mesa.git
14 years agor300g: fix misaligned generated offset for ubyte vertex indices
Marek Olšák [Sun, 21 Mar 2010 03:39:33 +0000 (04:39 +0100)]
r300g: fix misaligned generated offset for ubyte vertex indices

14 years agor300g: split the vertex buffer alignment validation
Marek Olšák [Sat, 20 Mar 2010 21:14:59 +0000 (22:14 +0100)]
r300g: split the vertex buffer alignment validation

14 years agor300g: skip null vertex buffers
Marek Olšák [Sat, 20 Mar 2010 20:26:19 +0000 (21:26 +0100)]
r300g: skip null vertex buffers

14 years agor300/compiler: fix assertion failure in the r500-fragprog emission path
Marek Olšák [Sat, 20 Mar 2010 17:31:11 +0000 (18:31 +0100)]
r300/compiler: fix assertion failure in the r500-fragprog emission path

14 years agor300g: Bump immediate limits.
Corbin Simpson [Sun, 21 Mar 2010 00:18:22 +0000 (17:18 -0700)]
r300g: Bump immediate limits.

Seems like a decent idea, especially since the big barrier now is
getting the VBOs back from the VRAM boundary.

14 years agor300g: Correctly hax max_index on pipe_vertex_buffers.
Corbin Simpson [Sun, 21 Mar 2010 00:16:46 +0000 (17:16 -0700)]
r300g: Correctly hax max_index on pipe_vertex_buffers.

Still not happy with this, but at least things seem to work.

14 years agoradeong: Use TRUE and FALSE.
Corbin Simpson [Sat, 20 Mar 2010 23:15:36 +0000 (16:15 -0700)]
radeong: Use TRUE and FALSE.

14 years agoradeong: Test the BO ptr for NULL instead of using a bool.
Corbin Simpson [Sat, 20 Mar 2010 23:14:20 +0000 (16:14 -0700)]
radeong: Test the BO ptr for NULL instead of using a bool.

Slight memory savings, but more importantly also avoids a potential NULL
dereference.

14 years agor300g: winsys fix tiling change flushes.
Dave Airlie [Sat, 20 Mar 2010 20:34:11 +0000 (06:34 +1000)]
r300g: winsys fix tiling change flushes.

If we change the tiling on a buffer we need to flush it, the old
radeon_buffer.c code had this but it crossed streams when I ported to
radeon_drm_buffer.c and I missed it. Should fix some piglit regressions.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agor300g: add buffer/texture referenced checks.
Dave Airlie [Sat, 20 Mar 2010 20:32:48 +0000 (06:32 +1000)]
r300g: add buffer/texture referenced checks.

I've no idea about the comments that were in there, just pass
this down to the winsys.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodocs: note GL_EXT_packed_float, GL_EXT_shared_exponent extensions
Brian Paul [Sat, 20 Mar 2010 18:05:25 +0000 (12:05 -0600)]
docs: note GL_EXT_packed_float, GL_EXT_shared_exponent extensions

14 years agomesa: added GL3 query for GL_TEXTURE_SHARED_SIZE
Brian Paul [Sat, 20 Mar 2010 18:04:52 +0000 (12:04 -0600)]
mesa: added GL3 query for GL_TEXTURE_SHARED_SIZE

14 years agomesa: added missing glGet query for GL_MAX_ARRAY_TEXTURE_LAYERS_EXT
Brian Paul [Sat, 20 Mar 2010 17:55:24 +0000 (11:55 -0600)]
mesa: added missing glGet query for GL_MAX_ARRAY_TEXTURE_LAYERS_EXT

14 years agomesa: added new GL3 buffer object queries
Brian Paul [Sat, 20 Mar 2010 17:52:12 +0000 (11:52 -0600)]
mesa: added new GL3 buffer object queries

And clean up the error checking code.

