nv-tegra-mirror/linux-nvgpu
1
0
Fork 0
mirror of git://nv-tegra.nvidia.com/linux-nvgpu.git synced 2025-12-23 18:16:01 +03:00
Code Issues Packages Projects Releases Wiki Activity

gpu: nvgpu: skip taking g->busy_lock in gk20a_idle

Browse Source 
We use g->busy_lock in gk20a_do_idle() to prevent submitting
more jobs to h/w and to wait for currently running jobs to
finish

But requesting this lock in gk20a_idle() prevents decrementing
runtime counter and hence gk20a_do_idle() can timeout with
below prints
[  148.904739] gk20a 17000000.gp10b: Timeout detected @
gk20a_do_idle+0x30/0x38
[  148.912185] gk20a 17000000.gp10b: __gk20a_do_idle: failed to idle -
refcount 4 != 1

Hence skip requesting this lock in gk20a_idle()

Bug 200294536

Change-Id: I060075fdee1b68e1b5fa11baa44a3f5ce4917d94
Signed-off-by: Deepak Nibade <dnibade@nvidia.com>
Reviewed-on: http://git-master/r/1480756
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
This commit is contained in:
Deepak Nibade
2017-05-12 16:28:47 +05:30
committed by mobile promotions
parent 6070d99a50
commit 326ed94d7a
1 changed files with 1 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
drivers/gpu/nvgpu/common/linux/module.c
View File

@@ -109,12 +109,11 @@ void gk20a_idle(struct gk20a *g)
struct device *dev; struct device *dev;
atomic_dec(&g->usage_count); atomic_dec(&g->usage_count);
down_read(&g->busy_lock);
dev = g->dev; dev = g->dev;
if (!(dev && gk20a_can_busy(g))) if (!(dev && gk20a_can_busy(g)))
goto fail; return;
if (pm_runtime_enabled(dev)) { if (pm_runtime_enabled(dev)) {
#ifdef CONFIG_PM #ifdef CONFIG_PM
@@ -128,8 +127,6 @@ void gk20a_idle(struct gk20a *g)
} else { } else {
gk20a_scale_notify_idle(dev); gk20a_scale_notify_idle(dev);
} }
fail:
up_read(&g->busy_lock);
} }
int gk20a_pm_finalize_poweron(struct device *dev) int gk20a_pm_finalize_poweron(struct device *dev)