON) + thresh=sharpsl_pm.machinfo->bat_levels_acin; + else + thresh=sharpsl_pm.machinfo->bat_levels_noac; + + while (i > 0 && (voltage > thresh[i].voltage)) + i--; + + return thresh[i].percentage; +} + +static int get_apm_status(int voltage) +{ + int low_thresh, high_thresh; + + if (sharpsl_pm.charge_mode == CHRG_ON) { + high_thresh = sharpsl_pm.machinfo->status_high_acin; + low_thresh = sharpsl_pm.machinfo->status_low_acin; + } else { + high_thresh = sharpsl_pm.machinfo->status_high_noac; + low_thresh = sharpsl_pm.machinfo->status_low_noac; + } + + if (voltage >= high_thresh) + return APM_BATTERY_STATUS_HIGH; + if (voltage >= low_thresh) + return APM_BATTERY_STATUS_LOW; + return APM_BATTERY_STATUS_CRITICAL; +} + +void sharpsl_battery_kick(void) +{ + schedule_delayed_work(&sharpsl_bat, msecs_to_jiffies(125)); +} +EXPORT_SYMBOL(sharpsl_battery_kick); + + +static void sharpsl_battery_thread(void *private_) +{ + int voltage, percent, apm_status, i = 0; + + if (!sharpsl_pm.machinfo) + return; + + sharpsl_pm.battstat.ac_status = (!(STATUS_AC_IN) ? APM_AC_OFFLINE : APM_AC_ONLINE); + + /* Corgi cannot confirm when battery fully charged so periodically kick! */ + if (machine_is_corgi() && (sharpsl_pm.charge_mode == CHRG_ON) + && time_after(jiffies, sharpsl_pm.charge_start_time + SHARPSL_CHARGE_ON_TIME_INTERVAL)) + schedule_work(&toggle_charger); + + while(1) { + voltage = sharpsl_read_MainBattery(); + if (voltage > 0) break; + if (i++ > 5) { + voltage = sharpsl_pm.machinfo->bat_levels_noac[0].voltage; + dev_warn(sharpsl_pm.dev, "Warning: Cannot read main battery!\n"); + break; + } + } + + voltage = sharpsl_average_value(voltage); + apm_status = get_apm_status(voltage); + percent = get_percentage(voltage); + + /* At low battery voltages, the voltage has a tendency to start + creeping back up so we try to avoid this here */ + if ((sharpsl_pm.battstat.ac_status == APM_AC_ONLINE) || (apm_status == APM_BATTERY_STATUS_HIGH) || percent <= sharpsl_pm.battstat.mainbat_percent) { + sharpsl_pm.battstat.mainbat_voltage = voltage; + sharpsl_pm.battstat.mainbat_status = apm_status; + sharpsl_pm.battstat.mainbat_percent = percent; + } + + dev_dbg(sharpsl_pm.dev, "Battery: voltage: %d, status: %d, percentage: %d, time: %d\n", voltage, + sharpsl_pm.battstat.mainbat_status, sharpsl_pm.battstat.mainbat_percent, jiffies); + + /* If battery is low. limit backlight intensity to save power. */ + if ((sharpsl_pm.battstat.ac_status != APM_AC_ONLINE) + && ((sharpsl_pm.battstat.mainbat_status == APM_BATTERY_STATUS_LOW) || + (sharpsl_pm.battstat.mainbat_status == APM_BATTERY_STATUS_CRITICAL))) { + if (!(sharpsl_pm.flags & SHARPSL_BL_LIMIT)) { + corgibl_limit_intensity(1); + sharpsl_pm.flags |= SHARPSL_BL_LIMIT; + } + } else if (sharpsl_pm.flags & SHARPSL_BL_LIMIT) { + corgibl_limit_intensity(0); + sharpsl_pm.flags &= ~SHARPSL_BL_LIMIT; + } + + /* Suspend if critical battery level */ + if ((sharpsl_pm.battstat.ac_status != APM_AC_ONLINE) + && (sharpsl_pm.battstat.mainbat_status == APM_BATTERY_STATUS_CRITICAL) + && !(sharpsl_pm.flags & SHARPSL_APM_QUEUED)) { + sharpsl_pm.flags |= SHARPSL_APM_QUEUED; + dev_err(sharpsl_pm.dev, "Fatal Off\n"); + apm_queue_event(APM_CRITICAL_SUSPEND); + } + + schedule_delayed_work(&sharpsl_bat, SHARPSL_BATCHK_TIME); +} + +static void sharpsl_charge_on(void) +{ + dev_dbg(sharpsl_pm.dev, "Turning Charger On\n"); + + sharpsl_pm.full_count = 0; + sharpsl_pm.charge_mode = CHRG_ON; + schedule_delayed_work(&toggle_charger, msecs_to_jiffies(250)); + schedule_delayed_work(&sharpsl_bat, msecs_to_jiffies(500)); +} + +static void sharpsl_charge_off(void) +{ + dev_dbg(sharpsl_pm.dev, "Turning Charger Off\n"); + + CHARGE_OFF(); + CHARGE_LED_OFF(); + sharpsl_pm.charge_mode = CHRG_OFF; + + schedule_work(&sharpsl_bat); +} + +static void sharpsl_charge_error(void) +{ + CHARGE_LED_ERR(); + CHARGE_OFF(); + sharpsl_pm.charge_mode = CHRG_ERROR; +} + +static void sharpsl_charge_toggle(void *private_) +{ + dev_dbg(sharpsl_pm.dev, "Toogling Charger at time: %lx\n", jiffies); + + if (STATUS_AC_IN == 0) { + sharpsl_charge_off(); + return; + } else if ((sharpsl_check_battery(1) < 0) || (sharpsl_ac_check() < 0)) { + sharpsl_charge_error(); + return; + } + + CHARGE_LED_ON(); + CHARGE_OFF(); + mdelay(SHARPSL_CHARGE_WAIT_TIME); + CHARGE_ON(); + + sharpsl_pm.charge_start_time = jiffies; +} + +static void sharpsl_ac_timer(unsigned long data) +{ + int acin = STATUS_AC_IN; + + dev_dbg(sharpsl_pm.dev, "AC Status: %d\n",acin); + + sharpsl_average_clear(); + if (acin && (sharpsl_pm.charge_mode != CHRG_ON)) + sharpsl_charge_on(); + else if (sharpsl_pm.charge_mode == CHRG_ON) + sharpsl_charge_off(); + + schedule_work(&sharpsl_bat); +} + + +static irqreturn_t sharpsl_ac_isr(int irq, void *dev_id, struct pt_regs *fp) +{ + /* Delay the event slightly to debounce */ + /* Must be a smaller delay than the chrg_full_isr below */ + mod_timer(&sharpsl_pm.ac_timer, jiffies + msecs_to_jiffies(250)); + + return IRQ_HANDLED; +} + +static void sharpsl_chrg_full_timer(unsigned long data) +{ + dev_dbg(sharpsl_pm.dev, "Charge Full at time: %lx\n", jiffies); + + sharpsl_pm.full_count++; + + if (STATUS_AC_IN == 0) { + dev_dbg(sharpsl_pm.dev, "Charge Full: AC removed - stop charging!\n"); + if (sharpsl_pm.charge_mode == CHRG_ON) + sharpsl_charge_off(); + } else if (sharpsl_pm.full_count < 2) { + dev_dbg(sharpsl_pm.dev, "Charge Full: Count too low\n"); + schedule_work(&toggle_charger); + } else if (time_after(jiffies, sharpsl_pm.charge_start_time + SHARPSL_CHARGE_FINISH_TIME)) { + dev_dbg(sharpsl_pm.dev, "Charge Full: Interrupt generated too slowly - retry.\n"); + schedule_work(&toggle_charger); + } else { + sharpsl_charge_off(); + sharpsl_pm.charge_mode = CHRG_DONE; + dev_dbg(sharpsl_pm.dev, "Charge Full: Charging Finished\n"); + } +} + +/* Charging Finished Interrupt (Not present on Corgi) */ +/* Can trigger at the same time as an AC staus change so + delay until after that has been processed */ +static irqreturn_t sharpsl_chrg_full_isr(int irq, void *dev_id, struct pt_regs *fp) +{ + if (sharpsl_pm.flags & SHARPSL_SUSPENDED) + return IRQ_HANDLED; + + /* delay until after any ac interrupt */ + mod_timer(&sharpsl_pm.chrg_full_timer, jiffies + msecs_to_jiffies(500)); + + return IRQ_HANDLED; +} + +static irqreturn_t sharpsl_fatal_isr(int irq, void *dev_id, struct pt_regs *fp) +{ + int is_fatal = 0; + + if (STATUS_BATT_LOCKED == 0) { + dev_err(sharpsl_pm.dev, "Battery now Unlocked! Suspending.\n"); + is_fatal = 1; + } + + if (sharpsl_pm.machinfo->gpio_fatal && (STATUS_FATAL == 0)) { + dev_err(sharpsl_pm.dev, "Fatal Batt Error! Suspending.\n"); + is_fatal = 1; + } + + if (!(sharpsl_pm.flags & SHARPSL_APM_QUEUED) && is_fatal) { + sharpsl_pm.flags |= SHARPSL_APM_QUEUED; + apm_queue_event(APM_CRITICAL_SUSPEND); + } + + return IRQ_HANDLED; +} + +/* + * Maintain an average of the last 10 readings + */ +#define SHARPSL_CNV_VALUE_NUM 10 +static int sharpsl_ad_index; + +static void sharpsl_average_clear(void) +{ + sharpsl_ad_index = 0; +} + +static int sharpsl_average_value(int ad) +{ + int i, ad_val = 0; + static int sharpsl_ad[SHARPSL_CNV_VALUE_NUM+1]; + + if (sharpsl_pm.battstat.mainbat_status != APM_BATTERY_STATUS_HIGH) { + sharpsl_ad_index = 0; + return ad; + } + + sharpsl_ad[sharpsl_ad_index] = ad; + sharpsl_ad_index++; + if (sharpsl_ad_index >= SHARPSL_CNV_VALUE_NUM) { + for (i=0; i < (SHARPSL_CNV_VALUE_NUM-1); i++) + sharpsl_ad[i] = sharpsl_ad[i+1]; + sharpsl_ad_index = SHARPSL_CNV_VALUE_NUM - 1; + } + for (i=0; i < sharpsl_ad_index; i++) + ad_val += sharpsl_ad[i]; + + return (ad_val / sharpsl_ad_index); +} + + +/* + * Read MAX1111 ADC + */ +static int read_max1111(int channel) +{ + return corgi_ssp_max1111_get((channel << MAXCTRL_SEL_SH) | MAXCTRL_PD0 | MAXCTRL_PD1 + | MAXCTRL_SGL | MAXCTRL_UNI | MAXCTRL_STR); +} + +static int sharpsl_read_MainBattery(void) +{ + return read_max1111(BATT_AD); +} + +static int sharpsl_read_Temp(void) +{ + int temp; + + sharpsl_pm.machinfo->measure_temp(1); + + mdelay(SHARPSL_CHECK_BATTERY_WAIT_TIME_TEMP); + temp = read_max1111(BATT_THM); + + sharpsl_pm.machinfo->measure_temp(0); + + return temp; +} + +static int sharpsl_read_jkvad(void) +{ + return read_max1111(JK_VAD); +} + +/* + * Take an array of 5 integers, remove the maximum and minimum values + * and return the average. + */ +static int get_select_val(int *val) +{ + int i, j, k, temp, sum = 0; + + /* Find MAX val */ + temp = val[0]; + j=0; + for (i=1; i<5; i++) { + if (temp < val[i]) { + temp = val[i]; + j = i; + } + } + + /* Find MIN val */ + temp = val[4]; + k=4; + for (i=3; i>=0; i--) { + if (temp > val[i]) { + temp = val[i]; + k = i; + } + } + + for (i=0; i<5; i++) + if (i != j && i != k ) + sum += val[i]; + + dev_dbg(sharpsl_pm.dev, "Average: %d from values: %d, %d, %d, %d, %d\n", sum/3, val[0], val[1], val[2], val[3], val[4]); + + return (sum/3); +} + +/* mode 0 - Check temperature and voltage + * 1 - Check temperature only */ +static int sharpsl_check_battery(int mode) +{ + int val, i, buff[5]; + + /* Check battery temperature */ + for (i=0; i<5; i++) { + mdelay(SHARPSL_CHECK_BATTERY_WAIT_TIME_TEMP); + buff[i] = sharpsl_read_Temp(); + } + + val = get_select_val(buff); + + dev_dbg(sharpsl_pm.dev, "Temperature: %d\n", val); + if (val > SHARPSL_CHARGE_ON_TEMP) + return -1; + if (mode == 1) + return 0; + + /* disable charge, enable discharge */ + CHARGE_OFF(); + DISCHARGE_ON(); + mdelay(SHARPSL_WAIT_DISCHARGE_ON); + + if (sharpsl_pm.machinfo->discharge1) + sharpsl_pm.machinfo->discharge1(1); + + /* Check battery voltage */ + for (i=0; i<5; i++) { + buff[i] = sharpsl_read_MainBattery(); + mdelay(SHARPSL_CHECK_BATTERY_WAIT_TIME_VOLT); + } + + if (sharpsl_pm.machinfo->discharge1) + sharpsl_pm.machinfo->discharge1(0); + + DISCHARGE_OFF(); + + val = get_select_val(buff); + dev_dbg(sharpsl_pm.dev, "Battery Voltage: %d\n", val); + + if (val < SHARPSL_CHARGE_ON_VOLT) + return -1; + + return 0; +} + +static int sharpsl_ac_check(void) +{ + int temp, i, buff[5]; + + for (i=0; i<5; i++) { + buff[i] = sharpsl_read_jkvad(); + mdelay(SHARPSL_CHECK_BATTERY_WAIT_TIME_JKVAD); + } + + temp = get_select_val(buff); + dev_dbg(sharpsl_pm.dev, "AC Voltage: %d\n",temp); + + if ((temp > SHARPSL_CHARGE_ON_JKVAD_HIGH) || (temp < SHARPSL_CHARGE_ON_JKVAD_LOW)) { + dev_err(sharpsl_pm.dev, "Error: AC check failed.\n"); + return -1; + } + + return 0; +} + +#ifdef CONFIG_PM +static int sharpsl_pm_suspend(struct device *dev, pm_message_t state) +{ + sharpsl_pm.flags |= SHARPSL_SUSPENDED; + flush_scheduled_work(); + + if (sharpsl_pm.charge_mode == CHRG_ON) + sharpsl_pm.flags |= SHARPSL_DO_OFFLINE_CHRG; + else + sharpsl_pm.flags &= ~SHARPSL_DO_OFFLINE_CHRG; + + return 0; +} + +static int sharpsl_pm_resume(struct device *dev) +{ + /* Clear the reset source indicators as they break the bootloader upon reboot */ + RCSR = 0x0f; + sharpsl_average_clear(); + sharpsl_pm.flags &= ~SHARPSL_APM_QUEUED; + sharpsl_pm.flags &= ~SHARPSL_SUSPENDED; + + return 0; +} + +static void corgi_goto_sleep(unsigned long alarm_time, unsigned int alarm_enable, suspend_state_t state) +{ + dev_dbg(sharpsl_pm.dev, "Time is: %08x\n",RCNR); + + dev_dbg(sharpsl_pm.dev, "Offline Charge Activate = %d\n",sharpsl_pm.flags & SHARPSL_DO_OFFLINE_CHRG); + /* not charging and AC-IN! */ + + if ((sharpsl_pm.flags & SHARPSL_DO_OFFLINE_CHRG) && (STATUS_AC_IN != 0)) { + dev_dbg(sharpsl_pm.dev, "Activating Offline Charger...\n"); + sharpsl_pm.charge_mode = CHRG_OFF; + sharpsl_pm.flags &= ~SHARPSL_DO_OFFLINE_CHRG; + sharpsl_off_charge_battery(); + } + + sharpsl_pm.machinfo->presuspend(); + + PEDR = 0xffffffff; /* clear it */ + + sharpsl_pm.flags &= ~SHARPSL_ALARM_ACTIVE; + if ((sharpsl_pm.charge_mode == CHRG_ON) && ((alarm_enable && ((alarm_time - RCNR) > (SHARPSL_BATCHK_TIME_SUSPEND + 30))) || !alarm_enable)) { + RTSR &= RTSR_ALE; + RTAR = RCNR + SHARPSL_BATCHK_TIME_SUSPEND; + dev_dbg(sharpsl_pm.dev, "Charging alarm at: %08x\n",RTAR); + sharpsl_pm.flags |= SHARPSL_ALARM_ACTIVE; + } else if (alarm_enable) { + RTSR &= RTSR_ALE; + RTAR = alarm_time; + dev_dbg(sharpsl_pm.dev, "User alarm at: %08x\n",RTAR); + } else { + dev_dbg(sharpsl_pm.dev, "No alarms set.\n"); + } + + pxa_pm_enter(state); + + sharpsl_pm.machinfo->postsuspend(); + + dev_dbg(sharpsl_pm.dev, "Corgi woken up from suspend: %08x\n",PEDR); +} + +static int corgi_enter_suspend(unsigned long alarm_time, unsigned int alarm_enable, suspend_state_t state) +{ + if (!sharpsl_pm.machinfo->should_wakeup(!(sharpsl_pm.flags & SHARPSL_ALARM_ACTIVE) && alarm_enable) ) + { + if (!(sharpsl_pm.flags & SHARPSL_ALARM_ACTIVE)) { + dev_dbg(sharpsl_pm.dev, "No user triggered wakeup events and not charging. Strange. Suspend.\n"); + corgi_goto_sleep(alarm_time, alarm_enable, state); + return 1; + } + if(sharpsl_off_charge_battery()) { + dev_dbg(sharpsl_pm.dev, "Charging. Suspend...\n"); + corgi_goto_sleep(alarm_time, alarm_enable, state); + return 1; + } + dev_dbg(sharpsl_pm.dev, "User triggered wakeup in offline charger.\n"); + } + + if ((STATUS_BATT_LOCKED == 0) || (sharpsl_fatal_check() < 0) ) + { + dev_err(sharpsl_pm.dev, "Fatal condition. Suspend.\n"); + corgi_goto_sleep(alarm_time, alarm_enable, state); + return 1; + } + + return 0; +} + +static int corgi_pxa_pm_enter(suspend_state_t state) +{ + unsigned long alarm_time = RTAR; + unsigned int alarm_status = ((RTSR & RTSR_ALE) != 0); + + dev_dbg(sharpsl_pm.dev, "SharpSL suspending for first time.\n"); + + corgi_goto_sleep(alarm_time, alarm_status, state); + + while (corgi_enter_suspend(alarm_time,alarm_status,state)) + {} + + dev_dbg(sharpsl_pm.dev, "SharpSL resuming...