14 years agomesa: added GL3 buffer attachment aliases
Brian Paul [Sat, 20 Mar 2010 17:50:55 +0000 (11:50 -0600)]
mesa: added GL3 buffer attachment aliases

14 years agomesa: added GL3 ContextFlags field and query code
Brian Paul [Sat, 20 Mar 2010 17:50:29 +0000 (11:50 -0600)]
mesa: added GL3 ContextFlags field and query code

14 years agogallivm: simplify and clean-up Z/stencil bit mask/shift code
Brian Paul [Fri, 19 Mar 2010 16:02:16 +0000 (10:02 -0600)]
gallivm: simplify and clean-up Z/stencil bit mask/shift code

Refactor the code into two helper functions which compute the bit mask
and shift terms for Z and stencil.  Plus add a bunch of new comments to
explain everything.

14 years agocso: remove cso_set/save/restore_sampler_textures
Roland Scheidegger [Fri, 19 Mar 2010 15:30:47 +0000 (16:30 +0100)]
cso: remove cso_set/save/restore_sampler_textures

no longer used after all statetrackers have been converted.

14 years agost/vega: fix up vega state tracker to use cso changes
Roland Scheidegger [Fri, 19 Mar 2010 15:29:22 +0000 (16:29 +0100)]
st/vega: fix up vega state tracker to use cso changes

use cso fragment sampler views instead of sampler textures.
since we don't really change views, try to store sampler views instead
of the textures to avoid having to recreate views most of the time.

14 years agoes: added -I$(TOP)/include path
Brian Paul [Fri, 19 Mar 2010 13:36:50 +0000 (07:36 -0600)]
es: added -I$(TOP)/include path

14 years agogallium/docs: Clarify sampler view descriptions.
Michal Krol [Fri, 19 Mar 2010 08:08:33 +0000 (09:08 +0100)]
gallium/docs: Clarify sampler view descriptions.

14 years agogallium/docs: Fix a couple ReST errors.
Corbin Simpson [Fri, 19 Mar 2010 07:07:55 +0000 (00:07 -0700)]
gallium/docs: Fix a couple ReST errors.

14 years agor300g: lod min/max clamping should be relative to the base level
Marek Olšák [Fri, 19 Mar 2010 03:46:15 +0000 (04:46 +0100)]
r300g: lod min/max clamping should be relative to the base level

14 years agor300g: fix breakage after the gallium-sampler-view merge
Marek Olšák [Fri, 19 Mar 2010 00:13:57 +0000 (01:13 +0100)]
r300g: fix breakage after the gallium-sampler-view merge

14 years agor300g: add PIPE_FORMAT_B5G5R5X1_UNORM colorbuffer support, cleanups
Marek Olšák [Thu, 18 Mar 2010 20:18:36 +0000 (21:18 +0100)]
r300g: add PIPE_FORMAT_B5G5R5X1_UNORM colorbuffer support, cleanups

14 years agor300g: mark SRGB colorbuffers as unsupported
Marek Olšák [Thu, 18 Mar 2010 19:58:04 +0000 (20:58 +0100)]
r300g: mark SRGB colorbuffers as unsupported

I can't find the register bits for SRGB-aware blending and it's not even
exposed by fglrx so it's most probably not supported by hw.

14 years agor300g: add generating texture coordinates for point sprites (WIP)
Marek Olšák [Thu, 24 Dec 2009 20:29:41 +0000 (21:29 +0100)]
r300g: add generating texture coordinates for point sprites (WIP)

14 years agogallivm: support PIPE_FORMAT_Z24S8_UNORM in depth/stencil code
Brian Paul [Thu, 18 Mar 2010 23:27:39 +0000 (17:27 -0600)]
gallivm: support PIPE_FORMAT_Z24S8_UNORM in depth/stencil code

14 years agoprogs/tests: also test stencil INCR_WRAP mode if supported
Brian Paul [Thu, 18 Mar 2010 21:04:31 +0000 (15:04 -0600)]
progs/tests: also test stencil INCR_WRAP mode if supported

14 years agogallivm: fix broken INCR/DECR stencil modes
Brian Paul [Thu, 18 Mar 2010 21:02:13 +0000 (15:02 -0600)]
gallivm: fix broken INCR/DECR stencil modes

We were mistakenly using the wrong data type for stencil values before.

