Brian Paul [Wed, 18 Nov 2009 02:22:11 +0000 (19:22 -0700)]
progs/util: warn when GLSL not supported
Ian Romanick [Wed, 18 Nov 2009 00:10:24 +0000 (16:10 -0800)]
Merge branch 'outputswritten64'
Add a GLbitfield64 type and several macros to operate on 64-bit
fields. The OutputsWritten field of gl_program is changed to use that
type. This results in a fair amount of fallout in drivers that use
programs.
No changes are strictly necessary at this point as all bits used are
below the 32-bit boundary. Fairly soon several bits will be added for
clip distances written by a vertex shader. This will cause several
bits used for varyings to be pushed above the 32-bit boundary. This
will affect any drivers that support GLSL.
At this point, only the i965 driver has been modified to support this
eventuality.
I did this as a "squash" merge. There were several places through the
outputswritten64 branch where things were broken. I foresee this
causing difficulties later for bisecting. The history is still
available in the branch.
Conflicts:
src/mesa/drivers/dri/i965/brw_wm.h
Ian Romanick [Thu, 12 Nov 2009 21:28:12 +0000 (13:28 -0800)]
ARB_fbo: Enable extensions related to GL_ARB_framebuffer_object for GLX
Ian Romanick [Thu, 12 Nov 2009 21:22:12 +0000 (13:22 -0800)]
ARB_fbo: Add missing GLX render opcode for glFramebufferTextureLayerEXT
Also regenerate the GLX protocol files from this change.
Ian Romanick [Thu, 12 Nov 2009 19:59:35 +0000 (11:59 -0800)]
Add missing XML files to API_XML
When the files missing from the list were modified, the generated files
weren't regenerated.
Ian Romanick [Wed, 11 Nov 2009 23:33:23 +0000 (15:33 -0800)]
ARB_fbo: Enable GL_EXT_framebuffer_multisample
All of this functionality is already included in ARB_fbo. This just
enables the string.
I was a bit lazy in using FEATURE_ARB_framebuffer_object for this
feature as well. I don't think it makes much difference in the long run.
Ian Romanick [Thu, 12 Nov 2009 19:50:28 +0000 (11:50 -0800)]
ARB_fbo: Regenerate files from previous commits.
Ian Romanick [Wed, 11 Nov 2009 23:18:22 +0000 (15:18 -0800)]
ARB_fbo: Move EXT_fb_blit to EXT_framebuffer_object.xml
This extension is layered on GL_EXT_framebuffer_object, so it should
live in the same XML file.
Ian Romanick [Wed, 11 Nov 2009 23:15:35 +0000 (15:15 -0800)]
ARB_fbo: Add GL_EXT_framebuffer_multisample
Add GL_EXT_framebuffer_multisample. Make
glRenderbufferStorageMultisampleEXT in GL_EXT_framebuffer_object alias
glRenderbufferStorageMultisample. Also add the missing GLX render
opcode to glRenderbufferStorageMultisample.
Since this extension is layered on GL_EXT_framebuffer_object, I put it
in EXT_framebuffer_object.xml.
Ian Romanick [Wed, 11 Nov 2009 23:09:44 +0000 (15:09 -0800)]
EXT_pds: Alias GL_NV_packed_depth_stencil to GL_EXT_packed_depth_stencil
GL_EXT_packed_depth_stencil is a functional superset of
GL_NV_packed_depth_stencil. If a driver enables EXT_pds, make NV_pds
available as well.
Ian Romanick [Wed, 11 Nov 2009 23:08:49 +0000 (15:08 -0800)]
EXT_pds: Add GL_EXT_packed_depth_stencil
Ian Romanick [Wed, 11 Nov 2009 23:01:25 +0000 (15:01 -0800)]
EXT_fbo_blit: Add missing GLX protocol render opcode
Ian Romanick [Wed, 11 Nov 2009 22:51:29 +0000 (14:51 -0800)]
ARB_fbo: Add missing protocol "Get" information
Ian Romanick [Wed, 11 Nov 2009 22:44:43 +0000 (14:44 -0800)]
ARB_fbo: Remove _EXT from enum names
Brian Paul [Tue, 17 Nov 2009 23:32:05 +0000 (16:32 -0700)]
docs: libGL documentation, from the xorg-docs tree
This is pretty old and needs some updating but might be useful or
interesting to some people.