\n"); + + return 0; +} +#endif + + +/* + * Check for fatal battery errors + * Fatal returns -1 + */ +static int sharpsl_fatal_check(void) +{ + int buff[5], temp, i, acin; + + dev_dbg(sharpsl_pm.dev, "sharpsl_fatal_check entered\n"); + + /* Check AC-Adapter */ + acin = STATUS_AC_IN; + + if (acin && (sharpsl_pm.charge_mode == CHRG_ON)) { + CHARGE_OFF(); + udelay(100); + DISCHARGE_ON(); /* enable discharge */ + mdelay(SHARPSL_WAIT_DISCHARGE_ON); + } + + if (sharpsl_pm.machinfo->discharge1) + sharpsl_pm.machinfo->discharge1(1); + + /* Check battery : check inserting battery ? */ + for (i=0; i<5; i++) { + buff[i] = sharpsl_read_MainBattery(); + mdelay(SHARPSL_CHECK_BATTERY_WAIT_TIME_VOLT); + } + + if (sharpsl_pm.machinfo->discharge1) + sharpsl_pm.machinfo->discharge1(0); + + if (acin && (sharpsl_pm.charge_mode == CHRG_ON)) { + udelay(100); + CHARGE_ON(); + DISCHARGE_OFF(); + } + + temp = get_select_val(buff); + dev_dbg(sharpsl_pm.dev, "sharpsl_fatal_check: acin: %d, discharge voltage: %d, no discharge: %d\n", acin, temp, sharpsl_read_MainBattery()); + + if ((acin && (temp < SHARPSL_FATAL_ACIN_VOLT)) || + (!acin && (temp < SHARPSL_FATAL_NOACIN_VOLT))) + return -1; + return 0; +} + +static int sharpsl_off_charge_error(void) +{ + dev_err(sharpsl_pm.dev, "Offline Charger: Error occured.\n"); + CHARGE_OFF(); + CHARGE_LED_ERR(); + sharpsl_pm.charge_mode = CHRG_ERROR; + return 1; +} + +/* + * Charging Control while suspended + * Return 1 - go straight to sleep + * Return 0 - sleep or wakeup depending on other factors + */ +static int sharpsl_off_charge_battery(void) +{ + int time; + + dev_dbg(sharpsl_pm.dev, "Charge Mode: %d\n", sharpsl_pm.charge_mode); + + if (sharpsl_pm.charge_mode == CHRG_OFF) { + dev_dbg(sharpsl_pm.dev, "Offline Charger: Step 1\n"); + + /* AC Check */ + if ((sharpsl_ac_check() < 0) || (sharpsl_check_battery(1) < 0)) + return sharpsl_off_charge_error(); + + /* Start Charging */ + CHARGE_LED_ON(); + CHARGE_OFF(); + mdelay(SHARPSL_CHARGE_WAIT_TIME); + CHARGE_ON(); + + sharpsl_pm.charge_mode = CHRG_ON; + sharpsl_pm.full_count = 0; + + return 1; + } else if (sharpsl_pm.charge_mode != CHRG_ON) { + return 1; + } + + if (sharpsl_pm.full_count == 0) { + int time; + + dev_dbg(sharpsl_pm.dev, "Offline Charger: Step 2\n"); + + if (sharpsl_check_battery(0) < 0) + return sharpsl_off_charge_error(); + + CHARGE_OFF(); + mdelay(SHARPSL_CHARGE_WAIT_TIME); + CHARGE_ON(); + sharpsl_pm.charge_mode = CHRG_ON; + + mdelay(SHARPSL_CHARGE_CO_CHECK_TIME); + + time = RCNR; + while(1) { + /* Check if any wakeup event had occured */ + if (sharpsl_pm.machinfo->charger_wakeup() != 0) + return 0; + /* Check for timeout */ + if ((RCNR - time) > SHARPSL_WAIT_CO_TIME) + return 1; + if (STATUS_CHRG_FULL) { + dev_dbg(sharpsl_pm.dev, "Offline Charger: Charge full occured. Retrying to check\n"); + sharpsl_pm.full_count++; + CHARGE_OFF(); + mdelay(SHARPSL_CHARGE_WAIT_TIME); + CHARGE_ON(); + return 1; + } + } + } + + dev_dbg(sharpsl_pm.dev, "Offline Charger: Step 3\n"); + + mdelay(SHARPSL_CHARGE_CO_CHECK_TIME); + + time = RCNR; + while(1) { + /* Check if any wakeup event had occured */ + if (sharpsl_pm.machinfo->charger_wakeup() != 0) + return 0; + /* Check for timeout */ + if ((RCNR-time) > SHARPSL_WAIT_CO_TIME) { + if (sharpsl_pm.full_count > SHARPSL_CHARGE_RETRY_CNT) { + dev_dbg(sharpsl_pm.dev, "Offline Charger: Not charged sufficiently. Retrying.\n"); + sharpsl_pm.full_count = 0; + } + sharpsl_pm.full_count++; + return 1; + } + if (STATUS_CHRG_FULL) { + dev_dbg(sharpsl_pm.dev, "Offline Charger: Charging complete.\n"); + CHARGE_LED_OFF(); + CHARGE_OFF(); + sharpsl_pm.charge_mode = CHRG_DONE; + return 1; + } + } +} + + +static ssize_t battery_percentage_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + return sprintf(buf, "%d\n",sharpsl_pm.battstat.mainbat_percent); +} + +static ssize_t battery_voltage_show(struct device *dev, struct device_attribute *attr, char *buf) +{ + return sprintf(buf, "%d\n",sharpsl_pm.battstat.mainbat_voltage); +} + +static DEVICE_ATTR(battery_percentage, 0444, battery_percentage_show, NULL); +static DEVICE_ATTR(battery_voltage, 0444, battery_voltage_show, NULL); + +extern void (*apm_get_power_status)(struct apm_power_info *); + +static void sharpsl_apm_get_power_status(struct apm_power_info *info) +{ + info->ac_line_status = sharpsl_pm.battstat.ac_status; + + if (sharpsl_pm.charge_mode == CHRG_ON) + info->battery_status = APM_BATTERY_STATUS_CHARGING; + else + info->battery_status = sharpsl_pm.battstat.mainbat_status; + + info->battery_flag = (1 << info->battery_status); + info->battery_life = sharpsl_pm.battstat.mainbat_percent; +} + +static struct pm_ops sharpsl_pm_ops = { + .pm_disk_mode = PM_DISK_FIRMWARE, + .prepare = pxa_pm_prepare, + .enter = corgi_pxa_pm_enter, + .finish = pxa_pm_finish, +}; + +static int __init sharpsl_pm_probe(struct device *dev) +{ + if (!dev->platform_data) + return -EINVAL; + + sharpsl_pm.dev = dev; + sharpsl_pm.machinfo = dev->platform_data; + sharpsl_pm.charge_mode = CHRG_OFF; + sharpsl_pm.flags = 0; + + sharpsl_pm.machinfo->init(); + + init_timer(&sharpsl_pm.ac_timer); + sharpsl_pm.ac_timer.function = sharpsl_ac_timer; + + init_timer(&sharpsl_pm.chrg_full_timer); + sharpsl_pm.chrg_full_timer.function = sharpsl_chrg_full_timer; + + pxa_gpio_mode(sharpsl_pm.machinfo->gpio_acin | GPIO_IN); + pxa_gpio_mode(sharpsl_pm.machinfo->gpio_batfull | GPIO_IN); + pxa_gpio_mode(sharpsl_pm.machinfo->gpio_batlock | GPIO_IN); + + /* Register interrupt handlers */ + if (request_irq(IRQ_GPIO(sharpsl_pm.machinfo->gpio_acin), sharpsl_ac_isr, SA_INTERRUPT, "AC Input Detect", sharpsl_ac_isr)) { + dev_err(sharpsl_pm.dev, "Could not get irq %d.\n", IRQ_GPIO(sharpsl_pm.machinfo->gpio_acin)); + } + else set_irq_type(IRQ_GPIO(sharpsl_pm.machinfo->gpio_acin),IRQT_BOTHEDGE); + + if (request_irq(IRQ_GPIO(sharpsl_pm.machinfo->gpio_batlock), sharpsl_fatal_isr, SA_INTERRUPT, "Battery Cover", sharpsl_fatal_isr)) { + dev_err(sharpsl_pm.dev, "Could not get irq %d.\n", IRQ_GPIO(sharpsl_pm.machinfo->gpio_batlock)); + } + else set_irq_type(IRQ_GPIO(sharpsl_pm.machinfo->gpio_batlock),IRQT_FALLING); + + if (sharpsl_pm.machinfo->gpio_fatal) { + if (request_irq(IRQ_GPIO(sharpsl_pm.machinfo->gpio_fatal), sharpsl_fatal_isr, SA_INTERRUPT, "Fatal Battery", sharpsl_fatal_isr)) { + dev_err(sharpsl_pm.dev, "Could not get irq %d.\n", IRQ_GPIO(sharpsl_pm.machinfo->gpio_fatal)); + } + else set_irq_type(IRQ_GPIO(sharpsl_pm.machinfo->gpio_fatal),IRQT_FALLING); + } + + if (!machine_is_corgi()) + { + /* Register interrupt handler. */ + if (request_irq(IRQ_GPIO(sharpsl_pm.machinfo->gpio_batfull), sharpsl_chrg_full_isr, SA_INTERRUPT, "CO", sharpsl_chrg_full_isr)) { + dev_err(sharpsl_pm.dev, "Could not get irq %d.\n", IRQ_GPIO(sharpsl_pm.machinfo->gpio_batfull)); + } + else set_irq_type(IRQ_GPIO(sharpsl_pm.machinfo->gpio_batfull),IRQT_RISING); + } + + device_create_file(dev, &dev_attr_battery_percentage); + device_create_file(dev, &dev_attr_battery_voltage); + + apm_get_power_status = sharpsl_apm_get_power_status; + + pm_set_ops(&sharpsl_pm_ops); + + mod_timer(&sharpsl_pm.ac_timer, jiffies + msecs_to_jiffies(250)); + + return 0; +} + +static int sharpsl_pm_remove(struct device *dev) +{ + pm_set_ops(NULL); + + device_remove_file(dev, &dev_attr_battery_percentage); + device_remove_file(dev, &dev_attr_battery_voltage); + + free_irq(IRQ_GPIO(sharpsl_pm.machinfo->gpio_acin), sharpsl_ac_isr); + free_irq(IRQ_GPIO(sharpsl_pm.machinfo->gpio_batlock), sharpsl_fatal_isr); + + if (sharpsl_pm.machinfo->gpio_fatal) + free_irq(IRQ_GPIO(sharpsl_pm.machinfo->gpio_fatal), sharpsl_fatal_isr); + + if (!machine_is_corgi()) + free_irq(IRQ_GPIO(sharpsl_pm.machinfo->gpio_batfull), sharpsl_chrg_full_isr); + + del_timer_sync(&sharpsl_pm.chrg_full_timer); + del_timer_sync(&sharpsl_pm.ac_timer); + + return 0; +} + +static struct device_driver sharpsl_pm_driver = { + .name = "sharpsl-pm", + .bus = &platform_bus_type, + .probe = sharpsl_pm_probe, + .remove = sharpsl_pm_remove, + .suspend = sharpsl_pm_suspend, + .resume = sharpsl_pm_resume, +}; + +static int __devinit sharpsl_pm_init(void) +{ + return driver_register(&sharpsl_pm_driver); +} + +static void sharpsl_pm_exit(void) +{ + driver_unregister(&sharpsl_pm_driver); +} + +late_initcall(sharpsl_pm_init); +module_exit(sharpsl_pm_exit); diff --git a/drivers/video/backlight/corgi_bl.c b/drivers/video/backlight/corgi_bl.c index 4867498f68e8..bd9a6996aee7 100644 --- a/drivers/video/backlight/corgi_bl.c +++ b/drivers/video/backlight/corgi_bl.c @@ -48,6 +48,12 @@ static void corgibl_send_intensity(int intensity) corgibl_mach_set_intensity(intensity); spin_unlock_irqrestore(&bl_lock, flags); + + corgi_kick_batt = symbol_get(sharpsl_battery_kick); + if (corgi_kick_batt) { + corgi_kick_batt(); + symbol_put(sharpsl_battery_kick); + } } static void corgibl_blank(int blank) diff --git a/include/asm-arm/arch-pxa/sharpsl.h b/include/asm-arm/arch-pxa/sharpsl.h index 311f2bb5386a..0b43495d24b4 100644 --- a/include/asm-arm/arch-pxa/sharpsl.h +++ b/include/asm-arm/arch-pxa/sharpsl.h @@ -21,12 +21,18 @@ struct corgits_machinfo { void (*wait_hsync)(void); }; + /* * SharpSL Backlight */ - struct corgibl_machinfo { int max_intensity; void (*set_bl_intensity)(int intensity); }; +extern void corgibl_limit_intensity(int limit); + +/* + * SharpSL Battery/PM Driver + */ +extern void sharpsl_battery_kick(void); From b216c01829d0b73a468204e2e763c0a818b77a46 Mon Sep 17 00:00:00 2001 From: Liam Girdwood Date: Thu, 10 Nov 2005 17:45:39 +0000 Subject: [PATCH 098/121] [ARM] 3098/1: pxa2xx disable ssp irq Patch from Liam Girdwood This patch allows users of the pxa SSP driver to register their own irq handlers instead of using the default SSP handler. It also cleans up the CKEN clock and irq detection as the values are now stored in a table. This patch replaces 2845/1 Changes:- o Added flags parameter to ssp_init() o Added SSP_NO_IRQ flag to disable registering of ssp irq handler (for drivers that want to register their own handler) o Cleaned up clock and irq detection, values are now stored in table. o Added build changes to allow other drivers (e.g audio) to select the ssp driver. o corgi_ssp.c changed to use new interface. Signed-off-by: Liam Girdwood Signed-off-by: Richard Purdie Signed-off-by: Russell King --- arch/arm/mach-pxa/Kconfig | 7 ++ arch/arm/mach-pxa/Makefile | 5 +- arch/arm/mach-pxa/corgi_ssp.c | 2 +- arch/arm/mach-pxa/ssp.c | 128 +++++++++++---------------------- include/asm-arm/arch-pxa/ssp.h | 8 ++- 5 files changed, 58 insertions(+), 92 deletions(-) diff --git a/arch/arm/mach-pxa/Kconfig b/arch/arm/mach-pxa/Kconfig index b380a438e68f..e201aa9765b9 100644 --- a/arch/arm/mach-pxa/Kconfig +++ b/arch/arm/mach-pxa/Kconfig @@ -60,6 +60,7 @@ config MACH_CORGI bool "Enable Sharp SL-C700 (Corgi) Support" depends PXA_SHARPSL_25x select PXA_SHARP_C7xx + select PXA_SSP config MACH_SHEPHERD bool "Enable Sharp SL-C750 (Shepherd) Support" @@ -102,12 +103,18 @@ config IWMMXT config PXA_SHARP_C7xx bool + select PXA_SSP help Enable support for all Sharp C7xx models config PXA_SHARP_Cxx00 bool + select PXA_SSP help Enable common support for Sharp Cxx00 models +config PXA_SSP + tristate + help + Enable support for PXA2xx SSP ports endif diff --git a/arch/arm/mach-pxa/Makefile b/arch/arm/mach-pxa/Makefile index 8bc72d07cea8..d210bd5032ce 100644 --- a/arch/arm/mach-pxa/Makefile +++ b/arch/arm/mach-pxa/Makefile @@ -11,8 +11,8 @@ obj-$(CONFIG_PXA27x) += pxa27x.o obj-$(CONFIG_ARCH_LUBBOCK) += lubbock.o obj-$(CONFIG_MACH_MAINSTONE) += mainstone.o obj-$(CONFIG_ARCH_PXA_IDP) += idp.o -obj-$(CONFIG_PXA_SHARP_C7xx) += corgi.o corgi_ssp.o corgi_lcd.o ssp.o -obj-$(CONFIG_PXA_SHARP_Cxx00) += spitz.o corgi_ssp.o corgi_lcd.o ssp.o +obj-$(CONFIG_PXA_SHARP_C7xx) += corgi.o corgi_ssp.o corgi_lcd.o +obj-$(CONFIG_PXA_SHARP_Cxx00) += spitz.o corgi_ssp.o corgi_lcd.o obj-$(CONFIG_MACH_POODLE) += poodle.o obj-$(CONFIG_MACH_TOSA) += tosa.o @@ -26,6 +26,7 @@ obj-$(CONFIG_LEDS) += $(led-y) # Misc features obj-$(CONFIG_PM) += pm.o sleep.o +obj-$(CONFIG_PXA_SSP) += ssp.o ifeq ($(CONFIG_PXA27x),y) obj-$(CONFIG_PM) += standby.o diff --git a/arch/arm/mach-pxa/corgi_ssp.c b/arch/arm/mach-pxa/corgi_ssp.c index 591e5f32dbec..bdf10cfa9440 100644 --- a/arch/arm/mach-pxa/corgi_ssp.c +++ b/arch/arm/mach-pxa/corgi_ssp.c @@ -203,7 +203,7 @@ static int __init corgi_ssp_probe(struct device *dev) GPDR(ssp_machinfo->cs_ads7846) |= GPIO_bit(ssp_machinfo->cs_ads7846); /* output */ GPSR(ssp_machinfo->cs_ads7846) = GPIO_bit(ssp_machinfo->cs_ads7846); /* High - Disable ADS7846*/ - ret = ssp_init(&corgi_ssp_dev,ssp_machinfo->port); + ret = ssp_init(&corgi_ssp_dev, ssp_machinfo->port, 0); if (ret) printk(KERN_ERR "Unable to register SSP handler!\n"); diff --git a/arch/arm/mach-pxa/ssp.c b/arch/arm/mach-pxa/ssp.c index 4d826c021315..a68b30eff4d2 100644 --- a/arch/arm/mach-pxa/ssp.c +++ b/arch/arm/mach-pxa/ssp.c @@ -19,6 +19,8 @@ * 22nd Aug 2003 Initial version. * 20th Dec 2004 Added ssp_config for changing port config without * closing the port. + * 4th Aug 2005 Added option to disable irq handler registration and + * cleaned up irq and clock detection. */ #include @@ -37,6 +39,26 @@ #define PXA_SSP_PORTS 3 +struct ssp_info_ { + int irq; + u32 clock; +}; + +/* + * SSP port clock and IRQ settings + */ +static const struct ssp_info_ ssp_info[PXA_SSP_PORTS] = { +#if defined (CONFIG_PXA27x) + {IRQ_SSP, CKEN23_SSP1}, + {IRQ_SSP2, CKEN3_SSP2}, + {IRQ_SSP3, CKEN4_SSP3}, +#else + {IRQ_SSP, CKEN3_SSP}, + {IRQ_NSSP, CKEN9_NSSP}, + {IRQ_ASSP, CKEN10_ASSP}, +#endif +}; + static DECLARE_MUTEX(sem); static int use_count[PXA_SSP_PORTS] = {0, 0, 0}; @@ -210,9 +232,9 @@ int ssp_config(struct ssp_dev *dev, u32 mode, u32 flags, u32 psp_flags, u32 spee * %-EBUSY if the resources are already in use * %0 on success */ -int ssp_init(struct ssp_dev *dev, u32 port) +int ssp_init(struct ssp_dev *dev, u32 port, u32 init_flags) { - int ret, irq; + int ret; if (port > PXA_SSP_PORTS || port == 0) return -ENODEV; @@ -229,61 +251,20 @@ int ssp_init(struct ssp_dev *dev, u32 port) up(&sem); return -EBUSY; } - - switch (port) { - case 1: - irq = IRQ_SSP; - break; -#if defined (CONFIG_PXA27x) - case 2: - irq = IRQ_SSP2; - break; - case 3: - irq = IRQ_SSP3; - break; -#else - case 2: - irq = IRQ_NSSP; - break; - case 3: - irq = IRQ_ASSP; - break; -#endif - default: - return -ENODEV; - } - dev->port = port; - ret = request_irq(irq, ssp_interrupt, 0, "SSP", dev); - if (ret) - goto out_region; + /* do we need to get irq */ + if (!