14 years agoprogs/tests: comments and clean-ups
Brian Paul [Thu, 18 Mar 2010 20:53:43 +0000 (14:53 -0600)]
progs/tests: comments and clean-ups

14 years agogallivm/llvmpipe: added lp_rast_shader_inputs::facing and pass through
Brian Paul [Thu, 18 Mar 2010 19:02:53 +0000 (13:02 -0600)]
gallivm/llvmpipe: added lp_rast_shader_inputs::facing and pass through

The triangle rasterizer sets this field to indicate front/back-facing.
It gets passed into the generated fragment code as another parameter.
Used now for stencil front/back selection but will also be used for
fragment shaders in general (see TGSI_SEMANTIC_FACE).

With this commit two-sided stenciling mostly works but there's
still a bug or two...

14 years agollvmpipe: defines for RAST_WHOLE, RAST_EDGE_TEST
Brian Paul [Thu, 18 Mar 2010 18:04:50 +0000 (12:04 -0600)]
llvmpipe: defines for RAST_WHOLE, RAST_EDGE_TEST

14 years agoAdd format B5G5R5X1
Alan Hourihane [Thu, 18 Mar 2010 18:01:05 +0000 (18:01 +0000)]
Add format B5G5R5X1

14 years agogallivm/llvmpipe: simplify front/back stencil ref value handling
Brian Paul [Thu, 18 Mar 2010 17:31:38 +0000 (11:31 -0600)]
gallivm/llvmpipe: simplify front/back stencil ref value handling

Instead of passing an array, just pass two scalar values.

14 years agogallivm: checkpoint WIP two-sided stencil test
Brian Paul [Thu, 18 Mar 2010 17:18:01 +0000 (11:18 -0600)]
gallivm: checkpoint WIP two-sided stencil test

14 years agollvmpipe: set opaque = FALSE if stencil enabled
Brian Paul [Thu, 18 Mar 2010 15:30:13 +0000 (09:30 -0600)]
llvmpipe: set opaque = FALSE if stencil enabled

Fixes occasional bad tiles seen in some demos like progs/demos/reflect.c

14 years agodri/nouveau: Flush after texture validation.
Francisco Jerez [Thu, 18 Mar 2010 12:46:20 +0000 (13:46 +0100)]
dri/nouveau: Flush after texture validation.

Swizzling needs the destination surface in VRAM, but the subsequent
rendering operations making use of it are likely to not care. Fire the
ring after validation to leave the memory manager more room for
maneuvering.

14 years agodri/nouveau: Avoid pushbuf flushes in the middle of LMA setup.
Francisco Jerez [Thu, 18 Mar 2010 13:19:10 +0000 (14:19 +0100)]
dri/nouveau: Avoid pushbuf flushes in the middle of LMA setup.

14 years agodri/nouveau: Implement texcoord generation.
Francisco Jerez [Thu, 18 Mar 2010 13:18:55 +0000 (14:18 +0100)]
dri/nouveau: Implement texcoord generation.

14 years agodri/nouveau: Implement texture matrices.
Francisco Jerez [Thu, 18 Mar 2010 13:13:36 +0000 (14:13 +0100)]
dri/nouveau: Implement texture matrices.

14 years agodri/nouveau: Some minor vertex submission fixes.
Francisco Jerez [Thu, 11 Mar 2010 20:48:51 +0000 (21:48 +0100)]
dri/nouveau: Some minor vertex submission fixes.

