[ANNOUNCE] pixman release 0.23.4 now available
A new pixman release 0.23.4 is now available. This is a development
release leading up to a stable release 0.24.0.
Note: If you run the test suite on this released tarball, this message:
*** BUG ***
In pixman_region32_union_rect: Invalid rectangle passed
Set a breakpoint on '_pixman_log_error' to debug
will be printed a number of times when the region-contains-test is
running. The test is still expected to pass, however.
Please report bugs to pix...@lists.freedesktop.org or file them at
https://bugs.freedesktop.org/enter_bug.cgi?product=pixman
Søren
tar.gz:
http://cairographics.org/snapshots/pixman-0.23.4.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.23.4.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.23.4.tar.bz2
Hashes:
MD5: 389ce497fcfb26eb99396ca5ed6e17ab pixman-0.23.4.tar.gz
MD5: 1fb7c1ab150cfbc417290d5888d68ce8 pixman-0.23.4.tar.bz2
SHA1: 539f4875070ce836a109fa0fae7a3421d503ca46 pixman-0.23.4.tar.gz
SHA1: 5c85fff3f474436fdcc0c29139ce4ed87e38e24b pixman-0.23.4.tar.bz2
GPG signature:
http://cairographics.org/snapshots/pixman-0.23.4.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen sandm...@cs.au.dk)
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.23.4
Log:
Andrea Canciani (4):
radial: Improve documentation and naming
radial: Fix typos and trailing whitespace
win32: Build benchmarks
Workaround bug in llvm-gcc
Chris Wilson (1):
bits: optimise fetching width==1 repeats
Siarhei Siamashka (2):
C fast path for scaled src_x888_ with nearest filter
ARM: workaround binutils bug #12931 (code sections alignment)
Søren Sandmann Pedersen (12):
Post-release version bump to 0.23.3
Makefile.am: Add pix...@lists.freedesktop.org to
RELEASE_ANNOUNCE_LIST
Fix lcg_rand_u32() to return 32 random bits.
New test of pixman_region_contains_{rectangle,point}
Speed up pixman_region{,32}_contains_rectangle()
Use find_box_for_y() in pixman_region_contains_point() too
Don't include stdint.h in lowlevel-blt-bench.c
In pixman_image_create_bits() allow images larger than 2GB
Rename pixman-fast-path.h to pixman-inlines.h
Use repeat() function from pixman-inlines.h in pixman-bits-image.c
Move bilinear interpolation to pixman-inlines.h
Pre-release version bump to 0.23.4
Taekyun Kim (3):
ARM NEON: Standard fast path out_reverse_8_
ARM: NEON better instruction scheduling of over_n_8_
ARM: NEON better instruction scheduling of over_n_
___
xorg-announce mailing list
xorg-announce@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] xf86-video-qxl 0.0.14
Alon Levy (15): gcc-4.6.0 fix: cast unmatching function pointers qxl_driver: remove unused variable (found by gcc 4.6.0) build: add spice-protocol dep use spice-protocol qxl_dev.h enums use structs from spice-protocol qxl_dev.h s/qxl_ram_header/QXLRam/ qxl.h: double include protection qxl_mem.c: fix pointer arithmatic warning (-Wpointer-arith) qxl_image.c: fix defined but unused warning qxl.h: move spice include right after system includes use spice-protocol defines for ring sizes qxl_driver: fix initialization from wrong pointer warning qxl_driver: fix three incompatible pointer passed warnings qxl_driver: whitespace fixes cleanup (add eol between system and non system includes) Søren Sandmann Pedersen (5): Don't re-upload read-only surfaces after a software fallback. Tiled upload of images after software fallback. Generate tighter damage for PolyFillRect Fix VT switching 0.0.14 git tag: xf86-video-qxl-0.0.14 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.14.tar.bz2 MD5: 3e1098302cc2efc7950322d9341537f4 xf86-video-qxl-0.0.14.tar.bz2 SHA1: 49648337fa7ad3f098811f04855228829769f79b xf86-video-qxl-0.0.14.tar.bz2 SHA256: 7283a3719cc71fe98045d92ed0d480aaccd3a5e202fc30cbe0103458a7ff2e60 xf86-video-qxl-0.0.14.tar.bz2 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.14.tar.gz MD5: 8a9d49ade004a99a71380c78dc753d9d xf86-video-qxl-0.0.14.tar.gz SHA1: 7d4652e5949e3436a49e71141aaa4292de127e79 xf86-video-qxl-0.0.14.tar.gz SHA256: eec959b27231cefdf55a81624535ac2f753323baa95d70e4dec1a2ca37ab80c9 xf86-video-qxl-0.0.14.tar.gz ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] xf86-video-qxl 0.0.14
Alon Levy (15): gcc-4.6.0 fix: cast unmatching function pointers qxl_driver: remove unused variable (found by gcc 4.6.0) build: add spice-protocol dep use spice-protocol qxl_dev.h enums use structs from spice-protocol qxl_dev.h s/qxl_ram_header/QXLRam/ qxl.h: double include protection qxl_mem.c: fix pointer arithmatic warning (-Wpointer-arith) qxl_image.c: fix defined but unused warning qxl.h: move spice include right after system includes use spice-protocol defines for ring sizes qxl_driver: fix initialization from wrong pointer warning qxl_driver: fix three incompatible pointer passed warnings qxl_driver: whitespace fixes cleanup (add eol between system and non system includes) Søren Sandmann Pedersen (5): Don't re-upload read-only surfaces after a software fallback. Tiled upload of images after software fallback. Generate tighter damage for PolyFillRect Fix VT switching 0.0.14 git tag: xf86-video-qxl-0.0.14 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.14.tar.bz2 MD5: 3e1098302cc2efc7950322d9341537f4 xf86-video-qxl-0.0.14.tar.bz2 SHA1: 49648337fa7ad3f098811f04855228829769f79b xf86-video-qxl-0.0.14.tar.bz2 SHA256: 7283a3719cc71fe98045d92ed0d480aaccd3a5e202fc30cbe0103458a7ff2e60 xf86-video-qxl-0.0.14.tar.bz2 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.14.tar.gz MD5: 8a9d49ade004a99a71380c78dc753d9d xf86-video-qxl-0.0.14.tar.gz SHA1: 7d4652e5949e3436a49e71141aaa4292de127e79 xf86-video-qxl-0.0.14.tar.gz SHA256: eec959b27231cefdf55a81624535ac2f753323baa95d70e4dec1a2ca37ab80c9 xf86-video-qxl-0.0.14.tar.gz ___ xorg@lists.freedesktop.org: X.Org support Archives: http://lists.freedesktop.org/archives/xorg Info: http://lists.freedesktop.org/mailman/listinfo/xorg Your subscription address: arch...@mail-archive.com
[ANNOUNCE] pixman release 0.22.2 now available
A new pixman release 0.22.2 is now available. This is a stable bug fix release. News: - Fix for bugs causing glyph corruption on ARM - Build fixes for Windows and OS/2. Soren tar.gz: http://cairographics.org/releases/pixman-0.22.2.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.22.2.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.22.2.tar.bz2 Hashes: MD5: 3dd0d9ed05dbf0e5e75d526ebae42e11 pixman-0.22.2.tar.gz MD5: 863b87e7a7d6e1d032fd583edc1df5dd pixman-0.22.2.tar.bz2 SHA1: b42a7aec5119b1e5ce8eed06452645fa3a4892ac pixman-0.22.2.tar.gz SHA1: ad2b828ce4280472f5933d8bb5f0f4d583aed7f3 pixman-0.22.2.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.22.2.tar.gz.sha1.asc (signed by Soren Sandmann sandm...@daimi.au.dk) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.22.2 Log: Andrea Canciani (1): test: Fix compilation on win32 Dave Yeo (1): Check for working mmap() Søren Sandmann (1): ARM: Fix two bugs in neon_composite_over_n__0565_ca(). Søren Sandmann Pedersen (2): Post-release version bump to 0.22.1 Pre-release version bump to 0.22.2 ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] pixman release 0.23.2 now available
A new pixman release 0.23.2 is now available. This is the first
development release leading up to a stable 0.24 release.
- Improved support for tiled bilinear scaling on SSE2 and ARM [Taekyun Kim]
- Fix for a bug causing glyph corruption on ARM devices with a 16 bit
frame buffer [Søren Sandmann]
Please report bugs to pixman at lists.freedesktop.org or file them at
https://bugs.freedesktop.org/enter_bug.cgi?product=pixman
Søren
tar.gz:
http://cairographics.org/snapshots/pixman-0.23.2.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.23.2.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.23.2.tar.bz2
Hashes:
MD5: 775c5ecde7485bd69936dcc978f5aa58 pixman-0.23.2.tar.gz
MD5: 2e2805f5ca02edeb15a7862779670069 pixman-0.23.2.tar.bz2
SHA1: 5100c1e2d4566e507ead5c638d2a66bb165b2e63 pixman-0.23.2.tar.gz
SHA1: 62568ed6eb84c0312cc0e6f9c5f4bc12b8202492 pixman-0.23.2.tar.bz2
GPG signature:
http://cairographics.org/snapshots/pixman-0.23.2.tar.gz.sha1.asc
(signed by Soren Sandmann sandm...@daimi.au.dk)
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.23.2
Log:
Andrea Canciani (3):
test: Fix compilation on win32
Include noop in win32 builds
Silence autoconf warnings
Dave Yeo (1):
Check for working mmap()
Nis Martensen (1):
Fix a few typos in pixman-combine.c.template
Søren Sandmann (5):
mmx: Delete some unused variables
sse2: Delete some unused variables
demos: Comment out some unused variables
ARM: Fix two bugs in neon_composite_over_n__0565_ca().
test: Make fuzzer-find-diff.pl executable
Søren Sandmann Pedersen (12):
Post-release version bump to 0.23.1
Add a noop implementation.
Add a noop composite function for the DST operator
Move noop dest fetching to noop implementation
Add a noop src iterator
Move NULL iterator into pixman-noop.c
Move NOP src iterator into noop implementation.
Replace instances of dst_* with dest_*
In pixman-general.c rename image_parameters to {src, mask,
dest}_image
Replace argumentxs to composite functions with a pointer to a
struct
blitters-test: Make common formats more likely to be tested.
Pre-release version bump to 0.23.2
Taekyun Kim (6):
Replace boolean arguments with flags for bilinear fast path
template
REPEAT_NORMAL support for bilinear fast path template
sse2: Declare bilinear src__ REPEAT_NORMAL composite
function
ARM: Add REPEAT_NORMAL functions to bilinear BIND macros
Enable REPEAT_NORMAL bilinear fast path entries
Bilinear REPEAT_NORMAL source line extension for too short
src_width
___
xorg-announce mailing list
xorg-announce@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] pixman stable release 0.22.0 now available
A new stable pixman release 0.22.0 is now available. Highlights
of this release:
- New r8g8b8a8 and r8g8b8x8 image formats [Alexandros Frantzis]
- Much faster image scaling on ARM and x86 [Siarhei Siamashka, Taekyun Kim]
- Faster 90/270 degree image rotation [Siarhei Siamashka]
- More comprehensive support for compositing triangles and
trapezoids [Søren Sandmann Pedersen]
Plus a large number of other performance improvements, bug and
portability fixes, and improvements to the test suite.
Thanks to everybody who contributed to pixman 0.22.0, including
Alan Coopersmith, Alexandros Frantzis, Andrea Canciani, Cyril
Brulebois, Gilles Espinasse, Jon TURNEY, Rolland Dudemaine,
Siarhei Siamashka, Søren Sandmann Pedersen, and Taekyun Kim.
Søren
tar.gz:
http://cairographics.org/releases/pixman-0.22.0.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.22.0.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.22.0.tar.bz2
Hashes:
MD5: cb8f3cb5ce2c8d7294f73ecb7021fda6 pixman-0.22.0.tar.gz
MD5: 307fe4d7dc83b1a558c362907097c0d0 pixman-0.22.0.tar.bz2
SHA1: da0a9c63fa315f163a32c7e40dd0b2a0f88c0d21 pixman-0.22.0.tar.gz
SHA1: d24ea233755d7dce9f0d93136ad99fba8d4e4fa0 pixman-0.22.0.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.22.0.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen sandm...@daimi.au.dk)
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.22.0
Log:
Alan Coopersmith (1):
Sun's copyrights belong to Oracle now
Alexandros Frantzis (2):
Add simple support for the r8g8b8a8 and r8g8b8x8 formats.
Add support for the r8g8b8a8 and r8g8b8x8 formats to the tests.
Andrea Canciani (10):
Remove unused stop_range field
Fix opacity check
Improve conical gradients opacity check
Improve handling of tangent circles
Add a test for radial gradients
Fix compilation on Win32
test: Fix tests for compilation on Windows
test: Add Makefile for Win32
Do not include unused headers
test: Silence MSVC warnings
Cyril Brulebois (2):
Fix argument quoting for AC_INIT.
Fix linking issues when HAVE_FEENABLEEXCEPT is set.
