linux/drivers/gpu/drm
Mukul Joshi d69fd951e6 drm/amdkfd: Fix circular locking dependency warning
[  150.887733] ======================================================
[  150.893903] WARNING: possible circular locking dependency detected
[  150.905917] ------------------------------------------------------
[  150.912129] kfdtest/4081 is trying to acquire lock:
[  150.917002] ffff8f7f3762e118 (&mm->mmap_sem#2){++++}, at:
                                 __might_fault+0x3e/0x90
[  150.924490]
               but task is already holding lock:
[  150.930320] ffff8f7f49d229e8 (&dqm->lock_hidden){+.+.}, at:
                                destroy_queue_cpsch+0x29/0x210 [amdgpu]
[  150.939432]
               which lock already depends on the new lock.

[  150.947603]
               the existing dependency chain (in reverse order) is:
[  150.955074]
               -> #3 (&dqm->lock_hidden){+.+.}:
[  150.960822]        __mutex_lock+0xa1/0x9f0
[  150.964996]        evict_process_queues_cpsch+0x22/0x120 [amdgpu]
[  150.971155]        kfd_process_evict_queues+0x3b/0xc0 [amdgpu]
[  150.977054]        kgd2kfd_quiesce_mm+0x25/0x60 [amdgpu]
[  150.982442]        amdgpu_amdkfd_evict_userptr+0x35/0x70 [amdgpu]
[  150.988615]        amdgpu_mn_invalidate_hsa+0x41/0x60 [amdgpu]
[  150.994448]        __mmu_notifier_invalidate_range_start+0xa4/0x240
[  151.000714]        copy_page_range+0xd70/0xd80
[  151.005159]        dup_mm+0x3ca/0x550
[  151.008816]        copy_process+0x1bdc/0x1c70
[  151.013183]        _do_fork+0x76/0x6c0
[  151.016929]        __x64_sys_clone+0x8c/0xb0
[  151.021201]        do_syscall_64+0x4a/0x1d0
[  151.025404]        entry_SYSCALL_64_after_hwframe+0x49/0xbe
[  151.030977]
               -> #2 (&adev->notifier_lock){+.+.}:
[  151.036993]        __mutex_lock+0xa1/0x9f0
[  151.041168]        amdgpu_mn_invalidate_hsa+0x30/0x60 [amdgpu]
[  151.047019]        __mmu_notifier_invalidate_range_start+0xa4/0x240
[  151.053277]        copy_page_range+0xd70/0xd80
[  151.057722]        dup_mm+0x3ca/0x550
[  151.061388]        copy_process+0x1bdc/0x1c70
[  151.065748]        _do_fork+0x76/0x6c0
[  151.069499]        __x64_sys_clone+0x8c/0xb0
[  151.073765]        do_syscall_64+0x4a/0x1d0
[  151.077952]        entry_SYSCALL_64_after_hwframe+0x49/0xbe
[  151.083523]
               -> #1 (mmu_notifier_invalidate_range_start){+.+.}:
[  151.090833]        change_protection+0x802/0xab0
[  151.095448]        mprotect_fixup+0x187/0x2d0
[  151.099801]        setup_arg_pages+0x124/0x250
[  151.104251]        load_elf_binary+0x3a4/0x1464
[  151.108781]        search_binary_handler+0x6c/0x210
[  151.113656]        __do_execve_file.isra.40+0x7f7/0xa50
[  151.118875]        do_execve+0x21/0x30
[  151.122632]        call_usermodehelper_exec_async+0x17e/0x190
[  151.128393]        ret_from_fork+0x24/0x30
[  151.132489]
               -> #0 (&mm->mmap_sem#2){++++}:
[  151.138064]        __lock_acquire+0x11a1/0x1490
[  151.142597]        lock_acquire+0x90/0x180
[  151.146694]        __might_fault+0x68/0x90
[  151.150879]        read_sdma_queue_counter+0x5f/0xb0 [amdgpu]
[  151.156693]        update_sdma_queue_past_activity_stats+0x3b/0x90 [amdgpu]
[  151.163725]        destroy_queue_cpsch+0x1ae/0x210 [amdgpu]
[  151.169373]        pqm_destroy_queue+0xf0/0x250 [amdgpu]
[  151.174762]        kfd_ioctl_destroy_queue+0x32/0x70 [amdgpu]
[  151.180577]        kfd_ioctl+0x223/0x400 [amdgpu]
[  151.185284]        ksys_ioctl+0x8f/0xb0
[  151.189118]        __x64_sys_ioctl+0x16/0x20
[  151.193389]        do_syscall_64+0x4a/0x1d0
[  151.197569]        entry_SYSCALL_64_after_hwframe+0x49/0xbe
[  151.203141]
               other info that might help us debug this:

