platform/upstream/mesa.git
13 years agor600g: fix cache flushes on r6xx
Alex Deucher [Thu, 5 May 2011 07:17:16 +0000 (03:17 -0400)]
r600g: fix cache flushes on r6xx

r6xx asics have some problems with the surface
sync logic for the CB and DB.  It's recommended
to use the event write interface for flushing
the DB/CB caches rather than the sync packets.
A single event write flush flushes all dst
caches, so we only need one for all CBs and DB.

Should fix:
https://bugs.freedesktop.org/show_bug.cgi?id=35312

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agomesa: don't call git if it's not git repository
Marcin Slusarz [Thu, 5 May 2011 20:37:22 +0000 (22:37 +0200)]
mesa: don't call git if it's not git repository

Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
13 years agomesa: don't touch git_sha1.h if sha1 didn't change
Marcin Slusarz [Thu, 5 May 2011 20:37:21 +0000 (22:37 +0200)]
mesa: don't touch git_sha1.h if sha1 didn't change

Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
13 years agor600g: Match alpha ref precision to color format precision.
Henri Verbeet [Thu, 5 May 2011 18:54:36 +0000 (20:54 +0200)]
r600g: Match alpha ref precision to color format precision.

This fixes piglit fbo-alphatest-formats on Evergreen.

Signed-off-by: Henri Verbeet <hverbeet@gmail.com>
13 years agoRevert "DRI2/GLX: make swap event handling match spec"
Dave Airlie [Thu, 5 May 2011 06:23:46 +0000 (16:23 +1000)]
Revert "DRI2/GLX: make swap event handling match spec"

This reverts commit 70227e21bbd4411956ceeb5039640140e64a11a8.

revert this until we sort things out, with a new glproto release
with compatible struct names.

13 years agor600g: do RV6xx base updates inline with state updates.
Dave Airlie [Tue, 3 May 2011 18:45:39 +0000 (20:45 +0200)]
r600g: do RV6xx base updates inline with state updates.

This seems more in line with what the documentation suggests we should be
doing. It doesn't fix the rv635 regression, though I thought it might,
so it means I've no idea whats actually going wrong there.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
13 years agoDRI2/GLX: make swap event handling match spec
Jesse Barnes [Tue, 3 May 2011 19:17:28 +0000 (12:17 -0700)]
DRI2/GLX: make swap event handling match spec

We only handle a 32 bit swap count, so use the new structure definitions.

Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
13 years agoGLX: don't crash when indirect clients try to set an event mask
Jesse Barnes [Tue, 3 May 2011 17:20:14 +0000 (10:20 -0700)]
GLX: don't crash when indirect clients try to set an event mask

After sending the GLXChangeDrawableAttributes request, we also set a
local set of attributes on the DRI drawable.  But in the indirect case
this array won't be present, so skip the setting in that case to avoid a
crash.

Reviewed-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
13 years agoscons: Add a few more human friendly messages.
José Fonseca [Wed, 4 May 2011 13:10:24 +0000 (14:10 +0100)]
scons: Add a few more human friendly messages.

13 years agomesa/gdi: Silence gcc warning about unused result.
José Fonseca [Wed, 4 May 2011 13:10:15 +0000 (14:10 +0100)]
mesa/gdi: Silence gcc warning about unused result.

13 years agost/wgl: Allow to create pbuffers bigger than the desktop.
José Fonseca [Tue, 3 May 2011 18:09:53 +0000 (19:09 +0100)]
st/wgl: Allow to create pbuffers bigger than the desktop.

We use a hidden window for pbuffer contexts, but Windows limits window
sizes to the desktop size by default. This means that creating a big
pbuffer on a small resolution single monitor would truncate the pbuffer
size to the desktop.

This change overrides the windows maximum size, allow to create windows
arbitrarily large.

13 years agost/wgl: Cope with zero width/height windows.
José Fonseca [Mon, 18 Apr 2011 15:41:25 +0000 (16:41 +0100)]
st/wgl: Cope with zero width/height windows.

While ensuring the framebuffer area is never zero.