Gilles Espinasse (2):
Fix missing AC_MSG_RESULT value from Werror test
Fix OpenMP not supported case
Jon TURNEY (1):
Remove stray #include fenv.h
Rolland Dudemaine (4):
test: Fix for mismatched 'fence_malloc' prototype/implementation
Correct the initialization of 'max_vx'
test: Use the right enum types instead of int to fix warnings
Fix variable was set but never used warnings
Siarhei Siamashka (69):
Fixed broken configure check for __thread support
Do CPU features detection from 'constructor' function when
compiled with
ARM: fix 'vld1.8'-'vld1.32' typo in add__ NEON fast path
ARM: NEON: source image pixel fetcher can be overrided now
ARM: nearest scaling support for NEON scanline compositing
functions
ARM: macro template in C code to simplify using scaled fast paths
ARM: performance tuning of NEON nearest scaled pixel fetcher
ARM: NEON optimization for scaled over__ with nearest
filter
ARM: NEON optimization for scaled over__0565 with nearest
filter
ARM: NEON optimization for scaled src__0565 with nearest
filter
ARM: NEON optimization for scaled src_0565_ with nearest
filter
ARM: optimization for scaled src_0565_0565 with nearest filter
C fast path for a1 fill operation
ARM: added 'neon_composite_over_n_8_8' fast path
ARM: introduced 'fetch_mask_pixblock' macro to simplify code
ARM: better NEON instructions scheduling for over_n_8_0565
ARM: added 'neon_composite_over__n_0565' fast path
ARM: reuse common NEON code for over_{n_8|_n|_8}_0565
ARM: added 'neon_composite_over_0565_n_0565' fast path
ARM: added 'neon_composite_add__8_' fast path
ARM: better NEON instructions scheduling for add___
ARM: added 'neon_composite_add_n_8_' fast path
ARM: added 'neon_composite_add__n_' fast path
ARM: added flags parameter to some asm fast path wrapper macros
ARM: added 'neon_composite_in_n_8' fast path
ARM: added 'neon_src_rpixbuf_' fast path
Fix for potential unaligned memory accesses
[ANNOUNCE] pixman release 0.21.4 now available
A new pixman release 0.21.4 is now available. This is a development
snapshot leading up to a stable 0.22.0 release. News:
- ARM performance improvements and bug fixes [Siarhei Siamashka]
- Performance improvements for bitmap filling [Siarhei]
- Improvements to the test suite [Andrea Canciani, Siarhei, Søren Sandmann]
- Refactoring of image access code [Søren]
Plus bug fixes and other updates [Alan Coopersmith, Andrea, Cyril
Brulebois, Jon TURNEY, Siarhei, Søren]
Please file bugs at
https://bugs.freedesktop.org/enter_bug.cgi?product=pixman
or report them at the pixman mailing list:
pix...@lists.freedesktop.org
Søren
tar.gz:
http://cairographics.org/releases/pixman-0.21.4.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.21.4.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.21.4.tar.bz2
Hashes:
MD5: e50975ace979cd416a505827c15191b4 pixman-0.21.4.tar.gz
MD5: ae57f4fa9121ab9895b9789b2e81f1e0 pixman-0.21.4.tar.bz2
SHA1: 3736d354b99b318628f49b5128a02e3bf8be9e0f pixman-0.21.4.tar.gz
SHA1: ca2c6f0b84500d4521b8ce39f62aea5b5d433974 pixman-0.21.4.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.21.4.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen sandm...@daimi.au.dk)
Git:
git://git.freedesktop.org/git/pixmantag: pixman-0.21.4
Log:
Alan Coopersmith (1):
Sun's copyrights belong to Oracle now
Andrea Canciani (2):
Improve handling of tangent circles
Add a test for radial gradients
Cyril Brulebois (2):
Fix argument quoting for AC_INIT.
Fix linking issues when HAVE_FEENABLEEXCEPT is set.
Jon TURNEY (1):
Remove stray #include fenv.h
Siarhei Siamashka (23):
C fast path for a1 fill operation
ARM: added 'neon_composite_over_n_8_8' fast path
ARM: introduced 'fetch_mask_pixblock' macro to simplify code
ARM: better NEON instructions scheduling for over_n_8_0565
ARM: added 'neon_composite_over__n_0565' fast path
ARM: reuse common NEON code for over_{n_8|_n|_8}_0565
ARM: added 'neon_composite_over_0565_n_0565' fast path
ARM: added 'neon_composite_add__8_' fast path
ARM: better NEON instructions scheduling for add___
ARM: added 'neon_composite_add_n_8_' fast path
ARM: added 'neon_composite_add__n_' fast path
ARM: added flags parameter to some asm fast path wrapper macros
ARM: added 'neon_composite_in_n_8' fast path
ARM: added 'neon_src_rpixbuf_' fast path
Fix for potential unaligned memory accesses
COPYING: added Nokia to the list of copyright holders
Revert Fix syntax error: empty declaration warnings.
Fix for syntax error: empty declaration Solaris Studio warnings
Workaround for a preprocessor issue in old Sun Studio
Bugfix for a corner case in 'pixman_transform_is_inverse'
Make 'fast_composite_scaled_nearest_*' less suspicious
A new configure option --enable-static-testprogs
ARM: do /proc/self/auxv based cpu features detection only in linux
Søren Sandmann Pedersen (28):
Post-release version bump to 0.21.3
test: Make composite test use some existing macros instead of
defining it
Add enable_fp_exceptions() function in utils.[ch]
Extend gradient-crash-test
test: Move palette initialization to utils.[ch]
test/utils.c: Initialize palette-rgba to 0.
Make the argument to fence_malloc() an int64_t
Add a stress-test program.
Add a test compositing with the various PDF operators.
Fix divide-by-zero in set_lum().
sse2: Skip src pixels that are zero in
sse2_composite_over__n_()
Add iterators in the general implementation
Move initialization of iterators for bits images to
pixman-bits-image.c
Eliminate the _pixman_image_store_scanline_32/64 functions
Move iterator initialization to the respective image files
Virtualize iterator initialization
Use an iterator in pixman_image_get_solid()
Move get_scanline_32/64 to the bits part of the image struct
Allow NULL property_changed function
Consolidate the various get_scanline_32() into
get_scanline_narrow()
Linear: Optimize for horizontal gradients
Get rid of the classify methods
Add direct-write optimization back
Skip fetching pixels when
[ANNOUNCE] pixman release 0.21.2 now available
A new pixman release 0.21.2 is now available. This is the first
development snapshot leading up to a stable 0.22 release.
News:
ARM: Performance improvements for image scaling [Siarhei Siamashka]
Performance improvements for affine transformations [Soren Sandmann]
Plus bug fixes and other improvements [Andrea Canciani, Siarhei, Soren].
NOTE: In this release a workaround for a bug in older version of the X
server has been removed. If your X server is version 1.6 or older, you
may see image corruption bugs with this version of pixman.
Thanks,
Soren
tar.gz:
http://cairographics.org/releases/pixman-0.21.2.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.21.2.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.21.2.tar.bz2
Hashes:
MD5: 9e09fd6e58cbf9717140891e0b7d4a7a pixman-0.21.2.tar.gz
MD5: 4bc4cf052635265f7a98ad3e890ae329 pixman-0.21.2.tar.bz2
SHA1: c0ff07d7e4877dd4d0d369ca09e50ca956e3386e pixman-0.21.2.tar.gz
SHA1: bb5cba514fe6756f9264a6995ec9dfed4d7439b5 pixman-0.21.2.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.21.2.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen sandm...@daimi.au.dk
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.21.2
Log:
Andrea Canciani (3):
Remove unused stop_range field
Fix opacity check
Improve conical gradients opacity check
Siarhei Siamashka (12):
Fixed broken configure check for __thread support
Do CPU features detection from 'constructor' function when
compiled with
ARM: fix 'vld1.8'-'vld1.32' typo in add__ NEON fast path
ARM: NEON: source image pixel fetcher can be overrided now
ARM: nearest scaling support for NEON scanline compositing
functions
ARM: macro template in C code to simplify using scaled fast paths
ARM: performance tuning of NEON nearest scaled pixel fetcher
ARM: NEON optimization for scaled over__ with nearest
filter
ARM: NEON optimization for scaled over__0565 with nearest
filter
ARM: NEON optimization for scaled src__0565 with nearest
filter
ARM: NEON optimization for scaled src_0565_ with nearest
filter
ARM: optimization for scaled src_0565_0565 with nearest filter
Søren Sandmann Pedersen (8):
Post-release version bump to 0.20.1
Version bump 0.21.1.
COPYING: Stop saying that a modification is currently under
discussion.
Remove workaround for a bug in the 1.6 X server.
[mmx] Mark some of the output variables as early-clobber.
Delete the source_image_t struct.
Generate {a,x}8r8g8b8, a8, 565 fetchers for nearest/affine images
Pre-release version bump
___
xorg-announce mailing list
xorg-announce@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg-announce
Re: Guarantees of order of X events vs requests
Eirik Byrkjeflot Anonsen ei...@opera.com writes: What guarantees does X give when it comes to the order of events generated in relation to processing of the requests sent by the client? (Also, of course: To which degree does various implementations of X actually fulfill these guarantees?) It would be a major bug if any X server didn't order events correctly. It should be safe to rely on, though you mileage and level of cynicism may vary. X events have a serial value. I expect that any event delivered by X will reflect the state after the request number serial (and all preceding requests) have been processed. Is this correct? Can I also assume that the X event will reflect the state before any requests with a later serial number is processed? That is correct. The protocol spec says: Whether or not a server is implemented with internal concurrency, the overall effect must be as if individual requests are executed to completion in some serial order, and requests from a given connection must be executed in delivery order (that is, the total execution order is a shuffle of the individual streams). The execution of a request includes validating all arguments, collecting all data for any reply, and generating and queueing all required events. Given an application that frequently performs a sequence of XCopyArea() calls on the contents of a window. When this application receives Expose events or GraphicsExpose events, it is necessary for the application to know exactly which XCopyArea calls have taken effect to be able to correctly calculate which area of the window has become invalid. Right. GTK+ used to have (and likely still has) a scheme where a CopyArea generates an internal record that a piece of the window has moved at a certain serial number. These records are stored in a queue. Then, when an expose arrives, entries in the queue with a serial number before the expose are processed. If the area touched by the record intersects the area exposed, the expose is treated as if it exposed the region (original_rect - intersected_rectangle) + (translated intersected_rectangle) Unless you have a similar mechanism, you will probably have to call XSync() after each XCopyArea() to ensure that the copying and associated exposes have been processed on the server, before procesing any exposes. Soren ___ xorg@lists.freedesktop.org: X.Org support Archives: http://lists.freedesktop.org/archives/xorg Info: http://lists.freedesktop.org/mailman/listinfo/xorg Your subscription address: arch...@mail-archive.com
[ANNOUNCE] pixman release 0.19.2 now available
A new pixman release 0.19.2 is now available. This is a development
snapshot leading up to a stable 0.20.0 release.
News:
- Gradient improvements [Andrea Canciani, Soren Sandmann]
- Altivec detection on OpenBSD [Brad Smith]
- Improved configure-time detection of pthreads [M Joonas Pihlaja]
- Support for x14r6g6b6 format [Marek Vasut]
- Improvements to the test suite [Siarhei Siamashka]
- NEON based combiner for OUT_REVERSE operator [Siarhei]
- Improvements to the SSE2 implementation [Soren]
- Alpha map loops are now caught before they can cause problems [Soren]
- Apply operator strength reduction in more cases [Soren]
Plus bug fixes and some internal reorganisation [Jeff Muizelaar,
Siarhei, Siddarth Agarwal, Soren].
Please report bugs by sending mail to pix...@lists.freedesktop.org or
by filing an issue in bugzilla:
https://bugs.freedesktop.org/enter_bug.cgi?product=pixman
Thanks,
Soren
tar.gz:
http://cairographics.org/releases/pixman-0.19.2.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.19.2.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.19.2.tar.bz2
Hashes:
MD5: dfb84c9f402fbba5047f0e0a69c17518 pixman-0.19.2.tar.gz
MD5: 4ed74b6aa9139541718c0d61a581fd50 pixman-0.19.2.tar.bz2
SHA1: beef0a10373f4c551698a376d62af158ac167d87 pixman-0.19.2.tar.gz
SHA1: c3d69deca35a1fd4f3ca9d7ca4344d805f56e205 pixman-0.19.2.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.19.2.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen sandm...@daimi.au.dk)
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.19.2
Log:
Andrea Canciani (2):
support single-stop gradients
test: verify that gradients do not crash pixman
Brad Smith (1):
Add support for AltiVec detection for OpenBSD/PowerPC.
Jeff Muizelaar (3):
Add missing HAVE_CONFIG_H guards for config.h inclusion
Add support for compiling pixman without thread/tls support
create getter for component alpha
M Joonas Pihlaja (6):
Try harder to find suitable flags for pthreads.
Don't trust OpenBSD's gcc to produce working code for __thread.
Check that the OpenMP pragmas don't cause link errors.
Check for specific flags by actually trying to compile and link.
Avoid trailing slashes on automake install dirs.
Fix thinko in configure.ac's macro to test linking.
Marek Vasut (1):
Add support for 32bpp X14R6G6B6 format.
Siarhei Siamashka (9):
test: main loop from blitters-test added as a new function to
utils.c
test: blitters-test-bisect.rb converted to perl
test: blitters-test updated to use new fuzzer_test_main() function
test: scaling-test updated to use new fuzzer_test_main() function
test: added OpenMP support for better utilization of multiple CPU
cores
test: 'scaling-crash-test' added
test: 'scaling-test' updated to provide better coverage
Code simplification (no need advancing 'vx' at the end of
scanline)
ARM: 'neon_combine_out_reverse_u' combiner
Siddharth Agarwal (1):
Visual Studio 2010 includes stdint.h
Søren Sandmann Pedersen (57):
Post-release version bump to 0.19.1
Fix uninitialized cache when pthreads are used
Fixes for pthread thread local storage.
[mmx] Fix mask creation bugs
Update README to mention the pixman mailing list
Some minor updates to README
Add pixman_image_get_format() accessor
Remove alphamap from the GTK+ part of tests/Makefile.am
Rename fast_composite_src__x888 to fast_composite_src_memcpy()
Add pixman_region{,32}_intersect_rect()
Add an over___ sse2 fast path.
