GitLab

Added a new shader aswizlle_prog to wired the alpha to 1 when
the image color depth is 24 (xrgb). Then we don't need to fallback
the xrgb source/mask to software composite in render phase. Also
don't wire the alpha bit to 1 in the render phase. This can get
about 2x performance gain with the cairo performance trace's
firefox-planet case.
Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>

those xcalloc/xfree/xalloc/XNFprintf/... are deprecated. Replace
then with the new one. And fix some other minor problems.

Due to the coordinate system on EGL is different from FBO
object. To support EGL surface well, we add this new feature.
When calling glamor_init from EGL ddx driver, it should use
the new flag GLAMOR_INVERTED_Y_AXIS.

This increases us from 23000 to 27000/sec on rgb24text.

This shaves CPU time in GL setup.  Performance of rgb24text went from
18400/sec to 23500/sec.

Brings x11perf -rgb24text from 230/sec to 18400/sec

They're stored just like a8, but the values are set to either 0.0 or
1.0.  Because they're a8 with only two legal values, we can't use them
as destinations, but nobody's rendering to a1 dests anyway (we hope).

Fixes (except for small bit differences) cairo source-pattern.

It's not an offset from pixmap coords to composited pixmap coords,
it's an offset from screen-relative window drawable coords to
composited pixmap coords.

There's a limitation still for RepeatNone, but this fixes a bunch of
fallbacks for gnome-terminal.

Fixes failure with rendercheck.

This doesn't yet have an optimized glamor_composite_rects()
implementation, but it does triple the speed of x11perf -aa10text.

Window dragging with metacity+gnome-terminal+xcompmgr is almost credible.

Bringup is really not flying when I can't see anything.  So dump back
to all software so I can turn on a bit at a time.

This brings in idr's glu3 code.  We'll probably want to move to linking to
it as a library, once an ABI-stable release is out.