Skip to content

cloud-kernel
cloud-kernel

openanolis / cloud-kernel
接近 2 年 前同步成功

通知 170
Star 36
Fork 7
cloud-kernel
cloud-kernel
切换分支/标签
  1. 04 2月, 2017 3 次提交
    • T
      base/memory, hotplug: fix a kernel oops in show_valid_zones() · a96dfddb
      由 Toshi Kani 提交于 
      Reading a sysfs "memoryN/valid_zones" file leads to the following oops
when the first page of a range is not backed by struct page.
show_valid_zones() assumes that 'start_pfn' is always valid for
page_zone().

 BUG: unable to handle kernel paging request at ffffea017a000000
 IP: show_valid_zones+0x6f/0x160

This issue may happen on x86-64 systems with 64GiB or more memory since
their memory block size is bumped up to 2GiB.  [1] An example of such
systems is desribed below.  0x3240000000 is only aligned by 1GiB and
this memory block starts from 0x3200000000, which is not backed by
struct page.

 BIOS-e820: [mem 0x0000003240000000-0x000000603fffffff] usable

Since test_pages_in_a_zone() already checks holes, fix this issue by
extending this function to return 'valid_start' and 'valid_end' for a
given range.  show_valid_zones() then proceeds with the valid range.

[1] 'Commit bdee237c ("x86: mm: Use 2GB memory block size on
    large-memory x86-64 systems")'

Link: http://lkml.kernel.org/r/20170127222149.30893-3-toshi.kani@hpe.comSigned-off-by: NToshi Kani <toshi.kani@hpe.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Zhang Zhen <zhenzhang.zhang@huawei.com>
Cc: Reza Arbab <arbab@linux.vnet.ibm.com>
Cc: David Rientjes <rientjes@google.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: <stable@vger.kernel.org>	[4.4+]
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>
      a96dfddb
    • H
      vhost: fix initialization for vq->is_le · cda8bba0
      由 Halil Pasic 提交于 
      Currently, under certain circumstances vhost_init_is_le does just a part
of the initialization job, and depends on vhost_reset_is_le being called
too. For this reason vhost_vq_init_access used to call vhost_reset_is_le
when vq->private_data is NULL. This is not only counter intuitive, but
also real a problem because it breaks vhost_net. The bug was introduced to
vhost_net with commit 2751c988 ("vhost: cross-endian support for
legacy devices"). The symptom is corruption of the vq's used.idx field
(virtio) after VHOST_NET_SET_BACKEND was issued as a part of the vhost
shutdown on a vq with pending descriptors.

Let us make sure the outcome of vhost_init_is_le never depend on the state
it is actually supposed to initialize, and fix virtio_net by removing the
reset from vhost_vq_init_access.

With the above, there is no reason for vhost_reset_is_le to do just half
of the job. Let us make vhost_reset_is_le reinitialize is_le.
Signed-off-by: NHalil Pasic <pasic@linux.vnet.ibm.com>
Reported-by: NMichael A. Tebolt <miket@us.ibm.com>
Reported-by: NDr. David Alan Gilbert <dgilbert@redhat.com>
Fixes: commit 2751c988 ("vhost: cross-endian support for legacy devices")
Cc: <stable@vger.kernel.org>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Reviewed-by: NGreg Kurz <groug@kaod.org>
Tested-by: NMichael A. Tebolt <miket@us.ibm.com>
      cda8bba0
    • M
      Revert "vring: Force use of DMA API for ARM-based systems with legacy devices" · 0d5415b4
      由 Michael S. Tsirkin 提交于 
      This reverts commit c7070619.

This has been shown to regress on some ARM systems:

by forcing on DMA API usage for ARM systems, we have inadvertently
kicked open a hornets' nest in terms of cache-coherency. Namely that
unless the virtio device is explicitly described as capable of coherent
DMA by firmware, the DMA APIs on ARM and other DT-based platforms will
assume it is non-coherent. This turns out to cause a big problem for the
likes of QEMU and kvmtool, which generate virtio-mmio devices in their
guest DTs but neglect to add the often-overlooked "dma-coherent"
property; as a result, we end up with the guest making non-cacheable
accesses to the vring, the host doing so cacheably, both talking past
each other and things going horribly wrong.