(init_flags & SSP_NO_IRQ)) { + ret = request_irq(ssp_info[port-1].irq, ssp_interrupt, + 0, "SSP", dev); + if (ret) + goto out_region; + dev->irq = ssp_info[port-1].irq; + } else + dev->irq = 0; /* turn on SSP port clock */ - switch (dev->port) { -#if defined (CONFIG_PXA27x) - case 1: - pxa_set_cken(CKEN23_SSP1, 1); - break; - case 2: - pxa_set_cken(CKEN3_SSP2, 1); - break; - case 3: - pxa_set_cken(CKEN4_SSP3, 1); - break; -#else - case 1: - pxa_set_cken(CKEN3_SSP, 1); - break; - case 2: - pxa_set_cken(CKEN9_NSSP, 1); - break; - case 3: - pxa_set_cken(CKEN10_ASSP, 1); - break; -#endif - } - + pxa_set_cken(ssp_info[port-1].clock, 1); up(&sem); return 0; @@ -301,46 +282,17 @@ out_region: */ void ssp_exit(struct ssp_dev *dev) { - int irq; - down(&sem); SSCR0_P(dev->port) &= ~SSCR0_SSE; - /* find irq, save power and turn off SSP port clock */ - switch (dev->port) { -#if defined (CONFIG_PXA27x) - case 1: - irq = IRQ_SSP; - pxa_set_cken(CKEN23_SSP1, 0); - break; - case 2: - irq = IRQ_SSP2; - pxa_set_cken(CKEN3_SSP2, 0); - break; - case 3: - irq = IRQ_SSP3; - pxa_set_cken(CKEN4_SSP3, 0); - break; -#else - case 1: - irq = IRQ_SSP; - pxa_set_cken(CKEN3_SSP, 0); - break; - case 2: - irq = IRQ_NSSP; - pxa_set_cken(CKEN9_NSSP, 0); - break; - case 3: - irq = IRQ_ASSP; - pxa_set_cken(CKEN10_ASSP, 0); - break; -#endif - default: - printk(KERN_WARNING "SSP: tried to close invalid port\n"); - return; + if (dev->port > PXA_SSP_PORTS || dev->port == 0) { + printk(KERN_WARNING "SSP: tried to close invalid port\n"); + return; } - free_irq(irq, dev); + pxa_set_cken(ssp_info[dev->port-1].clock, 0); + if (dev->irq) + free_irq(dev->irq, dev); release_mem_region(__PREG(SSCR0_P(dev->port)), 0x2c); use_count[dev->port - 1]--; up(&sem); diff --git a/include/asm-arm/arch-pxa/ssp.h b/include/asm-arm/arch-pxa/ssp.h index 6ec67b018c09..949878c0d908 100644 --- a/include/asm-arm/arch-pxa/ssp.h +++ b/include/asm-arm/arch-pxa/ssp.h @@ -18,6 +18,11 @@ #ifndef SSP_H #define SSP_H +/* + * SSP initialisation flags + */ +#define SSP_NO_IRQ 0x1 /* don't register an irq handler in SSP driver */ + struct ssp_state { u32 cr0; u32 cr1; @@ -31,6 +36,7 @@ struct ssp_dev { u32 flags; u32 psp_flags; u32 speed; + int irq; }; int ssp_write_word(struct ssp_dev *dev, u32 data); @@ -40,7 +46,7 @@ void ssp_enable(struct ssp_dev *dev); void ssp_disable(struct ssp_dev *dev); void ssp_save_state(struct ssp_dev *dev, struct ssp_state *ssp); void ssp_restore_state(struct ssp_dev *dev, struct ssp_state *ssp); -int ssp_init(struct ssp_dev *dev, u32 port); +int ssp_init(struct ssp_dev *dev, u32 port, u32 init_flags); int ssp_config(struct ssp_dev *dev, u32 mode, u32 flags, u32 psp_flags, u32 speed); void ssp_exit(struct ssp_dev *dev); From 7ed0175a462c4c30f6df6fac1cccac058f997739 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Thu, 10 Nov 2005 17:22:18 +0300 Subject: [PATCH 099/121] [PATCH] Don't auto-reap traced children If a task is being traced we never auto-reap it even if it might look like its parent doesn't care. The tracer obviously _does_ care. Signed-off-by: Linus Torvalds --- kernel/signal.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/signal.c b/kernel/signal.c index 1bf3c39d6109..80789a59b4db 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -1499,7 +1499,7 @@ void do_notify_parent(struct task_struct *tsk, int sig) psig = tsk->parent->sighand; spin_lock_irqsave(&psig->siglock, flags); - if (sig == SIGCHLD && + if (!tsk->ptrace && sig == SIGCHLD && (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN || (psig->action[SIGCHLD-1].sa.sa_flags & SA_NOCLDWAIT))) { /* From 1732b0ef3b3a02e3df328086fb3018741c5476da Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Mon, 7 Nov 2005 10:33:11 -0800 Subject: [PATCH 100/121] [IPoIB] add path record information in debugfs Add ibX_path files to debugfs that contain information about the IPoIB path cache. IPoIB ARP only gives GIDs, which the IPoIB driver must resolve to real IB paths through the ib_sa module. For debugging, when the ARP table looks OK but traffic isn't flowing, it's useful to be able to see if the resolution from GID to path worked. Also clean up the formatting of the existing _mcg debugfs files. Signed-off-by: Roland Dreier --- drivers/infiniband/ulp/ipoib/ipoib.h | 15 +- drivers/infiniband/ulp/ipoib/ipoib_fs.c | 179 +++++++++++++++--- drivers/infiniband/ulp/ipoib/ipoib_main.c | 72 ++++++- .../infiniband/ulp/ipoib/ipoib_multicast.c | 9 +- drivers/infiniband/ulp/ipoib/ipoib_vlan.c | 7 +- 5 files changed, 232 insertions(+), 50 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index 0095acc0fbbe..9923a15a9996 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -179,6 +179,7 @@ struct ipoib_dev_priv { #ifdef CONFIG_INFINIBAND_IPOIB_DEBUG struct list_head fs_list; struct dentry *mcg_dentry; + struct dentry *path_dentry; #endif }; @@ -270,7 +271,6 @@ void ipoib_mcast_dev_flush(struct net_device *dev); #ifdef CONFIG_INFINIBAND_IPOIB_DEBUG struct ipoib_mcast_iter *ipoib_mcast_iter_init(struct net_device *dev); -void ipoib_mcast_iter_free(struct ipoib_mcast_iter *iter); int ipoib_mcast_iter_next(struct ipoib_mcast_iter *iter); void ipoib_mcast_iter_read(struct ipoib_mcast_iter *iter, union ib_gid *gid, @@ -278,6 +278,11 @@ void ipoib_mcast_iter_read(struct ipoib_mcast_iter *iter, unsigned int *queuelen, unsigned int *complete, unsigned int *send_only); + +struct ipoib_path_iter *ipoib_path_iter_init(struct net_device *dev); +int ipoib_path_iter_next(struct ipoib_path_iter *iter); +void ipoib_path_iter_read(struct ipoib_path_iter *iter, + struct ipoib_path *path); #endif int ipoib_mcast_attach(struct net_device *dev, u16 mlid, @@ -299,13 +304,13 @@ void ipoib_pkey_poll(void *dev); int ipoib_pkey_dev_delay_open(struct net_device *dev); #ifdef CONFIG_INFINIBAND_IPOIB_DEBUG -int ipoib_create_debug_file(struct net_device *dev); -void ipoib_delete_debug_file(struct net_device *dev); +void ipoib_create_debug_files(struct net_device *dev); +void ipoib_delete_debug_files(struct net_device *dev); int ipoib_register_debugfs(void); void ipoib_unregister_debugfs(void); #else -static inline int ipoib_create_debug_file(struct net_device *dev) { return 0; } -static inline void ipoib_delete_debug_file(struct net_device *dev) { } +static inline void ipoib_create_debug_files(struct net_device *dev) { } +static inline void ipoib_delete_debug_files(struct net_device *dev) { } static inline int ipoib_register_debugfs(void) { return 0; } static inline void ipoib_unregister_debugfs(void) { } #endif diff --git a/drivers/infiniband/ulp/ipoib/ipoib_fs.c b/drivers/infiniband/ulp/ipoib/ipoib_fs.c index 38b150f775e7..685258e34034 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_fs.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_fs.c @@ -43,6 +43,18 @@ struct file_operations; static struct dentry *ipoib_root; +static void format_gid(union ib_gid *gid, char *buf) +{ + int i, n; + + for (n = 0, i = 0; i < 8; ++i) { + n += sprintf(buf + n, "%x", + be16_to_cpu(((__be16 *) gid->raw)[i])); + if (i < 7) + buf[n++] = ':'; + } +} + static void *ipoib_mcg_seq_start(struct seq_file *file, loff_t *pos) { struct ipoib_mcast_iter *iter; @@ -54,7 +66,7 @@ static void *ipoib_mcg_seq_start(struct seq_file *file, loff_t *pos) while (n--) { if (ipoib_mcast_iter_next(iter)) { - ipoib_mcast_iter_free(iter); + kfree(iter); return NULL; } } @@ -70,7 +82,7 @@ static void *ipoib_mcg_seq_next(struct seq_file *file, void *iter_ptr, (*pos)++; if (ipoib_mcast_iter_next(iter)) { - ipoib_mcast_iter_free(iter); + kfree(iter); return NULL; } @@ -87,32 +99,32 @@ static int ipoib_mcg_seq_show(struct seq_file *file, void *iter_ptr) struct ipoib_mcast_iter *iter = iter_ptr; char gid_buf[sizeof "ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff"]; union ib_gid mgid; - int i, n; unsigned long created; unsigned int queuelen, complete, send_only; - if (iter) { - ipoib_mcast_iter_read(iter, &mgid, &created, &queuelen, - &complete, &send_only); + if (!iter) + return 0; - for (n = 0, i = 0; i < sizeof mgid / 2; ++i) { - n += sprintf(gid_buf + n, "%x", - be16_to_cpu(((__be16 *) mgid.raw)[i])); - if (i < sizeof mgid / 2 - 1) - gid_buf[n++] = ':'; - } - } + ipoib_mcast_iter_read(iter, &mgid, &created, &queuelen, + &complete, &send_only); - seq_printf(file, "GID: %*s", -(1 + (int) sizeof gid_buf), gid_buf); + format_gid(&mgid, gid_buf); seq_printf(file, - " created: %10ld queuelen: %4d complete: %d send_only: %d\n", - created, queuelen, complete, send_only); + "GID: %s\n" + " created: %10ld\n" + " queuelen: %9d\n" + " complete: %9s\n" + " send_only: %8s\n" + "\n", + gid_buf, created, queuelen, + complete ? "yes" : "no", + send_only ? "yes" : "no"); return 0; } -static struct seq_operations ipoib_seq_ops = { +static struct seq_operations ipoib_mcg_seq_ops = { .start = ipoib_mcg_seq_start, .next = ipoib_mcg_seq_next, .stop = ipoib_mcg_seq_stop, @@ -124,7 +136,7 @@ static int ipoib_mcg_open(struct inode *inode, struct file *file) struct seq_file *seq; int ret; - ret = seq_open(file, &ipoib_seq_ops); + ret = seq_open(file, &ipoib_mcg_seq_ops); if (ret) return ret; @@ -134,7 +146,7 @@ static int ipoib_mcg_open(struct inode *inode, struct file *file) return 0; } -static struct file_operations ipoib_fops = { +static struct file_operations ipoib_mcg_fops = { .owner = THIS_MODULE, .open = ipoib_mcg_open, .read = seq_read, @@ -142,25 +154,138 @@ static struct file_operations ipoib_fops = { .release = seq_release }; -int ipoib_create_debug_file(struct net_device *dev) +static void *ipoib_path_seq_start(struct seq_file *file, loff_t *pos) { - struct ipoib_dev_priv *priv = netdev_priv(dev); - char name[IFNAMSIZ + sizeof "_mcg"]; + struct ipoib_path_iter *iter; + loff_t n = *pos; - snprintf(name, sizeof name, "%s_mcg", dev->name); + iter = ipoib_path_iter_init(file->private); + if (!iter) + return NULL; - priv->mcg_dentry = debugfs_create_file(name, S_IFREG | S_IRUGO, - ipoib_root, dev, &ipoib_fops); + while (n--) { + if (ipoib_path_iter_next(iter)) { + kfree(iter); + return NULL; + } + } - return priv->mcg_dentry ? 0 : -ENOMEM; + return iter; } -void ipoib_delete_debug_file(struct net_device *dev) +static void *ipoib_path_seq_next(struct seq_file *file, void *iter_ptr, + loff_t *pos) +{ + struct ipoib_path_iter *iter = iter_ptr; + + (*pos)++; + + if (ipoib_path_iter_next(iter)) { + kfree(iter); + return NULL; + } + + return iter; +} + +static void ipoib_path_seq_stop(struct seq_file *file, void *iter_ptr) +{ + /* nothing for now */ +} + +static int ipoib_path_seq_show(struct seq_file *file, void *iter_ptr) +{ + struct ipoib_path_iter *iter = iter_ptr; + char gid_buf[sizeof "ffff:ffff:ffff:ffff:ffff:ffff:ffff:ffff"]; + struct ipoib_path path; + int rate; + + if (!iter) + return 0; + + ipoib_path_iter_read(iter, &path); + + format_gid(&path.pathrec.dgid, gid_buf); + + seq_printf(file, + "GID: %s\n" + " complete: %6s\n", + gid_buf, path.pathrec.dlid ? "yes" : "no"); + + if (path.pathrec.dlid) { + rate = ib_sa_rate_enum_to_int(path.pathrec.rate) * 25; + + seq_printf(file, + " DLID: 0x%04x\n" + " SL: %12d\n" + " rate: %*d%s Gb/sec\n", + be16_to_cpu(path.pathrec.dlid), + path.pathrec.sl, + 10 - ((rate % 10) ? 