Jakob Bornecrantz [Tue, 27 May 2008 10:59:46 +0000 (12:59 +0200)]
draw: Fix for EMIT_4UB case
Keith Whitwell [Tue, 27 May 2008 11:26:23 +0000 (12:26 +0100)]
draw: hook up viewport / rhw emit to varient key state
Keith Whitwell [Tue, 27 May 2008 10:58:55 +0000 (11:58 +0100)]
draw: restore extras path in draw_pt_vcache.c, keep pipeline flags out of non-pipeline elts
Keith Whitwell [Tue, 27 May 2008 10:15:31 +0000 (11:15 +0100)]
draw: dump individual instructions as they are processed
Keith Whitwell [Tue, 27 May 2008 10:12:42 +0000 (11:12 +0100)]
tgsi: export utils for dumping individual instructions
Keith Whitwell [Tue, 27 May 2008 10:10:50 +0000 (11:10 +0100)]
draw: fix writemask/shufps confusion
Keith Whitwell [Tue, 27 May 2008 09:42:58 +0000 (10:42 +0100)]
draw: some possible fixes for spilling
Keith Whitwell [Tue, 27 May 2008 09:35:33 +0000 (10:35 +0100)]
ffvertex: emit full LIT when attenuating (needs the 1 in X position)
Keith Whitwell [Tue, 27 May 2008 08:48:32 +0000 (09:48 +0100)]
ffvertex: don't compute whole eye vector if only eye.z is required
Keith Whitwell [Mon, 26 May 2008 18:10:44 +0000 (19:10 +0100)]
draw: defensively flush pipeline backend when setting primitive
Keith Whitwell [Mon, 26 May 2008 17:54:35 +0000 (18:54 +0100)]
draw: defensively reset render primitive some more
Keith Whitwell [Mon, 26 May 2008 17:37:34 +0000 (18:37 +0100)]
draw: defensively reset render primitive, which can get clobbered by clipping
Keith Whitwell [Mon, 26 May 2008 17:29:47 +0000 (18:29 +0100)]
draw: add missing break
Keith Whitwell [Mon, 26 May 2008 10:03:00 +0000 (11:03 +0100)]
draw: extend precision in RSQ opcode
Keith Whitwell [Sun, 25 May 2008 23:09:02 +0000 (00:09 +0100)]
draw: more aos tweaks
Keith Whitwell [Sun, 25 May 2008 14:47:04 +0000 (15:47 +0100)]
draw: slight tweak for XPD opcode
Keith Whitwell [Sun, 25 May 2008 14:45:27 +0000 (15:45 +0100)]
draw: remove EXP & LOG from vs_aos.c
These don't get hit & look like bug magnets to me...
Keith Whitwell [Sun, 25 May 2008 14:44:17 +0000 (15:44 +0100)]
draw: special case for writing out scalar results
Keith Whitwell [Sun, 25 May 2008 14:37:47 +0000 (15:37 +0100)]
draw: clean up some of the xmm register manipulation function names
Keith Whitwell [Sun, 25 May 2008 14:28:30 +0000 (15:28 +0100)]
cso: use memcpy rather than structure asignment for copying
Apparently gcc will omit to copy hidden padding bytes under some circumstances,
which means later on memcmp() will indicate a difference between structs
even though all the visible members are identical.
Keith Whitwell [Sun, 25 May 2008 10:25:09 +0000 (11:25 +0100)]
draw: add viewport support to generic vs varient code
Keith Whitwell [Sun, 25 May 2008 10:20:38 +0000 (11:20 +0100)]
draw: fix input vs output typo in emit
Keith Whitwell [Sat, 24 May 2008 15:32:08 +0000 (16:32 +0100)]
Revert "mesa: save a temp on normalizes"
This reverts commit
feceb43948f76cc4d4c8ecbb86b1b1f438c6daee.
