mesa.git
16 years agoAdd E7221 variant to i915.
Adam Jackson [Fri, 15 Feb 2008 19:46:25 +0000 (14:46 -0500)]
Add E7221 variant to i915.

16 years agoTrivial SELinux awareness. Enable with --enable-selinux.
Adam Jackson [Fri, 15 Feb 2008 18:49:12 +0000 (13:49 -0500)]
Trivial SELinux awareness.  Enable with --enable-selinux.

Avoids AVC warnings when allocating executable memory by first checking
if the current process has permission to do so.

16 years agofix bug with generated fragment programs if vertex shader is active
Roland Scheidegger [Fri, 15 Feb 2008 16:26:06 +0000 (17:26 +0100)]
fix bug with generated fragment programs if vertex shader is active

When generating a fragment program from fixed function, checking
texUnit->_ReallyEnabled is not sufficient, need texUnit->Enabled too
since the former also represents texture enables from an active vertex
shader.

16 years agoi965: don't swizzle fogcoord if FogOption is FOG_NONE.
Xiang, Haihao [Fri, 15 Feb 2008 08:13:11 +0000 (16:13 +0800)]
i965: don't swizzle fogcoord if FogOption is FOG_NONE.
fix #10788 issue on 965.

16 years agoi915: set fogcoord to (f,0,0,1). fix #10788 issue on 915.
Xiang, Haihao [Fri, 15 Feb 2008 07:20:45 +0000 (15:20 +0800)]
i915: set fogcoord to (f,0,0,1). fix #10788 issue on 915.

16 years agoAdd TTM buffer object based texture from pixmap implementation.
Kristian Høgsberg [Fri, 15 Feb 2008 03:12:51 +0000 (22:12 -0500)]
Add TTM buffer object based texture from pixmap implementation.

Currently only implemented for intel hw.

16 years agoAdd GLXPixmap version of glxgears to show case new GLX Pixmap functionality.
Kristian Høgsberg [Fri, 15 Feb 2008 03:05:55 +0000 (22:05 -0500)]
Add GLXPixmap version of glxgears to show case new GLX Pixmap functionality.

16 years agoHook up i915 driver to new DRI2 infrastructure.
Kristian Høgsberg [Mon, 21 Jan 2008 22:07:33 +0000 (17:07 -0500)]
Hook up i915 driver to new DRI2 infrastructure.

16 years agoAdd new DRI2 infrastructure.
Kristian Høgsberg [Mon, 14 Jan 2008 23:31:05 +0000 (18:31 -0500)]
Add new DRI2 infrastructure.

16 years agominiglx doc updates
caner [Thu, 14 Feb 2008 16:33:05 +0000 (09:33 -0700)]
miniglx doc updates

16 years agoi965: use setup attributes as inputs when allocating registers
Xiang, Haihao [Thu, 14 Feb 2008 05:24:27 +0000 (13:24 +0800)]
i965: use setup attributes as inputs when allocating registers
for WM payload.  fix #10767

16 years ago[i965] flip point sprite
Zou Nan hai [Thu, 14 Feb 2008 03:01:34 +0000 (11:01 +0800)]
[i965] flip point sprite

16 years ago[i965] gl_FrontFacing support
Zou Nan hai [Thu, 14 Feb 2008 02:12:03 +0000 (10:12 +0800)]
[i965] gl_FrontFacing support

16 years agoi965: remove unused hal hooks
Dave Airlie [Thu, 14 Feb 2008 00:44:43 +0000 (10:44 +1000)]
i965: remove unused hal hooks

These don't appear to have ever been used.

16 years ago[965] Fix ARB_occlusion_query from intel_screen.c merge.
Eric Anholt [Thu, 14 Feb 2008 00:08:19 +0000 (16:08 -0800)]
[965] Fix ARB_occlusion_query from intel_screen.c merge.

It wasn't being initialized at screen setup, so we were getting stub
entrypoints even though it was exposed as enabled.  Fixes arbocclude mesa demo.

16 years ago[intel] Fix 965 rendering with non-TTM by merging intel_ioctl between 915/965.
Eric Anholt [Wed, 13 Feb 2008 18:34:20 +0000 (10:34 -0800)]
[intel] Fix 965 rendering with non-TTM by merging intel_ioctl between 915/965.