Don't consider indexed formats opaque.
sse2: Add sse2_composite_over_reverse_n_
Add support for 8bpp to pixman_fill_sse2()
Don't use __thread on MinGW.
Add macros for thread local storage on MinGW 32
test/gtk-utils: Set the size of the window to the size of the
image
Eliminate mask_bits from all the scanline fetchers.
When storing a g1 pixel, store the lowest bit, rather than
comparing with
Make separate gray scanline storers.
Fix conical gradients to match QConicalGradient from Qt
Store the conical angle in floating point radians, not fixed
point degree
Minor tweaks to README
Make the combiner macros less likely to cause name
[ANNOUNCE] Stable pixman release 0.18.4 now available
A new pixman release 0.18.4 is now available. This is a stable release which contains bug fixes since 0.18.2. NOTE: There is a known memory leak in this release when pixman is used in multi-threaded programs on Windows. This bug is present in 0.18.0 and 0.18.2 as well. If you know how to fix it, please do. A similar bug in the pthreads code was fixed in this release. Please report bugs to pix...@lists.freedesktop.org or file them at https://bugs.freedesktop.org/enter_bug.cgi?product=pixman Thanks, Soren tar.gz: http://cairographics.org/releases/pixman-0.18.4.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.18.4.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.18.4.tar.bz2 Hashes: MD5: 532db4194e18c108d0b7dd85f20d39b8 pixman-0.18.4.tar.gz MD5: c1d69aaddba8c1e046d26ac125da95bf pixman-0.18.4.tar.bz2 SHA1: a1f03504ffee182a16c49c822a9254d4bf92d6c0 pixman-0.18.4.tar.gz SHA1: 011b00fb33d6e5021a680bcc74f600f44d709d8d pixman-0.18.4.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.18.4.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk Git: git://git.freedesktop.org/git/pixman tag: pixman-0.18.4 Log: Søren Sandmann Pedersen (6): Post-release version bump to 0.18.3 Fix memory leak in the pthreads thread local storage code When storing a g1 pixel, store the lowest bit, rather than comparing with If we bail out of do_composite, make sure to undo any workarounds. Check for read accessors before taking the bilinear fast path Pre-release version bump to 0.18.4 ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
Re: [Pixman] [PATCH] Fix server crash in pixman (to be discussed)
Hi Matthias, The following patch fixes Novell bug 568811: VNC Installation aborts right in the middle due to an assertion in Xvnc/libpixman The bug seems occur only on *very* special occasions (in this case, only in SLES, but *not* in SLED, which is based on the same code basis...). Backtrace looks as follows: #2 0xb71f7baa in __assert_fail () from /lib/libc.so.6 #3 0xb781feab in pixman_region_append_non_o (y2=value optimized out, y1=value optimized out, r_end=value optimized out, r=value optimized out, region=value optimized out) at pixman-region.c:670 #4 pixman_op (y2=value optimized out, y1=value optimized out, r_end=value optimized out, r=value optimized out, region=value optimized out) at pixman-region.c:996 #5 0xb7820dbe in pixman_region_union (new_reg=0x82ee57c, reg1=0x82ee57c, reg2=0xbfd77c90) at pixman-region.c:1439 #6 0x080f023b in miUnion (newReg=0x82ee57c, reg1=0x82ee57c, reg2=0xbfd77c90) at miregion.c:1005 Thanks for tracking down this issue. Please note that while pixman is not as strict as the X server in who can push to the repository, it is not a complete free-for-all. Committing small, obvious patches that fixes typos or oversights is fine, but don't commit non-obvious stuff without sending it to the mailing list and giving people time to comment on it. When we are this close to stable release, we need to be even more careful about what goes in. Patches committed between now and next week will likely ship without no testing at all. This patch in particular, I don't think shold ship with no testing at all. So please revert it, and we can consider it again for 0.19.x. Comments and questions: - Why are you seeing an assert in the first place? The asserts have been disabled since 0.16.4/0.17.6 because pixman shouldn't take down the X server with asserts for things that would otherwise be harmless. If you are using an older version than those in SLES, then you might want to either upgrade to 0.16.4, or backport ec6de472d042bec05aaa53f9d14bfbe23edbe01e which is the commit that disables asserts. - How does the degenerate region get created? If you call pixman_init_rectangle() with an empty rectangle, the region will be correctly initialized to have the 'data' field point to pixman_region_empty_data_. I believe the various operations all make sure that if the resulting region is empty, they will store a pointer to pixman_region_empty_data_ in the data field. For better or worse, the region code currently relies on this invariant, so users shouldn't be filling out the region data structure themselves. You can argue that this is bad and pixman should cope with regions initialized directly with empty rectangles, and I probably would agree; it may make sense to fix it in pixman 0.20. It certainly does not make sense to make this change with no testing this close to a stable release unless it is demonstrably a serious bug in pixman that is causing crashes. - The patch does not look wrong to me - a region with empty extents should probably be considered empty rather than degenerate. However, unless this issue is truly causing crashes, changing it is the kind of thing that could cause all kinds of mysterious bugs to show up and therefore definitely should not happen between a release candidate and a stable release. Did you carefully review all the code to make sure it always uses PIXMAN_NIL to check for empty regions? - The region code is rather tricky, so when bugs are fixed in it, I'd like to have tests/region-test.c updated to test the issue. That way - we have a clear way to reproduce the bug - we can verify that the bug is really fixed - we make sure it doesn't get reintroduced - Finally, we need much more detail in commit messages than Fixes Novell bug xx. Is that even a public bug? I couldn't figure out how to look at it. I did not create a Novell Login though. Soren ___ xorg@lists.freedesktop.org: X.Org support Archives: http://lists.freedesktop.org/archives/xorg Info: http://lists.freedesktop.org/mailman/listinfo/xorg
Re: [Pixman] [PATCH] Fix server crash in pixman (to be discussed)
Matthias Hopf mh...@suse.de writes: This patch in particular, I don't think shold ship with no testing at all. So please revert it, and we can consider it again for 0.19.x. I'm fine with that if you consider it problematic. Given that the situation it changes should actually not occur at all, I doubt there's any fallout, though. I do consider it problematic. Thanks for reverting it. - Why are you seeing an assert in the first place? The asserts have been disabled since 0.16.4/0.17.6 because pixman shouldn't take down the X server with asserts for things that would otherwise be harmless. Apparently we're still using 0.16.0 in SLE11SP1. So that explains. I double checked that there were no related changes in the region code before fixing. However, what happens if the code would have been compiled with -NDEBUG? Is the code path stable with empty regions? If it is, it can be argued that the patch is not necessary, but it could also be argued that the assert() shouldn't have been there in the first place. Who knows? Who knows if it's stable even *with* the patch? That's why I don't want it in for 0.18.0. If you are using an older version than those in SLES, then you might want to either upgrade to 0.16.4, or backport No chance. We are version and feature freeze for quite some time now. Time cycles are much different in enterprise products :-/ Especially in SLE11SP1, as changes to SLE11 should be as minimal as possible (and SLE11 is over a year old). ec6de472d042bec05aaa53f9d14bfbe23edbe01e That sounds more realistic. However, we don't have any other issues with assert()s, and there is a slim chance that this backport introduces additional regressions (asserts with side effects etc.). If it were *my* enterprise product, I'd definitely get rid of the assertions, because they are known to take down the X server in various situations. That's your call of course. As of 0.17.6 the assertions are not even enabled in unstable releases because the only result were that they get triggered by the same old X server issues, which just causes people to not test the pixman releases. Soren ___ xorg@lists.freedesktop.org: X.Org support Archives: http://lists.freedesktop.org/archives/xorg Info: http://lists.freedesktop.org/mailman/listinfo/xorg
[ANNOUNCE] pixman beta release 0.17.12 now available
A new pixman release 0.17.12 is now available. This is a BETA release leading up to a stable 0.18.0 release. News: - Optimizations for rescaling with a nearest filter [Alexander Larsson, Siarhei Siamashka] - Overhead reduction by caching fast paths [Søren Sandmann] - Constant time operator strength reduction [Søren] - Misc. bug fixes [Alan Coopersmith, Egor Starkov, Loïc Minier] Please test this release thoroughly. If you are a distribution, it would be appropriate to put this release in your development/unstable version, if you have one. Testing on OS X and Windows is especially appreciated. Report bugs either at https://bugs.freedesktop.org/enter_bug.cgi?product=pixman or by sending mail to pix...@lists.freedesktop.org. Thanks, Søren tar.gz: http://cairographics.org/releases/pixman-0.17.12.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.12.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.12.tar.bz2 Hashes: MD5: 331415d7a110145cf27aa00e11d0a683 pixman-0.17.12.tar.gz MD5: fd4fdb27e7f2da69dd7e5084ab3021ae pixman-0.17.12.tar.bz2 SHA1: 0ad539300f52773967e1007b8ba29a90d53c813a pixman-0.17.12.tar.gz SHA1: 4a5e4e58828fb562def08b51e7f8c2bc0d3f2e3b pixman-0.17.12.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.17.12.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk Git: git://git.freedesktop.org/git/pixman tag: pixman-0.17.12 Log: Alan Coopersmith (2): Make .s target asm flag selection more portable Add Sun cc to thread-local support checks in pixman-compiler.h Alexander Larsson (6): Add CONVERT_0565_TO_ macro Add CONVERT__TO_ and CONVERT_0565_TO_0565 macros Add FAST_PATH_NO_NONE_REPEAT flag Add FAST_PATH_SAMPLES_COVER_CLIP and FAST_PATH_16BIT_SAFE Add specialized fast nearest scalers Use the right format for the OVER__565 fast path Egor Starkov (1): Eliminate trailing comma in enum Loïc Minier (1): ARM: SIMD: Try without any CFLAGS before forcing -mcpu= Søren Sandmann Pedersen (16): Post-release version bump to 0.17.11 Add a fast path cache Make the operator strength reduction constant time. Store the operator table more compactly. Strength reduce certain conjoint/disjoint to their normal counterparts. Change operator table to be an array of arrays of four bytes. pixman.h: Only define stdint types when PIXMAN_DONT_DEFINE_STDINT is unde test: Add support for indexed formats to blitters-test Fix typo: WORDS_BIG_ENDIAN = WORDS_BIGENDIAN in pixman-edge.c Add checks for various types of thread local storage. Add PIXMAN_DEFINE_THREAD_LOCAL() and PIXMAN_GET_THREAD_LOCAL() macros Fix contact address in configure.ac Ensure that only the low 4 bit of 4 bit pixels are stored. Add a FAST_PATH_X_UNIT_POSITIVE flag Specialize the fast_composite_scaled_nearest_* scalers to positive x unit Pre-release version bump to 0.17.12 ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
Re: Tinderbox update
Chris Ball c...@laptop.org writes: The i686/Linux machine (live) is running each X server it compiles, and testing a few apps (e.g. x11perf, cairo make test) against it. It's a radeon r100 card running with KMS enabled. It might make sense to also run make check for pixman. What happens if the tests fail? Thanks, Soren ___ xorg@lists.freedesktop.org: X.Org support Archives: http://lists.freedesktop.org/archives/xorg Info: http://lists.freedesktop.org/mailman/listinfo/xorg
[ANNOUNCE] pixman release 0.17.8 now available
A new pixman release 0.17.8 is now available. This is a development
snapshot leading up to a stable 0.18.0 release.
= Trapezoids =
This release contains a change to the behavior of a1 trapezoids. When
an edge of a trapezoid falls exactly on a sample point, the point now is
considered inside the trapezoid if the edge is a top or left edge,
where before, the opposite rule was in effect.
The new rule is similar to the X11 rule for filled shapes.
If you have an application that depends on these rules, please test
this version of pixman to make sure your application still works. If
it doesn't, please file a bug at
https://bugs.freedesktop.org/enter_bug.cgi?product=pixman
= Pixman mailing list =
There is now a mailing list for pixman related issues. To subscribe,
or to look at the archives, please see this page:
http://lists.freedesktop.org/mailman/listinfo/pixman
= News in this release =
- New pixman_region{32}_init_from_image() function [Alexander Larsson]
- Follow X11 rules for trapezoids [Soren Sandmann]
- Reduced overhead for pixman_image_composite() [Soren]
Soren
tar.gz:
http://cairographics.org/releases/pixman-0.17.8.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.8.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.8.tar.bz2
Hashes:
MD5: a7deb2ff6b286b676d67aa6ae91317ae pixman-0.17.8.tar.gz
MD5: 512c7346c2afa41121967c99891484d1 pixman-0.17.8.tar.bz2
SHA1: 3c6daddba709f4ddda67ebbf9e38ba2d1bc77c5e pixman-0.17.8.tar.gz
SHA1: c9064a518f8638b9042d8f259302905c11e57129 pixman-0.17.8.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.17.8.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen sandm...@daimi.au.dk
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.17.8
Log:
Alexander Larsson (3):
Move SCREEN_SHIFT_LEFT/RIGHT to pixman-private.h
Add pixman_region{32}_init_from_image
Test pixman_region32_init_from_image in region-test
Makoto Kato (1):
Compile by USE_SSE2 only without USE_MMX
Søren Sandmann Pedersen (24):
Fix slightly skewed sampling grid for antialiased traps
Fix rounding of top and bottom coordinates.
Round horizontal sampling points towards northwest.
Post-release version bump
Make general_composite_rect() just another fast path.
Make fast_composite_scaled_nearest() another fast path.
Store a pointer to the array of fast paths in the implementation
struct.
