gpu: nvgpu: gv100: init get_preempt_timeout hal

Currently gv100 is using very large timeout for polling
eng preempt done. This will prevent eng preempt polling
loop from timing out. Also before eng preempt loop times
out, there could be channel timeout getting kicked in.
Avoid this by using gr_idle_timeout defined by driver.

Change-Id: Icff5ae37f95f58f3195b9d630bdae42c08ced9a6
Signed-off-by: Seema Khowala <seemaj@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/1701059
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

1 files changed, 1 insertions, 1 deletions

diff --git a/drivers/gpu/nvgpu/gv100/hal_gv100.c b/drivers/gpu/nvgpu/gv100/hal_gv100.c
index fbf6e046..517c1ac6 100644
--- a/drivers/gpu/nvgpu/gv100/hal_gv100.c
+++ b/drivers/gpu/nvgpu/gv100/hal_gv100.c
@@ -509,7 +509,7 @@ static const struct gpu_ops gv100_ops = {
                         gr_gv100_pg_gr_load_gating_prod,
         },
         .fifo = {
-                .get_preempt_timeout = gv100_fifo_get_preempt_timeout,
+                .get_preempt_timeout = gv11b_fifo_get_preempt_timeout,
                 .init_fifo_setup_hw = gv11b_init_fifo_setup_hw,
                 .bind_channel = channel_gm20b_bind,
                 .unbind_channel = channel_gv11b_unbind,