mesa.git
14 years agoapple: Change ifdefs for DRI to be DRI && !APPLE
Jeremy Huddleston [Fri, 2 Apr 2010 08:35:19 +0000 (01:35 -0700)]
apple: Change ifdefs for DRI to be DRI && !APPLE

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoapple: Remove duplicate headers that already exist in mesa.
Jeremy Huddleston [Thu, 24 Dec 2009 17:51:17 +0000 (09:51 -0800)]
apple: Remove duplicate headers that already exist in mesa.

Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org>
14 years agoapple: Initial import of libGL for OSX from AppleSGLX svn repository.
Jeremy Huddleston [Thu, 1 Apr 2010 18:01:31 +0000 (11:01 -0700)]
apple: Initial import of libGL for OSX from AppleSGLX svn repository.

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoglapi: Fix loading of old DRI drivers.
Chia-I Wu [Tue, 27 Apr 2010 15:19:27 +0000 (23:19 +0800)]
glapi: Fix loading of old DRI drivers.

The removal of _glapi_noop_enable_warnings and _glapi_set_warning_func
in e4f168a6f4911a096be97d2e83ef8ad9c5862ec0 prevents DRI drivers built
before the commit from loading.  Add stub versions of the functions to
make them load again.

14 years agonv30/40g: fix calls to draw_create(), draw_set_rasterizer_state()
Pedro Maia [Tue, 27 Apr 2010 14:21:49 +0000 (08:21 -0600)]
nv30/40g: fix calls to draw_create(), draw_set_rasterizer_state()

Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agodocs: document updated glext.h and glxext.h files
Brian Paul [Tue, 27 Apr 2010 02:32:24 +0000 (20:32 -0600)]
docs: document updated glext.h and glxext.h files

14 years agoglxext.h: upgraded to 2010/02/10 version
Brian Paul [Tue, 27 Apr 2010 02:32:02 +0000 (20:32 -0600)]
glxext.h: upgraded to 2010/02/10 version

14 years agoglext: upgraded to version 61
Brian Paul [Tue, 27 Apr 2010 02:31:19 +0000 (20:31 -0600)]
glext: upgraded to version 61

14 years agodocs: add links to old mailing list archives
Brian Paul [Fri, 23 Apr 2010 21:36:29 +0000 (15:36 -0600)]
docs: add links to old mailing list archives

14 years agoFix __glXInitializeVisualConfigFromTags's handling of unrecognized fbconfig tags.
Aaron Plattner [Thu, 22 Apr 2010 16:30:42 +0000 (09:30 -0700)]
Fix __glXInitializeVisualConfigFromTags's handling of unrecognized fbconfig tags.

__glXInitializeVisualConfigFromTags doesn't skip the payload of
unrecognized tags.  Instead, it treats the value as if it were the
next tag, which can happen if the server's GLX extension is not
Mesa's.  For example, this falls down when NVIDIA sends a
GLX_FLOAT_COMPONENTS_NV = 0 pair, causing
__glXInitializeVisualConfigFromTags to bail out early.

Signed-off-by: Aaron Plattner <aplattner@nvidia.com>
Signed-off-by: Ian Romanick <ian.d.romanick@intel.com>
14 years agor600: don't enable depth test if there is no depth buffer
Jerome Glisse [Fri, 23 Apr 2010 09:56:06 +0000 (11:56 +0200)]
r600: don't enable depth test if there is no depth buffer

If there is no depth buffer bound to current context don't
enable depth test. GL states that if depth test is enabled
without depth buffer it's as if depth buffer always pass.

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
14 years agoglx/dri2: Fix build with dri2proto which doesn't define X_DRI2SwapInterval.
Michel Dänzer [Fri, 23 Apr 2010 09:14:13 +0000 (11:14 +0200)]
glx/dri2: Fix build with dri2proto which doesn't define X_DRI2SwapInterval.