Delete unused sources_cover() function
Don't call _pixman_implementation_composite() anymore.
Delete unused _pixman_walk_composite_region() function
Eliminate all the composite methods.
Optimize for the common case wrt. the workaround.
Move compositing to its own function, do_composite()
Move compositing functionality from pixman-utils.c into pixman.c
Manually inline _pixman_run_fast_path()
Move get_image_info() out of the loop in do_composite
Move region computation out of the loop in do_composite()
Move more things out of the inner loop in do_composite().
Add a fast path for non-repeating sources in
walk_region_internal().
Move NULL check out of get_image_info()
Move all code to do debugging spew into pixman-private.
Hide the C++ extern C declarations behind macros.
Add a1-trap-test
Pre-release version bump
___
xorg-announce mailing list
xorg-announce@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] pixman release 0.17.6 now available
A new pixman release 0.17.6 is now available. This is a development snapshot leading up to a stable 0.18.0 release. In this release, asserts are always turned off, so X server crashes in the region code should be gone (Bug 26314). News: 32 bit versions of some existing API [Benjamin Otte] pixman_fill_boxes32() pixman_image_composite_32() An accessor for the destroy data [Alexendar Larsson] pixman_image_get_destroy_data() Other news in this release [Søren Sandmann]: Optimizations for scaled images Some reorganisation of the fast path tables Plus a bug fix for ARM NEON detection [Siarhei Siamashka] Søren tar.gz: http://cairographics.org/releases/pixman-0.17.6.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.6.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.6.tar.bz2 Hashes: MD5: 32cbdf7a1ac1c275ee56230e26701456 pixman-0.17.6.tar.gz MD5: 643093dac601ee37f88ef73849c2f46b pixman-0.17.6.tar.bz2 SHA1: cf928635c29f476c67ac1c795617047ea86ab987 pixman-0.17.6.tar.gz SHA1: 5ea3ca7bfbdda7e5ace2ad8e9421bfb2ad40b9f1 pixman-0.17.6.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.17.6.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.17.6 Log: Alexander Larsson (2): Add extern C guards for c++ Add pixman_image_get_destroy_data() Benjamin Otte (5): Fix typo Make region argument to pixman_region(32)_init_rects() const Add pixman_image_composite32() Add pixman_image_fill_boxes() API Make pixman_image_fill_rectangles() call pixman_image_fill_boxes() Siarhei Siamashka (1): ARM: Remove any use of environment variables for cpu features detection Søren Sandmann Pedersen (17): Post-release version bump Fix some warnings Implement get_scanline_64() correctly for solid fill images. Turn some uint16_t variables to int32_t in the fast paths. Move the sanity checks for src, mask and destination into get_fast_path() Move pixbuf checks after src_format and mask_format have been computed. Consolidate the source and mask sanity checks in a function Fold get_fast_path() into _pixman_run_fast_path() Move calls to source_is_fastpathable() into get_source_format() Add src_, mask_, and dest_flags fields to fast path arrays Compute src, mask, dest flags and base fast path decisions on them. Move checks for src/mask repeat right before walking the region. Turn off asserts in development snapshots (bug 26314). Makefile.am: Remove 'check' from release-check Generalize and optimize fast_composite_src_scaled_nearest() Once unrolled version of fast_path_composite_nearest_scaled() Pre-release version bump ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
Re: [ANNOUNCE] xf86-video-qxl 0.0.11
Julien Cristau jcris...@debian.org writes: Hi Søren, your mails seem to have been corrupted on the way, or signed incorrectly (I guess due to the ø character in your name in the shortlog)... Yeah, I cutted and pasted it into my mail client. All kinds of things could have gone wrong in the process. I'll try doing better next time. Thanks, Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
[ANNOUNCE] pixman release 0.16.6 now available
A new stable pixman release 0.16.6 is now available with fixes for a couple of bugs related to alpha maps. Soren tar.gz: http://cairographics.org/releases/pixman-0.16.6.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.16.6.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.16.6.tar.bz2 Hashes: MD5: 17fed8b3888d38ec53ce35e0addfc34b pixman-0.16.6.tar.gz MD5: fbd6ee9cb09d0b26281557bc7afcb3ab pixman-0.16.6.tar.bz2 SHA1: 80226c9655e5a64af2d517175d4ec748b95f66fb pixman-0.16.6.tar.gz SHA1: a2beccb6b3bfbb55524a7a1c0b57354957b45af0 pixman-0.16.6.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.16.6.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk Git: git://git.freedesktop.org/git/pixman tag: pixman-0.16.6 Log: Søren Sandmann Pedersen (5): Post-release version bump When fetching from an alpha map, use the alpha map's fetch function. pixman_image_validate() needs to also validate the alpha map. When fetching from an alpha map, replace the alpha channel of the image Pre-release version bump ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] xf86-video-qxl 0.0.11
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Subject: [ANNOUNCE] xf86-video-qxl 0.0.11 To: xorg-announce@lists.freedesktop.org CC: x...@lists.freedesktop.org Alan Coopersmith (1): Fix build on OpenSolaris Søren Sandmann Pedersen (7): Store modes pointer out of qxl_map_memory() Do the undamage handling in qxl_copy_n_to_n() Remove some unused variables Only use QXL 16 bit modes when the screen depth is 15. Bump version number Eliminate some warnings about casting between u64 and pointers. Bump version number to 0.0.11 git tag: xf86-video-qxl-0.0.11 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.11.tar.bz2 MD5: 961d91ef8ac72f2cf86978015bbb862e xf86-video-qxl-0.0.11.tar.bz2 SHA1: 7536eafddda5b2975693d3d2fe33a20ec96113f2 xf86-video-qxl-0.0.11.tar.bz2 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.11.tar.gz MD5: 94755866c94b760047446bcc4ef67a5d xf86-video-qxl-0.0.11.tar.gz SHA1: 2e5f804f8c6f69685342f7562c237f6f449058a2 xf86-video-qxl-0.0.11.tar.gz -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (GNU/Linux) iEYEARECAAYFAktt8XkACgkQmxfmIW/3wajj2gCfe8U0oP0mg0rIG9iaPPdDufrJ XUMAn0xPjCoYAoFOYL+8OS4DdEChF5kB =+W3k -END PGP SIGNATURE- ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] xf86-video-qxl 0.0.12
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Subject: [ANNOUNCE] xf86-video-qxl 0.0.12 To: xorg-announce@lists.freedesktop.org CC: x...@lists.freedesktop.org Søren Sandmann Pedersen (3): Fixing 15/16 bit confusion accidentally broke 24 bits. Go back to drawing with QXL_EFFECT_OPAQUE instead of BLEND Version 0.0.12 git tag: xf86-video-qxl-0.0.12 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.12.tar.bz2 MD5: 542af8575b2870118786846c7d9d48c8 xf86-video-qxl-0.0.12.tar.bz2 SHA1: 4a190747a13da8ed6d645eefe2890e1908f16b3a xf86-video-qxl-0.0.12.tar.bz2 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.12.tar.gz MD5: 680383c588692dca55952a84f8b8f9e6 xf86-video-qxl-0.0.12.tar.gz SHA1: 0e8f8c1b1708abeeed7b2a299985b1cafbcbca2d xf86-video-qxl-0.0.12.tar.gz -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (GNU/Linux) iEYEARECAAYFAktt9+UACgkQmxfmIW/3waiq1gCaA/dSlMwcZ0tqmwLdGzrw4Sao 2HAAn3seEfyJxTcKnmzUPFcvL7uZ5jAp =GEW+ -END PGP SIGNATURE- ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] xf86-video-qxl 0.0.11
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Subject: [ANNOUNCE] xf86-video-qxl 0.0.11 To: xorg-annou...@lists.freedesktop.org CC: xorg@lists.freedesktop.org Alan Coopersmith (1): Fix build on OpenSolaris Søren Sandmann Pedersen (7): Store modes pointer out of qxl_map_memory() Do the undamage handling in qxl_copy_n_to_n() Remove some unused variables Only use QXL 16 bit modes when the screen depth is 15. Bump version number Eliminate some warnings about casting between u64 and pointers. Bump version number to 0.0.11 git tag: xf86-video-qxl-0.0.11 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.11.tar.bz2 MD5: 961d91ef8ac72f2cf86978015bbb862e xf86-video-qxl-0.0.11.tar.bz2 SHA1: 7536eafddda5b2975693d3d2fe33a20ec96113f2 xf86-video-qxl-0.0.11.tar.bz2 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.11.tar.gz MD5: 94755866c94b760047446bcc4ef67a5d xf86-video-qxl-0.0.11.tar.gz SHA1: 2e5f804f8c6f69685342f7562c237f6f449058a2 xf86-video-qxl-0.0.11.tar.gz -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (GNU/Linux) iEYEARECAAYFAktt8XkACgkQmxfmIW/3wajj2gCfe8U0oP0mg0rIG9iaPPdDufrJ XUMAn0xPjCoYAoFOYL+8OS4DdEChF5kB =+W3k -END PGP SIGNATURE- ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
[ANNOUNCE] xf86-video-qxl 0.0.12
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Subject: [ANNOUNCE] xf86-video-qxl 0.0.12 To: xorg-annou...@lists.freedesktop.org CC: xorg@lists.freedesktop.org Søren Sandmann Pedersen (3): Fixing 15/16 bit confusion accidentally broke 24 bits. Go back to drawing with QXL_EFFECT_OPAQUE instead of BLEND Version 0.0.12 git tag: xf86-video-qxl-0.0.12 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.12.tar.bz2 MD5: 542af8575b2870118786846c7d9d48c8 xf86-video-qxl-0.0.12.tar.bz2 SHA1: 4a190747a13da8ed6d645eefe2890e1908f16b3a xf86-video-qxl-0.0.12.tar.bz2 http://xorg.freedesktop.org/archive/individual/driver/xf86-video-qxl-0.0.12.tar.gz MD5: 680383c588692dca55952a84f8b8f9e6 xf86-video-qxl-0.0.12.tar.gz SHA1: 0e8f8c1b1708abeeed7b2a299985b1cafbcbca2d xf86-video-qxl-0.0.12.tar.gz -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.10 (GNU/Linux) iEYEARECAAYFAktt9+UACgkQmxfmIW/3waiq1gCaA/dSlMwcZ0tqmwLdGzrw4Sao 2HAAn3seEfyJxTcKnmzUPFcvL7uZ5jAp =GEW+ -END PGP SIGNATURE- ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: CopyArea : bug if areas overlap
ri...@happyleptic.org writes: There is a fastpath in fbBlt() using memcpy if the source and dest rasters are properly aligned. This fastpath rely on MEMCPY_WRAPPED, which is (sometime) a mere memcpy(). So, if the rasters do overlap the behavior is undefined. On my loongson (mips) at home, it segfaults, bus error or sometime hangs. This is easy to replace this memcpy by a memmove to solve this bug, but I wonder if these rasters are actually allowed to overlap ? They are allowed to overlap. The 'reverse' and 'upsidedown' arguments give the direction of the blt, precisely for this reason. So this is a real bug. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
[ANNOUNCE] pixman release 0.17.4 now available
A new pixman release 0.17.4 is now available. This is a development snapshot leading up to a stable 0.18.0 release. News: - Major performance improvements for ARM NEON [Siarhei Siamashka] - Test suite improvements [Marvin Schmidt] - Bug fixes [Benjamin Otte, Pierre-Loup A. Griffais, Søren Sandmann] Testing by KDE users is especially appreciated because there is a change in the behavior of alpha maps, which is a feature that is not used much outside of Qt. Thanks, Soren tar.gz: http://cairographics.org/releases/pixman-0.17.4.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.4.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.4.tar.bz2 Hashes: MD5: 734f70667d0a6454a1b1e332f8d50083 pixman-0.17.4.tar.gz MD5: fe103ff5f051f8f35c1d20b5c6d000bd pixman-0.17.4.tar.bz2 SHA1: 02176a0f4b19856f27aad410a89cb8dc8fa2c8e2 pixman-0.17.4.tar.gz SHA1: 85cfa05cc4ec2f4f09747c69903a2da52c344775 pixman-0.17.4.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.17.4.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.17.4 Log: Alan Coopersmith (1): Update Sun license notices to current X.Org standard form Benjamin Otte (3): Remove nonexistant function from header Fix bug that prevented pixman_fill MMX and SSE paths for 16 and 8bpp Make test program not throw warnings about undefined variables Marvin Schmidt (1): Build tests and run non-GTK+ ones on `make check` Pierre-Loup A. Griffais (1): Interpret the angle of a conical gradient in degrees. Siarhei Siamashka (18): test: support for testing pixbuf fast path functions in blitters-test ARM: NEON optimized pixman_blt ARM: added '.arch armv7a' directive to NEON assembly file ARM: minor source formatting changes ARM: added 'neon_composite_over___' fast path ARM: added 'neon_composite_over__8_' fast path ARM: added 'neon_composite_add__' fast path ARM: added 'neon_composite_add___' fast path ARM: added 'neon_composite_src_0565_' fast path ARM: added 'neon_composite_over_n_0565' fast path ARM: added 'neon_composite_over_n_' fast path ARM: added 'neon_src_0888__rev' fast path ARM: added 'neon_composite_src_0888_0565_rev' fast path ARM: added 'neon_composite_src_pixbuf_' fast path Use canonical pixman license notice for recently added ARM NEON assembly ARM: macro template for single scanline compositing functions ARM: added 'neon_combine_over_u' function ARM: added 'neon_combine_add_u' function Søren Sandmann Pedersen (15): Post-release version bump Reorder tests so that they fastest ones run first. Add 'check' to release-check make target mmx: Eliminate trailing whitespace. arm-simd: Whitespace fixes Some minor formatting fixes. region: Enable or disable fatal errors and selfchecks based on version nu fetch-test: Various formatting fixes fetch-test: Fix spelling error (pallete - palette) test: Add new alphamap test program. When fetching from an alpha map, use the alpha map's fetch function. pixman_image_validate() needs to also validate the alpha map. When fetching from an alpha map, replace the alpha channel of the image bits: Print an error if someone tries to create an image with bpp depth Pre-release version bump ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
