kernel/linux.git/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c, branch v6.19.12

drm/amdgpu: Fix fence put before wait in amdgpu_amdkfd_submit_ib

2026-04-02T11:25:36+00:00

[ Upstream commit 7150850146ebfa4ca998f653f264b8df6f7f85be ] amdgpu_amdkfd_submit_ib() submits a GPU job and gets a fence from amdgpu_ib_schedule(). This fence is used to wait for job completion. Currently, the code drops the fence reference using dma_fence_put() before calling dma_fence_wait(). If dma_fence_put() releases the last reference, the fence may be freed before dma_fence_wait() is called. This can lead to a use-after-free. Fix this by waiting on the fence first and releasing the reference only after dma_fence_wait() completes. Fixes the below: drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c:697 amdgpu_amdkfd_submit_ib() warn: passing freed memory 'f' (line 696) Fixes: 9ae55f030dc5 ("drm/amdgpu: Follow up change to previous drm scheduler change.") Cc: Felix Kuehling Cc: Dan Carpenter Cc: Christian König Cc: Alex Deucher Signed-off-by: Srinivasan Shanmugam Reviewed-by: Christian König Signed-off-by: Alex Deucher (cherry picked from commit 8b9e5259adc385b61a6590a13b82ae0ac2bd3482) Signed-off-by: Sasha Levin

drm/amdkfd: add proper handling for S0ix

2025-09-18T18:59:24+00:00

When in S0i3, the GFX state is retained, so all we need to do is stop the runlist so GFX can enter gfxoff. Reviewed-by: Mario Limonciello (AMD) Tested-by: David Perry Signed-off-by: Alex Deucher (cherry picked from commit 4bfa8609934dbf39bbe6e75b4f971469384b50b1) Cc: stable@vger.kernel.org

Merge tag 'amd-drm-next-6.17-2025-07-01' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

2025-07-04T00:06:29+00:00

amd-drm-next-6.17-2025-07-01: amdgpu: - FAMS2 fixes - OLED fixes - Misc cleanups - AUX fixes - DMCUB updates - SR-IOV hibernation support - RAS updates - DP tunneling fixes - DML2 fixes - Backlight improvements - Suspend improvements - Use scaling for non-native modes on eDP - SDMA 4.4.x fixes - PCIe DPM fixes - SDMA 5.x fixes - Cleaner shader updates for GC 9.x - Remove fence slab - ISP genpd support - Parition handling rework - SDMA FW checks for userq support - Add missing firmware declaration - Fix leak in amdgpu_ctx_mgr_entity_fini() - Freesync fix - Ring reset refactoring - Legacy dpm verbosity changes amdkfd: - GWS fix - mtype fix for ext coherent system memory - MMU notifier fix - gfx7/8 fix radeon: - CS validation support for additional GL extensions - Bump driver version for new CS validation checks From: Alex Deucher Link: https://lore.kernel.org/r/20250701194707.32905-1-alexander.deucher@amd.com Signed-off-by: Dave Airlie

drm/amdkfd: Move the process suspend and resume out of full access

2025-06-18T16:19:19+00:00

For the suspend and resume process, exclusive access is not required. Therefore, it can be moved out of the full access section to reduce the duration of exclusive access. v3: Move suspend processes before hardware fini. Remove twice call for bare metal. v4: Refine code Signed-off-by: Emily Deng Acked-by: Lijo Lazar Signed-off-by: Alex Deucher

drm/amdkfd: allow compute partition mode switch with cgroup exclusions

2025-06-18T16:19:17+00:00

The KFD currently bars a compute partition mode switch while a KFD process exists. Since cgroup excluded devices remain excluded for the lifetime of a KFD process and user space is able to mode switch single devices, allow users to mode switch a device with any running process that has been cgroup excluded from this device. Signed-off-by: Jonathan Kim Reviewed-by: Harish Kasiviswanathan Signed-off-by: Alex Deucher

Merge drm/drm-next into drm-misc-next

2025-06-11T07:01:34+00:00

Backmerging to forward to v6.16-rc1 Signed-off-by: Thomas Zimmermann

drm/amdkfd: Map wptr BO to GART unconditionally

2025-05-29T14:58:44+00:00

For simulation C models that don't run CP FW where adev->mes.sched_version is not populated correctly. This causes NULL dereference in amdgpu_amdkfd_free_gtt_mem(dev->adev, (void **)&pqn->q->wptr_bo_gart) and warning on unpinned BO in amdgpu_bo_gpu_offset(q->properties.wptr_bo). Compared with adding version check here and there, always map wptr BO to GART simplifies things. v2: Add NULL check in amdgpu_amdkfd_free_gtt_mem.(Philip) Signed-off-by: Lang Yu Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher

drm/sched: Store the drm client_id in drm_sched_fence

2025-05-28T14:15:58+00:00

This will be used in a later commit to trace the drm client_id in some of the gpu_scheduler trace events. This requires changing all the users of drm_sched_job_init to add an extra parameter. The newly added drm_client_id field in the drm_sched_fence is a bit of a duplicate of the owner one. One suggestion I received was to merge those 2 fields - this can't be done right now as amdgpu uses some special values (AMDGPU_FENCE_OWNER_*) that can't really be translated into a client id. Christian is working on getting rid of those; when it's done we should be able to squash owner/drm_client_id together. Reviewed-by: Christian König Signed-off-by: Pierre-Eric Pelloux-Prayer Signed-off-by: Philipp Stanner Link: https://lore.kernel.org/r/20250526125505.2360-3-pierre-eric.pelloux-prayer@amd.com

drm/amdgpu: simplify xgmi peer info calls

2025-02-25T16:45:12+00:00

Deprecate KFD XGMI peer info calls in favour of calling directly from simplified XGMI peer info functions. Signed-off-by: Jonathan Kim Reviewed-by: Lijo Lazar Signed-off-by: Alex Deucher

drm/amdkfd: add a new flag to manage where VRAM allocations go

2025-02-13T02:04:08+00:00

On big and small APUs we send KFD VRAM allocations to GTT since the carve out is either non-existent or relatively small. However, if someone sets the carve out size to be relatively large, we may end up using GTT rather than VRAM. No change of logic with this patch, but it allows the driver to determine which logic to use based on the carve out size in the future. Reviewed-by: Mario Limonciello Signed-off-by: Alex Deucher