summaryrefslogtreecommitdiff
path: root/src/gallium/auxiliary
AgeCommit message (Collapse)Author
2010-03-01util: Remove PIPE_FORMAT_A8B8G8R8_SNORM.José Fonseca
2010-03-01util: Generate linear<->sRGB conversion tables.Michal Krol
2010-03-01util: Fix PIPE_FORMAT_R8G8B8_SRGB swizzles.José Fonseca
The only user of PIPE_FORMAT_R8G8B8_SRGB is mesa, which per src/mesa/main/texstore.c interprets it as R8 G8 B8 and not the other way around.
2010-03-01util: Use PIPE_FORMAT_R8G8B8A8_UNORM_REV where appropriate.José Fonseca
2010-03-01util: Fix PIPE_FORMAT_R8G8B8_UNORM description and notes.José Fonseca
There is actually no ambiguity in use -- it is always used as R8 G8 B8
2010-03-01util: Remove inexisting formats.José Fonseca
Can't find these formats used in any state tracker or any API. For some of these probably the reverse notation was meant, for which formats already exist.
2010-03-01util: A few more links for D3D formats.José Fonseca
2010-03-01util: A few more todo marks.José Fonseca
2010-03-01util: More documentation for u_format.csvJosé Fonseca
2010-03-01util: Tweak comments regarding R8G8B8A8/X8_UNORM.José Fonseca
2010-03-01python: Don't mention deprecated drivers=trace option.José Fonseca
Now trace always built.
2010-02-27util: Fix X8/A8B8G8R8_SNORM's swizzle.José Fonseca
2010-02-26gallium: define GENERATED auxiliary filesBrian Paul
Don't need special default dependency on u_format_pack.h anymore.
2010-02-26util: Define SRGB's channels as normalized.José Fonseca
A better approximation than uscaled...
2010-02-26util: Code generate functions to pack and unpack a single pixel.José Fonseca
Should work correctly for all pixel formats except SRGB formats. Generated code made much simpler by defining the pixel format as a C structure. For example this is the generated structure for PIPE_FORMAT_B6UG5SR5S_NORM: union util_format_b6ug5sr5s_norm { uint16_t value; struct { int r:5; int g:5; unsigned b:6; } chan; }; Not used everywhere yet because it seems compiled code is slower than bitshift arithmetic by some misterious reason. So we should generate bitshift arithmetic at least for the simple UNORM pixel formats.
2010-02-26util: Factor some code into u_format_parse.pyJosé Fonseca
2010-02-26util: Add util_bswap16.José Fonseca
2010-02-26util: Use python names consistent with u_format.hJosé Fonseca
2010-02-26util: Add is_pot() method to formats.José Fonseca
2010-02-25util: remove a trailing semicolon in format table codeBrian Paul
2010-02-25util: Remove import of unused sys module.Vinson Lee
2010-02-25tgsi: Remove dead code.Michal Krol
2010-02-25util: Assert that pointer is not null before dereferencing.Vinson Lee
2010-02-24gallivm: Update for UTIL_FORMAT_LAYOUT_xxx changes.José Fonseca
2010-02-24util: Kill array vs arith layouts. Revamp UTIL_FORMAT_LAYOUT_xxx.José Fonseca
Too confusing. I now can get the same information through other means.
2010-02-24util: Refactor some code.José Fonseca
2010-02-24util: Cope with the fact that formats in u_format.csv are not ordered.José Fonseca
2010-02-24util: Factor out the code to shorten a format name.José Fonseca
2010-02-24util: Handle correctly 24bit formats.José Fonseca
2010-02-24util: Minor tweaks to ambigous/unused format descriptions.José Fonseca
2010-02-24util: Fix swizzles for SRGB formats.José Fonseca
Based on Marek Olšák's fix 8-bits-per-channel formats swizzles, with the exception that the notation for several formats (the vector formats) is actually reversed so they should be left alone.
2010-02-24util: Organize formats in groups. Add comments.José Fonseca
2010-02-24util: Allow '#' comments in u_format.csv.José Fonseca
2010-02-24util: Kill unused UTIL_FORMAT_LAYOUT_SCALAR.José Fonseca
2010-02-24util: Store more derived data in the the format description.José Fonseca
2010-02-23gallivm: added clamp and int_to_float functionsBrian Paul
2010-02-23pipebuffer: slab: if size < alignment, use alignment to find bucketLuca Barbieri
If the size is lower than the alignment, we must use the alignment to select the bucket. Otherwise, the selected bucket won't be able to satisfy our request and will fail. Signed-off-by: José Fonseca <jfonseca@vmware.com>
2010-02-23pipebuffer: fix inverted signalled checkingLuca Barbieri
A return of 0 means the fence is signalled. Signed-off-by: José Fonseca <jfonseca@vmware.com>
2010-02-23pipebuffer: avoid assert due to increasing a zeroed refcntLuca Barbieri
The cache manager stores buffers with a reference count that dropped to 0. pipe_reference asserts in this case on debug builds, so use pipe_reference_init instead. Signed-off-by: José Fonseca <jfonseca@vmware.com>
2010-02-22draw: Remove dead code.Vinson Lee
2010-02-22tgsi: Report an error when a destination has an empty writemask.José Fonseca
2010-02-22pipebuffer: Don't synchronize when checking for buffer overflows.José Fonseca
To avoid masking synchronization issues in debug builds.
2010-02-21draw: Remove dead code.Vinson Lee
2010-02-20tgsi: ifdef out unused function micro_sqrt.Vinson Lee
2010-02-19tgsi: Make more exec opcodes look pretty.Michal Krol
2010-02-19tgsi: Change prototypes of micro opcodes to explicitly indicates number of args.Michal Krol
2010-02-19tgsi: Rewrite exec implementations of NRM and NRM4.Michal Krol
2010-02-19Replace the _mesa_*printf() wrappers with the plain libc versionsKristian Høgsberg
2010-02-18pipebuffer: Include fenced buffer manager in buildJakob Bornecrantz
2010-02-17util: Fix typo of debug_dump_float_rgba_bmp.Vinson Lee