remarkable-linux/drivers/usb/host/ehci-pci.c
Benjamin Herrenschmidt 8de9840265 [PATCH] USB: Fix USB suspend/resume crasher (#2)
This patch closes the IRQ race and makes various other OHCI & EHCI code
path safer vs. suspend/resume.
I've been able to (finally !) successfully suspend and resume various
Mac models, with or without USB mouse plugged, or plugging while asleep,
or unplugging while asleep etc... all without a crash.

Alan, please verify the UHCI bit I did, I only verified that it builds.
It's very simple so I wouldn't expect any issue there. If you aren't
confident, then just drop the hunks that change uhci-hcd.c

I also made the patch a little bit more "safer" by making sure the store
to the interrupt register that disables interrupts is not posted before
I set the flag and drop the spinlock.

Without this patch, you cannot reliably sleep/wakeup any recent Mac, and
I suspect PCs have some more sneaky issues too (they don't frankly crash
with machine checks because x86 tend to silently swallow PCI errors but
that won't last afaik, at least PCI Express will blow up in those
situations, but the USB code may still misbehave).

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
2005-11-29 21:39:23 -08:00

421 lines
11 KiB
C

/*
* EHCI HCD (Host Controller Driver) PCI Bus Glue.
*
* Copyright (c) 2000-2004 by David Brownell
*
* 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., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
#ifndef CONFIG_PCI
#error "This file is PCI bus glue. CONFIG_PCI must be defined."
#endif
/*-------------------------------------------------------------------------*/
/* EHCI 0.96 (and later) section 5.1 says how to kick BIOS/SMM/...
* off the controller (maybe it can boot from highspeed USB disks).
*/
static int bios_handoff(struct ehci_hcd *ehci, int where, u32 cap)
{
struct pci_dev *pdev = to_pci_dev(ehci_to_hcd(ehci)->self.controller);
/* always say Linux will own the hardware */
pci_write_config_byte(pdev, where + 3, 1);
/* maybe wait a while for BIOS to respond */
if (cap & (1 << 16)) {
int msec = 5000;
do {
msleep(10);
msec -= 10;
pci_read_config_dword(pdev, where, &cap);
} while ((cap & (1 << 16)) && msec);
if (cap & (1 << 16)) {
ehci_err(ehci, "BIOS handoff failed (%d, %08x)\n",
where, cap);
// some BIOS versions seem buggy...
// return 1;
ehci_warn(ehci, "continuing after BIOS bug...\n");
/* disable all SMIs, and clear "BIOS owns" flag */
pci_write_config_dword(pdev, where + 4, 0);
pci_write_config_byte(pdev, where + 2, 0);
} else
ehci_dbg(ehci, "BIOS handoff succeeded\n");
}
return 0;
}
/* called after powerup, by probe or system-pm "wakeup" */
static int ehci_pci_reinit(struct ehci_hcd *ehci, struct pci_dev *pdev)
{
u32 temp;
int retval;
unsigned count = 256/4;
/* optional debug port, normally in the first BAR */
temp = pci_find_capability(pdev, 0x0a);
if (temp) {
pci_read_config_dword(pdev, temp, &temp);
temp >>= 16;
if ((temp & (3 << 13)) == (1 << 13)) {
temp &= 0x1fff;
ehci->debug = ehci_to_hcd(ehci)->regs + temp;
temp = readl(&ehci->debug->control);
ehci_info(ehci, "debug port %d%s\n",
HCS_DEBUG_PORT(ehci->hcs_params),
(temp & DBGP_ENABLED)
? " IN USE"
: "");
if (!(temp & DBGP_ENABLED))
ehci->debug = NULL;
}
}
temp = HCC_EXT_CAPS(readl(&ehci->caps->hcc_params));
/* EHCI 0.96 and later may have "extended capabilities" */
while (temp && count--) {
u32 cap;
pci_read_config_dword(pdev, temp, &cap);
ehci_dbg(ehci, "capability %04x at %02x\n", cap, temp);
switch (cap & 0xff) {
case 1: /* BIOS/SMM/... handoff */
if (bios_handoff(ehci, temp, cap) != 0)
return -EOPNOTSUPP;
break;
case 0: /* illegal reserved capability */
ehci_dbg(ehci, "illegal capability!\n");
cap = 0;
/* FALLTHROUGH */
default: /* unknown */
break;
}
temp = (cap >> 8) & 0xff;
}
if (!count) {
