[PATCH 18/26] drm/i915: use new iterator in i915_gem_object_last_write_engine

[Tvrtko Ursulin]

On 22/09/2021 12:00, Christian König wrote:
> Am 22.09.21 um 12:27 schrieb Tvrtko Ursulin:
>>
>> On 22/09/2021 10:10, Christian König wrote:
>>> This is maybe even a fix since the RCU usage here looks incorrect.
>>
>> I'm afraid I gazumped you here by removing this function shortly 
>> before you posted the respin.
> 
> Is that already landed in drm-misc-next? If not just give me an Acked-by 
> and it will be fixed when merging trees together again by just dropping 
> the change.
> 
> Alternatively if it is not in drm-next I will ping the drm-misc-next 
> maintainer for a merge.

Problem is you will never pass our CI with a series which does not apply 
to drm-tip. ;)

Regards,

Tvrtko

> Regards,
> Christian.
> 
>>
>> Regards,
>>
>> Tvrtko
>>
>>> Signed-off-by: Christian König <christian.koenig at amd.com>
>>> ---
>>>   drivers/gpu/drm/i915/gem/i915_gem_object.h | 16 ++++++++--------
>>>   1 file changed, 8 insertions(+), 8 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/gem/i915_gem_object.h 
>>> b/drivers/gpu/drm/i915/gem/i915_gem_object.h
>>> index 48112b9d76df..e20efffce3a9 100644
>>> --- a/drivers/gpu/drm/i915/gem/i915_gem_object.h
>>> +++ b/drivers/gpu/drm/i915/gem/i915_gem_object.h
>>> @@ -507,16 +507,16 @@ static inline struct intel_engine_cs *
>>>   i915_gem_object_last_write_engine(struct drm_i915_gem_object *obj)
>>>   {
>>>       struct intel_engine_cs *engine = NULL;
>>> +    struct dma_resv_iter cursor;
>>>       struct dma_fence *fence;
>>>   -    rcu_read_lock();
>>> -    fence = dma_resv_get_excl_unlocked(obj->base.resv);
>>> -    rcu_read_unlock();
>>> -
>>> -    if (fence && dma_fence_is_i915(fence) && 
>>> !dma_fence_is_signaled(fence))
>>> -        engine = to_request(fence)->engine;
>>> -    dma_fence_put(fence);
>>> -
>>> +    dma_resv_iter_begin(&cursor, obj->base.resv, false);
>>> +    dma_resv_for_each_fence_unlocked(&cursor, fence) {
>>> +        if (fence && dma_fence_is_i915(fence) &&
>>> +            !dma_fence_is_signaled(fence))
>>> +            engine = to_request(fence)->engine;
>>> +    }
>>> +    dma_resv_iter_end(&cursor);
>>>       return engine;
>>>   }
>>>
>