José Fonseca [Tue, 22 Sep 2009 16:40:20 +0000 (17:40 +0100)]
wgl: Flatten the source tree.
It is easier to have the WGL API on top of the ICD callbacks as
Microsoft's own implementation does, than to have a seperate shared
entity. This source reorganization is in antecipation of that.
José Fonseca [Tue, 22 Sep 2009 16:25:22 +0000 (17:25 +0100)]
wgl: Fix debug_printf format specifiers.
José Fonseca [Tue, 22 Sep 2009 16:16:35 +0000 (17:16 +0100)]
gallium: Update vendor string.
Brian Paul [Tue, 22 Sep 2009 22:06:18 +0000 (16:06 -0600)]
Merge branch 'mesa_7_6_branch'
Brian Paul [Tue, 22 Sep 2009 21:58:12 +0000 (15:58 -0600)]
mesa: don't re-use the meta glDrawPixels VBO; create a new one each time
This should help to work around bugs 24083 and 23670.
Brian Paul [Tue, 22 Sep 2009 21:28:34 +0000 (15:28 -0600)]
progs/perf: added glCopyTex[Sub]Image2D test
Richard Li [Tue, 22 Sep 2009 21:26:23 +0000 (17:26 -0400)]
r600 : add draw_prim support, make up one lost change.
Richard Li [Tue, 22 Sep 2009 20:39:11 +0000 (16:39 -0400)]
r600 : add draw_prim support.
Brian Paul [Tue, 22 Sep 2009 19:47:49 +0000 (13:47 -0600)]
mesa: fix more buffer object error messages
Brian Paul [Tue, 22 Sep 2009 19:44:43 +0000 (13:44 -0600)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch
Conflicts:
src/mesa/main/bufferobj.c
Brian Paul [Tue, 22 Sep 2009 19:19:05 +0000 (13:19 -0600)]
glx: include string.h to silence missing memset() prototype warning
Brian Paul [Tue, 22 Sep 2009 19:14:55 +0000 (13:14 -0600)]
mesa: fix error message text
Nicolai Hähnle [Tue, 22 Sep 2009 18:57:05 +0000 (20:57 +0200)]
r300: Fix crash reported in bug #24066
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Brian Paul [Tue, 22 Sep 2009 18:25:32 +0000 (12:25 -0600)]
progs/perf: test glGetTexImage() also
Jakob Bornecrantz [Tue, 22 Sep 2009 17:59:26 +0000 (10:59 -0700)]
st/xorg: Fix two leeks
We where leaking both surfaces in the composit code
and textures from pixmaps.
Jakob Bornecrantz [Tue, 22 Sep 2009 18:00:58 +0000 (11:00 -0700)]
i915g: Activate trace
Jakob Bornecrantz [Tue, 22 Sep 2009 17:55:41 +0000 (10:55 -0700)]
i915g: Do propper references of surfaces in context
Brian Paul [Tue, 22 Sep 2009 17:14:58 +0000 (11:14 -0600)]
progs/perf: add missing texture enable in fbobind.c
Brian Paul [Tue, 22 Sep 2009 17:03:43 +0000 (11:03 -0600)]
progs/perf: added fbobind.c test to test FBO binding speed
Brian Paul [Tue, 22 Sep 2009 17:02:04 +0000 (11:02 -0600)]
progs/perf: added PerfExtensionSupported() helper
Keith Whitwell [Tue, 22 Sep 2009 16:59:24 +0000 (17:59 +0100)]
teximage: add more tests, image formats
Add a test which creates a new texture from scratch before uploading.
Add more image formats.
Don't run all tests on all image formats.
Brian Paul [Tue, 22 Sep 2009 15:53:35 +0000 (09:53 -0600)]
progs/perf: add another VBO test for Create/Draw/Destroy pattern
Report both MB/sec and draw/sec.
Brian Paul [Tue, 22 Sep 2009 15:43:08 +0000 (09:43 -0600)]
progs/perf: simplify the code
Brian Paul [Tue, 22 Sep 2009 15:29:27 +0000 (09:29 -0600)]
progs/perf: fix comment
Brian Paul [Tue, 22 Sep 2009 13:50:12 +0000 (07:50 -0600)]
progs/perf: updated comments
Keith Whitwell [Tue, 22 Sep 2009 11:35:56 +0000 (12:35 +0100)]
progs/perf: make teximage results easier to read, more reproducible
Always run the same tests on different drivers, give zero results
where test image is too big for driver.
