summaryrefslogtreecommitdiffstats
path: root/src/mesa/x86/mmx_blend.S
Commit message (Collapse)AuthorAgeFilesLines
* mesa: Remove assyntax.h include from generated matypes.h.José Fonseca2008-09-211-1/+2
| | | | | | | matypes.h doesn't really depend on assyntax.h; assyntax.h is only present on x86; and this way we remove the need of grep being available (which is relevant for scons and non-unix platforms). (cherry picked from commit fc286ff3ba089af5f1aca2d35b93b53a9383b1cf)
* Fix TEXREL issues.Brian2007-03-101-5/+36
| | | | | Patch submitted by Christoph Brill. See http://www.gentoo.org/proj/en/hardened/pic-fix-guide.xml
* Add .note.GNU-stack section to assembler files to avoid the default behaviorKristian Høgsberg2006-04-171-0/+4
| | | | of requesting executable stacks.
* bring over build fixes from stable branchAlan Hourihane2004-04-261-1/+4
|
* Port over changes from XFree86/Mesa 5.0.2, mostly to silence compiler warnings.Brian Paul2003-12-041-5/+10
|
* Re-scheduling of the instructions according to the MMX pairing rules. This ↵Jose Fonseca2002-04-191-42/+36
| | | | is dificult to quantify but the improvemnet reaches about of 0.7% in some cases in a Pentium III. In older Pentiums the improvements could be quite substancial due less capable integer pipelines.
* Added min and max blend functions.Jose Fonseca2002-04-191-9/+68
| | | | Slight optimization of _mesa_mmx_blend_add.
* MMX add blending function added.Jose Fonseca2002-04-191-1/+16
|
* MMX modulate belding function added.Jose Fonseca2002-04-191-43/+51
| | | | Factorization of more common MMX code.
* Definition of several utility macros for self-contained MMX operations such ↵Jose Fonseca2002-04-181-305/+226
| | | | | | | | | as scaling and lerping. Restructured the MMX blending function to use a template, being only necessary to specify the main loop, which is also used for making the runin and runout sections. Optimization of the MMX function after remembering that the multiplication was commutative (how can somebody forget this..) resulting in less register usage. Now there is no need for generate or read from memory any constant inside the loop. Assemblers other than the GNU assembler can choke on the output of the C preprocessor since it was necessary to add line separators ';' to the defined macros.
* Removed code that was proven to have faster alternatives to reduce ↵Jose Fonseca2002-04-181-69/+0
| | | | complexity and facilitate reusability.
* Remaining patch for the 255x255=255 identity correction beJose Fonseca2002-04-141-0/+12
| | | | made also in the runin and runout section of the code.
* Jose's latest patch (GMBT_GEOMETRIC_CORRECTION)Brian Paul2002-04-121-1/+17
|
* new MMX blend code (Jose Fonseca)Brian Paul2002-04-101-336/+378
|
* more blending updates from JoseBrian Paul2002-04-021-19/+22
|
* updates from Jose FonsecaBrian Paul2002-04-021-25/+30
|
* New type system for assembly code. Asm files should now includeGareth Hughes2001-03-281-1/+1
| | | | | matypes.h, which includes assyntax.h and is generated from the core Mesa header files.
* more clean-upsBrian Paul2001-03-031-2/+2
|
* Major audit of all Mesa's x86 assembly code. This round is basicallyGareth Hughes2000-10-231-4/+0
| | | | | | | | general cleanups - more to come. Added P6 architecture timing to debug_xform routines. Still need to add test_all_vertex_functions test for the v16 asm. Dynamic reconfiguration of counter overhead for more accurate benchmarking.
* moved #define of LLBL into assyntax.h fileBrian Paul2000-06-141-16/+9
| | | | removed spaces from all uses of LLBL macro, for FreeBSD
* Initial revisionjtg1999-08-191-0/+363
generated by cgit v1.1 at 2024-09-11 23:17:04 +0000