Re: cairo performance and poulsbo driver
Johan Bilien j...@via.ecp.fr writes: So I'm left wondering where the overhead of the xlib backend comes from. If I run sysprof (profile attached) while running the trace (in the Composite disabled case), I can see that pixman gets only 27.5% of the CPU time, while 39.4% is spent in kernel. I'm vaguely thinking that these could be from moving pixmap data to and from the VRAM, but really I have no idea. Any idea on how to further investigate this? The latest version of sysprof (1.1.4) can trace into the kernel as well, but requires a 2.6.31 or later kernel though. If you don't have that, you may want to check out a7a1ab3081f58557a5997a667d713186f12 from sysprof git. That is the last version that still used a standalone module. It can also trace into the kernel. Profiling with oprofile shows that when running the firefox trace (xlib backend, Composite exa hook disabled), the CPU is idle a lot of the time. Maybe it's waiting for the GPU a lot? Here's the top of the profile: 1008233. vmlinux-2.6.24-24-lpia vmlinux-2.6.24-24-lpia acpi_idle_enter_bm I don't really know what this is, but I don't think oprofile reports CPU idle time like that. (Though I don't know oprofile very well). 3179 10.7207 vmlinux-2.6.24-24-lpia vmlinux-2.6.24-24-lpia get_page_from_freelist When I have seen this show up, it has been because of page faults, often triggered from pixman_fill_sse2(). 1662 5.6048 vmlinux-2.6.24-24-lpia vmlinux-2.6.24-24-lpia acpi_idle_enter_simple 1317 4.4414 libpixman-1.so.0.16.2libpixman-1.so.0.16.2 pixman_fill_sse2 1080 3.6421 libpixman-1.so.0.16.2libpixman-1.so.0.16.2 bits_image_fetch_transformed It doesn't show up very high, but FWIW, the pixman 0.17.2 development snapshot has faster transformed compositing, and there will likely be further improvements in a future 0.17.x release. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
[ANNOUNCE] Stable pixman release 0.16.4 now available
The only change since 0.16.2 is that asserts are now disabled by default, so that region bugs in the X server don't take down the entire desktop. Søren tar.gz: http://cairographics.org/releases/pixman-0.16.4.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.16.4.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.16.4.tar.bz2 Hashes: MD5: 7f430cf0de0bf05935d4e21906cb1e6b pixman-0.16.4.tar.gz MD5: 70c249a575ed02b7a5daa1bc9f9c6a4e pixman-0.16.4.tar.bz2 SHA1: 785dbd79818b5f9d62738d95c17e5af0c511dc4a pixman-0.16.4.tar.gz SHA1: 08fecbd065fa3792823818fcf715b97bc66d94cd pixman-0.16.4.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.16.4.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.16.4 Log: Søren Sandmann Pedersen (3): Post-release version bump region: Enable or disable fatal errors and selfchecks based on version nu Pre-release version bump ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] A new pixman development snapshot 0.17.2 is now available
A new pixman release 0.17.2 is now available. This is the first development release in the 0.17 series up to a stable 0.18.0 release. Highlights since the stable 0.16.2 version: - Faster bilinear image scaling [André Tupinambá] - Substantially improved ARM NEON fast paths [Siarhei Simashka] - Various other optimizations [Siarhei, Søren Sandmann] - Improved test suite [Chris Wilson] Søren A new pixman release 0.17.2 is now available tar.gz: http://cairographics.org/releases/pixman-0.17.2.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.2.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.17.2.tar.bz2 Hashes: MD5: 006c74f1450e5ed2e1bdaf88e9806faa pixman-0.17.2.tar.gz MD5: 64070b49e136f66640adf0f21c0b5fc8 pixman-0.17.2.tar.bz2 SHA1: 163e4e63a2f570690119ee40b8430eedb55ac8cc pixman-0.17.2.tar.gz SHA1: da17421f46b323af25906bb906cccd62f34799f6 pixman-0.17.2.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.17.2.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk Git: git://git.freedesktop.org/git/pixman tag: pixman-0.17.2 Log: Andrea Canciani (1): Fix composite on big-endian systems. André Tupinambá (2): Speed up bilinear interpolation. Add fast path scaled, bilinear fetcher. Benjamin Otte (3): Fix compile warnings Add default cases for all switch statements --enable-maintainer-mode is gone from configure, so remove it Chris Wilson (6): build: Suppress verbose compile lines [test] Add composite test from rendercheck [sse2] Don't emit prefetch 0 for an absent mask [build] Add rule to generate asm for inspection. [test] Exercise repeating patterns for composite. [Makefile] Set the SIMD specific CFLAGS for inspecting asm. Gaetan Nadon (1): Remove unused generated libcomp.pc #23801 Gerdus van Zyl (1): Fix build with Visual Studio 2008 Guillem Jover (1): pixman: Update .gitignore Jeremy Huddleston (1): This is not a GNU project, so declare it foreign. M Joonas Pihlaja (4): Work around differing _mm_prefetch() prototypes on Solaris. Work around a Sun Studio 12 code generation bug involving _mm_set_epi32() Default to optimised builds when using a Sun Studio compiler. Don't prefetch from NULL in the SSE2 fast paths. Makoto Kato (1): Add CPU detection for VC++ x64 Siarhei Siamashka (23): ARM: workaround for gcc bug in vshll_n_u8 intrinsic ARM: Remove fallback to ARMv6 implementation from NEON delegate chain Change CFLAGS order for PPC and ARM configure tests ARM: Removal of unused/broken NEON code blitters-test updated to also randomly generate mask_x/mask_y C fast path function for 'add_1000_1000' C fast path function for 'over_n_1_' ARM: removed old ARM NEON optimizations ARM: introduction of the new framework for NEON fast path optimizations ARM: enabled new implementation for pixman_fill_neon ARM: added 'bindings' for NEON assembly optimized functions ARM: enabled 'neon_composite_src_0565_0565' fast path ARM: enabled 'neon_composite_src__0565' fast path ARM: enabled 'neon_composite_src_0888_0888' fast path ARM: enabled 'neon_composite_over_n_8_0565' fast path ARM: enabled 'neon_composite_over_n_8_' fast path ARM: enabled 'neon_composite_over__n_' fast path ARM: enabled 'neon_composite_over__0565' fast path ARM: enabled 'neon_composite_over__' fast path ARM: enabled 'neon_composite_add_n_8_8' fast path ARM: enabled 'neon_composite_add_8_8_8' fast path ARM: enabled 'neon_composite_add_8000_8000' fast path C fast path function for 'over_n_1_0565' Søren Sandmann Pedersen (43): Post-release version bump Change names of add__8_8 fast paths to add_n_8_8 Enable the x888_8_ sse2 fast path. Fix off-by-one error in source_image_needs_out_of_bounds_workaround() Reformat test/composite.c to follow the standard coding style. Remove optimization for 0x and 0xff the add_n___ca fast p Revert Enable component alpha on solid masks. Return result from pixman_image_set_transform().
[ANNOUNCE] Stable pixman release 0.16.0 now available
A new major release 0.16.0 of the pixman library is now available. New features since the last stable release, 0.14.0, include: * Support for PDF blend modes [Benjamin Otte, Carlos Garcia Campos] * A number of fast paths for ARM NEON [Jeff Muizelaar, Jonathan Morton, Siarhei Siamashka, Ian Rickards] * Support for PIXMAN_REPEAT_REFLECT for images [Thomas Jaeger] * New image formats: a2r10g10b10, x2r10g10b10 [Ben Skeggs] b8g8r8a8, b8g8r8x8 [Mike Kettenis] * Support for MMX and SSE2 on Sun compilers and Solaris [Alan Coopersmith] * Support for SSE2 on Windows x64 [George Yohng] Also new in this release: * A reorganisation of pixman's internals to make future development and maintenance easier [Søren Sandmann Pedersen] * A substantial expansion of pixman's test suite [Siarhei Siamashka] Thanks to everyone who contributed to pixman 0.16.0, including Adam Jackson, Adrian Bunk, Alan Coopersmith, Andre Tupinamba, Ben Skeggs, Benjamin Otte, Carlos Garcia Campos, Chris Wilson, George Yohng, Guillaume Letellier, Ian Rickards, Jeff Muizelaar, Jonathan Morton, Loïc Minier, Luca Barbato, M Joonas Pihlaja, Magnus Kessler, Makoto Kato, Mark Kettenis, Michel Dänzer, Miha Vrhovnik, Siarhei Siamashka, Søren Sandmann Pedersen, Thomas Jaeger, and William Bonnet. Søren tar.gz: http://cairographics.org/releases/pixman-0.16.0.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.16.0.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.16.0.tar.bz2 Hashes: MD5: 46147b142ea530847e2228940f1cd535 pixman-0.16.0.tar.gz MD5: 183dea8a100c1d9732cdb2146ec90f68 pixman-0.16.0.tar.bz2 SHA1: 339112dc4d7b8d2d241b5077e18c46104cebc701 pixman-0.16.0.tar.gz SHA1: 7cf1f2261ba44b3c98cf3d2a32e9fd90b12ca19e pixman-0.16.0.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.16.0.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk Git: git://git.freedesktop.org/git/pixman tag: pixman-0.16.0 Log: Adam Jackson (1): Fix overflows during trap rasterization. [Bug 16560]. Adrian Bunk (2): Fix the NEON build after Convert CamelCase names to underscore_names. Fix NEON build for older ARM CPUs Alan Coopersmith (3): Fix MMX SSE intrinsics to work with Sun compilers Solaris Replace custom type ullong with standard uint64_t in pixman-mmx.c Add solaris-hwcap.mapfile to EXTRA_DIST Ben Skeggs (1): Add accessor functions for PIXMAN_x2r10g10b10 and PIXMAN_a2r10g10b10 Benjamin Otte (35): Add support for extended blend mode. First pass. use a pixman_op_t here Add INVERT and SUBTRACT blend modes used in Flash use PDF algorithm for soft-light fix ColorDodge and ColorBurn to conform to the PDF spec add nonseperable blend modes from PDF spec remove invert operator prefix HSL operators with HSL clean up seperable blend modes rewrite nonseperable blend modes the same way as seperable ones add component-alpha versions of the seperable blend-modes add non-seperable versions for component alpha invent a Subtract operator for component alpha remove a leftover debugging statement remove semicolon at end of macro typo fix: Seperable = Separable rework blend-mode documentation to match current code better rename operator SUBTRACT to FLASH_SUBTRACT fix component-alpha versions for seperable blend modes fix comment fix Multiply component-alpha version correct subtract implementation remove debugging leftovers fix component-alpha versions to set source to 0 when mask is 0 Fix burn and dodge operators to match acroread output Fix terms in comments describing the blend modes Clarify color dodge code - no semantical changes Clarify color burn code - no semantical changes Handle degenerate case in pixman_init_rect() Handle degenerate case in pixman_init_extents() Add a lot more sanity checks to region code Make the text when reporting a broken region more useful Simplify code that logs errors Log errors for invalid rectangles passed to region code Don't warn for empty rectangles, only degenerate ones Carlos Garcia Campos (5): [TEST] Update composite-test to test more operators Fix typo Do not use combineMask in component-alpha functions
[ANNOUNCE] pixman release 0.15.20 (Release Candidate) now available
A new pixman release 0.15.20 is now available. This is a Release Candidate for the 0.16.0 release. Unless bugs are reported in 0.15.20, this is what will become the stable 0.16.0. News since last time: - A new test program for compositing operators [Siarhei Siamashka] - Bugs and performance fixes [Benjamin Otte, Chris Wilson, Andre Tupinamba, Soren Sandmann] Please test, and report bugs at https://bugs.freedesktop.org/enter_bug.cgi?product=pixman Thanks, Soren tar.gz: http://cairographics.org/releases/pixman-0.15.20.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.20.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.20.tar.bz2 Hashes: MD5: 613c95e7ddc8069b7aa2708f93219b7d pixman-0.15.20.tar.gz MD5: 8cca7bdf45bd3c31cf95ecea0f05bddc pixman-0.15.20.tar.bz2 SHA1: a5f973b6895e269731964fbe328f61a8cbfdc931 pixman-0.15.20.tar.gz SHA1: 67a34a1fdc5431c9de5febd390057bdd9bb51df4 pixman-0.15.20.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.15.20.tar.gz.sha1.asc (signed by sandm...@daimi.au.dk) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.15.20 Log: Benjamin Otte (4): Make the text when reporting a broken region more useful Simplify code that logs errors Log errors for invalid rectangles passed to region code Don't warn for empty rectangles, only degenerate ones Chris Wilson (2): Fix inversion of radial gradients when r2 r1 Explain how we can simplify the radial gradient computation Siarhei Siamashka (3): Test program for stressing the use of different formats and operators HACK: updated test to better cover new neon optimizations One more update to blitters-test - use aligned memory Søren Sandmann Pedersen (56): Post-release version bump Replace a bunch of 'pict's with 'image' Rename source_pict_class_t to source_image_class_t Remove a couple of unused variables Add a CODING_STYLE document based on the one from cairo. Various updates to the CODING_STYLE document Only define PIXMAN_TIMERS if timers are actually enabled [bug 23169] Change bits_image_fetch_transformed() to work one pixel at a time. Make bits_image_fetch_nearest() return one pixel. Make the repeat routine work on only one coordinate at a time. Change bits_image_fetch_bilinear_pixels() to fetch one pixel at a time. Change bits_image_fetch_pixels_convolution() to fetch just one pixel. Change bits_image_fetch_alpha_pixels() to fetch just one pixel. Various renamings and clean-ups Add fetch_pixel_raw_32 and fetch_pixel_32 virtual functions. Change all the fetch_pixels() functions to only fetch one pixel. Misc formatting Remove unused function Remove leftover 0x in repeat() Move bounds checks for REPEAT_NONE to get_pixel() Add a fast path for the add_n__() operation. Add sse2 version of add_n__() Add a dirty bit to the image struct, and validate before using the image. Fix bugs in combine_over_reverse_ca(). Don't change the constant source in fast_composite_over_n__0565. Fix incorrect optimization in combine_over_ca(). Fix another case of changing the solid source. Fix bug in combine_mask_alpha_ca() Don't run fast paths if the format requires wide compositing. Fix a bunch of srca == 0 checks that should be src == 0 in pixman-mmx.c Rewrite the two-component arithmetic macros. Make pix_add_mul() in pixman-mmx.c produce exact results. Fix a couple more alpha==0 vs src==0 bugs in pixman-mmx.c Rename mmx_composite_add__8_8() to mmx_composite_add_n_8_8(). Fix a couple of alpha==0 vs src==0 issues in pixman-sse2.c Change the SSE2 versions of pix_add_multiply() to produce bit-exact resul Make pix_multiply bit-exact Fix bug in vmx_combine_xor_ca() Do the full four-component IN computation in vmx_combine_in_ca(). Fix vmx_combine_out_ca() In vmx_combine_out_ca() multiply with the alpha of the negated vdest. Fix vmx_combine_over_ca(). Fix vmx_combine_vmx_atop_ca() Multiply with the alpha of dest, not inverse alpha Fix vmx_over_reverse_ca() Fix vmx_combine_out_reverse_ca() Really
