mesa.git
14 years agost/mesa: use util_format_get_blocksize()
Brian Paul [Wed, 24 Feb 2010 18:03:58 +0000 (11:03 -0700)]
st/mesa: use util_format_get_blocksize()

14 years agost/mesa: code clean-ups in st_renderbuffer_alloc_storage()
Brian Paul [Wed, 24 Feb 2010 17:56:10 +0000 (10:56 -0700)]
st/mesa: code clean-ups in st_renderbuffer_alloc_storage()

14 years agollvmpipe: SRGB textures not supported yet
Brian Paul [Wed, 24 Feb 2010 17:47:58 +0000 (10:47 -0700)]
llvmpipe: SRGB textures not supported yet

14 years agollvmpipe: vertex shader textures not supported yet
Brian Paul [Wed, 24 Feb 2010 17:43:30 +0000 (10:43 -0700)]
llvmpipe: vertex shader textures not supported yet

14 years agor600: proper fix for 15601835361e2fdd34b38b265cfc3007749ee24d
Alex Deucher [Wed, 24 Feb 2010 16:43:33 +0000 (11:43 -0500)]
r600: proper fix for 15601835361e2fdd34b38b265cfc3007749ee24d

PRE_EMIT_STATE_BUFSZ accounts for the start 3d, idle, cd/db flush
not for state.  The relocs for CB_COLOR0_FRAG & CB_COLOR0_TILE
are part of the render target state.

14 years agor300: Update for UTIL_FORMAT_LAYOUT_xxx changes.
José Fonseca [Wed, 24 Feb 2010 16:11:08 +0000 (16:11 +0000)]
r300: Update for UTIL_FORMAT_LAYOUT_xxx changes.

14 years agomesa: Update for UTIL_FORMAT_LAYOUT_xxx changes.
José Fonseca [Wed, 24 Feb 2010 16:10:36 +0000 (16:10 +0000)]
mesa: Update for UTIL_FORMAT_LAYOUT_xxx changes.

14 years agollvmpipe: Update for UTIL_FORMAT_LAYOUT_xxx changes.
José Fonseca [Wed, 24 Feb 2010 16:10:14 +0000 (16:10 +0000)]
llvmpipe: Update for UTIL_FORMAT_LAYOUT_xxx changes.

14 years agogallivm: Update for UTIL_FORMAT_LAYOUT_xxx changes.
José Fonseca [Wed, 24 Feb 2010 16:10:06 +0000 (16:10 +0000)]
gallivm: Update for UTIL_FORMAT_LAYOUT_xxx changes.

14 years agoutil: Kill array vs arith layouts. Revamp UTIL_FORMAT_LAYOUT_xxx.
José Fonseca [Wed, 24 Feb 2010 16:09:44 +0000 (16:09 +0000)]
util: Kill array vs arith layouts. Revamp UTIL_FORMAT_LAYOUT_xxx.

Too confusing. I now can get the same information through other means.

14 years agoi915g: fix a comment
Brian Paul [Wed, 24 Feb 2010 15:52:20 +0000 (08:52 -0700)]
i915g: fix a comment

14 years agogallium/docs: clean up texture usage text, added texture geom info
Brian Paul [Wed, 24 Feb 2010 15:52:03 +0000 (08:52 -0700)]
gallium/docs: clean up texture usage text, added texture geom info

14 years agollvmpipe: Fix breakeage.
José Fonseca [Wed, 24 Feb 2010 15:45:11 +0000 (15:45 +0000)]
llvmpipe: Fix breakeage.

Due to the u_format.csv changes the unrolled code generator was being used
for formats it didn't support.

Bring some bugfixes from u_format_access.py too.

14 years agoutil: Refactor some code.
José Fonseca [Wed, 24 Feb 2010 15:41:24 +0000 (15:41 +0000)]
util: Refactor some code.

14 years agoutil: Cope with the fact that formats in u_format.csv are not ordered.
José Fonseca [Wed, 24 Feb 2010 15:11:28 +0000 (15:11 +0000)]
util: Cope with the fact that formats in u_format.csv are not ordered.