14 years agodri/nouveau: only reallocate texture when needed
Xavier Chantry [Sat, 13 Mar 2010 18:28:07 +0000 (19:28 +0100)]
dri/nouveau: only reallocate texture when needed

nouveau reallocated the mipmap tree on every MIN_FILTER call to account
for mipmap change. We only need to do this if the texture does not fit
in the existing mipmap tree. This gives a big performance boost for a
game like bzflag which changes MIN_FILTER all the time for its font
rendering.

Signed-off-by: Xavier Chantry <chantry.xavier@gmail.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
14 years agollvmpipe: Fix crashes when there is no depth buffer bound.
José Fonseca [Thu, 18 Mar 2010 10:24:10 +0000 (10:24 +0000)]
llvmpipe: Fix crashes when there is no depth buffer bound.

14 years agointel: Call _mesa_make_current() after getting initial buffers
Kristian Høgsberg [Thu, 18 Mar 2010 02:44:58 +0000 (22:44 -0400)]
intel: Call _mesa_make_current() after getting initial buffers

The default viewport is the window rectangle, which is set up by
_mesa_make_current().  To be able to do that we need to get the
window dimension (and buffers) first, so we have to call
intel_prepare_render() before we can call into _mesa_make_current().

Fixes #26676 and #26678.

14 years agoprogs/trivial: added comments
Brian Paul [Wed, 17 Mar 2010 22:26:41 +0000 (16:26 -0600)]
progs/trivial: added comments

14 years agogallivm/llvmpipe: basic stencil testing works
Brian Paul [Wed, 17 Mar 2010 22:24:12 +0000 (16:24 -0600)]
gallivm/llvmpipe: basic stencil testing works

Most stencil demos look OK (modulo some unrelated rendering glitches).
Only single-sided stencil test works at this point.
There are probably some bugs to be found...

14 years agollvmpipe: remove incorrect depth test check
Brian Paul [Wed, 17 Mar 2010 21:09:35 +0000 (15:09 -0600)]
llvmpipe: remove incorrect depth test check

14 years agogallivm: added lp_build_andc()
Brian Paul [Wed, 17 Mar 2010 21:07:45 +0000 (15:07 -0600)]
gallivm: added lp_build_andc()

14 years agollvmpipe: silence some pointer/casting warnings
Brian Paul [Wed, 17 Mar 2010 14:34:23 +0000 (08:34 -0600)]
llvmpipe: silence some pointer/casting warnings

14 years agogallivm/llvmpipe: more asst changes for stencil testing
Brian Paul [Wed, 17 Mar 2010 00:26:51 +0000 (18:26 -0600)]
gallivm/llvmpipe: more asst changes for stencil testing

14 years agogallivm/llmvpipe: pass stencil refs state into z/stencil build code
Brian Paul [Tue, 16 Mar 2010 20:32:18 +0000 (14:32 -0600)]
gallivm/llmvpipe: pass stencil refs state into z/stencil build code

14 years agollvmpipe: use new LP_JIT_CTX_ enums for jit context field positions
Brian Paul [Tue, 16 Mar 2010 20:11:43 +0000 (14:11 -0600)]
llvmpipe: use new LP_JIT_CTX_ enums for jit context field positions

Use the new enum values rather than integers in a few places.

14 years agollvmpipe: added stencil ref values to jit context state
Brian Paul [Tue, 16 Mar 2010 20:00:40 +0000 (14:00 -0600)]
llvmpipe: added stencil ref values to jit context state

14 years agollvmpipe: break lines
Brian Paul [Tue, 16 Mar 2010 19:50:19 +0000 (13:50 -0600)]
llvmpipe: break lines

14 years agointel: Replace mt->pitch with mt->region->pitch.
Eric Anholt [Wed, 17 Mar 2010 16:09:54 +0000 (09:09 -0700)]
intel: Replace mt->pitch with mt->region->pitch.