Add a newline between groups of tests.
Keith Whitwell [Tue, 22 Sep 2009 10:58:09 +0000 (11:58 +0100)]
progs/perf: reset row_length after subimage test
Also test fewer sizes in teximage generally.
Keith Whitwell [Tue, 22 Sep 2009 10:27:18 +0000 (11:27 +0100)]
progs/perf: clean up swapbuffers test a bit
Keith Whitwell [Tue, 22 Sep 2009 10:01:53 +0000 (11:01 +0100)]
progs/perf: break up long runs of fullscreen quads
Not all drivers cope gracefully with command-buffers with zillions of
fullscreen quads.
Keith Whitwell [Mon, 21 Sep 2009 15:55:12 +0000 (16:55 +0100)]
progs/perf: add first attempt at a swapbuffers rate test
This is pretty ugly as the original framework assumed you'd set
a single window size at startup and keep it throughout, but for
swapbuffers you want to test the rate at various window sizes.
With luck a nicer solution can be found, but this at least lays out
a marker.
Brian Paul [Mon, 21 Sep 2009 22:54:35 +0000 (16:54 -0600)]
vbo: added comment about max array index
Brian Paul [Mon, 21 Sep 2009 22:51:34 +0000 (16:51 -0600)]
vbo: restore some lost warning output
Tormod Volden [Sun, 20 Sep 2009 18:20:01 +0000 (20:20 +0200)]
GLX: Warn only once about applications calling GLX 1.3 functions
The warnings introduced in
1f309c40b8065b8729fce631540c66e4b50b84df
would pour out generously from some applications. This patch adds a
"warn once" wrapper macro, heavily inspired by
src/mesa/drivers/dri/r600/radeon_debug.h
Signed-off-by: Tormod Volden <debian.tormod@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Brian Paul [Mon, 21 Sep 2009 22:07:14 +0000 (16:07 -0600)]
vbo: disable the GL_ARB_draw_elements_base_vertex rebase path
This was introduced with commit
92d7ed8a20d4a018ce5324e6537ae7b478b9e5bf.
It causes rendering of stray polygons (with sw rendering at least) when
running the OGL Distilled / Picking demo (click on an object).
This needs additional debugging to fix/restore.
Found one suspect thing: in _tnl_draw_prims() there's some mixed signed/
unsigned arithmetic/comparing at line 422 that may be incorrect.
Adam Jackson [Mon, 21 Sep 2009 21:03:58 +0000 (17:03 -0400)]
selinux: Fix mmap() return value check
Brian Paul [Mon, 21 Sep 2009 20:23:07 +0000 (14:23 -0600)]
mesa: refine the error checking vbo_exec_DrawRangeElements()
If the 'end' index is out of bounds issue a warning as before. But instead
of just no-op'ing the draw call, examine the actual array indices to see
if they're OK. If the max array index is out of bounds, issue another
warning and no-op the draw call. Otherwise, draw normally. This is a
debug build-only feature since it could impact performance.
This "fixes" the missing torus in the OGL Distilled / Picking demo.
Brian Paul [Mon, 21 Sep 2009 20:07:35 +0000 (14:07 -0600)]
mesa: make max_buffer_index() a non-static function
Maciej Cencora [Sun, 20 Sep 2009 11:54:59 +0000 (13:54 +0200)]
Maciej Cencora [Sat, 19 Sep 2009 16:47:36 +0000 (18:47 +0200)]
mesa: add some debug info to teximage.c
Maciej Cencora [Sat, 19 Sep 2009 16:46:51 +0000 (18:46 +0200)]
r300: fix a typo
Brian Paul [Mon, 21 Sep 2009 14:32:43 +0000 (08:32 -0600)]
swrast: fix cube face selection
If arx and ary are equal, we still want to choose from one of them,
and not arz.
This is the same as Michal's softpipe fix.
