[Intel-gfx] [RFC 4/4] drm/i915: Remove trylock-looping from the shrinker

[Tvrtko Ursulin]

On 03/01/2019 12:56, Chris Wilson wrote:
> Quoting Tvrtko Ursulin (2019-01-03 12:23:29)
>> From: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
>>
>> In a patch named "drm/i915: Return immediately if trylock fails for
>> direct-reclaim" Chris Wilson implicitly suggests it is safe to use a plain
>> mutex_lock instead of trylock on the I915_SHRINK_ACTIVE code paths.
>>
>> Explore this to see how the lockdep traces will look.
>>
>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin at intel.com>
>> Cc: Chris Wilson <chris at chris-wilson.co.uk>
>> ---
>>   drivers/gpu/drm/i915/i915_gem_shrinker.c | 11 ++---------
>>   1 file changed, 2 insertions(+), 9 deletions(-)
>>
>> diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c
>> index 97deec775f40..2d4a54851c0a 100644
>> --- a/drivers/gpu/drm/i915/i915_gem_shrinker.c
>> +++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c
>> @@ -47,15 +47,8 @@ shrinker_lock(struct drm_i915_private *i915, unsigned int flags, bool *unlock)
>>          case MUTEX_TRYLOCK_FAILED:
>>                  *unlock = false;
>>                  if (flags & I915_SHRINK_ACTIVE) {
>> -                       preempt_disable();
>> -                       do {
>> -                               cpu_relax();
>> -                               if (mutex_trylock(&i915->drm.struct_mutex)) {
>> -                                       *unlock = true;
>> -                                       break;
>> -                               }
>> -                       } while (!need_resched());
>> -                       preempt_enable();
>> +                       mutex_lock(&i915->drm.struct_mutex);
> 
> You seemed to have lost some annotation here.

I deliberately omitted mutex_lock_nested from this series to avoid 
hiding any potential lockdep issues. If there won't be any hit then we 
really need to improve on the test coverage.

Regards,

Tvrtko