mesa.git
13 years agoswrast: Clean up header file inclusion in ss_context.h.
Vinson Lee [Fri, 13 Aug 2010 06:51:16 +0000 (23:51 -0700)]
swrast: Clean up header file inclusion in ss_context.h.

Remove mtypes.h and swrast_setup.h.
Include glheader.h for GL symbols.

13 years agomesa: Include missing header in programopt.h.
Vinson Lee [Fri, 13 Aug 2010 06:27:27 +0000 (23:27 -0700)]
mesa: Include missing header in programopt.h.

Include mtypes.h for GLcontext and gl_register_file symbols.

13 years agor300g: do not support separate depth/stencil clear in the driver
Marek Olšák [Fri, 13 Aug 2010 06:16:59 +0000 (08:16 +0200)]
r300g: do not support separate depth/stencil clear in the driver

It doesn't work well with Hyper-Z, so put the burden on the state tracker.

13 years agor300g: fix fastfill when color and Z clear are invoked separately
Marek Olšák [Fri, 13 Aug 2010 02:34:38 +0000 (04:34 +0200)]
r300g: fix fastfill when color and Z clear are invoked separately

This always restores the previous depth clear value after CBZB clear.

13 years agor600g: update shader caps
Marek Olšák [Fri, 13 Aug 2010 01:27:45 +0000 (03:27 +0200)]
r600g: update shader caps

Sent on ML by Владимир.

These values are what fglrx returns.

13 years agor300g: disable depth clamp for now
Marek Olšák [Fri, 13 Aug 2010 00:58:51 +0000 (02:58 +0200)]
r300g: disable depth clamp for now

It breaks Regnum Online in that it renders random triangles
all over the screen.

https://bugs.freedesktop.org/show_bug.cgi?id=29518

13 years agomesa: fpclassify is available with MinGW.
Vinson Lee [Fri, 13 Aug 2010 00:37:00 +0000 (17:37 -0700)]
mesa: fpclassify is available with MinGW.

This patch fixes the MinGW build.

13 years agor600g: fix warning in the winsys
Dave Airlie [Thu, 12 Aug 2010 23:12:49 +0000 (09:12 +1000)]
r600g: fix warning in the winsys

13 years agor600g: fix memory leaks running gears.
Dave Airlie [Fri, 13 Aug 2010 00:16:29 +0000 (10:16 +1000)]
r600g: fix memory leaks running gears.

I noticed gears memory usage was heading skywards, some r600 "states"
aren't properly refcounted, and the ctx->state is never freed.

13 years agotranslate_test: Fix compilation with MSVC.
Vinson Lee [Fri, 13 Aug 2010 00:12:56 +0000 (17:12 -0700)]
translate_test: Fix compilation with MSVC.

13 years agoscons: Add main/querymatrix.c to SCons build.
Vinson Lee [Thu, 12 Aug 2010 23:32:49 +0000 (16:32 -0700)]
scons: Add main/querymatrix.c to SCons build.

Commit 87eb66775949af6e9512daf7e4665c1cfa6b8745 added querymatrix.c to
make but not to SCons.

13 years agoglsl: print to stderr like other program printing code
Brian Paul [Thu, 12 Aug 2010 21:59:50 +0000 (15:59 -0600)]
glsl: print to stderr like other program printing code

13 years agomesa: Fix FreeBSD build with llvm enabled.
Vinson Lee [Thu, 12 Aug 2010 21:54:30 +0000 (14:54 -0700)]
mesa: Fix FreeBSD build with llvm enabled.

On FreeBSD LC_CTYPE_MASK is not available but 'llvm-config --cppflags'
adds the compiler flag -D_GNU_SOURCE to the build.

13 years agotranslate_test: fix compilation on non-POSIX platforms
Luca Barbieri [Thu, 12 Aug 2010 20:07:37 +0000 (22:07 +0200)]
translate_test: fix compilation on non-POSIX platforms

Use a kludgy function based on rand() instead of drand48()

13 years agotranslate_test: improve
Luca Barbieri [Thu, 12 Aug 2010 16:18:41 +0000 (18:18 +0200)]
translate_test: improve

1. Generate random data specifically for float and doubles, so that
   they end up in [0, 1] range
2. Don't test useless conversions like SCALED <-> NORM
3. Poison the buffers before testing

13 years agoRevert "u_blitter: unify clear_depth_stencil and flush_depth_stencil"
Marek Olšák [Thu, 12 Aug 2010 11:34:53 +0000 (13:34 +0200)]
Revert "u_blitter: unify clear_depth_stencil and flush_depth_stencil"

This reverts commit de4784e36505316c2a5ab34cc5b371d17f38d3c5.