14 years agoradeon: 9800 SE has only one quadpipe
Tormod Volden [Thu, 22 Apr 2010 19:52:15 +0000 (21:52 +0200)]
radeon: 9800 SE has only one quadpipe

Although these cards have 2 pipelines on the silicon only
the first passed the QA and the other should be disabled.

http://www.digital-daily.com/video/ati-radeon9800se/
http://www.rojakpot.com/showarticle.aspx?artno=101&pgno=1

Signed-off-by: Tormod Volden <debian.tormod@gmail.com>
14 years agost/mesa: add cases for MESA_FORMAT_Z24_X8, MESA_FORMAT_X8_Z24
Brian Paul [Thu, 22 Apr 2010 16:26:20 +0000 (10:26 -0600)]
st/mesa: add cases for MESA_FORMAT_Z24_X8, MESA_FORMAT_X8_Z24

14 years agomesa: fix conversion errors in signed_rgba8888[rev] texel fetch
Brian Paul [Thu, 22 Apr 2010 14:46:46 +0000 (08:46 -0600)]
mesa: fix conversion errors in signed_rgba8888[rev] texel fetch

Without the cast the returned texel colors were wrong.
Also, we don't need the "& 0xff" part anymore.
Bug found by Vinson Lee.

14 years agor300: fix vertex unit setup
Alex Deucher [Thu, 22 Apr 2010 06:33:10 +0000 (02:33 -0400)]
r300: fix vertex unit setup

RV3xx is 2, RV560,RV570 is 8

Noticed by Tormod Volden.

14 years agoprogs/demos: Fix the progs/demos/rain help text
Aaron Plattner [Thu, 22 Apr 2010 00:13:19 +0000 (18:13 -0600)]
progs/demos: Fix the progs/demos/rain help text

Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agoegl: Pass flags to locate Xlib headers and libraries
Dan Nicholson [Sun, 18 Apr 2010 16:33:03 +0000 (09:33 -0700)]
egl: Pass flags to locate Xlib headers and libraries

eglplatform.h pulls in Xlib.h on X11 platforms. Likewise, the egl glx
driver and egl programs needs to link to libX11. Make sure we use the
locations the user told us about.

Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
14 years agoprogs: Fix linking of Xlib demos for non-autoconf builds
Dan Nicholson [Thu, 1 Apr 2010 21:59:10 +0000 (14:59 -0700)]
progs: Fix linking of Xlib demos for non-autoconf builds

The Xlib demos were fixed to use $(X11_LIBS) so that configure could
detect the proper directory to link the library from, but this broke
the non-autoconf builds. Give X11_LIBS a default value to fallback on.
(cherry picked from commit e40fce13e152ecf929c899ed38a8e5f7d198c738)

14 years agoChange libX11 variables to not conflict with AC_PATH_XTRA
Dan Nicholson [Tue, 16 Mar 2010 03:53:56 +0000 (20:53 -0700)]
Change libX11 variables to not conflict with AC_PATH_XTRA

The variable X_LIBS from AC_PATH_XTRA contains only the -L searchdir
parameter and not the -lX11 to link to Xlib. Use X11 prefixed build vars
for linking with Xlib to avoid the conflict.

Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
(cherry picked from commit e725ef171b5a4d5425461f237d9ccab223806913)

14 years agogallium: Respect user's CFLAGS for including X headers
Dan Nicholson [Fri, 2 Apr 2010 13:33:48 +0000 (06:33 -0700)]
gallium: Respect user's CFLAGS for including X headers

This can break on systems that don't have a system X installation.

Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
(cherry picked from commit de4ee20578a79e024b0de83c40648112f42c994e)

Conflicts:

src/gallium/winsys/xlib/Makefile

14 years agoUse X_LIBS from pkg-config, instead of libdir, for locating libX11
Jeff Smith [Sat, 13 Mar 2010 00:55:09 +0000 (18:55 -0600)]
Use X_LIBS from pkg-config, instead of libdir, for locating libX11

Signed-off-by: Jeff Smith <whydoubt@yahoo.com>
Signed-off-by: Dan Nicholson <dbn.lists@gmail.com>
(cherry picked from commit 8d86d395dcf6a5f192b6987485bb7aef49f1fefc)

14 years agoDRI2: check for swapAvailable before using swap interval protocol
Jesse Barnes [Mon, 19 Apr 2010 21:19:48 +0000 (14:19 -0700)]
DRI2: check for swapAvailable before using swap interval protocol

This should have been part of the last change...