2 : 0), + rate / 10, rate % 10 ? ".5" : ""); + } + + seq_putc(file, '\n'); + + return 0; +} + +static struct seq_operations ipoib_path_seq_ops = { + .start = ipoib_path_seq_start, + .next = ipoib_path_seq_next, + .stop = ipoib_path_seq_stop, + .show = ipoib_path_seq_show, +}; + +static int ipoib_path_open(struct inode *inode, struct file *file) +{ + struct seq_file *seq; + int ret; + + ret = seq_open(file, &ipoib_path_seq_ops); + if (ret) + return ret; + + seq = file->private_data; + seq->private = inode->u.generic_ip; + + return 0; +} + +static struct file_operations ipoib_path_fops = { + .owner = THIS_MODULE, + .open = ipoib_path_open, + .read = seq_read, + .llseek = seq_lseek, + .release = seq_release +}; + +void ipoib_create_debug_files(struct net_device *dev) +{ + struct ipoib_dev_priv *priv = netdev_priv(dev); + char name[IFNAMSIZ + sizeof "_path"]; + + snprintf(name, sizeof name, "%s_mcg", dev->name); + priv->mcg_dentry = debugfs_create_file(name, S_IFREG | S_IRUGO, + ipoib_root, dev, &ipoib_mcg_fops); + if (!priv->mcg_dentry) + ipoib_warn(priv, "failed to create mcg debug file\n"); + + snprintf(name, sizeof name, "%s_path", dev->name); + priv->path_dentry = debugfs_create_file(name, S_IFREG | S_IRUGO, + ipoib_root, dev, &ipoib_path_fops); + if (!priv->path_dentry) + ipoib_warn(priv, "failed to create path debug file\n"); +} + +void ipoib_delete_debug_files(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); if (priv->mcg_dentry) debugfs_remove(priv->mcg_dentry); + if (priv->path_dentry) + debugfs_remove(priv->path_dentry); } int ipoib_register_debugfs(void) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index ce0296273e76..2fa30751f362 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -58,6 +58,11 @@ module_param_named(debug_level, ipoib_debug_level, int, 0644); MODULE_PARM_DESC(debug_level, "Enable debug tracing if > 0"); #endif +struct ipoib_path_iter { + struct net_device *dev; + struct ipoib_path path; +}; + static const u8 ipv4_bcast_addr[] = { 0x00, 0xff, 0xff, 0xff, 0xff, 0x12, 0x40, 0x1b, 0x00, 0x00, 0x00, 0x00, @@ -250,6 +255,64 @@ static void path_free(struct net_device *dev, struct ipoib_path *path) kfree(path); } +#ifdef CONFIG_INFINIBAND_IPOIB_DEBUG + +struct ipoib_path_iter *ipoib_path_iter_init(struct net_device *dev) +{ + struct ipoib_path_iter *iter; + + iter = kmalloc(sizeof *iter, GFP_KERNEL); + if (!iter) + return NULL; + + iter->dev = dev; + memset(iter->path.pathrec.dgid.raw, 0, 16); + + if (ipoib_path_iter_next(iter)) { + kfree(iter); + return NULL; + } + + return iter; +} + +int ipoib_path_iter_next(struct ipoib_path_iter *iter) +{ + struct ipoib_dev_priv *priv = netdev_priv(iter->dev); + struct rb_node *n; + struct ipoib_path *path; + int ret = 1; + + spin_lock_irq(&priv->lock); + + n = rb_first(&priv->path_tree); + + while (n) { + path = rb_entry(n, struct ipoib_path, rb_node); + + if (memcmp(iter->path.pathrec.dgid.raw, path->pathrec.dgid.raw, + sizeof (union ib_gid)) < 0) { + iter->path = *path; + ret = 0; + break; + } + + n = rb_next(n); + } + + spin_unlock_irq(&priv->lock); + + return ret; +} + +void ipoib_path_iter_read(struct ipoib_path_iter *iter, + struct ipoib_path *path) +{ + *path = iter->path; +} + +#endif /* CONFIG_INFINIBAND_IPOIB_DEBUG */ + void ipoib_flush_paths(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev); @@ -763,7 +826,7 @@ void ipoib_dev_cleanup(struct net_device *dev) { struct ipoib_dev_priv *priv = netdev_priv(dev), *cpriv, *tcpriv; - ipoib_delete_debug_file(dev); + ipoib_delete_debug_files(dev); /* Delete any child interfaces first */ list_for_each_entry_safe(cpriv, tcpriv, &priv->child_intfs, list) { @@ -972,8 +1035,7 @@ static struct net_device *ipoib_add_port(const char *format, goto register_failed; } - if (ipoib_create_debug_file(priv->dev)) - goto debug_failed; + ipoib_create_debug_files(priv->dev); if (ipoib_add_pkey_attr(priv->dev)) goto sysfs_failed; @@ -987,9 +1049,7 @@ static struct net_device *ipoib_add_port(const char *format, return priv->dev; sysfs_failed: - ipoib_delete_debug_file(priv->dev); - -debug_failed: + ipoib_delete_debug_files(priv->dev); unregister_netdev(priv->dev); register_failed: diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index 3ecf78a9493a..87096939e0b6 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -928,21 +928,16 @@ struct ipoib_mcast_iter *ipoib_mcast_iter_init(struct net_device *dev) return NULL; iter->dev = dev; - memset(iter->mgid.raw, 0, sizeof iter->mgid); + memset(iter->mgid.raw, 0, 16); if (ipoib_mcast_iter_next(iter)) { - ipoib_mcast_iter_free(iter); + kfree(iter); return NULL; } return iter; } -void ipoib_mcast_iter_free(struct ipoib_mcast_iter *iter) -{ - kfree(iter); -} - int ipoib_mcast_iter_next(struct ipoib_mcast_iter *iter) { struct ipoib_dev_priv *priv = netdev_priv(iter->dev); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c index 332d730e60c2..d280b341a37f 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_vlan.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_vlan.c @@ -113,8 +113,7 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey) priv->parent = ppriv->dev; - if (ipoib_create_debug_file(priv->dev)) - goto debug_failed; + ipoib_create_debug_files(priv->dev); if (ipoib_add_pkey_attr(priv->dev)) goto sysfs_failed; @@ -130,9 +129,7 @@ int ipoib_vlan_add(struct net_device *pdev, unsigned short pkey) return 0; sysfs_failed: - ipoib_delete_debug_file(priv->dev); - -debug_failed: + ipoib_delete_debug_files(priv->dev); unregister_netdev(priv->dev); register_failed: From 2f76e82947b977a1008cfd2868351a701c93c69c Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Mon, 7 Nov 2005 10:41:29 -0800 Subject: [PATCH 101/121] [IB] umad: avoid potential deadlock when unregistering MAD agents ib_unregister_mad_agent() completes all pending MAD sends and waits for the agent's send_handler routine to return. umad's send_handler() calls queue_packet(), which does down_read() on the port mutex to look up the agent ID. This means that the port mutex cannot be held for writing while calling ib_unregister_mad_agent(), or else it will deadlock. This patch fixes all the calls to ib_unregister_mad_agent() in the umad module to avoid this deadlock. Signed-off-by: Roland Dreier --- drivers/infiniband/core/user_mad.c | 29 +++++++++++++++++------------ 1 file changed, 17 insertions(+), 12 deletions(-) diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index 6aefeed42ab3..f5ed36c2a06c 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -505,8 +505,6 @@ found: goto out; } - file->agent[agent_id] = agent; - file->mr[agent_id] = ib_get_dma_mr(agent->qp->pd, IB_ACCESS_LOCAL_WRITE); if (IS_ERR(file->mr[agent_id])) { ret = -ENOMEM; @@ -519,14 +517,15 @@ found: goto err_mr; } + file->agent[agent_id] = agent; ret = 0; + goto out; err_mr: ib_dereg_mr(file->mr[agent_id]); err: - file->agent[agent_id] = NULL; ib_unregister_mad_agent(agent); out: @@ -536,27 +535,33 @@ out: static int ib_umad_unreg_agent(struct ib_umad_file *file, unsigned long arg) { + struct ib_mad_agent *agent = NULL; + struct ib_mr *mr = NULL; u32 id; int ret = 0; - down_write(&file->port->mutex); + if (get_user(id, (u32 __user *) arg)) + return -EFAULT; - if (get_user(id, (u32 __user *) arg)) { - ret = -EFAULT; - goto out; - } + down_write(&file->port->mutex); if (id < 0 || id >= IB_UMAD_MAX_AGENTS || !file->agent[id]) { ret = -EINVAL; goto out; } - ib_dereg_mr(file->mr[id]); - ib_unregister_mad_agent(file->agent[id]); + agent = file->agent[id]; + mr = file->mr[id]; file->agent[id] = NULL; out: up_write(&file->port->mutex); + + if (agent) { + ib_unregister_mad_agent(agent); + ib_dereg_mr(mr); + } + return ret; } @@ -623,16 +628,16 @@ static int ib_umad_close(struct inode *inode, struct file *filp) struct ib_umad_packet *packet, *tmp; int i; - down_write(&file->port->mutex); for (i = 0; i < IB_UMAD_MAX_AGENTS; ++i) if (file->agent[i]) { - ib_dereg_mr(file->mr[i]); ib_unregister_mad_agent(file->agent[i]); + ib_dereg_mr(file->mr[i]); } list_for_each_entry_safe(packet, tmp, &file->recv_list, list) kfree(packet); + down_write(&file->port->mutex); list_del(&file->port_list); up_write(&file->port->mutex); From 8c608a32e3cd7ff14498ad996ca32d1452245a97 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Mon, 7 Nov 2005 10:49:38 -0800 Subject: [PATCH 102/121] [IPoIB] no need to set skb->dev right before freeing skb For cut-and-paste reasons, the IPoIB driver was setting skb->dev right before calling dev_kfree_skb_any(). Get rid of this. Signed-off-by: Roland Dreier --- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 17 ++++------------- 1 file changed, 4 insertions(+), 13 deletions(-) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index 87096939e0b6..c33ed87f9dff 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -120,12 +120,8 @@ static void ipoib_mcast_free(struct ipoib_mcast *mcast) if (mcast->ah) ipoib_put_ah(mcast->ah); - while (!skb_queue_empty(&mcast->pkt_queue)) { - struct sk_buff *skb = skb_dequeue(&mcast->pkt_queue); - - skb->dev = dev; - dev_kfree_skb_any(skb); - } + while (!skb_queue_empty(&mcast->pkt_queue)) + dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue)); kfree(mcast); } @@ -317,13 +313,8 @@ ipoib_mcast_sendonly_join_complete(int status, IPOIB_GID_ARG(mcast->mcmember.mgid), status); /* Flush out any queued packets */ - while (!skb_queue_empty(&mcast->pkt_queue)) { - struct sk_buff *skb = skb_dequeue(&mcast->pkt_queue); - - skb->dev = dev; - - dev_kfree_skb_any(skb); - } + while (!skb_queue_empty(&mcast->pkt_queue)) + dev_kfree_skb_any(skb_dequeue(&mcast->pkt_queue)); /* Clear the busy flag so we try again */ clear_bit(IPOIB_MCAST_FLAG_BUSY, &mcast->flags); From 0b4ff2c0e624089ad87dc1604e239b7c3201c53f Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Mon, 7 Nov 2005 22:01:02 -0800 Subject: [PATCH 103/121] [IB] mthca: fix typo in catastrophic error polling Fix a typo in the rearming of the catastrophic error polling timer: we should rearm the timer as long as the stop flag is _not_ set. Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_catas.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mthca/mthca_catas.c b/drivers/infiniband/hw/mthca/mthca_catas.c index 7ac52af43b99..3447cd70ce32 100644 --- a/drivers/infiniband/hw/mthca/mthca_catas.c +++ b/drivers/infiniband/hw/mthca/mthca_catas.c @@ -94,7 +94,7 @@ static void poll_catas(unsigned long dev_ptr) } spin_lock_irqsave(&catas_lock, flags); - if (dev->catas_err.stop) + if (!dev->catas_err.stop) mod_timer(&dev->catas_err.timer, jiffies + MTHCA_CATAS_POLL_INTERVAL); spin_unlock_irqrestore(&catas_lock, flags); From 40de2e548c225e3ef859e3c60de9785e37e1b5b1 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Tue, 8 Nov 2005 11:10:25 -0800 Subject: [PATCH 104/121] [IB] Have cq_resize() method take an int, not int* Change the struct ib_device.resize_cq() method to take a plain integer that holds the new CQ size, rather than a pointer to an integer that it uses to return the new size. This makes the interface match the exported ib_resize_cq() signature, and allows the low-level driver to update the CQ size with proper locking if necessary. No in-tree drivers are exporting this method yet. Signed-off-by: Roland Dreier --- drivers/infiniband/core/verbs.c | 12 ++---------- include/rdma/ib_verbs.h | 2 +- 2 files changed, 3 insertions(+), 11 deletions(-) diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 72d3ef786db5..4f51d797f84a 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -324,16 +324,8 @@ EXPORT_SYMBOL(ib_destroy_cq); int ib_resize_cq(struct ib_cq *cq, int cqe) { - int ret; - - if (!cq->device->resize_cq) - return -ENOSYS; - - ret = cq->device->resize_cq(cq, &cqe); - if (!ret) - cq->cqe = cqe; - - return ret; + return cq->device->resize_cq ? + cq->device->resize_cq(cq, cqe) : -ENOSYS; } EXPORT_SYMBOL(ib_resize_cq); diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index f72d46d54e0a..a7f4c355a91f 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -881,7 +881,7 @@ struct ib_device { struct ib_ucontext *context, struct ib_udata *udata); int (*destroy_cq)(struct ib_cq *cq); - int (*resize_cq)(struct ib_cq *cq, int *cqe); + int (*resize_cq)(struct ib_cq *cq, int cqe); int (*poll_cq)(struct ib_cq *cq, int num_entries, struct ib_wc *wc); int (*peek_cq)(struct ib_cq *cq, int wc_cnt); From ec914c52d6208d8752dfd85b48a9aff304911434 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Wed, 9 Nov 2005 09:58:10 -0800 Subject: [PATCH 105/121] [IB] umad: get rid of unused mr array Now that ib_umad uses the new MAD sending interface, it no longer needs its own L_Key. So just delete the array of MRs that it keeps. Signed-off-by: Roland Dreier --- drivers/infiniband/core/user_mad.c | 29 ++++------------------------- 1 file changed, 4 insertions(+), 25 deletions(-) diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index f5ed36c2a06c..d61f544f19e0 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -116,7 +116,6 @@ struct ib_umad_file { spinlock_t recv_lock; wait_queue_head_t recv_wait; struct ib_mad_agent *agent[IB_UMAD_MAX_AGENTS]; - struct ib_mr *mr[IB_UMAD_MAX_AGENTS]; }; struct ib_umad_packet { @@ -505,29 +504,16 @@ found: goto out; } - file->mr[agent_id] = ib_get_dma_mr(agent->qp->pd, IB_ACCESS_LOCAL_WRITE); - if (IS_ERR(file->mr[agent_id])) { - ret = -ENOMEM; - goto err; - } - if (put_user(agent_id, (u32 __user *) (arg + offsetof(struct ib_user_mad_reg_req, id)))) { ret = -EFAULT; - goto err_mr; + ib_unregister_mad_agent(agent); + goto out; } file->agent[agent_id] = agent; ret = 0; - goto out; - -err_mr: - ib_dereg_mr(file->mr[agent_id]); - -err: - ib_unregister_mad_agent(agent); - out: up_write(&file->port->mutex); return ret; @@ -536,7 +522,6 @@ out: static int ib_umad_unreg_agent(struct ib_umad_file *file, unsigned long arg) { struct ib_mad_agent *agent = NULL; - struct ib_mr *mr = NULL; u32 id; int ret = 0; @@ -551,16 +536,13 @@ static int ib_umad_unreg_agent(struct ib_umad_file *file, unsigned long arg) } agent = file->agent[id]; - mr = file->mr[id]; file->agent[id] = NULL; out: up_write(&file->port->mutex); - if (agent) { + if (agent) ib_unregister_mad_agent(agent); - ib_dereg_mr(mr); - } return ret; } @@ -629,10 +611,8 @@ static int ib_umad_close(struct inode *inode, struct file *filp) int i; for (i = 0; i < IB_UMAD_MAX_AGENTS; ++i) - if (file->agent[i]) { + if (file->agent[i]) ib_unregister_mad_agent(file->agent[i]); - ib_dereg_mr(file->mr[i]); - } list_for_each_entry_safe(packet, tmp, &file->recv_list, list) kfree(packet); @@ -872,7 +852,6 @@ static void ib_umad_kill_port(struct ib_umad_port *port) for (id = 0; id < IB_UMAD_MAX_AGENTS; ++id) { if (!file->agent[id]) continue; - ib_dereg_mr(file->mr[id]); ib_unregister_mad_agent(file->agent[id]); file->agent[id] = NULL; } From 77369ed31daac51f4827c50d30f233c45480235a Mon Sep 17 00:00:00 2001 From: Jack Morgenstein Date: Wed, 9 Nov 2005 11:26:07 -0800 Subject: [PATCH 106/121] [IB] uverbs: have kernel return QP capabilities Move the computation of QP capabilities (max scatter/gather entries, max inline data, etc) into the kernel, and have the uverbs module return the values as part of the create QP response. This keeps precise knowledge of device limits in the low-level kernel driver. This requires an ABI bump, so while we're making changes, get rid of the max_sge parameter for the modify SRQ command -- it's not used and shouldn't be there. Signed-off-by: Jack Morgenstein Signed-off-by: Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/core/uverbs_cmd.c | 12 ++- drivers/infiniband/hw/mthca/mthca_cmd.c | 2 + drivers/infiniband/hw/mthca/mthca_dev.h | 1 + drivers/infiniband/hw/mthca/mthca_main.c | 1 + drivers/infiniband/hw/mthca/mthca_provider.c | 2 +- drivers/infiniband/hw/mthca/mthca_provider.h | 1 + drivers/infiniband/hw/mthca/mthca_qp.c | 86 ++++++++++++++++++-- include/rdma/ib_user_verbs.h | 9 +- 8 files changed, 98 insertions(+), 16 deletions(-) diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 63a74151c60b..ed45da892b1c 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -708,7 +708,7 @@ ssize_t ib_uverbs_poll_cq(struct ib_uverbs_file *file, resp->wc[i].opcode = wc[i].opcode; resp->wc[i].vendor_err = wc[i].vendor_err; resp->wc[i].byte_len = wc[i].byte_len; - resp->wc[i].imm_data = wc[i].imm_data; + resp->wc[i].imm_data = (__u32 __force) wc[i].imm_data; resp->wc[i].qp_num = wc[i].qp_num; resp->wc[i].src_qp = wc[i].src_qp; resp->wc[i].wc_flags = wc[i].wc_flags; @@ -908,7 +908,12 @@ retry: if (ret) goto err_destroy; - resp.qp_handle = uobj->uobject.id; + resp.qp_handle = uobj->uobject.id; + resp.max_recv_sge = attr.cap.max_recv_sge; + resp.max_send_sge = attr.cap.max_send_sge; + resp.max_recv_wr = attr.cap.max_recv_wr; + resp.max_send_wr = attr.cap.max_send_wr; + resp.max_inline_data = attr.cap.max_inline_data; if (copy_to_user((void __user *) (unsigned long) cmd.response, &resp, sizeof resp)) { @@ -1135,7 +1140,7 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, next->num_sge = user_wr->num_sge; next->opcode = user_wr->opcode; next->send_flags = user_wr->send_flags; - next->imm_data = user_wr->imm_data; + next->imm_data = (__be32 __force) user_wr->imm_data; if (qp->qp_type == IB_QPT_UD) { next->wr.ud.ah = idr_find(&ib_uverbs_ah_idr, @@ -1701,7 +1706,6 @@ ssize_t ib_uverbs_modify_srq(struct ib_uverbs_file *file, } attr.max_wr = cmd.max_wr; - attr.max_sge = cmd.max_sge; attr.srq_limit = cmd.srq_limit; ret = ib_modify_srq(srq, &attr, cmd.attr_mask); diff --git a/drivers/infiniband/hw/mthca/mthca_cmd.c b/drivers/infiniband/hw/mthca/mthca_cmd.c index 49f211d55df7..9ed34587fc5c 100644 --- a/drivers/infiniband/hw/mthca/mthca_cmd.c +++ b/drivers/infiniband/hw/mthca/mthca_cmd.c @@ -1060,6 +1060,8 @@ int mthca_QUERY_DEV_LIM(struct mthca_dev *dev, dev_lim->hca.arbel.resize_srq = field & 1; MTHCA_GET(field, outbox, QUERY_DEV_LIM_MAX_SG_RQ_OFFSET); dev_lim->max_sg = min_t(int, field, dev_lim->max_sg); + MTHCA_GET(size, outbox, QUERY_DEV_LIM_MAX_DESC_SZ_RQ_OFFSET); + dev_lim->max_desc_sz = min_t(int, size, dev_lim->max_desc_sz); MTHCA_GET(size, outbox, QUERY_DEV_LIM_MPT_ENTRY_SZ_OFFSET); dev_lim->mpt_entry_sz = size; MTHCA_GET(field, outbox, QUERY_DEV_LIM_PBL_SZ_OFFSET); diff --git a/drivers/infiniband/hw/mthca/mthca_dev.h b/drivers/infiniband/hw/mthca/mthca_dev.h index 808037f25c78..497ff794ef6a 100644 --- a/drivers/infiniband/hw/mthca/mthca_dev.h +++ b/drivers/infiniband/hw/mthca/mthca_dev.h @@ -131,6 +131,7 @@ struct mthca_limits { int max_sg; int num_qps; int max_wqes; + int max_desc_sz; int max_qp_init_rdma; int reserved_qps; int num_srqs; diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c index 16594d1342df..147f248a8073 100644 --- a/drivers/infiniband/hw/mthca/mthca_main.c +++ b/drivers/infiniband/hw/mthca/mthca_main.c @@ -168,6 +168,7 @@ static int __devinit mthca_dev_lim(struct mthca_dev *mdev, struct mthca_dev_lim mdev->limits.max_srq_wqes = dev_lim->max_srq_sz; mdev->limits.reserved_srqs = dev_lim->reserved_srqs; mdev->limits.reserved_eecs = dev_lim->reserved_eecs; + mdev->limits.max_desc_sz = dev_lim->max_desc_sz; /* * Subtract 1 from the limit because we need to allocate a * spare CQE so the HCA HW can tell the difference between an diff --git a/drivers/infiniband/hw/mthca/mthca_provider.c b/drivers/infiniband/hw/mthca/mthca_provider.c index e78259b2664b..4cc7e2846df1 100644 --- a/drivers/infiniband/hw/mthca/mthca_provider.c +++ b/drivers/infiniband/hw/mthca/mthca_provider.c @@ -616,11 +616,11 @@ static struct ib_qp *mthca_create_qp(struct ib_pd *pd, return ERR_PTR(err); } - init_attr->cap.max_inline_data = 0; init_attr->cap.max_send_wr = qp->sq.max; init_attr->cap.max_recv_wr = qp->rq.max; init_attr->cap.max_send_sge = qp->sq.max_gs; init_attr->cap.max_recv_sge = qp->rq.max_gs; + init_attr->cap.max_inline_data = qp->max_inline_data; return &qp->ibqp; } diff --git a/drivers/infiniband/hw/mthca/mthca_provider.h b/drivers/infiniband/hw/mthca/mthca_provider.h index bcd4b01a339c..1e73947b4702 100644 --- a/drivers/infiniband/hw/mthca/mthca_provider.h +++ b/drivers/infiniband/hw/mthca/mthca_provider.h @@ -251,6 +251,7 @@ struct mthca_qp { struct mthca_wq sq; enum ib_sig_type sq_policy; int send_wqe_offset; + int max_inline_data; u64 *wrid; union mthca_buf queue; diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index 8852ea477c21..7f39af44b274 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -885,6 +885,48 @@ int mthca_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask) return err; } +static void mthca_adjust_qp_caps(struct mthca_dev *dev, + struct mthca_pd *pd, + struct mthca_qp *qp) +{ + int max_data_size; + + /* + * Calculate the maximum size of WQE s/g segments, excluding + * the next segment and other non-data segments. + */ + max_data_size = min(dev->limits.max_desc_sz, 1 << qp->sq.wqe_shift) - + sizeof (struct mthca_next_seg); + + switch (qp->transport) { + case MLX: + max_data_size -= 2 * sizeof (struct mthca_data_seg); + break; + + case UD: + if (mthca_is_memfree(dev)) + max_data_size -= sizeof (struct mthca_arbel_ud_seg); + else + max_data_size -= sizeof (struct mthca_tavor_ud_seg); + break; + + default: + max_data_size -= sizeof (struct mthca_raddr_seg); + break; + } + + /* We don't support inline data for kernel QPs (yet). */ + if (!pd->ibpd.uobject) + qp->max_inline_data = 0; + else + qp->max_inline_data = max_data_size - MTHCA_INLINE_HEADER_SIZE; + + qp->sq.max_gs = max_data_size / sizeof (struct mthca_data_seg); + qp->rq.max_gs = (min(dev->limits.max_desc_sz, 1 << qp->rq.wqe_shift) - + sizeof (struct mthca_next_seg)) / + sizeof (struct mthca_data_seg); +} + /* * Allocate and register buffer for WQEs. qp->rq.max, sq.max, * rq.max_gs and sq.max_gs must all be assigned. @@ -902,27 +944,53 @@ static int mthca_alloc_wqe_buf(struct mthca_dev *dev, size = sizeof (struct mthca_next_seg) + qp->rq.max_gs * sizeof (struct mthca_data_seg); + if (size > dev->limits.max_desc_sz) + return -EINVAL; + for (qp->rq.wqe_shift = 6; 1 << qp->rq.wqe_shift < size; qp->rq.wqe_shift++) ; /* nothing */ - size = sizeof (struct mthca_next_seg) + - qp->sq.max_gs * sizeof (struct mthca_data_seg); + size = qp->sq.max_gs * sizeof (struct mthca_data_seg); switch (qp->transport) { case MLX: size += 2 * sizeof (struct mthca_data_seg); break; + case UD: - if (mthca_is_memfree(dev)) - size += sizeof (struct mthca_arbel_ud_seg); - else - size += sizeof (struct mthca_tavor_ud_seg); + size += mthca_is_memfree(dev) ? + sizeof (struct mthca_arbel_ud_seg) : + sizeof (struct mthca_tavor_ud_seg); break; + + case UC: + size += sizeof (struct mthca_raddr_seg); + break; + + case RC: + size += sizeof (struct mthca_raddr_seg); + /* + * An atomic op will require an atomic segment, a + * remote address segment and one scatter entry. + */ + size = max_t(int, size, + sizeof (struct mthca_atomic_seg) + + sizeof (struct mthca_raddr_seg) + + sizeof (struct mthca_data_seg)); + break; + default: - /* bind seg is as big as atomic + raddr segs */ - size += sizeof (struct mthca_bind_seg); + break; } + /* Make sure that we have enough space for a bind request */ + size = max_t(int, size, sizeof (struct mthca_bind_seg)); + + size += sizeof (struct mthca_next_seg); + + if (size > dev->limits.max_desc_sz) + return -EINVAL; + for (qp->sq.wqe_shift = 6; 1 << qp->sq.wqe_shift < size; qp->sq.wqe_shift++) ; /* nothing */ @@ -1066,6 +1134,8 @@ static int mthca_alloc_qp_common(struct mthca_dev *dev, return ret; } + mthca_adjust_qp_caps(dev, pd, qp); + /* * If this is a userspace QP, we're done now. The doorbells * will be allocated and buffers will be initialized in diff --git a/include/rdma/ib_user_verbs.h b/include/rdma/ib_user_verbs.h index 072f3a2edace..5ff1490c08db 100644 --- a/include/rdma/ib_user_verbs.h +++ b/include/rdma/ib_user_verbs.h @@ -43,7 +43,7 @@ * Increment this value if any changes that break userspace ABI * compatibility are made. */ -#define IB_USER_VERBS_ABI_VERSION 3 +#define IB_USER_VERBS_ABI_VERSION 4 enum { IB_USER_VERBS_CMD_GET_CONTEXT, @@ -333,6 +333,11 @@ struct ib_uverbs_create_qp { struct ib_uverbs_create_qp_resp { __u32 qp_handle; __u32 qpn; + __u32 max_send_wr; + __u32 max_recv_wr; + __u32 max_send_sge; + __u32 max_recv_sge; + __u32 max_inline_data; }; /* @@ -552,9 +557,7 @@ struct ib_uverbs_modify_srq { __u32 srq_handle; __u32 attr_mask; __u32 max_wr; - __u32 max_sge; __u32 srq_limit; - __u32 reserved; __u64 driver_data[0]; }; From 62abb8416f1923f4cef50ce9ce841b919275e3fb Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Wed, 9 Nov 2005 11:30:14 -0800 Subject: [PATCH 107/121] [IB] mthca: fix posting of atomic operations The size of work requests for atomic operations was computed incorrectly in mthca: all sizeofs need to be divided by 16. Signed-off-by: Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_qp.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index 7f39af44b274..190c1dc0ee27 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -1556,8 +1556,8 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, } wqe += sizeof (struct mthca_atomic_seg); - size += sizeof (struct mthca_raddr_seg) / 16 + - sizeof (struct mthca_atomic_seg); + size += (sizeof (struct mthca_raddr_seg) + + sizeof (struct mthca_atomic_seg)) / 16; break; case IB_WR_RDMA_WRITE: @@ -1876,8 +1876,8 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, } wqe += sizeof (struct mthca_atomic_seg); - size += sizeof (struct mthca_raddr_seg) / 16 + - sizeof (struct mthca_atomic_seg); + size += (sizeof (struct mthca_raddr_seg) + + sizeof (struct mthca_atomic_seg)) / 16; break; case IB_WR_RDMA_READ: From 64044bcf75063cb5a6d42712886a712449df2ce3 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Wed, 9 Nov 2005 12:23:17 -0800 Subject: [PATCH 108/121] [IB] mthca: fix wraparound handling in mthca_cq_clean() Handle case where prod_index has wrapped around and become less than cq->cons_index by checking that their difference as a signed int is positive rather than comparing directly. Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_cq.c | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-) diff --git a/drivers/infiniband/hw/mthca/mthca_cq.c b/drivers/infiniband/hw/mthca/mthca_cq.c index f98e23555826..4a8adcef2079 100644 --- a/drivers/infiniband/hw/mthca/mthca_cq.c +++ b/drivers/infiniband/hw/mthca/mthca_cq.c @@ -258,7 +258,7 @@ void mthca_cq_clean(struct mthca_dev *dev, u32 cqn, u32 qpn, { struct mthca_cq *cq; struct mthca_cqe *cqe; - int prod_index; + u32 prod_index; int nfreed = 0; spin_lock_irq(&dev->cq_table.lock); @@ -293,19 +293,15 @@ void mthca_cq_clean(struct mthca_dev *dev, u32 cqn, u32 qpn, * Now sweep backwards through the CQ, removing CQ entries * that match our QP by copying older entries on top of them. */ - while (prod_index > cq->cons_index) { - cqe = get_cqe(cq, (prod_index - 1) & cq->ibcq.cqe); + while ((int) --prod_index - (int) cq->cons_index >= 0) { + cqe = get_cqe(cq, prod_index & cq->ibcq.cqe); if (cqe->my_qpn == cpu_to_be32(qpn)) { if (srq) mthca_free_srq_wqe(srq, be32_to_cpu(cqe->wqe)); ++nfreed; - } - else if (nfreed) - memcpy(get_cqe(cq, (prod_index - 1 + nfreed) & - cq->ibcq.cqe), - cqe, - MTHCA_CQ_ENTRY_SIZE); - --prod_index; + } else if (nfreed) + memcpy(get_cqe(cq, (prod_index + nfreed) & cq->ibcq.cqe), + cqe, MTHCA_CQ_ENTRY_SIZE); } if (nfreed) { From ae57e24a4006fd46b73d842ee99db9580ef74a02 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Wed, 9 Nov 2005 14:59:57 -0800 Subject: [PATCH 109/121] [IB] mthca: fix posting long lists of receive work requests In Tavor mode, when posting a long list of receive work requests, a doorbell must be rung every 256 requests. Add code to do this when required. Signed-off-by: Michael S. Tsirkin Signed-off-by: Roland Dreier --- drivers/infiniband/hw/mthca/mthca_qp.c | 19 +++++++++++++++++-- drivers/infiniband/hw/mthca/mthca_srq.c | 22 ++++++++++++++++++++-- drivers/infiniband/hw/mthca/mthca_wqe.h | 3 ++- 3 files changed, 39 insertions(+), 5 deletions(-) diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index 190c1dc0ee27..760c418d5bc9 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -1707,6 +1707,7 @@ int mthca_tavor_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr, { struct mthca_dev *dev = to_mdev(ibqp->device); struct mthca_qp *qp = to_mqp(ibqp); + __be32 doorbell[2]; unsigned long flags; int err = 0; int nreq; @@ -1724,6 +1725,22 @@ int mthca_tavor_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr, ind = qp->rq.next_ind; for (nreq = 0; wr; ++nreq, wr = wr->next) { + if (unlikely(nreq == MTHCA_TAVOR_MAX_WQES_PER_RECV_DB)) { + nreq = 0; + + doorbell[0] = cpu_to_be32((qp->rq.next_ind << qp->rq.wqe_shift) | size0); + doorbell[1] = cpu_to_be32(qp->qpn << 8); + + wmb(); + + mthca_write64(doorbell, + dev->kar + MTHCA_RECEIVE_DOORBELL, + MTHCA_GET_DOORBELL_LOCK(&dev->doorbell_lock)); + + qp->rq.head += MTHCA_TAVOR_MAX_WQES_PER_RECV_DB; + size0 = 0; + } + if (mthca_wq_overflow(&qp->rq, nreq, qp->ibqp.recv_cq)) { mthca_err(dev, "RQ %06x full (%u head, %u tail," " %d max, %d nreq)\n", qp->qpn, @@ -1781,8 +1798,6 @@ int mthca_tavor_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr, out: if (likely(nreq)) { - __be32 doorbell[2]; - doorbell[0] = cpu_to_be32((qp->rq.next_ind << qp->rq.wqe_shift) | size0); doorbell[1] = cpu_to_be32((qp->qpn << 8) | nreq); diff --git a/drivers/infiniband/hw/mthca/mthca_srq.c b/drivers/infiniband/hw/mthca/mthca_srq.c