The pitch is not really an inherent part of the miptree, since it's
not part of any of the layout calculations, and it's dictated by the
libdrm-allocated region pitch now.

14 years agointel: Return false like other blit failure paths if out of aperture.
Eric Anholt [Wed, 17 Mar 2010 16:27:48 +0000 (09:27 -0700)]
intel: Return false like other blit failure paths if out of aperture.

The primary consumer of this (miptree relayout) already has this code
for handling failure, and the other paths want to know if failure
actually occurs and do something appropriate, which may not include
memcpy.

14 years agointel: Respect src pitch in _mesa_copy_rect().
Eric Anholt [Wed, 17 Mar 2010 17:10:37 +0000 (10:10 -0700)]
intel: Respect src pitch in _mesa_copy_rect().

If a non-zero src_y was used, this would break piglit
depth-level-clamp.

14 years agointel: Rely on allocated region pitch for the miptree pitch.
Eric Anholt [Tue, 16 Mar 2010 23:20:03 +0000 (16:20 -0700)]
intel: Rely on allocated region pitch for the miptree pitch.

Bug #26966: 945 miptree pitch disagreement with libdrm.

14 years agointel: Assert that the linear blits succeed.
Eric Anholt [Wed, 17 Mar 2010 16:26:37 +0000 (09:26 -0700)]
intel: Assert that the linear blits succeed.

We don't have any fallback code here, and we want to avoid this path
if failure would happen, so just assert.

14 years agointel: Remove level_offset now that it's unused.
Eric Anholt [Tue, 16 Mar 2010 23:11:05 +0000 (16:11 -0700)]
intel: Remove level_offset now that it's unused.

This is the last pitch-dependent part of miptree setup.

14 years agointel: Remove extra tiling setting after allocating a tiled region.
Eric Anholt [Tue, 16 Mar 2010 23:05:53 +0000 (16:05 -0700)]
intel: Remove extra tiling setting after allocating a tiled region.

14 years agointel: Rename the z24_x8 depth spans to z24_s8 since they do stencil too.
Eric Anholt [Tue, 16 Mar 2010 21:02:22 +0000 (14:02 -0700)]
intel: Rename the z24_x8 depth spans to z24_s8 since they do stencil too.

14 years agollvmpipe: remove -m32 flag from linux-llvm config
Brian Paul [Wed, 17 Mar 2010 16:53:55 +0000 (10:53 -0600)]
llvmpipe: remove -m32 flag from linux-llvm config

14 years agost/mesa: plug in default for pipe_context::surface_copy() if needed
Brian Paul [Wed, 17 Mar 2010 16:31:57 +0000 (10:31 -0600)]
st/mesa: plug in default for pipe_context::surface_copy() if needed

This lets us avoid conditionals and duplicated code in several places.

14 years agoprogs/samples: improve copy.c demo
Brian Paul [Wed, 17 Mar 2010 16:17:04 +0000 (10:17 -0600)]
progs/samples: improve copy.c demo

If the test image was larger than the window, nothing was drawn because
of invalid raster position.  Use glWindowPos instead of glRasterPos.
Also, use integer src/dst coordinates to avoid grabbing black pixels
outside of the src image region.

14 years agoMerge branch '7.8'
Brian Paul [Wed, 17 Mar 2010 16:13:51 +0000 (10:13 -0600)]
Merge branch '7.8'

Conflicts:

src/mesa/state_tracker/st_cb_drawpixels.c

14 years agoprogs/samples: silence warnings
Brian Paul [Wed, 17 Mar 2010 16:11:09 +0000 (10:11 -0600)]
progs/samples: silence warnings

14 years agost/mesa: fix glCopyPixels bugs/crashes when src region need clipping
Brian Paul [Wed, 17 Mar 2010 16:06:27 +0000 (10:06 -0600)]
st/mesa: fix glCopyPixels bugs/crashes when src region need clipping

Use the _mesa_clip_readpixels() function to clip the src region against
the buffer's bounds.  Neatly, the resulting pixel unpack object's
SkipPixels/SkipRows fields can be used to determine the position of the
region in the destination texture.