14 years agoDRI2: synchronize swap interval with server at startup time
Jesse Barnes [Mon, 19 Apr 2010 16:54:08 +0000 (09:54 -0700)]
DRI2: synchronize swap interval with server at startup time

In the direct rendered case, we need to tell the server our initial swap
interval.  If we don't, the local and server values will be out of sync,
since the server and client defaults may be different (as they were
before this patch).

14 years agost/mesa: invert scissor rect depending on FB orientation
Brian Paul [Mon, 19 Apr 2010 16:10:41 +0000 (10:10 -0600)]
st/mesa: invert scissor rect depending on FB orientation

Fixes fd.o bug 27715

14 years agomesa: Fix build with gcc 3.3.
Matthieu Herrb [Mon, 19 Apr 2010 15:32:37 +0000 (09:32 -0600)]
mesa: Fix build with gcc 3.3.

Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agoradeon: Let this build with gcc 3.3
Matthieu Herrb [Sun, 18 Apr 2010 23:55:38 +0000 (16:55 -0700)]
radeon: Let this build with gcc 3.3

Declaring the loop index inside for () is not supported by this version.

Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agoradeon: Fix command type for DRM_RADEON_IRQ_EMIT ioctl.
Owain Ainsworth [Sun, 18 Apr 2010 23:55:37 +0000 (16:55 -0700)]
radeon: Fix command type for DRM_RADEON_IRQ_EMIT ioctl.

This should be drmCommandWriteRead to avoid an EINVAL error on systems
that strictly check ioctl args. This command has been r/w for ever.
Discussion with airlied agreed that this was the correct course.

Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agomesa: Use __OpenBSD__ to check for OpenBSD.
Matthieu Herrb [Sun, 18 Apr 2010 23:55:36 +0000 (16:55 -0700)]
mesa: Use __OpenBSD__ to check for OpenBSD.

Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agogallium/draw: use correct rasterization state for wide/AA points/lines
Brian Paul [Mon, 19 Apr 2010 14:35:53 +0000 (08:35 -0600)]
gallium/draw: use correct rasterization state for wide/AA points/lines

When points or lines are decomposed into triangles, we need to be sure
to disable polygon culling, stippling, "un-filled" modes, etc.

This patch sets the rasterization state to disable those things prior to
drawing points/lines with triangles, then restores the previous state
afterward.

The new piglit point-no-line-cull test checks this problem & solution.

14 years agodocs: update links for new mesa-users and mesa-announce lists
Brian Paul [Fri, 16 Apr 2010 14:25:58 +0000 (08:25 -0600)]
docs: update links for new mesa-users and mesa-announce lists

14 years agosavage: call _mesa_meta_init/free()
Tormod Volden [Thu, 15 Apr 2010 17:13:53 +0000 (10:13 -0700)]
savage: call _mesa_meta_init/free()

Fixes crash when using _mesa_CopyTexImage2D.
Bug #27652.

Signed-off-by: Tormod Volden <debian.tormod@gmail.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agosoftpipe: Fix division by zero
Arpad Borsos [Wed, 14 Apr 2010 20:26:32 +0000 (14:26 -0600)]
softpipe: Fix division by zero

This can be triggered by running the cairo tests using the gl backend
and softpipe.

Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agoInitialize DRI2 swap interval to 0
Michael Schmidt [Wed, 14 Apr 2010 14:12:42 +0000 (10:12 -0400)]
Initialize DRI2 swap interval to 0

https://bugs.freedesktop.org/show_bug.cgi?id=27628

14 years agoDRI2: report swap events correctly in direct rendered case
Jesse Barnes [Tue, 13 Apr 2010 16:38:41 +0000 (09:38 -0700)]
DRI2: report swap events correctly in direct rendered case

In the direct rendered case, we need to convert DRI2 swap complete
events to GLX events for the client to consume.  This path had what
looks like a stray "& 0x75" from some earlier debugging that prevented
clients from seeing the right event code.

