mesa.git
13 years agomesa: implement new texture format ARGB2101010
Marek Olšák [Tue, 21 Dec 2010 17:54:50 +0000 (18:54 +0100)]
mesa: implement new texture format ARGB2101010

Radeon GPUs do support GL_RGB10_A2.

13 years agost/mesa: if Z32 is unsupported, prefer Z24 to Z16
Marek Olšák [Tue, 21 Dec 2010 04:50:58 +0000 (05:50 +0100)]
st/mesa: if Z32 is unsupported, prefer Z24 to Z16

13 years agost/mesa: use RGBA16 for RGB12 and RGB16
Marek Olšák [Tue, 21 Dec 2010 04:45:46 +0000 (05:45 +0100)]
st/mesa: use RGBA16 for RGB12 and RGB16

To provide enough precision if a user wants it.

13 years agost/mesa: use DXT SRGB formats for COMPRESSED_SRGB
Marek Olšák [Tue, 21 Dec 2010 05:41:59 +0000 (06:41 +0100)]
st/mesa: use DXT SRGB formats for COMPRESSED_SRGB

And also check if the formats are supported to return something meaningful
if compression cannot be used.

13 years agoi965: Keep around a copy of the VS constant surface dumping code.
Eric Anholt [Thu, 23 Dec 2010 06:48:24 +0000 (22:48 -0800)]
i965: Keep around a copy of the VS constant surface dumping code.

Just like everywhere else, I never trust my constant uploads to
correctly put constants in the right places, even though that's so
rarely where the issue is.

13 years agoi965: Correct the dp_read message descriptor setup on g4x.
Eric Anholt [Thu, 23 Dec 2010 06:46:12 +0000 (22:46 -0800)]
i965: Correct the dp_read message descriptor setup on g4x.

It's mostly like gen4 message descriptor setup, except that the sizes
of type/control changed to be like gen5.  Fixes 21 piglit cases on
gm45, including the regressions in bug #32311 from increased VS
constant buffer usage.

13 years agoi965: upload multisample state for fragment program change
Zhenyu Wang [Thu, 23 Dec 2010 08:30:20 +0000 (16:30 +0800)]
i965: upload multisample state for fragment program change

This makes conformance tests stable on sandybridge D0 to track
multisample state before SF/WM state.

13 years agoi965: Use MI_FLUSH_DW for blt ring flush on sandybridge
Zhenyu Wang [Thu, 23 Dec 2010 07:06:33 +0000 (15:06 +0800)]
i965: Use MI_FLUSH_DW for blt ring flush on sandybridge

Old MI_FLUSH command is deprecated on sandybridge blt.

13 years agost/mesa: Remove unnecessary header.
Vinson Lee [Thu, 23 Dec 2010 09:03:32 +0000 (01:03 -0800)]
st/mesa: Remove unnecessary header.

13 years agogallivm: Disable MMX-disabling code on llvm-2.9.
Vinson Lee [Thu, 23 Dec 2010 03:44:00 +0000 (19:44 -0800)]
gallivm: Disable MMX-disabling code on llvm-2.9.

The disable-mmx option was removed in llvm-2.9svn by revisions 122188
and 122189.

Fixes FDO bug 32564.

13 years agogallivm: Fix 'cast from pointer to integer of different size' warning.
Vinson Lee [Thu, 23 Dec 2010 00:48:19 +0000 (16:48 -0800)]
gallivm: Fix 'cast from pointer to integer of different size' warning.

Fixes this GCC warning.
lp_bld_const.h: In function 'lp_build_const_int_pointer':
lp_bld_const.h:137: warning: cast from pointer to integer of different size

13 years agoi915g: Remove unnecessary header.
Vinson Lee [Wed, 22 Dec 2010 08:57:52 +0000 (00:57 -0800)]
i915g: Remove unnecessary header.

13 years agollvmpipe: Remove unnecessary headers.
Vinson Lee [Wed, 22 Dec 2010 08:55:41 +0000 (00:55 -0800)]
llvmpipe: Remove unnecessary headers.

13 years agor300g: Remove unnecessary headers.
Vinson Lee [Wed, 22 Dec 2010 08:52:05 +0000 (00:52 -0800)]
r300g: Remove unnecessary headers.

13 years agosvga: Remove unnecessary header.
Vinson Lee [Wed, 22 Dec 2010 08:42:23 +0000 (00:42 -0800)]
svga: Remove unnecessary header.