Fixes crash in progs/samples/copy.c and probably other cases.

14 years agomesa: rename params in prototype to match implementation
Brian Paul [Wed, 17 Mar 2010 15:04:26 +0000 (09:04 -0600)]
mesa: rename params in prototype to match implementation

14 years agoglapi: fix assorted warnings
Brian Paul [Wed, 17 Mar 2010 14:44:09 +0000 (08:44 -0600)]
glapi: fix assorted warnings

And replace some instances of GLuint with unsigned int to avoid pulling in
GL/gl.h

14 years agoswrast: remove unused compute_coveragei() function
Brian Paul [Wed, 17 Mar 2010 14:41:47 +0000 (08:41 -0600)]
swrast: remove unused compute_coveragei() function

14 years agost/mesa: Fix build breakage.
Corbin Simpson [Wed, 17 Mar 2010 11:35:14 +0000 (04:35 -0700)]
st/mesa: Fix build breakage.

Nearly certain this is what was intended; it compiles, but I'm not sure
this path is ever hit in my tests.

14 years agoMerge branch '7.8' into master
Pauli Nieminen [Wed, 17 Mar 2010 09:26:48 +0000 (11:26 +0200)]
Merge branch '7.8' into master

Conflicts:
Makefile
src/mesa/main/version.h

14 years agocell: build identity driver too
Brian Paul [Tue, 16 Mar 2010 22:45:54 +0000 (16:45 -0600)]
cell: build identity driver too

14 years agocell: add missing semicolon
Brian Paul [Tue, 16 Mar 2010 22:45:36 +0000 (16:45 -0600)]
cell: add missing semicolon

14 years agonv50: support more formats in surface_copy,fill
Christoph Bumiller [Tue, 16 Mar 2010 21:32:42 +0000 (22:32 +0100)]
nv50: support more formats in surface_copy,fill

Fixes corrupted fonts in bzFlag, where we've been silently
failing to copy I8 mipmaps to a new miptree.

Print an error message on unsupported format now, since we
can't return failure.

14 years agoi965: Fix readpixels from ReadBuffer != DrawBuffer.
Eric Anholt [Tue, 16 Mar 2010 20:23:23 +0000 (13:23 -0700)]
i965: Fix readpixels from ReadBuffer != DrawBuffer.

Fixes piglit fbo-readdrawpix.

14 years agoi965: Fix inversion for glCopyPixels to/from FBOs.
Eric Anholt [Tue, 16 Mar 2010 20:14:38 +0000 (13:14 -0700)]
i965: Fix inversion for glCopyPixels to/from FBOs.

fixes piglit fbo-copypix.

14 years agointel: Remove more code for x8z24 visuals, since we only do s8z24.
Eric Anholt [Tue, 16 Mar 2010 19:59:59 +0000 (12:59 -0700)]
intel: Remove more code for x8z24 visuals, since we only do s8z24.

14 years agoRevert "i965: Do VS SGT, SLT, and friends using CMP, SEL instead of CMP, MOV, MOV."
Eric Anholt [Tue, 16 Mar 2010 18:23:08 +0000 (11:23 -0700)]
Revert "i965: Do VS SGT, SLT, and friends using CMP, SEL instead of CMP, MOV, MOV."

This reverts commit 8ef3b1834a896927bdd4f2aea552cdb732849da9.  Fixes
piglit glsl-vs-if.

14 years agoRevert "i965: Do FS SLT, SGT, and friends using CMP, SEL instead of CMP, MOV, MOV."
Eric Anholt [Tue, 16 Mar 2010 18:22:29 +0000 (11:22 -0700)]
Revert "i965: Do FS SLT, SGT, and friends using CMP, SEL instead of CMP, MOV, MOV."

