mesa.git
14 years agodri: make unresolved symbol test link work even without a libGL.so
Luca Barbieri [Tue, 23 Mar 2010 20:10:07 +0000 (21:10 +0100)]
dri: make unresolved symbol test link work even without a libGL.so

Currently the test link uses -lGL to define the glapi symbols.

This makes it impossible to build DRI drivers on systems without
Mesa installed and without building the libGL from the Mesa tree
first.

Some automated build systems trigger this problem.

This commit removes -lGL and instead adds a dummy implementation of
glapi to dri_test.c

This, along with Kristian's commit, should fix all known regressions
due to the addition of unresolved symbol checking.

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>
14 years agodri: Add $(DRI_LIB_DEPS) when linking the test binary
Kristian Høgsberg [Tue, 23 Mar 2010 19:47:49 +0000 (15:47 -0400)]
dri: Add $(DRI_LIB_DEPS) when linking the test binary

This will make sure we pick up libdrm_$(chipset).so from the right place.

14 years agoi965: Stop abusing ctx->NewState flags for storing driver internal changes.
Eric Anholt [Wed, 24 Mar 2010 03:52:06 +0000 (20:52 -0700)]
i965: Stop abusing ctx->NewState flags for storing driver internal changes.

We're still abusing the flags by putting them where our driver stores
ctx->NewState changes.  Making them into more restricted state change
flags would be a project for later.

Fixes a failure where calling intel_draw_buffer() too often would trip
up Mesa assertions about when Mesa state could get changed, when it hadn't.

Bug #27034.

14 years agomesa: Also print _NEW_STENCIL in _mesa_print_state()
Kristian Høgsberg [Tue, 23 Mar 2010 16:07:45 +0000 (12:07 -0400)]
mesa: Also print _NEW_STENCIL in _mesa_print_state()

14 years agonvfx: add NOUVEAU_VTXIDX_IN_VRAM variable to put vertex/index buffers in VRAM
Luca Barbieri [Mon, 11 Jan 2010 02:13:42 +0000 (03:13 +0100)]
nvfx: add NOUVEAU_VTXIDX_IN_VRAM variable to put vertex/index buffers in VRAM

On some systems, putting vertex and index buffers in VRAM instead of GART
memory eliminates massive graphics corruption which is otherwise present,
due to unclear causes.

This patch adds an environment variable that does that, along with helpful
messages.

It turns it on by default on G7x, as it is what I am seeing corruption
on and some other reports also seemed to pinpoint these cards.

14 years agonvfx: delay allocation of buffers in GART/VRAM to validation time
Luca Barbieri [Tue, 23 Mar 2010 15:34:51 +0000 (16:34 +0100)]
nvfx: delay allocation of buffers in GART/VRAM to validation time

Currently we allocate buffers in GART or VRAM at creation time.

However, when using swtnl, this results in reads from uncached
memory, which drastically impair performance.

So, for now, cause nouveau_screen.c to not pass any placement flags
to buffer creation, so that the buffers are moved later.

Previously libdrm itself did this, but was changed to not to do it.

This may introduce an extra copy in normal usage, but this currently
does not seem to introduce significant performance degradation.

This will be revisited when pipebuffer is integrated.

Note that for AGP systems, properly solving this may be complex
since currently there is no fast way of reading from GART/VRAM.

We will probably need to try mapping AGP as writethrough and, in
addition, make buffer creation more aware of future buffer usage.

14 years agost/mesa: make st_manager.c set have[Stencil|Depth]Buffer only if bits > 0
Luca Barbieri [Tue, 23 Mar 2010 16:40:42 +0000 (17:40 +0100)]
st/mesa: make st_manager.c set have[Stencil|Depth]Buffer only if bits > 0

Fixes a segfault when clearing a non-existent stencil buffer.

14 years agodri: use mv -f instead of mv to ensure no prompting occurs
Luca Barbieri [Tue, 23 Mar 2010 16:55:38 +0000 (17:55 +0100)]
dri: use mv -f instead of mv to ensure no prompting occurs

Using just mv may cause prompts on some systems/configurations.