13 years agoi915g: add const qualifier to silence warning
Brian Paul [Wed, 4 May 2011 00:56:20 +0000 (18:56 -0600)]
i915g: add const qualifier to silence warning

13 years agomesa: remove unused restart.[ch] files
Brian Paul [Wed, 4 May 2011 00:19:49 +0000 (18:19 -0600)]
mesa: remove unused restart.[ch] files

_mesa_PrimitiveRestartIndex() is in varray.c and glPrimitiveRestart()
is handled in the vbo module.

13 years agoglsl: Remove extra newline from error message
Ian Romanick [Tue, 3 May 2011 16:50:53 +0000 (09:50 -0700)]
glsl: Remove extra newline from error message

13 years agor600c: add some new pci ids
Alex Deucher [Tue, 3 May 2011 20:12:04 +0000 (16:12 -0400)]
r600c: add some new pci ids

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
13 years agor600g: add some new pci ids
Alex Deucher [Tue, 3 May 2011 19:26:07 +0000 (15:26 -0400)]
r600g: add some new pci ids

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
13 years agost/xorg: flush after loading the cursor
Marcin Slusarz [Tue, 3 May 2011 09:45:01 +0000 (11:45 +0200)]
st/xorg: flush after loading the cursor

We need cursor data to land in destination buffer before drmModeSetCursor.
It fixes "cursor lag" on nv50.

13 years agogallium/draw: Fix enum type taken by draw_get_shader_param().
Michel Dänzer [Tue, 3 May 2011 11:23:14 +0000 (13:23 +0200)]
gallium/draw: Fix enum type taken by draw_get_shader_param().

Pointed out by clang:

src/gallium/auxiliary/draw/draw_context.h:251:41: warning: implicit conversion
from enumeration type 'enum pipe_cap' to different enumeration type
'enum pipe_shader_cap' [-Wconversion]
      return tgsi_exec_get_shader_param(param);
             ~~~~~~~~~~~~~~~~~~~~~~~~~~ ^~~~~

13 years agomesa: implement AMD_shader_stencil_export
Marek Olšák [Mon, 2 May 2011 14:41:04 +0000 (16:41 +0200)]
mesa: implement AMD_shader_stencil_export

It's just an alias of the ARB variant with some GLSL compiler changes.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agor600g: reorder caps
Marek Olšák [Tue, 3 May 2011 09:54:40 +0000 (11:54 +0200)]
r600g: reorder caps

13 years agor600g: set correct PIPE_CAP_MAX_COMBINED_SAMPLERS
Marek Olšák [Tue, 3 May 2011 09:54:07 +0000 (11:54 +0200)]
r600g: set correct PIPE_CAP_MAX_COMBINED_SAMPLERS

13 years agomesa: make AMD_draw_buffers_blend an alias of the ARB variant
Marek Olšák [Mon, 2 May 2011 10:59:24 +0000 (12:59 +0200)]
mesa: make AMD_draw_buffers_blend an alias of the ARB variant

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
13 years agoglapi: regenerate files for AMD_draw_buffers_blend
Marek Olšák [Mon, 2 May 2011 10:42:52 +0000 (12:42 +0200)]
glapi: regenerate files for AMD_draw_buffers_blend

13 years agoglapi: add functions for AMD_draw_buffers_blend
Marek Olšák [Mon, 2 May 2011 10:33:08 +0000 (12:33 +0200)]
glapi: add functions for AMD_draw_buffers_blend

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
13 years agomesa: flush vertices before changing GL_RASTERIZER_DISCARD state, not after
Marek Olšák [Mon, 2 May 2011 15:13:59 +0000 (17:13 +0200)]
mesa: flush vertices before changing GL_RASTERIZER_DISCARD state, not after

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
13 years agomesa: make _NEW_TEXTURE dirty when changing GL_TEXTURE_CUBE_MAP_SEAMLESS
Marek Olšák [Mon, 2 May 2011 00:36:42 +0000 (02:36 +0200)]
mesa: make _NEW_TEXTURE dirty when changing GL_TEXTURE_CUBE_MAP_SEAMLESS

Otherwise there would be no way to know whether the state has been changed.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
13 years agonv50,nvc0: advertise GL_FIXED vertex buffers as supported
Christoph Bumiller [Mon, 2 May 2011 21:55:05 +0000 (23:55 +0200)]
nv50,nvc0: advertise GL_FIXED vertex buffers as supported

We'll handle them like f64 vertex buffers, by falling back to
copying vertex data to the command buffer through translate.