13 years agost/vega: Remove unnecessary headers.
Vinson Lee [Wed, 22 Dec 2010 08:38:42 +0000 (00:38 -0800)]
st/vega: Remove unnecessary headers.

13 years agor600g: Remove the unused "pframebuffer" field from r600_pipe_context.
Henri Verbeet [Wed, 22 Dec 2010 08:18:05 +0000 (09:18 +0100)]
r600g: Remove the unused "pframebuffer" field from r600_pipe_context.

13 years agor600g: r600_new() and r600_delete() are unused.
Henri Verbeet [Wed, 22 Dec 2010 08:18:05 +0000 (09:18 +0100)]
r600g: r600_new() and r600_delete() are unused.

13 years agoi965: explicit tell header present for fb write on sandybridge
Zhenyu Wang [Wed, 22 Dec 2010 15:52:12 +0000 (10:52 -0500)]
i965: explicit tell header present for fb write on sandybridge

Determine header present for fb write by msg length is not right
for SIMD16 dispatch, and if there're more output attributes, header
present is not easy to tell from msg length. This explicitly adds
new param for fb write to say header present or not.

Fixes many cases' hang and failure in GL conformance test.

13 years agost/egl: Assorted fixes for dri2_display_get_configs.
Chia-I Wu [Wed, 22 Dec 2010 07:13:52 +0000 (15:13 +0800)]
st/egl: Assorted fixes for dri2_display_get_configs.

Set window_bit only when the visual id is greater than zero.  Correct
visual types.  Skip slow configs as they are not relevant.  Finally, do
not return duplicated configs.

13 years agor600g: remove useless switch statements
Alex Deucher [Wed, 22 Dec 2010 06:30:41 +0000 (01:30 -0500)]
r600g: remove useless switch statements

Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
13 years agost/egl: Fix eglCopyBuffers.
Chia-I Wu [Wed, 22 Dec 2010 06:21:21 +0000 (14:21 +0800)]
st/egl: Fix eglCopyBuffers.

Flush before presenting.

13 years agost/egl: Plug pbuffer leaks.
Chia-I Wu [Wed, 22 Dec 2010 05:59:58 +0000 (13:59 +0800)]
st/egl: Plug pbuffer leaks.

Unreference validated resources or remove unnecessary validations.

13 years agost/egl: Allow single-buffered pixmaps.
Chia-I Wu [Wed, 22 Dec 2010 05:47:50 +0000 (13:47 +0800)]
st/egl: Allow single-buffered pixmaps.

All single-buffered configs were ignored before to make sure
EGL_RENDER_BUFFER is settable for window surfaces.  It is better to
allow single-buffered configs and set EGL_WINDOW_BIT only for
double-buffered ones.  This way there can be single-buffered pixmaps.

13 years agor600g: drop unused code in evergreen.
Dave Airlie [Wed, 22 Dec 2010 05:58:29 +0000 (15:58 +1000)]
r600g: drop unused code in evergreen.

this code was pretty much duplicated, thanks to Henri Verbeet on irc for
pointing it out.

Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agost/egl: Remove native_config::samples.
Chia-I Wu [Wed, 22 Dec 2010 04:27:35 +0000 (12:27 +0800)]
st/egl: Remove native_config::samples.

Multisample buffers are never requested.

13 years agost/egl: Remove native_config::slow_config.
Chia-I Wu [Wed, 22 Dec 2010 04:22:42 +0000 (12:22 +0800)]
st/egl: Remove native_config::slow_config.

In direct rendering scenario, it is not needed until an EGLDisplay can
support both HW and SW pipe screens.

13 years agost/egl: Remove unnecessary egl_g3d_find_pixmap_config.
Chia-I Wu [Wed, 22 Dec 2010 04:19:12 +0000 (12:19 +0800)]
st/egl: Remove unnecessary egl_g3d_find_pixmap_config.

It was used to find a compatible config for a given pixmap.  Now that a
config is optional for pixmap surface creation, the function is not
needed.

13 years agost/egl: Make config optional for create_pixmap_surface.
Chia-I Wu [Wed, 22 Dec 2010 04:14:44 +0000 (12:14 +0800)]
st/egl: Make config optional for create_pixmap_surface.

eglCopyBuffers or EGL_KHR_image_pixmap require creating a pixmap surface
without a config.  Make it just work without relying on
is_pixmap_supported.

