[Mesa-dev] [PATCH 1/1] gallium/gallivm: code generation options for LLVM 3.1+

2012-09-22 Thread Alexander V. Nikolaev
From: "Alexander V. Nikolaev" LLVM 3.1+ haven't more "extern unsigned llvm::StackAlignmentOverride" and friends for configuring code generation options, like stack alignment. So I restrict assiging of lvm::StackAlignmentOverride and other variables to LLVM 3.0 only, and wrote similiar code using

[Mesa-dev] [PATCH 6/6] mesa: remove FEATURE_ES1 tests in enable.c code

2012-09-22 Thread Brian Paul
From: Brian Paul --- src/mesa/main/enable.c |6 -- 1 files changed, 0 insertions(+), 6 deletions(-) diff --git a/src/mesa/main/enable.c b/src/mesa/main/enable.c index 3643cfb..fe2870b 100644 --- a/src/mesa/main/enable.c +++ b/src/mesa/main/enable.c @@ -717,7 +717,6 @@ _mesa_set_enable(s

[Mesa-dev] [PATCH 5/6] mesa: remove FEATURE_ES test in _mesa_get_compressed_formats()

2012-09-22 Thread Brian Paul
From: Brian Paul --- src/mesa/main/texcompress.c |2 -- 1 files changed, 0 insertions(+), 2 deletions(-) diff --git a/src/mesa/main/texcompress.c b/src/mesa/main/texcompress.c index bb61e6e..957cc6d 100644 --- a/src/mesa/main/texcompress.c +++ b/src/mesa/main/texcompress.c @@ -275,7 +275,6

[Mesa-dev] [PATCH 4/6] mesa: remove FEATURE_ES test in _mesa_is_compressed_format()

2012-09-22 Thread Brian Paul
From: Brian Paul The code already has a runtime ES1 test. --- src/mesa/main/glformats.c |2 -- 1 files changed, 0 insertions(+), 2 deletions(-) diff --git a/src/mesa/main/glformats.c b/src/mesa/main/glformats.c index 04029c0..b15e344 100644 --- a/src/mesa/main/glformats.c +++ b/src/mesa/mai

[Mesa-dev] [PATCH 3/6] mesa: remove FEATURE_GL test from updated_drawbuffers()

2012-09-22 Thread Brian Paul
From: Brian Paul There's already a runtime test for full OpenGL. --- src/mesa/main/buffers.c |2 -- 1 files changed, 0 insertions(+), 2 deletions(-) diff --git a/src/mesa/main/buffers.c b/src/mesa/main/buffers.c index 216b6ee..cc268b6 100644 --- a/src/mesa/main/buffers.c +++ b/src/mesa/main

[Mesa-dev] [PATCH 2/6] mesa: remove #if _HAVE_FULL_GL checks

2012-09-22 Thread Brian Paul
From: Brian Paul This is basically more of the "remove FEATURE_x" clean-up. --- src/mesa/main/blend.c | 10 +- src/mesa/main/clear.c |2 -- src/mesa/main/context.c |8 src/mesa/main/debug.h | 15 --- src/mesa/main/depth.h |

[Mesa-dev] [PATCH 1/6] mesa: remove ASSERT_NO_FEATURE macro

2012-09-22 Thread Brian Paul
From: Brian Paul Was only used in one place. --- src/mesa/main/api_validate.c |2 +- src/mesa/main/mfeatures.h|3 --- 2 files changed, 1 insertions(+), 4 deletions(-) diff --git a/src/mesa/main/api_validate.c b/src/mesa/main/api_validate.c index db9d14a..6f02508 100644 --- a/src/mes

[Mesa-dev] [PATCH] llvmpipe: Fix creation of shared and scanout textures

2012-09-22 Thread John Kåre Alsaker
--- src/gallium/drivers/llvmpipe/lp_texture.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/src/gallium/drivers/llvmpipe/lp_texture.c b/src/gallium/drivers/llvmpipe/lp_texture.c index 841df00..df4417f 100644 --- a/src/gallium/drivers/llvmpipe/lp_texture.c +++ b/src/galliu

[Mesa-dev] [PATCH 07/10] i965: Replace brw_wm_* with dumping code into the fs_visitor.

2012-09-22 Thread Eric Anholt
This makes a giant pile of code newly dead. It also fixes TXB on newer chipsets, which has been totally broken (I now have a piglit test for that). It passes the same set of Ian's ARB_fragment_program tests. It also improves high-settings ETQW performance by 3.2 +/- 1.9% (n=3), thanks to better o

[Mesa-dev] [PATCH 02/10] i965/fs: Pass fragment depth to the fb write as a fs_reg, not an ir_variable.