14 years agoMerge branch '7.8'
Brian Paul [Tue, 23 Mar 2010 15:14:17 +0000 (09:14 -0600)]
Merge branch '7.8'

Conflicts:

Makefile
src/gallium/state_trackers/glx/xlib/xm_api.c
src/mesa/drivers/dri/intel/intel_blit.c
src/mesa/drivers/dri/intel/intel_pixel_copy.c
src/mesa/main/version.h

14 years agosoftpipe: comments, re-formatting, etc
Brian Paul [Tue, 23 Mar 2010 15:08:35 +0000 (09:08 -0600)]
softpipe: comments, re-formatting, etc

14 years agosoftpipe: add special cases for all Z compare modes for 16-bit Z buffer
Brian Paul [Tue, 23 Mar 2010 15:00:19 +0000 (09:00 -0600)]
softpipe: add special cases for all Z compare modes for 16-bit Z buffer

We had fast paths for PIPE_FUNC_LESS and LEQUAL before.  To satisfy
OpenGL invariance rules, all depth compare modes should produce the
same fragment Z values.

Fixes progs/demos/singlebuffer.c

14 years agost/glx: better format selection in xmesa_choose_z_stencil_format()
Brian Paul [Tue, 23 Mar 2010 14:58:09 +0000 (08:58 -0600)]
st/glx: better format selection in xmesa_choose_z_stencil_format()

This is a back-port of commit ef2664da6c4db1b52ef351641e3ee949b87f9c7b
from master.

14 years agonvfx: don't crash on empty fragment program
Luca Barbieri [Sun, 17 Jan 2010 21:27:53 +0000 (22:27 +0100)]
nvfx: don't crash on empty fragment program

14 years agonvfx: fix coding style in nvfx_transfer.c
Luca Barbieri [Wed, 17 Mar 2010 21:31:29 +0000 (22:31 +0100)]
nvfx: fix coding style in nvfx_transfer.c

14 years agonvfx: stop incessantly spewing debug messages on the terminal
Luca Barbieri [Wed, 17 Mar 2010 19:31:56 +0000 (20:31 +0100)]
nvfx: stop incessantly spewing debug messages on the terminal

Currently we are continuously spewing messages about these variables
since we call debug_get_bool_option everytime we want to check their value.

This is annoying, slows things down due to terminal rerendering
and obscures useful messages.

This patch only calls debug_get_bool_option once and caches the result in a
static variable.

14 years agonvfx: fix/workaround nv3x hwtnl issues
Luca Barbieri [Wed, 17 Mar 2010 14:44:45 +0000 (15:44 +0100)]
nvfx: fix/workaround nv3x hwtnl issues

This patch re-emits the viewport state on framebuffer or rasterizer
change.

This seems to be necessary on nv3x, but the reason is not fully
understood.

It is quite likely that this isn't really the correct fix, but seems
to work, and makes nv3x much better.

14 years agodri: test whether the built drivers have unresolved symbols
Luca Barbieri [Mon, 15 Mar 2010 17:36:27 +0000 (18:36 +0100)]
dri: test whether the built drivers have unresolved symbols

This is a different approach to solving this problem that the patch
I previously posted, and unlike that, should not cause any problems.

Right now undefined symbols in DRI drivers will still allow the
build to succeed.

As a result, people modifying drivers they cannot test risk creating
unloadable drivers with no easy way of automatically avoiding it.

For instance, the modifications to nv50 for context transfers caused
such an issue recently.

Unfortunately, just adding -Wl,--no-undefined doesn't work, because
the DRI drivers depend on glapi symbols, but do not depend on
libGL.so.1

Adding -lGL is not the correct solution since DRI drivers are not loaded
just by libGL, but also by X and possibly by other clients.

So, this patch simply tries to build an executable linked to the DRI
driver and to libGL.
If the DRI driver contains any undefined symbols not satisfied by its
dependencies or by libGL, this will fail.

This solution does not alter the built drivers, and does not significantly
slow down the build process.

All classic DRI drivers as well as all the Gallium drivers with configure
options compiled successfully with this change.