We are working on a safer work-around.

Fixes: c7070619 ("vring: Force use of DMA API for ARM-based systems with legacy devices")
Reported-by: NRobin Murphy <robin.murphy@arm.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NWill Deacon <will.deacon@arm.com>
Signed-off-by: NMichael S. Tsirkin <mst@redhat.com>
Acked-by: NMarc Zyngier <marc.zyngier@arm.com>
      0d5415b4
  3. 03 2月, 2017 2 次提交
  5. 02 2月, 2017 3 次提交
    • A
      efi/fdt: Avoid FDT manipulation after ExitBootServices() · c8f325a5
      由 Ard Biesheuvel 提交于 
      Some AArch64 UEFI implementations disable the MMU in ExitBootServices(),
after which unaligned accesses to RAM are no longer supported.

Commit:

  abfb7b68 ("efi/libstub/arm*: Pass latest memory map to the kernel")

fixed an issue in the memory map handling of the stub FDT code, but
inadvertently created an issue with such firmware, by moving some
of the FDT manipulation to after the invocation of ExitBootServices().

Given that the stub's libfdt implementation uses the ordinary, accelerated
string functions, which rely on hardware handling of unaligned accesses,
manipulating the FDT with the MMU off may result in alignment faults.

So fix the situation by moving the update_fdt_memmap() call into the
callback function invoked by efi_exit_boot_services() right before it
calls the ExitBootServices() UEFI service (which is arguably a better
place for it anyway)

Note that disabling the MMU in ExitBootServices() is not compliant with
the UEFI spec, and carries great risk due to the fact that switching from
cached to uncached memory accesses halfway through compiler generated code
(i.e., involving a stack) can never be done in a way that is architecturally
safe.

Fixes: abfb7b68 ("efi/libstub/arm*: Pass latest memory map to the kernel")
Signed-off-by: NArd Biesheuvel <ard.biesheuvel@linaro.org>
Tested-by: NRiku Voipio <riku.voipio@linaro.org>
Cc: <stable@vger.kernel.org>
Cc: mark.rutland@arm.com
Cc: linux-efi@vger.kernel.org
Cc: matt@codeblueprint.co.uk
Cc: leif.lindholm@linaro.org
Cc: linux-arm-kernel@lists.infradead.org
Link: http://lkml.kernel.org/r/1485971102-23330-2-git-send-email-ard.biesheuvel@linaro.orgSigned-off-by: NIngo Molnar <mingo@kernel.org>
      c8f325a5
    • A
      vfio/spapr: Fix missing mutex unlock when creating a window · 2da64d20
      由 Alexey Kardashevskiy 提交于 
      Commit d9c72894 ("vfio/spapr: Postpone default window creation")
added an additional exit to the VFIO_IOMMU_SPAPR_TCE_CREATE case and
made it possible to return from tce_iommu_ioctl() without unlocking
container->lock; this fixes the issue.

Fixes: d9c72894 ("vfio/spapr: Postpone default window creation")
Signed-off-by: NAlexey Kardashevskiy <aik@ozlabs.ru>
Reviewed-by: NDavid Gibson <david@gibson.dropbear.id.au>
Signed-off-by: NAlex Williamson <alex.williamson@redhat.com>
      2da64d20
    • I
      be2net: fix initial MAC setting · 4993b39a
      由 Ivan Vecera 提交于 
      Recent commit 34393529 ("be2net: fix MAC addr setting on privileged
BE3 VFs") allows privileged BE3 VFs to set its MAC address during
initialization. Although the initial MAC for such VFs is already
programmed by parent PF the subsequent setting performed by VF is OK,
but in certain cases (after fresh boot) this command in VF can fail.