13 years agor600: add support for draw_elements_base_vertex
Andre Maasikas [Wed, 11 Aug 2010 11:30:18 +0000 (14:30 +0300)]
r600: add support for draw_elements_base_vertex

use VTX_BASE_VTX_LOC for offset, last time using INDEX_OFFSET was
probably a wrong register for this

13 years agomesa: Clean up header file inclusion in prog_uniform.h.
Vinson Lee [Thu, 12 Aug 2010 07:33:48 +0000 (00:33 -0700)]
mesa: Clean up header file inclusion in prog_uniform.h.

Remove mtypes.h.
Remove prog_statevars.h.
Include glheader.h for GL symbols.

13 years agomesa: Include missing headers in prog_print.h.
Vinson Lee [Thu, 12 Aug 2010 07:27:01 +0000 (00:27 -0700)]
mesa: Include missing headers in prog_print.h.

Include stdio.h for FILE symbol.
Include glheader.h for GL symbols.
Include mtypes.h for GLcontext symbol.
Add forward declarations.

13 years agomesa: Include missing header in prog_optimize.h.
Vinson Lee [Thu, 12 Aug 2010 07:18:06 +0000 (00:18 -0700)]
mesa: Include missing header in prog_optimize.h.

Include mtypes.h for GLcontext symbol.

13 years agor600g: fix typo in stencil translate.
Dave Airlie [Thu, 12 Aug 2010 06:32:19 +0000 (16:32 +1000)]
r600g: fix typo in stencil translate.

fixes piglit stencil-twoside and stencil-wrap

13 years agor600g: fix provoking-vertex piglit test.
Dave Airlie [Thu, 12 Aug 2010 06:29:04 +0000 (16:29 +1000)]
r600g: fix provoking-vertex piglit test.

13 years agor600g: improve texture format checker.
Dave Airlie [Thu, 12 Aug 2010 06:06:40 +0000 (16:06 +1000)]
r600g: improve texture format checker.

This takes the r300g texture format checker and fixes it up for r600g,
it passes glean texSwizzle, pixelformats, and texture_srgb tests,

however I think it L8S8_SRGB is broken as is L8_SRGB, need to investigate.

Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agou_blitter: unify clear_depth_stencil and flush_depth_stencil
Marek Olšák [Thu, 12 Aug 2010 03:06:21 +0000 (05:06 +0200)]
u_blitter: unify clear_depth_stencil and flush_depth_stencil

No need to enable depth test for clear.

13 years agoscons: Fix freebsd8 build.
Vinson Lee [Thu, 12 Aug 2010 01:57:27 +0000 (18:57 -0700)]
scons: Fix freebsd8 build.

13 years agoscons: Add freebsd8 to list of accepted platforms.
Vinson Lee [Thu, 12 Aug 2010 00:58:06 +0000 (17:58 -0700)]
scons: Add freebsd8 to list of accepted platforms.

13 years agor300/compiler: remove an unused variable
Marek Olšák [Wed, 11 Aug 2010 22:39:04 +0000 (00:39 +0200)]
r300/compiler: remove an unused variable

13 years agotranslate_test: fix segfault on x86-64
Luca Barbieri [Wed, 11 Aug 2010 20:26:57 +0000 (22:26 +0200)]
translate_test: fix segfault on x86-64

13 years agou_staging: remove useless inline keyword
Luca Barbieri [Wed, 11 Aug 2010 20:11:19 +0000 (22:11 +0200)]
u_staging: remove useless inline keyword

13 years agotranslate_test: fix, reindent, and improve
Luca Barbieri [Wed, 11 Aug 2010 19:28:33 +0000 (21:28 +0200)]
translate_test: fix, reindent, and improve

1. Fix the segfault due to the reverted commit using the new interface
2. Reindent to Mesa 3 spaces style
3. Improve output and return success/failure with error code
4. Add much better support for testing translate_sse

13 years agotranslate: allow clients to ask for supported output formats
Luca Barbieri [Wed, 11 Aug 2010 19:19:12 +0000 (21:19 +0200)]
translate: allow clients to ask for supported output formats

Currently translate asserts on unsupported output formats, making
it impossible to use for some purposes, such as testing whether it
actually works on all formats it supports.