Thanks to Xavier Chantry <chantry.xavier@gmail.com> and
Michel Daenzer <michel@daenzer.net> for helping with this.

Signed-off-by: Luca Barbieri <luca@luca-barbieri.com>
Acked-by: Brian Paul <brian.e.paul@gmail.com>
14 years agoradeong: fix build issue with libdrm < 2.4.19
Marek Olšák [Tue, 23 Mar 2010 12:33:16 +0000 (13:33 +0100)]
radeong: fix build issue with libdrm < 2.4.19

14 years agor300g: fix glean occlusion query test
Dave Airlie [Mon, 22 Mar 2010 03:35:20 +0000 (13:35 +1000)]
r300g: fix glean occlusion query test

14 years agoAdd missing EGL files to the tarballs.
Chia-I Wu [Tue, 23 Mar 2010 06:37:24 +0000 (14:37 +0800)]
Add missing EGL files to the tarballs.

Add the Makefile of Gallium EGL drivers and demos using EGL to the
tarballs.

14 years agoglx: Suppress BadDrawable from DRI2CopyRegion
Kristian Høgsberg [Tue, 23 Mar 2010 01:57:26 +0000 (21:57 -0400)]
glx: Suppress BadDrawable from DRI2CopyRegion

This can happen when an X window is destroyed behind our back.  We use
DRI2CopyRegion behind the scenes in many places (like flushing the fake
front to the real front) so we have to ignore X errors triggered in that
case.

The glean test cases trigger this consistently as they don't destroy the
GLX drawable nicely, they just destroy the X window.

14 years agoglx: Suppress BadDrawable from DRI2CopyRegion
Kristian Høgsberg [Tue, 23 Mar 2010 01:57:26 +0000 (21:57 -0400)]
glx: Suppress BadDrawable from DRI2CopyRegion

This can happen when an X window is destroyed behind our back.  We use
DRI2CopyRegion behind the scenes in many places (like flushing the fake
front to the real front) so we have to ignore X errors triggered in that
case.

The glean test cases trigger this consistently as they don't destroy the
GLX drawable nicely, they just destroy the X window.

14 years agogallium/docs: Cleanup and clarify point ras info.
Corbin Simpson [Tue, 23 Mar 2010 01:12:16 +0000 (18:12 -0700)]
gallium/docs: Cleanup and clarify point ras info.

...This state's interdependent? Really? Needs moar cleanup.

14 years agoradeong: Always initialize this variable.
Corbin Simpson [Tue, 23 Mar 2010 01:11:40 +0000 (18:11 -0700)]
radeong: Always initialize this variable.

May or may not be responsible for slight increases in ipers FPS.

14 years agor300g: Cleanup a few old warnings.
Corbin Simpson [Mon, 22 Mar 2010 10:10:19 +0000 (03:10 -0700)]
r300g: Cleanup a few old warnings.

Those paths aren't important anymore, and a debugging warning won't stop
a hardlock anyway.

14 years agomesa: set version string to 7.8-rc2
Ian Romanick [Tue, 23 Mar 2010 01:09:15 +0000 (18:09 -0700)]
mesa: set version string to 7.8-rc2

14 years agoGLX/OML: honor OML semantics even if target, divisor and remainder are 0
Jesse Barnes [Mon, 22 Mar 2010 23:39:11 +0000 (16:39 -0700)]
GLX/OML: honor OML semantics even if target, divisor and remainder are 0

This change passes a remainder of 1 to the server with the
DRI2SwapBuffers request, causing it to honor the OML semantics for the
swap rather than falling through to glXSwapBuffers behavior.  The
remainder actually ends up ignored since the divisor is 0, but we need
to differentiate the OML and standard behavior somehow.

Reported-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de>
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
14 years agoi965: Add disasm for SNB MATH opcode.
Eric Anholt [Mon, 22 Mar 2010 17:17:19 +0000 (10:17 -0700)]
i965: Add disasm for SNB MATH opcode.

14 years agoi965: Enable normal clipping on SNB.
Eric Anholt [Mon, 22 Mar 2010 16:04:25 +0000 (09:04 -0700)]
i965: Enable normal clipping on SNB.