The MAC should be initialized only when:
1) no MAC is programmed (always except BE3 VFs during first init)
2) programmed MAC is different from requested (e.g. MAC is set when
   interface is down). In this case the initial MAC programmed by PF
   needs to be deleted.

The adapter->dev_mac contains MAC address currently programmed in HW so
it should be zeroed when the MAC is deleted from HW and should not be
filled when MAC is set when interface is down in be_mac_addr_set() as
no programming is performed in this case.

Example of failure without the fix (immediately after fresh boot):

# ip link set eth0 up  <- eth0 is BE3 PF
be2net 0000:01:00.0 eth0: Link is Up

# echo 1 > /sys/class/net/eth0/device/sriov_numvfs  <- Create 1 VF
...
be2net 0000:01:04.0: Emulex OneConnect(be3): VF  port 0

# ip link set eth8 up  <- eth8 is created privileged VF
be2net 0000:01:04.0: opcode 59-1 failed:status 1-76
RTNETLINK answers: Input/output error

# echo 0 > /sys/class/net/eth0/device/sriov_numvfs  <- Delete VF
iommu: Removing device 0000:01:04.0 from group 33
...

# echo 1 > /sys/class/net/eth0/device/sriov_numvfs  <- Create it again
iommu: Removing device 0000:01:04.0 from group 33
...

# ip link set eth8 up
be2net 0000:01:04.0 eth8: Link is Up

Initialization is now OK.

v2 - Corrected the comment and condition check suggested by Suresh & Harsha

Fixes: 34393529 ("be2net: fix MAC addr setting on privileged BE3 VFs")
Cc: Sathya Perla <sathya.perla@broadcom.com>
Cc: Ajit Khaparde <ajit.khaparde@broadcom.com>
Cc: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Cc: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: NIvan Vecera <cera@cera.cz>
Acked-by: NSriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      4993b39a
  7. 01 2月, 2017 11 次提交
  9. 31 1月, 2017 15 次提交
    • M
      drm/atomic: Fix double free in drm_atomic_state_default_clear · 92c715fc
      由 Maarten Lankhorst 提交于 
      drm_atomic_helper_page_flip and drm_atomic_ioctl set their own events
in crtc_state->event. But when it's set the event is freed in 2 places.

Solve this by only freeing the event in the atomic ioctl when it
allocated its own event.

This has been broken twice. The first time when the code was introduced,
but only in the corner case when an event is allocated, but more crtc's
were included by atomic check and then failing. This can mostly
happen when you do an atomic modeset in i915 and the display clock is
changed, which forces all crtc's to be included to the state.

This has been broken worse by adding in-fences support, which caused
the double free to be done unconditionally.

[IGT] kms_rotation_crc: starting subtest primary-rotation-180
=============================================================================
BUG kmalloc-128 (Tainted: G     U         ): Object already free
-----------------------------------------------------------------------------

Disabling lock debugging due to kernel taint
INFO: Allocated in drm_atomic_helper_setup_commit+0x285/0x2f0 [drm_kms_helper] age=0 cpu=3 pid=1529
 ___slab_alloc+0x308/0x3b0
 __slab_alloc+0xd/0x20
 kmem_cache_alloc_trace+0x92/0x1c0
 drm_atomic_helper_setup_commit+0x285/0x2f0 [drm_kms_helper]
 intel_atomic_commit+0x35/0x4f0 [i915]
 drm_atomic_commit+0x46/0x50 [drm]
 drm_mode_atomic_ioctl+0x7d4/0xab0 [drm]
 drm_ioctl+0x2b3/0x490 [drm]
 do_vfs_ioctl+0x69c/0x700
 SyS_ioctl+0x4e/0x80
 entry_SYSCALL_64_fastpath+0x13/0x94
INFO: Freed in drm_event_cancel_free+0xa3/0xb0 [drm] age=0 cpu=3 pid=1529
 __slab_free+0x48/0x2e0
 kfree+0x159/0x1a0
 drm_event_cancel_free+0xa3/0xb0 [drm]
 drm_mode_atomic_ioctl+0x86d/0xab0 [drm]
 drm_ioctl+0x2b3/0x490 [drm]
 do_vfs_ioctl+0x69c/0x700
 SyS_ioctl+0x4e/0x80
 entry_SYSCALL_64_fastpath+0x13/0x94
INFO: Slab 0xffffde1f0997b080 objects=17 used=2 fp=0xffff92fb65ec2578 flags=0x200000000008101
INFO: Object 0xffff92fb65ec2578 @offset=1400 fp=0xffff92fb65ec2ae8