14 years agoutil: Factor out the code to shorten a format name.
José Fonseca [Wed, 24 Feb 2010 15:10:46 +0000 (15:10 +0000)]
util: Factor out the code to shorten a format name.

14 years agollvmpipe: Handle correctly 24bit formats.
José Fonseca [Wed, 24 Feb 2010 14:51:47 +0000 (14:51 +0000)]
llvmpipe: Handle correctly 24bit formats.

14 years agoutil: Handle correctly 24bit formats.
José Fonseca [Wed, 24 Feb 2010 14:51:31 +0000 (14:51 +0000)]
util: Handle correctly 24bit formats.

14 years agoutil: Minor tweaks to ambigous/unused format descriptions.
José Fonseca [Wed, 24 Feb 2010 14:36:54 +0000 (14:36 +0000)]
util: Minor tweaks to ambigous/unused format descriptions.

14 years agoutil: Fix swizzles for SRGB formats.
José Fonseca [Wed, 24 Feb 2010 14:33:45 +0000 (14:33 +0000)]
util: Fix swizzles for SRGB formats.

Based on Marek Olšák's fix 8-bits-per-channel formats swizzles, with the
exception that the notation for several formats (the vector formats)
is actually reversed so they should be left alone.

14 years agoutil: Organize formats in groups. Add comments.
José Fonseca [Wed, 24 Feb 2010 14:15:09 +0000 (14:15 +0000)]
util: Organize formats in groups. Add comments.

14 years agoutil: Allow '#' comments in u_format.csv.
José Fonseca [Wed, 24 Feb 2010 13:47:42 +0000 (13:47 +0000)]
util: Allow '#' comments in u_format.csv.

14 years agomesa: disable unreachable meta mipmap gen code
Brian Paul [Wed, 24 Feb 2010 15:13:55 +0000 (08:13 -0700)]
mesa: disable unreachable meta mipmap gen code

More work is needed to support 3D mipmap generation.  Disable unreachable
code until then.

See bug 26722.

14 years agoutil: Kill unused UTIL_FORMAT_LAYOUT_SCALAR.
José Fonseca [Tue, 23 Feb 2010 20:20:19 +0000 (20:20 +0000)]
util: Kill unused UTIL_FORMAT_LAYOUT_SCALAR.

14 years agollvmpipe: Kill unused UTIL_FORMAT_LAYOUT_SCALAR.
José Fonseca [Tue, 23 Feb 2010 20:19:50 +0000 (20:19 +0000)]
llvmpipe: Kill unused UTIL_FORMAT_LAYOUT_SCALAR.

14 years agoutil: Store more derived data in the the format description.
José Fonseca [Tue, 23 Feb 2010 19:51:42 +0000 (19:51 +0000)]
util: Store more derived data in the the format description.

14 years agoradeon/r200: fix the state emission before kernel clear
Dave Airlie [Wed, 24 Feb 2010 09:13:07 +0000 (19:13 +1000)]
radeon/r200: fix the state emission before kernel clear

this moves the emission outside the lock and adds r200 support.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agor600: Assert that array index is not negative.
Vinson Lee [Wed, 24 Feb 2010 08:24:51 +0000 (00:24 -0800)]
r600: Assert that array index is not negative.

14 years agoradeon: fix bad state emission causes kernel to do bad depth clear
Dave Airlie [Wed, 24 Feb 2010 07:19:09 +0000 (17:19 +1000)]
radeon: fix bad state emission causes kernel to do bad depth clear

The kernel lets you clear depth without getting a depth offset
from userspace, mesa used to emit state before clear, but that got
lost in the refactoring, which made the kernel bug show up. Fix
mesa driver to emit the state properly now.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agointel: Call intel_prepare_render() in intelMakeCurrent()
Kristian Høgsberg [Wed, 24 Feb 2010 04:15:20 +0000 (23:15 -0500)]
intel: Call intel_prepare_render() in intelMakeCurrent()

This restores old behaviour, where we end up doing a DRI2GetBuffers()
call from intelMakeCurrent().  The idea was that we could do this
lazily, just before we start rendering.  However, if we don't do the
DRI2GetBuffers() round-trip we don't get the drawable size and higher
level mesa ends up short-cutting a number of GL calls, such as glClear().

