We can't initialize the list head on deletion as this causes the node to
point to itself, which causes an infinite loop if vmd_irq() happens to be
servicing that node.
The list initialization was trying to fix a bug from multiple calls to
disable the same IRQ. Fix this instead by having the VMD driver track if
the interrupt is enabled.
[bhelgaas: changelog, add "Fixes"]
Fixes: 97e9230635 ("x86/PCI: VMD: Initialize list item in IRQ disable")
Reported-by: Grzegorz Koczot <grzegorz.koczot@intel.com>
Tested-by: Miroslaw Drost <miroslaw.drost@intel.com>
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by Jon Derrick: <jonathan.derrick@intel.com>
The dma-mapping core and the implementations do not change the DMA
attributes passed by pointer. Thus the pointer can point to const data.
However the attributes do not have to be a bitfield. Instead unsigned
long will do fine:
1. This is just simpler. Both in terms of reading the code and setting
attributes. Instead of initializing local attributes on the stack
and passing pointer to it to dma_set_attr(), just set the bits.
2. It brings safeness and checking for const correctness because the
attributes are passed by value.
Semantic patches for this change (at least most of them):
virtual patch
virtual context
@r@
identifier f, attrs;
@@
f(...,
- struct dma_attrs *attrs
+ unsigned long attrs
, ...)
{
...
}
@@
identifier r.f;
@@
f(...,
- NULL
+ 0
)
and
// Options: --all-includes
virtual patch
virtual context
@r@
identifier f, attrs;
type t;
@@
t f(..., struct dma_attrs *attrs);
@@
identifier r.f;
@@
f(...,
- NULL
+ 0
)
Link: http://lkml.kernel.org/r/1468399300-5399-2-git-send-email-k.kozlowski@samsung.com
Signed-off-by: Krzysztof Kozlowski <k.kozlowski@samsung.com>
Acked-by: Vineet Gupta <vgupta@synopsys.com>
Acked-by: Robin Murphy <robin.murphy@arm.com>
Acked-by: Hans-Christian Noren Egtvedt <egtvedt@samfundet.no>
Acked-by: Mark Salter <msalter@redhat.com> [c6x]
Acked-by: Jesper Nilsson <jesper.nilsson@axis.com> [cris]
Acked-by: Daniel Vetter <daniel.vetter@ffwll.ch> [drm]
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com>
Acked-by: Joerg Roedel <jroedel@suse.de> [iommu]
Acked-by: Fabien Dessenne <fabien.dessenne@st.com> [bdisp]
Reviewed-by: Marek Szyprowski <m.szyprowski@samsung.com> [vb2-core]
Acked-by: David Vrabel <david.vrabel@citrix.com> [xen]
Acked-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> [xen swiotlb]
Acked-by: Joerg Roedel <jroedel@suse.de> [iommu]
Acked-by: Richard Kuo <rkuo@codeaurora.org> [hexagon]
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org> [m68k]
Acked-by: Gerald Schaefer <gerald.schaefer@de.ibm.com> [s390]
Acked-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Acked-by: Hans-Christian Noren Egtvedt <egtvedt@samfundet.no> [avr32]
Acked-by: Vineet Gupta <vgupta@synopsys.com> [arc]
Acked-by: Robin Murphy <robin.murphy@arm.com> [arm64 and dma-iommu]
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Child devices in a VMD domain that want to use MSI are slowing down MSI-X
using devices sharing the same vectors. Move all MSI usage to a single VMD
vector, and MSI-X devices can share the rest.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Jon Derrick <jonathan.derrick@intel.com>
Otherwise APIC code assumes VMD's IRQ domain can be managed by the APIC,
resulting in an invalid cast of irq_data during irq_force_complete_move().
Signed-off-by: Jon Derrick <jonathan.derrick@intel.com>
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Enabling interrupts may result in an interrupt raised and serviced while
VMD holds a lock, resulting in contention with the spin lock held while
enabling interrupts.
The solution is to disable preemption and save/restore the state during
interrupt enable and disable.