Redzone ffff92fb65ec2570: bb bb bb bb bb bb bb bb                          ........
Object ffff92fb65ec2578: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
Object ffff92fb65ec2588: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
Object ffff92fb65ec2598: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
Object ffff92fb65ec25a8: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
Object ffff92fb65ec25b8: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
Object ffff92fb65ec25c8: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
Object ffff92fb65ec25d8: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b  kkkkkkkkkkkkkkkk
Object ffff92fb65ec25e8: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b a5  kkkkkkkkkkkkkkk.
Redzone ffff92fb65ec25f8: bb bb bb bb bb bb bb bb                          ........
Padding ffff92fb65ec2738: 5a 5a 5a 5a 5a 5a 5a 5a                          ZZZZZZZZ
CPU: 3 PID: 180 Comm: kworker/3:2 Tainted: G    BU          4.10.0-rc6-patser+ #5039
Hardware name:                  /NUC5PPYB, BIOS PYBSWCEL.86A.0031.2015.0601.1712 06/01/2015
Workqueue: events intel_atomic_helper_free_state [i915]
Call Trace:
 dump_stack+0x4d/0x6d
 print_trailer+0x20c/0x220
 free_debug_processing+0x1c6/0x330
 ? drm_atomic_state_default_clear+0xf7/0x1c0 [drm]
 __slab_free+0x48/0x2e0
 ? drm_atomic_state_default_clear+0xf7/0x1c0 [drm]
 kfree+0x159/0x1a0
 drm_atomic_state_default_clear+0xf7/0x1c0 [drm]
 ? drm_atomic_state_clear+0x30/0x30 [drm]
 intel_atomic_state_clear+0xd/0x20 [i915]
 drm_atomic_state_clear+0x1a/0x30 [drm]
 __drm_atomic_state_free+0x13/0x60 [drm]
 intel_atomic_helper_free_state+0x5d/0x70 [i915]
 process_one_work+0x260/0x4a0
 worker_thread+0x2d1/0x4f0
 kthread+0x127/0x130
 ? process_one_work+0x4a0/0x4a0
 ? kthread_stop+0x120/0x120
 ret_from_fork+0x29/0x40
FIX kmalloc-128: Object at 0xffff92fb65ec2578 not freed

Fixes: 3b24f7d6 ("drm/atomic: Add struct drm_crtc_commit to track async updates")
Fixes: 96260142 ("drm/fence: add in-fences support")
Cc: <stable@vger.kernel.org> # v4.8+
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: NGustavo Padovan <gustavo.padovan@collabora.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1485854725-27640-1-git-send-email-maarten.lankhorst@linux.intel.com
      92c715fc
    • J
      HID: cp2112: fix gpio-callback error handling · 8e9faa15
      由 Johan Hovold 提交于 
      In case of a zero-length report, the gpio direction_input callback would
currently return success instead of an errno.

Fixes: 1ffb3c40 ("HID: cp2112: make transfer buffers DMA capable")
Cc: stable <stable@vger.kernel.org>     # 4.9
Signed-off-by: NJohan Hovold <johan@kernel.org>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>
      8e9faa15
    • J
      HID: cp2112: fix sleep-while-atomic · 7a7b5df8
      由 Johan Hovold 提交于 
      A recent commit fixing DMA-buffers on stack added a shared transfer
buffer protected by a spinlock. This is broken as the USB HID request
callbacks can sleep. Fix this up by replacing the spinlock with a mutex.