13 years agomesa,st/mesa: fix WPOS adjustment
Christoph Bumiller [Mon, 2 May 2011 22:00:55 +0000 (00:00 +0200)]
mesa,st/mesa: fix WPOS adjustment

Tested-by: Marek Olšák <maraeo@gmail.com>
13 years agoDefault to --disable-gallium-r300 if not --with-driver=dri
Jon TURNEY [Fri, 29 Apr 2011 11:52:10 +0000 (12:52 +0100)]
Default to --disable-gallium-r300 if not --with-driver=dri

Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
Signed-off-by: Marek Olšák <maraeo@gmail.com>
13 years agogallium: reorder fields of pipe_rasterizer_state to pack it more tightly
Marek Olšák [Mon, 2 May 2011 00:04:17 +0000 (02:04 +0200)]
gallium: reorder fields of pipe_rasterizer_state to pack it more tightly

sizeof(struct pipe_rasterizer_state):
    Before: 32 bytes
    After: 28 bytes

Reviewed-by: Brian Paul <brianp@vmare.com>
13 years agor600g: document some new evergreen formats
Marek Olšák [Mon, 2 May 2011 15:11:46 +0000 (17:11 +0200)]
r600g: document some new evergreen formats

13 years agor600g: remove some more FIXME comments
Marek Olšák [Mon, 2 May 2011 09:39:47 +0000 (11:39 +0200)]
r600g: remove some more FIXME comments

13 years agost/xorg: Fix two more instances of typo 'accle'.
Michel Dänzer [Mon, 2 May 2011 07:36:49 +0000 (09:36 +0200)]
st/xorg: Fix two more instances of typo 'accle'.

13 years agost/xorg: fix typos
Marcin Slusarz [Sun, 1 May 2011 21:51:10 +0000 (23:51 +0200)]
st/xorg: fix typos

13 years agoi965/gen6: Align interleaved URB writes for overflow outputs as well.
Eric Anholt [Fri, 29 Apr 2011 22:59:30 +0000 (15:59 -0700)]
i965/gen6: Align interleaved URB writes for overflow outputs as well.

Fixes glsl-max-varyings.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=35614
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agotargets/xorg-nouveau: load nouveau_dri.so instead of i915_dri.so
Marcin Slusarz [Mon, 25 Apr 2011 21:59:40 +0000 (23:59 +0200)]
targets/xorg-nouveau: load nouveau_dri.so instead of i915_dri.so

13 years agotargets/xorg-nouveau: add libnvc0.a to nouveau libs
Marcin Slusarz [Mon, 25 Apr 2011 21:57:11 +0000 (23:57 +0200)]
targets/xorg-nouveau: add libnvc0.a to nouveau libs

13 years agor600g: remove some FIXME comments
Marek Olšák [Sun, 1 May 2011 23:43:33 +0000 (01:43 +0200)]
r600g: remove some FIXME comments

All texture LOD tests pass.

13 years agoRevert "r600g: truncate point sampled texture coordinates"
Marek Olšák [Sun, 1 May 2011 23:10:19 +0000 (01:10 +0200)]
Revert "r600g: truncate point sampled texture coordinates"

This reverts commit 1dc204d145dc8c0b19473a7814c201a8954b6274.

MC_COORD_TRUNCATE is for MPEG and produces quite an interesting behavior
on regular textures. Anyway that commit broke filtering in demos/cubemap.