The 965 path wasn't setting pClipRects for batch submission since it didn't
want kernel cliprect handling before.  The 915 path also grew the INTEL_NO_HW=1
option for testing just driver overhead.

16 years agoWhen mapping, wait on the buffer's fence, not hardware idle, in bufmgr_fake
Eric Anholt [Wed, 13 Feb 2008 17:26:06 +0000 (09:26 -0800)]
When mapping, wait on the buffer's fence, not hardware idle, in bufmgr_fake

16 years agoRemove O(n^2) debugging code from non-debug path of dri_bufmgr_fake.
Eric Anholt [Wed, 13 Feb 2008 17:16:07 +0000 (09:16 -0800)]
Remove O(n^2) debugging code from non-debug path of dri_bufmgr_fake.

16 years ago_mesa_swizzle_ubyt_image: Don't use single swizzle_copy call
Xiang, Haihao [Wed, 13 Feb 2008 07:57:44 +0000 (15:57 +0800)]
_mesa_swizzle_ubyt_image: Don't use single swizzle_copy call
if components don't match. fix #13508

16 years ago[intel] Remove cached reloc data buffer now that it's not a BO.
Eric Anholt [Wed, 13 Feb 2008 00:03:58 +0000 (16:03 -0800)]
[intel] Remove cached reloc data buffer now that it's not a BO.

It's not worth the extra effort to avoid a free/malloc, and we'd rather
auto-size the reloc data buffer at some point so we don't need to have
max_relocs.

16 years ago[intel] Fix type of some more flags variables for uint64_t flags.
Eric Anholt [Wed, 13 Feb 2008 00:01:26 +0000 (16:01 -0800)]
[intel] Fix type of some more flags variables for uint64_t flags.

Harmless since we don't yet have any bits above 31 for flags.

16 years ago[intel] Note when BO map/unmap fail with TTM.
Eric Anholt [Mon, 11 Feb 2008 16:29:05 +0000 (08:29 -0800)]
[intel] Note when BO map/unmap fail with TTM.

16 years ago[intel] Fix INTEL_DEBUG=bufmgr after relocation interface fixups.
Eric Anholt [Fri, 8 Feb 2008 23:58:45 +0000 (15:58 -0800)]
[intel] Fix INTEL_DEBUG=bufmgr after relocation interface fixups.

16 years ago[965] Remove stale brw_state_cache.c comment and function export.
Eric Anholt [Thu, 7 Feb 2008 23:12:59 +0000 (15:12 -0800)]
[965] Remove stale brw_state_cache.c comment and function export.

16 years agoFix broken test.
Brian [Tue, 12 Feb 2008 23:56:18 +0000 (16:56 -0700)]
Fix broken test.

As-is, if the texture was too large for the target, an assertion would fail.
Now check proxy texture first and if it works, then test non-proxy target.

16 years agoSet Min/MaxLod with glTexParameterf, not glTexParameteri
Brian [Tue, 12 Feb 2008 22:53:37 +0000 (15:53 -0700)]
Set Min/MaxLod with glTexParameterf, not glTexParameteri

16 years agotweak incr/decrements amounts for keyboard options
Brian [Tue, 12 Feb 2008 21:59:07 +0000 (14:59 -0700)]
tweak incr/decrements amounts for keyboard options

16 years agopress 0,1,2,etc keys for specific bias values
Brian [Mon, 11 Feb 2008 16:46:10 +0000 (09:46 -0700)]
press 0,1,2,etc keys for specific bias values

16 years agoglapi: Correct prerequisites for gl_and_glX_API.xml
Dan Nicholson [Tue, 15 Jan 2008 18:21:31 +0000 (10:21 -0800)]
glapi: Correct prerequisites for gl_and_glX_API.xml

The indirect_dispatch.h and indirect_table.c source files use
gl_and_glX_API.xml in their generation rather than glX_API.xml, but it
wasn't listed in their prerequisites. In turn, gl_and_glX_API.xml uses
glX_API.xml, but this is already listed in $(COMMON_GLX).

