Ben Avison (7):
Tweaks to lowlevel-blt-bench
ARMv6: Lay the groundwork for later patches in the series
ARMv6: New fill routines
ARMv6: New blit routines
ARMv6: New conversion routines
ARMv6: Replacement add_8_8, over_8888_8888, over_8888_n_8888 and over_n_8_8888 routines
Improve L1 and L2 benchmark tests for caches that don't use allocate-on-write
Benjamin Gilbert (1):
Fix thread safety on mingw-w64 and clang
Chris Wilson (6):
sse2: Add a fast path for add_n_8888
sse2: Add a fast path for add_n_8_8888
sse2: Add fast paths for bilinear source with a solid mask
Always return a valid function from lookup_composite()
Always return a valid function from lookup_combiner()
Eliminate duplicate copies of channel flags for pixman_image_composite32()
Jeff Muizelaar (1):
Add a version of bilinear_interpolation for precision <=4
Joshua Root (1):
Fix undeclared variable use and sysctlbyname error handling on ppc
Marko Lindqvist (1):
Fix build with automake-1.13
Matt Turner (2):
Add new demos and tests to .gitignore
Convert INCLUDES to AM_CPPFLAGS
Nemanja Lukic (5):
MIPS: DSPr2: Added more fast-paths for SRC operation:
MIPS: DSPr2: Added more fast-paths for OVER operation:
MIPS: DSPr2: Added several nearest neighbor fast paths with a8 mask:
MIPS: DSPr2: Added more fast-paths for REVERSE operation: - out_reverse_8_0565 - out_reverse_8_8888
MIPS: DSPr2: Added more fast-paths: - over_reverse_n_8888 - in_n_8_8
Siarhei Siamashka (21):
test: Change is_little_endian() into inline function
test: Added a better PRNG (pseudorandom number generator)
test: Search/replace 'lcg_*' -> 'prng_*'
test: Switch to the new PRNG instead of old LCG
test: Get rid of the obsolete 'prng_rand_N' and 'prng_rand_u32'
Improve performance of combine_over_u
test: Workaround unaligned MOVDQA bug (http://gcc.gnu.org/PR55614)
test: add "src_0565_8888" to lowlevel-blt-bench
Change CONVERT_XXXX_TO_YYYY macros into inline functions
Faster conversion from a8r8g8b8 to r5g6b5 in C code
ARM: make use of UQADD8 instruction even in generic C code paths
Rename 'xor' variable to 'filler' (because 'xor' is a C++ keyword)
Define SIZE_MAX if it is not provided by the standard C headers
Use more appropriate types and remove a magic constant
Added C variants of r5g6b5 fetch/write-back iterators
Faster write-back for the C variant of r5g6b5 dest iterator
Faster fetch for the C variant of r5g6b5 src/dest iterator
Add higher precision "pixman_transform_point_*" functions
configure.ac: Added detection for __float128 support
test: Added matrix-test for testing projective transform accuracy
Use pixman_transform_point_31_16() from pixman_transform_point()
Stefan Weil (3):
Always use xmmintrin.h for 64 bit Windows
pixman: Use uintptr_t in type casts from pointer to integral value
test: Fix compiler warnings caused by unused code
Søren Sandmann Pedersen (37):
Post-release version bump to 0.27.5
Pre-release version bump to 0.28.0
Post-release version bump to 0.29.1
pixman_image_composite: Reduce opaque masks to NULL
pixman.h: Add typedefs for pixman_f_transform and pixman_f_vector
Allow src and dst to be identical in pixman_f_transform_invert()
Round fixed-point multiplication
Convolution filter: round color values instead of truncating
Add text file rounding.txt describing how rounding works
Add new filter PIXMAN_FILTER_SEPARABLE_CONVOLUTION
rounding.txt: Describe how SEPARABLE_CONVOLUTION filter works
Add new pixman_filter_create_separable_convolution() API
demos/gtk-utils.[ch]: Add pixman_image_from_file()
demos: Add new demo program, "scale"
Add demos/zone_plate.png
Add demo program for conical gradients
Add fast paths for separable convolution
demos/radial-test: Add zero-radius circles to demonstrate rendering bugs
radial: When comparing t to mindr, use >= rather than >
test/utils.[ch]: Add utility function to draw a checkerboard
demos/conical-test: Use the draw_checkerboard() utility function
demos/radial-test: Add checkerboard to display the alpha channel
Add testing of trapezoids to stress-test
pixman_composite_trapezoids: Return early if mask_format is not of TYPE_ALPHA
pixman_composite_trapezoids(): Check for NULL return from create_bits()
float-combiner.c: Change tests for x == 0.0 tests to - FLT_MIN < x < FLT_MIN
Handle solid images in the noop iterator
pixman-filter.c: Cope with NULL returns from malloc()
README: Add guidelines on how to contribute patches
test/utils.[ch]: Add new function operator_name()
utils.[ch]: Add new format_name() function
test: Use operator_name() and format_name() in composite.c
demo/scale: Add a spin button to set the number of subsample bits
pixman-combine-float.c: Use IS_ZERO() in clip_color() and set_sat()
Change default GPGKEY to 3892336E, which is soren.sandmann@gmail.com
stresstest: Ensure that the rasterizer is only given alpha formats
Pre-release version bump to 0.29.2
Our gfxPattern cache is currently way too specific, which causes lots of
unnecessary cache misses. The only thing that we actually need to key on is the
color stops.
This switches the cache to hold GradientStops instead of gfxPatterns. This
improves our cache hit rate and is simpler. It also avoids doing caching when
not using Azure, which currently has no benefit.
Average paint times results (in ms)
(on windows 7 x64, i7-3630qm, HD4000)
-------------------------------------
m-c
open: 3.5 close: 3.3
m-c cache-key
open: 2.6 close: 2.4
ux
open: 7.3 close: 5.2
ux cache-key
open: 6.9 close: 5.3