[  151.211140] Chain exists of:
                 &mm->mmap_sem#2 --> &adev->notifier_lock --> &dqm->lock_hidden

[  151.222535]  Possible unsafe locking scenario:

[  151.228447]        CPU0                    CPU1
[  151.232971]        ----                    ----
[  151.237502]   lock(&dqm->lock_hidden);
[  151.241254]                                lock(&adev->notifier_lock);
[  151.247774]                                lock(&dqm->lock_hidden);
[  151.254038]   lock(&mm->mmap_sem#2);

This commit fixes the warning by ensuring get_user() is not called
while reading SDMA stats with dqm_lock held as get_user() could cause a
page fault which leads to the circular locking scenario.

Signed-off-by: Mukul Joshi <mukul.joshi@amd.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2020-07-01 01:59:27 -04:00
..
amd drm/amdkfd: Fix circular locking dependency warning 2020-07-01 01:59:27 -04:00
arc
arm drm/komeda: use devm_drm_dev_alloc 2020-04-28 16:04:00 +02:00
armada
aspeed drm/aspeed: Use devm_drm_dev_alloc 2020-04-28 16:13:03 +02:00
ast drm/ast: Don't check new mode if CRTC is being disabled 2020-05-08 09:29:18 +02:00
atmel-hlcdc
bochs drm/bochs: Remove explicit drm_connector_register 2020-04-28 16:14:17 +02:00
bridge drm/bridge: chrontel-ch7033: Add a new driver 2020-05-05 09:48:56 +02:00
etnaviv
exynos drm/exynos-vidi: convert platform driver to use dev_groups 2020-05-18 13:19:18 +09:00
fsl-dcu
gma500 drm/gma500: Remove dead code 2020-04-21 09:02:29 +02:00
hisilicon drm: kirin: Revert change to add register connect helper functions 2020-04-13 01:46:02 +00:00
i2c drm/i2c/tda998x: Make tda998x_audio_digital_mute static 2020-05-09 23:06:07 +02:00
i810
i915 Merge tag 'drm-intel-next-fixes-2020-05-20' of git://anongit.freedesktop.org/drm/drm-intel into drm-next 2020-05-21 10:44:33 +10:00
imx
ingenic drm/ingenic: Don't set drm_device->dev_private 2020-04-28 16:03:53 +02:00
lib
lima drm/lima: enable runtime pm 2020-04-24 20:51:24 +08:00
mcde drm/mcde: dsi: Fix return value check in mcde_dsi_bind() 2020-05-12 13:48:24 +02:00
mediatek Mediatek DRM Next for Linux 5.8 2020-05-22 12:20:18 +10:00
meson drm/meson: add mode selection limits against specific SoC revisions 2020-05-05 10:19:33 +02:00
mga
mgag200 drm/mgag200: Embed connector instance in struct mga_device 2020-05-11 16:40:44 +02:00
msm drm-misc-next for 5.8: 2020-04-22 10:41:35 +10:00
mxsfb
nouveau drm/nouveau: use correct conflicting framebuffer API 2020-05-26 14:41:03 +10:00
omapdrm drm/omap: change default signal polarities and drives 2020-05-05 09:58:37 +03:00
panel drm/panel-simple: Support hpd-gpios for delaying prepare() 2020-05-09 22:11:45 +02:00
panfrost
pl111 drm: pl111: Move VExpress setup into versatile init 2020-04-28 14:06:20 -05:00
qxl drm/qxl: Don't use drm_device->dev_private 2020-04-28 20:04:01 +02:00
r128
radeon drm/radeon: fix array out-of-bounds read and write issues 2020-07-01 01:59:26 -04:00
rcar-du drm: rcar-du: Set primary plane zpos immutably at initializing 2020-05-14 02:10:52 +03:00
rockchip drm/rockchip: Remove unneeded semicolon 2020-04-27 21:15:55 +02:00
savage
scheduler
selftests
shmobile
sis
sti
stm drm/stm: ltdc: check number of endpoints 2020-04-27 15:42:12 +02:00
sun4i
tdfx
tegra drm: Fix misspellings of "Analog Devices" 2020-04-21 08:15:03 +02:00
tidss drm/tidss: remove AM65x PG1 YUV erratum code 2020-05-05 10:00:07 +03:00
tilcdc drm/tilcdc: Set up fbdev after fully registering device 2020-04-08 16:24:56 +02:00