14 years agodocs: update name, link for mesa-dev list
Brian Paul [Tue, 13 Apr 2010 00:45:10 +0000 (18:45 -0600)]
docs: update name, link for mesa-dev list

14 years agoi915g: Fix assert.
Vinson Lee [Mon, 12 Apr 2010 00:56:21 +0000 (17:56 -0700)]
i915g: Fix assert.

14 years agor300: respect radeon common code fallbacks
Maciej Cencora [Sun, 11 Apr 2010 10:39:54 +0000 (12:39 +0200)]
r300: respect radeon common code fallbacks

Fixes progs/demos/shadowtex under KMS

14 years agor300: set proper vertex index limits also in non indexed mode
Maciej Cencora [Fri, 9 Apr 2010 19:14:15 +0000 (21:14 +0200)]
r300: set proper vertex index limits also in non indexed mode

Fixes #27521, broken menus in UT2004 and broken water refraction in Sauerbraten.

14 years agor600: add new r7xx pci ids
Alex Deucher [Sat, 10 Apr 2010 00:20:42 +0000 (20:20 -0400)]
r600: add new r7xx pci ids

14 years agoglx: Fix config chooser logic for 'mask' matching
Kristian Høgsberg [Fri, 9 Apr 2010 02:09:11 +0000 (22:09 -0400)]
glx: Fix config chooser logic for 'mask' matching

When matching attributes using the 'mask' matching criteria, the spec
says that

  "Only GLXFBConfigs for which the set bits of attribute include all
   the bits that are set in the requested value are
   considered. (Additional bits might be set in the attribute)."

The current test returns true if the two bit masks have bits in
common, specifically it matches even if the requested value has bits
set that are not set in the fbconfig attribute.  For example, an
application asking for

  GLX_DRAWABLE_TYPE, GLX_PIXMAP_BIT | GLX_PBUFFER_BIT,

as glxpbdemo does, will match fbconfigs that don't support pbuffer
rendering, as long as they support pixmap rendering.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
14 years agomesa: fix instruction indexing bugs
Brian Paul [Fri, 9 Apr 2010 16:03:31 +0000 (10:03 -0600)]
mesa: fix instruction indexing bugs

We were looping over instructions but only looking at the 0th instruction's
opcode.

Fixes fd.o bug 27566.

14 years agodocs: link to 7.8.2 release notes
Brian Paul [Thu, 8 Apr 2010 19:11:05 +0000 (13:11 -0600)]
docs: link to 7.8.2 release notes

14 years agodocs: initial 7.8.2 release notes
Brian Paul [Thu, 8 Apr 2010 19:10:51 +0000 (13:10 -0600)]
docs: initial 7.8.2 release notes

14 years agoprogs/tests: added some debug code (disabled)
Brian Paul [Thu, 8 Apr 2010 19:09:01 +0000 (13:09 -0600)]
progs/tests: added some debug code (disabled)

14 years agost/mesa: fix glDrawPixels(GL_DEPTH_COMPONENT) regression
Brian Paul [Thu, 8 Apr 2010 19:05:35 +0000 (13:05 -0600)]
st/mesa: fix glDrawPixels(GL_DEPTH_COMPONENT) regression

Commit 1454f20a991ddda35f1a2ffda953012078b407ba caused the regression.
When the vertex shader emitted both a texcoord and color we were grabbing
the wrong vertex attributes.  Fix the draw_quad() code to put texcoords
in slot[1] and color in slot[2].  That's a bit cleaner than changing
the vertex shader code.

Tested with progs/tests/zreaddraw.c

14 years agointel: Call intel_prepare_render() in intelClear()
Li Peng [Thu, 8 Apr 2010 17:52:55 +0000 (13:52 -0400)]
intel: Call intel_prepare_render() in intelClear()

Make sure we have up to date buffers before we start looking at
the tiling bits to determine how to clear.