index 292f55be8cbd..c3c0331a1f6e 100644 --- a/drivers/infiniband/hw/mthca/mthca_srq.c +++ b/drivers/infiniband/hw/mthca/mthca_srq.c @@ -414,6 +414,7 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, { struct mthca_dev *dev = to_mdev(ibsrq->device); struct mthca_srq *srq = to_msrq(ibsrq); + __be32 doorbell[2]; unsigned long flags; int err = 0; int first_ind; @@ -429,6 +430,25 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, first_ind = srq->first_free; for (nreq = 0; wr; ++nreq, wr = wr->next) { + if (unlikely(nreq == MTHCA_TAVOR_MAX_WQES_PER_RECV_DB)) { + nreq = 0; + + doorbell[0] = cpu_to_be32(first_ind << srq->wqe_shift); + doorbell[1] = cpu_to_be32(srq->srqn << 8); + + /* + * Make sure that descriptors are written + * before doorbell is rung. + */ + wmb(); + + mthca_write64(doorbell, + dev->kar + MTHCA_RECEIVE_DOORBELL, + MTHCA_GET_DOORBELL_LOCK(&dev->doorbell_lock)); + + first_ind = srq->first_free; + } + ind = srq->first_free; if (ind < 0) { @@ -491,8 +511,6 @@ int mthca_tavor_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, } if (likely(nreq)) { - __be32 doorbell[2]; - doorbell[0] = cpu_to_be32(first_ind << srq->wqe_shift); doorbell[1] = cpu_to_be32((srq->srqn << 8) | nreq); diff --git a/drivers/infiniband/hw/mthca/mthca_wqe.h b/drivers/infiniband/hw/mthca/mthca_wqe.h index 1f4c0ff28f79..73f1c0b9021e 100644 --- a/drivers/infiniband/hw/mthca/mthca_wqe.h +++ b/drivers/infiniband/hw/mthca/mthca_wqe.h @@ -49,7 +49,8 @@ enum { }; enum { - MTHCA_INVAL_LKEY = 0x100 + MTHCA_INVAL_LKEY = 0x100, + MTHCA_TAVOR_MAX_WQES_PER_RECV_DB = 256 }; struct mthca_next_seg { From 94382f3562e350ed7c8f7dcd6fc968bdece31328 Mon Sep 17 00:00:00 2001 From: Roland Dreier Date: Thu, 10 Nov 2005 10:18:23 -0800 Subject: [PATCH 110/121] [IB] umad: further ib_unregister_mad_agent() deadlock fixes The previous umad deadlock fix left ib_umad_kill_port() still vulnerable to deadlocking. This patch fixes that by downgrading our lock to a read lock when we might end up trying to reacquire the lock for reading. Signed-off-by: Roland Dreier --- drivers/infiniband/core/user_mad.c | 87 +++++++++++++++++++++--------- 1 file changed, 63 insertions(+), 24 deletions(-) diff --git a/drivers/infiniband/core/user_mad.c b/drivers/infiniband/core/user_mad.c index d61f544f19e0..5ea741f47fc8 100644 --- a/drivers/infiniband/core/user_mad.c +++ b/drivers/infiniband/core/user_mad.c @@ -31,7 +31,7 @@ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE * SOFTWARE. * - * $Id: user_mad.c 2814 2005-07-06 19:14:09Z halr $ + * $Id: user_mad.c 4010 2005-11-09 23:11:56Z roland $ */ #include @@ -110,12 +110,13 @@ struct ib_umad_device { }; struct ib_umad_file { - struct ib_umad_port *port; - struct list_head recv_list; - struct list_head port_list; - spinlock_t recv_lock; - wait_queue_head_t recv_wait; - struct ib_mad_agent *agent[IB_UMAD_MAX_AGENTS]; + struct ib_umad_port *port; + struct list_head recv_list; + struct list_head port_list; + spinlock_t recv_lock; + wait_queue_head_t recv_wait; + struct ib_mad_agent *agent[IB_UMAD_MAX_AGENTS]; + int agents_dead; }; struct ib_umad_packet { @@ -144,6 +145,12 @@ static void ib_umad_release_dev(struct kref *ref) kfree(dev); } +/* caller must hold port->mutex at least for reading */ +static struct ib_mad_agent *__get_agent(struct ib_umad_file *file, int id) +{ + return file->agents_dead ? NULL : file->agent[id]; +} + static int queue_packet(struct ib_umad_file *file, struct ib_mad_agent *agent, struct ib_umad_packet *packet) @@ -151,10 +158,11 @@ static int queue_packet(struct ib_umad_file *file, int ret = 1; down_read(&file->port->mutex); + for (packet->mad.hdr.id = 0; packet->mad.hdr.id < IB_UMAD_MAX_AGENTS; packet->mad.hdr.id++) - if (agent == file->agent[packet->mad.hdr.id]) { + if (agent == __get_agent(file, packet->mad.hdr.id)) { spin_lock_irq(&file->recv_lock); list_add_tail(&packet->list, &file->recv_list); spin_unlock_irq(&file->recv_lock); @@ -326,7 +334,7 @@ static ssize_t ib_umad_write(struct file *filp, const char __user *buf, down_read(&file->port->mutex); - agent = file->agent[packet->mad.hdr.id]; + agent = __get_agent(file, packet->mad.hdr.id); if (!agent) { ret = -EINVAL; goto err_up; @@ -480,7 +488,7 @@ static int ib_umad_reg_agent(struct ib_umad_file *file, unsigned long arg) } for (agent_id = 0; agent_id < IB_UMAD_MAX_AGENTS; ++agent_id) - if (!file->agent[agent_id]) + if (!__get_agent(file, agent_id)) goto found; ret = -ENOMEM; @@ -530,7 +538,7 @@ static int ib_umad_unreg_agent(struct ib_umad_file *file, unsigned long arg) down_write(&file->port->mutex); - if (id < 0 || id >= IB_UMAD_MAX_AGENTS || !file->agent[id]) { + if (id < 0 || id >= IB_UMAD_MAX_AGENTS || !__get_agent(file, id)) { ret = -EINVAL; goto out; } @@ -608,21 +616,29 @@ static int ib_umad_close(struct inode *inode, struct file *filp) struct ib_umad_file *file = filp->private_data; struct ib_umad_device *dev = file->port->umad_dev; struct ib_umad_packet *packet, *tmp; + int already_dead; int i; - for (i = 0; i < IB_UMAD_MAX_AGENTS; ++i) - if (file->agent[i]) - ib_unregister_mad_agent(file->agent[i]); + down_write(&file->port->mutex); + + already_dead = file->agents_dead; + file->agents_dead = 1; list_for_each_entry_safe(packet, tmp, &file->recv_list, list) kfree(packet); - down_write(&file->port->mutex); list_del(&file->port_list); - up_write(&file->port->mutex); + + downgrade_write(&file->port->mutex); + + if (!already_dead) + for (i = 0; i < IB_UMAD_MAX_AGENTS; ++i) + if (file->agent[i]) + ib_unregister_mad_agent(file->agent[i]); + + up_read(&file->port->mutex); kfree(file); - kref_put(&dev->ref, ib_umad_release_dev); return 0; @@ -848,13 +864,36 @@ static void ib_umad_kill_port(struct ib_umad_port *port) port->ib_dev = NULL; - list_for_each_entry(file, &port->file_list, port_list) - for (id = 0; id < IB_UMAD_MAX_AGENTS; ++id) { - if (!file->agent[id]) - continue; - ib_unregister_mad_agent(file->agent[id]); - file->agent[id] = NULL; - } + /* + * Now go through the list of files attached to this port and + * unregister all of their MAD agents. We need to hold + * port->mutex while doing this to avoid racing with + * ib_umad_close(), but we can't hold the mutex for writing + * while calling ib_unregister_mad_agent(), since that might + * deadlock by calling back into queue_packet(). So we + * downgrade our lock to a read lock, and then drop and + * reacquire the write lock for the next iteration. + * + * We do list_del_init() on the file's list_head so that the + * list_del in ib_umad_close() is still OK, even after the + * file is removed from the list. + */ + while (!list_empty(&port->file_list)) { + file = list_entry(port->file_list.next, struct ib_umad_file, + port_list); + + file->agents_dead = 1; + list_del_init(&file->port_list); + + downgrade_write(&port->mutex); + + for (id = 0; id < IB_UMAD_MAX_AGENTS; ++id) + if (file->agent[id]) + ib_unregister_mad_agent(file->agent[id]); + + up_read(&port->mutex); + down_write(&port->mutex); + } up_write(&port->mutex); From 393b07258766130146b962bc294d66615a47468a Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Thu, 10 Nov 2005 12:47:50 -0800 Subject: [PATCH 111/121] [SPARC64]: Re-export uts_sem for solaris compat module. Revert: b26b9bc58263acda274f82a9dde8b6d96559878a Signed-off-by: David S. Miller --- kernel/sys.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/kernel/sys.c b/kernel/sys.c index c43b3e22bbda..bce933ebb29f 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -1497,6 +1497,8 @@ EXPORT_SYMBOL(in_egroup_p); DECLARE_RWSEM(uts_sem); +EXPORT_SYMBOL(uts_sem); + asmlinkage long sys_newuname(struct new_utsname __user * name) { int errno = 0; From e623a3b8086208eb118d9501b8bfb3e656e754d4 Mon Sep 17 00:00:00 2001 From: Zachary Amsden Date: Mon, 7 Nov 2005 15:27:51 -0800 Subject: [PATCH 112/121] [PATCH] kbuild: build breaks after -imacros was introduced I have to revert the recent addition of -imacros to the Makefile to get my tool chain to build. Without the change, below, I get: Note that this looks entirely like a toolchain bug. Here is the offending command: [pid 12163] execve("/usr/lib/gcc-lib/i386-redhat-linux/3.2.2/tradcpp0", ["/usr/lib/gcc-lib/i386-redhat-linux/3.2.2/tradcpp0", "-lang-asm", "-nostdinc", "-Iinclude", "-Iinclude/asm-i386/mach-default", "-D__GNUC__=3", "-D__GNUC_MINOR__=2", "-D__GNUC_PATCHLEVEL__=2", "-D__GXX_ABI_VERSION=102", "-D__ELF__", "-Dunix", "-D__gnu_linux__", "-Dlinux", "-D__ELF__", "-D__unix__", "-D__gnu_linux__", "-D__linux__", "-D__unix", "-D__linux", "-Asystem=posix", "-D__NO_INLINE__", "-D__STDC_HOSTED__=1", "-Acpu=i386", "-Amachine=i386", "-Di386", "-D__i386", "-D__i386__", "-D__tune_i386__", "-D__KERNEL__", "-D__ASSEMBLY__", "-isystem", "/usr/lib/gcc-lib/i386-redhat-linux/3.2.2/include", "-imacros", "include/linux/autoconf.h", "-MD", "arch/i386/kernel/.entry.o.d", "arch/i386/kernel/entry.S", "-o", "/tmp/ccOlsFJR.s"] Which should execute properly, I think. But it does not: zach-dev:linux-2.6.14-zach-work $ make CHK include/linux/version.h CHK include/linux/compile.h CHK usr/initramfs_list AS arch/i386/kernel/entry.o /usr/lib/gcc-lib/i386-redhat-linux/3.2.2/tradcpp0: output filename specified twice make[1]: *** [arch/i386/kernel/entry.o] Error 1 make: *** [arch/i386/kernel] Error 2 gcc (GCC) 3.2.2 20030222 (Red Hat Linux 3.2.2-5) Deprecating the -imacros fixes the build for me. It does not appear to be a simple argument overflow problem in trapcpp0, since deprecating all the defines reproduces the problem as well. Also, switching -imacros to -include fixes the problem. Signed-off-by: Zachary Amsden Signed-off-by: Sam Ravnborg --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 3152d6377eb2..83a3185670f1 100644 --- a/Makefile +++ b/Makefile @@ -347,7 +347,7 @@ AFLAGS_KERNEL = # Needed to be compatible with the O= option LINUXINCLUDE := -Iinclude \ $(if $(KBUILD_SRC),-Iinclude2 -I$(srctree)/include) \ - -imacros include/linux/autoconf.h + -include include/linux/autoconf.h CPPFLAGS := -D__KERNEL__ $(LINUXINCLUDE) From b925556cc9e82b32ab68a7620b247f47193501a7 Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Thu, 10 Nov 2005 12:55:52 -0800 Subject: [PATCH 113/121] [ATM]: [horizon] fix sparse warnings Signed-off-by: Dave Jones Signed-off-by: Chas Williams Signed-off-by: David S. Miller --- drivers/atm/horizon.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/atm/horizon.c b/drivers/atm/horizon.c index 0cded0468003..821c81e8cd38 100644 --- a/drivers/atm/horizon.c +++ b/drivers/atm/horizon.c @@ -1511,8 +1511,8 @@ static inline short setup_idle_tx_channel (hrz_dev * dev, hrz_vcc * vcc) { // a.k.a. prepare the channel and remember that we have done so. tx_ch_desc * tx_desc = &memmap->tx_descs[tx_channel]; - u16 rd_ptr; - u16 wr_ptr; + u32 rd_ptr; + u32 wr_ptr; u16 channel = vcc->channel; unsigned long flags; From 2f23c523f32324e5b5f39565cbcb0a8ff8923019 Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Thu, 10 Nov 2005 12:57:33 -0800 Subject: [PATCH 114/121] [BNX2]: output driver name as prefix in error message Output driver name as prefix to "Unknown flash/EEPROM type." message. Signed-off-by: John W. Linville Signed-off-by: David S. Miller --- drivers/net/bnx2.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c index 8f464271664d..65aae2828f39 100644 --- a/drivers/net/bnx2.c +++ b/drivers/net/bnx2.c @@ -2707,7 +2707,7 @@ bnx2_init_nvram(struct bnx2 *bp) if (j == entry_count) { bp->flash_info = NULL; - printk(KERN_ALERT "Unknown flash/EEPROM type.\n"); + printk(KERN_ALERT PFX "Unknown flash/EEPROM type.\n"); rc = -ENODEV; } From b6cbc3b6fe588c0ea1341d10413e12c2a96a6032 Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Thu, 10 Nov 2005 12:58:00 -0800 Subject: [PATCH 115/121] [BNX2]: check return of dev_alloc_skb in bnx2_test_loopback Check return of dev_alloc_skb in bnx2_test_loopback, and handle appropriately. Signed-off-by: John W. Linville Signed-off-by: David S. Miller --- drivers/net/bnx2.