Brian Paul [Mon, 21 Sep 2009 20:49:46 +0000 (14:49 -0600)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch
Alex Deucher [Mon, 21 Sep 2009 20:48:55 +0000 (16:48 -0400)]
r600: fix typo in the last commit
128 gprs, 256 reg-based consts
Alex Deucher [Mon, 21 Sep 2009 20:30:14 +0000 (16:30 -0400)]
r600: various cleanups
- max texture size is 8k, but mesa doesn't support
that at the moment.
- attempt to set shader limits to what the hw actually
supports
- clean up some old r300 cruft
- no need to explicitly disable irqs. This is fixed
in the drm now.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Brian Paul [Mon, 21 Sep 2009 20:23:07 +0000 (14:23 -0600)]
mesa: refine the error checking vbo_exec_DrawRangeElements()
If the 'end' index is out of bounds issue a warning as before. But instead
of just no-op'ing the draw call, examine the actual array indices to see
if they're OK. If the max array index is out of bounds, issue another
warning and no-op the draw call. Otherwise, draw normally. This is a
debug build-only feature since it could impact performance.
This "fixes" the missing torus in the OGL Distilled / Picking demo.
Brian Paul [Mon, 21 Sep 2009 20:07:35 +0000 (14:07 -0600)]
mesa: make max_buffer_index() a non-static function
Maciej Cencora [Sun, 20 Sep 2009 11:54:59 +0000 (13:54 +0200)]
Maciej Cencora [Sat, 19 Sep 2009 16:47:36 +0000 (18:47 +0200)]
mesa: add some debug info to teximage.c
Maciej Cencora [Sat, 19 Sep 2009 16:46:51 +0000 (18:46 +0200)]
r300: fix a typo
José Fonseca [Mon, 21 Sep 2009 18:57:57 +0000 (19:57 +0100)]
mesa: Ensure TGSI tokens are freed with gallium's free.
To avoid breaking the gallium's builtin malloc debugging.
Brian Paul [Mon, 21 Sep 2009 18:17:49 +0000 (12:17 -0600)]
progs/perf: added fill-rate test
Many more fill modes could be tested, but this hits the basics including
blending, texturing and shaders.
Brian Paul [Mon, 21 Sep 2009 17:58:03 +0000 (11:58 -0600)]
progs/perf: added helpers for creating simple textures and shader programs
Brian Paul [Mon, 21 Sep 2009 17:09:00 +0000 (11:09 -0600)]
progs/perf: offsetof() should be defined in stddef.h, include it
If this breaks mingw, feel free to revert this.
Eric Anholt [Wed, 20 May 2009 21:05:03 +0000 (14:05 -0700)]
intel: Mark the FBO as incomplete if there's no intel_renderbuffer for it.
This happens to rendering with textures with a border, which had resulted
in a segfault on dereferencing the irb.
(cherry-picked from commit
8bba183b9eeb162661a287bf2e118c6dd419dd24)
Nicolai Hähnle [Mon, 21 Sep 2009 15:35:10 +0000 (17:35 +0200)]
r300g: Fix bad formatting parameters in calls to debug_printf
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Nicolai Hähnle [Mon, 21 Sep 2009 15:28:37 +0000 (17:28 +0200)]
gallium debug: Add gcc printf hint to debug_printf
This causes gcc to issue warnings when format parameters do not match up
with the format string in calls to debug_printf.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Keith Whitwell [Mon, 21 Sep 2009 14:56:17 +0000 (15:56 +0100)]
progs/perf: a few more vbo upload modes
Some tests, eg small SubData are probably overwhelmed by the cost of
performing the draw after each upload. Add a varient which does a lot
of subdata uploads and then a single draw.
Also try to avoid cache-artifacts in the upload timings.
Keith Whitwell [Mon, 21 Sep 2009 14:54:28 +0000 (15:54 +0100)]
progs/perf: human-readable drawoverhead output
Keith Whitwell [Mon, 21 Sep 2009 14:52:17 +0000 (15:52 +0100)]
progs/perf: human-format vertexrate output
Keith Whitwell [Mon, 21 Sep 2009 14:51:26 +0000 (15:51 +0100)]
progs/perf: add human-readable float formatter
Vinson Lee [Mon, 21 Sep 2009 14:44:53 +0000 (08:44 -0600)]
progs/glsl: Include local headers before installed headers during compilation.
Fixes compilation errors on platforms with insufficient older installed
GL headers.
Brian Paul [Mon, 21 Sep 2009 14:34:00 +0000 (08:34 -0600)]
softpipe: Fix cube face selection.