Rejecting all doesn't seem to be helping get the pipeline lit up.

14 years agoi965: Force single program flow in SNB GS, to match gen4 GS.
Eric Anholt [Sat, 20 Mar 2010 00:14:12 +0000 (17:14 -0700)]
i965: Force single program flow in SNB GS, to match gen4 GS.

14 years agoi965: Correct copy and wasted field shifts for SNB GS URB.
Eric Anholt [Sat, 20 Mar 2010 00:08:42 +0000 (17:08 -0700)]
i965: Correct copy and wasted field shifts for SNB GS URB.

14 years agoi965: Enable VS on SNB.
Eric Anholt [Fri, 19 Mar 2010 23:00:31 +0000 (16:00 -0700)]
i965: Enable VS on SNB.

It appears that the thing that was killing VS threads was the
gratuitous NOP that replaced the gratuitous jump from OPCODE_END to
the nearby OPCODE_END implementation.  With that gone, we can move on
to the rest of the pipeline.

14 years agoi965: Remove gratuitous jump or nop from OPCODE_END to vertex emit.
Eric Anholt [Fri, 19 Mar 2010 22:57:50 +0000 (15:57 -0700)]
i965: Remove gratuitous jump or nop from OPCODE_END to vertex emit.

Just emit the URB write at END time.  Subroutine code that sits after
OPCODE_END won't be executed since we've ended the thread at the point
that the URB write is done.

14 years agoi965: Ignore execution mask for the mov(m0, g0) of VS URB write header on SNB.
Eric Anholt [Fri, 19 Mar 2010 22:04:17 +0000 (15:04 -0700)]
i965: Ignore execution mask for the mov(m0, g0) of VS URB write header on SNB.

Otherwise, we may not get the FFTID set up which would break freeing
of resources.

14 years agoi965: Allow FS constants to be used as immediates instead of push/pull.
Eric Anholt [Fri, 19 Mar 2010 19:34:53 +0000 (12:34 -0700)]
i965: Allow FS constants to be used as immediates instead of push/pull.

The hope is to later take advantage of the reduced constant usage to
free up regs.  This only covers the GLSL path at the moment, because
the brw_wm_emit path doesn't get the information as to whether a float
value is a constant or a uniform.

14 years agoi965: Add INTEL_DEBUG=glsl_force to force brw_wm_glsl.c.
Eric Anholt [Fri, 19 Mar 2010 19:27:38 +0000 (12:27 -0700)]
i965: Add INTEL_DEBUG=glsl_force to force brw_wm_glsl.c.

I keep finding the desire to force this path to debug it instead of
cooking up goofy-looking testcases to do so.

14 years agoi965: Optimize OPCODE_CMP by using BRW_SEL to choose results.
Eric Anholt [Fri, 19 Mar 2010 18:44:43 +0000 (11:44 -0700)]
i965: Optimize OPCODE_CMP by using BRW_SEL to choose results.

Tested with piglit glsl-fs-sqrt-branch, fp-cmp.vpfp.

14 years agost/glx: fix depth/stencil format selection code
Brian Paul [Mon, 22 Mar 2010 21:55:43 +0000 (15:55 -0600)]
st/glx: fix depth/stencil format selection code

This fixes a pretty big performance regression caused by commit
3475e88442c16fb2b50b903fe246b3ebe49da226.

When the user does not request a stencil buffer it's important that we
don't use a depth/stencil format (or at least make it our last choice).
If the user calls glClear(GL_DEPTH_BUFFER_BIT) when we have a combined
depth/stencil buffer, that causes us to hit the clear_with_quad() path
which can be much, much slower than calling pipe_context::clear().

Also, try to use a shallower depth format before a deeper one.

14 years agost/mesa: rename st_framebuffer() to st_ws_framebuffer()
Brian Paul [Mon, 22 Mar 2010 21:49:38 +0000 (15:49 -0600)]
st/mesa: rename st_framebuffer() to st_ws_framebuffer()

Be clear that this function is not just a cast wrapper.

