On Thu, Aug 03, 2017 at 07:27:32PM +0200, Daniel Vetter wrote:
> Back when we used cs flips it made sense to go through different
> engines, since a buffer busy on an engine that we couldnt' use for
> cs flipping ended up in different paths.
>
> But with atomic we use a worker for all flips, and going through the
> combinatorial growth of engines just wastes precious machine time.
> More so the more modern the platform is.
>
> Of course gem tests should still do some diagonal testing across all
> engines, but the kms side can afford to be a bit cheaper.
>
> v2: Make it compile. Oops. Also make sure we don't break
> fast-feedback.testlist.
>
> Adding a mode where we depend upon all engines isn't really possible,
> because there's only 1 exclusive fence and that's the only thing a
> flip waits for.
>
> Cc: Chris Wilson <[email protected]>
> Cc: Maarten Lankhorst <[email protected]>
> Cc: Arkadiusz Hiler <[email protected]>
> Acked-by: Maarten Lankhorst <[email protected]>
> Signed-off-by: Daniel Vetter <[email protected]>
Reviewed-by: Arkadiusz Hiler <[email protected]>
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
