platform/upstream/mesa.git
11 years agoilo: add debug flags settable through ILO_DEBUG 23/6223/1
Chia-I Wu [Wed, 12 Dec 2012 20:24:40 +0000 (04:24 +0800)]
ilo: add debug flags settable through ILO_DEBUG

11 years agoilo: new pipe driver for Intel GEN6+ 22/6222/1
Chia-I Wu [Wed, 12 Dec 2012 20:07:16 +0000 (04:07 +0800)]
ilo: new pipe driver for Intel GEN6+

This commit adds some boilerplate code.  The header files found under include/
are copied from i965.

11 years agowinsys/intel: new winsys for intel 21/6221/1
Chia-I Wu [Wed, 12 Dec 2012 19:52:50 +0000 (03:52 +0800)]
winsys/intel: new winsys for intel

This is a wrapper for libdrm_intel to allow the pipe driver to stay OS
agnostic.

11 years agogallivm: Fix trivial out-of-bounds indirection in lp_build_cube_lookup(). 20/6220/1
José Fonseca [Fri, 26 Apr 2013 07:43:00 +0000 (08:43 +0100)]
gallivm: Fix trivial out-of-bounds indirection in lp_build_cube_lookup().

Courtesy of clang:

  src/gallium/auxiliary/gallivm/lp_bld_sample.c:1483:10: warning: array index of '2' indexes past the end of an array (that contains 2 elements) [-Warray-bounds]
           tmp[2] = lp_build_swizzle_aos(coord_bld, ddx_ddy[1], swizzle02);
           ^   ~
  src/gallium/auxiliary/gallivm/lp_bld_sample.c:1430:10: note: array 'tmp' declared here
           LLVMValueRef ddx_ddy[2], tmp[2], rho_vec;
           ^
  src/gallium/auxiliary/gallivm/lp_bld_sample.c:1487:56: warning: array index of '2' indexes past the end of an array (that contains 2 elements) [-Warray-bounds]
              rho_vec = lp_build_add(coord_bld, rho_vec, tmp[2]);
                                                       ^   ~
  src/gallium/auxiliary/gallivm/lp_bld_sample.c:1430:10: note: array 'tmp' declared here
           LLVMValueRef ddx_ddy[2], tmp[2], rho_vec;
           ^
  src/gallium/auxiliary/gallivm/lp_bld_sample.c:1491:56: warning: array index of '2' indexes past the end of an array (that contains 2 elements) [-Warray-bounds]
              rho_vec = lp_build_max(coord_bld, rho_vec, tmp[2]);
                                                       ^   ~
  src/gallium/auxiliary/gallivm/lp_bld_sample.c:1430:10: note: array 'tmp' declared here
           LLVMValueRef ddx_ddy[2], tmp[2], rho_vec;
           ^

11 years agoi965/vs: Add support for LRP instruction. 19/6219/1
Matt Turner [Thu, 25 Apr 2013 18:03:38 +0000 (11:03 -0700)]
i965/vs: Add support for LRP instruction.

Only 13 affected programs in shader-db, but they were all helped.

total instructions in shared programs: 368877 -> 368851 (-0.01%)
instructions in affected programs:     1576 -> 1550 (-1.65%)

Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agoi965/vs: Add a function to fix-up uniform arguments for 3-src insts. 18/6218/1
Matt Turner [Thu, 25 Apr 2013 18:02:02 +0000 (11:02 -0700)]
i965/vs: Add a function to fix-up uniform arguments for 3-src insts.

Three-source instructions have a vertical stride overloaded to 4, which
prevents directly using vec4 uniforms as arguments. Instead we need to
insert a MOV instruction to do the replication for the three-source
instruction.

With this in place, we can use three-source instructions in the vertex
shader. While some thought needs to go into deciding whether its better
to use a three-source instruction rather than a sequence of equivalent
instructions (when one or more sources are uniforms or immediates), this
will allow us to skip a lot of ugly lowering code and use the BFE and
BFI2 instructions directly.

Reviewed-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agowinsys/radeon: consolidate tracing into winsys v2 17/6217/1
Jerome Glisse [Tue, 23 Apr 2013 23:22:33 +0000 (19:22 -0400)]
winsys/radeon: consolidate tracing into winsys v2

This move the tracing timeout and printing into winsys and add
an debug environement variable for it (R600_DEBUG=trace_cs).

