Brian [Tue, 24 Apr 2007 04:03:11 +0000 (22:03 -0600)]
remove SWvertex->fog field, use attrib field
Brian [Tue, 24 Apr 2007 04:01:34 +0000 (22:01 -0600)]
remove SWvertex->fog field, use attrib field
Brian [Tue, 24 Apr 2007 03:27:13 +0000 (21:27 -0600)]
some re-org, clean-up
Brian [Tue, 24 Apr 2007 03:21:52 +0000 (21:21 -0600)]
Improve the code for interpolating fragment attributes a little. More to come...
Brian [Sun, 22 Apr 2007 21:32:10 +0000 (15:32 -0600)]
import glxext.h version 19
Sean D'Epagnier [Sun, 22 Apr 2007 03:18:42 +0000 (20:18 -0700)]
Fixed fallout, glfbdev driver was calling a nop function that was removed.
Brian [Sat, 21 Apr 2007 21:58:37 +0000 (15:58 -0600)]
Fix color sum bug 10688.
Brian [Sat, 21 Apr 2007 21:56:10 +0000 (15:56 -0600)]
specularColor.alpha should default to 1, not 0
Brian [Sat, 21 Apr 2007 20:12:57 +0000 (14:12 -0600)]
fix SkipPixels bugs in _mesa_pack_bitmap(), bug 10690
Brian [Sat, 21 Apr 2007 19:18:06 +0000 (13:18 -0600)]
Remove the !rb->Data check that was added a few months ago.
Was changed while debugging #7205. Broke the shadowtext demo. Revisit this
if the problem w/ bug 7205 returns...
Brian [Sat, 21 Apr 2007 18:54:23 +0000 (12:54 -0600)]
s/occlude.h/queryobj.h/
Brian [Sat, 21 Apr 2007 18:42:54 +0000 (12:42 -0600)]
Rename occlude.[ch] to queryobj.[ch]
Brian [Sat, 21 Apr 2007 18:36:39 +0000 (12:36 -0600)]
Rename occlude.[ch] to queryobj.[ch] to better reflect contents.
Brian [Sat, 21 Apr 2007 18:31:51 +0000 (12:31 -0600)]
prep for 6.5.3 rc2
Brian [Sat, 21 Apr 2007 18:30:54 +0000 (12:30 -0600)]
memory leaks fixed
Brian [Sat, 21 Apr 2007 18:29:46 +0000 (12:29 -0600)]
Remove all the USE_MEMPOOL debug code.
Brian [Sat, 21 Apr 2007 18:08:56 +0000 (12:08 -0600)]
Use new memory pool allocator. Lots of debug code still in place...
Ian Romanick [Sat, 21 Apr 2007 17:47:54 +0000 (10:47 -0700)]
Fix FBO completeness bug in shadowtex.
Brian [Sat, 21 Apr 2007 16:18:49 +0000 (10:18 -0600)]
added MemPool field
Brian [Fri, 20 Apr 2007 18:59:26 +0000 (12:59 -0600)]
call grammar_alloc_free() to fix memleaks
Brian [Fri, 20 Apr 2007 17:59:18 +0000 (11:59 -0600)]
destroy window upon exit
Brian [Fri, 20 Apr 2007 17:54:30 +0000 (11:54 -0600)]
remove more unused vars, assertions
Brian [Fri, 20 Apr 2007 17:53:48 +0000 (11:53 -0600)]
remove unused vars, assertions
Brian [Fri, 20 Apr 2007 14:30:42 +0000 (08:30 -0600)]
added slang_mem.c
Brian [Fri, 20 Apr 2007 14:30:24 +0000 (08:30 -0600)]
Memory pool manager.
Instead of doing intricate malloc/free tracking during compiling, allocate
everything (which would be discarded after compiling) from a pool that can
be freed en masse when done.
Brian [Fri, 20 Apr 2007 14:21:49 +0000 (08:21 -0600)]
undo a test/debug change
Brian [Fri, 20 Apr 2007 14:12:17 +0000 (08:12 -0600)]
fix instruction comment code
Brian [Fri, 20 Apr 2007 14:11:51 +0000 (08:11 -0600)]
another bit of debug code
Brian [Thu, 19 Apr 2007 21:37:24 +0000 (15:37 -0600)]
free subroutine array (fix mem leak)
Brian [Thu, 19 Apr 2007 21:30:11 +0000 (15:30 -0600)]
new varnames in slang_operation_insert()
Brian [Thu, 19 Apr 2007 21:28:42 +0000 (15:28 -0600)]
fix memleak in slang_operation_insert()
Brian [Thu, 19 Apr 2007 21:23:34 +0000 (15:23 -0600)]
free shProg->Attributes in _mesa_free_shader_program_data()
Brian [Thu, 19 Apr 2007 21:22:04 +0000 (15:22 -0600)]
move allocation of shProg->Attributes earlier in function
Brian [Sat, 21 Apr 2007 15:47:03 +0000 (09:47 -0600)]
Call _mesa_warning() instead of _mesa_problem() when external s3tc lib not found.