Vinson Lee [Tue, 17 Nov 2009 23:24:40 +0000 (15:24 -0800)]
progs/tests: Fix memory leak in arbgpuprog.c.
Brian Paul [Tue, 17 Nov 2009 23:18:29 +0000 (16:18 -0700)]
mesa: remove old comment
Brian Paul [Tue, 17 Nov 2009 23:16:16 +0000 (16:16 -0700)]
mesa: fix assorted compiler warnings
Brian Paul [Tue, 17 Nov 2009 23:15:56 +0000 (16:15 -0700)]
mesa: fix incorrect type in store_texel_al1616()
Brian Paul [Tue, 17 Nov 2009 23:15:29 +0000 (16:15 -0700)]
i915g: replace //-style comments
Brian Paul [Tue, 17 Nov 2009 23:15:21 +0000 (16:15 -0700)]
gallium/util: replace //-style comments
Brian Paul [Tue, 17 Nov 2009 23:14:54 +0000 (16:14 -0700)]
i915g: remove trailing commas in enum lists to silence warnings
Brian Paul [Tue, 17 Nov 2009 23:14:20 +0000 (16:14 -0700)]
progs/xdemos: change //-style comments
Brian Paul [Tue, 17 Nov 2009 23:14:09 +0000 (16:14 -0700)]
progs/glsl: change //-style comments
Brian Paul [Tue, 17 Nov 2009 23:13:12 +0000 (16:13 -0700)]
progs/demos: silence warnings
Brian Paul [Tue, 17 Nov 2009 23:05:59 +0000 (16:05 -0700)]
mesa: remove trailing comment to silence warning
Brian Paul [Tue, 17 Nov 2009 23:05:24 +0000 (16:05 -0700)]
mesa: remove a bit of old code
Vinson Lee [Tue, 17 Nov 2009 22:22:35 +0000 (14:22 -0800)]
progs/slang: Fix potential null pointer dereference in cltest.c.
Vinson Lee [Tue, 17 Nov 2009 22:10:35 +0000 (14:10 -0800)]
progs/redbook: Fix memory leak in quadric.c.
Jakob Bornecrantz [Tue, 17 Nov 2009 11:04:17 +0000 (12:04 +0100)]
svga: More -Werror removal
Vinson Lee [Tue, 17 Nov 2009 21:49:31 +0000 (13:49 -0800)]
progs/util: Fix memory leak if fopen fails in RawImageOpen.
Jakob Bornecrantz [Tue, 17 Nov 2009 08:07:15 +0000 (09:07 +0100)]
svga: Remove -Werror for now as GCC 4.4.x raises a bunch of warnings
Maciej Cencora [Tue, 17 Nov 2009 20:27:31 +0000 (21:27 +0100)]
r300: fix reads and writes for MESA_FORMAT_S8Z24 buffer
Regression was introduced by texformat-rework branch merge.
Brian Paul [Tue, 17 Nov 2009 20:39:13 +0000 (13:39 -0700)]
progs/xdemos: added -lX11 -lpthread for GNU gold linker
Vinson Lee [Tue, 17 Nov 2009 20:00:22 +0000 (12:00 -0800)]
progs/util: Fix memory leak if fread returns 0 in CompileShaderFile.
Brian Paul [Tue, 17 Nov 2009 19:47:12 +0000 (12:47 -0700)]
mesa: move -ansi -pedantic to CFLAGS
Vinson Lee [Tue, 17 Nov 2009 19:37:48 +0000 (11:37 -0800)]
progs/glsl: Change tangentAttrib from GLuint to GLint in bump.c.
tangentAtrrib is assigned the result of glGetAttribLocation.
The assertion 'assert(tangentAtrrib >= 0)' would be a no-op if
tangentAttrib is a GLuint.
Vinson Lee [Tue, 17 Nov 2009 19:04:24 +0000 (11:04 -0800)]
progs/util: Fix memory leak if malloc fails in tkRGBImageLoad.