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c index 65aae2828f39..32267d538b17 100644 --- a/drivers/net/bnx2.c +++ b/drivers/net/bnx2.c @@ -3903,6 +3903,8 @@ bnx2_test_loopback(struct bnx2 *bp) pkt_size = 1514; skb = dev_alloc_skb(pkt_size); + if (!skb) + return -ENOMEM; packet = skb_put(skb, pkt_size); memcpy(packet, bp->mac_addr, 6); memset(packet + 6, 0x0, 8); From 1064e944d03eb7a72c0fa11236d5e69cfd877a71 Mon Sep 17 00:00:00 2001 From: "John W. Linville" Date: Thu, 10 Nov 2005 12:58:24 -0800 Subject: [PATCH 116/121] [BNX2]: simplify parameter checks in bnx2_{get,set}_eeprom Remove the superfluous parameter checking in bnx2_{get,set}_eeprom. The parameters are already validated in ethtool_{get,set}_eeprom. Signed-off-by: John W. Linville Signed-off-by: David S. Miller --- drivers/net/bnx2.c | 12 ++---------- 1 file changed, 2 insertions(+), 10 deletions(-) diff --git a/drivers/net/bnx2.c b/drivers/net/bnx2.c index 32267d538b17..49fa1e4413fa 100644 --- a/drivers/net/bnx2.c +++ b/drivers/net/bnx2.c @@ -4800,11 +4800,7 @@ bnx2_get_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom, struct bnx2 *bp = dev->priv; int rc; - if (eeprom->offset > bp->flash_info->total_size) - return -EINVAL; - - if ((eeprom->offset + eeprom->len) > bp->flash_info->total_size) - eeprom->len = bp->flash_info->total_size - eeprom->offset; + /* parameters already validated in ethtool_get_eeprom */ rc = bnx2_nvram_read(bp, eeprom->offset, eebuf, eeprom->len); @@ -4818,11 +4814,7 @@ bnx2_set_eeprom(struct net_device *dev, struct ethtool_eeprom *eeprom, struct bnx2 *bp = dev->priv; int rc; - if (eeprom->offset > bp->flash_info->total_size) - return -EINVAL; - - if ((eeprom->offset + eeprom->len) > bp->flash_info->total_size) - eeprom->len = bp->flash_info->total_size - eeprom->offset; + /* parameters already validated in ethtool_set_eeprom */ rc = bnx2_nvram_write(bp, eeprom->offset, eebuf, eeprom->len); From fb286bb2990a107009dbf25f6ffebeb7df77f9be Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Thu, 10 Nov 2005 13:01:24 -0800 Subject: [PATCH 117/121] [NET]: Detect hardware rx checksum faults correctly Here is the patch that introduces the generic skb_checksum_complete which also checks for hardware RX checksum faults. If that happens, it'll call netdev_rx_csum_fault which currently prints out a stack trace with the device name. In future it can turn off RX checksum. I've converted every spot under net/ that does RX checksum checks to use skb_checksum_complete or __skb_checksum_complete with the exceptions of: * Those places where checksums are done bit by bit. These will call netdev_rx_csum_fault directly. * The following have not been completely checked/converted: ipmr ip_vs netfilter dccp This patch is based on patches and suggestions from Stephen Hemminger and David S. Miller. Signed-off-by: Herbert Xu Signed-off-by: David S. Miller --- include/linux/netdevice.h | 7 ++++ include/linux/skbuff.h | 27 ++++++++++++- include/net/tcp.h | 3 +- net/core/datagram.c | 21 ++++++++-- net/core/dev.c | 12 ++++++ net/core/netpoll.c | 18 +++++---- net/ipv4/icmp.c | 6 +-- net/ipv4/igmp.c | 19 ++++++--- net/ipv4/ip_gre.c | 15 +++---- net/ipv4/netfilter/ip_conntrack_proto_icmp.c | 11 ++--- net/ipv4/tcp_ipv4.c | 24 +++++------ net/ipv4/udp.c | 7 +--- net/ipv6/icmp.c | 21 +++++----- net/ipv6/raw.c | 42 ++++++++------------ net/ipv6/tcp_ipv6.c | 20 +++++----- net/ipv6/udp.c | 25 +++++------- net/rxrpc/transport.c | 15 +++---- net/sunrpc/socklib.c | 5 +++ net/sunrpc/svcsock.c | 9 ++--- 19 files changed, 173 insertions(+), 134 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index c6efce4a04a4..936f8b76114e 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -927,6 +927,13 @@ extern int netdev_max_backlog; extern int weight_p; extern int netdev_set_master(struct net_device *dev, struct net_device *master); extern int skb_checksum_help(struct sk_buff *skb, int inward); +#ifdef CONFIG_BUG +extern void netdev_rx_csum_fault(struct net_device *dev); +#else +static inline void netdev_rx_csum_fault(struct net_device *dev) +{ +} +#endif /* rx skb timestamps */ extern void net_enable_timestamp(void); extern void net_disable_timestamp(void); diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 83010231db99..0a8ea8b35816 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -1236,8 +1236,7 @@ extern unsigned int datagram_poll(struct file *file, struct socket *sock, extern int skb_copy_datagram_iovec(const struct sk_buff *from, int offset, struct iovec *to, int size); -extern int skb_copy_and_csum_datagram_iovec(const - struct sk_buff *skb, +extern int skb_copy_and_csum_datagram_iovec(struct sk_buff *skb, int hlen, struct iovec *iov); extern void skb_free_datagram(struct sock *sk, struct sk_buff *skb); @@ -1305,6 +1304,30 @@ static inline void skb_set_timestamp(struct sk_buff *skb, const struct timeval * extern void __net_timestamp(struct sk_buff *skb); +extern unsigned int __skb_checksum_complete(struct sk_buff *skb); + +/** + * skb_checksum_complete - Calculate checksum of an entire packet + * @skb: packet to process + * + * This function calculates the checksum over the entire packet plus + * the value of skb->csum. The latter can be used to supply the + * checksum of a pseudo header as used by TCP/UDP. It returns the + * checksum. + * + * For protocols that contain complete checksums such as ICMP/TCP/UDP, + * this function can be used to verify that checksum on received + * packets. In that case the function should return zero if the + * checksum is correct. In particular, this function will return zero + * if skb->ip_summed is CHECKSUM_UNNECESSARY which indicates that the + * hardware has already verified the correctness of the checksum. + */ +static inline unsigned int skb_checksum_complete(struct sk_buff *skb) +{ + return skb->ip_summed != CHECKSUM_UNNECESSARY && + __skb_checksum_complete(skb); +} + #ifdef CONFIG_NETFILTER static inline void nf_conntrack_put(struct nf_conntrack *nfct) { diff --git a/include/net/tcp.h b/include/net/tcp.h index c24339c4e310..96cc3b434e40 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -27,6 +27,7 @@ #include #include #include +#include #include #include @@ -852,7 +853,7 @@ static __inline__ u16 tcp_v4_check(struct tcphdr *th, int len, static __inline__ int __tcp_checksum_complete(struct sk_buff *skb) { - return (unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum)); + return __skb_checksum_complete(skb); } static __inline__ int tcp_checksum_complete(struct sk_buff *skb) diff --git a/net/core/datagram.c b/net/core/datagram.c index d219435d086c..1bcfef51ac58 100644 --- a/net/core/datagram.c +++ b/net/core/datagram.c @@ -350,6 +350,20 @@ fault: return -EFAULT; } +unsigned int __skb_checksum_complete(struct sk_buff *skb) +{ + unsigned int sum; + + sum = (u16)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum)); + if (likely(!sum)) { + if (unlikely(skb->ip_summed == CHECKSUM_HW)) + netdev_rx_csum_fault(skb->dev); + skb->ip_summed = CHECKSUM_UNNECESSARY; + } + return sum; +} +EXPORT_SYMBOL(__skb_checksum_complete); + /** * skb_copy_and_csum_datagram_iovec - Copy and checkum skb to user iovec. * @skb: skbuff @@ -363,7 +377,7 @@ fault: * -EFAULT - fault during copy. Beware, in this case iovec * can be modified! */ -int skb_copy_and_csum_datagram_iovec(const struct sk_buff *skb, +int skb_copy_and_csum_datagram_iovec(struct sk_buff *skb, int hlen, struct iovec *iov) { unsigned int csum; @@ -376,8 +390,7 @@ int skb_copy_and_csum_datagram_iovec(const struct sk_buff *skb, iov++; if (iov->iov_len < chunk) { - if ((unsigned short)csum_fold(skb_checksum(skb, 0, chunk + hlen, - skb->csum))) + if (__skb_checksum_complete(skb)) goto csum_error; if (skb_copy_datagram_iovec(skb, hlen, iov, chunk)) goto fault; @@ -388,6 +401,8 @@ int skb_copy_and_csum_datagram_iovec(const struct sk_buff *skb, goto fault; if ((unsigned short)csum_fold(csum)) goto csum_error; + if (unlikely(skb->ip_summed == CHECKSUM_HW)) + netdev_rx_csum_fault(skb->dev); iov->iov_len -= chunk; iov->iov_base += chunk; } diff --git a/net/core/dev.c b/net/core/dev.c index 8d1541595277..0b48e294aafe 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -1108,6 +1108,18 @@ out: return ret; } +/* Take action when hardware reception checksum errors are detected. */ +#ifdef CONFIG_BUG +void netdev_rx_csum_fault(struct net_device *dev) +{ + if (net_ratelimit()) { + printk(KERN_ERR "%s: hw csum failure.\n", dev->name); + dump_stack(); + } +} +EXPORT_SYMBOL(netdev_rx_csum_fault); +#endif + #ifdef CONFIG_HIGHMEM /* Actually, we should eliminate this check as soon as we know, that: * 1. IOMMU is present and allows to map all the memory. diff --git a/net/core/netpoll.c b/net/core/netpoll.c index 802fe11efad0..49424a42a2c0 100644 --- a/net/core/netpoll.c +++ b/net/core/netpoll.c @@ -101,16 +101,20 @@ void netpoll_queue(struct sk_buff *skb) static int checksum_udp(struct sk_buff *skb, struct udphdr *uh, unsigned short ulen, u32 saddr, u32 daddr) { - if (uh->check == 0) + unsigned int psum; + + if (uh->check == 0 || skb->ip_summed == CHECKSUM_UNNECESSARY) return 0; - if (skb->ip_summed == CHECKSUM_HW) - return csum_tcpudp_magic( - saddr, daddr, ulen, IPPROTO_UDP, skb->csum); + psum = csum_tcpudp_nofold(saddr, daddr, ulen, IPPROTO_UDP, 0); - skb->csum = csum_tcpudp_nofold(saddr, daddr, ulen, IPPROTO_UDP, 0); + if (skb->ip_summed == CHECKSUM_HW && + !(u16)csum_fold(csum_add(psum, skb->csum))) + return 0; - return csum_fold(skb_checksum(skb, 0, skb->len, skb->csum)); + skb->csum = psum; + + return __skb_checksum_complete(skb); } /* @@ -489,7 +493,7 @@ int __netpoll_rx(struct sk_buff *skb) if (ulen != len) goto out; - if (checksum_udp(skb, uh, ulen, iph->saddr, iph->daddr) < 0) + if (checksum_udp(skb, uh, ulen, iph->saddr, iph->daddr)) goto out; if (np->local_ip && np->local_ip != ntohl(iph->daddr)) goto out; diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index 175e093ec564..e3eceecd0496 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -934,11 +934,11 @@ int icmp_rcv(struct sk_buff *skb) case CHECKSUM_HW: if (!(u16)csum_fold(skb->csum)) break; - LIMIT_NETDEBUG(KERN_DEBUG "icmp v4 hw csum failure\n"); + /* fall through */ case CHECKSUM_NONE: - if ((u16)csum_fold(skb_checksum(skb, 0, skb->len, 0))) + skb->csum = 0; + if (__skb_checksum_complete(skb)) goto error; - default:; } if (!pskb_pull(skb, sizeof(struct icmphdr))) diff --git a/net/ipv4/igmp.c b/net/ipv4/igmp.c index c6247fc84060..c04607b49212 100644 --- a/net/ipv4/igmp.c +++ b/net/ipv4/igmp.c @@ -872,11 +872,18 @@ int igmp_rcv(struct sk_buff *skb) return 0; } - if (!pskb_may_pull(skb, sizeof(struct igmphdr)) || - (u16)csum_fold(skb_checksum(skb, 0, len, 0))) { - in_dev_put(in_dev); - kfree_skb(skb); - return 0; + if (!pskb_may_pull(skb, sizeof(struct igmphdr))) + goto drop; + + switch (skb->ip_summed) { + case CHECKSUM_HW: + if (!(u16)csum_fold(skb->csum)) + break; + /* fall through */ + case CHECKSUM_NONE: + skb->csum = 0; + if (__skb_checksum_complete(skb)) + goto drop; } ih = skb->h.igmph; @@ -906,6 +913,8 @@ int igmp_rcv(struct sk_buff *skb) default: NETDEBUG(KERN_DEBUG "New IGMP type=%d, why we do not know about it?\n", ih->type); } + +drop: in_dev_put(in_dev); kfree_skb(skb); return 0; diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c index 896ce3f8f53a..4e9c74b54b15 100644 --- a/net/ipv4/ip_gre.c +++ b/net/ipv4/ip_gre.c @@ -577,15 +577,16 @@ static int ipgre_rcv(struct sk_buff *skb) goto drop_nolock; if (flags&GRE_CSUM) { - if (skb->ip_summed == CHECKSUM_HW) { + switch (skb->ip_summed) { + case CHECKSUM_HW: csum = (u16)csum_fold(skb->csum); - if (csum) - skb->ip_summed = CHECKSUM_NONE; - } - if (skb->ip_summed == CHECKSUM_NONE) { - skb->csum = skb_checksum(skb, 0, skb->len, 0); + if (!csum) + break; + /* fall through */ + case CHECKSUM_NONE: + skb->csum = 0; + csum = __skb_checksum_complete(skb); skb->ip_summed = CHECKSUM_HW; - csum = (u16)csum_fold(skb->csum); } offset += 4; } diff --git a/net/ipv4/netfilter/ip_conntrack_proto_icmp.c b/net/ipv4/netfilter/ip_conntrack_proto_icmp.c index 5198f3a1e2cd..e4d6b268e8c4 100644 --- a/net/ipv4/netfilter/ip_conntrack_proto_icmp.c +++ b/net/ipv4/netfilter/ip_conntrack_proto_icmp.c @@ -13,6 +13,7 @@ #include #include #include +#include #include #include #include @@ -230,19 +231,15 @@ icmp_error(struct sk_buff *skb, enum ip_conntrack_info *ctinfo, case CHECKSUM_HW: if (!(u16)csum_fold(skb->csum)) break; - if (LOG_INVALID(IPPROTO_ICMP)) - nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, - "ip_ct_icmp: bad HW ICMP checksum "); - return -NF_ACCEPT; + /* fall through */ case CHECKSUM_NONE: - if ((u16)csum_fold(skb_checksum(skb, 0, skb->len, 0))) { + skb->csum = 0; + if (__skb_checksum_complete(skb)) { if (LOG_INVALID(IPPROTO_ICMP)) nf_log_packet(PF_INET, 0, skb, NULL, NULL, NULL, "ip_ct_icmp: bad ICMP checksum "); return -NF_ACCEPT; } - default: - break; } checksum_skipped: diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 634dabb558fd..ac1fcf5b4ebc 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -1110,24 +1110,18 @@ static struct sock *tcp_v4_hnd_req(struct sock *sk, struct sk_buff *skb) static int tcp_v4_checksum_init(struct sk_buff *skb) { if (skb->ip_summed == CHECKSUM_HW) { - skb->ip_summed = CHECKSUM_UNNECESSARY; if (!tcp_v4_check(skb->h.th, skb->len, skb->nh.iph->saddr, - skb->nh.iph->daddr, skb->csum)) + skb->nh.iph->daddr, skb->csum)) { + skb->ip_summed = CHECKSUM_UNNECESSARY; return 0; - - LIMIT_NETDEBUG(KERN_DEBUG "hw tcp v4 csum failed\n"); - skb->ip_summed = CHECKSUM_NONE; + } } + + skb->csum = csum_tcpudp_nofold(skb->nh.iph->saddr, skb->nh.iph->daddr, + skb->len, IPPROTO_TCP, 0); + if (skb->len <= 76) { - if (tcp_v4_check(skb->h.th, skb->len, skb->nh.iph->saddr, - skb->nh.iph->daddr, - skb_checksum(skb, 0, skb->len, 0))) - return -1; - skb->ip_summed = CHECKSUM_UNNECESSARY; - } else { - skb->csum = ~tcp_v4_check(skb->h.th, skb->len, - skb->nh.iph->saddr, - skb->nh.iph->daddr, 0); + return __skb_checksum_complete(skb); } return 0; } @@ -1219,7 +1213,7 @@ int tcp_v4_rcv(struct sk_buff *skb) * provided case of th->doff==0 is elimineted. * So, we defer the checks. */ if ((skb->ip_summed != CHECKSUM_UNNECESSARY && - tcp_v4_checksum_init(skb) < 0)) + tcp_v4_checksum_init(skb))) goto bad_packet; th = skb->h.th; diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c index e0bd1013cb0d..2422a5f7195d 100644 --- a/net/ipv4/udp.c +++ b/net/ipv4/udp.c @@ -761,7 +761,7 @@ int udp_ioctl(struct sock *sk, int cmd, unsigned long arg) static __inline__ int __udp_checksum_complete(struct sk_buff *skb) { - return (unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum)); + return __skb_checksum_complete(skb); } static __inline__ int udp_checksum_complete(struct sk_buff *skb) @@ -1100,11 +1100,8 @@ static int udp_checksum_init(struct sk_buff *skb, struct udphdr *uh, if (uh->check == 0) { skb->ip_summed = CHECKSUM_UNNECESSARY; } else if (skb->ip_summed == CHECKSUM_HW) { - skb->ip_summed = CHECKSUM_UNNECESSARY; if (!udp_check(uh, ulen, saddr, daddr, skb->csum)) - return 0; - LIMIT_NETDEBUG(KERN_DEBUG "udp v4 hw csum failure.