13 years agor600g: fix evergreen segfaults.
Dave Airlie [Wed, 22 Dec 2010 04:54:17 +0000 (14:54 +1000)]
r600g: fix evergreen segfaults.

evergreen was crashing running even gears here.

This is a 7.10 candidate if its broken the same.

Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agor300g: fix precision issues with B10G10R10A2
Marek Olšák [Wed, 22 Dec 2010 01:22:09 +0000 (02:22 +0100)]
r300g: fix precision issues with B10G10R10A2

13 years agor300g: support B10G10R10A2 render targets only with DRM 2.8.0 or later versions
Marek Olšák [Tue, 21 Dec 2010 18:26:27 +0000 (19:26 +0100)]
r300g: support B10G10R10A2 render targets only with DRM 2.8.0 or later versions

13 years agoi965: Avoid using float type for raw moves, to work around SNB issue.
Eric Anholt [Tue, 21 Dec 2010 21:04:07 +0000 (13:04 -0800)]
i965: Avoid using float type for raw moves, to work around SNB issue.

The SNB alt-mode math does the denorm and inf reduction even for a
"raw MOV" like we do for g0 message header setup, where we are moving
values that aren't actually floats.  Just use UD type, where raw MOVs
really are raw MOVs.

Fixes glxgears since c52adfc2e1d130effea940e75690897eb5d3ceaa, but no
piglit tests had regressed(!)

13 years agor600g: avoid segfault
Jerome Glisse [Tue, 21 Dec 2010 15:49:53 +0000 (10:49 -0500)]
r600g: avoid segfault

Candidates 7.10

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agointel: Check for unsupported texture when finishing using as a render target
Chris Wilson [Tue, 21 Dec 2010 11:55:11 +0000 (11:55 +0000)]
intel: Check for unsupported texture when finishing using as a render target

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=32541
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
13 years agost/mesa: Clean up header file inclusion in st_format.h.
Vinson Lee [Tue, 21 Dec 2010 09:25:04 +0000 (01:25 -0800)]
st/mesa: Clean up header file inclusion in st_format.h.

13 years agost/mesa: Clean up header file inclusion in st_draw.h.
Vinson Lee [Tue, 21 Dec 2010 09:17:37 +0000 (01:17 -0800)]
st/mesa: Clean up header file inclusion in st_draw.h.

13 years agonouveau: fix includes for latest libdrm
Ben Skeggs [Mon, 20 Dec 2010 03:39:36 +0000 (13:39 +1000)]
nouveau: fix includes for latest libdrm

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
13 years agoRegenerate gl_mangle.h.
Tom Fogal [Tue, 21 Dec 2010 02:24:29 +0000 (19:24 -0700)]
Regenerate gl_mangle.h.

NOTE: This is a candidate for the 7.10 branch.

13 years agor600g: properly unset vertex buffer
Jerome Glisse [Mon, 20 Dec 2010 20:30:42 +0000 (15:30 -0500)]
r600g: properly unset vertex buffer

Fix bug http://bugs.freedesktop.org/show_bug.cgi?id=32455

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agost/python: remove unused 'buf' parameter in pipe_buffer_unmap
Vinson Lee [Mon, 20 Dec 2010 19:40:54 +0000 (11:40 -0800)]
st/python: remove unused 'buf' parameter in pipe_buffer_unmap

This is a follow-up to commit ec51092a72e2dff1e9b1362d813fe4691cda89b7.

Fixes SCons build.

13 years agogallium: remove unused 'buf' parameter in pipe_buffer_unmap
Marek Olšák [Sun, 19 Dec 2010 03:17:43 +0000 (04:17 +0100)]
gallium: remove unused 'buf' parameter in pipe_buffer_unmap

13 years agost/mesa: Remove comment cruft from st_context.h.
Vinson Lee [Mon, 20 Dec 2010 09:24:26 +0000 (01:24 -0800)]
st/mesa: Remove comment cruft from st_context.h.

This was unintentionally added by commit
1525fb4afec0adc164948b4060ec1c0359441cd3.

13 years agost/mesa: Clean up header file inclusion in st_cb_texture.h.
Vinson Lee [Mon, 20 Dec 2010 09:15:04 +0000 (01:15 -0800)]
st/mesa: Clean up header file inclusion in st_cb_texture.h.

