Merge lp://staging/~robertcarr/mir/nested-dont-require-opaque-surface into lp://staging/mir

PASSED: Continuous integration, rev:1245
http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-ci/411/
Executed test runs:
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-android-trusty-i386-build/247
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-clang-trusty-amd64-build/244
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-amd64-ci/140
        deb: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-amd64-ci/140/artifact/work/output/*zip*/output.zip
    SUCCESS: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-armhf-ci/140
        deb: http://jenkins.qa.ubuntu.com/job/mir-team-mir-development-branch-trusty-armhf-ci/140/artifact/work/output/*zip*/output.zip

Click here to trigger a rebuild:
http://s-jenkins.ubuntu-ci:8080/job/mir-team-mir-development-branch-ci/411/rebuild

Tested and it works for me.

i'm still wondering about why we had nested selecting an opaque pixel format in the first place... is there something in the compositor that needs to change too?

I think your assertion is out of date. Nexus4 acquired an opaque pixel format recently in r1229 which resolved bug 1240833.

As such, I don't think this is required any more (?)

I don't think opaque is a hard requirement - but is desirable for a desktop session (as we'd like to enable bypass).

The current code is a simplistic POC that got nested running. (Although we're still waiting for some supporting elements - e.g. in Mesa - to land.)

I don't like this approach but, if it is actually useful, don't really mind it landing either.

Going forward I'd prefer to see a mechanism by which the client code can identify an acceptable format for frame buffers.

> I think your assertion is out of date. Nexus4 acquired an opaque pixel format recently in r1229 which resolved bug 1240833.
>
> As such, I don't think this is required any more (?)

Oh, hm. I bet you are right. I reported this problem to Robert based on libmirserver10, which didn't include that fix. Looks like that fix will land in libmirserver11.

> > I think your assertion is out of date. Nexus4 acquired an opaque pixel
> > format recently in r1229 which resolved bug 1240833.
> >
> > As such, I don't think this is required any more (?)
>
> Oh, hm. I bet you are right. I reported this problem to Robert based on
> libmirserver10, which didn't include that fix. Looks like that fix will land
> in libmirserver11.

libmirserver11 is based on r1192 (which is somewhat earlier)

Reviewing the change again, in context, it actually looks almost harmless despite not really being required any more.

Although if it were to land then we might have to fix this too:
    EGL_ALPHA_SIZE, 0,