If arx and ary are equal, we still want to choose from one of them,
and not arz.
(cherry picked from commit
de685b37a91bc95dd4093a44a49b7b47385b1f7c)
Brian Paul [Mon, 21 Sep 2009 14:32:43 +0000 (08:32 -0600)]
swrast: fix cube face selection
If arx and ary are equal, we still want to choose from one of them,
and not arz.
This is the same as Michal's softpipe fix.
Andre Maasikas [Mon, 21 Sep 2009 14:14:25 +0000 (10:14 -0400)]
r600: fix some issues with LIT instruction
- MUL_LIT is ALU.Trans instruction
- some Trans instructions can take 3 arguments
- don't clobber dst.x, use dst.z as temp, it'll get written correct
value in last insn
- respect source swizzles
Nicolai Hähnle [Mon, 21 Sep 2009 11:08:34 +0000 (13:08 +0200)]
Merge branch 'mesa_7_6_branch'
Nicolai Hähnle [Mon, 21 Sep 2009 10:50:33 +0000 (12:50 +0200)]
r300: Zero-initialize register for NV_vertex_program
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Nicolai Hähnle [Wed, 9 Sep 2009 17:56:57 +0000 (19:56 +0200)]
r300: Fix handling of NV_vertex_program parameters
The handling is a bit inefficient, unfortunately, but I don't want to make
any intrusive changes for Mesa 7.6.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Michel Dänzer [Mon, 21 Sep 2009 08:48:32 +0000 (10:48 +0200)]
Merge branch 'mesa_7_6_branch'
Michel Dänzer [Mon, 21 Sep 2009 08:39:20 +0000 (10:39 +0200)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch
Michel Dänzer [Mon, 21 Sep 2009 08:08:11 +0000 (10:08 +0200)]
intel: Fix crash in intel_flush().
Since commit
2921a2555d0a76fa649b23c31e3264bbc78b2ff5 ('intel: Deassociated
drawables from private context struct in intelUnbindContext'),
intel->driDrawable may be NULL in intel_flush().
Ben Skeggs [Mon, 21 Sep 2009 04:51:25 +0000 (14:51 +1000)]
nouveau: allow building modesetting_drv.so
Ben Skeggs [Mon, 21 Sep 2009 04:46:50 +0000 (14:46 +1000)]
nouveau: drm_api create_screen()'s 'arg' argument can be NULL
Brian Paul [Mon, 21 Sep 2009 04:19:28 +0000 (22:19 -0600)]
windows: call _mesa_meta_init/free()
Brian Paul [Mon, 21 Sep 2009 04:19:06 +0000 (22:19 -0600)]
xlib: always call _mesa_meta_init/free()
Brian Paul [Mon, 21 Sep 2009 04:13:28 +0000 (22:13 -0600)]
dri/swrast: call _mesa_meta_init/free()
Brian Paul [Mon, 21 Sep 2009 04:13:09 +0000 (22:13 -0600)]
osmesa: call _mesa_meta_init/free()
Brian Paul [Mon, 21 Sep 2009 04:08:30 +0000 (22:08 -0600)]
Revert "st/mesa: Compile in meta.c."
This reverts commit
6c5726cd39ab12b86fae391d075fa74bc24b615c.
Brian Paul [Mon, 21 Sep 2009 04:08:21 +0000 (22:08 -0600)]
Revert "scons: add meta.c to sources"
This reverts commit
41fefe88c50376a57876b498c8619c8c9f535de6.
Brian Paul [Mon, 21 Sep 2009 04:06:03 +0000 (22:06 -0600)]
Revert "mesa: move _mesa_meta_init/free() calls to core Mesa"
This reverts commit
651cffd626a82d9bf539437ca4bdf8ea4b396fab.
The commit inadvertantly introduced a new gallium dependency on the meta code.
Brian Paul [Mon, 21 Sep 2009 03:31:38 +0000 (21:31 -0600)]
scons: add meta.c to sources
Chia-I Wu [Mon, 21 Sep 2009 02:57:14 +0000 (10:57 +0800)]
st/mesa: Compile in meta.c.
meta was moved to core Mesa since
651cffd626a82d9bf539437ca4bdf8ea4b396fab.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Chia-I Wu [Mon, 21 Sep 2009 02:52:07 +0000 (10:52 +0800)]
intel: meta clear has a new name.