\n"); - skb->ip_summed = CHECKSUM_NONE; + skb->ip_summed = CHECKSUM_UNNECESSARY; } if (skb->ip_summed != CHECKSUM_UNNECESSARY) skb->csum = csum_tcpudp_nofold(saddr, daddr, ulen, IPPROTO_UDP, 0); diff --git a/net/ipv6/icmp.c b/net/ipv6/icmp.c index 23e540365a14..1bdf0fb8bf8a 100644 --- a/net/ipv6/icmp.c +++ b/net/ipv6/icmp.c @@ -585,17 +585,16 @@ static int icmpv6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) daddr = &skb->nh.ipv6h->daddr; /* Perform checksum. */ - if (skb->ip_summed == CHECKSUM_HW) { - skb->ip_summed = CHECKSUM_UNNECESSARY; - if (csum_ipv6_magic(saddr, daddr, skb->len, IPPROTO_ICMPV6, - skb->csum)) { - LIMIT_NETDEBUG(KERN_DEBUG "ICMPv6 hw checksum failed\n"); - skb->ip_summed = CHECKSUM_NONE; - } - } - if (skb->ip_summed == CHECKSUM_NONE) { - if (csum_ipv6_magic(saddr, daddr, skb->len, IPPROTO_ICMPV6, - skb_checksum(skb, 0, skb->len, 0))) { + switch (skb->ip_summed) { + case CHECKSUM_HW: + if (!csum_ipv6_magic(saddr, daddr, skb->len, IPPROTO_ICMPV6, + skb->csum)) + break; + /* fall through */ + case CHECKSUM_NONE: + skb->csum = ~csum_ipv6_magic(saddr, daddr, skb->len, + IPPROTO_ICMPV6, 0); + if (__skb_checksum_complete(skb)) { LIMIT_NETDEBUG(KERN_DEBUG "ICMPv6 checksum failed [%04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x > %04x:%04x:%04x:%04x:%04x:%04x:%04x:%04x]\n", NIP6(*saddr), NIP6(*daddr)); goto discard_it; diff --git a/net/ipv6/raw.c b/net/ipv6/raw.c index 651c79b41eeb..8e9628f1c4c5 100644 --- a/net/ipv6/raw.c +++ b/net/ipv6/raw.c @@ -298,13 +298,10 @@ void rawv6_err(struct sock *sk, struct sk_buff *skb, static inline int rawv6_rcv_skb(struct sock * sk, struct sk_buff * skb) { if ((raw6_sk(sk)->checksum || sk->sk_filter) && - skb->ip_summed != CHECKSUM_UNNECESSARY) { - if ((unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum))) { - /* FIXME: increment a raw6 drops counter here */ - kfree_skb(skb); - return 0; - } - skb->ip_summed = CHECKSUM_UNNECESSARY; + skb_checksum_complete(skb)) { + /* FIXME: increment a raw6 drops counter here */ + kfree_skb(skb); + return 0; } /* Charge it to the socket. */ @@ -337,32 +334,25 @@ int rawv6_rcv(struct sock *sk, struct sk_buff *skb) if (!rp->checksum) skb->ip_summed = CHECKSUM_UNNECESSARY; - if (skb->ip_summed != CHECKSUM_UNNECESSARY) { - if (skb->ip_summed == CHECKSUM_HW) { - skb_postpull_rcsum(skb, skb->nh.raw, - skb->h.raw - skb->nh.raw); + if (skb->ip_summed == CHECKSUM_HW) { + skb_postpull_rcsum(skb, skb->nh.raw, + skb->h.raw - skb->nh.raw); + if (!csum_ipv6_magic(&skb->nh.ipv6h->saddr, + &skb->nh.ipv6h->daddr, + skb->len, inet->num, skb->csum)) skb->ip_summed = CHECKSUM_UNNECESSARY; - if (csum_ipv6_magic(&skb->nh.ipv6h->saddr, - &skb->nh.ipv6h->daddr, - skb->len, inet->num, skb->csum)) { - LIMIT_NETDEBUG(KERN_DEBUG "raw v6 hw csum failure.\n"); - skb->ip_summed = CHECKSUM_NONE; - } - } - if (skb->ip_summed == CHECKSUM_NONE) - skb->csum = ~csum_ipv6_magic(&skb->nh.ipv6h->saddr, - &skb->nh.ipv6h->daddr, - skb->len, inet->num, 0); } + if (skb->ip_summed != CHECKSUM_UNNECESSARY) + skb->csum = ~csum_ipv6_magic(&skb->nh.ipv6h->saddr, + &skb->nh.ipv6h->daddr, + skb->len, inet->num, 0); if (inet->hdrincl) { - if (skb->ip_summed != CHECKSUM_UNNECESSARY && - (unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum))) { + if (skb_checksum_complete(skb)) { /* FIXME: increment a raw6 drops counter here */ kfree_skb(skb); return 0; } - skb->ip_summed = CHECKSUM_UNNECESSARY; } rawv6_rcv_skb(sk, skb); @@ -407,7 +397,7 @@ static int rawv6_recvmsg(struct kiocb *iocb, struct sock *sk, if (skb->ip_summed==CHECKSUM_UNNECESSARY) { err = skb_copy_datagram_iovec(skb, 0, msg->msg_iov, copied); } else if (msg->msg_flags&MSG_TRUNC) { - if ((unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum))) + if (__skb_checksum_complete(skb)) goto csum_copy_err; err = skb_copy_datagram_iovec(skb, 0, msg->msg_iov, copied); } else { diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c index d746d3b27efb..62c0e5bd931c 100644 --- a/net/ipv6/tcp_ipv6.c +++ b/net/ipv6/tcp_ipv6.c @@ -1401,20 +1401,18 @@ out: static int tcp_v6_checksum_init(struct sk_buff *skb) { if (skb->ip_summed == CHECKSUM_HW) { - skb->ip_summed = CHECKSUM_UNNECESSARY; if (!tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr, - &skb->nh.ipv6h->daddr,skb->csum)) + &skb->nh.ipv6h->daddr,skb->csum)) { + skb->ip_summed = CHECKSUM_UNNECESSARY; return 0; - LIMIT_NETDEBUG(KERN_DEBUG "hw tcp v6 csum failed\n"); + } } + + skb->csum = ~tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr, + &skb->nh.ipv6h->daddr, 0); + if (skb->len <= 76) { - if (tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr, - &skb->nh.ipv6h->daddr,skb_checksum(skb, 0, skb->len, 0))) - return -1; - skb->ip_summed = CHECKSUM_UNNECESSARY; - } else { - skb->csum = ~tcp_v6_check(skb->h.th,skb->len,&skb->nh.ipv6h->saddr, - &skb->nh.ipv6h->daddr,0); + return __skb_checksum_complete(skb); } return 0; } @@ -1575,7 +1573,7 @@ static int tcp_v6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) goto discard_it; if ((skb->ip_summed != CHECKSUM_UNNECESSARY && - tcp_v6_checksum_init(skb) < 0)) + tcp_v6_checksum_init(skb))) goto bad_packet; th = skb->h.th; diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c index bf9519341fd3..e671153b47b2 100644 --- a/net/ipv6/udp.c +++ b/net/ipv6/udp.c @@ -248,7 +248,7 @@ try_again: err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov, copied); } else if (msg->msg_flags&MSG_TRUNC) { - if ((unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum))) + if (__skb_checksum_complete(skb)) goto csum_copy_err; err = skb_copy_datagram_iovec(skb, sizeof(struct udphdr), msg->msg_iov, copied); @@ -363,13 +363,10 @@ static inline int udpv6_queue_rcv_skb(struct sock * sk, struct sk_buff *skb) return -1; } - if (skb->ip_summed != CHECKSUM_UNNECESSARY) { - if ((unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum))) { - UDP6_INC_STATS_BH(UDP_MIB_INERRORS); - kfree_skb(skb); - return 0; - } - skb->ip_summed = CHECKSUM_UNNECESSARY; + if (skb_checksum_complete(skb)) { + UDP6_INC_STATS_BH(UDP_MIB_INERRORS); + kfree_skb(skb); + return 0; } if (sock_queue_rcv_skb(sk,skb)<0) { @@ -491,13 +488,10 @@ static int udpv6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) uh = skb->h.uh; } - if (skb->ip_summed==CHECKSUM_HW) { + if (skb->ip_summed == CHECKSUM_HW && + !csum_ipv6_magic(saddr, daddr, ulen, IPPROTO_UDP, skb->csum)) skb->ip_summed = CHECKSUM_UNNECESSARY; - if (csum_ipv6_magic(saddr, daddr, ulen, IPPROTO_UDP, skb->csum)) { - LIMIT_NETDEBUG(KERN_DEBUG "udp v6 hw csum failure.\n"); - skb->ip_summed = CHECKSUM_NONE; - } - } + if (skb->ip_summed != CHECKSUM_UNNECESSARY) skb->csum = ~csum_ipv6_magic(saddr, daddr, ulen, IPPROTO_UDP, 0); @@ -521,8 +515,7 @@ static int udpv6_rcv(struct sk_buff **pskb, unsigned int *nhoffp) if (!xfrm6_policy_check(NULL, XFRM_POLICY_IN, skb)) goto discard; - if (skb->ip_summed != CHECKSUM_UNNECESSARY && - (unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum))) + if (skb_checksum_complete(skb)) goto discard; UDP6_INC_STATS_BH(UDP_MIB_NOPORTS); diff --git a/net/rxrpc/transport.c b/net/rxrpc/transport.c index 122c086ee2db..dbe6105e83a5 100644 --- a/net/rxrpc/transport.c +++ b/net/rxrpc/transport.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #include #if defined(CONFIG_IPV6) || defined (CONFIG_IPV6_MODULE) @@ -475,15 +476,11 @@ void rxrpc_trans_receive_packet(struct rxrpc_transport *trans) /* we'll probably need to checksum it (didn't call * sock_recvmsg) */ - if (pkt->ip_summed != CHECKSUM_UNNECESSARY) { - if ((unsigned short) - csum_fold(skb_checksum(pkt, 0, pkt->len, - pkt->csum))) { - kfree_skb(pkt); - rxrpc_krxiod_queue_transport(trans); - _leave(" CSUM failed"); - return; - } + if (skb_checksum_complete(pkt)) { + kfree_skb(pkt); + rxrpc_krxiod_queue_transport(trans); + _leave(" CSUM failed"); + return; } addr = pkt->nh.iph->saddr; diff --git a/net/sunrpc/socklib.c b/net/sunrpc/socklib.c index 8f97e90f36c8..eb330d4f66d6 100644 --- a/net/sunrpc/socklib.c +++ b/net/sunrpc/socklib.c @@ -6,6 +6,9 @@ * Copyright (C) 1995, 1996 Olaf Kirch */ +#include +#include +#include #include #include #include @@ -165,6 +168,8 @@ int csum_partial_copy_to_xdr(struct xdr_buf *xdr, struct sk_buff *skb) return -1; if ((unsigned short)csum_fold(desc.csum)) return -1; + if (unlikely(skb->ip_summed == CHECKSUM_HW)) + netdev_rx_csum_fault(skb->dev); return 0; no_checksum: if (xdr_partial_copy_from_skb(xdr, 0, &desc, skb_read_bits) < 0) diff --git a/net/sunrpc/svcsock.c b/net/sunrpc/svcsock.c index f16e7cdd6150..e50e7cf43737 100644 --- a/net/sunrpc/svcsock.c +++ b/net/sunrpc/svcsock.c @@ -623,12 +623,9 @@ svc_udp_recvfrom(struct svc_rqst *rqstp) /* we can use it in-place */ rqstp->rq_arg.head[0].iov_base = skb->data + sizeof(struct udphdr); rqstp->rq_arg.head[0].iov_len = len; - if (skb->ip_summed != CHECKSUM_UNNECESSARY) { - if ((unsigned short)csum_fold(skb_checksum(skb, 0, skb->len, skb->csum))) { - skb_free_datagram(svsk->sk_sk, skb); - return 0; - } - skb->ip_summed = CHECKSUM_UNNECESSARY; + if (skb_checksum_complete(skb)) { + skb_free_datagram(svsk->sk_sk, skb); + return 0; } rqstp->rq_skbuff = skb; } From ac57d04267383829ce817e522e4ebbb39f50ae75 Mon Sep 17 00:00:00 2001 From: Pavel Roskin Date: Thu, 10 Nov 2005 13:03:08 -0800 Subject: [PATCH 118/121] [NET]: Annotate h_proto in struct ethhdr The protocol field in ethernet headers is big-endian and should be annotated as such. This patch allows detection of missing ntohs() calls on the ethernet protocol field when sparse is run with __CHECK_ENDIAN__ defined. This is a revised version that includes so that the userspace programs are not confused by __be16. Thanks to David S. Miller. Signed-off-by: Pavel Roskin Signed-off-by: David S. Miller --- include/linux/if_ether.h | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/include/linux/if_ether.h b/include/linux/if_ether.h index d21c305c6c64..fe26d431de87 100644 --- a/include/linux/if_ether.h +++ b/include/linux/if_ether.h @@ -21,6 +21,8 @@ #ifndef _LINUX_IF_ETHER_H #define _LINUX_IF_ETHER_H +#include + /* * IEEE 802.3 Ethernet magic constants. The frame sizes omit the preamble * and FCS/CRC (frame check sequence). @@ -100,7 +102,7 @@ struct ethhdr { unsigned char h_dest[ETH_ALEN]; /* destination eth addr */ unsigned char h_source[ETH_ALEN]; /* source ether addr */ - unsigned short h_proto; /* packet type ID field */ + __be16 h_proto; /* packet type ID field */ } __attribute__((packed)); #ifdef __KERNEL__ From 98069ff4ec9c73a03e6c61dbb17f8e5a0b85ef6e Mon Sep 17 00:00:00 2001 From: Ian McDonald Date: Thu, 10 Nov 2005 13:04:33 -0800 Subject: [PATCH 119/121] [DCCP]: Create Documentation/networking/dccp.txt This patch is a first go at some documentation. Please advise if gmail has mangled patch and I will revert to an attachment: Signed-off-by: Ian McDonald Signed-off-by: David S. Miller --- Documentation/networking/dccp.txt | 56 +++++++++++++++++++++++++++++++ 1 file changed, 56 insertions(+) create mode 100644 Documentation/networking/dccp.txt diff --git a/Documentation/networking/dccp.txt b/Documentation/networking/dccp.txt new file mode 100644 index 000000000000..c45daabd3bfe --- /dev/null +++ b/Documentation/networking/dccp.txt @@ -0,0 +1,56 @@ +DCCP protocol +============ + +Last updated: 10 November 2005 + +Contents +======== + +- Introduction +- Missing features +- Socket options +- Notes + +Introduction +============ + +Datagram Congestion Control Protocol (DCCP) is an unreliable, connection +based protocol designed to solve issues present in UDP and TCP particularly +for real time and multimedia traffic. + +It has a base protocol and pluggable congestion control IDs (CCIDs). + +It is at draft RFC status and the homepage for DCCP as a protocol is at: + http://www.icir.org/kohler/dcp/ + +Missing features +================ + +The DCCP implementation does not currently have all the features that are in +the draft RFC. + +In particular the following are missing: +- CCID2 support +- feature negotiation + +When testing against other implementations it appears that elapsed time +options are not coded compliant to the specification. + +Socket options +============== + +DCCP_SOCKOPT_PACKET_SIZE is used for CCID3 to set default packet size for +calculations. + +DCCP_SOCKOPT_SERVICE sets the service. This is compulsory as per the +specification. If you don't set it you will get EPROTO. + +Notes +===== + +SELinux does not yet have support for DCCP. You will need to turn it off or +else you will get EACCES. + +DCCP does not travel through NAT successfully at present. This is because +the checksum covers the psuedo-header as per TCP and UDP. It should be +relatively trivial to add Linux NAT support for DCCP. From f3b84ec2a67bbfbe44cf907203ee99b23b38fc30 Mon Sep 17 00:00:00 2001 From: Ian McDonald Date: Thu, 10 Nov 2005 13:05:01 -0800 Subject: [PATCH 120/121] [DCCP]: Alter website for DCCP in MAINTAINERS Website for DCCP is now hosted at OSDL Signed-off-by: Ian McDonald Signed-off-by: David S. Miller --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 5541f9970b87..0b03a88e88be 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -707,7 +707,7 @@ DCCP PROTOCOL P: Arnaldo Carvalho de Melo M: acme@mandriva.com L: dccp@vger.kernel.org -W: http://www.wlug.org.nz/DCCP +W: http://linux-net.osdl.org/index.php/DCCP S: Maintained DECnet NETWORK LAYER From 6b482c6779daaa893b277fc9b70767a7c2e7c5eb Mon Sep 17 00:00:00 2001 From: Dave Jones Date: Thu, 10 Nov 2005 15:45:56 -0500 Subject: [PATCH 121/121] [PATCH] Don't print per-cpu vm stats for offline cpus. I just hit a page allocation error on a kernel configured to support 64 CPUs. It spewed 60 completely useless unnecessary lines of info. Signed-off-by: Dave Jones Signed-off-by: Linus Torvalds --- mm/page_alloc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/page_alloc.c b/mm/page_alloc.c index ff81b5c65511..987225bdd661 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -1330,7 +1330,7 @@ void show_free_areas(void) } else printk("\n"); - for_each_cpu(cpu) { + for_each_online_cpu(cpu) { struct per_cpu_pageset *pageset; pageset = zone_pcp(zone, cpu);