Removing the assert was met with opposition, so this change allows
clients to ask whether an output format is supported, and they are thus
able to avoid attempting to use it.

Since this is just an addition to the API, no adverse effect is
possible, and it makes the testsuite work again.

13 years agoauxiliary: Make u_staging.c MSVC compatible.
Vinson Lee [Wed, 11 Aug 2010 19:45:02 +0000 (12:45 -0700)]
auxiliary: Make u_staging.c MSVC compatible.

Fixes MSVC build.

13 years agotranslate: Add translate_test.c to SCons.
Vinson Lee [Wed, 11 Aug 2010 19:18:25 +0000 (12:18 -0700)]
translate: Add translate_test.c to SCons.

This is a follow-up to commit 945e38c73b46afc12b0655f9fc8200f216a8f8c1,
which added translate_test.c to make.

13 years agoauxiliary: Add u_staging.c to SCons build.
Vinson Lee [Wed, 11 Aug 2010 19:04:54 +0000 (12:04 -0700)]
auxiliary: Add u_staging.c to SCons build.

This is a follow-up to commit b85c71d4e1e4ed788be834dff5b7b3c0cd0402ac
which added u_staging.c to make.

13 years agor300/compiler: Implement the CONT opcode.
Tom Stellard [Wed, 11 Aug 2010 16:25:45 +0000 (09:25 -0700)]
r300/compiler: Implement the CONT opcode.

13 years agor300/compiler: Handle loops in the register allocator.
Tom Stellard [Wed, 11 Aug 2010 16:24:07 +0000 (09:24 -0700)]
r300/compiler: Handle loops in the register allocator.

13 years agor600g: accept empty frag prog shader
Jerome Glisse [Wed, 11 Aug 2010 18:26:07 +0000 (14:26 -0400)]
r600g: accept empty frag prog shader

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agor600g: add src negation support
Jerome Glisse [Wed, 11 Aug 2010 17:49:06 +0000 (13:49 -0400)]
r600g: add src negation support

Should fix few glBitmap cases.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agogallivm: Fix and enable the extra Newton/Raphson step in lp_build_rcp().
José Fonseca [Wed, 11 Aug 2010 17:44:17 +0000 (18:44 +0100)]
gallivm: Fix and enable the extra Newton/Raphson step in lp_build_rcp().

Thanks to Michal for spotting this.

13 years agor600g: add point/sprite rendering support
Jerome Glisse [Wed, 11 Aug 2010 16:19:33 +0000 (12:19 -0400)]
r600g: add point/sprite rendering support

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agoRevert "translate_generic: return NULL instead of assert(0) if format not supported"
Luca Barbieri [Wed, 11 Aug 2010 14:59:25 +0000 (16:59 +0200)]
Revert "translate_generic: return NULL instead of assert(0) if format not supported"

This reverts commit 16b45ca7cefb3432b4133fe9d0b1dbfe3f286131.

José Fonseca asked for a revert.

Note that the testsuite will now segfault since it attempts to test
all possible formats.

13 years agotranslate_generic: fix broken A8R8G8B8_UNORM output
Luca Barbieri [Wed, 11 Aug 2010 14:00:32 +0000 (16:00 +0200)]
translate_generic: fix broken A8R8G8B8_UNORM output

translate was attempting to output A8R8G8B8_UNORM as if it were
R8G8B8A8_UNORM.

Now the tests just added pass.

13 years agotranslate: add testsuite
Luca Barbieri [Wed, 11 Aug 2010 13:57:37 +0000 (15:57 +0200)]
translate: add testsuite

Currently only checks all possible format conversions, and doesn't
attempt to test whether multiple buffers/elements or indices work.

13 years agotranslate_generic: return NULL instead of assert(0) if format not supported
Luca Barbieri [Wed, 11 Aug 2010 13:59:43 +0000 (15:59 +0200)]
translate_generic: return NULL instead of assert(0) if format not supported

This gives the caller a chance to recover (or crash anyway otherwise).

