[Mesa-dev] [PATCH mesa 0/2] egl: wire up swap_with_damage() and enable it on Android

Emil Velikov emil.l.velikov at gmail.com
Wed Jun 14 15:46:16 UTC 2017


On 13 June 2017 at 07:03, Tapani Pälli <tapani.palli at intel.com> wrote:
>
>
> On 06/12/2017 04:30 PM, Eric Engestrom wrote:
>>
>> On Monday, 2017-06-12 09:39:15 +0300, Tapani Pälli wrote:
>>>
>>> Hi Eric;
>>>
>>> EGL_KHR_swap_buffers_with_damage is implemented by Android's EGL wrapper
>>> library so these tests are passing already, no implementation required in
>>> Mesa side.
>>
>>
>> Right, I keep forgetting that android intercepts some stuff and lets the
>> rest through to mesa.
>> So 2/2 is unnecessary, but should we still land 1/2?
>>
>> At worst, it's an unnecessary call to a `return false` function, and at
>> best it sets the correct damage before swapping.
>>
>> There's an alternative "worst", which would be a platform on which
>> swap_buffers_with_damage != set_damage_region + swap_buffers,
>> but on such a platform the fallback shouldn't be used anyway, so using
>> it would be the bug.
>>
>> What do you think?
>>
>
> Sure, I'm OK going ahead with patch 1 (r-b)!
>
I was going to mention "hey we'll assert in
eglSwapBuffersWithDamage{EXT,KHR} if 2/2 is missing" only to notice
that the assert is missing :-\

We might want to even drop the asserts and return early, since asserts
are omitted on release builds :-\

Mildly related: there's some subtle difference in the "surface"
meaning in EGL_KHR_swap_buffers_with_damage vs EGL_KHR_partial_update
but we could address that as a follow-up.

Tl;Dr: AFAICT we want both patches, plus some small  follow-up fixes
Reviewed-by: Emil Velikov <emil.velikov at collabora.com>

-Emil


More information about the mesa-dev mailing list