14 years agointel: Call intel_prepare_render() before looking up regions.
Kristian Høgsberg [Mon, 22 Mar 2010 21:54:35 +0000 (17:54 -0400)]
intel: Call intel_prepare_render() before looking up regions.

Fixes #27213.

14 years agointel: Call intel_prepare_render() before looking up regions.
Kristian Høgsberg [Mon, 22 Mar 2010 21:54:35 +0000 (17:54 -0400)]
intel: Call intel_prepare_render() before looking up regions.

Fixes #27213.

14 years agost/mesa: Remove unnecessary headers.
Vinson Lee [Mon, 22 Mar 2010 20:02:09 +0000 (13:02 -0700)]
st/mesa: Remove unnecessary headers.

14 years agoglslcompiler: fix build again
Brian Paul [Mon, 22 Mar 2010 19:36:27 +0000 (13:36 -0600)]
glslcompiler: fix build again

Simply use the libglapi.a archive instead of individual .o files.
Fixes the non-debug build.

14 years agocso: Remove unnecessary header.
Vinson Lee [Mon, 22 Mar 2010 18:34:18 +0000 (11:34 -0700)]
cso: Remove unnecessary header.

14 years agoFix potential compilation issue in visual studio project file
Karl Schultz [Mon, 22 Mar 2010 18:18:02 +0000 (12:18 -0600)]
Fix potential compilation issue in visual studio project file

Add quotes around filespec in post-build event to allow paths
with embedded spaces.

14 years agodri/nouveau: Rectangle texture fixes.
Francisco Jerez [Mon, 22 Mar 2010 17:10:25 +0000 (18:10 +0100)]
dri/nouveau: Rectangle texture fixes.

14 years agodri/nouveau: Some render to texture fixes.
Francisco Jerez [Mon, 22 Mar 2010 17:28:51 +0000 (18:28 +0100)]
dri/nouveau: Some render to texture fixes.

14 years agodri/nouveau: Fix swrast fallbacks when the read and draw buffers aren't the same.
Francisco Jerez [Mon, 22 Mar 2010 17:25:30 +0000 (18:25 +0100)]
dri/nouveau: Fix swrast fallbacks when the read and draw buffers aren't the same.

14 years agodri/nouveau: Expose EXT_framebuffer_blit.
Francisco Jerez [Mon, 22 Mar 2010 17:21:13 +0000 (18:21 +0100)]
dri/nouveau: Expose EXT_framebuffer_blit.

14 years agodri/nouveau: Plug in some mesa_meta functions instead of the swrast variants.
Francisco Jerez [Mon, 22 Mar 2010 13:23:40 +0000 (14:23 +0100)]
dri/nouveau: Plug in some mesa_meta functions instead of the swrast variants.

14 years agodri/nouveau: Emit dirty states on nouveau_validate_framebuffer.
Francisco Jerez [Mon, 22 Mar 2010 13:15:52 +0000 (14:15 +0100)]
dri/nouveau: Emit dirty states on nouveau_validate_framebuffer.

Fixes broken intermediate frames when a window is being resized
(regression caused by 878eef8c4).

14 years agonvfx: fix sampler views support
Luca Barbieri [Mon, 22 Mar 2010 17:36:16 +0000 (18:36 +0100)]
nvfx: fix sampler views support

The code was half converted, resulting in texturing being totally broken.

14 years agoglslcompiler: fix build breakage
Brian Paul [Mon, 22 Mar 2010 15:08:09 +0000 (09:08 -0600)]
glslcompiler: fix build breakage

14 years agoMerge branch '7.8'
Brian Paul [Mon, 22 Mar 2010 15:07:46 +0000 (09:07 -0600)]
Merge branch '7.8'

Conflicts:

src/gallium/drivers/cell/ppu/cell_screen.c
src/mesa/state_tracker/st_cb_drawpixels.c

14 years agoglslcompiler: fix build breakage
Brian Paul [Mon, 22 Mar 2010 15:00:13 +0000 (09:00 -0600)]
glslcompiler: fix build breakage

14 years agomesa: remove return, do as the comment says
Brian Paul [Mon, 22 Mar 2010 14:23:14 +0000 (08:23 -0600)]
mesa: remove return, do as the comment says

14 years agomesa: Add end of line to the end of a debug output.
Pauli Nieminen [Mon, 22 Mar 2010 13:14:16 +0000 (15:14 +0200)]
mesa: Add end of line to the end of a debug output.