13 years agoauxiliary: fix util_framebuffer_copy
Luca Barbieri [Fri, 6 Aug 2010 09:08:45 +0000 (11:08 +0200)]
auxiliary: fix util_framebuffer_copy

util_framebuffer_copy was attempting to copy all elements of the
source framebuffer state.

However, this breaks if the user does not zero initialize the structure.
Instead, only copy the elements up to nr_cbufs, and clear elements up
to dst->nr_cbufs, if the destination was larger than the source.

13 years agogallivm: Use lp_build_div instead of lp_build_mul + lp_build_rcp.
José Fonseca [Wed, 11 Aug 2010 14:13:17 +0000 (15:13 +0100)]
gallivm: Use lp_build_div instead of lp_build_mul + lp_build_rcp.

Single divide, so let lp_build_div decide how to implement this.

This will save a multiplication in architectures which don't have
a RCP intrinsic.

13 years agogallivm: Use unsigned shift in lp_build_minify.
José Fonseca [Wed, 11 Aug 2010 14:11:12 +0000 (15:11 +0100)]
gallivm: Use unsigned shift in lp_build_minify.

Texture dimensions are unsigned.

13 years agollvmpipe: Debug code to dump interpolation coefficients.
José Fonseca [Wed, 11 Aug 2010 14:10:28 +0000 (15:10 +0100)]
llvmpipe: Debug code to dump interpolation coefficients.

13 years agollvmpipe: Use single precision divide for one over area computation.
José Fonseca [Wed, 11 Aug 2010 14:09:54 +0000 (15:09 +0100)]
llvmpipe: Use single precision divide for one over area computation.

13 years agoutil: copy the u_staging commit message to the code
Marek Olšák [Wed, 11 Aug 2010 13:11:37 +0000 (15:11 +0200)]
util: copy the u_staging commit message to the code

13 years agoauxiliary: support for transfers using staging resources
Luca Barbieri [Tue, 3 Aug 2010 19:20:53 +0000 (21:20 +0200)]
auxiliary: support for transfers using staging resources

Direct3D 10/11 has no concept of transfers. Applications instead
create resources with a STAGING or DYNAMIC usage, copy between them
and the real resource and use Map to map the STAGING/DYNAMIC resource.

This util module allows to implement Gallium drivers as a Direct3D
driver would be implemented: transfers allocate a resource with
PIPE_USAGE_STAGING, and copy the data between it and the real resource
with resource_copy_region.

13 years agogallium: add ALWAYS_INLINE
Luca Barbieri [Tue, 10 Aug 2010 08:26:14 +0000 (10:26 +0200)]
gallium: add ALWAYS_INLINE

Used when we want to be sure the compiler inlines a large function into
an inner loop.

13 years agou_surfaces: add util_surfaces_peek
Luca Barbieri [Tue, 3 Aug 2010 20:19:30 +0000 (22:19 +0200)]
u_surfaces: add util_surfaces_peek

Used to find out if a surface exists without creating one.

13 years agou_surfaces: use cso_hash instead of util_hash_table
Luca Barbieri [Fri, 6 Aug 2010 07:46:14 +0000 (09:46 +0200)]
u_surfaces: use cso_hash instead of util_hash_table

Using cso_hash directly is the right thing since util_hash_table
adds useless overhead and is harder to use for this application.

13 years agou_surfaces: fix surface leak due to off by one
Luca Barbieri [Fri, 6 Aug 2010 07:53:15 +0000 (09:53 +0200)]
u_surfaces: fix surface leak due to off by one

13 years agoauxiliary: make primitive splitter assert on unimplemented adjacency prims
Luca Barbieri [Wed, 11 Aug 2010 08:54:31 +0000 (10:54 +0200)]
auxiliary: make primitive splitter assert on unimplemented adjacency prims

They are unimplemented, even though the framework makes it possible to
implement them well, and nv50 needs them.

13 years agoauxiliary: fix u_split_prim naming convention
Luca Barbieri [Wed, 11 Aug 2010 08:51:28 +0000 (10:51 +0200)]
auxiliary: fix u_split_prim naming convention

Current practice is to start identifiers with "util_" instead of "u_".

