mesa.git
15 years agosoftpipe: Grab a ref when the fb is set.
Brian Paul [Tue, 29 Sep 2009 14:21:54 +0000 (08:21 -0600)]
softpipe: Grab a ref when the fb is set.

Nasty bug when the surface is freed and another is allocated right on
top of it. The next time we set the fb state SP thinks it's the same
surface and doesn't flush, and when the flush eventually happens the
surface belongs to a completely different texture.

(cherry picked from commit a77226071f6814a53358a5d6caff685889d0e4ec)

Conflicts:
src/gallium/drivers/softpipe/sp_context.c

15 years agodocs: update news.html file with 7.5.2 and 7.6 release
Brian Paul [Mon, 28 Sep 2009 19:19:57 +0000 (13:19 -0600)]
docs: update news.html file with 7.5.2 and 7.6 release

15 years agoMerge branch 'mesa_7_5_branch' into mesa_7_6_branch
Ian Romanick [Mon, 28 Sep 2009 19:02:39 +0000 (12:02 -0700)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch

Conflicts:
src/mesa/main/version.h

15 years agodocs: 7.5.2 md5 sums
Ian Romanick [Mon, 28 Sep 2009 18:41:23 +0000 (11:41 -0700)]
docs: 7.5.2 md5 sums

15 years agoPrep for 7.5.2 release
Ian Romanick [Mon, 28 Sep 2009 18:33:52 +0000 (11:33 -0700)]
Prep for 7.5.2 release

15 years agodocs: 7.6 md5 sums
Ian Romanick [Mon, 28 Sep 2009 18:29:25 +0000 (11:29 -0700)]
docs: 7.6 md5 sums

15 years agoPrep for 7.6 release
Ian Romanick [Mon, 28 Sep 2009 18:22:54 +0000 (11:22 -0700)]
Prep for 7.6 release

15 years agoMerge branch 'mesa_7_5_branch' into mesa_7_6_branch
Brian Paul [Mon, 28 Sep 2009 15:59:59 +0000 (09:59 -0600)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch

15 years agoFix build on non GLIBC platforms (FreeBSD at least)
Robert Noland [Tue, 22 Sep 2009 18:49:57 +0000 (11:49 -0700)]
Fix build on non GLIBC platforms (FreeBSD at least)

Build was broken by commit 9666529b5a5be1fcde82caadc2fe2efa5ea81e49

I'm not certain that this is entirely the correct fix since the demo
from bug #23774 seemed to work before the commit that broke the build.

Signed-off-by: Robert Noland <rnoland@2hip.net>
Signed-off-by: Brian Paul <brianp@vmware.com>
15 years agodocs: list additional 7.5.2 bug fixes
Brian Paul [Mon, 28 Sep 2009 15:43:42 +0000 (09:43 -0600)]
docs: list additional 7.5.2 bug fixes

15 years agodocs: document gallium mipmap generation fix
Brian Paul [Mon, 28 Sep 2009 15:38:50 +0000 (09:38 -0600)]
docs: document gallium mipmap generation fix

15 years agost/mesa: fix st_generate_mipmap() issues
Brian Paul [Fri, 25 Sep 2009 23:24:27 +0000 (17:24 -0600)]
st/mesa: fix st_generate_mipmap() issues

The main issue is we didn't always have a gallium texture object with
enough space to store the to-be-generated mipmap levels.  When that's
the case, allocate a new gallium texture and use st_texure_finalize()
to copy images from the old texture to the new one.

We also had the baseLevel parameter to st_render_mipmap() wrong.

15 years agost/mesa: fix/simplify st_texture_object::lastLevel calculation
Brian Paul [Mon, 28 Sep 2009 15:28:50 +0000 (09:28 -0600)]
st/mesa: fix/simplify st_texture_object::lastLevel calculation

Don't compute the st_texture_object::lastLevel field based on the texture
filters.  Use the _MaxLevel value that core Mesa computes for us.
When called from the GenerateMipmap path, we'll use the lastLevel field
as-is.