This reverts commit 46450c1f3f93bf4dc96696fc7e0f0eb808d9c08a.  I was
wrong about null reg behavior -- it reads undefined, not 0.  And
they're not kidding.

14 years agometa: Fix up restoration of state if _mesa_map_pbo_source() fails.
Eric Anholt [Sat, 13 Mar 2010 01:12:14 +0000 (17:12 -0800)]
meta: Fix up restoration of state if _mesa_map_pbo_source() fails.

14 years agometa: Properly refcount our saved programs and texobjs.
Eric Anholt [Sat, 13 Mar 2010 01:09:51 +0000 (17:09 -0800)]
meta: Properly refcount our saved programs and texobjs.

Found while debugging bug #24119.

14 years agogallium: add target-helpers/wrap_screen.c to C_SOURCES
Brian Paul [Tue, 16 Mar 2010 19:31:42 +0000 (13:31 -0600)]
gallium: add target-helpers/wrap_screen.c to C_SOURCES

Was commented out before.

14 years agoi965: Fix ENDLOOP to only patch up this loop's BREAK and CONT.
Eric Anholt [Tue, 9 Mar 2010 22:22:51 +0000 (14:22 -0800)]
i965: Fix ENDLOOP to only patch up this loop's BREAK and CONT.

Corresponds to d225a25e21a24508aea3b877c78beb35502e942d and fixes
piglit glsl-fs-loop-nested.  Bug #25173.
(cherry picked from commit a81836ee2fe5092d695b717addf8cec91f569777)

14 years agoi965: Unalias all GLSL source regs from the destination regs used.
Eric Anholt [Tue, 9 Mar 2010 19:56:14 +0000 (11:56 -0800)]
i965: Unalias all GLSL source regs from the destination regs used.

We were doing it ad-hoc before, as instructions with potential
aliasing problems were identified.  But thanks to swizzling basically
anything can have aliasing, so just do it generally at source reg
setup time.  This is somewhat inefficient, because sometimes an
operation doesn't need unaliasing protection if the swizzling is safe,
but the unaliasing before didn't cover those cases either.

Fixes piglit glsl-fs-loop.
(cherry picked from commit 6b194dab6b4d9f12cdd54c699b23c0d3420a49c2)

14 years agoi965: Fix nested loops in the VS.
Eric Anholt [Tue, 9 Mar 2010 17:56:42 +0000 (09:56 -0800)]
i965: Fix nested loops in the VS.

We were patching up all the break and continues between the start of
our loop and the end of our loop, even if they were breaks/continues
for an inner loop.  Avoiding patching already patched breaks/continues
fixes piglit glsl-vs-loop-nested.
(cherry picked from commit f6f547d87ea68f44c50a0b0231b7360ca94b2975)

14 years agoi965: Fix up VP constbuf leak on program delete.
Eric Anholt [Tue, 9 Mar 2010 00:08:33 +0000 (16:08 -0800)]
i965: Fix up VP constbuf leak on program delete.
(cherry picked from commit 7f6d2754d586545ab6c970acffdd897294879039)

14 years agoi965: Fix up the handling of point sprite coordinate replacement.
Eric Anholt [Mon, 8 Mar 2010 20:12:31 +0000 (12:12 -0800)]
i965: Fix up the handling of point sprite coordinate replacement.

The code was walking over the regs of pairs of attributes and checking
whether the attribute with a given reg index had point sprite enabled.
So the point sprite setup code was rarely even getting executed.
Instead, we need to determine which channels of a reg need point
sprite coordinate replacement.  In addition, it was multiplying the
attribute by 1/w, when it's supposed to cover (0, 1) in each direction
regardless of w, and it wasn't filling in the Z and W components of
the texcoord as specified.

