[PATCH 6.1] drm/amdkfd: Skip handle mapping SVM range with no GPU access
Murad Masimov
m.masimov at maxima.ru
Wed Sep 18 08:47:16 UTC 2024
From: Philip Yang <Philip.Yang at amd.com>
commit 8c45b31909b730f9c7b146588e038f9c6553394d upstream.
If the SVM range has no GPU access nor access-in-place attribute,
validate and map to GPU should skip the range.
Add NULL pointer check if find_first_bit(ctx->bitmap, MAX_GPU_INSTANCE)
returns MAX_GPU_INSTANCE as gpuidx if ctx->bitmap is empty.
Signed-off-by: Philip Yang <Philip.Yang at amd.com>
Reviewed-by: Alex Sierra <alex.sierra at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
[m.masimov at maxima.ru: In order to adapt this patch to branch 6.1
ctx was treated as a variable and not as a pointer.]
Signed-off-by: Murad Masimov <m.masimov at maxima.ru>
---
drivers/gpu/drm/amd/amdkfd/kfd_svm.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
index 7fa5e70f1aac..a44781b66af9 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
@@ -1491,6 +1491,8 @@ static void *kfd_svm_page_owner(struct kfd_process *p, int32_t gpuidx)
struct kfd_process_device *pdd;
pdd = kfd_process_device_from_gpuidx(p, gpuidx);
+ if (!pdd)
+ return NULL;
return SVM_ADEV_PGMAP_OWNER(pdd->dev->adev);
}
@@ -1561,10 +1563,10 @@ static int svm_range_validate_and_map(struct mm_struct *mm,
}
if (bitmap_empty(ctx.bitmap, MAX_GPU_INSTANCE)) {
- if (!prange->mapped_to_gpu)
- return 0;
-
bitmap_copy(ctx.bitmap, prange->bitmap_access, MAX_GPU_INSTANCE);
+ if (!prange->mapped_to_gpu ||
+ bitmap_empty(ctx.bitmap, MAX_GPU_INSTANCE))
+ return 0;
}
if (prange->actual_loc && !prange->ttm_res) {
--
2.39.2
More information about the dri-devel
mailing list