kernel/linux.git/drivers/gpu/drm/amd/amdgpu/amdgpu_vm_sdma.c, branch v6.6.150Linux kernel stable tree (mirror)https://git.radix-linux.su/kernel/linux.git/atom?h=v6.6.1502024-08-14T11:58:44+00:00drm/amdgpu: fix potential resource leak warning2024-08-14T11:58:44+00:00Tim HuangTim.Huang@amd.com2024-04-25T03:09:00+00:00urn:sha1:bc93cfde69b7c786fcdc70463e153fdf633caff9
[ Upstream commit 22a5daaec0660dd19740c4c6608b78f38760d1e6 ]
Clear resource leak warning that when the prepare fails,
the allocated amdgpu job object will never be released.
Signed-off-by: Tim Huang <Tim.Huang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drm/amdgpu: drop the extra sign extension2023-03-22T04:47:58+00:00Alex Deucheralexander.deucher@amd.com2023-03-15T17:51:43+00:00urn:sha1:aef98f2e1bc411a27f25db81574309b7415602ca
amdgpu_bo_gpu_offset_no_check() already calls
amdgpu_gmc_sign_extend() so no need to call it twice.
Reviewed-by: Yang Wang <kevinyang.wang@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: grab extra fence reference for drm_sched_job_add_dependency2022-12-21T20:34:03+00:00Christian Königckoenig.leichtzumerken@gmail.com2022-12-19T10:47:18+00:00urn:sha1:c1c4a8b217213c1924eabf4f28385bbee9cc50c1
That function consumes the reference.
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Reported-by: Borislav Petkov (AMD) <bp@alien8.de>
Tested-by: Borislav Petkov (AMD) <bp@alien8.de>
Signed-off-by: Christian König <christian.koenig@amd.com>
Fixes: aab9cf7b6954 ("drm/amdgpu: use scheduler dependencies for VM updates")
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Merge drm/drm-next into drm-misc-next2022-11-05T15:08:36+00:00Thomas Zimmermanntzimmermann@suse.de2022-11-05T15:02:42+00:00urn:sha1:8e4e4c2f53ffcb0ef746dc3b87ce1a57c5c94c7d
Backmerging drm/drm-next to get the latest changes in the xlnx driver.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
drm/amdgpu: use scheduler dependencies for VM updates2022-11-03T11:45:20+00:00Christian Königchristian.koenig@amd.com2022-09-29T11:34:34+00:00urn:sha1:aab9cf7b6954136f4339136a1a7fc0602a2c4d8b
Instead of putting that into the job sync object.
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221014084641.128280-9-christian.koenig@amd.com
drm/amdgpu: cleanup scheduler job initialization v22022-11-03T11:45:20+00:00Christian Königchristian.koenig@amd.com2022-09-28T18:31:38+00:00urn:sha1:f7d66fb2ea43a3016e78a700a2ca6c77a74579f9
Init the DRM scheduler base class while allocating the job.
This makes the whole handling much more cleaner.
v2: fix coding style
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Luben Tuikov <luben.tuikov@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221014084641.128280-7-christian.koenig@amd.com
Merge drm/drm-fixes into drm-misc-fixes2022-10-20T07:09:00+00:00Thomas Zimmermanntzimmermann@suse.de2022-10-20T07:09:00+00:00urn:sha1:1aca5ce036e3499336d1a2ace3070f908381c055
Backmerging to get v6.1-rc1.
Signed-off-by: Thomas Zimmermann <tzimmermann@suse.de>
drm/amdgpu: use DRM_SCHED_FENCE_DONT_PIPELINE for VM updates2022-10-19T10:45:00+00:00Christian Königchristian.koenig@amd.com2022-10-07T08:59:58+00:00urn:sha1:01f2cf53844b01e691516b465df1b6ab01b03230
Make sure that we always have a CPU round trip to let the submission
code correctly decide if a TLB flush is necessary or not.
Signed-off-by: Christian König <christian.koenig@amd.com>
Bug: https://gitlab.freedesktop.org/drm/amd/-/issues/2113#note_1579296
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Acked-by: Luben Tuikov <luben.tuikov@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20221014081553.114899-2-christian.koenig@amd.com
drm/amdgpu: SDMA update use unlocked iterator2022-09-19T19:05:54+00:00Philip YangPhilip.Yang@amd.com2022-09-13T19:46:30+00:00urn:sha1:3913f0179ba366f7d7d160c506ce00de1602bbc4
SDMA update page table may be called from unlocked context, this
generate below warning. Use unlocked iterator to handle this case.
WARNING: CPU: 0 PID: 1475 at
drivers/dma-buf/dma-resv.c:483 dma_resv_iter_next
Call Trace:
dma_resv_iter_first+0x43/0xa0
amdgpu_vm_sdma_update+0x69/0x2d0 [amdgpu]
amdgpu_vm_ptes_update+0x29c/0x870 [amdgpu]
amdgpu_vm_update_range+0x2f6/0x6c0 [amdgpu]
svm_range_unmap_from_gpus+0x115/0x300 [amdgpu]
svm_range_cpu_invalidate_pagetables+0x510/0x5e0 [amdgpu]
__mmu_notifier_invalidate_range_start+0x1d3/0x230
unmap_vmas+0x140/0x150
unmap_region+0xa8/0x110
Signed-off-by: Philip Yang <Philip.Yang@amd.com>
Suggested-by: Felix Kuehling <felix.kuehling@amd.com>
Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
drm/amdgpu: use DMA_RESV_USAGE_BOOKKEEP v22022-09-13T18:33:01+00:00Christian Königchristian.koenig@amd.com2022-04-05T08:35:31+00:00urn:sha1:4247084057cffcd8f9897110d8d827d550858dc6
Use DMA_RESV_USAGE_BOOKKEEP for VM page table updates and KFD preemption fence.
v2: actually update all usages for KFD
Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>