<div dir='auto'><div><br><div class="gmail_extra"><br><div class="gmail_quote">On Nov 6, 2018 13:51, "Juan A. Suarez Romero" <jasuarez@igalia.com> wrote:<br type="attribution"><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><p dir="ltr">On Tue, 2018-11-06 at 10:32 +0000, Emil Velikov wrote:
<br>
> On Fri, 2 Nov 2018 at 16:06, Erik Faye-Lund
<br>
> <erik.faye-lund@collabora.com> wrote:
<br>
> > On Fri, 2018-11-02 at 15:40 +0000, Emil Velikov wrote:
<br>
> > > On Tue, 30 Oct 2018 at 17:11, Erik Faye-Lund
<br>
> > > <erik.faye-lund@collabora.com> wrote:
<br>
> > > > EXT_shader_implicit_conversions is a useful extension that adds
<br>
> > > > implicit
<br>
> > > > conversions to OpenGL ES 3.1. Since it's tested excensively in
<br>
> > > > dEQP, and
<br>
> > > > Mesa already has support for implicit conversions, it seems
<br>
> > > > reasonable to
<br>
> > > > allow for the extension. This ended up mostly as code-cleanups
<br>
> > > > anyway.
<br>
> > > >
<br>
> > > > While enabling this, one bug was discorvered due to a failing dEQP
<br>
> > > > test
<br>
> > > > (see the last patch).
<br>
> > > >
<br>
> > > > This makes 2068 dEQP-GLE31 tests go from NotSupported to Pass on
<br>
> > > > i965.
<br>
> > > >
<br>
> > > > No piglit regressions observed.
<br>
> > > >
<br>
> > > > Erik Faye-Lund (5):
<br>
> > > > glsl: add has_implicit_conversions()-helper
<br>
> > > > glsl: add has_implicit_uint_to_int_conversion()-helper
<br>
> > > > glsl: fall back to inexact function-match
<br>
> > > > mesa/glsl: add support for EXT_shader_implicit_conversions
<br>
> > > > glsl: do not allow implicit casts of unsized array initializers
<br>
> > > >
<br>
> > > Erik, I think we'd want the hole series minus the ES extension?
<br>
> > >
<br>
> > > The issues fixed with the final patch affect GLSL 1.30-4.60 at least.
<br>
> > > Even if there's no desktop gl tests for it.
<br>
> > >
<br>
> >
<br>
> > Do you mean for the 18.3 release? If so, yes that might be good. The
<br>
> > last patch alone should also fix it, it shouldn't depend on the other
<br>
> > patches...
<br>
> >
<br>
> Technically it should be applicable for earlier series. I'll leave it
<br>
> to the guys to decide if we'd want this in 18.2 though.
<br>
>
<br>
<br>
FWIW, all the 5 patches apply cleanly in 18.2 branch, except the former, which
<br>
requires a trivial conflict resolution.
<br>
<br>
As I'm not sure if the patches are worth for 18.2, I'll wait to know if you want
<br>
them in 18.2 or not.
<br>
</p></blockquote></div></div></div><div dir="auto">As they only add missing error cases, I doubt they are worth back porting... </div><div dir="auto"><br></div><div dir="auto"><div class="gmail_extra"><div class="gmail_quote"><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><p dir="ltr"><br>
J.A.
<br>
<br>
> -Emil
<br>
> _______________________________________________
<br>
> mesa-stable mailing list
<br>
> mesa-stable@lists.freedesktop.org
<br>
> https://lists.freedesktop.org/mailman/listinfo/mesa-stable
<br>
<br>
</p>
</blockquote></div><br></div></div></div>