15 years agogallium/util: add sanity check assertions
Brian Paul [Fri, 25 Sep 2009 23:19:25 +0000 (17:19 -0600)]
gallium/util: add sanity check assertions

15 years agointel: Handle GL_RGB8 for glCopyTex(Sub)Image.
Michel Dänzer [Fri, 25 Sep 2009 18:59:44 +0000 (20:59 +0200)]
intel: Handle GL_RGB8 for glCopyTex(Sub)Image.

Avoids an unnecessary fallback.

15 years agomesa: move declaration before code
Vinson Lee [Fri, 25 Sep 2009 14:51:57 +0000 (08:51 -0600)]
mesa: move declaration before code

15 years agoi915: Fix GetBufferSubData in the case of a system-memory BO.
Eric Anholt [Fri, 25 Sep 2009 03:03:21 +0000 (20:03 -0700)]
i915: Fix GetBufferSubData in the case of a system-memory BO.

Bug #23760 (crashes in wine)

15 years agointel: Flush the batch when we're about to subdata into a VBO.
Eric Anholt [Thu, 24 Sep 2009 23:15:52 +0000 (16:15 -0700)]
intel: Flush the batch when we're about to subdata into a VBO.

This fixes the clears in openarena with the new metaops clear code, and
the new piglit vbo-subdata-sync test.

Bug #23857.

15 years agoi965: Clean up some mess with the batch cache.
Eric Anholt [Thu, 24 Sep 2009 21:53:49 +0000 (14:53 -0700)]
i965: Clean up some mess with the batch cache.

Its flagging of extra state that's already flagged by the vtbl new_batch
when appropriate was confusing my tracking down of the OA clear bug.

15 years agoMerge branch 'mesa_7_5_branch' into mesa_7_6_branch
Brian Paul [Thu, 24 Sep 2009 22:01:26 +0000 (16:01 -0600)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch

15 years agosoftpipe: Increase GL_MAX_3D_TEXTURE_SIZE to 256.
Vinson Lee [Thu, 24 Sep 2009 21:59:57 +0000 (15:59 -0600)]
softpipe: Increase GL_MAX_3D_TEXTURE_SIZE to 256.

15 years agoi965: Emit zero initialization for NV VP temporaries as required.
Eric Anholt [Thu, 24 Sep 2009 18:58:33 +0000 (11:58 -0700)]
i965: Emit zero initialization for NV VP temporaries as required.

This is similar to what r300 does inside the driver, but I've added it as
a generic option since it seems most hardware will want it.

Fixes piglit nv-init-zero-reg.vpfp and nv-init-zero-addr.vpfp.

15 years agoi965: Remove assert about NV_vp now that it somewhat works.
Eric Anholt [Wed, 23 Sep 2009 23:50:59 +0000 (16:50 -0700)]
i965: Remove assert about NV_vp now that it somewhat works.

15 years agoi965: Load NV program matrices when required.
Eric Anholt [Thu, 24 Sep 2009 17:40:32 +0000 (10:40 -0700)]
i965: Load NV program matrices when required.

15 years agomesa: Initialize NV_vertex_program fields for the parameter lists and such.
Eric Anholt [Wed, 23 Sep 2009 23:49:52 +0000 (16:49 -0700)]
mesa: Initialize NV_vertex_program fields for the parameter lists and such.

This helps let drivers treat NV_vp like ARB_vp.

15 years agomesa: remove glEnable(GL_DEPTH_BOUNDS_TEST_EXT) check/warning
Brian Paul [Thu, 24 Sep 2009 20:24:14 +0000 (14:24 -0600)]
mesa: remove glEnable(GL_DEPTH_BOUNDS_TEST_EXT) check/warning

At the time of the enable there may not be a Z buffer, but one
may be attached to the FBO later.