ehci_err(ehci, "bogus capabilities ... PCI problems!\n");
return -EIO;
}
/* PCI Memory-Write-Invalidate cycle support is optional (uncommon) */
retval = pci_set_mwi(pdev);
if (!retval)
ehci_dbg(ehci, "MWI active\n");
ehci_port_power(ehci, 0);
return 0;
}
/* called by khubd or root hub (re)init threads; leaves HC in halt state */
static int ehci_pci_reset(struct usb_hcd *hcd)
{
struct ehci_hcd *ehci = hcd_to_ehci(hcd);
struct pci_dev *pdev = to_pci_dev(hcd->self.controller);
u32 temp;
int retval;
ehci->caps = hcd->regs;
ehci->regs = hcd->regs + HC_LENGTH(readl(&ehci->caps->hc_capbase));
dbg_hcs_params(ehci, "reset");
dbg_hcc_params(ehci, "reset");
/* cache this readonly data; minimize chip reads */
ehci->hcs_params = readl(&ehci->caps->hcs_params);
retval = ehci_halt(ehci);
if (retval)
return retval;
/* NOTE: only the parts below this line are PCI-specific */
switch (pdev->vendor) {
case PCI_VENDOR_ID_TDI:
if (pdev->device == PCI_DEVICE_ID_TDI_EHCI) {
ehci->is_tdi_rh_tt = 1;
tdi_reset(ehci);
}
break;
case PCI_VENDOR_ID_AMD:
/* AMD8111 EHCI doesn't work, according to AMD errata */
if (pdev->device == 0x7463) {
ehci_info(ehci, "ignoring AMD8111 (errata)\n");
return -EIO;
}
break;
case PCI_VENDOR_ID_NVIDIA:
/* NVidia reports that certain chips don't handle
* QH, ITD, or SITD addresses above 2GB. (But TD,
* data buffer, and periodic schedule are normal.)
*/
switch (pdev->device) {
case 0x003c: /* MCP04 */
case 0x005b: /* CK804 */
case 0x00d8: /* CK8 */
case 0x00e8: /* CK8S */
if (pci_set_consistent_dma_mask(pdev,
DMA_31BIT_MASK) < 0)
ehci_warn(ehci, "can't enable NVidia "
"workaround for >2GB RAM\n");
break;
}
break;
}
if (ehci_is_TDI(ehci))
ehci_reset(ehci);
/* at least the Genesys GL880S needs fixup here */
temp = HCS_N_CC(ehci->hcs_params) * HCS_N_PCC(ehci->hcs_params);
temp &= 0x0f;
if (temp && HCS_N_PORTS(ehci->hcs_params) > temp) {
ehci_dbg(ehci, "bogus port configuration: "
"cc=%d x pcc=%d < ports=%d\n",
HCS_N_CC(ehci->hcs_params),
HCS_N_PCC(ehci->hcs_params),
HCS_N_PORTS(ehci->hcs_params));
switch (pdev->vendor) {
case 0x17a0: /* GENESYS */
/* GL880S: should be PORTS=2 */
temp |= (ehci->hcs_params & ~0xf);
ehci->hcs_params = temp;
break;
case PCI_VENDOR_ID_NVIDIA:
/* NF4: should be PCC=10 */
break;
}
}
/* Serial Bus Release Number is at PCI 0x60 offset */
pci_read_config_byte(pdev, 0x60, &ehci->sbrn);
/* REVISIT: per-port wake capability (PCI 0x62) currently unused */
retval = ehci_pci_reinit(ehci, pdev);
/* finish init */
return ehci_init(hcd);
}
/*-------------------------------------------------------------------------*/
#ifdef CONFIG_PM
/* suspend/resume, section 4.3 */
/* These routines rely on the PCI bus glue
* to handle powerdown and wakeup, and currently also on
* transceivers that don't need any software attention to set up
* the right sort of wakeup.
* Also they depend on separate root hub suspend/resume.
*/
static int ehci_pci_suspend(struct usb_hcd *hcd, pm_message_t message)
{
struct ehci_hcd *ehci = hcd_to_ehci(hcd);
unsigned long flags;
int rc = 0;
if (time_before(jiffies, ehci->next_statechange))
msleep(10);
/* Root hub was already suspended. Disable irq emission and
* mark HW unaccessible, bail out if RH has been resumed. Use
* the spinlock to properly synchronize with possible pending
* RH suspend or resume activity.
*
* This is still racy as hcd->state is manipulated outside of
* any locks =P But that will be a different fix.
*/
spin_lock_irqsave (&ehci->lock, flags);
if (hcd->state != HC_STATE_SUSPENDED) {
rc = -EINVAL;
goto bail;
}
writel (0, &ehci->regs->intr_enable);
(void)readl(&ehci->regs->intr_enable);
clear_bit(HCD_FLAG_HW_ACCESSIBLE, &hcd->flags);
bail:
spin_unlock_irqrestore (&ehci->lock, flags);
// could save FLADJ in case of Vaux power loss
// ... we'd only use it to handle clock skew