14 years agogallivm: added clamp and int_to_float functions
Brian Paul [Wed, 24 Feb 2010 04:16:18 +0000 (21:16 -0700)]
gallivm: added clamp and int_to_float functions

14 years agollvmpipe: add lp_tile_image.c to sources
Brian Paul [Mon, 22 Feb 2010 22:19:45 +0000 (15:19 -0700)]
llvmpipe: add lp_tile_image.c to sources

14 years agollvmpipe: whole image tile/untile functions
Brian Paul [Mon, 22 Feb 2010 22:18:49 +0000 (15:18 -0700)]
llvmpipe: whole image tile/untile functions

14 years agost/mesa: whitespace fixes
Brian Paul [Wed, 24 Feb 2010 00:13:13 +0000 (17:13 -0700)]
st/mesa: whitespace fixes

14 years agollvmpipe: remove Z16 format override hack
Brian Paul [Wed, 24 Feb 2010 01:56:58 +0000 (18:56 -0700)]
llvmpipe: remove Z16 format override hack

14 years agollvmpipe: Z16 format is not supported
Brian Paul [Wed, 24 Feb 2010 01:56:24 +0000 (18:56 -0700)]
llvmpipe: Z16 format is not supported

14 years agost/glx: improved depth/stencil format selection code
Brian Paul [Wed, 24 Feb 2010 01:55:28 +0000 (18:55 -0700)]
st/glx: improved depth/stencil format selection code

Actually ask the gallium screen what Z/stencil format is supported.
This will let us remove some hacks in the llvmpipe driver.

14 years agost/glx: added PIPE_FORMAT_Z24S8_UNORM code
Brian Paul [Wed, 24 Feb 2010 01:23:15 +0000 (18:23 -0700)]
st/glx: added PIPE_FORMAT_Z24S8_UNORM code

14 years agollvmpipe: added assertions to verify that scene is empty
Brian Paul [Thu, 11 Feb 2010 01:10:16 +0000 (18:10 -0700)]
llvmpipe: added assertions to verify that scene is empty

14 years agollvmpipe: added some assertions
Brian Paul [Thu, 11 Feb 2010 01:07:38 +0000 (18:07 -0700)]
llvmpipe: added some assertions

14 years agoglapi: fix es after mv
George Sapountzis [Mon, 22 Feb 2010 20:27:49 +0000 (22:27 +0200)]
glapi: fix es after mv

14 years agoglapi: fix code generation Makefile after mv
George Sapountzis [Mon, 22 Feb 2010 20:27:48 +0000 (22:27 +0200)]
glapi: fix code generation Makefile after mv

14 years agoglapi: mv code generation to subdir
George Sapountzis [Mon, 22 Feb 2010 20:27:48 +0000 (22:27 +0200)]
glapi: mv code generation to subdir

This just moves files, code generation Makefile will be fixed in next commit for easier review.

14 years agor300g: fix texture swizzling for the SRGB formats
Marek Olšák [Tue, 23 Feb 2010 08:08:28 +0000 (09:08 +0100)]
r300g: fix texture swizzling for the SRGB formats

14 years agoglx: Fix some typos in pointer checking for OML_sync_control entry points.
kleinerm [Fri, 19 Feb 2010 20:28:26 +0000 (12:28 -0800)]
glx: Fix some typos in pointer checking for OML_sync_control entry points.

Signed-off-by: Mario Kleiner <mario.kleiner@tuebingen.mpg.de>
Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agopipebuffer: slab: if size < alignment, use alignment to find bucket
Luca Barbieri [Tue, 23 Feb 2010 16:55:24 +0000 (08:55 -0800)]
pipebuffer: slab: if size < alignment, use alignment to find bucket

If the size is lower than the alignment, we must use the alignment to
select the bucket.

Otherwise, the selected bucket won't be able to satisfy our request
and will fail.

Signed-off-by: José Fonseca <jfonseca@vmware.com>
14 years agopipebuffer: fix inverted signalled checking
Luca Barbieri [Tue, 23 Feb 2010 16:55:38 +0000 (08:55 -0800)]
pipebuffer: fix inverted signalled checking

A return of 0 means the fence is signalled.

