nv-tegra-mirror/linux-nvgpu
1
0
Fork 0
mirror of git://nv-tegra.nvidia.com/linux-nvgpu.git synced 2025-12-22 17:36:20 +03:00
Code Issues Packages Projects Releases Wiki Activity

gpu: nvgpu: fix for worker thread unit test

Browse Source 
Fix a possible issue which could arise due to invoking the
function nvgpu_worker_init twice on same worker thread. A worker
poll task thread could be already using the mutex/condition variable
when the second init invocation is done, leading to an undefined
behaviour. Second invocation is done for branch coverage, which
can be done using a different worker with thread running status set.

Bug 3180383

Change-Id: Ife2917e349aa039a4e0baa399c696a7c9ab3015e
Signed-off-by: ajeshkv <akv@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2453218
Reviewed-by: automaticguardword <automaticguardword@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: Sagar Kamble <skamble@nvidia.com>
Reviewed-by: Philip Elcan <pelcan@nvidia.com>
Reviewed-by: Vaibhav Kachore <vkachore@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This commit is contained in:
ajeshkv
2020-12-01 16:34:21 +05:30
committed by Alex Waterman
parent bfa712ab3d
commit 2435b93cb8
1 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
userspace/units/interface/worker/worker.c
View File

@@ -95,6 +95,7 @@ static u32 wakeup_timeout(struct nvgpu_worker *worker)
}
_Thread_local struct nvgpu_worker worker;
_Thread_local struct nvgpu_worker worker_branch;
_Thread_local struct nvgpu_worker_ops worker_ops = {
/* pre_process is NULL for branch testing for NULL when thread starts. */
.pre_process = NULL,
@@ -136,7 +137,9 @@ int test_init(struct unit_module *m, struct gk20a *g, void *args)
while (!nvgpu_thread_is_running(&worker.poll_task)) {
nvgpu_udelay(5);
}
err = nvgpu_worker_init(g, &worker, &worker_ops);
nvgpu_atomic_set(&worker_branch.poll_task.running, 1);
err = nvgpu_worker_init(g, &worker_branch, &worker_ops);
unit_assert(err == 0, return UNIT_FAIL);
return UNIT_SUCCESS;