Patrice Mandin [Tue, 17 Nov 2009 18:50:37 +0000 (19:50 +0100)]
nouveau: nv30: Check for NULL front (happens with DRI2)
Patrice Mandin [Tue, 17 Nov 2009 18:49:56 +0000 (19:49 +0100)]
nouveau: nv30: Add missing include to fix warning
Michel Dänzer [Tue, 17 Nov 2009 18:41:29 +0000 (19:41 +0100)]
Remove unconditional use of glibc specific bswap_16() macro.
Fixes unresolved symbol bswap_16 on non-glibc or little endian glibc platforms.
Brian Paul [Tue, 17 Nov 2009 18:23:03 +0000 (11:23 -0700)]
docs: list the new VMware SVGA Gallium driver in release notes
Dan Nicholson [Mon, 9 Nov 2009 13:51:23 +0000 (05:51 -0800)]
dri: Ensure subdirs have finished before linking driver
Recursive make is hard. If there are subdirectories in the DRI drivers,
it's pretty certain we want to finish building in them before linking
the driver. Add a new target to serialize the rules.
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
Vinson Lee [Tue, 17 Nov 2009 18:11:50 +0000 (10:11 -0800)]
progs/util: Fix memory leak if LoadRGBImage fails.
Michel Dänzer [Tue, 17 Nov 2009 18:02:03 +0000 (19:02 +0100)]
r600: More span breakage fixes.
At least now the compiler doesn't complain about implicitly declared functions
anymore...
Vinson Lee [Tue, 17 Nov 2009 17:53:19 +0000 (09:53 -0800)]
progs/glsl: Fix memory leak in deriv.c.
Michel Dänzer [Tue, 17 Nov 2009 17:47:24 +0000 (18:47 +0100)]
r600: Attempt to fix span breakage introduced by big endian fixes.
Only compile tested; I happened to notice people on IRC reporting
.../r600_dri.so: undefined symbol: radeon_ptr_2byte_8x2
Vinson Lee [Tue, 17 Nov 2009 16:29:36 +0000 (08:29 -0800)]
progs/util: Fix memory if LoadRGBMipmaps2 fails.
Michel Dänzer [Tue, 17 Nov 2009 07:05:24 +0000 (08:05 +0100)]
radeon: Depth/stencil span code fixes for big endian.
Fixes e.g. text in progs/demos/arbocclude.
Michel Dänzer [Tue, 17 Nov 2009 07:05:24 +0000 (08:05 +0100)]
radeon: Fix occlusion queries on big endian.
Michel Dänzer [Tue, 17 Nov 2009 07:05:23 +0000 (08:05 +0100)]
radeon: Fix software fallbacks with KMS on big endian.
Michel Dänzer [Tue, 17 Nov 2009 07:05:22 +0000 (08:05 +0100)]
radeon: FBO fixes for big endian.
Michel Dänzer [Tue, 17 Nov 2009 07:05:22 +0000 (08:05 +0100)]
Add MESA_FORMAT_XRGB8888_REV.
Dave Airlie [Tue, 17 Nov 2009 06:06:26 +0000 (16:06 +1000)]
radeon: rn50's have no 3D engine so don't try and init 3D driver.
Vinson Lee [Tue, 17 Nov 2009 02:22:26 +0000 (18:22 -0800)]
progs/util: Fix memory leak if LoadYUVImage fails.
Jakob Bornecrantz [Thu, 12 Nov 2009 00:28:26 +0000 (01:28 +0100)]
svga: Build svga driver
Jakob Bornecrantz [Tue, 17 Nov 2009 01:56:04 +0000 (02:56 +0100)]
svga: Add vmwgfx_drm.h file from vmwgfx kernel driver
Add the vmwgfx_drm.h header for now, this allows the svga driver
to be enabled by default without forcing people to install the
vmwgfx_drm.h header on their system.
To be removed once vmwgfx_drm.h is in libdrm.
Jakob Bornecrantz [Mon, 16 Nov 2009 18:56:18 +0000 (19:56 +0100)]
svga: Add svga driver
Vinson Lee [Tue, 17 Nov 2009 02:06:40 +0000 (18:06 -0800)]
progs/demos: Fix memory leak in ray.c.