13 years agost/mesa: remove set-but-unused variables
Marek Olšák [Sun, 1 May 2011 12:31:41 +0000 (14:31 +0200)]
st/mesa: remove set-but-unused variables

13 years agotgsi: remove set-but-unused variables
Marek Olšák [Sun, 1 May 2011 12:13:09 +0000 (14:13 +0200)]
tgsi: remove set-but-unused variables

13 years agodraw: remove set-but-unused variable
Marek Olšák [Sun, 1 May 2011 12:12:47 +0000 (14:12 +0200)]
draw: remove set-but-unused variable

13 years agorbug: remove set-but-unused variables
Marek Olšák [Sun, 1 May 2011 12:01:59 +0000 (14:01 +0200)]
rbug: remove set-but-unused variables

13 years agoir_to_mesa: remove set-but-unused variables
Marek Olšák [Sun, 1 May 2011 11:44:00 +0000 (13:44 +0200)]
ir_to_mesa: remove set-but-unused variables

13 years agomesa: remove set-but-unused variables in texcompress_s3tc
Marek Olšák [Sun, 1 May 2011 11:38:48 +0000 (13:38 +0200)]
mesa: remove set-but-unused variables in texcompress_s3tc

13 years agomesa: remove set-but-unused variable in bind_attrib_location
Marek Olšák [Sun, 1 May 2011 11:35:43 +0000 (13:35 +0200)]
mesa: remove set-but-unused variable in bind_attrib_location

13 years agomesa: remove unused-but-set variable in extract_uint_rgba
Marek Olšák [Sun, 1 May 2011 11:33:38 +0000 (13:33 +0200)]
mesa: remove unused-but-set variable in extract_uint_rgba

13 years agor300/compiler: remove set-but-unused variables
Marek Olšák [Sun, 1 May 2011 11:31:46 +0000 (13:31 +0200)]
r300/compiler: remove set-but-unused variables

13 years agor600g: remove set-but-unused variable
Marek Olšák [Sun, 1 May 2011 11:25:52 +0000 (13:25 +0200)]
r600g: remove set-but-unused variable

13 years agor600g: remove FIXME comment
Marek Olšák [Sat, 30 Apr 2011 22:04:06 +0000 (00:04 +0200)]
r600g: remove FIXME comment

DONTBLOCK is correctly handled in r600_bo_map.

13 years agomesa: GL_PROVOKING_VERTEX_EXT is a GLenum, not GLboolean.
José Fonseca [Sat, 30 Apr 2011 21:36:18 +0000 (22:36 +0100)]
mesa: GL_PROVOKING_VERTEX_EXT is a GLenum, not GLboolean.

13 years agor300/compiler: Cleanups from the regalloc merge
Tom Stellard [Sat, 30 Apr 2011 20:15:29 +0000 (13:15 -0700)]
r300/compiler: Cleanups from the regalloc merge

13 years agor300/compiler: Add return statement to rc_list()
Tom Stellard [Sat, 30 Apr 2011 19:52:32 +0000 (12:52 -0700)]
r300/compiler: Add return statement to rc_list()

13 years agor300g: Fix scons build
Tom Stellard [Sat, 30 Apr 2011 19:26:15 +0000 (12:26 -0700)]
r300g: Fix scons build

Broken by the dependency on ralloc introduced by
fe622bac0c1b5b9f2a9fcf9f35b51232a06bea42

13 years agoi915g: enable X-tiling for s3tc textures
Daniel Vetter [Sat, 30 Apr 2011 18:40:42 +0000 (20:40 +0200)]
i915g: enable X-tiling for s3tc textures

Tested-by: Christopher Egert <cme3000@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
13 years agor300/compiler: Enable swizzle packing in the allocator for r300 and r400
Tom Stellard [Tue, 26 Apr 2011 07:29:39 +0000 (00:29 -0700)]
r300/compiler: Enable swizzle packing in the allocator for r300 and r400

13 years agor300/compiler: Handle loops in the register allocator
Tom Stellard [Mon, 18 Apr 2011 05:33:04 +0000 (22:33 -0700)]
r300/compiler: Handle loops in the register allocator

13 years agor300/compiler: Handle loops in rc_get_readers()
Tom Stellard [Tue, 12 Apr 2011 23:08:25 +0000 (16:08 -0700)]
r300/compiler: Handle loops in rc_get_readers()

13 years agor300/compiler: Rewrite register allocator
Tom Stellard [Tue, 11 Jan 2011 08:05:08 +0000 (00:05 -0800)]
r300/compiler: Rewrite register allocator

The new allocator uses ra and does swizzle packing.