return rc;
}
static int ehci_pci_resume(struct usb_hcd *hcd)
{
struct ehci_hcd *ehci = hcd_to_ehci(hcd);
unsigned port;
struct usb_device *root = hcd->self.root_hub;
struct pci_dev *pdev = to_pci_dev(hcd->self.controller);
int retval = -EINVAL;
// maybe restore FLADJ
if (time_before(jiffies, ehci->next_statechange))
msleep(100);
/* Mark hardware accessible again as we are out of D3 state by now */
set_bit(HCD_FLAG_HW_ACCESSIBLE, &hcd->flags);
/* If CF is clear, we lost PCI Vaux power and need to restart. */
if (readl(&ehci->regs->configured_flag) != FLAG_CF)
goto restart;
/* If any port is suspended (or owned by the companion),
* we know we can/must resume the HC (and mustn't reset it).
* We just defer that to the root hub code.
*/
for (port = HCS_N_PORTS(ehci->hcs_params); port > 0; ) {
u32 status;
port--;
status = readl(&ehci->regs->port_status [port]);
if (!(status & PORT_POWER))
continue;
if (status & (PORT_SUSPEND | PORT_RESUME | PORT_OWNER)) {
usb_hcd_resume_root_hub(hcd);
return 0;
}
}
restart:
ehci_dbg(ehci, "lost power, restarting\n");
for (port = HCS_N_PORTS(ehci->hcs_params); port > 0; ) {
port--;
if (!root->children [port])
continue;
usb_set_device_state(root->children[port],
USB_STATE_NOTATTACHED);
}
/* Else reset, to cope with power loss or flush-to-storage
* style "resume" having let BIOS kick in during reboot.
*/
(void) ehci_halt(ehci);
(void) ehci_reset(ehci);
(void) ehci_pci_reinit(ehci, pdev);
/* emptying the schedule aborts any urbs */
spin_lock_irq(&ehci->lock);
if (ehci->reclaim)
ehci->reclaim_ready = 1;
ehci_work(ehci, NULL);
spin_unlock_irq(&ehci->lock);
/* restart; khubd will disconnect devices */
retval = ehci_run(hcd);
/* here we "know" root ports should always stay powered */
ehci_port_power(ehci, 1);
return retval;
}
#endif
static const struct hc_driver ehci_pci_hc_driver = {
.description = hcd_name,
.product_desc = "EHCI Host Controller",
.hcd_priv_size = sizeof(struct ehci_hcd),
/*
* generic hardware linkage
*/
.irq = ehci_irq,
.flags = HCD_MEMORY | HCD_USB2,
/*
* basic lifecycle operations
*/
.reset = ehci_pci_reset,
.start = ehci_run,
#ifdef CONFIG_PM
.suspend = ehci_pci_suspend,
.resume = ehci_pci_resume,
#endif
.stop = ehci_stop,
/*
* managing i/o requests and associated device resources
*/
.urb_enqueue = ehci_urb_enqueue,
.urb_dequeue = ehci_urb_dequeue,
.endpoint_disable = ehci_endpoint_disable,
/*
* scheduling support
*/
.get_frame_number = ehci_get_frame,
/*
* root hub support
*/
.hub_status_data = ehci_hub_status_data,
.hub_control = ehci_hub_control,
.bus_suspend = ehci_bus_suspend,
.bus_resume = ehci_bus_resume,
};
/*-------------------------------------------------------------------------*/
/* PCI driver selection metadata; PCI hotplugging uses this */
static const struct pci_device_id pci_ids [] = { {
/* handle any USB 2.0 EHCI controller */
PCI_DEVICE_CLASS(((PCI_CLASS_SERIAL_USB << 8) | 0x20), ~0),
.driver_data = (unsigned long) &ehci_pci_hc_driver,
},
{ /* end: all zeroes */ }
};
MODULE_DEVICE_TABLE(pci, pci_ids);
/* pci driver glue; this is a "new style" PCI driver module */
static struct pci_driver ehci_pci_driver = {
.name = (char *) hcd_name,
.id_table = pci_ids,
.probe = usb_hcd_pci_probe,
.remove = usb_hcd_pci_remove,
#ifdef CONFIG_PM
.suspend = usb_hcd_pci_suspend,
.resume = usb_hcd_pci_resume,
#endif
};
static int __init ehci_hcd_pci_init(void)
{
if (usb_disabled())
return -ENODEV;
pr_debug("%s: block sizes: qh %Zd qtd %Zd itd %Zd sitd %Zd\n",
hcd_name,
sizeof(struct ehci_qh), sizeof(struct ehci_qtd),
sizeof(struct ehci_itd), sizeof(struct ehci_sitd));
return pci_register_driver(&ehci_pci_driver);
}
module_init(ehci_hcd_pci_init);
static void __exit ehci_hcd_pci_cleanup(void)
{
pci_unregister_driver(&ehci_pci_driver);
}
module_exit(ehci_hcd_pci_cleanup);