On 2022-10-17 21:27, Rafael Mendonca wrote:
If the number of pages from the userptr BO differs from the SG BO then the allocated memory for the SG table doesn't get freed before returning -EINVAL, which may lead to a memory leak in some error paths. Fix this by checking the number of pages before allocating memory for the SG table.Fixes: 264fb4d332f5 ("drm/amdgpu: Add multi-GPU DMA mapping helpers") Signed-off-by: Rafael Mendonca <rafaelmen...@gmail.com>
Thank you for catching this. The patch is Reviewed-by: Felix Kuehling <felix.kuehl...@amd.com> I submitted it to our amd-staging-drm-next branch. Regards, Felix
--- drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c index 978d3970b5cc..84f44f7e4111 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c @@ -510,13 +510,13 @@ kfd_mem_dmamap_userptr(struct kgd_mem *mem, struct ttm_tt *ttm = bo->tbo.ttm; int ret;+ if (WARN_ON(ttm->num_pages != src_ttm->num_pages))+ return -EINVAL; + ttm->sg = kmalloc(sizeof(*ttm->sg), GFP_KERNEL); if (unlikely(!ttm->sg)) return -ENOMEM;- if (WARN_ON(ttm->num_pages != src_ttm->num_pages))- return -EINVAL; - /* Same sequence as in amdgpu_ttm_tt_pin_userptr */ ret = sg_alloc_table_from_pages(ttm->sg, src_ttm->pages, ttm->num_pages, 0,