Also, a data structure (struct rc_variable) and associated functions have
been added for generating UD and DU chains.

13 years agora: Add ra_set_node_reg()
Tom Stellard [Sun, 27 Mar 2011 05:56:08 +0000 (22:56 -0700)]
ra: Add ra_set_node_reg()

This function can be used to avoid creating single register classes for
input/payload registers.  This makes optimistic coloring less likely
to fail.

Reviewed-by: Eric Anholt <eric@anholt.net>
13 years agor300/compiler: Use RC_MASK_W when converting RGB to Alpha instructions
Tom Stellard [Sun, 20 Mar 2011 07:09:30 +0000 (00:09 -0700)]
r300/compiler: Use RC_MASK_W when converting RGB to Alpha instructions

13 years agor300/compiler: rc_reader_data: Add ExitOnAbort flag
Tom Stellard [Sun, 30 Jan 2011 23:29:29 +0000 (15:29 -0800)]
r300/compiler: rc_reader_data: Add ExitOnAbort flag

13 years agor300/compiler: Add more info to struct rc_reader
Tom Stellard [Sat, 19 Mar 2011 04:34:56 +0000 (21:34 -0700)]
r300/compiler: Add more info to struct rc_reader

For pair instructions we need a reference to both the arg
and source.

13 years agor300/compiler: Add remove dead sources pass
Tom Stellard [Sat, 26 Mar 2011 00:05:53 +0000 (17:05 -0700)]
r300/compiler: Add remove dead sources pass

The instruction scheduler will sometimes leave orphaned sources when
converting instructions from RGB to Alpha.  If one of these orphaned
sources has an index greater than the maximum temporary register index,
then the compiler will incorrectly report "Too many hardware temporaries
used".  The dead sources pass cleans up these orphaned sources.

13 years agodocs: add ARB_ES2_compatibility in relnotes-7.11
Marek Olšák [Sat, 30 Apr 2011 10:00:18 +0000 (12:00 +0200)]
docs: add ARB_ES2_compatibility in relnotes-7.11

13 years agost/mesa: expose ARB_ES2_compatibility if GL_FIXED vertex format is supported
Marek Olšák [Thu, 28 Apr 2011 18:01:02 +0000 (20:01 +0200)]
st/mesa: expose ARB_ES2_compatibility if GL_FIXED vertex format is supported

Tested with softpipe and llvmpipe.

Reviewed-by: Brian Paul <brianp@vmware.com>
13 years agomesa: accept GL_FIXED in glVertexAttribPointer (ARB_ES2_compatibility)
Marek Olšák [Fri, 29 Apr 2011 12:23:15 +0000 (14:23 +0200)]
mesa: accept GL_FIXED in glVertexAttribPointer (ARB_ES2_compatibility)

GL_FIXED should not be accepted in the other gl*Pointer calls in OpenGL.

There is a new piglit for this: arb_es2_compatibility-fixed-type.

Reviewed-by: Brian Paul <brianp@vmware.com>
13 years agomesa: fix error string, remove out of date comment
Brian Paul [Sat, 30 Apr 2011 00:42:09 +0000 (18:42 -0600)]
mesa: fix error string, remove out of date comment

13 years agomesa: remove commented-out #include
Brian Paul [Sat, 30 Apr 2011 00:40:56 +0000 (18:40 -0600)]
mesa: remove commented-out #include

13 years agomesa: fix tarball creation
Thierry Vignaud [Fri, 29 Apr 2011 23:56:35 +0000 (17:56 -0600)]
mesa: fix tarball creation

Signed-off-by: Brian Paul <brianp@vmware.com>
13 years agoi965/gen6: Fix LogicOp handling for GL_COPY and/or floating-point RTs.
Eric Anholt [Fri, 29 Apr 2011 22:11:11 +0000 (15:11 -0700)]
i965/gen6: Fix LogicOp handling for GL_COPY and/or floating-point RTs.

We were accidentally leaving blending enabled for LogicOp GL_COPY,
which ARB_color_buffer_float/GL_RGBA32F-render (and friends) caught.
Additionally, the GL spec says that no LogicOp should be done to
floating-point targets, and the GPU gets really angry even if you say
to LogicOp GL_COPY to float.