13 years agoauxiliary: move Ben Skeggs' primitive splitter to common code
Luca Barbieri [Wed, 11 Aug 2010 08:46:12 +0000 (10:46 +0200)]
auxiliary: move Ben Skeggs' primitive splitter to common code

This is a simple framework that handles splitting primitives in an
abstract way.

The user has to specify the primitive start, start index and count.

Then, it can ask the primitive splitter to "draw" a chunk of the
primitive, staying under a given vertex/index budget.

The primitive splitter will then call user-supplied functions to
emit a range of vertices/indices, as well as switch the edgeflag
on or off.

This is particularly useful for hardware that either has limits
on the vertex count field, or where vertices are pushed on a FIFO
or temporary buffer of limited size.

Note that unlike other splitters, it does not manipulate data in
any way, and merely asks a callback to do so, in vertex intervals.

13 years agomesa: Include missing header in prog_noise.h.
Vinson Lee [Wed, 11 Aug 2010 08:10:54 +0000 (01:10 -0700)]
mesa: Include missing header in prog_noise.h.

Include glheader.h for GLfloat symbol.

13 years agomesa: Clean up header file inclusion in prog_instruction.h.
Vinson Lee [Wed, 11 Aug 2010 08:05:49 +0000 (01:05 -0700)]
mesa: Clean up header file inclusion in prog_instruction.h.

Remove mfeatures.h.
Include glheader.h for GL symbols.

13 years agomesa: Include missing header in prog_execute.h.
Vinson Lee [Wed, 11 Aug 2010 07:43:28 +0000 (00:43 -0700)]
mesa: Include missing header in prog_execute.h.

Include mtypes.h for GLcontext symbol.

13 years agomesa: Include missing header in prog_cache.h.
Vinson Lee [Wed, 11 Aug 2010 07:35:34 +0000 (00:35 -0700)]
mesa: Include missing header in prog_cache.h.

Include mtypes.h for GLcontext symbol.

13 years agomesa: Include missing header in nvvertparse.h.
Vinson Lee [Wed, 11 Aug 2010 07:33:02 +0000 (00:33 -0700)]
mesa: Include missing header in nvvertparse.h.

Include mtypes.h for GLcontext symbol.

13 years agomesa: Include missing header in nvfragparse.h.
Vinson Lee [Wed, 11 Aug 2010 07:29:26 +0000 (00:29 -0700)]
mesa: Include missing header in nvfragparse.h.

Include mtypes.h for GLcontext symbol.

13 years agomesa: Remove unnecessary header from hash_table.h.
Vinson Lee [Wed, 11 Aug 2010 07:26:05 +0000 (00:26 -0700)]
mesa: Remove unnecessary header from hash_table.h.

13 years agoswrast: fix span color array pointer assignment for 32-bit/channel rendering
M.Froehlich@science-computing.de [Wed, 11 Aug 2010 03:19:46 +0000 (21:19 -0600)]
swrast: fix span color array pointer assignment for 32-bit/channel rendering

See fd.o bug 29487.

NOTE: This is a candidate for the 7.8 branch.

Signed-off-by: Brian Paul <brianp@vmware.com>
13 years agor300g: initialize VAP_VTX_STATE_CNTL
Marek Olšák [Wed, 11 Aug 2010 00:58:50 +0000 (02:58 +0200)]
r300g: initialize VAP_VTX_STATE_CNTL

This got lost during the rasterizer rewrite.

13 years agor300g: implement gl_FrontFacing
Marek Olšák [Tue, 10 Aug 2010 12:21:05 +0000 (14:21 +0200)]
r300g: implement gl_FrontFacing

13 years agoswrast: Remove unnecessary header.
Vinson Lee [Wed, 11 Aug 2010 01:05:52 +0000 (18:05 -0700)]
swrast: Remove unnecessary header.

13 years agosvga: Remove unnecessary headers.
Vinson Lee [Wed, 11 Aug 2010 00:53:08 +0000 (17:53 -0700)]
svga: Remove unnecessary headers.

13 years agor300g: Remove unnecessary header.
Vinson Lee [Wed, 11 Aug 2010 00:39:13 +0000 (17:39 -0700)]
r300g: Remove unnecessary header.