Fixes piglit point-sprite and the spriteblast demo.  Bug #24431, #22245.
(cherry picked from commit bc632d04370566c1156cbd0345fe303834f0b910)

14 years agoi965: Add support for the CMP opcode in the GLSL path.
Eric Anholt [Wed, 10 Mar 2010 19:00:40 +0000 (11:00 -0800)]
i965: Add support for the CMP opcode in the GLSL path.

This would be triggered by use of sqrt() along with control flow.
Fixes piglit-fs-sqrt-branch and a bug in Yo Frankie!.
(cherry picked from commit 48dca99feb394febc3af44e14f23fb12a9cc9204)

14 years agoi965: Fix the response len of masked sampler messages for 8-wide dispatch.
Eric Anholt [Wed, 10 Mar 2010 18:38:20 +0000 (10:38 -0800)]
i965: Fix the response len of masked sampler messages for 8-wide dispatch.

The bad response length would hang the GPU with a masked sample in a
shader using control flow.  For 8-wide, the response length is always
4, and masked slots are just not written to.  brw_wm_glsl.c already
allocates registers in the right locations.

Fixes piglit glsl-fs-bug25902 (fd.o bug #25902).
(cherry picked from commit f6d210c284751ac50a8d6358de7e75a1ff1e4ac7)
(cherry picked from commit dc8c0359448cdae7b367552ba58783c04b199778)

14 years agost/mesa: Invalidate sampler view when texture object changes.
Michal Krol [Tue, 16 Mar 2010 18:39:09 +0000 (19:39 +0100)]
st/mesa: Invalidate sampler view when texture object changes.

14 years agoRemove nonexistant files from tarballs rule of Makefile
Ian Romanick [Tue, 16 Mar 2010 14:48:14 +0000 (07:48 -0700)]
Remove nonexistant files from tarballs rule of Makefile

The Makefile should eventually be fixed to stop building tarballs if
one of the individual calls to tar fails.

14 years agomesa: set version string to 7.8-rc1
Ian Romanick [Tue, 16 Mar 2010 14:38:51 +0000 (07:38 -0700)]
mesa: set version string to 7.8-rc1

14 years agoCorrect GL_EQUIV code in r67/7xx.
Matthew W. S. Bell [Mon, 15 Mar 2010 01:18:02 +0000 (01:18 +0000)]
Correct GL_EQUIV code in r67/7xx.

From 247e121106e8d3e389f2e5a6edf13ea70ac18df7 Mon Sep 17 00:00:00 2001

These seem to be documented in
<http://www.svgopen.org/2003/papers/RasterOperationsUsingFilterElements/index.html>.

14 years agost/xorg: fix up xorg state tracker to use cso changes
Roland Scheidegger [Tue, 16 Mar 2010 14:49:23 +0000 (15:49 +0100)]
st/xorg: fix up xorg state tracker to use cso changes

use cso_set_fragment_sampler_views instead of cso_set_sampler_textures

14 years agoradeon: Fix buffer object unmap to be called only once for dma buffers.
Pauli Nieminen [Mon, 15 Mar 2010 08:30:18 +0000 (10:30 +0200)]
radeon: Fix buffer object unmap to be called only once for dma buffers.

If flush happens inside radeonRefillCurrentMaRegion the last dma buffer would
be unmapped twice. Unmapping buffer when moving buffer to wait list fixes the
mapping error.

14 years agoregenerate gl_mangle.h to get new EGLImage functions
Brian Paul [Tue, 16 Mar 2010 13:56:56 +0000 (07:56 -0600)]
regenerate gl_mangle.h to get new EGLImage functions

14 years agolibgl-xlib: Use a simple GALLIUM_DRIVER env var to select the pipe driver.
José Fonseca [Tue, 16 Mar 2010 13:48:09 +0000 (13:48 +0000)]
libgl-xlib: Use a simple GALLIUM_DRIVER env var to select the pipe driver.

GALLIUM_DRIVER is being used in many other places, and it easier to
memorizing and understand than all the GALLIUM_NO_XXX.