Fixes: 1ffb3c40 ("HID: cp2112: make transfer buffers DMA capable")
Cc: stable <stable@vger.kernel.org>	# 4.9
Signed-off-by: NJohan Hovold <johan@kernel.org>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NJiri Kosina <jkosina@suse.cz>
      7a7b5df8
    • G
      mmc: sdhci: Ignore unexpected CARD_INT interrupts · 161e6d44
      由 Gabriel Krisman Bertazi 提交于 
      One of our kernelCI boxes hanged at boot because a faulty eSDHC device
was triggering spurious CARD_INT interrupts for SD cards, causing CMD52
reads, which are not allowed for SD devices.  This adds a sanity check
to the interruption path, preventing that illegal command from getting
sent if the CARD_INT interruption should be disabled.

This quirk allows that particular machine to resume boot despite the
faulty hardware, instead of getting hung dealing with thousands of
mishandled interrupts.
Suggested-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NGabriel Krisman Bertazi <krisman@collabora.co.uk>
Acked-by: NAdrian Hunter <adrian.hunter@intel.com>
Signed-off-by: NUlf Hansson <ulf.hansson@linaro.org>
Cc: <stable@vger.kernel.org>
      161e6d44
    • B
      drm/nouveau/kms/nv50: request vblank events for commits that send completion events · 2b507893
      由 Ben Skeggs 提交于 
      This somehow fixes an issue where sync-to-vblank longer works correctly
after resume from suspend.

From a HW perspective, we don't need the IRQs turned on to be able to
detect flip completion, so it's assumed that this is required for the
voodoo in the core DRM vblank core.
Signed-off-by: NBen Skeggs <bskeggs@redhat.com>
      2b507893
    • I
      drm/nouveau/nv1a,nv1f/disp: fix memory clock rate retrieval · 24bf7ae3
      由 Ilia Mirkin 提交于 
      Based on the xf86-video-nv code, NFORCE (NV1A) and NFORCE2 (NV1F) have a
different way of retrieving clocks. See the
nv_hw.c:nForceUpdateArbitrationSettings function in the original code
for how these clocks were accessed.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=54587
Cc: stable@vger.kernel.org
Signed-off-by: NIlia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: NBen Skeggs <bskeggs@redhat.com>
      24bf7ae3
    • A
      drm/nouveau/disp/gt215: Fix HDA ELD handling (thus, HDMI audio) on gt215 · d347583a
      由 Alastair Bridgewater 提交于 
      Store the ELD correctly, not just enough copies of the first byte
to pad out the given ELD size.
Signed-off-by: NAlastair Bridgewater <alastair.bridgewater@gmail.com>
Fixes: 120b0c39 ("drm/nv50-/disp: audit and version SOR_HDA_ELD method")
Cc: stable@vger.kernel.org # v3.17+
Reviewed-by: NIlia Mirkin <imirkin@alum.mit.edu>
Signed-off-by: NBen Skeggs <bskeggs@redhat.com>
      d347583a
    • M
      drm/nouveau/nouveau/led: prevent compiling the led-code if nouveau=y and leds=m · d72498ca
      由 Martin Peres 提交于 
      The proper fix would have been to select LEDS_CLASS but this can lead
to a circular dependency, as found out by Arnd.