The later incorrectly encourages filing a bug (10703, for example).
Jeremy Kolb [Sat, 21 Apr 2007 01:53:38 +0000 (21:53 -0400)]
Make xcb backend compile.
Add nouveau to match linux-dri config.
Brian [Sat, 21 Apr 2007 01:18:13 +0000 (19:18 -0600)]
Remove unused Imakefiles.
Brian [Sat, 21 Apr 2007 01:01:21 +0000 (19:01 -0600)]
Long obsolete.
Brian [Sat, 21 Apr 2007 00:55:13 +0000 (18:55 -0600)]
added new xm_glide.c file
George Sapountzis [Fri, 20 Apr 2007 15:51:20 +0000 (18:51 +0300)]
xmesa: call _glapi_set_dispatch() for all xserver DDXes.
This is to unify the xmesa code across xserver DDX'es. The call is intented for
XGL, but it does not hurt to call for other DDX'es. In fact it was not guarded
against XGL when it was first added in xserver.
George Sapountzis [Wed, 18 Apr 2007 20:44:01 +0000 (23:44 +0300)]
xmesa: minor cosmetic
mainly drop 'client' argument from initialize_visual_and_buffer().
George Sapountzis [Fri, 20 Apr 2007 15:25:42 +0000 (18:25 +0300)]
xmesa: split FX functions to separate file, part 2.
George Sapountzis [Fri, 20 Apr 2007 15:12:38 +0000 (18:12 +0300)]
xmesa: split FX functions to separate file, part 1.
George Sapountzis [Fri, 20 Apr 2007 14:51:55 +0000 (17:51 +0300)]
xmesa: spilt FX code to separate functions.
George Sapountzis [Fri, 20 Apr 2007 14:35:01 +0000 (17:35 +0300)]
Revert "xmesa: drop glide (FX) backend."
This reverts commit
2a2f8d806f74619f0a7cf97fdc7f7b3ad1cad81b.
Brian [Thu, 19 Apr 2007 20:24:58 +0000 (14:24 -0600)]
Notes about shared lib version. Document depth-peel fixes.
Brian [Thu, 19 Apr 2007 20:24:29 +0000 (14:24 -0600)]
dFdx(), etc. don't work yet
Brian [Thu, 19 Apr 2007 20:24:10 +0000 (14:24 -0600)]
Put gl_program_machine into swrast structure rather than using a local variable.
Basically an easy way to make sure the memory gets initialized once (to zero)
to avoid lots of valgrind warnings.
Brian [Thu, 19 Apr 2007 20:15:11 +0000 (14:15 -0600)]
Fix valgrind problem caused by reading text[-1].
Brian [Thu, 19 Apr 2007 20:07:16 +0000 (14:07 -0600)]
remove invalid assertion (span->facing may be set because of polygonmode)
Brian [Thu, 19 Apr 2007 20:06:43 +0000 (14:06 -0600)]
remove invalid assertion
Brian [Thu, 19 Apr 2007 20:06:27 +0000 (14:06 -0600)]
fix crash when program has invalid structure field
Brian [Thu, 19 Apr 2007 20:04:30 +0000 (14:04 -0600)]
init A.curFuncEndLabel = NULL
Brian [Thu, 19 Apr 2007 17:23:26 +0000 (11:23 -0600)]
In _mesa_unpack_depth_span() look for special cases of GLuint->GLushort and GLushort->GLuint conversion.
This improves performance and avoids int/float/int conversion problems that
can introduce errors during glCopyTexImage(). Another fix for the depth peeling
algorithm.
Brian [Thu, 19 Apr 2007 17:21:14 +0000 (11:21 -0600)]
Don't allow deferredTexture if using occlusion query and a frag shader.
Occlusion query might depend on the shader killing/discarding fragments.
Helps fix depth peeling technique.
Also, minor tweaks in interpolate_wpos().
