[Mesa-dev] [PATCH] loader/dri3: Overhaul dri3_update_num_back

[Eric Anholt]

Michel Dänzer <michel at daenzer.net> writes:

> On 20/08/16 04:42 AM, Eric Anholt wrote:
>> Michel Dänzer <michel at daenzer.net> writes:
>> 
>>> From: Michel Dänzer <michel.daenzer at amd.com>
>>>
>>> Always use 3 buffers when flipping. With only 2 buffers, we have to wait
>>> for a flip to complete (which takes non-0 time even with asynchronous
>>> flips) before we can start working on the next frame. We were previously
>>> only using 2 buffers for flipping if the X server supports asynchronous
>>> flips, even when we're not using asynchronous flips. This could result
>>> in bad performance (the referenced bug report is an extreme case, where
>>> the inter-frame stalls were preventing the GPU from reaching its maximum
>>> clocks).
>>>
>>> I couldn't measure any performance boost using 4 buffers with flipping.
>>> Performance actually seemed to go down slightly, but that might have
>>> been just noise.
>>>
>>> Without flipping, a single back buffer is enough for swap interval 0,
>>> but we need to use 2 back buffers when the swap interval is non-0,
>>> otherwise we have to wait for the swap interval to pass before we can
>>> start working on the next frame. This condition was previously reversed.
>>>
>>> Cc: "12.0 11.2" <mesa-stable at lists.freedesktop.org>
>>> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97260
>>> Signed-off-by: Michel Dänzer <michel.daenzer at amd.com>
>> 
>> Reviewed-by: Eric Anholt <eric at anholt.net>
>
> Thanks.
>
> Note that four piglit EGL tests fail with this change, but AFAICT all of
> those tests are broken. I submitted fixes for three of them:
>
> https://patchwork.freedesktop.org/patch/106809/
> https://patchwork.freedesktop.org/patch/106807/
> https://patchwork.freedesktop.org/patch/106808/
>
> The last one is egl_chromium_sync_control. It calls eglSwapBuffers twice
> in a row and expects that the SBC is higher than before immediately
> after that. I.e. it expects that there is only one back buffer, so that
> at least the second eglSwapBuffers has to block until the first swap
> finishes. Which is no longer true with this DRI3 change. Not sure what
> to do about this one.

From a quick glance, change the test to wait on the SBC before querying
the new one, maybe?  It does seem questionable to assume that a swap
waits on the previous swap.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 800 bytes
Desc: not available
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20160823/32e7a4d4/attachment.sig>