This patch implements Arnd's suggestion instead, at the cost of some
auto-magic for a fringe feature.
Reported-by: NArnd Bergmann <arnd@arndb.de>
Reported-by: Intel's 0-DAY
Fixes: 8d021d71 ("drm/nouveau/drm/nouveau: add a LED driver for the NVIDIA logo")
Signed-off-by: NMartin Peres <martin.peres@free.fr>
Signed-off-by: NBen Skeggs <bskeggs@redhat.com>
      d72498ca
    • B
      drm/nouveau/disp/mcp7x: disable dptmds workaround · 7dfee682
      由 Ben Skeggs 提交于 
      The workaround appears to cause regressions on these boards, and from
inspection of RM traces, NVIDIA don't appear to do it on them either.
Signed-off-by: NBen Skeggs <bskeggs@redhat.com>
Tested-by: NRoy Spliet <nouveau@spliet.org>
      7dfee682
    • B
      drm/nouveau: prevent userspace from deleting client object · c966b627
      由 Ben Skeggs 提交于 
      Signed-off-by: NBen Skeggs <bskeggs@redhat.com>
      c966b627
    • B
      drm/nouveau/fence/g84-: protect against concurrent access to semaphore buffers · 96692b09
      由 Ben Skeggs 提交于 
      Signed-off-by: NBen Skeggs <bskeggs@redhat.com>
      96692b09
    • D
      Drivers: hv: vmbus: finally fix hv_need_to_signal_on_read() · 433e19cf
      由 Dexuan Cui 提交于 
      Commit a389fcfd ("Drivers: hv: vmbus: Fix signaling logic in
hv_need_to_signal_on_read()")
added the proper mb(), but removed the test "prev_write_sz < pending_sz"
when making the signal decision.

As a result, the guest can signal the host unnecessarily,
and then the host can throttle the guest because the host
thinks the guest is buggy or malicious; finally the user
running stress test can perceive intermittent freeze of
the guest.

This patch brings back the test, and properly handles the
in-place consumption APIs used by NetVSC (see get_next_pkt_raw(),
put_pkt_raw() and commit_rd_index()).

Fixes: a389fcfd ("Drivers: hv: vmbus: Fix signaling logic in
hv_need_to_signal_on_read()")
Signed-off-by: NDexuan Cui <decui@microsoft.com>
Reported-by: NRolf Neugebauer <rolf.neugebauer@docker.com>
Tested-by: NRolf Neugebauer <rolf.neugebauer@docker.com>
Cc: "K. Y. Srinivasan" <kys@microsoft.com>
Cc: Haiyang Zhang <haiyangz@microsoft.com>
Cc: Stephen Hemminger <sthemmin@microsoft.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NK. Y. Srinivasan <kys@microsoft.com>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>
      433e19cf
    • C
      Input: synaptics-rmi4 - fix reversed conditions in enable/disable_irq_wake · 05e0be7c
      由 Christophe JAILLET 提交于 
      These tests are reversed.  A warning should be displayed if an error is
returned, not on success.
Signed-off-by: NChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: NBenjamin Tissoires <benjamin.tissoires@redhat.com>
Signed-off-by: NDmitry Torokhov <dmitry.torokhov@gmail.com>
      05e0be7c
    • J
      net/mlx4_core: Avoid command timeouts during VF driver device shutdown · d585df1c
      由 Jack Morgenstein 提交于 
      Some Hypervisors detach VFs from VMs by instantly causing an FLR event
to be generated for a VF.

In the mlx4 case, this will cause that VF's comm channel to be disabled
before the VM has an opportunity to invoke the VF device's "shutdown"
method.

The result is that the VF driver on the VM will experience a command
timeout during the shutdown process when the Hypervisor does not deliver
a command-completion event to the VM.

To avoid FW command timeouts on the VM when the driver's shutdown method
is invoked, we detect the absence of the VF's comm channel at the very
start of the shutdown process. If the comm-channel has already been
disabled, we cause all FW commands during the device shutdown process to
immediately return success (and thus avoid all command timeouts).
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      d585df1c
    • A
      gianfar: synchronize DMA API usage by free_skb_rx_queue w/ gfar_new_page · 4af0e5bb
      由 Arseny Solokha 提交于 
      In spite of switching to paged allocation of Rx buffers, the driver still
called dma_unmap_single() in the Rx queues tear-down path.