14 years agotexenvprogram: fix for ARB_draw_buffers.
Dave Airlie [Sat, 3 Apr 2010 03:34:29 +0000 (13:34 +1000)]
texenvprogram: fix for ARB_draw_buffers.

piglit has a test called fbo-drawbuffers, this fails for me on r300g,
and fixing the texenv program to use the DATA outputs fixes it.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agodocs: Update 7.8.1 release MD5 sums
Ian Romanick [Mon, 5 Apr 2010 19:45:20 +0000 (12:45 -0700)]
docs: Update 7.8.1 release MD5 sums

14 years agomesa: bump MESA_TINY version
Brian Paul [Mon, 5 Apr 2010 17:22:32 +0000 (11:22 -0600)]
mesa: bump MESA_TINY version

14 years agodocs: added news item for 7.8.1 release
Brian Paul [Mon, 5 Apr 2010 17:12:59 +0000 (11:12 -0600)]
docs: added news item for 7.8.1 release

14 years agodocs: add link to 7.8.1 release notes
Brian Paul [Mon, 5 Apr 2010 17:10:51 +0000 (11:10 -0600)]
docs: add link to 7.8.1 release notes

14 years agosoftpipe: index the correct blend/mask state index
Brian Paul [Mon, 5 Apr 2010 17:00:52 +0000 (11:00 -0600)]
softpipe: index the correct blend/mask state index

Need to check pipe_blend_state::independent_blend_enable to determine
which render target/index to use when checking blend enable and colormask
state.

This is part of the fix for piglit/fbo-drawbuffers

14 years agodocs: Add 7.8.1 release MD5 sums
Ian Romanick [Mon, 5 Apr 2010 16:51:18 +0000 (09:51 -0700)]
docs: Add 7.8.1 release MD5 sums

14 years agomesa: set version string to 7.8.1
Ian Romanick [Mon, 5 Apr 2010 16:49:08 +0000 (09:49 -0700)]
mesa: set version string to 7.8.1

14 years agoInitial 7.8.1 release notes
Ian Romanick [Mon, 5 Apr 2010 16:12:33 +0000 (09:12 -0700)]
Initial 7.8.1 release notes

14 years agomesa: update_arrays() depends on program state.
Henri Verbeet [Sun, 4 Apr 2010 17:24:46 +0000 (10:24 -0700)]
mesa: update_arrays() depends on program state.

It uses ctx->VertexProgram._Current.

Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agoUpdate to final names from GLX_INTEL_swap_event spec
Ian Romanick [Mon, 5 Apr 2010 06:06:01 +0000 (23:06 -0700)]
Update to final names from GLX_INTEL_swap_event spec

Fixes bug #27454.

14 years agogl: updated glxext.h to version 27
Ian Romanick [Sun, 4 Apr 2010 23:53:41 +0000 (16:53 -0700)]
gl: updated glxext.h to version 27

14 years agoutil: Use GCC atomic bultins on GCC 4.1 and higher only.
Vinson Lee [Sat, 3 Apr 2010 19:14:21 +0000 (12:14 -0700)]
util: Use GCC atomic bultins on GCC 4.1 and higher only.

14 years agoxdemos: Build object files first
Jeremy Huddleston [Thu, 1 Apr 2010 19:28:41 +0000 (12:28 -0700)]
xdemos: Build object files first

This helps debugging on darwin.

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoglx: Fix compilation errors on non-GLX_DIRECT_RENDERING builds.
Vinson Lee [Fri, 2 Apr 2010 07:12:59 +0000 (00:12 -0700)]
glx: Fix compilation errors on non-GLX_DIRECT_RENDERING builds.

Add ifdef guards around variables of types defined only for
GLX_DIRECT_RENDERING.

14 years agost/vega: Do not depend on libm.
Chia-I Wu [Fri, 2 Apr 2010 02:27:57 +0000 (10:27 +0800)]
st/vega: Do not depend on libm.