15 years agomesa: remove rgbMode check in enable_texture()
Brian Paul [Thu, 24 Sep 2009 20:19:06 +0000 (14:19 -0600)]
mesa: remove rgbMode check in enable_texture()

If the currently bound FBO isn't yet validated it's possible for
rgbMode to be zero so we'll lose the texture enable.
This could fix some FBO rendering glitches, but I don't know of
any specific instances.

15 years agointel: use default array/element buffers in intel_generate_mipmap()
Brian Paul [Thu, 24 Sep 2009 18:41:14 +0000 (12:41 -0600)]
intel: use default array/element buffers in intel_generate_mipmap()

If there happened to be a bound VBO when intel_generate_mipmap() was
called we blew up because of a bad vertex array pointer.

Fixes regnumonline, bug 23859.

15 years agomesa: replace assertion with no-op function assignment
Brian Paul [Thu, 24 Sep 2009 18:37:34 +0000 (12:37 -0600)]
mesa: replace assertion with no-op function assignment

15 years agomesa: added comment
Brian Paul [Thu, 24 Sep 2009 18:37:06 +0000 (12:37 -0600)]
mesa: added comment

15 years agovbo: limit number of warnings to 10
Brian Paul [Thu, 24 Sep 2009 18:35:51 +0000 (12:35 -0600)]
vbo: limit number of warnings to 10

Otherwise some apps will emit tons of warnings.

15 years agoradeon: Fix scissors for r600 KMS.
Pauli Nieminen [Thu, 24 Sep 2009 16:58:09 +0000 (19:58 +0300)]
radeon: Fix scissors for r600 KMS.

Radeon generic scissors code had problem that some of code was using exclusive
and some inclusive bottom right corner. Only r600 driver is using exclusive
coordinate so changed generic code to pass inclusive coordinate and r600 driver
changes BR coordinate to be exclusive.

15 years agoMerge branch 'mesa_7_5_branch' into mesa_7_6_branch
Brian Paul [Thu, 24 Sep 2009 16:52:15 +0000 (10:52 -0600)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch

15 years agotgsi/sse: Pass the lodbias, not zero. More comments.
Brian Paul [Thu, 24 Sep 2009 16:26:56 +0000 (10:26 -0600)]
tgsi/sse: Pass the lodbias, not zero.  More comments.

This fixes the glean/glsl1 "texture2D(), with bias" test when using SSE.

15 years agomesa: added default case return to silence warning
Brian Paul [Wed, 23 Sep 2009 21:44:37 +0000 (15:44 -0600)]
mesa: added default case return to silence warning

15 years agoglsl: init var to silence warning
Brian Paul [Wed, 23 Sep 2009 21:44:18 +0000 (15:44 -0600)]
glsl: init var to silence warning

15 years agoglsl: fix missing initializers warning
Brian Paul [Wed, 23 Sep 2009 21:44:02 +0000 (15:44 -0600)]
glsl: fix missing initializers warning

15 years agor600: add support for CUBE textures, also TXP
Andre Maasikas [Wed, 23 Sep 2009 11:20:59 +0000 (14:20 +0300)]
r600: add support for CUBE textures, also TXP

seems to work here ...

15 years agor600: fix typo in the last commit
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

15 years agor600: various cleanups
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>
15 years agor600: fix some issues with LIT instruction
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

15 years agor600: fix point sizes
Alex Deucher [Sat, 19 Sep 2009 19:18:42 +0000 (15:18 -0400)]
r600: fix point sizes

registers takes radius

15 years agor600: fix polygon offset
Alex Deucher [Sat, 19 Sep 2009 18:46:06 +0000 (14:46 -0400)]
r600: fix polygon offset

15 years agoradeon: don't build non-r600 span code on r600
Alex Deucher [Tue, 15 Sep 2009 21:12:03 +0000 (17:12 -0400)]
radeon: don't build non-r600 span code on r600

