ARMv6: Support for very variable-hungry composite operations
73d2f8b
create_bits(): Cast the result of height * stride to size_t
857e40f
ARM: share pixman_asm_function definition
4b76bbf
ARMv6: Add fast path for over_reverse_n_8888
4ee85b0
test: Fix OpenMP clauses for the tolerance-test
5662214
configure.ac: Check if the compiler supports GCC vector extensions
840912b
lowlevel-blt-bench: add in_reverse_8888_8888 test
c343846
lowlevel-blt-bench: over_reverse_n_8888 needs solid source
898859f
ARMv6: remove 1 instr per row in generate_composite_function
38317cb
ARMv6: Fix indentation in the composite macros
763a6d3
Remove all the operators that use division from pixman-combine32.c
82d0946
Copy the comments from pixman-combine32.c to pixman-combine-float.c
ccb1df0
utils.c: Set DEVIATION to 0.0128
94244b0
Use floating point combiners for all operators that involve divisions
15aa37a
Soft Light: Consistent approach to division by zero
8f38243
pixman-combine32.c: Fix bugs related to integer promotion
89662ad
pixman/pixman-combine32.c: Bug fixes for separable blend modes
e7a99b3
pixel-test.c: Add a number of pixels that have failed at some point
fe3504d
test/tolerance-test: New test program
bd94c17
pixel-test: Command line argument to specify the regression to run
c2fd65d
pixel-test: Add support for mask pixels
a692e01
test/check-formats.c: Add support for separable blend modes
779ca46
test/utils.c: Add support for separable blend mode ops to do_composite()
a42af27
configure.ac: Check and use -Wno-unused-local-typedefs GCC option
b29d74e
Soft Light: The first comparison should be <=, not <
945ab7a
general: Support component alpha for all image types
9ba3a34
test/utils.c: Make the stack unaligned only on 32 bit Windows
f473fd1
Fix the SSSE3 CPUID detection.
8487dfb
demos/Makefile.am: Move EXTRA_DIST outside "if HAVE_GTK"
7410073
test: Fix the win32 build
9bab46e