Signed-off-by: José Fonseca <jfonseca@vmware.com>
14 years agopipebuffer: avoid assert due to increasing a zeroed refcnt
Luca Barbieri [Tue, 23 Feb 2010 16:50:15 +0000 (08:50 -0800)]
pipebuffer: avoid assert due to increasing a zeroed refcnt

The cache manager stores buffers with a reference count that dropped to 0.
pipe_reference asserts in this case on debug builds,
so use pipe_reference_init instead.

Signed-off-by: José Fonseca <jfonseca@vmware.com>
14 years agoi965: Enable GL_ARB_fragment_coord_conventions now that the GLSL is fixed.
Eric Anholt [Tue, 23 Feb 2010 18:48:19 +0000 (13:48 -0500)]
i965: Enable GL_ARB_fragment_coord_conventions now that the GLSL is fixed.

Tested with piglit glsl-arb-fragment-coord-conventions.

14 years agoegl: Remove unused _EGL_SKIP_HANDLE_CHECK.
Chia-I Wu [Mon, 22 Feb 2010 13:59:14 +0000 (21:59 +0800)]
egl: Remove unused _EGL_SKIP_HANDLE_CHECK.

It was added to skip checking EGLDisplay, EGLSurface, and etc.  It is
never defined and the spec does not allow the checks to be skipped.
Remove it for good.

14 years agost/egl: Unify surface creation.
Chia-I Wu [Mon, 22 Feb 2010 13:56:58 +0000 (21:56 +0800)]
st/egl: Unify surface creation.

Add a new function egl_g3d_create_surface and use it to create window,
pixmap, buffer, and screen surfaces.

14 years agomesa: Assert that array index is not negative.
Vinson Lee [Tue, 23 Feb 2010 08:19:30 +0000 (00:19 -0800)]
mesa: Assert that array index is not negative.

14 years agoglsl/apps: Assert that ftell does not return an error.
Vinson Lee [Tue, 23 Feb 2010 07:49:50 +0000 (23:49 -0800)]
glsl/apps: Assert that ftell does not return an error.

14 years agoi965g: Add fallthrough comments in switch statement.
Vinson Lee [Tue, 23 Feb 2010 07:32:43 +0000 (23:32 -0800)]
i965g: Add fallthrough comments in switch statement.

Silences warnings about missing break statements in static analysis.

14 years agodraw: Remove dead code.
Vinson Lee [Tue, 23 Feb 2010 07:23:57 +0000 (23:23 -0800)]
draw: Remove dead code.

14 years agor300g: fix draw_elements for "start" != 0
Marek Olšák [Mon, 22 Feb 2010 12:51:38 +0000 (13:51 +0100)]
r300g: fix draw_elements for "start" != 0

Reported-by: Andre Maasikas <amaasikas@gmail.com>
14 years agosvga: Upload user buffers only once.
José Fonseca [Mon, 22 Feb 2010 19:24:18 +0000 (19:24 +0000)]
svga: Upload user buffers only once.

14 years agosvga: Fix the guest offset of piecewise buffer DMAs.
José Fonseca [Mon, 22 Feb 2010 19:20:25 +0000 (19:20 +0000)]
svga: Fix the guest offset of piecewise buffer DMAs.

It was being erroneously set equal to the host offset, but it should be
zero.

14 years agosvga: Emit a scalar rcp in shadow maps.
José Fonseca [Mon, 22 Feb 2010 18:57:09 +0000 (18:57 +0000)]
svga: Emit a scalar rcp in shadow maps.

Small improvement in Lightsmark 2008.

14 years agosvga: Don't emit zero writemasks.
José Fonseca [Mon, 22 Feb 2010 18:01:32 +0000 (18:01 +0000)]
svga: Don't emit zero writemasks.

This fixes a regression with Lightsmark, where more compact TGSI from Mesa
was causing a zero mask MOV to be emitted for shadow map compare, causing
problems in some backends.

Add a few more assertions to catch cases like this.

14 years agotgsi: Report an error when a destination has an empty writemask.
José Fonseca [Mon, 22 Feb 2010 21:47:07 +0000 (21:47 +0000)]
tgsi: Report an error when a destination has an empty writemask.