Jakob Bornecrantz [Mon, 16 Nov 2009 23:39:09 +0000 (00:39 +0100)]
st/xorg: Fix copy-pasto
Thanks Alan.
Vinson Lee [Tue, 17 Nov 2009 00:31:34 +0000 (16:31 -0800)]
progs/demos: Fix memory leak in projtex.c.
Vinson Lee [Mon, 16 Nov 2009 23:44:52 +0000 (15:44 -0800)]
progs/demos: Fix memory leak in ipers.c.
Ian Romanick [Mon, 16 Nov 2009 23:09:24 +0000 (15:09 -0800)]
i965: Use MESA_FORMAT_AL1616 when appropriate
Ian Romanick [Mon, 16 Nov 2009 22:56:17 +0000 (14:56 -0800)]
Move 'static' to start of declaration to silence compiler warning
Ian Romanick [Mon, 16 Nov 2009 22:53:36 +0000 (14:53 -0800)]
AL1616: Enable MESA_FORMAT_AL1616 for software paths
Ian Romanick [Mon, 16 Nov 2009 22:53:12 +0000 (14:53 -0800)]
AL1616: Add TexImage storage path
Ian Romanick [Mon, 16 Nov 2009 22:52:43 +0000 (14:52 -0800)]
AL1616: Add texel fetch / store routines
Ian Romanick [Mon, 16 Nov 2009 22:49:40 +0000 (14:49 -0800)]
AL1616: Add formats for GL_LUMINANCE16_ALPHA16 textures
Ian Romanick [Mon, 16 Nov 2009 22:30:47 +0000 (14:30 -0800)]
AL1616: Add macros to pack two GLushorts into a texel
Vinson Lee [Mon, 16 Nov 2009 22:56:07 +0000 (14:56 -0800)]
progs/demos: Fix memory leak in fslight.c.
Jakob Bornecrantz [Mon, 16 Nov 2009 21:57:43 +0000 (22:57 +0100)]
st/xorg: Fix comp alpha code and deal with luminance masks
There are two fixes in here one is a one liner that fixes
component alpha logic. The other deals better with luminance
formats used for masks, sources not yet implemented.
Fixes component alpha text and icons in gnome. There are a one
or two cases that this code misses. Like if src_luminance is set
but no mask image is given.
Jakob Bornecrantz [Thu, 12 Nov 2009 10:16:30 +0000 (11:16 +0100)]
st/xorg: Incase of format missmatch swizzle channels and set alpha
This path is only hit for icons in gnome so far
Alex Deucher [Mon, 16 Nov 2009 20:03:48 +0000 (15:03 -0500)]
r600: don't force Z order
Let the hw decide (early vs late Z)
fixes fdo bug 25092
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Corbin Simpson [Mon, 16 Nov 2009 18:46:20 +0000 (10:46 -0800)]
docs: Update r300g status.
Brian Paul [Fri, 13 Nov 2009 06:17:14 +0000 (23:17 -0700)]
docs: fixes, changes for Mesa 7.7
Brian Paul [Fri, 13 Nov 2009 06:06:54 +0000 (23:06 -0700)]
progs/tests: silence warning
Brian Paul [Fri, 13 Nov 2009 06:04:26 +0000 (23:04 -0700)]
mesa: remove unused vertex array driver hooks
Brian Paul [Mon, 16 Nov 2009 15:25:17 +0000 (08:25 -0700)]
mesa: check BaseLevel, MaxLevel in _mesa_GenerateMipmapEXT()
Brian Paul [Mon, 16 Nov 2009 15:21:28 +0000 (08:21 -0700)]
mesa: use _mesa_get_current_tex_object()
Brian Paul [Mon, 16 Nov 2009 15:14:23 +0000 (08:14 -0700)]
mesa: added another check in check_gen_mipmap()
We don't need to call ctx->Driver.GenerateMipmap() if we're updating
a texture level >= MAX_LEVEL.
Michel Dänzer [Mon, 16 Nov 2009 10:59:39 +0000 (11:59 +0100)]
st/xorg: Only reference new picture formats when they're defined.
Fixes http://bugs.freedesktop.org/show_bug.cgi/?id=25094 .
