diff options
| author | Qu Huang <jinsdb@126.com> | 2021-01-28 15:14:25 +0300 |
|---|---|---|
| committer | Alex Deucher <alexander.deucher@amd.com> | 2021-04-09 23:47:06 +0300 |
| commit | b010affea45d812d8d386cc49c3b2bafd74b4154 (patch) | |
| tree | 2d933d8e8411e93eedf740ebcb2c761cf08cf520 /drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h | |
| parent | 84e070f58ad22879787602bb18df656c326f6492 (diff) | |
| download | linux-b010affea45d812d8d386cc49c3b2bafd74b4154.tar.xz | |
drm/amdkfd: dqm fence memory corruption
Amdgpu driver uses 4-byte data type as DQM fence memory,
and transmits GPU address of fence memory to microcode
through query status PM4 message. However, query status
PM4 message definition and microcode processing are all
processed according to 8 bytes. Fence memory only allocates
4 bytes of memory, but microcode does write 8 bytes of memory,
so there is a memory corruption.
Changes since v1:
* Change dqm->fence_addr as a u64 pointer to fix this issue,
also fix up query_status and amdkfd_fence_wait_timeout function
uses 64 bit fence value to make them consistent.
Signed-off-by: Qu Huang <jinsdb@126.com>
Reviewed-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Diffstat (limited to 'drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h')
| -rw-r--r-- | drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h index dd0d8834b432..71e2fde56b2b 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.h @@ -187,7 +187,7 @@ struct device_queue_manager { uint16_t vmid_pasid[VMID_NUM]; uint64_t pipelines_addr; uint64_t fence_gpu_addr; - unsigned int *fence_addr; + uint64_t *fence_addr; struct kfd_mem_obj *fence_mem; bool active_runlist; int sched_policy; |