[ANNOUNCE] pixman beta release 0.15.18 now available
A new pixman release 0.15.18 is now available. This is a beta release leading up to a stable 0.16.0 release. News: - ARM bug fixes [Siarhei Siamashka] - No more asserts on malformed regions [Søren Sandmann Pedersen] - SSE2 support for 64 bit windows [George Yohng] - Improvements to the region code [Benjamin Otte] - Misc. bug fixes [Adrian Bunk, Chris Wilson, M Joonas Pihlaja, Miha Vrhovnik] Søren tar.gz: http://cairographics.org/releases/pixman-0.15.18.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.18.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.18.tar.bz2 Hashes: MD5: 9b60fe7623621b546b76777b8b7d2490 pixman-0.15.18.tar.gz MD5: 87d256db0440e579f113cd4fb67cbfb0 pixman-0.15.18.tar.bz2 SHA1: a4c99678679f3838de06afa8fa1ccd947fcbe33f pixman-0.15.18.tar.gz SHA1: 30d953ff32338740dfdba8764f4cc8545b96d370 pixman-0.15.18.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.15.18.tar.gz.sha1.asc (signed by sandm...@daimi.au.dk) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.15.18 Log: Adrian Bunk (1): Fix NEON build for older ARM CPUs Benjamin Otte (6): Fix terms in comments describing the blend modes Clarify color dodge code - no semantical changes Clarify color burn code - no semantical changes Handle degenerate case in pixman_init_rect() Handle degenerate case in pixman_init_extents() Add a lot more sanity checks to region code Chris Wilson (1): Fix read of BITS members from a solid image. George Yohng (1): Add implementation of MMX __m64 functions for MSVC x64. M Joonas Pihlaja (1): Check whether the linker understands the hwcap file before using it. Miha Vrhovnik (1): Update Makefile.win32 to make it work again. Siarhei Siamashka (4): ARM: Use Ian's variant of 'neon_composite_over_n_8_0565' function again ARM: Commented out the rest of buggy NEON optimizations ARM: Workaround cs2007q3 compiler bug for NEON registers clobber list ARM: Fixes for the inline assembly constraints in pixman_fill_neon Søren Sandmann Pedersen (19): Post-release version bump Add -fno-strict-aliasing Change name of macro from RGB16_TO_ENTRY to RGB15_TO_ENTRY Various minor formatting changes Change composite test to use a rainbow gradient as source. Fix combine_src_ca() to fill out all of the destination line. Add back check for need_workaround that got removed during reindenting Move read and write functions to the bits_image_t struct. Align the stack in _pixman_implementation_create_sse2() Replace // comments with /* */ comments in various places Some formatting changes to pixman-arm-neon.c Fix typo in sse2_combine_over_pixbuf_0565() Add test case for bug 22844. Rework the workaround for bogus X server images. Only apply the workaround if the clip region extends beyond the drawable. Fix search-and-replace issue pointed out by Koen Kooi. Fix another search and replace issue Don't assert when malformed regions are detected. Pre-release version bump ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] pixman release 0.15.14 now available
A new pixman release 0.15.14 is now available. This is a development
release leading up to a stable 0.16.0 release.
News:
- Support for PDF blend modes [Benjamin Otte, Carlos Garcia Campos].
- NEON updates [Jonathan Morton].
- Fix for an out-of-bounds memory read [Søren Sandmann Pedersen].
- Refactoring and code cleanup [Søren Sandmann Pedersen].
Testing is appreciated as always. If you are a distribution, consider
adding this release to your development version if you have one.
If find a bug, please either send mail to ca...@cairographics.org, or
file a bug at:
https://bugs.freedesktop.org/enter_bug.cgi?product=pixman
Thanks,
Søren
tar.gz:
http://cairographics.org/releases/pixman-0.15.14.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.14.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.14.tar.bz2
Hashes:
MD5: bfbb075dd60402e86528add13f8e27ad pixman-0.15.14.tar.gz
MD5: bb98ff30947c171f5e396b66f5714e2d pixman-0.15.14.tar.bz2
SHA1: 46065c0fee9f8c5bf63ce68ca1e03ff1ec7c2f80
pixman-0.15.14.tar.gz
SHA1: a3155f5e286c11031adcf0056eee8a2b3c1f00b8
pixman-0.15.14.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.15.14.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen sandm...@daimi.au.dk)
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.15.14
Log:
Benjamin Otte (24):
Add support for extended blend mode. First pass.
use a pixman_op_t here
Add INVERT and SUBTRACT blend modes used in Flash
use PDF algorithm for soft-light
fix ColorDodge and ColorBurn to conform to the PDF spec
add nonseperable blend modes from PDF spec
remove invert operator
prefix HSL operators with HSL
clean up seperable blend modes
rewrite nonseperable blend modes the same way as seperable ones
add component-alpha versions of the seperable blend-modes
add non-seperable versions for component alpha
invent a Subtract operator for component alpha
remove a leftover debugging statement
remove semicolon at end of macro
typo fix: Seperable = Separable
rework blend-mode documentation to match current code better
rename operator SUBTRACT to FLASH_SUBTRACT
fix component-alpha versions for seperable blend modes
fix comment
fix Multiply component-alpha version
correct subtract implementation
remove debugging leftovers
fix component-alpha versions to set source to 0 when mask is 0
Carlos Garcia Campos (5):
[TEST] Update composite-test to test more operators
Fix typo
Do not use combineMask in component-alpha functions
Use floating point in SetLum
Fix BlendColorBurn
Jonathan Morton (3):
[NEON] Replace Ian's glyph-blitter with a better one.
[NEON] Add transparent rect blitter.
[NEON] Add ARGB8-over-RGB565 compositing blitter.
M Joonas Pihlaja (1):
Remove redundant NULL checks from general_composite_rect().
Michel Dänzer (1):
Fix the build on big endian machines.
Søren Sandmann Pedersen (92):
Post-release version bump
Eliminate trivial READ and WRITE macros in pixman-fast-path.c
Remove dstMask from pixman-fast-path.c
Remove unused pixmanFetchGradient() declaration
Delete FbComposeData type
Delete empty FASTCALL macro
Rename bits_image.fetch_pixels_{32,64} to
fetch_pixels_raw_{32,64}
Delete unused CombineFunc{32,64} types
Remove useless FbBits typedef
Remove commented-out fbAddTriangles
Delete unused IS_SOURCE_IMAGE() macro
Rename GradientWalker to pixman_gradient_walker_t
Rename PixmanTimer to pixman_timer_t
Get rid of pixman_region_internal_set_static_pointers()
Rename pixman-pict.c to pixman.c
Only use force_align_arg_pointer on gcc/x86-32
Rename combine.inc and combine.h.inc to
pixman-combine.{c,h}.template.
Eliminate MSBFirst, LSBFirst, IMAGE_BYTE_ORDER, and
BITMAP_BIT_ORDER.
Rename FastPathInfo to pixman_fast_path_t
Move compiler dependencies to a new pixman-compiler.h file
Move edge utilities into pixman-trap.c
Use DIV instead of _div In pixman-trap.c.
Move FbGen macro into pixman-combine.c
Implement pixman_format_supported_destination() in
Re: Current tinderbox regression (pixman, ppc)
Chris Ball c...@laptop.org writes: http://tinderbox.x.org/builds/2009-05-31-0014/logs/pixman/#build pixman-fast-path.c: In function 'fbOver': pixman-fast-path.c:42: warning: implicit declaration of function 'FbByteMulAdd' [...] ../pixman/.libs/libpixman-1.so: undefined reference to `FbByteMulAdd' This seems to have fixed itself. The next build succeeded even though nothing changed. FWIW, FbByteMulAdd is a macro defined in a generated file pixman-combine32.h. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Current tinderbox regression (pixman, ppc only)
Chris Ball c...@laptop.org writes: Hi, pixman-vmx.h:40: error: 'vmx_fast_path_array' undeclared here (not in a function) I think this patch is sufficient; the assignment happens in pixman-vmx.c. Yeah, please push. Thanks, Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
[ANNOUNCE] pixman release 0.15.4 now available
A new pixman release 0.15.4 is now available. This is a development release leading up to a stable 0.16.0 release. News: - New ARM NEON fast paths [Ian Rickards, Jeff Muizelaar] - The first part of a large refactoring of the pixman code base [Søren Sandmann] - Bug fixes [Adam Jackson, Alan Coopersmith, Siarhei Siamashka, Jonathan Morton] Testing is extra appreciated for this and the upcoming 0.15.6 and 0.15.8 releases since the refactoring makes new bugs more likely than usual. Thanks, Søren tar.gz: http://cairographics.org/releases/pixman-0.15.4.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.4.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.15.4.tar.bz2 Hashes: MD5: 243255cdcee641675e00649f1b048338 pixman-0.15.4.tar.gz MD5: 1464bf6f422a4d1747ed3273c7a531d3 pixman-0.15.4.tar.bz2 SHA1: 0936126b4d7e9148e5a54dc2687c63c92b1fe93f pixman-0.15.4.tar.gz SHA1: 1388fcb88b51df27404fadcf32cabebf8a4992e0 pixman-0.15.4.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.15.4.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen sandm...@daimi.au.dk) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.15.4 Log: Adam Jackson (1): Fix overflows during trap rasterization. [Bug 16560]. Alan Coopersmith (1): Add solaris-hwcap.mapfile to EXTRA_DIST Ian Rickards (1): Add support for ARM NEON fast paths Jeff Muizelaar (10): Add support for doing ARM simd detection on Windows Add support for doing ARM simd detection on Windows Enable fbCompositeSrcAdd_x8x8neon Enable fbCompositeSrcAdd_8000x8000neon Enable fbCompositeSrc_x888x0565neon Enable fbCompositeSrc_xneon Enable fbCompositeSrc_x8xneon Enable fbCompositeSolidMask_nx8x0565neon Enable fbCompositeSolidMask_nx8xneon Enable NEON assembly when we can build it Jonathan Morton (1): Test USE_GCC_INLINE_ASM instead of USE_NEON_INLINE_ASM. Siarhei Siamashka (2): Updated scaling test to support rgb565, source clipping, etc. Fixed rendering bug for source alpha == 0 in OVER fastpath functions Søren Sandmann Pedersen (53): Post-release version bump Add trap-crasher.c test program Include string.h and fix warning in trap-test.c Move all the GTK+ test code into its own file, utils.c Mention utils.h in test/Makefile.am so that make dist will pick it up Add refactoring notes More refactoring notes Move SourcePictClassify into pixman-image.c Move calls of the classification out of the if statements. Add notes on how Render actually works More refactoring notes Get rid of SOURCE_IMAGE_CLASS_NEITHER again Factor out duplicated fetcher finding code Formatting Simplify logic to compute store function Move fetch logic to pixman-image.c Move store logic into pixman-image.c Remove unused xoff/yoff variables Remove accessor version of pixman-compose.c Make combineC function type take const args Change type of combineU to take a mask Get rid of separate combineU and combineC types Have the generic code go through the component alpha path always Use memcpy() in fbCombineSrcU when there is no mask Implement the sse2 combiners with masks (disabled) Enable sse2CombineOverU Enable sse2CombineOverReverseU Enable sse2CombineInU Enable sse2CombineInReverseU Enable sse2CombineOutU Enable sse2CombineOutReverseU Enable sse2CombineAtopU Enable sse2CombineAtopReverseU Enable sse2CombineAtopXorU Enable sse2CombineAddU Enable sse2CombineSaturateU Implement the mmx combiners with masks (disabled) Enable mmxCombineOverU Enable mmxCombineOverReverseU Enable mmxCombineInU Enable mmxCombineInReverseU Enable mmxCombineOutU Enable mmxCombineOutReverseU Enable mmxCombineAtopU Enable mmxCombineAtopReverseU Enable mmxCombineXorU Enable mmxCombineAddU Enable mmxCombineSaturateU Optimize source reading in combine4 Remove unused combineMaskU functions Fix
Re: Xrandr loop with gnome-settings-daemon [WAS: Re: Intel GM45: Loop of continuously triggered output detections]
Peter Clifton pc...@cam.ac.uk writes: Should gnome-settings-daemon be avoiding retaliating to a notification by requesting XRRGetScreenSizeRange, or should XRRGetScreenSizeRange avoid calling a procedure which will emit another notification? I'm pretty sure gnome-settings-daemon is doing what it's supposed to do here. RandR is designed so that clients are supposed to update their information in response to notifications. It's been a while since I looked at it though. I don't think XRRGetScreenSizeRange should generate notifications. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: [PATCH] Don't treat PIXMAN_TYPE_YUY2 and PIXMAN_TYPE_YV12 as PIXMAN_FORMAT_COLOR.