Brian [Thu, 19 Apr 2007 00:05:33 +0000 (18:05 -0600)]
fix an uninitialized variable and a warning
Brian [Wed, 18 Apr 2007 22:48:55 +0000 (16:48 -0600)]
comments, assertions
Brian [Wed, 18 Apr 2007 22:46:53 +0000 (16:46 -0600)]
new comments
Brian [Wed, 18 Apr 2007 22:44:02 +0000 (16:44 -0600)]
record proper datatypes for uniforms/samplers
Brian [Wed, 18 Apr 2007 22:43:40 +0000 (16:43 -0600)]
added _slang_gltype_from_specifier()
Brian [Wed, 18 Apr 2007 22:27:35 +0000 (16:27 -0600)]
Start fixing some issues with uniform variables and their types.
Brian [Wed, 18 Apr 2007 22:05:53 +0000 (16:05 -0600)]
Fix some bugs related to querying active uniforms.
Brian [Wed, 18 Apr 2007 21:22:43 +0000 (15:22 -0600)]
fix error strings, add a sanity check
George Sapountzis [Wed, 18 Apr 2007 20:54:00 +0000 (23:54 +0300)]
xmesa: drop glide (FX) backend.
glide is no longer compiled with stand-alone libGL, so this will not link.
There are still the glide config files. some code in demos and the
GLX_MESA_set_3dfx_mode code which could be removed.
Brian [Wed, 18 Apr 2007 20:55:18 +0000 (14:55 -0600)]
Remove bad assertions, decrement NumShaders in _mesa_detach_shader().
Fixes crashes w/ Brad King's depth peeling test.
Brian [Wed, 18 Apr 2007 20:53:23 +0000 (14:53 -0600)]
s/GL_SHADER_PROGRAM/GL_SHADER_PROGRAM_MESA/ (a Mesa-specific token)
Brian [Wed, 18 Apr 2007 20:19:17 +0000 (14:19 -0600)]
Replace _mesa_parameter_longest_name() with _mesa_longest_parameter_name().
The later takes a type parameter so we can match uniforms or attributes/inputs.
Used by the GL_ACTIVE_ATTRIBUTE_MAX_LENGTH and GL_ACTIVE_UNIFORM_MAX_LENGTH
queries. Fixes problem reported by Brad King in VTK.
Brian [Wed, 18 Apr 2007 18:41:16 +0000 (12:41 -0600)]
call ProgramStringNotify() after linking
Brian [Wed, 18 Apr 2007 18:09:40 +0000 (12:09 -0600)]
check _PreferPixelFog in _swrast_span_default_fog(), see bug 10669
Brian [Wed, 18 Apr 2007 18:08:18 +0000 (12:08 -0600)]
improved fog comment
Brian [Wed, 18 Apr 2007 15:30:07 +0000 (09:30 -0600)]
glean glsl test now does over 150 tests
Brian [Wed, 18 Apr 2007 15:28:38 +0000 (09:28 -0600)]
added href to shading language info
Brian [Wed, 18 Apr 2007 15:07:56 +0000 (09:07 -0600)]
prep for 6.5.3 release
Brian [Wed, 18 Apr 2007 14:52:18 +0000 (08:52 -0600)]
add progs/glsl/ to DEMO_FILES, s/pre/rc1/
Xiang, Haihao [Wed, 18 Apr 2007 04:37:09 +0000 (12:37 +0800)]
fig segment fault issue in TAG(triangle), (see bug 10589)
Bit SS_TWOSIDE_BIT is set if gl_FrontFacing lives in
fragment input (see commit
10b5895597d5e069183cb647d17eb412effceb4f).
However, VB->ColorPtr[1] isn't assigned after that.
Brian [Tue, 17 Apr 2007 22:00:17 +0000 (16:00 -0600)]
omit the glide/svga sources from stand-alone libGL (anyone using them?)
Brian [Tue, 17 Apr 2007 21:56:46 +0000 (15:56 -0600)]
Enable texture sampling for vertex programs/shaders.
This is a bit of a hack for now because the tnl module is using the swrast
module to fetch texels. The texture fetch/filter code should probably be
moved into the main/ module since it doesn't really depend upon other
swrast code.
Brian [Tue, 17 Apr 2007 16:19:47 +0000 (10:19 -0600)]
In bind_inputs() set VB->EdgeFlag to NULL if it's not needed.
Otherwise, the clip/interp code was finding VB->EdgeFlag to be non-null and
reading/writing it when the memory may have been freed earlier in free_space().