14 years agodocs: Update the path to build libgl-xlib.
Chia-I Wu [Mon, 22 Mar 2010 05:30:54 +0000 (13:30 +0800)]
docs: Update the path to build libgl-xlib.

14 years agoprogs/tests: Add cva_huge to SCons build.
Vinson Lee [Mon, 22 Mar 2010 02:21:11 +0000 (19:21 -0700)]
progs/tests: Add cva_huge to SCons build.

14 years agoprogs/tests: Include stddef.h for ptrdiff_t on all platforms.
Vinson Lee [Mon, 22 Mar 2010 02:09:54 +0000 (19:09 -0700)]
progs/tests: Include stddef.h for ptrdiff_t on all platforms.

stddef.h is the standard C header that defines ptrdiff_t.

Fixes build of cva_huge on Mac OS X.

14 years agor200: Fix emit size prediction to account elt splitting.
Pauli Nieminen [Sun, 21 Mar 2010 21:23:21 +0000 (23:23 +0200)]
r200: Fix emit size prediction to account elt splitting.

Emit sizes prediction didn't account for render splitting in
hwtnl path.

14 years agor200: Don't flush when closing elts in KMS.
Pauli Nieminen [Sun, 21 Mar 2010 20:55:13 +0000 (22:55 +0200)]
r200: Don't flush when closing elts in KMS.

Flush in middle of rendering in KMS is not allowed because
buffers are discarded in flush.

Fixes crash when emiting split indices with RADEON_DEBUG=all.

14 years agor300g: disable tiling for YUV formats
Marek Olšák [Sun, 21 Mar 2010 21:17:14 +0000 (22:17 +0100)]
r300g: disable tiling for YUV formats

14 years agor300g: fix scons build
Marek Olšák [Sun, 21 Mar 2010 21:05:27 +0000 (22:05 +0100)]
r300g: fix scons build

14 years agor300g: accelerate blitting for all 64-bit texture formats
Marek Olšák [Sun, 21 Mar 2010 20:47:41 +0000 (21:47 +0100)]
r300g: accelerate blitting for all 64-bit texture formats

14 years agor300g: add and enable square microtiling
Marek Olšák [Sun, 14 Feb 2010 16:43:32 +0000 (17:43 +0100)]
r300g: add and enable square microtiling

It requires DRM 2.1.0 (e.g. kernel 2.6.34) and is disabled on older ones.

Finally, the texture tiling implementation is now complete. Uff.

14 years agor300g: cleanup tiling flags propagation
Marek Olšák [Sun, 21 Mar 2010 20:34:57 +0000 (21:34 +0100)]
r300g: cleanup tiling flags propagation

14 years agor300g: put common defines into one file
Marek Olšák [Sun, 21 Mar 2010 20:34:18 +0000 (21:34 +0100)]
r300g: put common defines into one file

14 years agor300g: remove unused code
Marek Olšák [Sun, 21 Mar 2010 20:22:21 +0000 (21:22 +0100)]
r300g: remove unused code

14 years agor200: Fix swtnl fallback to flush pending rendering before transition.
Pauli Nieminen [Sun, 21 Mar 2010 18:16:17 +0000 (20:16 +0200)]
r200: Fix swtnl fallback to flush pending rendering before transition.

Flush after transition would emit wrong state that could cause
wrong state emited for pending rendering operation.

Fixes wan once from extrement tuxracer that is using per vertex
materials.

14 years agor200: Fix mixed indetion in r200TclFallback.
Pauli Nieminen [Sun, 21 Mar 2010 18:02:10 +0000 (20:02 +0200)]
r200: Fix mixed indetion in r200TclFallback.

14 years agoRevert "drm/sw: does not need sw_winsys"
George Sapountzis [Sun, 21 Mar 2010 17:36:28 +0000 (19:36 +0200)]
Revert "drm/sw: does not need sw_winsys"

This reverts commit 97a694e2211dc804090f282d8e096c028f29579f.