13 years agoi965: Remove dead entrypoints to state cache, rename the one that's left.
Eric Anholt [Mon, 25 Apr 2011 03:48:33 +0000 (20:48 -0700)]
i965: Remove dead entrypoints to state cache, rename the one that's left.

As we expanded the usage of the state cache, it grew extra
functionality.  However, with the recent state streaming rework, we're
back to the state cache being used only for shader kernels, which is
the piece of GPU state that's actually expensive to compute again from
scratch, since it involves compiling.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen4: Move the GS state to state streaming.
Eric Anholt [Mon, 25 Apr 2011 03:36:22 +0000 (20:36 -0700)]
i965/gen4: Move the GS state to state streaming.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen4: Move clip state to state streaming
Eric Anholt [Mon, 25 Apr 2011 03:20:16 +0000 (20:20 -0700)]
i965/gen4: Move clip state to state streaming

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965: Drop the now unused brw_cache_data() function.
Eric Anholt [Mon, 25 Apr 2011 03:05:02 +0000 (20:05 -0700)]
i965: Drop the now unused brw_cache_data() function.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965: Fix up state dumping for SF state.
Eric Anholt [Mon, 25 Apr 2011 02:57:33 +0000 (19:57 -0700)]
i965: Fix up state dumping for SF state.

It was moved to state streaming a while back and this was left over.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen4: Move VS state to state streaming.
Eric Anholt [Mon, 25 Apr 2011 02:47:37 +0000 (19:47 -0700)]
i965/gen4: Move VS state to state streaming.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agomesa: Add a bunch of documentation to the register allocator.
Eric Anholt [Sun, 24 Apr 2011 20:44:32 +0000 (13:44 -0700)]
mesa: Add a bunch of documentation to the register allocator.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Use the dynamic state base address to reduce relocations.
Eric Anholt [Fri, 22 Apr 2011 23:00:14 +0000 (16:00 -0700)]
i965/gen6: Use the dynamic state base address to reduce relocations.

Now that all the dynamic state is streamed through the top of the
batchbuffer, we can cut out many of our relocations to that state by
using the base address.

Improves 3DMMES taiji performance 3.3% +/- 0.4% (n=15).

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Move scissor state to state streaming.
Eric Anholt [Mon, 25 Apr 2011 03:02:38 +0000 (20:02 -0700)]
i965/gen6: Move scissor state to state streaming.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965: Move sampler state to state streaming.
Eric Anholt [Fri, 22 Apr 2011 22:30:42 +0000 (15:30 -0700)]
i965: Move sampler state to state streaming.

Overall, across this series since the last set of numbers, gen6 3DMMES
taiji performance has dropped 0.8% +/- 0.3% (n=15), probably due to
the increased reissuing of state from some of the state objects that
otherwise never changed, and increased occurrence of the per-batch
overhead as we've increased how much we put in the batch BO without
increasing the batch BO's size.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen4: Move WM state to state streaming.
Eric Anholt [Sat, 23 Apr 2011 07:33:43 +0000 (00:33 -0700)]
i965/gen4: Move WM state to state streaming.

The samplers are about to become streamed for gen6 performance, which
would cause this unit to blow out the state cache.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Move the depth/stencil state to state streaming.
Eric Anholt [Fri, 22 Apr 2011 21:26:12 +0000 (14:26 -0700)]
i965/gen6: Move the depth/stencil state to state streaming.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Move the blend state to state streaming.
Eric Anholt [Fri, 22 Apr 2011 21:07:30 +0000 (14:07 -0700)]
i965/gen6: Move the blend state to state streaming.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Move the color calc state to state streaming.
Eric Anholt [Fri, 22 Apr 2011 20:54:10 +0000 (13:54 -0700)]
i965/gen6: Move the color calc state to state streaming.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965: Move the CC VP to state streaming.
Eric Anholt [Fri, 22 Apr 2011 20:44:00 +0000 (13:44 -0700)]
i965: Move the CC VP to state streaming.