This fixes the "no rule to make target -lm" error on darwin, reported by
Jeremy Huddleston.

14 years agoGLX/DRI2: pass GLX drawable ID to dri2InvalidateBuffers
Jesse Barnes [Fri, 2 Apr 2010 00:26:04 +0000 (17:26 -0700)]
GLX/DRI2: pass GLX drawable ID to dri2InvalidateBuffers

The IDs will be the same in the case where an X window is used directly
as a GLX drawable, but will fail if a new GLX drawable is created
explicitly, as with glxgears_fbconfig.

Fixes fdo bug #27190.

14 years agoRevert accidental commits from the xquartz tree
Jeremy Huddleston [Fri, 2 Apr 2010 00:00:22 +0000 (17:00 -0700)]
Revert accidental commits from the xquartz tree

This reverts commit 9aadc793f3db64cefa0b08f18abad424a659dacc.
This reverts commit 69ea4e7718efb60b6b0d795a355cebd6712ceac1.
This reverts commit dbe8b013936d977ec63d6607bfd2fc6772d29787.
This reverts commit 23215ef4d60a86d9f3b3fdc08e3fdadc59e98890.
This reverts commit 9495e3703062d1ddaf3161f4efc23f0b51284d9b.
This reverts commit 0594cf70883b64692ba617d85f4f9b4e636e5c2b.
This reverts commit 86a7978d37393ee34f876569ac06ffdb8d7289ae.
This reverts commit 437902ce978cde9a0e1aa260f12dc232a8501c42.

14 years agoapple: Don't build gallium
Jeremy Huddleston [Thu, 1 Apr 2010 19:14:08 +0000 (12:14 -0700)]
apple: Don't build gallium

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoapple: Use mesa gl.h rather than generating one.
Jeremy Huddleston [Sat, 2 Jan 2010 05:03:53 +0000 (00:03 -0500)]
apple: Use mesa gl.h rather than generating one.

Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org>
14 years agoapple: Integrate our libGL into the existing build system better
Jeremy Huddleston [Thu, 1 Apr 2010 18:18:42 +0000 (11:18 -0700)]
apple: Integrate our libGL into the existing build system better

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoapple: Change ifdefs for DRI to be DRI && !APPLE
Jeremy Huddleston [Thu, 1 Apr 2010 18:17:04 +0000 (11:17 -0700)]
apple: Change ifdefs for DRI to be DRI && !APPLE

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoapple: Remove duplicate headers that already exist in mesa.
Jeremy Huddleston [Thu, 24 Dec 2009 17:51:17 +0000 (09:51 -0800)]
apple: Remove duplicate headers that already exist in mesa.

Signed-off-by: Jeremy Huddleston <jeremyhu@freedesktop.org>
14 years agoapple: Initial import of libGL for OSX from AppleSGLX svn repository.
Jeremy Huddleston [Thu, 1 Apr 2010 18:01:31 +0000 (11:01 -0700)]
apple: Initial import of libGL for OSX from AppleSGLX svn repository.

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoxdemos: Build object files first
Jeremy Huddleston [Thu, 1 Apr 2010 19:28:41 +0000 (12:28 -0700)]
xdemos: Build object files first

This helps debugging on darwin.

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoxdemos: Fix a build failure for non-autoconf configs
Jeremy Huddleston [Thu, 1 Apr 2010 18:51:15 +0000 (11:51 -0700)]
xdemos: Fix a build failure for non-autoconf configs

Signed-off-by: Jeremy Huddleston <jeremyhu@apple.com>
14 years agoglx: Fix compilation errors on non-GLX_DIRECT_RENDERING builds.
Vinson Lee [Thu, 1 Apr 2010 07:10:28 +0000 (00:10 -0700)]
glx: Fix compilation errors on non-GLX_DIRECT_RENDERING builds.

driContext field for __GLXcontextRec struct is only defined if
GLX_DIRECT_RENDERING is set.