14 years agosvga: Upload the dirty ranges instead of the whole buffer.
José Fonseca [Thu, 18 Feb 2010 17:21:53 +0000 (17:21 +0000)]
svga: Upload the dirty ranges instead of the whole buffer.

14 years agosvga: Break uploads of big buffers into smaller pieces.
José Fonseca [Thu, 18 Feb 2010 16:40:14 +0000 (16:40 +0000)]
svga: Break uploads of big buffers into smaller pieces.

Fixes crash in Homeworld2 which tries to create a 14MB buffer, because we
now avoid creating GMR buffers larger than 8MB to ensure progress given
we have only a 16MB pool.

14 years agosvga: Note ranges even when there is no hardware buffer around.
José Fonseca [Thu, 18 Feb 2010 14:12:31 +0000 (14:12 +0000)]
svga: Note ranges even when there is no hardware buffer around.

Not having a hardware buffer around doesn't change the fact that the
range is dirty and needs to be uploaded eventually.

14 years agosvga: Further cleanup/comment svga buffer code.
José Fonseca [Thu, 18 Feb 2010 14:05:57 +0000 (14:05 +0000)]
svga: Further cleanup/comment svga buffer code.

Just cosmetic changes -- no behavior change.

14 years agosvga: Temporarily create a sw vertex buf when failed to create a hw buf.
José Fonseca [Thu, 18 Feb 2010 18:11:50 +0000 (18:11 +0000)]
svga: Temporarily create a sw vertex buf when failed to create a hw buf.

Many apps don't check the return of map buffer so it is better not to
fail.

14 years agosvga: Remove unused buffer download code.
José Fonseca [Tue, 16 Feb 2010 16:43:53 +0000 (16:43 +0000)]
svga: Remove unused buffer download code.

Unnecessary now that we never destroy buffer storage.

14 years agopipebuffer: Don't synchronize when checking for buffer overflows.
José Fonseca [Tue, 16 Feb 2010 15:51:34 +0000 (15:51 +0000)]
pipebuffer: Don't synchronize when checking for buffer overflows.

To avoid masking synchronization issues in debug builds.

14 years agopython/retrace: Dump all vertices for verbosity level 3.
José Fonseca [Thu, 28 Jan 2010 17:11:02 +0000 (17:11 +0000)]
python/retrace: Dump all vertices for verbosity level 3.

14 years agosvga: Prevent buffer overflow in buffer ranges.
José Fonseca [Wed, 27 Jan 2010 15:41:25 +0000 (15:41 +0000)]
svga: Prevent buffer overflow in buffer ranges.

Do this by extending the nearest range to cover the new range.

This fixes an access fault in Call of Duty which was doing many disjoint
glBufferSubData calls.

14 years agosvga: Fix PIPE_LOGICOP_INVERT.
José Fonseca [Wed, 27 Jan 2010 14:45:56 +0000 (14:45 +0000)]
svga: Fix PIPE_LOGICOP_INVERT.

Actually the current XOR implementation is an INVERT.

This fixes rectangle selection in Maya.

14 years agobin/mklib: Clear CDPATH to avoid damaging expand_archive output
Keith Packard [Sun, 21 Feb 2010 22:30:00 +0000 (14:30 -0800)]
bin/mklib: Clear CDPATH to avoid damaging expand_archive output

The bash 'cd' command tends to emit random stuff to stdout when the
CDPATH variable is set, so clear it to keep extra filenames from being
emitted from the expand_archive function, which would otherwise cause
mklib to fail.

Signed-off-by: Keith Packard <keithp@keithp.com>
Reviewed-by: Dan Nicholson <dbn.lists@gmail.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
14 years agoglxinfo: Fix User error in glGetProgramivARB
Brian Paul [Mon, 22 Feb 2010 15:56:18 +0000 (08:56 -0700)]
glxinfo: Fix User error in glGetProgramivARB

glxinfo needed fixing after commit 4bccd69.

Move fragment program only parameters into their own list
so that they are not queried for a vertex program.