It was renamed to _mesa_meta_Clear.
Signed-off-by: Chia-I Wu <olvaffe@gmail.com>
Dave Airlie [Mon, 21 Sep 2009 03:26:48 +0000 (13:26 +1000)]
xorg/st: fixup builds against later dpms headers.
Pauli Nieminen [Sun, 20 Sep 2009 19:24:35 +0000 (22:24 +0300)]
radeon: Fix legacy bo not to reuse dma buffers before refcount is 1.
This should help detecting possible memory leaks with dma buffers and prevent
possible visual corruption if data would be overwriten too early.
Nicolai Hähnle [Sun, 20 Sep 2009 18:40:03 +0000 (20:40 +0200)]
r300/compiler: Fix trig instructions in R300 fp
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Pauli Nieminen [Sun, 20 Sep 2009 18:08:42 +0000 (21:08 +0300)]
radeon: Fix typo in variable name.
Pauli Nieminen [Sun, 20 Sep 2009 17:07:35 +0000 (20:07 +0300)]
radeon: Improve WARN_ONCE macro to appear as single statement.
Do-while makes macro safe to be used with if and for constructions.
Also remove __LINE__ macro from variable name because scope is local to macro anyway.
José Fonseca [Sun, 20 Sep 2009 17:07:16 +0000 (18:07 +0100)]
scons: Drop gprof support for profile builds; tweak optimization flags instead.
gprof is useful for shared libraries, hence our drivers. Nevertheless
profilers like oprofile can benefit from disabling some relatively
minor optimizations for more accurate / complete results.
José Fonseca [Sun, 20 Sep 2009 17:04:00 +0000 (18:04 +0100)]
llvmpipe: Fix lp_get_cached_tile.
Align coordinates to tile boundaries.
José Fonseca [Sun, 20 Sep 2009 11:28:07 +0000 (12:28 +0100)]
llvmpipe: Update tile status on flush.
Nicolai Hähnle [Sun, 20 Sep 2009 16:45:32 +0000 (18:45 +0200)]
radeon: Fix "verts" debugging enable
Copy'n'paste apparently prevented the RADEON_VERTS flag from being enabled.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Brian Paul [Sun, 20 Sep 2009 16:36:45 +0000 (10:36 -0600)]
Merge branch 'mesa_7_6_branch'
Conflicts:
src/mesa/drivers/dri/intel/intel_clear.c
Nicolai Hähnle [Sun, 20 Sep 2009 14:59:03 +0000 (16:59 +0200)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch
Nicolai Hähnle [Sat, 12 Sep 2009 14:49:31 +0000 (16:49 +0200)]
mesa/st: Create front renderbuffer on the fly when supplied with a surface
Normally, the mesa/st would create a fake front buffer out of a
client-allocated surface.
In the DRI setting, however, st/dri provides a front buffer surface which is
created and maintained by the X server. Prefer to use this surface instead,
so that front buffer rendering and reading works correctly.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Nicolai Hähnle [Sat, 12 Sep 2009 10:13:35 +0000 (12:13 +0200)]
mesa/st: Initialize format bits of framebuffer renderbuffers
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Nicolai Hähnle [Sun, 20 Sep 2009 14:46:58 +0000 (16:46 +0200)]
docs: Document new features in radeon/r200/r300 drivers
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Nicolai Hähnle [Sun, 20 Sep 2009 14:33:59 +0000 (16:33 +0200)]
r300/compiler: Fix R300 fragment program regression introduced by 0723cd1...
We obviously need to move the code addr register backwards because their may
be overlap.
This bug affected in particular the Compiz water plugin.
Signed-off-by: Nicolai Hähnle <nhaehnle@gmail.com>
Brian Paul [Sat, 19 Sep 2009 23:26:14 +0000 (17:26 -0600)]
mesa: remove redundant readbuffer check
Brian Paul [Sat, 19 Sep 2009 23:06:15 +0000 (17:06 -0600)]
mesa: rename functions to be more consistant with rest of mesa
Brian Paul [Sat, 19 Sep 2009 22:42:36 +0000 (16:42 -0600)]
swrast: remove obsolete s_imaging.c file