16 years agoglapi: Use make automatic variables to clean up the commands
Dan Nicholson [Tue, 15 Jan 2008 18:15:36 +0000 (10:15 -0800)]
glapi: Use make automatic variables to clean up the commands

Make use of the make automatic variables $@ (the target) and $< (first
prerequisite) to clean up the commands for the glapi generation. This
improves readability and guards against typos since words are repeated
less frequently.

16 years agoglapi: Use variable for indent and flags
Dan Nicholson [Tue, 15 Jan 2008 17:52:17 +0000 (09:52 -0800)]
glapi: Use variable for indent and flags

Put the path to indent and the flags to call it with in configs/default
rather than in the Makefile. This makes it easier to change the values
globally.

16 years agonouveau: ddx versioning changed
Ben Skeggs [Tue, 12 Feb 2008 06:28:31 +0000 (17:28 +1100)]
nouveau: ddx versioning changed

16 years ago[directfb] Added RGB444 and RGB555.
Claudio Ciccani [Fri, 8 Feb 2008 11:36:45 +0000 (12:36 +0100)]
[directfb] Added RGB444 and RGB555.
Also added color expansion for RGB16, ARGB1555 and ARGB4444.

16 years ago[glut-directfb] Fixed displaying of single buffered window.
Claudio Ciccani [Fri, 8 Feb 2008 11:35:08 +0000 (12:35 +0100)]
[glut-directfb] Fixed displaying of single buffered window.

16 years ago[glut-directfb] When converting microseconds to milliseconds, round to the next integer.
Claudio Ciccani [Fri, 8 Feb 2008 11:03:53 +0000 (12:03 +0100)]
[glut-directfb] When converting microseconds to milliseconds, round to the next integer.

16 years ago[965] Flush icache on new batch, not just new context.
Eric Anholt [Thu, 7 Feb 2008 20:09:16 +0000 (12:09 -0800)]
[965] Flush icache on new batch, not just new context.

This is required since our buffer manager may now move our
instruction-containing buffers at any batchbuffer emit.

16 years agoadded -altopts to allow overriding all other opts
Andy Skinner [Thu, 7 Feb 2008 20:20:06 +0000 (13:20 -0700)]
added -altopts to allow overriding all other opts

16 years ago[915] Fix COS function using same plan as SIN.
Eric Anholt [Wed, 6 Feb 2008 23:41:04 +0000 (15:41 -0800)]
[915] Fix COS function using same plan as SIN.

The previous COS function failed badly outside of [-pi/2, pi/2].

16 years ago[915] Use a quartic term to improve the accuracy of SIN results.
Eric Anholt [Wed, 6 Feb 2008 23:38:16 +0000 (15:38 -0800)]
[915] Use a quartic term to improve the accuracy of SIN results.

This is described in the link in the comment, and is the same technique that
r300 uses.

16 years ago[915] Fix fp SIN function, and use a quadratic approximation instead of Taylor.
Eric Anholt [Wed, 6 Feb 2008 19:34:14 +0000 (11:34 -0800)]
[915] Fix fp SIN function, and use a quadratic approximation instead of Taylor.

The Taylor series notably fails at producing sin(pi) == 0, which leads to
discontinuity every 2*pi.  The quadratic gets us sin(pi) == 0 behavior, at the
expense of going from 2.4% THD with working Taylor series to 3.8% THD (easily
seen on comparative graphs of the two).  However, our previous implementation
was producing sin(pi) < -1 and worse, so any reasonable approximation is an
improvement.  This also fixes the repeating behavior, where the previous
implementation would repeat sin(x) for x>pi as sin(x % pi) and the opposite
for x < -pi.

16 years ago[965] Bug 14314: assertion failure with with !AIGLX and depth=24 visual.
Eric Anholt [Tue, 5 Feb 2008 19:00:04 +0000 (11:00 -0800)]
[965] Bug 14314: assertion failure with with !AIGLX and depth=24 visual.

16 years ago[965] Fix TTM relocation caching overzealousness.
Eric Anholt [Tue, 5 Feb 2008 17:38:02 +0000 (09:38 -0800)]
[965] Fix TTM relocation caching overzealousness.