Lot of file touched because of winsys API changes.

v2: Do not write lockup file if ib uniq id does not match last one

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
11 years agor600g/compute: Removed unused and untested code 16/6216/1
Tom Stellard [Mon, 22 Apr 2013 16:12:07 +0000 (09:12 -0700)]
r600g/compute: Removed unused and untested code

There was a lot of code in evergreen_compute_internal.c that was not
being used at all and most of it was duplicating code from other parts
of the driver.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
11 years agor600g/compute: Use a constant buffer to store kernel parameters v2 15/6215/1
Tom Stellard [Mon, 22 Apr 2013 15:38:40 +0000 (08:38 -0700)]
r600g/compute: Use a constant buffer to store kernel parameters v2

v2:
  - Fix usage of set_constant_buffer()
  - Fix typo in comment

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
11 years agor600g: Add evergreen_emit_cs_constant_buffers() v2 14/6214/1
Tom Stellard [Tue, 23 Apr 2013 03:06:54 +0000 (20:06 -0700)]
r600g: Add evergreen_emit_cs_constant_buffers() v2

v2:
  - Bump R600_NUM_ATOMS

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
11 years agor600g/compute: Don't use radeon_winsys::buffer_wait() after dispatching a kernel 13/6213/1
Tom Stellard [Mon, 22 Apr 2013 15:34:18 +0000 (08:34 -0700)]
r600g/compute: Don't use radeon_winsys::buffer_wait() after dispatching a kernel

The state tracker should be responsible for waiting for the kernel to
finish.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
11 years agor600g/compute: Fix input buffer size calculation 12/6212/1
Tom Stellard [Mon, 22 Apr 2013 14:32:10 +0000 (07:32 -0700)]
r600g/compute: Fix input buffer size calculation

Buffer size should be in bytes not dwords.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
11 years agolinux: Don't emit a .note.ABI-tag section anymore (#26663) 11/6211/1
Adam Jackson [Tue, 23 Apr 2013 18:07:33 +0000 (14:07 -0400)]
linux: Don't emit a .note.ABI-tag section anymore (#26663)

We don't support pre-2.6 kernels anyway - the install docs say 2.6.28
for DRI - and apparently this confuses ld.so's sorting when multiple
libGLs are installed.  Just remove it.

Note: this is a candidate for the stable branches.

Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Adam Jackson <ajax@redhat.com>
11 years agofreedreno: use writecombine buffers 10/6210/1
Rob Clark [Thu, 25 Apr 2013 19:00:58 +0000 (15:00 -0400)]
freedreno: use writecombine buffers

Better than uncached for writes, which are common for vertex buffer
upload, etc.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agofreedreno: don't patch and re-emit same shader as much 09/6209/1
Rob Clark [Thu, 25 Apr 2013 15:17:02 +0000 (11:17 -0400)]
freedreno: don't patch and re-emit same shader as much

New textures or vertex buffers don't always require patching and
re-emitting the shaders.  So do a better job of figuring out when we
actually have to patch the shader.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agoi965: Avoid recompiles for fragment clamping on non-clamping APIs. 08/6208/1
Eric Anholt [Mon, 15 Apr 2013 23:44:55 +0000 (16:44 -0700)]
i965: Avoid recompiles for fragment clamping on non-clamping APIs.

Removes 75/78 state-dependent recompiles in GLB2.7 (the remaining 3 are
due to FBO-rendering size predictions).  We currently expose
GL_ARB_color_buffer_float on GL core, so we may mis-predict there, but I'm
about to send a patch for removing that silly extension in that case.

Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agoradeonsi: add new SI pci ids 07/6207/1
Alex Deucher [Thu, 25 Apr 2013 18:22:46 +0000 (14:22 -0400)]
radeonsi: add new SI pci ids

Note: this is a candidate for the 9.1 branch.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agor600g: add new richland pci ids 06/6206/1
Alex Deucher [Thu, 25 Apr 2013 18:21:15 +0000 (14:21 -0400)]
r600g: add new richland pci ids

Note: this is a candidate for the stable branches.

Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agodraw: Yield zeros for LLVM fetches of non-existing vertex elements. 05/6205/1
José Fonseca [Thu, 25 Apr 2013 15:16:21 +0000 (16:16 +0100)]
draw: Yield zeros for LLVM fetches of non-existing vertex elements.

If a bug in an app/stater-tacker causes vertex buffer to fetch vertex
elements that are not bound, simply return zeros instead of crashing.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agotrace: Only close trace files on exit. 04/6204/1
José Fonseca [Thu, 25 Apr 2013 13:18:33 +0000 (14:18 +0100)]
trace: Only close trace files on exit.

Many applications don't exit cleanly, others may create and destroy a
screen multiple times, so we only write </trace> tag and close at exit
time.

11 years agograw: Set the vertex shader constant buffer. 03/6203/1
José Fonseca [Thu, 25 Apr 2013 13:06:50 +0000 (14:06 +0100)]
graw: Set the vertex shader constant buffer.

We were setting the fragment shader, which wasn't needed.

11 years agograw: Simple utilities to dump and disassemble TGSI tokens. 02/6202/1
José Fonseca [Wed, 24 Apr 2013 12:08:46 +0000 (13:08 +0100)]
graw: Simple utilities to dump and disassemble TGSI tokens.

Useful for core dumps, where calling tgsi_dump() from gdb is not an
alternative.

11 years agoscons: Support clang. 01/6201/1
José Fonseca [Wed, 24 Apr 2013 21:02:18 +0000 (22:02 +0100)]
scons: Support clang.

clang is supports most gcc options / extensions, with a some exceptions.

The biggest advantage of using clang is that compilation times are much
short.