Fixes lockdep:
======================================================
[ INFO: HARDIRQ-safe -> HARDIRQ-unsafe lock order detected ]
4.6.0-2016-06-16-lockdep+ #47 Tainted: G E
------------------------------------------------------
kworker/0:1/447 [HC0[0]:SC0[0]:HE0:SE1] is trying to acquire:
(list_lock){+.+...}, at: [<ffffffffa04eb8fc>] vmd_irq_enable+0x3c/0x70 [vmd]
and this task is already holding:
(&irq_desc_lock_class){-.-...}, at: [<ffffffff810e1ff6>] __setup_irq+0xa6/0x610
which would create a new lock dependency:
(&irq_desc_lock_class){-.-...} -> (list_lock){+.+...}
but this new dependency connects a HARDIRQ-irq-safe lock:
(&irq_desc_lock_class){-.-...}
... which became HARDIRQ-irq-safe at:
[<ffffffff810c9f21>] __lock_acquire+0x981/0xe00
[<ffffffff810cb039>] lock_acquire+0x119/0x220
[<ffffffff8167294d>] _raw_spin_lock+0x3d/0x80
[<ffffffff810e36d4>] handle_level_irq+0x24/0x110
[<ffffffff8101f20a>] handle_irq+0x1a/0x30
[<ffffffff81675fc1>] do_IRQ+0x61/0x120
[<ffffffff8167404c>] ret_from_intr+0x0/0x20
[<ffffffff81672e30>] _raw_spin_unlock_irqrestore+0x40/0x60
[<ffffffff810e21ee>] __setup_irq+0x29e/0x610
[<ffffffff810e25a1>] setup_irq+0x41/0x90
[<ffffffff81f5777f>] setup_default_timer_irq+0x1e/0x20
[<ffffffff81f57798>] hpet_time_init+0x17/0x19
[<ffffffff81f5775a>] x86_late_time_init+0xa/0x11
[<ffffffff81f51e9b>] start_kernel+0x382/0x436
[<ffffffff81f51308>] x86_64_start_reservations+0x2a/0x2c
[<ffffffff81f51445>] x86_64_start_kernel+0x13b/0x14a
to a HARDIRQ-irq-unsafe lock:
(list_lock){+.+...}
... which became HARDIRQ-irq-unsafe at:
... [<ffffffff810c9d8e>] __lock_acquire+0x7ee/0xe00
[<ffffffff810cb039>] lock_acquire+0x119/0x220
[<ffffffff8167294d>] _raw_spin_lock+0x3d/0x80
[<ffffffffa04eba42>] vmd_msi_init+0x72/0x150 [vmd]
[<ffffffff810e8597>] msi_domain_alloc+0xb7/0x140
[<ffffffff810e6b10>] irq_domain_alloc_irqs_recursive+0x40/0xa0
[<ffffffff810e6cea>] __irq_domain_alloc_irqs+0x14a/0x330
[<ffffffff810e8a8c>] msi_domain_alloc_irqs+0x8c/0x1d0
[<ffffffff813ca4e3>] pci_msi_setup_msi_irqs+0x43/0x70
[<ffffffff813cada1>] pci_enable_msi_range+0x131/0x280
[<ffffffff813bf5e0>] pcie_port_device_register+0x320/0x4e0
[<ffffffff813bf9a4>] pcie_portdrv_probe+0x34/0x60
[<ffffffff813b0e85>] local_pci_probe+0x45/0xa0
[<ffffffff813b226b>] pci_device_probe+0xdb/0x130
[<ffffffff8149e3cc>] driver_probe_device+0x22c/0x440
[<ffffffff8149e774>] __device_attach_driver+0x94/0x110
[<ffffffff8149bfad>] bus_for_each_drv+0x5d/0x90
[<ffffffff8149e030>] __device_attach+0xc0/0x140
[<ffffffff8149e0c0>] device_attach+0x10/0x20
[<ffffffff813a77f7>] pci_bus_add_device+0x47/0x90
[<ffffffff813a7879>] pci_bus_add_devices+0x39/0x70
[<ffffffff813aaba7>] pci_rescan_bus+0x27/0x30
[<ffffffffa04ec1af>] vmd_probe+0x68f/0x76c [vmd]
[<ffffffff813b0e85>] local_pci_probe+0x45/0xa0
[<ffffffff81088064>] work_for_cpu_fn+0x14/0x20
[<ffffffff8108c244>] process_one_work+0x1f4/0x740
[<ffffffff8108c9c6>] worker_thread+0x236/0x4f0
[<ffffffff810935c2>] kthread+0xf2/0x110
[<ffffffff816738f2>] ret_from_fork+0x22/0x50
other info that might help us debug this:
Possible interrupt unsafe locking scenario:
CPU0 CPU1
---- ----
lock(list_lock);
local_irq_disable();
lock(&irq_desc_lock_class);
lock(list_lock);
<Interrupt>
lock(&irq_desc_lock_class);
*** DEADLOCK ***
Signed-off-by: Jon Derrick <jonathan.derrick@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Keith Busch <keith.busch@intel.com>
There is no way to know which device in a VMD triggered an interrupt
without invoking every registered driver's actions. This uses the
untracked irq handler so that a less used device does not trigger
spurious interrupt.