The DMA region unmapping code in free_skb_rx_queue() basically predates
the introduction of paged allocation to the driver. While being refactored,
it apparently hasn't reflected the change in the DMA API usage by its
counterpart gfar_new_page().

As a result, setting an interface to the DOWN state now yields the following:

  # ip link set eth2 down
  fsl-gianfar ffe24000.ethernet: DMA-API: device driver frees DMA memory with wrong function [device address=0x000000001ecd0000] [size=40]
  ------------[ cut here ]------------
  WARNING: CPU: 1 PID: 189 at lib/dma-debug.c:1123 check_unmap+0x8e0/0xa28
  CPU: 1 PID: 189 Comm: ip Tainted: G           O    4.9.5 #1
  task: dee73400 task.stack: dede2000
  NIP: c02101e8 LR: c02101e8 CTR: c0260d74
  REGS: dede3bb0 TRAP: 0700   Tainted: G           O     (4.9.5)
  MSR: 00021000 <CE,ME>  CR: 28002222  XER: 00000000

  GPR00: c02101e8 dede3c60 dee73400 000000b6 dfbd033c dfbd36c4 1f622000 dede2000
  GPR08: 00000007 c05b1634 1f622000 00000000 22002484 100a9904 00000000 00000000
  GPR16: 00000000 db4c849c 00000002 db4c8480 00000001 df142240 db4c84bc 00000000
  GPR24: c0706148 c0700000 00029000 c07552e8 c07323b4 dede3cb8 c07605e0 db535540
  NIP [c02101e8] check_unmap+0x8e0/0xa28
  LR [c02101e8] check_unmap+0x8e0/0xa28
  Call Trace:
  [dede3c60] [c02101e8] check_unmap+0x8e0/0xa28 (unreliable)
  [dede3cb0] [c02103b8] debug_dma_unmap_page+0x88/0x9c
  [dede3d30] [c02dffbc] free_skb_resources+0x2c4/0x404
  [dede3d80] [c02e39b4] gfar_close+0x24/0xc8
  [dede3da0] [c0361550] __dev_close_many+0xa0/0xf8
  [dede3dd0] [c03616f0] __dev_close+0x2c/0x4c
  [dede3df0] [c036b1b8] __dev_change_flags+0xa0/0x174
  [dede3e10] [c036b2ac] dev_change_flags+0x20/0x60
  [dede3e30] [c03e130c] devinet_ioctl+0x540/0x824
  [dede3e90] [c0347dcc] sock_ioctl+0x134/0x298
  [dede3eb0] [c0111814] do_vfs_ioctl+0xac/0x854
  [dede3f20] [c0111ffc] SyS_ioctl+0x40/0x74
  [dede3f40] [c000f290] ret_from_syscall+0x0/0x3c
  --- interrupt: c01 at 0xff45da0
      LR = 0xff45cd0
  Instruction dump:
  811d001c 7c66482e 813d0020 9061000c 807f000c 5463103a 7cc6182e 3c60c052
  386309ac 90c10008 4cc63182 4826b845 <0fe00000> 4bfffa60 3c80c052 388402c4
  ---[ end trace 695ae6d7ac1d0c47 ]---
  Mapped at:
   [<c02e22a8>] gfar_alloc_rx_buffs+0x178/0x248
   [<c02e3ef0>] startup_gfar+0x368/0x570
   [<c036aeb4>] __dev_open+0xdc/0x150
   [<c036b1b8>] __dev_change_flags+0xa0/0x174
   [<c036b2ac>] dev_change_flags+0x20/0x60

Even though the issue was discovered in 4.9 kernel, the code in question
is identical in the current net and net-next trees.

Fixes: 75354148 ("gianfar: Add paged allocation and Rx S/G")
Signed-off-by: NArseny Solokha <asolokha@kb.kras.ru>
Acked-by: NClaudiu Manoil <claudiu.manoil@nxp.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>
      4af0e5bb
  11. 30 1月, 2017 6 次提交