[Pixman] [PATCH 00/14] Repost of (mostly) ARMv6 patches

[Ben Avison]

I posted a number of patches earlier this year. Some were pushed, one was
disputed (as I recall, it was questioned whether the Cairo traces it was
optimising weren't better fixed in the application instead), and a number
of following ones didn't get pushed because the profiling benchmarks
contained in their commit logs had been performed with the disputed patch
in place.

It's taken a while, but I've finally got round to rebasing the remaining
patches and re-profiling them all. The results are generally very similar,
other than the fact that the two over_n_8888 patches now don't show any
measurable improvement on the Cairo profiles. However, they also don't
show any measurable regression either, and the lowlevel-blt-bench scores
indicate that they may still be useful for any applications that do
actually use them.

A number of these patches do show a significant speed improvement on
ARMv6, so I hope they will be acceptable now.

Ben Avison (14):
  ARMv6: Fix some indentation in the composite macros
  ARMv6: Minor optimisation
  ARMv6: Support for very variable-hungry composite operations
  ARMv6: Add fast path flag to force no preload of destination buffer
  ARMv6: Force fast paths to have fixed alignment to the BTAC
  Add extra test to lowlevel-blt-bench and fix an existing one
  ARMv6: Macro to permit testing for early returns or alternate
    implementations
  Add generic fast path for over_n_8888
  ARMv6: add platform-specific fast path for over_n_8888
  ARMv6: Added fast path for over_n_8888_8888_ca
  ARMv6: Add fast path for in_reverse_8888_8888
  ARMv6: Add fast path for over_reverse_n_8888
  ARMv6: Add fast path for add_8888_8888
  ARMv6: Add fast path for src_x888_0565

 pixman/pixman-arm-simd-asm.S |  632 ++++++++++++++++++++++++++++++++++++++++++
 pixman/pixman-arm-simd-asm.h |  122 ++++++--
 pixman/pixman-arm-simd.c     |   43 +++
 pixman/pixman-fast-path.c    |   41 +++
 test/lowlevel-blt-bench.c    |    3 +-
 5 files changed, 815 insertions(+), 26 deletions(-)

-- 
1.7.10.4