The failure mode that was a available was:
reloc 1 -> target_buf
exec: PRESUMED_OFFSET wrong, buffer migrates, r1 entry updated.
reloc 2 -> target_buf
exec: suppose buffer migrates again.  PRESUMED_OFFSET wrong. r2 entry updated.
reloc 1 -> target_buf
exec: suppose buffer doesn't migrate.  PRESUMED_OFFSET right. no relocations
      performed.  r1 has stale pointer at original location.

Failures were reported with OGLconform's VBO test and SPECviewperf90, though
I haven't confirmed that this fixes it.

16 years agoi965: adjust the byte order of clear color. fix #14165
Xiang, Haihao [Tue, 5 Feb 2008 07:17:58 +0000 (15:17 +0800)]
i965: adjust the byte order of clear color. fix #14165

16 years agoReplace usage of DRM_BO_FLAG_MEM_TT in intel_regions.c with local/cached.
Eric Anholt [Tue, 5 Feb 2008 02:19:03 +0000 (18:19 -0800)]
Replace usage of DRM_BO_FLAG_MEM_TT in intel_regions.c with local/cached.

In addition to potentially binding when it was about to be mapped anyway,
failure to use CACHED_MAPPED means eating a full wbinvd on validate.  Thanks to
airlied for catching this.

16 years agoInclude glext.h in the cva test so that it actually uses CVAs.
Eric Anholt [Sun, 3 Feb 2008 09:04:46 +0000 (01:04 -0800)]
Include glext.h in the cva test so that it actually uses CVAs.

16 years agoAllow first != 0 in mesa CVA handling, and add more error checking.
Eric Anholt [Sat, 2 Feb 2008 10:54:13 +0000 (02:54 -0800)]
Allow first != 0 in mesa CVA handling, and add more error checking.

16 years ago[965] Convert brw_draw_upload to managing dri_bos, not gl_buffer_objects.
Eric Anholt [Sun, 3 Feb 2008 07:27:19 +0000 (23:27 -0800)]
[965] Convert brw_draw_upload to managing dri_bos, not gl_buffer_objects.

This helps us avoid a bunch of mess with gl_client_arrays that we filled
with unused data and confused readers.

16 years ago[965] Remove dead structure in brw_draw_upload.c.
Eric Anholt [Sat, 2 Feb 2008 10:59:55 +0000 (02:59 -0800)]
[965] Remove dead structure in brw_draw_upload.c.

16 years ago[965] Move temporary vbo array storage into the function using it.
Eric Anholt [Tue, 22 Jan 2008 22:03:24 +0000 (14:03 -0800)]
[965] Move temporary vbo array storage into the function using it.

16 years ago[965] Remove dead brw_vertex_element members.
Eric Anholt [Tue, 22 Jan 2008 19:35:04 +0000 (11:35 -0800)]
[965] Remove dead brw_vertex_element members.

16 years ago[965] Add a wrapper around interleaved copy_array_to_vbo_array for profiling.
Eric Anholt [Tue, 22 Jan 2008 19:34:46 +0000 (11:34 -0800)]
[965] Add a wrapper around interleaved copy_array_to_vbo_array for profiling.

If compiled with optimization, it shouldn't appear at all, and helps me for
now.

16 years ago[965] Avoid overloaded use of the term 'input' for clarity.
Eric Anholt [Tue, 22 Jan 2008 19:04:09 +0000 (11:04 -0800)]
[965] Avoid overloaded use of the term 'input' for clarity.

16 years ago[965] Replace VEP/VBP state structures with inline batch emits.
Eric Anholt [Tue, 22 Jan 2008 18:56:09 +0000 (10:56 -0800)]
[965] Replace VEP/VBP state structures with inline batch emits.

16 years agor300: fix isosurf on rs690
Dave Airlie [Mon, 4 Feb 2008 10:59:26 +0000 (21:59 +1100)]
r300: fix isosurf on rs690

16 years agoi965: fix potential NULL pointer dereference. The third region
Xiang, Haihao [Sun, 3 Feb 2008 12:36:17 +0000 (20:36 +0800)]
i965: fix potential NULL pointer dereference. The third region
isn't created at all for 965

16 years ago[965] Fix indentation.
Eric Anholt [Sat, 2 Feb 2008 00:04:38 +0000 (16:04 -0800)]
[965] Fix indentation.