Signed-off-by: Xavier Chantry <chantry.xavier@gmail.com>
Signed-off-by: Brian Paul <brianp@vmware.com>
(cherry picked from commit b0e5dcb859692707fec8752dd21fc61fc4c3dbcc)

14 years agoi965g: added missing texture cap queries
Brian Paul [Mon, 22 Feb 2010 15:14:14 +0000 (08:14 -0700)]
i965g: added missing texture cap queries

14 years agoi915g: added missing texture cap queries
Brian Paul [Mon, 22 Feb 2010 15:14:05 +0000 (08:14 -0700)]
i915g: added missing texture cap queries

14 years agost/mesa: change viewport Z scale/bias for glBitmap/glDrawPixels
Brian Paul [Mon, 22 Feb 2010 15:04:53 +0000 (08:04 -0700)]
st/mesa: change viewport Z scale/bias for glBitmap/glDrawPixels

This fixes incorrect Z position of glBitmap, glDraw/CopyPixels for the
svga driver.  Now we use 0.5, 0.5 as is typical for ordinary 3D rendering.

(cherry picked from commit bcd561c66777e58dbb29a573c4d2279772bac6c5)

14 years agost/mesa: Add checks for ST_SURFACE_x vs MESA_BUFFER_x
Brian Paul [Mon, 22 Feb 2010 15:04:39 +0000 (08:04 -0700)]
st/mesa: Add checks for ST_SURFACE_x vs MESA_BUFFER_x

The ST_SURFACE_x values should match the Mesa BUFFER_x values.
Added some assertions to prevent future mix-ups.

(cherry picked from commit 13cbb5fff68ef8831230638e9f0b29a217750e9d)

14 years agomesa/st: Make ST_SURFACE_DEPTH index consistent with mesa's BUFFER_DEPTH.
Brian Paul [Mon, 22 Feb 2010 15:04:16 +0000 (08:04 -0700)]
mesa/st: Make ST_SURFACE_DEPTH index consistent with mesa's BUFFER_DEPTH.

Some st functions assume that they are identical.

(cherry picked from commit 9d17ad2891b58de9e33e943ff918a678c6a3c2bd)

14 years agointel: assert that we do not overflow the batch buffer.
Chris Wilson [Mon, 22 Feb 2010 09:53:36 +0000 (09:53 +0000)]
intel: assert that we do not overflow the batch buffer.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
14 years agoi915: Fix logic !gen >= 3
Chris Wilson [Mon, 22 Feb 2010 09:51:33 +0000 (09:51 +0000)]
i915: Fix logic !gen >= 3

The effect of this was that all objects were aligned to 128 bytes
on all generations, rather than just gen2.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
14 years agoi915: Remove superfluous MI_NOOP from vertex emission
Chris Wilson [Mon, 22 Feb 2010 09:50:15 +0000 (09:50 +0000)]
i915: Remove superfluous MI_NOOP from vertex emission

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
14 years agointel: Check that we have a bufmgr or bail out when initializing the context.
Chris Wilson [Thu, 21 Jan 2010 17:39:01 +0000 (17:39 +0000)]
intel: Check that we have a bufmgr or bail out when initializing the context.

Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
14 years agoRevert "r300g: rebuild winsys/pipe buffer handling and add buffer map"
Dave Airlie [Mon, 22 Feb 2010 07:26:30 +0000 (17:26 +1000)]
Revert "r300g: rebuild winsys/pipe buffer handling and add buffer map"

This reverts commit fff5be8e7b4557c221f2425dcafc2e7cbbba76ba.

Probably went too soon with this, dileX reported OA not working for him
it works here fine, but the optimisations I wanted aren't working properly
yet so I'll fix that now.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agor300g: rebuild winsys/pipe buffer handling and add buffer map
Dave Airlie [Fri, 12 Feb 2010 05:39:51 +0000 (15:39 +1000)]
r300g: rebuild winsys/pipe buffer handling and add buffer map

This creates a cleaner winsys and drop the simple screen stuff.

It makes r300g use pb_bufmgr structs and adds usage of the cached
bufmgr for vertex/index buffers.

It also avoids mapping too often.

I'm not 100% sure this is perfect but it won't find its own bugs.