14 years agoRevert "gallium: add soft screen helper"
George Sapountzis [Sun, 21 Mar 2010 17:33:17 +0000 (19:33 +0200)]
Revert "gallium: add soft screen helper"

This reverts commit f87a5f6499f51f651c2a9f2d4682875b22926905.

14 years agoRevert "drm/sw: just s/softpipe_create_screen/gallium_soft_create_screen/"
George Sapountzis [Sun, 21 Mar 2010 17:33:09 +0000 (19:33 +0200)]
Revert "drm/sw: just s/softpipe_create_screen/gallium_soft_create_screen/"

This reverts commit 5d524cce9c4fcc18ed977801d59ba7bb911020db.

14 years agoRevert "fix scons build"
George Sapountzis [Sun, 21 Mar 2010 17:32:40 +0000 (19:32 +0200)]
Revert "fix scons build"

This reverts commit 29ec05164838f13d9535271796a50fa213d07912.

14 years agofix scons build
George Sapountzis [Sun, 21 Mar 2010 17:20:15 +0000 (19:20 +0200)]
fix scons build

14 years agoMerge branch 'master' of git+ssh://git.freedesktop.org/git/mesa/mesa
George Sapountzis [Sun, 21 Mar 2010 13:37:40 +0000 (15:37 +0200)]
Merge branch 'master' of git+ssh://git.freedesktop.org/git/mesa/mesa

14 years agoRevert "Makefile.template: respect LIBRARY_DEFINES in mkdep"
George Sapountzis [Sun, 21 Mar 2010 13:32:50 +0000 (15:32 +0200)]
Revert "Makefile.template: respect LIBRARY_DEFINES in mkdep"

This reverts commit 9ec29e31919e85f9230867f43841c0e74be930d3.

14 years agonv50: report driver as GLSL capable
Christoph Bumiller [Sun, 21 Mar 2010 13:05:20 +0000 (14:05 +0100)]
nv50: report driver as GLSL capable

There's still no hint of optimization in the shaders, but
we support conditionals, loops and even integer opcodes so
it isn't that big a lie.

14 years agoMerge branch 'master' of git://anongit.freedesktop.org/git/mesa/mesa
George Sapountzis [Sun, 21 Mar 2010 13:05:21 +0000 (15:05 +0200)]
Merge branch 'master' of git://anongit.freedesktop.org/git/mesa/mesa

14 years agodrm/sw: just s/softpipe_create_screen/gallium_soft_create_screen/
George Sapountzis [Sun, 21 Mar 2010 13:01:36 +0000 (15:01 +0200)]
drm/sw: just s/softpipe_create_screen/gallium_soft_create_screen/

This partially reverts commit f87a5f6499f51f651c2a9f2d4682875b22926905.

14 years agonv50: don't validate arrays on clear
Christoph Bumiller [Sun, 21 Mar 2010 11:26:08 +0000 (12:26 +0100)]
nv50: don't validate arrays on clear

The vertex elements and buffers might not be valid anymore.

14 years agonv50: get rid of the static_init stateobj
Christoph Bumiller [Sun, 21 Mar 2010 12:17:02 +0000 (13:17 +0100)]
nv50: get rid of the static_init stateobj

Relocations of per-screen buffers are now emitted directly,
and include the necessary method to get changes in constbuf
addresses committed to the hw.

It should also be a bit cheaper than the way stateobjs emit
relocation markers, use a little less pushbuf space.

14 years agogallium: add soft screen helper
George Sapountzis [Fri, 19 Mar 2010 00:38:11 +0000 (02:38 +0200)]
gallium: add soft screen helper

14 years agodri1_api: need not include drm.h
George Sapountzis [Fri, 19 Mar 2010 00:38:11 +0000 (02:38 +0200)]
dri1_api: need not include drm.h

14 years agodrm/sw: does not need sw_winsys
George Sapountzis [Fri, 19 Mar 2010 00:38:11 +0000 (02:38 +0200)]
drm/sw: does not need sw_winsys

14 years agodrm/sw: llvmpipe texture_from_handle
George Sapountzis [Fri, 19 Mar 2010 00:38:11 +0000 (02:38 +0200)]
drm/sw: llvmpipe texture_from_handle

Not sure, but judging by softpipe, this hook was forgotten.