This is in a way a revert of f5bb775fd1f333d8e579d07a5cac1ded2bd54a2f.
The tiny win that had will be overwhelmed by the win of using the gen6
dynamic state base address.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Move the SF VP to state streaming.
Eric Anholt [Fri, 22 Apr 2011 20:25:24 +0000 (13:25 -0700)]
i965/gen6: Move the SF VP to state streaming.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Move the clip VP to state streaming.
Eric Anholt [Fri, 22 Apr 2011 20:20:08 +0000 (13:20 -0700)]
i965/gen6: Move the clip VP to state streaming.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Stream the VS push constants.
Eric Anholt [Fri, 22 Apr 2011 16:55:25 +0000 (09:55 -0700)]
i965/gen6: Stream the VS push constants.

Improves 3DMMES taiji demo performance by 10.1% +/- 0.9% (n=15).

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965/gen6: Stream the WM push constants.
Eric Anholt [Fri, 22 Apr 2011 02:03:18 +0000 (19:03 -0700)]
i965/gen6: Stream the WM push constants.

Improves 3DMMES taiji demo performance by 5.1% +/- 1.9% (n=15), by
reducing CPU time spent thrashing around those tiny little constant BOs.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965: Give the push constants and pull constants unique function names.
Eric Anholt [Fri, 22 Apr 2011 01:33:22 +0000 (18:33 -0700)]
i965: Give the push constants and pull constants unique function names.

This helps clarify profiling results.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
13 years agoi965: Fix fragcoord_w on gen6 with 16-wide.
Eric Anholt [Fri, 29 Apr 2011 19:17:02 +0000 (12:17 -0700)]
i965: Fix fragcoord_w on gen6 with 16-wide.

The payload regs can go all the way up to register 60+, so just give
them 8 bits to be addressed by instead of 3-4 (which made source_w_reg
of 8 end up 0).  There's no reason to aggressively pack these fields,
as they are just used as compiler information, where being easier to
access is probably more important than shaving a byte or two off of
the structure.

Fixes piglit fragcoord_w.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=36649

13 years agometa: Don't ask for floating point textures if not ARB_texture_float.
Eric Anholt [Tue, 26 Apr 2011 19:56:37 +0000 (12:56 -0700)]
meta: Don't ask for floating point textures if not ARB_texture_float.

I was promoting to float for ARB_color_buffer_float unclamped, which
failed when ARB_texture_float wasn't present.  Since the metaops don't
need results outside of [0,1] when not drawing to a floating point
destination, they can just use a fixed point texture when floating
point destinations are impossible.

Fixes regression in fdo23670-depth_test when --enable-texture-float is
not present.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=36473

13 years agomesa: fix a few incorrect error messages
Brian Paul [Fri, 29 Apr 2011 18:03:28 +0000 (12:03 -0600)]
mesa: fix a few incorrect error messages

13 years agoegl/wayland: Fix possible lockup in drm initialization
Benjamin Franzke [Fri, 29 Apr 2011 15:04:11 +0000 (17:04 +0200)]
egl/wayland: Fix possible lockup in drm initialization

Lockup happens when wl_drm interface is not available.

13 years agoAdd git_sha1.h to .gitignore
Jon TURNEY [Tue, 26 Apr 2011 12:11:23 +0000 (13:11 +0100)]
Add git_sha1.h to .gitignore

Since commit de579a1 "Include GIT SHA1 in GL version string"

$ git status
 On branch master
 Your branch is ahead of 'origin/master' by 2 commits.

 Untracked files:
   (use "git add <file>..." to include in what will be committed)

       src/mesa/main/git_sha1.h
nothing added to commit but untracked files present (use "git add" to track)

Add git_sha1.h to .gitignore so git knows not to warn it is present but untracked

Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
13 years agor600g: print opcodes names instead of numbers
Marek Olšák [Fri, 29 Apr 2011 11:16:24 +0000 (13:16 +0200)]
r600g: print opcodes names instead of numbers

13 years agodocs: update relnotes-7.11
Marek Olšák [Thu, 28 Apr 2011 15:33:05 +0000 (17:33 +0200)]
docs: update relnotes-7.11

13 years agodocs: update GL3 status
Marek Olšák [Thu, 28 Apr 2011 15:30:55 +0000 (17:30 +0200)]
docs: update GL3 status