[PATCH v2 2/2] x11 backend: add option to use pixman renderer

[Vasily Khoruzhick]

On Mon, Jan 7, 2013 at 9:49 PM, Kristian Høgsberg <hoegsberg at gmail.com> wrote:
> On Mon, Jan 07, 2013 at 08:39:50PM +0300, Vasily Khoruzhick wrote:
>> When --use-shm is passed to weston and x11 backend is active,
>> it will use SHM surfaces with pixman renderer instead of EGL
>>
>> Signed-off-by: Vasily Khoruzhick <anarsoul at gmail.com>
>> ---
>> v2: - handle missing MIT-SHM extension gracefully (tested with Xnest)
>>     - follow libc convention of error handling
>>     - get rid of xcb-aux dependency
>
> Looks good, committed.  Is there a reason you don't just use
> visual_type->depth instead of get_depth_of_visual()?

There's no depth field in xcb_visualtype_t (libxcb-1.9)

> Maybe I'm
> missing something.  Also, to test the pixman renderer with desktop
> shell, I have to comment out
>
> #define HAVE_CAIRO_EGL 1
>
> in config.h, since we don't have a configure option to disable that.

Probably cairo in archlinux is compiled without EGL support, that's
why it works for me.

> If you feel like making a patch to add --disable-egl or such, that
> would be great.

Sure, will add it.

> And of course, it would be nice to make the
> pixman-renderer work with the kms backend.

My target is fbdev backend, but probably I can try to implement KMS too.
Hope to release some code during this week.

> Anyway, thanks a lot, I really appreciate this work.  It's a great
> validation of the weston_renderer infrastructure and of course it's
> very cool that we can now run weston without hw acceleration without
> having to rely on sw mesa.

No pb, and thanks for answering all my questions about weston and gl
renderer on IRC! It would
take more time without your and pq's help.

Regards
Vasily