16 years agoRevert "intel: don't apply the relocation optimization if a target"
Eric Anholt [Fri, 1 Feb 2008 02:43:51 +0000 (18:43 -0800)]
Revert "intel: don't apply the relocation optimization if a target"

This reverts commit e2cb905bc6e23eaafaeeb2abdc9480e70959ee3f.

It was a reversion of an optimization hidden as otherwise.
pre_target_buf_handle was always NULL, so the optimization was never enabled,
rather than fixing the important optimization (resulting in 25-50% performance
loss).

16 years ago[965] Replace XXX comment about constant swizzle with an assert.
Eric Anholt [Fri, 1 Feb 2008 00:16:29 +0000 (16:16 -0800)]
[965] Replace XXX comment about constant swizzle with an assert.

16 years ago[965] Fix some indentation in brw_vs_tnl.c.
Eric Anholt [Mon, 28 Jan 2008 06:21:01 +0000 (22:21 -0800)]
[965] Fix some indentation in brw_vs_tnl.c.

16 years agomesa: re-define NEED_SECONDARY_COLOR. fix #14310.
Xiang, Haihao [Fri, 1 Feb 2008 16:56:51 +0000 (00:56 +0800)]
mesa: re-define NEED_SECONDARY_COLOR. fix #14310.

16 years ago [intel] fix for previous fix
Zou Nan hai [Fri, 1 Feb 2008 12:28:26 +0000 (20:28 +0800)]
  [intel]  fix for previous fix

16 years ago [intel] use _mesa_copy_rect for upload compressed texture,
Zou Nan hai [Fri, 1 Feb 2008 09:36:56 +0000 (17:36 +0800)]
   [intel] use _mesa_copy_rect for upload compressed texture,
this fix bad texture issue in some games(UT and quake).

16 years agoi965: Don't emit state if fall back to software rendering. fix #14116
Xiang, Haihao [Fri, 1 Feb 2008 08:46:41 +0000 (16:46 +0800)]
i965: Don't emit state if fall back to software rendering. fix #14116

16 years ago[i965] renable regative rhw test
Zou Nan hai [Thu, 31 Jan 2008 10:22:19 +0000 (18:22 +0800)]
[i965] renable regative rhw test

16 years agointel: don't apply the relocation optimization if a target
Xiang, Haihao [Thu, 31 Jan 2008 09:17:03 +0000 (17:17 +0800)]
intel: don't apply the relocation optimization if a target
buffer is used for a relocatee in the former relocation process
then another target buffer is used for this relocatee at the same
offset in the current relocation process.

16 years agoregenerate glsl library functions
Roland Scheidegger [Thu, 31 Jan 2008 00:34:01 +0000 (01:34 +0100)]
regenerate glsl library functions

16 years agofix w component of glsl vec4 asin
Roland Scheidegger [Thu, 31 Jan 2008 00:32:46 +0000 (01:32 +0100)]
fix w component of glsl vec4 asin

16 years agocheck if fb->Delete is null (bugs 13507,14293)
Brian [Wed, 30 Jan 2008 15:08:23 +0000 (08:08 -0700)]
check if fb->Delete is null (bugs 13507,14293)

16 years agoAdd new RV380 pci id
Alex Deucher [Tue, 29 Jan 2008 15:14:04 +0000 (10:14 -0500)]
Add new RV380 pci id

bug 14289

16 years agoi965: new integrated graphics chipset support
Xiang, Haihao [Tue, 29 Jan 2008 03:13:53 +0000 (11:13 +0800)]
i965: new integrated graphics chipset support

16 years agopush out far clip plane to 200
Brian [Mon, 28 Jan 2008 19:41:47 +0000 (12:41 -0700)]
push out far clip plane to 200

16 years agoAdded d/D keys to change viewing distance, 'a' to toggle animation
Brian [Mon, 28 Jan 2008 19:35:20 +0000 (12:35 -0700)]
Added d/D keys to change viewing distance, 'a' to toggle animation

16 years agor300: add initial rs690 support to Mesa
Dave Airlie [Sun, 27 Jan 2008 02:16:06 +0000 (12:16 +1000)]
r300: add initial rs690 support to Mesa

The rs690 has an rs4xx style vertex-shader less 3D engine. It uses the new
r500 output engine though. It also needs a new drm with rs690 support,
which is just getting cleaned up.