Keith Whitwell [Sat, 24 May 2008 15:31:11 +0000 (16:31 +0100)]
draw: use lookup tables to avoid calling pow() in LIT opcode
Keith Whitwell [Sat, 24 May 2008 12:23:06 +0000 (13:23 +0100)]
mesa: pre-swizzle normal scale state value
Keith Whitwell [Sat, 24 May 2008 12:22:39 +0000 (13:22 +0100)]
mesa: evaluate _NeedEyeCoords prior to generating internal vertex shader
Keith Whitwell [Sat, 24 May 2008 12:22:29 +0000 (13:22 +0100)]
draw: add a debug-print which can be called from inside generated shaders
Keith Whitwell [Sat, 24 May 2008 12:22:15 +0000 (13:22 +0100)]
rtasm: add some helpers for calling out from generated code
Keith Whitwell [Fri, 23 May 2008 19:37:50 +0000 (20:37 +0100)]
mesa: save a temp on normalizes
Keith Whitwell [Fri, 23 May 2008 19:05:36 +0000 (20:05 +0100)]
mesa: further degenerate the special case lit substitute
Keith Whitwell [Fri, 23 May 2008 18:17:57 +0000 (19:17 +0100)]
glapi: fix include path & make build work
Keith Whitwell [Fri, 23 May 2008 18:17:02 +0000 (19:17 +0100)]
mesa: don't emit LIT instruction when mat shininess known to be zero
Use a faster path in that case & make gears go faster.
Keith Whitwell [Fri, 23 May 2008 08:14:17 +0000 (09:14 +0100)]
draw: faster LIT(), incorrect though
Keith Whitwell [Fri, 23 May 2008 08:10:59 +0000 (09:10 +0100)]
mesa: do object-space lighting in ffvertex_prog.c
Start pulling over some of the optimizations from the fixed function
paths.
Keith Whitwell [Thu, 22 May 2008 19:22:15 +0000 (20:22 +0100)]
draw: correct but slow LIT() in aos varient
Keith Whitwell [Thu, 22 May 2008 19:21:49 +0000 (20:21 +0100)]
draw: tweak x87_emit_ex2 to avoid changing x87 fpu settings
Keith Whitwell [Thu, 22 May 2008 14:24:02 +0000 (15:24 +0100)]
draw: use aligned movs within draw_vs_aos.c
Keith Whitwell [Thu, 22 May 2008 12:49:38 +0000 (13:49 +0100)]
draw: for debug, do rhw divide in aos_sse viewport calcs
Keith Whitwell [Thu, 22 May 2008 12:48:07 +0000 (13:48 +0100)]
draw: shortcircuit shuffle in aos_sse when possible
Keith Whitwell [Thu, 22 May 2008 12:47:08 +0000 (13:47 +0100)]
draw: remove FPU_MANIP ifdef
Keith Whitwell [Thu, 22 May 2008 12:46:06 +0000 (13:46 +0100)]
draw: clean up masked writes in aos sse, make some xmm function names clearer
Keith Whitwell [Thu, 22 May 2008 12:41:49 +0000 (13:41 +0100)]
draw: clean up internal immediates in aos sse
Keith Whitwell [Thu, 22 May 2008 12:37:48 +0000 (13:37 +0100)]
draw: fse works with elts, remove assert
Keith Whitwell [Wed, 21 May 2008 19:50:36 +0000 (20:50 +0100)]
rtasm: remove debug
Keith Whitwell [Wed, 21 May 2008 19:41:01 +0000 (20:41 +0100)]
draw: when preloading args to x87 stack, need to use reverse order
Keith Whitwell [Wed, 21 May 2008 19:38:23 +0000 (20:38 +0100)]
draw: avoid a pointless mov in many sse opcodes
Keith Whitwell [Wed, 21 May 2008 19:34:52 +0000 (20:34 +0100)]
draw: fix x87_ex2 and partially fix lit insn
Keith Whitwell [Wed, 21 May 2008 19:32:43 +0000 (20:32 +0100)]
draw: fix fpu control word manipulations
Keith Whitwell [Wed, 21 May 2008 19:31:08 +0000 (20:31 +0100)]
draw: fix vs aos internal/machine state
Keith Whitwell [Wed, 21 May 2008 19:28:56 +0000 (20:28 +0100)]
draw: add viewport to varient state
Keith Whitwell [Wed, 21 May 2008 19:14:55 +0000 (20:14 +0100)]
rtasm: add x87 instructions and debug-check for x87 stack usage
Keith Whitwell [Wed, 21 May 2008 12:17:48 +0000 (13:17 +0100)]
vp-tris: use test name as window name
Keith Whitwell [Wed, 21 May 2008 08:44:16 +0000 (09:44 +0100)]
draw: add aos vertex shader varient