Brian Paul [Fri, 13 Nov 2009 03:59:26 +0000 (20:59 -0700)]
st/mesa: comments for st_draw.c
Brian Paul [Fri, 13 Nov 2009 00:53:54 +0000 (17:53 -0700)]
st/egl: add some basic comments
Francisco Jerez [Sun, 15 Nov 2009 13:49:02 +0000 (14:49 +0100)]
nv20: Fix build for the last nouveau_class.h changes.
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Pekka Paalanen <pq@iki.fi>
Francisco Jerez [Thu, 5 Nov 2009 18:07:19 +0000 (19:07 +0100)]
nv10: Fix build for the last nouveau_class.h changes.
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Pekka Paalanen <pq@iki.fi>
Vinson Lee [Sun, 15 Nov 2009 05:36:18 +0000 (21:36 -0800)]
progs: Ignore Mac OS dSYM directories.
Eric Anholt [Wed, 19 Aug 2009 20:44:13 +0000 (13:44 -0700)]
i965: Share OPCODE_TXB between brw_wm_emit.c and brw_wm_glsl.c
This should fix TXB on G45 and older in the GLSL case.
Eric Anholt [Wed, 19 Aug 2009 21:48:11 +0000 (14:48 -0700)]
i965: Share OPCODE_TEX between brw_wm_emit.c and brw_wm_glsl.c.
New comments should explain some of the confusion about how this message
works.
Eric Anholt [Fri, 13 Nov 2009 23:16:17 +0000 (15:16 -0800)]
i965: Clean up emit_tex a bit.
Eric Anholt [Fri, 13 Nov 2009 22:45:29 +0000 (14:45 -0800)]
Merge remote branch 'origin/mesa_7_6_branch'
Eric Anholt [Fri, 13 Nov 2009 04:08:44 +0000 (20:08 -0800)]
i965: Flag BRW_NEW_CONTEXT on some context state.
Fixing this is a prereq for avoiding flagging all state at new
batch time. Eliminating that still causes problems, though (notably
glean logicOp fails on my GM965).
Eric Anholt [Fri, 13 Nov 2009 03:12:11 +0000 (19:12 -0800)]
intel: Remove some dead context structure fields.
Eric Anholt [Fri, 13 Nov 2009 02:27:12 +0000 (18:27 -0800)]
i965: Remove an unused cache_item field.
Eric Anholt [Fri, 13 Nov 2009 01:48:55 +0000 (17:48 -0800)]
i965: Remove long dead structures for ffvertex_prog.c.
Eric Anholt [Thu, 12 Nov 2009 22:57:30 +0000 (14:57 -0800)]
i965: Use bo_map instead of subdata to upload the bits of constant buffer.
Saves CPU time, resulting in a 2.5% FPS win on ETQW.
Eric Anholt [Thu, 12 Nov 2009 22:05:18 +0000 (14:05 -0800)]
i965: Validate the number of URB entries selected for the VS.
Eric Anholt [Thu, 12 Nov 2009 18:45:05 +0000 (10:45 -0800)]
intel: When subdataing a busy buffer, use a temporary and blit in.
This cuts a massive number of waits in ET:QW, which uses a VBO ringbuffer.
Unfortunately it doesn't BufferData when wrapping back to 0, so we can't
be clever with tracking what's been initialized.
Eric Anholt [Wed, 11 Nov 2009 23:29:34 +0000 (15:29 -0800)]
i965: Clean up Ironlake sampler type definitions.
They're the same regardless of execution width for 8, 4x2, and 16.
Eric Anholt [Wed, 11 Nov 2009 19:58:12 +0000 (11:58 -0800)]
i965: Avoid moving the current value back into the accumulator for MAD.
This is a 2.9% (+/-.3%) performance win for my GL demo, which hits MAD
sequences for matrix transforms.
Eric Anholt [Wed, 11 Nov 2009 21:26:26 +0000 (13:26 -0800)]
mesa: Improve the eliminate-move-use to work across multiple instructions.
This shaves more instructions off of the VS of my GL demo, but no
performance difference this time at n=6. This also fixes a regression
that was in this path, which is now piglit's glsl-vs-mov-after-deref.