13 years agost/mesa: Clean up header file inclusion in st_cb_readpixels.h.
Vinson Lee [Mon, 20 Dec 2010 09:00:26 +0000 (01:00 -0800)]
st/mesa: Clean up header file inclusion in st_cb_readpixels.h.

13 years agor300g: optimize the fallback for misaligned ushort indices
Marek Olšák [Sun, 19 Dec 2010 01:54:43 +0000 (02:54 +0100)]
r300g: optimize the fallback for misaligned ushort indices

13 years agost/mesa: Clean up header file inclusion in st_cb_program.h.
Vinson Lee [Sat, 18 Dec 2010 09:44:52 +0000 (01:44 -0800)]
st/mesa: Clean up header file inclusion in st_cb_program.h.

13 years agost/mesa: Clean up header file inclusion in st_cb_accum.h.
Vinson Lee [Sat, 18 Dec 2010 09:28:18 +0000 (01:28 -0800)]
st/mesa: Clean up header file inclusion in st_cb_accum.h.

13 years agomesa: Clean up header file inclusion in prog_statevars.h.
Vinson Lee [Sat, 18 Dec 2010 09:16:53 +0000 (01:16 -0800)]
mesa: Clean up header file inclusion in prog_statevars.h.

13 years agomesa: fix queryobj whitespace.
Dave Airlie [Sat, 18 Dec 2010 07:48:21 +0000 (17:48 +1000)]
mesa: fix queryobj whitespace.

Had done this before pushing but forgot to amend, doh.

13 years agomesa/swrast/st: add ARB_occlusion_query2 support.
Dave Airlie [Sat, 11 Sep 2010 20:31:30 +0000 (06:31 +1000)]
mesa/swrast/st: add ARB_occlusion_query2 support.

This gets my vote for most pointless extension of all time, I'm guessing
some driver could possibly optimise for this instead of counting it might
just get a true/false, but I'm not really sure.

need this to eventually advertise 3.3 despite its total uselessness.

Signed-off-by: Dave Airlie <airlied@redhat.com>
13 years agomapi: Clean up sources.mk.
Chia-I Wu [Fri, 10 Dec 2010 19:27:54 +0000 (03:27 +0800)]
mapi: Clean up sources.mk.

Rename MAPI_GLAPI_SOURCES to MAPI_UTIL_SOURCES.  Rename macro
MAPI_GLAPI_CURRENT to MAPI_MODE_UTIL.  Update the comments to make it
clear that mapi may be used in two ways and how.

13 years agomapi: Clean up u_current interface.
Chia-I Wu [Fri, 10 Dec 2010 09:14:04 +0000 (17:14 +0800)]
mapi: Clean up u_current interface.

Try not to use macros to make u_current.h appear to be glapi.h.  Use
u_current.h to implement glapi.h instead whenever possible.

13 years agomapi: Add ABI-tag note.
Chia-I Wu [Thu, 16 Dec 2010 19:11:53 +0000 (03:11 +0800)]
mapi: Add ABI-tag note.

TLS requires kernel >= 2.4.20.  Per glapi.

13 years agoRefresh autogenerated file builtin_function.cpp.
Kenneth Graunke [Sat, 18 Dec 2010 03:28:47 +0000 (19:28 -0800)]
Refresh autogenerated file builtin_function.cpp.

NOTE: The 7.9 and 7.10 branches will need their builtins refreshed too.
Rather than cherry-picking this commit, run 'make builtins'.

13 years agoglsl/builtins: Compute the correct value for smoothstep(vec, vec, vec).
Kenneth Graunke [Sat, 18 Dec 2010 03:25:37 +0000 (19:25 -0800)]
glsl/builtins: Compute the correct value for smoothstep(vec, vec, vec).

These mistakenly computed 't' instead of t * t * (3.0 - 2.0 * t).

Also, properly vectorize the smoothstep(float, float, vec) variants.

NOTE: This is a candidate for the 7.9 and 7.10 branches.

13 years agogallivm: Cleanup util_format_xxx_fetch_xxx call generation.
José Fonseca [Fri, 17 Dec 2010 20:08:35 +0000 (20:08 +0000)]
gallivm: Cleanup util_format_xxx_fetch_xxx call generation.

No need to register function prototypes in the module now that we call
the C function pointer directly -- less LLVM objects lying around.

Limited testing with lp_test_format.