13 years agoglut: also check for GLX_ARB_multisample
Nigel Stewart [Tue, 10 Aug 2010 23:27:37 +0000 (17:27 -0600)]
glut: also check for GLX_ARB_multisample

Signed-off-by: Brian Paul <brianp@vmware.com>
13 years agomesa: use switch stmt in init_program_limits()
Brian Paul [Tue, 10 Aug 2010 15:14:22 +0000 (09:14 -0600)]
mesa: use switch stmt in init_program_limits()

13 years agomesa: remove obsolete comments
Brian Paul [Tue, 10 Aug 2010 15:10:30 +0000 (09:10 -0600)]
mesa: remove obsolete comments

13 years agor300/compiler: Use predicate bit for IF statements in r500 vertex shaders
Tom Stellard [Sun, 8 Aug 2010 19:53:29 +0000 (12:53 -0700)]
r300/compiler: Use predicate bit for IF statements in r500 vertex shaders

13 years agor300/compiler: Implement hardware assisted loops for vertex shaders.
Tom Stellard [Thu, 5 Aug 2010 17:19:00 +0000 (10:19 -0700)]
r300/compiler: Implement hardware assisted loops for vertex shaders.

Single loops work, but nested loops do not.

13 years agor300/compiler: Correctly transform nested loops.
Tom Stellard [Fri, 6 Aug 2010 04:03:55 +0000 (21:03 -0700)]
r300/compiler: Correctly transform nested loops.

13 years agor600g: avoid reemiting literal, avoid scheduling empty cs
Jerome Glisse [Tue, 10 Aug 2010 15:52:00 +0000 (11:52 -0400)]
r600g: avoid reemiting literal, avoid scheduling empty cs

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agoutil: Add util_format_srgb().
José Fonseca [Tue, 10 Aug 2010 15:32:29 +0000 (16:32 +0100)]
util: Add util_format_srgb().

To convert RGB -> SRGB format.

13 years agogallivm: fix non-SSE4.1 case in lp_build_pack2()
Brian Paul [Tue, 10 Aug 2010 14:56:20 +0000 (08:56 -0600)]
gallivm: fix non-SSE4.1 case in lp_build_pack2()

Since there's no SSE instruction for this case, fall through to the
generic shuffle code.

Fixes bug fd.o 29468.

13 years agomesa: additional program limit assertions
Brian Paul [Tue, 10 Aug 2010 14:38:12 +0000 (08:38 -0600)]
mesa: additional program limit assertions

13 years agomesa: fix comment typo
Brian Paul [Tue, 10 Aug 2010 14:35:47 +0000 (08:35 -0600)]
mesa: fix comment typo

13 years agogallivm: Fix bitwise operations for floats, division for integers
nobled [Mon, 9 Aug 2010 21:25:18 +0000 (21:25 +0000)]
gallivm: Fix bitwise operations for floats, division for integers

http://bugs.freedesktop.org/29407

Signed-off-by: José Fonseca <jfonseca@vmware.com>
13 years agogallivm: Even more type checking
nobled [Mon, 9 Aug 2010 21:15:08 +0000 (21:15 +0000)]
gallivm: Even more type checking

http://bugs.freedesktop.org/29407

Signed-off-by: José Fonseca <jfonseca@vmware.com>
13 years agollvmpipe: Always use floating-point operators for floating-point types
nobled [Fri, 6 Aug 2010 17:32:29 +0000 (17:32 +0000)]
llvmpipe: Always use floating-point operators for floating-point types

See:
http://bugs.freedesktop.org/29404
http://bugs.freedesktop.org/29407

Signed-off-by: José Fonseca <jfonseca@vmware.com>
13 years agoradeon: Use MESA_FORMAT_SARGB8 for sRGB formats
Henri Verbeet [Mon, 9 Aug 2010 19:06:02 +0000 (15:06 -0400)]
radeon: Use MESA_FORMAT_SARGB8 for sRGB formats

This can be supported on r600 without using the endian swapper, and is a
better fit for (typical) uploads using GL_BGRA, GL_UNSIGNED_INT_8_8_8_8_REV
anyway.

