[Mesa-dev] [PATCH 0/7] EGL_MESA_configless_context

Kristian Høgsberg krh at bitplanet.net
Wed Mar 12 14:49:09 PDT 2014

On Fri, Mar 7, 2014 at 10:05 AM, Neil Roberts <neil at linux.intel.com> wrote:
> Here is a series of patches to add an extension which makes it
> possible to create an EGL context without specifying a config. A
> context created in this way can be bound with any surface using the
> same EGLDisplay rather than being restricted to those using the same
> config. The main use case is that it would then be possible for a
> Wayland compositor to drive two displays which may have different
> bit-depths without having to wastefully create two contexts.
> Mesa seems to mostly cope with this already so the patches are mostly
> just to add the extension and remove the restriction.
> There is also a patch for Piglit to try and test the extension. I
> tried running the branch though the 'all tests' test suite and apart
> from this new test now passing there was only one difference which is
> that the GLX_OML_sync_control/swapbuffersmsc-divisor-zero test failed.
> However if I run piglit-run.py again and set a regexp so it only runs
> that one test then it passes so I'm wondering if it might just be an
> intermittent failure.
> The main thorny issue with the extension is how to handle the initial
> value of glDrawBuffer when a configless context is used. If a config
> is used then we can just say the default is GL_BACK if the config is
> double-buffered and GL_FRONT otherwise. I have taken the approach that
> this decision is made the first time the context is bound rather than
> when it is first created. There should be no way to query the value of
> glDrawBuffer until the context is first bound so it shouldn't cause
> any harm that Mesa changes the value at that point. I think this is
> worth doing for the convenience to the application which would
> otherwise have to remember to explicitly set it to GL_BACK in the
> common case that double-buffering is used.
> I've also included a couple of patches to take advantage of the
> extension in Weston.

Thanks Neil, series applied to mesa and weston.


> Regards,
> - Neil
> _______________________________________________
> mesa-dev mailing list
> mesa-dev at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/mesa-dev

More information about the wayland-devel mailing list