Aaron Plattner aplatt...@nvidia.com writes: Various pieces of code expect PIXMAN_FORMAT_COLOR (and its less cool older brother, PICT_FORMAT_COLOR) formats to have ARGB bits, and the YUV formats do not. Looks good to me. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Blend modes take 3
Benjamin Otte [EMAIL PROTECTED] writes:
On Tue, Nov 25, 2008 at 10:10 PM, Soeren Sandmann [EMAIL PROTECTED] wrote:
- When mask is 0, there is no reason to read the source.
- There is no reason to read the destination if the inverse
combined src/mask is 0
Don't compilers take care of that?
The code looks better to me if the variable assignments all occur in one
place.
- The continue is incorrect; it needs to write out 0 in that case,
since the intermediate buffer is not zero-filled.
Oh, the code is supposed to modify the _source_ parameter, too?
Someone really needs to document what component alpha does properly so
I really grok it.
Fixed.
No, what I meant was that the destination buffer would need to be
modified in that case too, but that was just a thinko on my part.
However, I don't think the new code is correct either. In order to
skip the computation, it needs to ensure that both m and s are 0. So
let's not do this optimization for now.
Another problem I forgot to mention: fbByteMulAddC() produces a result
in its first argument, not its last.
if (min 0) {
tmp[0] = l + (tmp[0] - l) / 4 * l / (l - min) * 4;
tmp[1] = l + (tmp[1] - l) / 4 * l / (l - min) * 4;
tmp[2] = l + (tmp[2] - l) / 4 * l / (l - min) * 4;
}
if (max a) {
tmp[0] = l + (tmp[0] - l) / 4 * (a - l) / (max - l) * 4;
tmp[1] = l + (tmp[1] - l) / 4 * (a - l) / (max - l) * 4;
tmp[2] = l + (tmp[2] - l) / 4 * (a - l) / (max - l) * 4;
}
Where do those 4s come from?
Overflow protection. The tmp value can range from -COMP2_T_MAX .. 2 *
COMP2_T_MAX (I think), so I took the shortcut to just divide by a high
enough number to avoid it. Also, this code will break on 64bit cases
as I'm using ints there for lack of a signed comp4_t type.
It's one of the cases I asked about previously on IRC I think as I was
unsure if this is a case for doubles or how it should best be
handled.
I would be more comfortable with floating points, yes. In the
'blendmode' branch here:
git://freedesktop.org/~sandmann/pixman/
I fixed some of the above things and added a long comment about the
linearity of the HSL functions. Here is the comment:
/* For premultiplied colors, we need to know what happens when C is
* multiplied by a real number. Lum and Sat are linear:
*
*Lum (r * C) = r * Lum (C) Sat (r * C) = r * Sat (C)
*
* If we extend ClipColor with an extra argument a and change
*
*if x = 1.0
*
* into
*
*if x = a
*
* then ClipColor is also linear:
*
*r * ClipColor (C, a) = ClipColor (rC, ra);
*
* for positive r.
*
* Similarly, we can extend SetLum with an extra argument that is just passed
* on to ClipColor:
*
* r * SetLum ( C, l, a)
*
* = r * ClipColor ( C + l - Lum (C), a)
*
* = ClipColor ( r * C + r * l - r * Lum (C), r * a)
*
* = SetLum ( r * C, r * l, r * a)
*
* Finally, SetSat:
*
*r * SetSat (C, s) = SetSat (x * C, r * s)
*
* The above holds for all non-zero x, because they x'es in the fraction for
* C_mid cancel out. Specifically, it holds for x = r:
*
*r * SetSat (C, s) = SetSat (rC, rs)
*
*/
/* So, for the non-separable PDF blend modes, we have (using s, d for
non-premultiplied
* colors, and S, D for premultiplied:
*
* Color:
*
* a_s * a_d * B(s, d)
* = a_s * a_d * SetLum (S/a_s, Lum (D/a_d), 1)
* = SetLum (S * a_d, a_s * Lum (D), a_s * a_d)
*
*
* Luminosity:
*
* a_s * a_d * B(s, d)
* = a_s * a_d * SetLum (D/a_d, Lum(S/a_s), 1)
* = SetLum (a_s * D, a_d * Lum(S), a_s * a_d)
*
*
* Saturation:
*
* a_s * a_d * B(s, d)
* = a_s * a_d * SetLum (SetSat (D/a_d, Sat (S/a_s)), Lum (D/a_d), 1)
* = SetLum (a_s * a_d * SetSat (D/a_d, Sat (S/a_s)), a_s * Lum (D), a_s *
a_d)
* = SetLum (SetSat (a_s * D, a_d * Sat (S), a_s * Lum (D), a_s * a_d))
*
* Hue:
*
* a_s * a_d * B(s, d)
* = a_s * a_d * SetLum (SetSat (S/a_s, Sat (D/a_d)), Lum (D/a_d), 1)
* = a_s * a_d * SetLum (SetSat (a_d * S, a_s * Sat (D)), a_s * Lum (D), a_s
* a_d)
*
*/
A careful review of both the code in the branch and the math in the
comment would be appreciated.
The things remaining to be done on top of that branch:
- Review that what I did makes sense
- Use floating point in the HSL modes
- Getting rid of component alpha versions of HSL and
return_if_fail()ing if you to use them with a component
alpha mask. [1]
- There is an #undef Set which should be #undef Sat
Soren
[1] They could be done in ways that would be mathematically consistent
with the other blend modes, but I don't know that it would make any
visual sense. I could be convinced otherwise, though.
___
xorg mailing list
xorg@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg
Re: X server 1.6_beta1 pending pixman
Keith Packard [EMAIL PROTECTED] writes: So, we'll see if we can't get a bit of pixman review and perhaps a pixman release done tomorrow so that the X server beta can head out. Here are some comments on the matrix code. I didn't review all the numericals, but nothing jumped out at me either. - I'd like to have the interface const correct, for example in pixman_transform_multily (struct pixman_transform_t *dst, struct pixman_transform_t *l, struct pixman_transform_t *r); l and r could be const - For the rotation interfaces, maybe expand the names s and c to sin and cos? I first thought c meant center and was then mystified what s could mean. - The interfaces that take forward/reverse matrices should probably accept NULL's. - Pixman's version numbering scheme is similar to cairo's: The git master version has an odd micro number, released versions have even micro numbers. - The name pixman_f_transform bothers me, but I don't have a better suggestion since pixman_transformf would be worse. - Is there any particular reason for the fixed point epsilon of 2 (as opposed to 1)? - There is a comment about floating point interfaces, but the file contains both fixed and floating point interfaces. Thanks, Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: X server 1.6_beta1 pending pixman
Keith Packard [EMAIL PROTECTED] writes: - Pixman's version numbering scheme is similar to cairo's: The git master version has an odd micro number, released versions have even micro numbers. What version would you like? Before releasing, bump to 0.13.2, after releasing, bump to 0.13.3. Or if you don't want to release, but still want a number to depend on, just bump to 0.13.3. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: [PATCH] Export a bunch of matrix operations from render.
Keith Packard [EMAIL PROTECTED] writes: The render extension uses many matrix operations internally, this change exposes those functions to other parts of the server, drivers and extensions. The change is motivated by the 'transform' additions to the RandR extension but will likely be useful elsewhere. Can we put these in pixman and add a dependency instead? pixman_transform_point3d is already there; the only reason I left it in the server was for ABI reasons. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: [RFC] RandR 1.3 properties
Matthias Hopf [EMAIL PROTECTED] writes: ATOMs are obviously supported, but FLOATs seem harder as they aren't described in the core protocol anywhere. Thinking about that, adding floats was probably a bull idea. However, having semantics about ATOMs might be helpful (e.g. for xrandr or any general purpose property setting tool). We may need them in XRender as well to support higher-precision coordinates and transformations. They could be specified by just referencing IEEE 754. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: pixman 0.11.6 license
☂Josh Chiα (谢任中) [EMAIL PROTECTED] writes: Taking pixman-0.12.0, going through the .c, .h and .in files in the pixman directory, I extracted the following notices. Thanks a lot for doing this. Only the .c, .h and .in files have the word 'copyright' in them. I'm not setup to checkin. Could someone please put this in COPYING, which is currently empty? You can commit to your local git repository, then send a patch. If you do, I'll check it in. It's probably a good idea to model the COPYING file after the one in the xserver repository: http://cgit.freedesktop.org/xorg/xserver/tree/COPYING Thanks , Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Render spec clarification
Adam Jackson [EMAIL PROTECTED] writes: On Mon, 2008-10-13 at 11:19 -0700, Keith Packard wrote: On Mon, 2008-10-13 at 13:50 -0400, Adam Jackson wrote: like a copy semantics of Option B to make the results predictable. The core protocol uses 'like a copy' for CopyArea/CopyPlane. Actually creating a copy would be fairly simple inside the Render code -- just make a copy of the destination and use that as whichever source operands match the destination. Either or both source operands, remember. If the only real use of src/mask=dst is to support scrolling, then maybe XCopyArea is good enough. The only real drawback I can see is that XCopyArea is then one of the only remaining legitimate uses of GC's. Considering the GraphicsExpose issue below, this is probably a vote for BadMatch. Clients can always make a copy themselves if they need the feature. I'm not strongly opposed to like-a-copy though. The other possibly unpleasant bit about Option B is it introduces more cases where the server will need to do significant allocation to satisfy rendering. Probably not a huge deal, but worth noting. A pain, to be sure. If we do this, we would want to make sure the driver could recognize scrolling cases and make those work correctly. And, make GraphicsExpose events actually work too. My thinking there was that we'd just smash onto the CopyArea path like we do already. Technically (a SRC b) is more general than a plain blit since they can be different Picture formats pointing to the same Pixmap, but that seems like something we should just not allow for the scrolling loophole. I get what GraphicsExpose events are supposed to be for, but not really how the region is defined. It does seem like CopyArea's returned region is approximately the same thing though. Unfortunately it's more complicated than that because Render has repeating pictures. Suppose a repeating picture points to a partially obscured window; then the GraphicsExposed region must be a repeated copy of the hierarchy clip. Repeat here can be normal, reflect or pad and each will need to be handled differently by the GraphicsExpose code. Clipping is already quite complicated in pixman, so if possible, I'd prefer to avoid GraphicsExposes. I don't see using window sources as interesting except in compositing managers where the pixels are guaranteed to be available. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: pixman with and without SSE2 benchmarks?
Juliusz Chroboczek [EMAIL PROTECTED] writes: Note that this is on core 2, which has magnificent SSE performance. I haven't tried it, but I'm willing to bet that on Pentium-M and Pentium 4, the results are similar for SSE and MMX. Actually they are not, but that's because gcc 4 generates much worse code for MMX than it does for SSE2. The generated MMX code includes bizarre stuff like this: movq%mm0, -24(%ebp) movq-24(%ebp), %mm0 psllq $16, %mm0 movq-24(%ebp), %mm1 Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: pixman with and without SSE2 benchmarks?