13 years agoglsl: Expose a public glsl_type::void_type const pointer.
Kenneth Graunke [Mon, 13 Dec 2010 19:19:38 +0000 (11:19 -0800)]
glsl: Expose a public glsl_type::void_type const pointer.

This is analogous to glsl_type::int_type and all the others.

13 years agor300g: finally fix the texture corruption on r3xx-r4xx
Marek Olšák [Fri, 17 Dec 2010 07:59:32 +0000 (08:59 +0100)]
r300g: finally fix the texture corruption on r3xx-r4xx

Even though a bound texture stays bound when calling set_fragment_sampler_views,
it must be assigned a new cache region depending on the occupancy of other
texture units.

This fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=28800

Thanks to Álmos <aaalmosss@gmail.com> for finding the bug in the code.

NOTE: This is a candidate for both the 7.9 and 7.10 branches.

13 years agoRemove OES_compressed_paletted_texture from the ES2 extension list.
Kenneth Graunke [Fri, 17 Dec 2010 01:40:00 +0000 (17:40 -0800)]
Remove OES_compressed_paletted_texture from the ES2 extension list.

We don't support it.

13 years agosoftpipe: remove sp_tex_tile_cache_border_color()
Brian Paul [Fri, 17 Dec 2010 01:17:20 +0000 (18:17 -0700)]
softpipe: remove sp_tex_tile_cache_border_color()

With swizzling done at the end of texture sampling, we can greatly
simplify swizzling of the border color.

Fixes http://bugs.freedesktop.org/show_bug.cgi?id=32460

13 years agosoftpipe: fix depth texture sampling regression
Brian Paul [Fri, 17 Dec 2010 00:38:16 +0000 (17:38 -0700)]
softpipe: fix depth texture sampling regression

We need to keep using the pipe_get_tile_swizzle() even though there's
no swizzling because we need to explicitly pass in the surface format.

Fixes http://bugs.freedesktop.org/show_bug.cgi?id=32459

13 years agogallivm: fix copy&paste error from previous commit
Brian Paul [Thu, 16 Dec 2010 21:29:15 +0000 (14:29 -0700)]
gallivm: fix copy&paste error from previous commit

Fixes piglit regression, http://bugs.freedesktop.org/show_bug.cgi?id=32452

NOTE: This is a candidate for the 7.10 branch

13 years agor600c : inline vertex format is not updated in an app, switch to use vfetch constants...
richard [Thu, 16 Dec 2010 20:52:55 +0000 (15:52 -0500)]
r600c : inline vertex format is not updated in an app, switch to use vfetch constants. For the 7.9 and 7.10 branches as well.

13 years agointel: Support glCopyTexImage() from XRGB8888 to ARGB8888.
Eric Anholt [Wed, 15 Dec 2010 20:10:03 +0000 (12:10 -0800)]
intel: Support glCopyTexImage() from XRGB8888 to ARGB8888.

The only mismatch between the two is that we have to clear the
destination's alpha to 1.0.  Fixes WOW performance on my Ironlake,
from a few frames a second to almost playable.

13 years agointel: Try to sanely check that formats match for CopyTexImage.
Eric Anholt [Wed, 15 Dec 2010 19:54:38 +0000 (11:54 -0800)]
intel: Try to sanely check that formats match for CopyTexImage.

Before, we were going off of a couple of known (hopeful) matches
between internalFormats and the cpp of the read buffer.  Instead, we
can now just look at the gl_format of the two to see if they match.
We should avoid bad blits that might have been possible before, but
also allow different internalFormats to work without having to
enumerate each one.

13 years agointel: Drop commented intel_flush from copy_teximage.
Eric Anholt [Wed, 15 Dec 2010 19:46:37 +0000 (11:46 -0800)]
intel: Drop commented intel_flush from copy_teximage.

The blit that follows appears in the command stream so it's serialized
with previous rendering.  Any queued vertices in the tnl layer were
already flushed up in mesa/main/.

13 years agointel: Update renderbuffers before looking up CopyTexImage's read buffer.
Eric Anholt [Wed, 15 Dec 2010 19:37:45 +0000 (11:37 -0800)]
intel: Update renderbuffers before looking up CopyTexImage's read buffer.

Not fixing a particular bug, just noticed by code inspection.