16 years agoi965: valid message length includes message header.
Xiang, Haihao [Fri, 25 Jan 2008 08:52:08 +0000 (16:52 +0800)]
i965: valid message length includes message header.

16 years agoi965: re-define the type of reg.loopcount.
Xiang, Haihao [Fri, 25 Jan 2008 08:38:38 +0000 (16:38 +0800)]
i965: re-define the type of reg.loopcount.
  avoid some issues such that 1 + (-2) gets a big
positive value.

16 years agoBufmgr cleanup from intel-batchbuffer branch of 2d driver.
Eric Anholt [Thu, 24 Jan 2008 21:00:13 +0000 (13:00 -0800)]
Bufmgr cleanup from intel-batchbuffer branch of 2d driver.

16 years agoClean up comments/dead code from relocation buffer change.
Eric Anholt [Thu, 24 Jan 2008 20:29:04 +0000 (12:29 -0800)]
Clean up comments/dead code from relocation buffer change.

16 years agoMerge commit 'airlied/i915-ttm-cfu'
Eric Anholt [Thu, 24 Jan 2008 20:26:29 +0000 (12:26 -0800)]
Merge commit 'airlied/i915-ttm-cfu'

This requires current DRM which changes the relocation buffer from being a
buffer object to plain malloced memory.

16 years agoBring in previous 7.0.1/2 release notes, added Cell driver page.
Brian [Thu, 24 Jan 2008 16:15:31 +0000 (09:15 -0700)]
Bring in previous 7.0.1/2 release notes, added Cell driver page.

16 years agoi915: move to using copy from user for relocations
Dave Airlie [Thu, 24 Jan 2008 04:38:50 +0000 (14:38 +1000)]
i915: move to using copy from user for relocations

16 years agoAssorted patches for miniglx/linux-solo (Gavin Li <codeview@gmail.com>)
Brian [Wed, 23 Jan 2008 23:03:19 +0000 (16:03 -0700)]
Assorted patches for miniglx/linux-solo (Gavin Li <codeview@gmail.com>)

16 years agoFix some issues with glDrawBuffer(GL_NONE), bug 14198
Brian [Wed, 23 Jan 2008 04:11:55 +0000 (21:11 -0700)]
Fix some issues with glDrawBuffer(GL_NONE), bug 14198

Set _ColorDrawBuffers[0] = NULL if no renderbuffers enabled.
Check that _ColorDrawBuffers[0] is non-null before dereferencing in a few places.

16 years agoglxinfo: Fix multisample visual reporting.
Adam Jackson [Tue, 22 Jan 2008 19:57:20 +0000 (14:57 -0500)]
glxinfo: Fix multisample visual reporting.

strstr() == 0 tests for the string _not_ being present.  Originally
Red Hat bug #351871.

16 years ago[intel] Clean up references to screen buffer metrics.
Kristian Høgsberg [Tue, 22 Jan 2008 17:13:16 +0000 (12:13 -0500)]
[intel] Clean up references to screen buffer metrics.

The screen wide info such as pitch and cpp are obsoleted by the FBO
changes, so clean up the last few references to those, except for
setting up the legacy screen regions.

16 years agofix pc vs. gc->ps usage (bug 14197)
Brian [Tue, 22 Jan 2008 15:40:29 +0000 (08:40 -0700)]
fix pc vs. gc->ps usage (bug 14197)

16 years agoDon't build yuvrect_client by default
Brian [Tue, 22 Jan 2008 14:56:26 +0000 (07:56 -0700)]
Don't build yuvrect_client by default

Added 'extra' rule to build optional yuvrect_client, shape, xdemo programs

16 years agoremove unused var
Brian [Tue, 22 Jan 2008 14:54:08 +0000 (07:54 -0700)]
remove unused var

16 years agodefine M_PI if needed
Alexey Sokolov [Tue, 22 Jan 2008 14:43:43 +0000 (07:43 -0700)]
define M_PI if needed

16 years ago[965] Fix WM unit cache keying that broke line stipple and polygon offset.
Eric Anholt [Sun, 20 Jan 2008 07:22:45 +0000 (23:22 -0800)]
[965] Fix WM unit cache keying that broke line stipple and polygon offset.