This fixes several VTK segfaults/failures reported by Brad King @ Kitware.
Brian [Tue, 17 Apr 2007 15:16:59 +0000 (09:16 -0600)]
regenerated
Brian [Tue, 17 Apr 2007 15:16:30 +0000 (09:16 -0600)]
fix/simplify some texture functions
Brian [Tue, 17 Apr 2007 15:15:57 +0000 (09:15 -0600)]
more matrix function updates
Brian [Tue, 17 Apr 2007 14:29:37 +0000 (08:29 -0600)]
don't set GL_TEXTURE_MAX_LEVEL for GL_TEXTURE_RECTANGLE_ARB as that generates an error
Thomas Hellstrom [Tue, 17 Apr 2007 13:21:54 +0000 (15:21 +0200)]
Defer buffer pool creation to the first context creation.
This way we have a hw context so that we can take the hardware lock.
Also, at this point, AIGLX isn't locked with the X server context as it is
at screen creation.
Roland Scheidegger [Tue, 17 Apr 2007 12:01:42 +0000 (14:01 +0200)]
cleanups for t_vb_program.c
use VertexProgram._Current instead of VertexProgram.Current in a few more places.
Only fixup fogc and psiz in case this is really a nv program (others are fine
if undefined), and fix this case up so the values actually get written.
Oliver McFadden [Tue, 17 Apr 2007 07:05:15 +0000 (07:05 +0000)]
r300: r300_render.c:391: warning: unused variable 'i'
Brian [Mon, 16 Apr 2007 23:56:13 +0000 (17:56 -0600)]
lots of changes, fixes, clean-ups
Brian [Mon, 16 Apr 2007 23:36:39 +0000 (17:36 -0600)]
just clean-ups
Brian [Mon, 16 Apr 2007 23:18:39 +0000 (17:18 -0600)]
move GL_EXT_stencil_two_side into alphabetical position
Brian [Mon, 16 Apr 2007 23:16:46 +0000 (17:16 -0600)]
Fix glActiveStencilFaceEXT dispatch problem (bug 10523).
OK, _all_ extensions that might get enabled by the driver need to be in the
card_extensions[] list. driInitExtensions() is called at least twice: first
during screen creation, then once for each context that's created.
The first call sets up the dispatch table. The second call just sets the
extension enable/disable flags.
Brian [Mon, 16 Apr 2007 21:15:53 +0000 (15:15 -0600)]
remove _tnl_arb_vertex_program_stage
Brian [Mon, 16 Apr 2007 21:15:23 +0000 (15:15 -0600)]
remove _tnl_arb_vertex_program_stage
Brian [Mon, 16 Apr 2007 21:09:39 +0000 (15:09 -0600)]
need to clamp MAD for linear fog
Brian [Mon, 16 Apr 2007 17:03:14 +0000 (11:03 -0600)]
removed unused t_vb_arbprogram stage
Brian [Mon, 16 Apr 2007 16:59:00 +0000 (10:59 -0600)]
remove t_vb_arbprogram.c
Brian [Mon, 16 Apr 2007 16:58:30 +0000 (10:58 -0600)]
unhook t_vb_arbprogram.c code - it's going away
Brian [Mon, 16 Apr 2007 16:36:28 +0000 (10:36 -0600)]
Use generic program limits instead of NV-specific ones to init program constants.
Previously, this limited us to 12 temp regs for vertex programs. Many vertex
shaders could exceed that. This forces us to stop using t_vb_arbprogram.c
for now because of its particular register indexing scheme. Need to increase
bits allocated for register indexing, etc.
Brian [Mon, 16 Apr 2007 16:32:48 +0000 (10:32 -0600)]
Fix some assertions that could occur when an error was earlier logged.
Brian [Mon, 16 Apr 2007 15:11:20 +0000 (09:11 -0600)]
use b->display instead of b->xm_visual->display to fix some problems detected w/ valgrind
Thomas Hellstrom [Mon, 16 Apr 2007 14:04:12 +0000 (16:04 +0200)]
Make sure we are locked when creating drm buffer objects.
Don't place buffer objects on unfenced list when newly created.
Fix a buffer object wait-for-idle deadlock.
Brian [Sat, 14 Apr 2007 15:38:20 +0000 (09:38 -0600)]
illegal to set GL_TEXTURE_MAX_LEVEL w/ GL_TEXTURE_RECTANGLE_ARB
Brian [Sat, 14 Apr 2007 15:36:17 +0000 (09:36 -0600)]
implement shadow2DRect functions