Keith Whitwell [Wed, 21 May 2008 08:43:30 +0000 (09:43 +0100)]
draw: mark varient functions as PIPE_CDECL
Keith Whitwell [Wed, 21 May 2008 07:29:19 +0000 (08:29 +0100)]
translate: mark functions as PIPE_CDECL
Keith Whitwell [Wed, 21 May 2008 07:28:53 +0000 (08:28 +0100)]
gallium: define PIPE_CDECL calling convention, which really is cdecl everywhere
Keith Whitwell [Wed, 21 May 2008 07:28:16 +0000 (08:28 +0100)]
rtasm: export debug reg print function
Keith Whitwell [Tue, 20 May 2008 13:34:06 +0000 (14:34 +0100)]
translate: remove spurious comment
Keith Whitwell [Mon, 19 May 2008 15:28:53 +0000 (16:28 +0100)]
rtasm: remove unused struct member
Keith Whitwell [Thu, 15 May 2008 11:39:08 +0000 (12:39 +0100)]
draw: create specialized vs varients incorporating fetch & emit
Keith Whitwell [Tue, 13 May 2008 12:40:22 +0000 (13:40 +0100)]
draw: move some state into a new 'vs' area
Keith Whitwell [Tue, 13 May 2008 12:35:14 +0000 (13:35 +0100)]
draw: get rid of fetch-shade-emit frontend hack
The code is now living in it's intended place as a pt middle end.
Keith Whitwell [Tue, 13 May 2008 08:46:53 +0000 (09:46 +0100)]
draw: remove disabled non-sse swizzle code
Zack Rusin [Mon, 19 May 2008 16:10:42 +0000 (12:10 -0400)]
Merge commit 'origin/gallium-0.1' into gallium-vertex-linear
Brian Paul [Mon, 19 May 2008 15:34:28 +0000 (09:34 -0600)]
if x86_get_func() returns NULL, handle it properly instead of aborting
Brian Paul [Mon, 19 May 2008 15:04:02 +0000 (09:04 -0600)]
fix program delete/refcount bug
cherry-picked from gallium-tex-surfaces
Brian Paul [Mon, 19 May 2008 14:59:41 +0000 (08:59 -0600)]
Fix program refcounting assertion failure during context tear-down
When purging the program hash table, the refcount _should_ be one since
the program is referenced by the hash table. Need to explicitly set to
zero before calling delete().
Also, purge high-level shader hash tables before low-level program hash tabl
cherry-picked from master
Brian Paul [Sun, 18 May 2008 21:46:58 +0000 (15:46 -0600)]
print vertex.attrib[n] (from master)
Brian Paul [Sun, 18 May 2008 21:32:58 +0000 (15:32 -0600)]
Fix some bugs in material, depth range state
picked from master
Brian Paul [Sun, 18 May 2008 21:23:03 +0000 (15:23 -0600)]
Regenerated API dispatch files
Follow on to cherry-pick from master
Brian Paul [Sun, 18 May 2008 21:21:28 +0000 (15:21 -0600)]
alias ProgramEnvParameter4xyARB and ProgramParameter4xyNV (bug #12935)
these should be the same functions (as per spec).
cherry-picked from master (
86a4810b09097714942bf2b889e6c62357bba931)
Zack Rusin [Fri, 16 May 2008 21:56:38 +0000 (17:56 -0400)]
cosmetic changes
Zack Rusin [Fri, 16 May 2008 21:35:47 +0000 (17:35 -0400)]
fix rsq
Zack Rusin [Fri, 16 May 2008 21:10:52 +0000 (17:10 -0400)]
fix abs and start on rsq
Zack Rusin [Fri, 16 May 2008 20:06:59 +0000 (16:06 -0400)]
do the lit (some artifacts present)
Zack Rusin [Fri, 16 May 2008 18:54:40 +0000 (14:54 -0400)]
implement min/max and abstract ops on vectors
Zack Rusin [Thu, 15 May 2008 21:46:20 +0000 (17:46 -0400)]
llvm: implement sub and abs
Zack Rusin [Thu, 15 May 2008 18:11:19 +0000 (14:11 -0400)]
fix injections of functions from builtins into shaders
Zack Rusin [Thu, 15 May 2008 16:17:46 +0000 (12:17 -0400)]
move the swizzling code to gallivm in preperation for code-generating it
also some minor cleanups
Brian Paul [Sat, 17 May 2008 16:30:21 +0000 (10:30 -0600)]
gallium: in drivers, make copy of tokens passed to pipe->create_vs/fs_state()
The caller can then free the token array immediately.