tiny drm/cirrus: Move to drm/tiny 2020-04-28 16:05:15 +02:00
ttm drm-misc-next for 5.8: 2020-04-22 10:41:35 +10:00
tve200
udl drm/udl: Make udl_handle_damage static 2020-05-06 21:08:37 +02:00
v3d drm/v3d: Delete v3d_dev->pdev 2020-04-28 15:15:59 +02:00
vboxvideo drm/vboxvideo: Use devm_gen_pool_create 2020-04-28 15:08:51 +02:00
vc4 drm-misc-next for 5.8: 2020-04-22 10:41:35 +10:00
vgem
via
virtio
vkms drm/vkms: Hold gem object while still in-use 2020-05-06 21:51:46 -04:00
vmwgfx
xen drm-misc-next for 5.8: 2020-04-22 10:41:35 +10:00
zte
drm_agpsupport.c
drm_atomic.c
drm_atomic_helper.c
drm_atomic_state_helper.c
drm_atomic_uapi.c
drm_auth.c
drm_blend.c drm: plane: Verify that no or all planes have a zpos property 2020-05-14 02:10:52 +03:00
drm_bridge.c
drm_bridge_connector.c
drm_bufs.c
drm_cache.c
drm_client.c
drm_client_modeset.c
drm_color_mgmt.c
drm_connector.c
drm_context.c
drm_crtc.c
drm_crtc_helper.c
drm_crtc_helper_internal.h
drm_crtc_internal.h
drm_damage_helper.c
drm_debugfs.c
drm_debugfs_crc.c
drm_dma.c
drm_dp_aux_dev.c
drm_dp_cec.c
drm_dp_dual_mode_helper.c
drm_dp_helper.c Short summary of fixes pull (less than what git shortlog provides): 2020-05-28 15:38:39 +10:00
drm_dp_mst_topology.c drm/dp_mst: Kill the second sideband tx slot, save the world 2020-04-27 16:18:51 -04:00
drm_dp_mst_topology_internal.h
drm_drv.c drm: Add devm_drm_dev_alloc macro 2020-04-28 15:05:19 +02:00
drm_dsc.c
drm_dumb_buffers.c
drm_edid.c drm: Nuke mode->hsync 2020-04-29 18:44:26 +03:00
drm_edid_load.c
drm_encoder.c
drm_encoder_slave.c
drm_fb_cma_helper.c drm: Fix misspellings of "Analog Devices" 2020-04-21 08:15:03 +02:00
drm_fb_helper.c drm/fb-helper: Remove return value from drm_fbdev_generic_setup() 2020-04-08 16:27:08 +02:00
drm_file.c drm: make drm_file use keyed wakeups 2020-04-28 17:10:56 +02:00
drm_flip_work.c
drm_format_helper.c
drm_fourcc.c
drm_framebuffer.c
drm_gem.c
drm_gem_cma_helper.c
drm_gem_framebuffer_helper.c drm: Don't free a struct never allocated by drm_gem_fb_init() 2020-04-16 13:44:29 +02:00
drm_gem_shmem_helper.c
drm_gem_ttm_helper.c
drm_gem_vram_helper.c drm/vram-helpers: Merge code into a single file 2020-04-09 09:56:33 +02:00
drm_hashtab.c
drm_hdcp.c
drm_internal.h
drm_ioc32.c
drm_ioctl.c
drm_irq.c
drm_kms_helper_common.c
drm_lease.c
drm_legacy.h
drm_legacy_misc.c
drm_lock.c
drm_managed.c
drm_memory.c
drm_mipi_dbi.c
drm_mipi_dsi.c
drm_mm.c drm/mm: optimize rb_hole_addr rbtree search 2020-05-05 13:39:38 +02:00
drm_mode_config.c
drm_mode_object.c drm: Replace drm_modeset_lock/unlock_all with DRM_MODESET_LOCK_ALL_* helpers 2020-05-05 10:51:56 +02:00
drm_modes.c drm: Nuke mode->hsync 2020-04-29 18:44:26 +03:00
drm_modeset_helper.c
drm_modeset_lock.c
drm_of.c
drm_panel.c
drm_panel_orientation_quirks.c
drm_pci.c
drm_plane.c drm: plane: Verify that no or all planes have a zpos property 2020-05-14 02:10:52 +03:00
drm_plane_helper.c
drm_prime.c
drm_print.c
drm_probe_helper.c
drm_property.c
drm_rect.c
drm_scatter.c
drm_scdc_helper.c
drm_self_refresh_helper.c
drm_simple_kms_helper.c
drm_syncobj.c
drm_sysfs.c
drm_trace.h
drm_trace_points.c
drm_vblank.c Short summary of fixes pull (less than what git shortlog provides): 2020-05-28 15:38:39 +10:00
drm_vm.c
drm_vma_manager.c
drm_writeback.c
Kconfig drm/cirrus: Move to drm/tiny 2020-04-28 16:05:15 +02:00
Makefile drm/cirrus: Move to drm/tiny 2020-04-28 16:05:15 +02:00