gpu: nvgpu: gv11b: Load FECS via DMA controller.

ctxsw ucode loading time is reduced to ~3ms from ~9ms
if we use DMA controller to load the FECS falcon
instead of using PRI writes.

Bug 2400729

Change-Id: I0e5758c857cf76cde9abb460c8915e15e41b03d2
Signed-off-by: Deepak <dgoyal@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1846819
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: svc-misra-checker <svc-misra-checker@nvidia.com>
Reviewed-by: svc-mobile-coverity <svc-mobile-coverity@nvidia.com>
GVS: Gerrit_Virtual_Submit
Reviewed-by: Seshendra Gadagottu <sgadagottu@nvidia.com>
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This commit is contained in:
Deepak
2018-09-26 10:33:25 +05:30
committed by Abdul Salam
parent 180e8be77c
commit 7a7666a8dd

View File

@@ -167,7 +167,7 @@ bool gv11b_is_priv_load(u32 falcon_id)
switch (falcon_id) { switch (falcon_id) {
case LSF_FALCON_ID_FECS: case LSF_FALCON_ID_FECS:
enable_status = true; enable_status = false;
break; break;
case LSF_FALCON_ID_GPCCS: case LSF_FALCON_ID_GPCCS:
enable_status = true; enable_status = true;