13 years agogallivm: work around LLVM 2.6 bug when calling C functions
Brian Paul [Thu, 16 Dec 2010 17:13:02 +0000 (10:13 -0700)]
gallivm: work around LLVM 2.6 bug when calling C functions

Create a constant int pointer to the C function, then cast it to the
function's type.  This avoids using trampoline code which seem to be
inadvertantly freed by LLVM in some situations (which leads to segfaults).
The root issue and work-around were found by José.

NOTE: This is a candidate for the 7.10 branch

13 years agodraw: s/varient/variant/
Brian Paul [Thu, 16 Dec 2010 15:50:42 +0000 (08:50 -0700)]
draw: s/varient/variant/

13 years agosvga: s/varient/variant/
Brian Paul [Thu, 16 Dec 2010 15:50:30 +0000 (08:50 -0700)]
svga: s/varient/variant/

13 years agoi965g: s/varient/variant/
Brian Paul [Thu, 16 Dec 2010 15:47:37 +0000 (08:47 -0700)]
i965g: s/varient/variant/

13 years agoi915g: s/varient/variant/
Brian Paul [Thu, 16 Dec 2010 15:46:35 +0000 (08:46 -0700)]
i915g: s/varient/variant/

13 years agosoftpipe: s/varient/variant
Brian Paul [Thu, 16 Dec 2010 15:44:12 +0000 (08:44 -0700)]
softpipe: s/varient/variant

13 years agost/mesa: s/varient/variant
Brian Paul [Thu, 16 Dec 2010 15:43:21 +0000 (08:43 -0700)]
st/mesa: s/varient/variant

13 years agoi965: Set the alternative floating point mode on gen6 VS and WM.
Eric Anholt [Wed, 15 Dec 2010 23:15:02 +0000 (15:15 -0800)]
i965: Set the alternative floating point mode on gen6 VS and WM.

This matches how we did the math instructions pre-gen6, though it
applies to non-math as well.

Fixes vp1-LIT test 2 (degenerate case: 0 ^ 0 -> 1)

13 years agoi915: Fix INTEL_DEBUG=wm segmentation fault
Shuang He [Wed, 15 Dec 2010 03:16:49 +0000 (11:16 +0800)]
i915: Fix INTEL_DEBUG=wm segmentation fault

The program should be disassembled after it's uploaded

13 years agonvfx: fix fragprog word swapping on big-endian machines
Ben Skeggs [Thu, 16 Dec 2010 01:12:52 +0000 (11:12 +1000)]
nvfx: fix fragprog word swapping on big-endian machines

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
13 years agogallium: properly check for src->dst blit compatibilities
Jerome Glisse [Wed, 15 Dec 2010 20:27:16 +0000 (15:27 -0500)]
gallium: properly check for src->dst blit compatibilities

Spotted by Christoph Bumiller & Jose Fonseca

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agor600g: fix pow(0, 0) evaluating to NaN
Fredrik Höglund [Wed, 15 Dec 2010 19:00:42 +0000 (20:00 +0100)]
r600g: fix pow(0, 0) evaluating to NaN

We have to use the non-IEEE compliant version of MUL here, since
log2(0) is -inf, and 0 * -inf is NaN in IEEE arithmetic.

candidates for 7.10 branch

13 years agor600g: need to reference upload buffer as the might still live accross flush
Jerome Glisse [Wed, 15 Dec 2010 17:07:09 +0000 (12:07 -0500)]
r600g: need to reference upload buffer as the might still live accross flush

Can't get away from referencing upload buffer as after flush a vertex buffer
using the upload buffer might still be active. Likely need to simplify the
pipe_refence a bit so we don't waste so much cpu time in it.

candidates for 7.10 branch

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agost/mesa: fix incorrect prev pointer in destroy_program_variants()
Brian Paul [Tue, 14 Dec 2010 21:15:20 +0000 (14:15 -0700)]
st/mesa: fix incorrect prev pointer in destroy_program_variants()

13 years agosoftpipe: do texture swizzle during texture sampling
Brian Paul [Tue, 14 Dec 2010 20:01:00 +0000 (13:01 -0700)]
softpipe: do texture swizzle during texture sampling

Instead of when we read texture tiles.  Now swizzling happens after
the shadow depth compare step.  This fixes the piglit glsl-fs-shadow2d*
tests (except for proj+bias because of a GLSL bug).