15 years agor600: minor span cleanups
Alex Deucher [Tue, 15 Sep 2009 20:58:37 +0000 (16:58 -0400)]
r600: minor span cleanups

15 years agor600: support position_invariant programs
Andre Maasikas [Tue, 15 Sep 2009 15:27:51 +0000 (11:27 -0400)]
r600: support position_invariant programs

15 years agor600: add span support for 1D tiles
Alex Deucher [Mon, 14 Sep 2009 22:05:15 +0000 (18:05 -0400)]
r600: add span support for 1D tiles

1D tile span support for depth/stencil/color/textures

Z and stencil buffers are always tiled, so this fixes
sw access to Z and stencil buffers.  color and textures
are currently linear, but this adds span support when we
implement 1D tiling.

This fixes the text in progs/demos/engine and progs/tests/z*

15 years agor600: fix warning
Alex Deucher [Mon, 14 Sep 2009 21:08:26 +0000 (17:08 -0400)]
r600: fix warning

Noticed by rnoland on IRC.

15 years agor600: fix texcoords from constants
Andre Maasikas [Fri, 11 Sep 2009 19:59:55 +0000 (15:59 -0400)]
r600: fix texcoords from constants

with some minor updates from Richard.

15 years agor600: enable caching of vertex programs
Andre Maasikas [Fri, 11 Sep 2009 14:59:05 +0000 (10:59 -0400)]
r600: enable caching of vertex programs

15 years agor600: check if textures are actually enabled before submission
Alex Deucher [Wed, 9 Sep 2009 15:14:17 +0000 (11:14 -0400)]
r600: check if textures are actually enabled before submission

noticed by taiu on IRC.

15 years agor600: fix ftp for dri1
Alex Deucher [Wed, 9 Sep 2009 05:41:46 +0000 (01:41 -0400)]
r600: fix ftp for dri1

We use t->bo for dri1 since r600 uses CS for dri1.

15 years agor600: don't setup hardware state if TFP
Dave Airlie [Wed, 9 Sep 2009 05:02:16 +0000 (15:02 +1000)]
r600: don't setup hardware state if TFP

if we have a BO here it means TFP and we should have set it
up already.

tested by b0le on #radeon

15 years agor600: fix dri2 clipping
Alex Deucher [Tue, 8 Sep 2009 20:03:25 +0000 (16:03 -0400)]
r600: fix dri2 clipping

15 years agor300: fallback to software rendering if we are out of free texcoords
Maciej Cencora [Sat, 19 Sep 2009 16:45:59 +0000 (18:45 +0200)]
r300: fallback to software rendering if we are out of free texcoords

Fixes #22741

15 years agoswrast: add lod bias when texture sampling
Brian Paul [Wed, 23 Sep 2009 19:35:03 +0000 (13:35 -0600)]
swrast: add lod bias when texture sampling

Mostly fixes progs/demos/lodbias when MESA_TEX_PROG=1.  But the LOD still
seems off by -1 or so.
May be an issue with the params passed to _swrast_compute_lambda()

15 years agomesa: don't bias LOD in shader interpreter; do it in swrast
Brian Paul [Wed, 23 Sep 2009 19:34:30 +0000 (13:34 -0600)]
mesa: don't bias LOD in shader interpreter; do it in swrast

15 years agoswrast: fix typo in partial derivatives parameter passing
Brian Paul [Wed, 23 Sep 2009 18:54:14 +0000 (12:54 -0600)]
swrast: fix typo in partial derivatives parameter passing

15 years agost/mesa: trim calculated userbuffer size
Keith Whitwell [Wed, 23 Sep 2009 13:40:45 +0000 (14:40 +0100)]
st/mesa: trim calculated userbuffer size

In get_array_bounds we were previously defining a user buffer sized as
(nr_vertices * stride).  The trouble is that if the vertex data
occupies less than stride bytes, the extra tailing (stride - size)
bytes may extend outside the memory actually allocated by the app and
caused a segfault.

