gpu: nvgpu: NVGPU abstraction for ACCESS_ONCE

Construct a wrapper macro NV_ACCESS_ONCE(x) which uses OS specific
versions of ACCESS_ONCE. e.g for linux, ACCESS_ONCE(x) is used.

Jira NVGPU-125

Change-Id: Ia5c67baae111c1a7978c530bf279715fc808287d
Signed-off-by: Debarshi Dutta <ddutta@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1549928
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
Reviewed-by: svccoveritychecker <svccoveritychecker@nvidia.com>
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>

1 files changed, 1 insertions, 1 deletions

diff --git a/drivers/gpu/nvgpu/common/pmu/pmu.c b/drivers/gpu/nvgpu/common/pmu/pmu.c
index 63597d10..89216ea2 100644
--- a/drivers/gpu/nvgpu/common/pmu/pmu.c
+++ b/drivers/gpu/nvgpu/common/pmu/pmu.c
@@ -413,7 +413,7 @@ static int nvgpu_pg_init_task(void *arg)
                         (pg_init->state_change == true), 0);
 
                 pmu->pg_init.state_change = false;
-                pmu_state = ACCESS_ONCE(pmu->pmu_state);
+                pmu_state = NV_ACCESS_ONCE(pmu->pmu_state);
 
                 if (pmu_state == PMU_STATE_EXIT) {
                         nvgpu_pmu_dbg(g, "pmu state exit");