13 years agomesa: more program debug code
Brian Paul [Tue, 14 Dec 2010 19:46:01 +0000 (12:46 -0700)]
mesa: more program debug code

13 years agotgsi: document texture opcodes
Brian Paul [Tue, 14 Dec 2010 19:45:36 +0000 (12:45 -0700)]
tgsi: document texture opcodes

13 years agoglsl: new glsl_strtod() wrapper to fix decimal point interpretation
Brian Paul [Mon, 13 Dec 2010 15:41:08 +0000 (08:41 -0700)]
glsl: new glsl_strtod() wrapper to fix decimal point interpretation

We always want to use '.' as the decimal point.

See http://bugs.freedesktop.org/show_bug.cgi?id=24531

NOTE: this is a candidate for the 7.10 branch.

13 years agogallivm: do texture swizzle after shadow compare
Brian Paul [Tue, 14 Dec 2010 17:38:15 +0000 (10:38 -0700)]
gallivm: do texture swizzle after shadow compare

We need to swizzle after the shadow comparison so that the GL_DEPTH_MODE
functionality is handled properly.

This fixes all the piglit glsl-fs-shadow2d*.shader_test cases, except
for glsl-fs-shadow2dproj-bias.shader_test which fails because of a
bug in the GLSL compiler (fd.o 32395).

13 years agost/mesa: rename the varient release functions
Brian Paul [Tue, 14 Dec 2010 15:36:39 +0000 (08:36 -0700)]
st/mesa: rename the varient release functions

13 years agor600g: fix segfault when translating vertex buffer
Jerome Glisse [Tue, 14 Dec 2010 18:50:46 +0000 (13:50 -0500)]
r600g: fix segfault when translating vertex buffer

Note the support for non float vertex draw likely regressed need to
find what we want to do there.

candidates for 7.10 branches

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
13 years agomesa: Clean up header file inclusion in prog_optimize.h.
Vinson Lee [Tue, 14 Dec 2010 08:39:57 +0000 (00:39 -0800)]
mesa: Clean up header file inclusion in prog_optimize.h.

13 years agomesa: Clean up header file inclusion in prog_cache.h.
Vinson Lee [Tue, 14 Dec 2010 08:30:24 +0000 (00:30 -0800)]
mesa: Clean up header file inclusion in prog_cache.h.

13 years agomesa: Clean up header file inclusion in nvvertparse.h.
Vinson Lee [Tue, 14 Dec 2010 08:22:27 +0000 (00:22 -0800)]
mesa: Clean up header file inclusion in nvvertparse.h.

13 years agoi965: Add support for using the BLT ring on gen6.
Eric Anholt [Mon, 13 Dec 2010 19:32:19 +0000 (11:32 -0800)]
i965: Add support for using the BLT ring on gen6.

13 years agoi965: Improve the hacks for ARB_fp scalar^scalar POW on gen6.
Eric Anholt [Mon, 13 Dec 2010 23:18:13 +0000 (15:18 -0800)]
i965: Improve the hacks for ARB_fp scalar^scalar POW on gen6.

This is still awful, but my ability to care about reworking the old
backend so we can just get a temporary value into a POW is awfully low
since the new backend does this all sensibly.

Fixes:
fp1-LIT test 1
fp1-LIT test 3 (case x < 0)
fp1-POW test (exponentiation)
fp-lit-mask

13 years agost/mesa: 80-columns wrapping, whitespace fixes in st_cb_bitmap.c
Brian Paul [Tue, 14 Dec 2010 00:34:07 +0000 (17:34 -0700)]
st/mesa: 80-columns wrapping, whitespace fixes in st_cb_bitmap.c

13 years agost/mesa: add geom program code in destroy_program_variants()
Brian Paul [Tue, 14 Dec 2010 00:29:56 +0000 (17:29 -0700)]
st/mesa: add geom program code in destroy_program_variants()

13 years agost/mesa: program struct comments
Brian Paul [Tue, 14 Dec 2010 00:28:02 +0000 (17:28 -0700)]
st/mesa: program struct comments

13 years agost/mesa: use st_fragment_program() instead of cast
Brian Paul [Tue, 14 Dec 2010 00:25:29 +0000 (17:25 -0700)]
st/mesa: use st_fragment_program() instead of cast

13 years agost/mesa: rename variable
Brian Paul [Tue, 14 Dec 2010 00:25:10 +0000 (17:25 -0700)]
st/mesa: rename variable

13 years agost/mesa: minor re-indenting
Brian Paul [Tue, 14 Dec 2010 00:18:47 +0000 (17:18 -0700)]
st/mesa: minor re-indenting