2012-09-22 Thread Eric Anholt
This will be used for the ARB_fp change to use this backend. --- src/mesa/drivers/dri/i965/brw_fs.h |2 +- src/mesa/drivers/dri/i965/brw_fs_visitor.cpp |9 +++-- 2 files changed, 4 insertions(+), 7 deletions(-) diff --git a/src/mesa/drivers/dri/i965/brw_fs.h b/src/mesa/driv

[Mesa-dev] [PATCH 08/10] i965: Remove the old ARB_fragment_program backend.

2012-09-22 Thread Eric Anholt
--- I've trimmed the removed files here because the patch was huge. src/mesa/drivers/dri/i965/Makefile.sources |6 - src/mesa/drivers/dri/i965/brw_wm.c | 55 - src/mesa/drivers/dri/i965/brw_wm.h | 364 -- src/mesa/drivers/dri/i965/brw_wm_debug.c | 174 --- src/me

[Mesa-dev] [PATCH 09/10] i965/fs: Move constant propagation to the same codebase as copy prop.

2012-09-22 Thread Eric Anholt
This means that we don't get constant prop across into the first block after a BRW_OPCODE_IF or a BRW_OPCODE_DO, but we have hope for properly doing it across control flow at some point. More importantly, it avoids the O(n^2) with instruction count runtime for shaders that have many constant moves

[Mesa-dev] [PATCH 06/10] i965: Reduce maximum GL_ARB_fragment_program instruction count to 1024.

2012-09-22 Thread Eric Anholt
I don't know of any programs that would need more than this. The larger programs I've seen have neared 100 instructions. This prevent excessive runtimes of automatic tests that attempt to test up to the exposed maximums (like fp-long-alu). --- src/mesa/drivers/dri/i965/brw_context.c |8 -

[Mesa-dev] [PATCH 10/10] i965/fs: Improve performance of copy/constant propagation.

2012-09-22 Thread Eric Anholt
Use a simple chaining hash table for the ACP. This is not really very good, because we still do a full walk of the tree per destination write, but it still reduces fp-long-alu runtime from 5.3 to 3.9s. --- src/mesa/drivers/dri/i965/brw_fs.h |2 +- .../drivers/dri/i965/brw_fs_c

[Mesa-dev] [PATCH 05/10] i965/fs: Add a couple more algebraic cases that help some ARB_fp patterns.

2012-09-22 Thread Eric Anholt
ARB_fp doesn't go through the GLSL optimizer, and these were things you see frequently thanks to conditionals being lowered to SLT/SGE and MUL. --- src/mesa/drivers/dri/i965/brw_fs.cpp | 23 +++ 1 file changed, 23 insertions(+) diff --git a/src/mesa/drivers/dri/i965/brw_fs.c

[Mesa-dev] [PATCH 04/10] i965/fs: Pull ir_binop_min/ir_binop_max handling to a separate function.

2012-09-22 Thread Eric Anholt
This will be reused from the ARB_fp compiler. I touched up the pre-gen6 path to not overwrite dst in the first instruction, which prevents the need for aliasing checks (we'll need that in the ARB_fp compiler, but it actually hasn't been needed in this codebase since the revert of the nasty old MOV

[Mesa-dev] [PATCH 01/10] mesa: Note that OPCODE_RFL is not part of ARB_fp (it's NV_fp only).

2012-09-22 Thread Eric Anholt
--- src/mesa/program/prog_instruction.h |2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/mesa/program/prog_instruction.h b/src/mesa/program/prog_instruction.h index 09659ce..656556d 100644 --- a/src/mesa/program/prog_instruction.h +++ b/src/mesa/program/prog_instruction.

[Mesa-dev] [PATCH 03/10] i965/fs: Refactor rectangle/GL_CLAMP texture coordinate adjustment.

2012-09-22 Thread Eric Anholt
We'll want to reuse this for ARB_fp handling. --- src/mesa/drivers/dri/i965/brw_fs.h |2 ++ src/mesa/drivers/dri/i965/brw_fs_visitor.cpp | 44 +++--- 2 files changed, 27 insertions(+), 19 deletions(-) diff --git a/src/mesa/drivers/dri/i965/brw_fs.h b/src/mesa/

Re: [Mesa-dev] [PATCH 2/2] mesa: move _mesa_es_error_check_format_and_type() to glformats.c

2012-09-22 Thread Kenneth Graunke
On 09/22/2012 08:35 AM, Brian Paul wrote: > From: Brian Paul > > Where the non-ES _mesa_error_check_format_and_type() function lives. LGTM! For the series: Reviewed-by: Kenneth Graunke ___ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://l

[Mesa-dev] [Bug 55224] NV50 Gallium codegen RA incorrectly tracks interference for 64-bit / 128-bit registers

