kernel/linux.git/drivers/gpu/drm/amd/amdkfd, branch v6.5.12

drm/amdkfd: Handle errors from svm validate and map

2023-11-20T10:57:01+00:00

[ Upstream commit eb3c357bcb286e89386e89302061fe717fe4e562 ] If new range is splited to multiple pranges with max_svm_range_pages alignment and added to update_list, svm validate and map should keep going after error to make sure prange->mapped_to_gpu flag is up to date for the whole range. svm validate and map update set prange->mapped_to_gpu after mapping to GPUs successfully, otherwise clear prange->mapped_to_gpu flag (for update mapping case) instead of setting error flag, we can remove the redundant error flag to simpliy code. Refactor to remove goto and update prange->mapped_to_gpu flag inside svm_range_lock, to guarant we always evict queues or unmap from GPUs if there are invalid ranges. After svm validate and map return error -EAGIN, the caller retry will update the mapping for the whole range again. Fixes: c22b04407097 ("drm/amdkfd: flag added to handle errors from svm validate and map") Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Tested-by: James Zhu Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdkfd: Remove svm range validated_once flag

2023-11-20T10:57:01+00:00

[ Upstream commit c99b16128082de519975aa147d9da3e40380de67 ] The validated_once flag is not used after the prefault was removed, The prefault was needed to ensure validate all system memory pages at least once before mapping or migrating the range to GPU. Signed-off-by: Philip Yang Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher Stable-dep-of: eb3c357bcb28 ("drm/amdkfd: Handle errors from svm validate and map") Signed-off-by: Sasha Levin

drm/amdkfd: retry after EBUSY is returned from hmm_ranges_get_pages

2023-11-20T10:57:00+00:00

[ Upstream commit ebac9414a56a5f7c336db5f5c7cc34713b649407 ] if hmm_range_get_pages returns EBUSY error during svm_range_validate_and_map, within the context of a page fault interrupt. This should retry through svm_range_restore_pages callback. Therefore we treat this as EAGAIN error instead, and defer it to restore pages fallback. Signed-off-by: Alex Sierra Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher Stable-dep-of: eb3c357bcb28 ("drm/amdkfd: Handle errors from svm validate and map") Signed-off-by: Sasha Levin

drm/amdkfd: fix some race conditions in vram buffer alloc/free of svm code

2023-11-20T10:57:00+00:00

[ Upstream commit 7bfaa160caed8192f8262c4638f552cad94bcf5a ] This patch fixes: 1: ref number of prange's svm_bo got decreased by an async call from hmm. When wait svm_bo of prange got released we shoul also wait prang->svm_bo become NULL, otherwise prange->svm_bo may be set to null after allocate new vram buffer. 2: During waiting svm_bo of prange got released in a while loop should reschedule current task to give other tasks oppotunity to run, specially the the workque task that handles svm_bo ref release, otherwise we may enter to softlock. Signed-off-by: Xiaogang.Chen Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdkfd: Use gpu_offset for user queue's wptr

2023-10-06T11:16:29+00:00

commit cc39f9ccb82426e576734b493e1777ea01b144a8 upstream. Directly use tbo's start address will miss the domain start offset. Need to use gpu_offset instead. Signed-off-by: YuBiao Wang Reviewed-by: Christian König Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman

drm/amdkfd: Checkpoint and restore queues on GFX11

2023-10-06T11:16:12+00:00

[ Upstream commit 9296da8c40900b4dae3d973aa22be306e2a77671 ] The code in kfd_mqd_manager_v11.c to support criu dump and restore of queue state was missing. Added it; should be equivalent to kfd_mqd_manager_v10.c. CC: Felix Kuehling Reviewed-by: Harish Kasiviswanathan Acked-by: Alex Deucher Signed-off-by: David Francis Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdkfd: Update CU masking for GFX 9.4.3

2023-10-06T11:16:11+00:00

[ Upstream commit fc6efed2c728c9c10b058512fc9c1613f870a8e8 ] The CU mask passed from user-space will change based on different spatial partitioning mode. As a result, update CU masking code for GFX9.4.3 to work for all partitioning modes. Signed-off-by: Mukul Joshi Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdkfd: Update cache info reporting for GFX v9.4.3

2023-10-06T11:16:11+00:00

[ Upstream commit 0752e66e91fa86fa5481b04b22053363833ffb85 ] Update cache info reporting in sysfs to report the correct number of CUs and associated cache information based on different spatial partitioning modes. Signed-off-by: Mukul Joshi Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdgpu: Store CU info from all XCCs for GFX v9.4.3

2023-10-06T11:16:11+00:00

[ Upstream commit 97e3c6a853f2af9145daf0c6ca25bcdf55c759d4 ] Currently, we store CU info only for a single XCC assuming that it is the same for all XCCs. However, that may not be true. As a result, store CU info for all XCCs. This info is later used for CU masking. Signed-off-by: Mukul Joshi Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher Signed-off-by: Sasha Levin

drm/amdkfd: Insert missing TLB flush on GFX10 and later

2023-09-23T09:14:38+00:00

commit edcfe22985d09ee8e2346c9217f5a52ab150099f upstream. Heavy-weight TLB flush is required after unmap on all GPUs for correctness and security. Signed-off-by: Harish Kasiviswanathan Reviewed-by: Felix Kuehling Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org Signed-off-by: Greg Kroah-Hartman