[Mesa-dev] change 396cbab, configs with accumulation buffer

[Kenneth Graunke]

On Monday, February 22, 2016 12:25:45 PM PST Tapani Pälli wrote:
> 
> On 02/22/2016 11:41 AM, Kenneth Graunke wrote:
> > On Monday, February 22, 2016 8:06:35 AM PST Tapani Pälli wrote:
> >> Hi Marek;
> >>
> >> Was this commit fixing some issues/problems? Why would we not expose
> >> configs with accumulation buffer?
> >>
> >> --- 8< ---
> >> commit 396cbabbefaae64deac6d33c79898bb07db8a621
> >> Author: Marek Olšák <marek.olsak at amd.com>
> >> Date:   Thu Apr 9 23:25:07 2015 +0200
> >>
> >>       egl/dri: don't expose configs with an accumulation buffer
> >> --- 8< ---
> >>
> >> Thanks;
> >>
> >> // Tapani
> >
> > Assuming it's legal not to expose them, it would certainly be nice.
> > Accumulation buffers only exist in legacy GL (neither core profile nor
> > GLES have ever had them), and our GLX implementation has marked them
> > as "SLOW" for at least 12 years.
> 
> Right, I guess the only issue here is that even if you happen to have 
> such buffer (capability?) in the visual it does not necessarily mean 
> that it is in use. This change will remove all such configs, I'm not 
> sure if that is generally a problem though, I guess it is up to the 
> driver what it wants to expose.
> 
> I'm asking this because this patch causes a problem in internal project 
> where they simply revert this patch currently, I will query more but I 
> don't think they use such legacy feature.

I believe that Intel project has a managerial requirement to be passing
some percentage of an internal test suite.  That suite includes tests
for legacy functionality, so enabling old fairly useless features which
have a lot of tests can improve their pass rate.

If I were to hazard a guess, this is just them gaming the system: add
useless features that have a lot of passing tests, boost the percentage,
look good for the boss without doing much work.  I've seen it happen
before.

Of course, they could have actually found a use for this, too.
If they actually participated in an open source development
process, we might be able to find out...

--Ken
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: This is a digitally signed message part.
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20160222/19dccb0b/attachment.sig>