We have been previously recommending users to enable "noirqdebug", but do
not want to force a system setting just to keep this domain functional.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Acked-by: Bjorn Helgaas <bhelgaas@google.com>
Cc: linux-pci@vger.kernel.org
Cc: Jon Derrick <jonathan.derrick@intel.com>
Link: http://lkml.kernel.org/r/1466200821-29159-2-git-send-email-keith.busch@intel.com
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Multiple calls to disable an IRQ would have caused the driver to
dereference a poisoned list item. This re-initializes the list to allow
multiple requests to disable the IRQ.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by Jon Derrick: <jonathan.derrick@intel.com>
VMD device doesn't usually have device archdata specific dma_ops, so we
need to override the default ops for VMD devices.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by Jon Derrick: <jonathan.derrick@intel.com>
The bus always starts at 0. Due to alignment and down-casting, this
happened to work before, but looked alarmingly incorrect in kernel logs.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Comment the less obvious portion of the code for setting up memory windows,
and the platform dependency for initializing the h/w with appropriate
resources.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
The Intel Volume Management Device (VMD) is a Root Complex Integrated
Endpoint that acts as a host bridge to a secondary PCIe domain. BIOS can
reassign one or more Root Ports to appear within a VMD domain instead of
the primary domain. The immediate benefit is that additional PCIe domains
allow more than 256 buses in a system by letting bus numbers be reused
across different domains.
VMD domains do not define ACPI _SEG, so to avoid domain clashing with host
bridges defining this segment, VMD domains start at 0x10000, which is
greater than the highest possible 16-bit ACPI defined _SEG.
This driver enumerates and enables the domain using the root bus
configuration interface provided by the PCI subsystem. The driver provides
configuration space accessor functions (pci_ops), bus and memory resources,
an MSI IRQ domain with irq_chip implementation, and DMA operations
necessary to use devices through the VMD endpoint's interface.
VMD routes I/O as follows:
1) Configuration Space: BAR 0 ("CFGBAR") of VMD provides the base
address and size for configuration space register access to VMD-owned
root ports. It works similarly to MMCONFIG for extended configuration
space. Bus numbering is independent and does not conflict with the
primary domain.
2) MMIO Space: BARs 2 and 4 ("MEMBAR1" and "MEMBAR2") of VMD provide the
base address, size, and type for MMIO register access. These addresses
are not translated by VMD hardware; they are simply reservations to be
distributed to root ports' memory base/limit registers and subdivided
among devices downstream.
3) DMA: To interact appropriately with an IOMMU, the source ID DMA read
and write requests are translated to the bus-device-function of the VMD
endpoint. Otherwise, DMA operates normally without VMD-specific address
translation.
4) Interrupts: Part of VMD's BAR 4 is reserved for VMD's MSI-X Table and
PBA. MSIs from VMD domain devices and ports are remapped to appear as
if they were issued using one of VMD's MSI-X table entries. Each MSI
and MSI-X address of VMD-owned devices and ports has a special format
where the address refers to specific entries in the VMD's MSI-X table.
As with DMA, the interrupt source ID is translated to VMD's
bus-device-function.
The driver provides its own MSI and MSI-X configuration functions
specific to how MSI messages are used within the VMD domain, and
provides an irq_chip for independent IRQ allocation to relay interrupts
from VMD's interrupt handler to the appropriate device driver's handler.
5) Errors: PCIe error message are intercepted by the root ports normally
(e.g., AER), except with VMD, system errors (i.e., firmware first) are
disabled by default. AER and hotplug interrupts are translated in the
same way as endpoint interrupts.
6) VMD does not support INTx interrupts or IO ports. Devices or drivers
requiring these features should either not be placed below VMD-owned
root ports, or VMD should be disabled by BIOS for such endpoints.
[bhelgaas: add VMD BAR #defines, factor out vmd_cfg_addr(), rework VMD
resource setup, whitespace, changelog]
Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Thomas Gleixner <tglx@linutronix.de> (IRQ-related parts)
Keith Busch