Signed-off-by: Dave Airlie <airlied@redhat.com>
14 years agor3OOg: support rendering of more than 65535 vertices per draw call (almost)
Marek Olšák [Sun, 21 Feb 2010 19:49:29 +0000 (20:49 +0100)]
r3OOg: support rendering of more than 65535 vertices per draw call (almost)

The path for VAP_ALT_NUM_VERTICES is also in place (and tested)
but not enabled by default due to the missing support of this reg
in the upstream kernel.

Also, a non-zero BUFFER_BASE in the INDX_BUFFER packet3 hangs the machine.
Am I missing something? Because of this, only draw_arrays can render
more than 65535 vertices without the use of VAP_ALT_NUM_VERTICES.

14 years agoi965g: Fix memory leak.
Vinson Lee [Sun, 21 Feb 2010 23:15:52 +0000 (15:15 -0800)]
i965g: Fix memory leak.

14 years agost/mesa: Fix memory leak on out-of-memory error path.
Vinson Lee [Sun, 21 Feb 2010 23:10:50 +0000 (15:10 -0800)]
st/mesa: Fix memory leak on out-of-memory error path.

14 years agor300g: Remove unnecessary header.
Vinson Lee [Sun, 21 Feb 2010 22:26:53 +0000 (14:26 -0800)]
r300g: Remove unnecessary header.

14 years agoLink egl progs to libm as necessary.
Ray Kohler [Sun, 21 Feb 2010 03:26:43 +0000 (22:26 -0500)]
Link egl progs to libm as necessary.

Signed-off-by: Ray Kohler <ataraxia937@gmail.com>
14 years agor300g: precompute framebuffer register values
Marek Olšák [Sat, 20 Feb 2010 20:12:45 +0000 (21:12 +0100)]
r300g: precompute framebuffer register values

14 years agor300g: add all missing colorbuffer formats
Marek Olšák [Sat, 20 Feb 2010 23:17:04 +0000 (00:17 +0100)]
r300g: add all missing colorbuffer formats

14 years agor300g: make r300_translate_texformat private
Marek Olšák [Sat, 20 Feb 2010 20:33:53 +0000 (21:33 +0100)]
r300g: make r300_translate_texformat private

Unlikely to increase performance from inlining.

And partially expose it through r300_is_sampler_format_supported.

14 years agor300g: re-enable SRGB formats
Marek Olšák [Sat, 20 Feb 2010 17:52:20 +0000 (18:52 +0100)]
r300g: re-enable SRGB formats

Ouch.

14 years agomesa: Add assert to check for null pointer dereference.
Vinson Lee [Sun, 21 Feb 2010 10:13:15 +0000 (02:13 -0800)]
mesa: Add assert to check for null pointer dereference.

14 years agodraw: Remove dead code.
Vinson Lee [Sun, 21 Feb 2010 09:57:16 +0000 (01:57 -0800)]
draw: Remove dead code.

14 years agoi965g: Remove dead code.
Vinson Lee [Sun, 21 Feb 2010 09:30:35 +0000 (01:30 -0800)]
i965g: Remove dead code.

14 years agoi965g: Fix assertions that were always true.
Vinson Lee [Sun, 21 Feb 2010 09:19:47 +0000 (01:19 -0800)]
i965g: Fix assertions that were always true.

The surface_format field is 9 bits wide which has a maximum value
of 511. BRW_SURFACEFORMAT_INVALID has a value of 4095 which exceeds
the width of the surface_format field.

This patch asserts that the format is not BRW_SURFACEFORMAT_INVALID
before assignment to the surface_format field.

14 years agoglsl/cl: Prevent possible buffer overflow.
Vinson Lee [Sun, 21 Feb 2010 08:36:50 +0000 (00:36 -0800)]
glsl/cl: Prevent possible buffer overflow.

14 years agoi965g: Fix use of out-of-scope variable.
Vinson Lee [Sun, 21 Feb 2010 08:24:05 +0000 (00:24 -0800)]
i965g: Fix use of out-of-scope variable.

14 years agost/egl: Disable SHM in ximage backend.
Chia-I Wu [Sun, 21 Feb 2010 07:47:45 +0000 (15:47 +0800)]
st/egl: Disable SHM in ximage backend.

With pipe_screen::update_buffer being called in st_validate_state, it is
likely that softpipe will flush tile caches to a detached SHM segment.
Disable SHM for now until a better solution is found.