2012-09-22 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=55224 --- Comment #1 from Jay Cornwall 2012-09-22 16:32:40 UTC --- Created attachment 67549 --> https://bugs.freedesktop.org/attachment.cgi?id=67549 Fix NV50 RA interference register tracking. -- Configure bugmail: https://bugs.freedesktop.org/user

[Mesa-dev] [Bug 55224] New: NV50 Gallium codegen RA incorrectly tracks interference for 64-bit / 128-bit registers

2012-09-22 Thread bugzilla-daemon
https://bugs.freedesktop.org/show_bug.cgi?id=55224 Bug #: 55224 Summary: NV50 Gallium codegen RA incorrectly tracks interference for 64-bit / 128-bit registers Classification: Unclassified Product: Mesa Version: git Pl

[Mesa-dev] [PATCH 2/2] mesa: move _mesa_es_error_check_format_and_type() to glformats.c

2012-09-22 Thread Brian Paul
From: Brian Paul Where the non-ES _mesa_error_check_format_and_type() function lives. --- src/mesa/main/glformats.c | 69 + src/mesa/main/glformats.h |4 ++ src/mesa/main/readpix.c |1 - src/mesa/main/teximage.c | 65

[Mesa-dev] [PATCH 1/2] mesa: move GL_HALF_FLOAT_OES definition to glheader.h

2012-09-22 Thread Brian Paul
From: Brian Paul --- src/mesa/main/glheader.h |7 +++ src/mesa/main/readpix.c |5 - src/mesa/main/teximage.c |6 -- 3 files changed, 7 insertions(+), 11 deletions(-) diff --git a/src/mesa/main/glheader.h b/src/mesa/main/glheader.h index 3152323..e93ca30 100644 --- a/src

[Mesa-dev] [PATCH 6/6] mesa: minor fix to glTexSubImage error message

2012-09-22 Thread Brian Paul
From: Brian Paul --- src/mesa/main/teximage.c |3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/src/mesa/main/teximage.c b/src/mesa/main/teximage.c index f298ba5..fdd6129 100644 --- a/src/mesa/main/teximage.c +++ b/src/mesa/main/teximage.c @@ -2204,7 +2204,8 @@ texsubima

[Mesa-dev] [PATCH 5/6] mesa: consolidate sub-texture error checking code

2012-09-22 Thread Brian Paul
From: Brian Paul Do all error checking of glTexSubImage, glCopyTexSubImage and glCompressedTexSubImage's xoffset, yoffset, zoffset, width, height, and depth params in one place. --- src/mesa/main/teximage.c | 270 +- 1 files changed, 98 insertions(+),

[Mesa-dev] [PATCH 4/6] mesa: consolidate glTexSubImage() error checking

2012-09-22 Thread Brian Paul
From: Brian Paul --- src/mesa/main/teximage.c | 72 ++--- 1 files changed, 29 insertions(+), 43 deletions(-) diff --git a/src/mesa/main/teximage.c b/src/mesa/main/teximage.c index 6abe607..fe3e024 100644 --- a/src/mesa/main/teximage.c +++ b/src/mesa/mai

[Mesa-dev] [PATCH 3/6] mesa: consolidate glCompressedTexSubImage() error checking

2012-09-22 Thread Brian Paul
From: Brian Paul Do all the checking in one function instead of two and fix up some of the error checking.alignment check --- src/mesa/main/teximage.c | 158 +++--- 1 files changed, 80 insertions(+), 78 deletions(-) diff --git a/src/mesa/main/teximage.c

[Mesa-dev] [PATCH 2/6] mesa: consolidate subtexture xoffset/yoffset/width/height error checking code

2012-09-22 Thread Brian Paul
From: Brian Paul This is the code that checks if a subtexture region is aligned to the compressed format's block size. --- src/mesa/main/teximage.c | 161 +- 1 files changed, 73 insertions(+), 88 deletions(-) diff --git a/src/mesa/main/teximage.c b/s

[Mesa-dev] [PATCH 1/6] mesa: consolidate glCopyTexSubImage error checking

2012-09-22 Thread Brian Paul
From: Brian Paul Do all the checking in one function instead of two. --- src/mesa/main/teximage.c | 127 ++--- 1 files changed, 51 insertions(+), 76 deletions(-) diff --git a/src/mesa/main/teximage.c b/src/mesa/main/teximage.c index c22e853..37218f5 1006

[Mesa-dev] [PATCH] mesa: fix incorrect error for glCompressedSubTexImage

2012-09-22 Thread Brian Paul
From: Brian Paul If a subtexture region isn't aligned to the compressed block size, return GL_INVALID_OPERATION, not gl_INVALID_VALUE. NOTE: This is a candidate for the stable branches. --- src/mesa/main/teximage.c |6 +++--- 1 files changed, 3 insertions(+), 3 deletions(-) diff --git a/sr