16 years ago[intel] Fix memory leak with fake bufmgr.
Eric Anholt [Fri, 18 Jan 2008 19:39:35 +0000 (11:39 -0800)]
[intel] Fix memory leak with fake bufmgr.

16 years ago[965] Do a little bit rotation in state hash to reduce collisions.
Eric Anholt [Fri, 18 Jan 2008 18:15:28 +0000 (10:15 -0800)]
[965] Do a little bit rotation in state hash to reduce collisions.

This was around 3% improvement in OA.

16 years ago[intel] Use a static array for the validation list instead of a linked list.
Eric Anholt [Sat, 12 Jan 2008 02:54:31 +0000 (18:54 -0800)]
[intel] Use a static array for the validation list instead of a linked list.

Around 10% of a CPU was being wasted to create the linked list which we
threw out immediately after passing it to the kernel.

16 years ago[intel] Make the no_rast option be standard driconf instead of INTEL_NO_RAST.
Eric Anholt [Thu, 17 Jan 2008 22:23:04 +0000 (14:23 -0800)]
[intel] Make the no_rast option be standard driconf instead of INTEL_NO_RAST.

16 years ago[i915] Fix driver from cliprects changes, and clean up state emission.
Eric Anholt [Mon, 14 Jan 2008 19:03:05 +0000 (11:03 -0800)]
[i915] Fix driver from cliprects changes, and clean up state emission.

The fix for pageflipping with cliprects ended up causing a batch flush at
an inopportune time, which is fixed by moving it up.

Additionally, the recovery code for handling batch wraps at bad times is
replaced by just checking for the space up front, and using a no_batch_wrap
assert like on 965 to make sure that we weren't wrong about how much space that
was.

16 years ago[965] Fix whitespace in c9b1fef0c9c5018efd825c42782f19ad0618696a
Eric Anholt [Thu, 17 Jan 2008 19:45:37 +0000 (11:45 -0800)]
[965] Fix whitespace in c9b1fef0c9c5018efd825c42782f19ad0618696a

16 years ago[965] Fix potential segfaults from bad realloc.
Eric Anholt [Thu, 17 Jan 2008 19:25:04 +0000 (11:25 -0800)]
[965] Fix potential segfaults from bad realloc.

C has no order of evaluation restrictions on function arguments, so we
attempted to realloc from new-size to new-size.

16 years agoi965: always call dri_emit_reloc when creating clip unit state
Xiang, Haihao [Thu, 17 Jan 2008 03:05:24 +0000 (11:05 +0800)]
i965: always call dri_emit_reloc when creating clip unit state

This fix ensures it gets the starting location of the clip program
if a clip unit state is same as a unit which is created when metaops
is actived and it doesn't impact metaops because the clip state offset
isn't emitted when metaops is actived.

16 years ago[965] Fix inversion of SLT/SGE results in vertex programs.
Eric Anholt [Wed, 16 Jan 2008 23:07:10 +0000 (15:07 -0800)]
[965] Fix inversion of SLT/SGE results in vertex programs.

The WM code had this right, so copy its behavior.  This reverts a flipping
of the arguments to SLT in brw_vs_tnl which came in with the GLSL code that
probably occurred to work around the flipped results, and brings the code back
in line with t_vp_build.c.

16 years ago[965] Rename depth_mask in CC key to depth_write, since it's a boolean enable.
Eric Anholt [Wed, 16 Jan 2008 18:50:28 +0000 (10:50 -0800)]
[965] Rename depth_mask in CC key to depth_write, since it's a boolean enable.

16 years ago[965] Fix the type of alpha_ref in CC key, fixing ppracer rendering.
Eric Anholt [Wed, 16 Jan 2008 18:46:35 +0000 (10:46 -0800)]
[965] Fix the type of alpha_ref in CC key, fixing ppracer rendering.