14 years agoMakefile.template: respect LIBRARY_DEFINES in mkdep
George Sapountzis [Fri, 19 Mar 2010 00:38:10 +0000 (02:38 +0200)]
Makefile.template: respect LIBRARY_DEFINES in mkdep

14 years agorename dri_sw to drisw_util for consistency
George Sapountzis [Fri, 19 Mar 2010 00:38:10 +0000 (02:38 +0200)]
rename dri_sw to drisw_util for consistency

14 years agodri_inteface: add define for checking presence of drm.h
George Sapountzis [Fri, 19 Mar 2010 00:38:09 +0000 (02:38 +0200)]
dri_inteface: add define for checking presence of drm.h

__NOT_HAVE_DRM_H is a like a feature, defined by default on specific platforms
and allows to be defined externally as well.

__NOT_HAVE_DRM_H should only be used by xserver and mesa swrast_dri drivers

14 years agoglx: minor cosmetic
George Sapountzis [Fri, 19 Mar 2010 23:51:00 +0000 (01:51 +0200)]
glx: minor cosmetic

14 years agoglx: swapBuffers prototype has changed
George Sapountzis [Fri, 19 Mar 2010 00:38:09 +0000 (02:38 +0200)]
glx: swapBuffers prototype has changed

14 years agovbo: Fix vbo_split_copy to pass correct max_index to draw.
Pauli Nieminen [Sun, 21 Mar 2010 10:42:10 +0000 (12:42 +0200)]
vbo: Fix vbo_split_copy to pass correct max_index to draw.

vbo_split_copy was passing one past the max_index to draw function
which caused _tnl_draw_prims function to read uninitialized values
from copied array.

Bug was spoted in valgrind report of progs/tests/cva_huge.

14 years agotests: Add test for huge client arrays that has to be split.
Pauli Nieminen [Sun, 21 Mar 2010 11:11:47 +0000 (13:11 +0200)]
tests: Add test for huge client arrays that has to be split.

When running this test case in valgrind report includes read of unitialized
value in _tnl_draw_prims. The bug doesn't cause any vissible errors.

Bug is caused by vbo_split_copy that is calling draw function with max_index
one past the end instead of the end.

14 years agodri/nouveau: Reemit the light model state when lighting goes on/off.
Francisco Jerez [Sun, 21 Mar 2010 10:49:46 +0000 (11:49 +0100)]
dri/nouveau: Reemit the light model state when lighting goes on/off.

14 years agodri/nouveau: Random cleanups.
Francisco Jerez [Sun, 21 Mar 2010 10:49:20 +0000 (11:49 +0100)]
dri/nouveau: Random cleanups.

14 years agor300g: fix occlusion query hangs.
Dave Airlie [Sun, 21 Mar 2010 09:36:15 +0000 (19:36 +1000)]
r300g: fix occlusion query hangs.

Okay need to revist the whole OQ stuff anyways, glean test asserts
which is never good.

I'm liking the cached bufmgr restrictions less and less, I think I'll
probably play with the fence and/or busy stuff ASAP and try and clean it up.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agor300g: add bufmgr cached for vertex/index buffers.
Dave Airlie [Sun, 21 Mar 2010 08:25:31 +0000 (18:25 +1000)]
r300g: add bufmgr cached for vertex/index buffers.

This adds the cached bufmgr for vertex/index buffers on top of the kernel
manager.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agogallium: cached bufmgr add map busy check
Dave Airlie [Sun, 21 Mar 2010 08:23:22 +0000 (18:23 +1000)]
gallium: cached bufmgr add map busy check

Suggested by Jose on the list, probably not perfect but will let me get
past this for now, testing with a fenced bufmgr on top of this, was slower,

Also this doesn't let you do the busy early exit optimisation either from
what I can see.

Signed-off-by: Dave Airlie <airlied@redhat.com>