14 years agost/mesa: use BITFIELD64_BIT to access shader OutputsWritten in more places
Ben Skeggs [Wed, 31 Mar 2010 05:44:25 +0000 (15:44 +1000)]
st/mesa: use BITFIELD64_BIT to access shader OutputsWritten in more places

14 years agoglx/dri2: Fix debug build with old dri2proto.
Michel Dänzer [Tue, 30 Mar 2010 09:50:13 +0000 (11:50 +0200)]
glx/dri2: Fix debug build with old dri2proto.

Apparently the higher compiler optimization level in non-debug builds was
eliminating the unused functions referencing the unresolved DRI2 symbols...

14 years agodocs: fix 7.7.1 release date
Brian Paul [Mon, 29 Mar 2010 18:01:28 +0000 (12:01 -0600)]
docs: fix 7.7.1 release date

(cherry picked from commit e6f5ca0fa3bff975f2acb3a825d77f095bc9f43e)

14 years agomesa: Add 7.8 release MD5 sums
Ian Romanick [Sun, 28 Mar 2010 23:35:56 +0000 (16:35 -0700)]
mesa: Add 7.8 release MD5 sums

14 years agointel: Bump intel driver date to reflect status as 2010Q1 release
Ian Romanick [Sun, 28 Mar 2010 23:33:53 +0000 (16:33 -0700)]
intel: Bump intel driver date to reflect status as 2010Q1 release

14 years agomesa: set version string to 7.8
Ian Romanick [Sun, 28 Mar 2010 23:31:08 +0000 (16:31 -0700)]
mesa: set version string to 7.8

Also set the correct release date.

14 years agomesa: Add 7.7.1 release MD5 sums
Ian Romanick [Sun, 28 Mar 2010 23:25:47 +0000 (16:25 -0700)]
mesa: Add 7.7.1 release MD5 sums
(cherry picked from commit 0c88e340499c961cc7a06107a727710a67e280ed)

14 years agomesa: set version string to 7.7.1
Ian Romanick [Sun, 28 Mar 2010 23:22:44 +0000 (16:22 -0700)]
mesa: set version string to 7.7.1

Also set the correct release date.
(cherry picked from commit 663642b435af2f8ab4b16360783eb367f42486c3)

14 years agoprogs/tests: Remove duplicate texcompress2 in SConscript.
Vinson Lee [Sat, 27 Mar 2010 21:40:41 +0000 (14:40 -0700)]
progs/tests: Remove duplicate texcompress2 in SConscript.

14 years agoi965g: Add brw_winsys_debug.c to SCons build.
Vinson Lee [Sat, 27 Mar 2010 21:25:03 +0000 (14:25 -0700)]
i965g: Add brw_winsys_debug.c to SCons build.

14 years agoidentity: Add id_drm.c to SCons build.
Vinson Lee [Sat, 27 Mar 2010 17:07:14 +0000 (10:07 -0700)]
identity: Add id_drm.c to SCons build.

This was missed in commit f7cbaae13d67c55abe81ac230de37f564365099f.
(cherry picked from commit 02ee7c29502966dffa44243bfc8c20c15907b880)

14 years agomesa: move/update hash function comments
Brian Paul [Sat, 27 Mar 2010 14:59:17 +0000 (08:59 -0600)]
mesa: move/update hash function comments

(cherry picked from commit 535742d75f0096b22d1b8ff203ae561167af18f7)

14 years agomesa: fix deadlock in _mesa_HashFindFreeKeyBlock()
Brian Paul [Sat, 27 Mar 2010 14:58:59 +0000 (08:58 -0600)]
mesa: fix deadlock in _mesa_HashFindFreeKeyBlock()

Fixes fd.o bug 27340.
(cherry picked from commit 8fe3b3f66ae57a1a6eca7f6dcb0455e14ad92075)

14 years agodri/swrast: Fix frontbuffer rendering.
Michel Dänzer [Fri, 26 Mar 2010 15:29:59 +0000 (16:29 +0100)]
dri/swrast: Fix frontbuffer rendering.

