drm/amdgpu: Initialize fences array entries in amdgpu_sa_bo_next_hole

The entries were only initialized once in amdgpu_sa_bo_new. If a fence
wasn't signalled yet in the first amdgpu_sa_bo_next_hole call, but then
got signalled before a later amdgpu_sa_bo_next_hole call, it could
destroy the fence but leave its pointer in the array, resulting in
use-after-free in amdgpu_sa_bo_new.

Reviewed-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This commit is contained in:
Michel Dänzer 2018-09-12 18:07:10 +02:00 committed by Alex Deucher
parent 520cbe0f4a
commit 6a15f3ff19

View file

@ -226,6 +226,8 @@ static bool amdgpu_sa_bo_next_hole(struct amdgpu_sa_manager *sa_manager,
for (i = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i) {
struct amdgpu_sa_bo *sa_bo;
fences[i] = NULL;
if (list_empty(&sa_manager->flist[i]))
continue;
@ -296,10 +298,8 @@ int amdgpu_sa_bo_new(struct amdgpu_sa_manager *sa_manager,
spin_lock(&sa_manager->wq.lock);
do {
for (i = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i) {
fences[i] = NULL;
for (i = 0; i < AMDGPU_SA_NUM_FENCE_LISTS; ++i)
tries[i] = 0;
}
do {
amdgpu_sa_bo_try_free(sa_manager);