Brian Paul [Sat, 17 May 2008 14:47:56 +0000 (08:47 -0600)]
gallium: remove deprecated format names
Brian Paul [Sat, 17 May 2008 14:46:43 +0000 (08:46 -0600)]
s/PIPE_FORMAT_U_S8/PIPE_FORMAT_S8_UNORM/
Brian Paul [Fri, 16 May 2008 22:04:30 +0000 (16:04 -0600)]
gallium: release textures during context tear-down (fix mem leak)
Brian Paul [Fri, 16 May 2008 22:04:04 +0000 (16:04 -0600)]
gallium: ref counting clean-ups
Brian Paul [Fri, 16 May 2008 22:02:52 +0000 (16:02 -0600)]
mesa: call glutDestroyWindow() on exit to help find mem leaks
Brian Paul [Fri, 16 May 2008 21:48:11 +0000 (15:48 -0600)]
bring in fixes/changes from master
Brian Paul [Fri, 16 May 2008 15:56:59 +0000 (09:56 -0600)]
Fix a program refcounting error, don't share program parameter lists.
The refcounting bug was causing a memleak (unfreed programs).
The old parameter list sharing is not needed since the change in how
uniforms are handled.
Brian Paul [Fri, 16 May 2008 21:34:22 +0000 (15:34 -0600)]
fix an attr/src mix-up when setting-up/binding vertex arrays
This fixes problems with incorrect material coefficients when glMaterial
is called per-vertex.
Brian Paul [Thu, 15 May 2008 01:19:53 +0000 (19:19 -0600)]
fix parsing of state.texenv.color (bug 14931)
cherry-picked from master
Brian Paul [Thu, 15 May 2008 01:14:41 +0000 (19:14 -0600)]
Consolidate texture fetch code and use partial derivatives when possible.
picked from master
Brian Paul [Wed, 14 May 2008 22:44:08 +0000 (16:44 -0600)]
mesa: fix InfoLog memleak, misc sync with master
Eric Anholt [Tue, 15 Jan 2008 23:08:34 +0000 (15:08 -0800)]
Bug #13492: Only call ProgramStringNotify if program parsing succeeded.
Wine intentionally tries some out-of-spec programs to test strictness, and
calling ProgramStringNotify on the results of a failed program parse resulted
in crashes in the 965 driver.
Brian [Tue, 7 Aug 2007 20:42:34 +0000 (21:42 +0100)]
fix swizzle error test (bug 11881)
Brian Paul [Wed, 14 May 2008 22:32:18 +0000 (16:32 -0600)]
fix-up inlined/non-inlined function inconsistencies
cherry-picked from master
Brian Paul [Wed, 14 May 2008 22:30:15 +0000 (16:30 -0600)]
mesa: rm unneeded file
Brian Paul [Wed, 14 May 2008 22:27:34 +0000 (16:27 -0600)]
mesa: prefix rm command with -
Brian Paul [Wed, 14 May 2008 22:20:28 +0000 (16:20 -0600)]
rewrite some of the mat*mat, mat*vec intrinsics
Also, remove obsolete matrix codegen code.
cherry-picked from master
Brian Paul [Wed, 14 May 2008 22:07:18 +0000 (16:07 -0600)]
mesa: regenerate slang built-in data (asin(vec4) fix)
Brian Paul [Wed, 14 May 2008 22:06:10 +0000 (16:06 -0600)]
mesa: fix missing w assignment in asin(vec4)
Brian Paul [Wed, 14 May 2008 22:05:48 +0000 (16:05 -0600)]
mesa: use CALLOC_STRUCT()