To fix this, define a the buffer bounds to be:

   ptr .. ptr + (nr-1)*stride + element_size

15 years agosoftpipe: added max texture/surface size sanity check
Brian Paul [Wed, 23 Sep 2009 16:50:38 +0000 (10:50 -0600)]
softpipe: added max texture/surface size sanity check

15 years agosoftpipe: increase MAX_WIDTH/HEIGTH 4096 to match max texture size
Brian Paul [Tue, 22 Sep 2009 21:35:05 +0000 (15:35 -0600)]
softpipe: increase MAX_WIDTH/HEIGTH 4096 to match max texture size

15 years agomesa: don't re-use the meta glDrawPixels VBO; create a new one each time
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.

15 years agomesa: fix more buffer object error messages
Brian Paul [Tue, 22 Sep 2009 19:47:49 +0000 (13:47 -0600)]
mesa: fix more buffer object error messages

15 years agoMerge branch 'mesa_7_5_branch' into mesa_7_6_branch
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

15 years agoglx: include string.h to silence missing memset() prototype warning
Brian Paul [Tue, 22 Sep 2009 19:19:05 +0000 (13:19 -0600)]
glx: include string.h to silence missing memset() prototype warning

15 years agomesa: fix error message text
Brian Paul [Tue, 22 Sep 2009 19:14:55 +0000 (13:14 -0600)]
mesa: fix error message text

15 years agor300: Fix crash reported in bug #24066
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>
15 years agoGLX: Warn only once about applications calling GLX 1.3 functions
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>
15 years agoMerge branch 'mesa_7_5_branch' into mesa_7_6_branch
Brian Paul [Mon, 21 Sep 2009 20:49:46 +0000 (14:49 -0600)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch

15 years agomesa: refine the error checking vbo_exec_DrawRangeElements()
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.

15 years agomesa: make max_buffer_index() a non-static function
Brian Paul [Mon, 21 Sep 2009 20:07:35 +0000 (14:07 -0600)]
mesa: make max_buffer_index() a non-static function

15 years agoradeon: update buffer map/unmap code for changes introduced in 92033a9516942d7272ce4b...
Maciej Cencora [Sun, 20 Sep 2009 11:54:59 +0000 (13:54 +0200)]
radeon: update buffer map/unmap code for changes introduced in 92033a9516942d7272ce4bf36ecd422009bbaf60 and 822c7964819ca1fcc270880d4ca8b3de8a4276d0

15 years agomesa: add some debug info to teximage.c
Maciej Cencora [Sat, 19 Sep 2009 16:47:36 +0000 (18:47 +0200)]
mesa: add some debug info to teximage.c

15 years agor300: fix a typo
Maciej Cencora [Sat, 19 Sep 2009 16:46:51 +0000 (18:46 +0200)]
r300: fix a typo

15 years agointel: Mark the FBO as incomplete if there's no intel_renderbuffer for it.
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)

15 years agosoftpipe: Fix cube face selection.
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)

15 years agoswrast: fix cube face selection
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.

15 years agor300: Zero-initialize register for NV_vertex_program
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>
15 years agor300: Fix handling of NV_vertex_program parameters
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>
15 years agoMerge branch 'mesa_7_5_branch' into 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

15 years agointel: Fix crash in intel_flush().
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().

15 years agoradeon: Fix legacy bo not to reuse dma buffers before refcount is 1.
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.

15 years agor300/compiler: Fix trig instructions in R300 fp
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>
15 years agoradeon: Fix typo in variable name.
Pauli Nieminen [Sun, 20 Sep 2009 18:08:42 +0000 (21:08 +0300)]
radeon: Fix typo in variable name.

15 years agoradeon: Improve WARN_ONCE macro to appear as single statement.
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.

