A new pixman release 0.15.8 is now available. This is a development
release leading up to a stable 0.16.0 release.
News:
- Phase three of the refactoring [S?ren Sandmann Pedersen]
- Fixes for refactoring fallout [Jonathan Morton, Jeff Muizelaar,
Luca Barbato, Magnus Kessler]
- Improved testing code [Siarhei Siamashka]
Soren
tar.gz:
http://cairographics.org/releases/pixman-0.15.8.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.8.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.8.tar.bz2
Hashes:
MD5: 22318888bdc56c28b742ebdd93931c41 pixman-0.15.8.tar.gz
MD5: 10230271d06a4cdc0f218f6be2664d13 pixman-0.15.8.tar.bz2
SHA1: 59d9d7d2eac9ba070a104357ecc8ef6b70b30b5a
pixman-0.15.8.tar.gz
SHA1: f2c153da11a92f2475ac03e0422d99b799fb35e1
pixman-0.15.8.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.15.8.tar.gz.sha1.asc
(signed by S?ren Sandmann Pedersen <sandmann at daimi.au.dk>)
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.15.8
Log:
Jeff Muizelaar (2):
Fix uses of dst_keep
NEON: Remove some unneeded casts
Jonathan Morton (1):
Fixup the arm-simd and arm-neon implementations.
Luca Barbato (1):
Update vmxCombine*U to the new interface
Magnus Kessler (1):
pixman: define pixman_have_{mmx,sse2} on 64-bit Linux
Siarhei Siamashka (1):
Scaling test updated to provide better coverage for problematic
cases
S?ren Sandmann Pedersen (41):
Post-release version bump
Add new store_scanline_{32,64} in bits_image_t
Move fbStoreExternalAlpha{,64} into pixman-bits-image.c
Beginning of pluggable implementations
Formatting
Add beginning of general implementation
Move entire C implementation into pixman-general.c
Beginning of MMX implementation
Add component alpha combiners in pixman-implementation.c
Change prototypes for compositing functions to use 32 bit integers
Initial fast path implementation
Copy fast path lookup code into pixman-utils.c
Beginning of sse2 implementation
Remove fast path lookup code from pixman-general
Add _pixman_choose_implementation
Change pixman_lookup_fast_path() to actually run the fast path
Move sse2 fast path running to the sse2 implementation
Move mmx fast path code to pixman-mmx.c
Move SSE2 variable initializations to
pixman_implementations_create_sse2
Use the implementation's combiner's
Make pixman_implementation call the right combiner
Set up SSE2 combiners
Move gcc alignment workaround to pixman-sse2.c
Move sse2 and mmx blt implementations to their respective files
Add a general_blt() that just returns FALSE
Add alignment workaround to sse2
Split fill implementations out in the implementations
Make the fast_path implementation run the c_fast_paths
Delete pixman-sse2.h and pixman-mmx.h
Make a couple of functions static
Move the argument struct into pixman_image_composite_rect
Consolidate the general implementation into one function
Set up combiner functions for an implementation directly in
combine.inc.
Call the toplevel implementation for combining
In _pixman_implementation_fill() don't call the delegate; call
the actual
Add back pixman_composeFunctions since vmx is not ported to
implementatio
Really fix PPC build.
In the mmx implementation, check for source == 0 rather than alpha
== 0.
In pixman-sse2.c test for non-zero source, not just non-zero
source alpha
Create a vmx pixman_implementation_t
Pre-release version bump.