[Mesa-dev] [RFC] New gallium flags for using different contexts in several threads

[Axel Davy]

On 18/12/2016 18:34, Nicolai Hähnle wrote:
>
>
> Then again, why not just call flush unconditionally? If the flush was 
> unnecessary, it should be a no-op, and the driver should already have 
> a fast path for that anyway.

I just looked at radeon source with amdgpu, and it looks like to me a flush
will wait:
1) current commands in the context are submitted to the thread 
submitting the cs to the context
2) Wait the submitting thread has finished all its work.

While 1) would be noop, 2) wouldn't be in my understanding, because it 
would wait on work done in other contexts.

Axel