13 years agor600c: Disable alpha test during blits
Henri Verbeet [Mon, 9 Aug 2010 19:03:27 +0000 (15:03 -0400)]
r600c: Disable alpha test during blits

13 years agogallivm: More type checks.
José Fonseca [Mon, 9 Aug 2010 16:30:33 +0000 (17:30 +0100)]
gallivm: More type checks.

13 years agogallivm: Don't call LLVMBuildFNeg on llvm-2.6.
José Fonseca [Mon, 9 Aug 2010 16:26:18 +0000 (17:26 +0100)]
gallivm: Don't call LLVMBuildFNeg on llvm-2.6.

It didn't exist yet.

13 years agogallivm: Always use floating-point operators for floating-point types
nobled [Sun, 8 Aug 2010 19:44:54 +0000 (19:44 +0000)]
gallivm: Always use floating-point operators for floating-point types

This fixes the assert added in LLVM 2.8:
assert(getType()->isIntOrIntVectorTy() &&
       "Tried to create an integer operation on a non-integer type!")

But it also fixes some subtle bugs, since we should've been doing this
since LLVM 2.6 anyway.

Includes a modified patch from steckdenis@yahoo.fr for the
FNeg instructions in emit_fetch(); thanks for pointing those out.

http://bugs.freedesktop.org/29404
http://bugs.freedesktop.org/29407

Signed-off-by: José Fonseca <jfonseca@vmware.com>
13 years agoutil: Move _mm_shuffle_epi8() to u_sse.h.
José Fonseca [Mon, 9 Aug 2010 15:41:44 +0000 (16:41 +0100)]
util: Move _mm_shuffle_epi8() to u_sse.h.

It's bound to be useful elsewhere.

13 years agor600g: fix r600 context structure, avoid segfault when no scissor
Jerome Glisse [Mon, 9 Aug 2010 15:32:45 +0000 (11:32 -0400)]
r600g: fix r600 context structure, avoid segfault when no scissor

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agor300g: do not print shader compiler errors by default
Marek Olšák [Mon, 9 Aug 2010 15:15:37 +0000 (17:15 +0200)]
r300g: do not print shader compiler errors by default

13 years agor600g: fix some warnings
Marek Olšák [Mon, 9 Aug 2010 12:57:56 +0000 (14:57 +0200)]
r600g: fix some warnings

13 years agor600g: fill out some missing caps and sort them
Marek Olšák [Mon, 9 Aug 2010 12:53:58 +0000 (14:53 +0200)]
r600g: fill out some missing caps and sort them

The shader caps need additional corrections.

(based on a patch from netkas at Phoronix)

13 years agoradeon: fix npot mipmap alignment on r600
Andre Maasikas [Mon, 9 Aug 2010 11:37:01 +0000 (14:37 +0300)]
radeon: fix npot mipmap alignment on r600

seems it got lost in commit 0d383547
have no earlier radeons to test, but npot mipmaps were not supported there?

13 years agor600: bump glsl version
Andre Maasikas [Mon, 9 Aug 2010 11:28:00 +0000 (14:28 +0300)]
r600: bump glsl version

from the tests i couldn't find any new driver faults

13 years agor300g: fix hiz/zmask offset emissions.
Dave Airlie [Mon, 9 Aug 2010 09:56:45 +0000 (19:56 +1000)]
r300g: fix hiz/zmask offset emissions.

ofs is in dwords, so need to shift it for registers.

Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agonouveau: fix maps with PIPE_TRANSFER_UNSYNCHRONIZED | PIPE_TRANSFER_DONTBLOCK
Luca Barbieri [Mon, 9 Aug 2010 02:56:03 +0000 (04:56 +0200)]
nouveau: fix maps with PIPE_TRANSFER_UNSYNCHRONIZED | PIPE_TRANSFER_DONTBLOCK

In this case, we were incorrectly prioritizing PIPE_TRANSFER_DONTBLOCK over
PIPE_TRANSFER_UNSYNCHRONIZED.

This can lead to failure in the Mesa VBO draw paths that end up specifying
both, but don't expect map to fail (in particular, the problem manifested as
a leak of buffer objects in teapot with other changes).

13 years agor300g: remove a flush
Marek Olšák [Sun, 8 Aug 2010 21:09:46 +0000 (23:09 +0200)]
r300g: remove a flush

Ooops, it wasn't supposed to be there.