Matthieu Herrb [EMAIL PROTECTED] writes: Hi Does someone have some benchmarks or numbers from real life X usage showing the benefits of using the SSE2 code in pixman 0.12.0 over non SSE2 on x86 and/or x86_64 cpus? Not from real-life X usage, but from the cairo performance test suite. See this thread: http://lists.cairographics.org/archives/cairo/2008-April/013610.html There is a summary of the performance difference compared to the MMX code: http://www.daimi.au.dk/~sandmann/perf-diff I'm asking because OpenBSD is stuck with an old gcc version which can't compile the SSE2 code, and I wonder if it's worth hacking something to be able ship a version compiled by gcc 4.2 or later with SSE2 support with OpenBSD binary distributions. Part of the performance win of the SSE2 code is that gcc has regressed quite badly for MMX code since 3.4, so the performance difference above may apply to your version of gcc. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
Re: Optimization idea: soft XvPutImage
Adam Richter [EMAIL PROTECTED] writes: Even if you do not want to do stretch, I believe that the X Render extension would require first copying the YUV data to a drawable and then doing a drawable-drawable block transfer operation to do the YUV transformation. In comparison, XvPutImage is a single call takes an XImage, which can be in shared memory, and would normally be in YUV, and specifies the YUV-RGB conversion and stretch in a single operation. As other people pointed out, XRender does allow arbitrary 3x3 transformations of source images, but you are right that the XRender protocol would require putting the data in a drawable first. A shared memory pixmap would be a possibility, perhaps, though the shared memory extension should eventually be replaced with something based on the DRM memory manager. Soren ___ xorg mailing list xorg@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg
[ANNOUNCE] Stable pixman release 0.12.0 now available
A new major release 0.12.0 of the pixman library is now available. New features since 0.10.0 include: - New image formats with 10 bits per channel (Aaron Plattner) - SSE2 optimizations for many operations (André Tupinambá) - Opacity based optimizations for operators (Antoine Azar) - Altivec optimizations for many operations (Luca Barbato) - A 32 bit region implementation (Søren Sandmann) Thanks to everyone who contributed to this release, including Aaron Plattner, Alan Coopersmith, André Tupinambá, Antoine Azar, Behdad Esfahbod, Benjamin Otte, Damien Carbery, David Sharp, Dimitrios Apostolou, Frédéric Plourde, Ginn Chen, Jeff Muizelaar, Julien Cristau, Luca Barbato, Luo Jinghua, Maximilian Grothusmann, Oswald Buddenhagen, Peter O'Gorman, Richard Hult, Søren Sandmann, Thomas Zimmermann, and Vladimir Vukicevic. Søren tar.gz: http://cairographics.org/releases/pixman-0.12.0.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.12.0.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.12.0.tar.bz2 Hashes: MD5: 09357cc74975b01714e00c5899ea1881 pixman-0.12.0.tar.gz MD5: 494af78c1c7d825c9ad6815d7b91f17d pixman-0.12.0.tar.bz2 SHA1: 595487766c3814afeaaf18e1021b312dd546dfec pixman-0.12.0.tar.gz SHA1: 0c57c6c7e21e2545257068f52517c460e6e51ae4 pixman-0.12.0.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.12.0.tar.gz.sha1.asc (signed by Søren Sandmann Pedersen [EMAIL PROTECTED]) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.12.0 Log: Aaron Plattner (22): Fix test build when srcdir != builddir. Get rid of pixman_composeFunctions_accessors. Move combining macros into pixman-combine.c. Move combining routines into combine.inc and add a Perl rule to generate Tack 32 onto the ends of the combining function types. Replace hardcoded numbers and uint*_t in combine.inc. Generate 64-bit combining functions. Use a macro to append _accessors to things. Move the scanline buffer allocation logic into pixman_composite_rect_gene Split fetch/fetchPixel/store proc types into 32-bit and 64-bit versions. Add infrastructure for allocating wide scanline buffers. Not yet used. Use wide compositing functions when wide == 1. Fix wide alpha fetch macro. Add expand and contract functions to convert between ARGB8 and ARGB16. Add wide fetch/store functions. Add wide source picture, external alpha, and transformed image routines. Make expansion and contraction loops clearer. Take the source format into account in pixman_expand. Decide based on the image formats whether we need wide compositing. Use pixman_malloc_ab instead of plain malloc for the fbStore64_generic sc Add a lossy 32-bit generic pixel fetch function for wide surfaces. Add depth 30 formats to pixman_format_supported_source. Alan Coopersmith (1): Fix pixman/Makefile.am to work with Solaris make Andre Tupinamba (1): [sse2] Change pixman-sse to pass __mm128i args as pointers, so as not to André Tupinambá (4): Add SSE2 implementations of many compositing operations. Fix bug 16310 in the SSE2 fast path in function fbCompositeSolidMask_nx88 Fix SSE2 bug where x888 pixels were treated as Use CopyAreasse2, plus a compatibility fix Antoine Azar (1): Optimize operators based on source or dest opacity. Behdad Esfahbod (1): Fix fbCompositeSrc_x888xnxmmx to properly ignore source alpha Benjamin Otte (1): fix Altivec detection Damien Carbery (1): Add pixman-1-uninstalled.pc file David Sharp (1): pixman-sse.c: silence pointer-cast compiler warnings. Dimitrios Apostolou (1): Fix compilation on SGI Frederic Plourde (2): Add CopyAreammx fast path for argb32 SRC xrgb32 and abgr32 SRC xbgr32 Win32 Makefile fix Frédéric Plourde (1): Win32 build system fixes Ginn Chen (1): Use hidden attribute for private functions when compiling with Sun Studio Jeff Muizelaar (2): Comment FbByteMulAdd Speed up fbOver Julien Cristau (7): Remove prototype for nonexistent pixman_image_set_filter_params pixman-version.h is generated, don't distribute it Revert pixman-version.h is
[ANNOUNCE] pixman release 0.11.8 now available
A new pixman release 0.11.8 is now available.
This is mainly a bugfix release. See the short-log below for the details.
Søren
tar.gz:
http://cairographics.org/releases/pixman-0.11.8.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.11.8.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.11.8.tar.bz2
Hashes:
MD5: 7b5db768c51337b5e5e954fc9c961cd3 pixman-0.11.8.tar.gz
MD5: 57949578526a796c724d0c79b045eff0 pixman-0.11.8.tar.bz2
SHA1: 9117a904c81673c728e4f78f18600602d0b77adc pixman-0.11.8.tar.gz
SHA1: 8db43d9fe094073cfcb03c1b1c86296b5a7df1aa pixman-0.11.8.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.11.8.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen [EMAIL PROTECTED])
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.11.8
Log:
Aaron Plattner (1):
Add a lossy 32-bit generic pixel fetch function for wide surfaces.
André Tupinambá (2):
Fix bug 16310 in the SSE2 fast path in function
fbCompositeSolidMask_nx88
Fix SSE2 bug where x888 pixels were treated as
Behdad Esfahbod (1):
Fix fbCompositeSrc_x888xnxmmx to properly ignore source alpha
Benjamin Otte (1):
fix Altivec detection
Damien Carbery (1):
Add pixman-1-uninstalled.pc file
Jeff Muizelaar (2):
Comment FbByteMulAdd
Speed up fbOver
Julien Cristau (2):
make --{en,dis}able-gtk work correctly
fix --enable-{mmx,sse2,vmx}
Søren Sandmann (5):
Only export the 16 bit version of
pixman_region_set_static_pointers()
Unexport or delete various functions
Remove unused pixman-combine.c
Update TODO
Use -no-undefined -- fixes bug 15927
Søren Sandmann Pedersen (8):
Post-release version bump
Explicitly use my GPG key for signing.
Update release instructions
Update versioning documentation
Update TODO
Use inttypes.h on most types of Unix.
Make a couple of functions static
Pre-release version bump
___
xorg-announce mailing list
xorg-announce@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] pixman release 0.11.6 now available
A new pixman release 0.11.6 is now available. This is a development
release leading up to a stable 0.12.0 release.
Notable news:
- Support for new image formats with 10 bits per channel,
including infrastructure to composite with 32 bit
intermediate precision. (Aaron Plattner)
- Fix for a bad memory leak. (Maximilian Grothusmann)
tar.gz:
http://cairographics.org/releases/pixman-0.11.6.tar.gz
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.11.6.tar.gz
tar.bz2:
http://xorg.freedesktop.org/archive/individual/lib/pixman-0.11.6.tar.bz2
Hashes:
MD5: 1b404bb31bf419a86531927492b5b95d pixman-0.11.6.tar.gz
MD5: 7d619cd89b05fefb87ca006ca6788a9e pixman-0.11.6.tar.bz2
SHA1: f79e03dc9efa01585a09006c662db637eec3c10d pixman-0.11.6.tar.gz
SHA1: 9ee97f75d98678d50d4a400a9266bf05d3835481 pixman-0.11.6.tar.bz2
GPG signature:
http://cairographics.org/releases/pixman-0.11.6.tar.gz.sha1.asc
(signed by Søren Sandmann Pedersen)
Git:
git://git.freedesktop.org/git/pixman
tag: pixman-0.11.6
Log:
Aaron Plattner (7):
Add expand and contract functions to convert between ARGB8 and
ARGB16.
Add wide fetch/store functions.
Add wide source picture, external alpha, and transformed image
routines.
Make expansion and contraction loops clearer.
Take the source format into account in pixman_expand.
Decide based on the image formats whether we need wide
compositing.
Use pixman_malloc_ab instead of plain malloc for the
fbStore64_generic sc
David Sharp (1):
pixman-sse.c: silence pointer-cast compiler warnings.
Luo Jinghua (1):
Fix implicit declaration of function 'free'.
Maximilian Grothusmann (1):
Fix memory leak by freeing boxes{16,32}.
Søren Sandmann Pedersen (6):
Post-release version bump
Move PIXMAN_FORMAT_16BPC to pixman-private.h
Update TODO
Add configure time options to disable mmx/sse2/vmx
TODO
Pre-release version bump
___
xorg-announce mailing list
xorg-announce@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] pixman release 0.11.4 now available
A new pixman release 0.11.4 is now available. This is a brown paper bag release to restore the ABI which broke in 0.11.2. Notable news: - ABI is unbroken (Julien Cristau) tar.gz: http://cairographics.org/releases/pixman-0.11.4.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.11.4.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.11.4.tar.bz2 Hashes: MD5: 8c8afc259af61d42aa9b3d8ec5a9a03b pixman-0.11.4.tar.gz MD5: 781ea328ad7340470235587b5101f343 pixman-0.11.4.tar.bz2 SHA1: 736b0d9ee1e594ce7eb2b51185950162c800457b pixman-0.11.4.tar.gz SHA1: e74dfcc6435edf7a397e8c42ffb4d581484e05d3 pixman-0.11.4.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.11.4.tar.gz.sha1.asc (signed by Soren Sandmann Pedersen) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.11.4 Log: Julien Cristau (2): Add PIXMAN_EXPORT for pixman_region_equal Add a couple more missing PIXMAN_EXPORTs Søren Sandmann Pedersen (3): Post-release version bump Add back prototypes for pixman_version() and pixman_version_string() to p Pre-release version bump ___ xorg-announce mailing list xorg-announce@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/xorg-announce
[ANNOUNCE] pixman release 0.11.2 now available
A new pixman release 0.11.2 is now available. This is a development version leading up to a stable 0.12.0 release. Notable new features in this release include: - SSE2 implementations of many compositing ops (André Tupinambá) - VMX implementations for many compositing ops (Luca Barbato) - Opacity based optimization for operators (Antoine Azar) - A 32 bit region implmentation (Søren Sandmann) tar.gz: http://cairographics.org/releases/pixman-0.11.2.tar.gz http://xorg.freedesktop.org/archive/individual/lib/pixman-0.11.2.tar.gz tar.bz2: http://xorg.freedesktop.org/archive/individual/lib/pixman-0.11.2.tar.bz2 Hashes: MD5: 1c9008358b86d1ea72ef31ff46510b3c pixman-0.11.2.tar.gz MD5: 7ba370232a93cddfdaac1a0b180b6a92 pixman-0.11.2.tar.bz2 SHA1: 348a6ed3bbfa353cdd22603ad30e877b17689a06 pixman-0.11.2.tar.gz SHA1: de333547164ddfffe5e611d21b614ab9251a90b4 pixman-0.11.2.tar.bz2 GPG signature: http://cairographics.org/releases/pixman-0.11.2.tar.gz.sha1.asc (signed by Soren Sandmann Pedersen) Git: git://git.freedesktop.org/git/pixman tag: pixman-0.11.2 Log: Aaron Plattner (13): Fix test build when srcdir != builddir. Get rid of pixman_composeFunctions_accessors. Move combining macros into pixman-combine.c. Move combining routines into combine.inc and add a Perl rule to generate Tack 32 onto the ends of the combining function types. Replace hardcoded numbers and uint*_t in combine.inc. Generate 64-bit combining functions. Use a macro to append _accessors to things. Move the scanline buffer allocation logic into pixman_composite_rect_gene Split fetch/fetchPixel/store proc types into 32-bit and 64-bit versions. Add infrastructure for allocating wide scanline buffers. Not yet used. Use wide compositing functions when wide == 1. Fix wide alpha fetch macro. Alan Coopersmith (1): Fix pixman/Makefile.am to work with Solaris make André Tupinambá (1): Add SSE2 implementations of many compositing operations. Antoine Azar (1): Optimize operators based on source or dest opacity. Dimitrios Apostolou (1): Fix compilation on SGI Frederic Plourde (1): Add CopyAreammx fast path for argb32 SRC xrgb32 and abgr32 SRC xbgr32 Julien Cristau (3): Remove prototype for nonexistent pixman_image_set_filter_params pixman-version.h is generated, don't distribute it Revert pixman-version.h is generated, don't distribute it Luca Barbato (11): update patch Make configure message alike the mmx/sse/sse2 ones Unbreak vmx pixman Force inlining Refactor path selection Simplify cpu feature check Remove VMX from CPUFeatures, ppc isn't using it at all Remove unused macro Split combine.inc generated files in source and header Use combine macros from the generated header Use sigaction instead of signal to restore the previous handler Oswald Buddenhagen (1): fix pixman compile with srcdir != builddir Richard Hult (1): Add --disable-gtk switch to configure Søren Sandmann (33): Fix bug in FbStore_b2g3r3 Require GCC 3.4 for SSE and SSE2. Simplify handling compiler handling in MMX code Update TODO Add support for 8 bit fills in pixman_fill_mmx() Work around GCC bug causing crashes in Mozilla with SSE2 Update TODO macroize pixman-region.c Add pixman-region16.c; compile that instead of pixman-region.c Replace pixman_box16_t with box_type_t Add point_type_t More data types; remove unused stuff Delete more unused stuff Delete more unused stuff Add prefixes to global primitive regions Add pixman-region32.c Add prototypes for 32 bit region methods Fix up types in pixman32-region.c Rename pixman-combine.h - pixman-combin32.h Fix forgotten use of pixman-combine.h Move PIXMAN_EXPORT into pixman-region.c Move all the PIXMAN_EXPORT into .c files Some formatting fixing Add pixman_region32_copy_from_region16