Was broken since the endianness fixes.
(cherry picked from commit 4cf14fa80bda5f4ea65bef3a64e748e064d0bde1)

14 years agodri/swrast: Fix missed conversion of one pixel pointer increment.
Michel Dänzer [Fri, 26 Mar 2010 15:29:59 +0000 (16:29 +0100)]
dri/swrast: Fix missed conversion of one pixel pointer increment.

This probably broke the swrast DRI driver when running X in depth 16.
(cherry picked from commit 6ec259eb17dfbb74972b8cffb4e02a9dbab288cc)

14 years agoi965: Fix readpixels from ReadBuffer != DrawBuffer.
Eric Anholt [Tue, 16 Mar 2010 20:23:23 +0000 (13:23 -0700)]
i965: Fix readpixels from ReadBuffer != DrawBuffer.

Fixes piglit fbo-readdrawpix.
(cherry picked from commit 5782b2a968bb979b651e49bb5fc4162faa842050)

14 years agoi965: Fix inversion for glCopyPixels to/from FBOs.
Eric Anholt [Wed, 24 Mar 2010 21:28:45 +0000 (14:28 -0700)]
i965: Fix inversion for glCopyPixels to/from FBOs.

fixes piglit fbo-copypix.
(cherry picked from commit a589da14dee0c2a32e6e529f1a390b01a3ee4001)

14 years agointel: Rely on allocated region pitch for the miptree pitch.
Eric Anholt [Tue, 16 Mar 2010 23:20:03 +0000 (16:20 -0700)]
intel: Rely on allocated region pitch for the miptree pitch.

Bug #26966: 945 miptree pitch disagreement with libdrm.
(cherry picked from commit da011faf48155a5c02ebc1fe1fa20a4f54b8c657)

14 years agointel: Remove extra tiling setting after allocating a tiled region.
Eric Anholt [Tue, 16 Mar 2010 23:05:53 +0000 (16:05 -0700)]
intel: Remove extra tiling setting after allocating a tiled region.
(cherry picked from commit 32f143b4327521a058dc05f0ab9087a5696b9618)

14 years agointel: Respect src pitch in _mesa_copy_rect().
Eric Anholt [Wed, 17 Mar 2010 17:10:37 +0000 (10:10 -0700)]
intel: Respect src pitch in _mesa_copy_rect().

If a non-zero src_y was used, this would break piglit
depth-level-clamp.
(cherry picked from commit e1e48ea15c1fe448f0b69e086b66c1123dc98bb7)

14 years agoRegenerate gl_mangle.h
Tom Fogal [Thu, 25 Mar 2010 22:48:59 +0000 (16:48 -0600)]
Regenerate gl_mangle.h

14 years agoglapi: return int = 0 from NoOpGeneric()
Brian Paul [Thu, 25 Mar 2010 22:03:13 +0000 (16:03 -0600)]
glapi: return int = 0 from NoOpGeneric()

If a GL function is called w/out a current rendering context, this stub
gets called.  It should return 0 so that non-void-valued functions return
0/NULL instead of a random number.

14 years agoglslcompiler: Fix build.
Vinson Lee [Thu, 25 Mar 2010 19:58:36 +0000 (12:58 -0700)]
glslcompiler: Fix build.

14 years agor300: report correct state atom size
Maciej Cencora [Sun, 21 Mar 2010 11:43:38 +0000 (12:43 +0100)]
r300: report correct state atom size

Spotted by Pauli Nieminen

14 years agor300: fix off by one
Maciej Cencora [Sun, 21 Mar 2010 11:12:05 +0000 (12:12 +0100)]
r300: fix off by one

R300_PVS_MAX_CONST_ADDR field holds highest const addr, not
const count.

Fixes missing models and others rendering errors for vertex
program using 256 params.

14 years agor300: fix vertex programs with big number of params (>255) under KMS
Maciej Cencora [Sun, 21 Mar 2010 11:10:06 +0000 (12:10 +0100)]
r300: fix vertex programs with big number of params (>255) under KMS

UMS will probably require some kernel work