15 years agoradeon: Fix "verts" debugging enable
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>
15 years agoMerge branch 'mesa_7_5_branch' into mesa_7_6_branch
Nicolai Hähnle [Sun, 20 Sep 2009 14:59:03 +0000 (16:59 +0200)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch

15 years agomesa/st: Create front renderbuffer on the fly when supplied with a surface
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>
15 years agomesa/st: Initialize format bits of framebuffer renderbuffers
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>
15 years agodocs: Document new features in radeon/r200/r300 drivers
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>
15 years agor300/compiler: Fix R300 fragment program regression introduced by 0723cd1...
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>
15 years ago[i965] add a missing header file
Zou Nan hai [Fri, 18 Sep 2009 08:04:41 +0000 (16:04 +0800)]
[i965] add a missing header file

15 years ago [i965] use intel_batchbuffer_flush to flush the clear
Zou Nan hai [Fri, 18 Sep 2009 05:29:28 +0000 (13:29 +0800)]
 [i965] use intel_batchbuffer_flush to flush the clear

15 years agomesa: fix clip plane, fog issues
Brian Paul [Thu, 17 Sep 2009 03:21:42 +0000 (21:21 -0600)]
mesa: fix clip plane, fog issues

15 years agoglx: Use initstate_r / random_r instead of corrupting global random number state
Ian Romanick [Wed, 16 Sep 2009 23:43:50 +0000 (16:43 -0700)]
glx: Use initstate_r / random_r instead of corrupting global random number state

Previously srandom and random were used.  This cause the global random
number generator state to be modified.  This caused problems for
applications that called srandom before calling into GLX.  By using
local state the global state is left unmodified.

This should fix bug #23774.

15 years agost/mesa: fix some incorrect branching/clean-up code in TexImage functions
Brian Paul [Wed, 16 Sep 2009 19:07:12 +0000 (13:07 -0600)]
st/mesa: fix some incorrect branching/clean-up code in TexImage functions

We need to be sure to call the _mesa_unmap_teximage_pbo() function if we
called _mesa_validate_pbo_teximage().

15 years agost/mesa: fix texture memory allocation bug
Brian Paul [Wed, 16 Sep 2009 18:57:26 +0000 (12:57 -0600)]
st/mesa: fix texture memory allocation bug

The following example caused an incorrect GL_OUT_OF_MEMORY error to be
raised in glTexSubImage2D:

   glTexImage2D(level=0, width=32, height=32, pixels=NULL);
   glTexImage2D(level=0, width=64, height=64, pixels=NULL);
   glTexSubImage2D(level=0, pixels!=NULL);

The second glTexImage2D() call needs to cause the first image to be
deallocated then reallocated at the new size.  This was not happening
because we were testing for pixels==NULL too early.

15 years agoMerge branch 'mesa_7_5_branch' into mesa_7_6_branch
Ian Romanick [Wed, 16 Sep 2009 14:57:19 +0000 (07:57 -0700)]
Merge branch 'mesa_7_5_branch' into mesa_7_6_branch

Conflicts:
src/mesa/main/dlist.c

15 years agointel: Deassociated drawables from private context struct in intelUnbindContext
Ian Romanick [Wed, 16 Sep 2009 14:39:58 +0000 (07:39 -0700)]
intel: Deassociated drawables from private context struct in intelUnbindContext

The generic DRI infrastructure makes sure that __DRIcontextRec::driDrawablePriv
and __DRIcontextRec::driReadablePriv are set to NULL after unbinding a
context.  However, the intel_context structure keeps cached copies of
these pointers.  If these cached pointers are not NULLed and the
drawable is actually destroyed after unbinding the context (typically
by way of glXDestroyWindow), freed memory will be dereferenced in
intelDestroyContext.

This should fix bug #23418.

15 years agoi965: do a flush in clear, fix openarena render issue,
Zou Nan hai [Wed, 16 Sep 2009 05:25:46 +0000 (13:25 +0800)]
i965: do a flush in clear, fix openarena render issue,
      fd.o bug# 23857