One can tell scons to use clang when building by invoking it as

   CC=clang CXX=clang++ scons libgl-xlib

11 years agoutil/u_sse: Fix _mm_shuffle_epi8 prototype for clang. 00/6200/1
José Fonseca [Wed, 24 Apr 2013 20:58:20 +0000 (21:58 +0100)]
util/u_sse: Fix _mm_shuffle_epi8 prototype for clang.

Clang does not support __artificial__. Instead match precisely what's
in the clang headers.

11 years agoscons: Remove redundant code. 99/6199/1
José Fonseca [Wed, 24 Apr 2013 21:00:32 +0000 (22:00 +0100)]
scons: Remove redundant code.

-fvisibility=hidden is already elsewhere for the whole tree.

11 years agomesa: fix bogus comment about PrimitiveRestart fields 98/6198/1
Chris Forbes [Mon, 22 Apr 2013 05:07:57 +0000 (17:07 +1200)]
mesa: fix bogus comment about PrimitiveRestart fields

Signed-off-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agoi965: report correct sample positions 97/6197/1
Chris Forbes [Tue, 23 Apr 2013 18:44:02 +0000 (06:44 +1200)]
i965: report correct sample positions

From low to high bits, the sample positions are packed y0,x0,y1,x1...

Fixes arb_texture_multisample-sample-position piglit.

Signed-off-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Paul Berry <stereotype441@gmail.com>
11 years agofreedreno: fix bogus IMM const reg index 96/6196/1
Rob Clark [Wed, 24 Apr 2013 21:45:00 +0000 (17:45 -0400)]
freedreno: fix bogus IMM const reg index

We were assigning incorrect const register for immediates, and
potentially writing immediate const to the wrong location.  This fixes
an incorrect-rendering bug with xonotic.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agofreedreno: clear fixes and debugging 95/6195/1
Rob Clark [Wed, 24 Apr 2013 14:50:51 +0000 (10:50 -0400)]
freedreno: clear fixes and debugging

Set a few extra registers to make sure we are in proper state for
clearing.  And also add some debug options to mark all state dirty in
clear and gmem operations to aid in debugging.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agofreedreno: fix texture fetch type 94/6194/1
Rob Clark [Wed, 24 Apr 2013 14:48:59 +0000 (10:48 -0400)]
freedreno: fix texture fetch type

There is a bit we need to set for 2D vs 3D fetch, to tell the hw whether
there are two or there valid input components.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agofreedreno: fix temp register usage 93/6193/1
Rob Clark [Wed, 24 Apr 2013 14:44:56 +0000 (10:44 -0400)]
freedreno: fix temp register usage

The previous approach of using the dst register as an intermediate
temporary doesn't work in a lot of cases.  For example, if the dst
register is the same as one of the src registers.

For now, just simplify it and always allocate a new register to use as
an intermediate.  In some cases this will result in more registers used
than required.  I think the best solution would be to implement an
optimization pass to reduce the number of registers used, which would
also solve the problem we have now of not being able to use GPRs that
are assigned for TGSI_FILE_INPUT.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agofreedreno: add noop driver 92/6192/1
Rob Clark [Tue, 23 Apr 2013 15:20:25 +0000 (11:20 -0400)]
freedreno: add noop driver

It is useful for debugging.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agofreedreno: use u_math macros/helpers more 91/6191/1
Rob Clark [Mon, 22 Apr 2013 17:55:14 +0000 (13:55 -0400)]
freedreno: use u_math macros/helpers more

Get rid of a few self-defined macros:
  ALIGN() -> align()
  min() -> MIN2()
  max() -> MAX2()

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agofreedreno: implement fd_screen_destroy() 90/6190/1
Rob Clark [Mon, 22 Apr 2013 17:42:55 +0000 (13:42 -0400)]
freedreno: implement fd_screen_destroy()

Opps, didn't notice that I had left it stubbed out.

Also, make things fail a bit more gracefully when things go wrong.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agofreedreno: set SWAP bit based on format 89/6189/1
Rob Clark [Mon, 22 Apr 2013 17:21:21 +0000 (13:21 -0400)]
freedreno: set SWAP bit based on format

Really this should be set based on buffer format, not on color vs
depth/stencil.  Probably there should be more formats that set the bit
as we add support for more render target formats.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agoradeon/llvm: Fix segfault with a specifc libelf implementation 88/6188/1
Tom Stellard [Tue, 23 Apr 2013 15:08:30 +0000 (08:08 -0700)]
radeon/llvm: Fix segfault with a specifc libelf implementation

The libelf implementation that is distributed here:
http://www.mr511.de/software/english.html
requires calling elf_version() prior to calling elf_memory()

Tested-by: Michel Dänzer <michel.daenzer@amd.com>
11 years agor600g: use CP DMA for buffer clears on evergreen+ 87/6187/1
Alex Deucher [Wed, 24 Apr 2013 16:26:52 +0000 (12:26 -0400)]
r600g: use CP DMA for buffer clears on evergreen+

Lighter weight then using streamout.  Only evergreen
and newer asics support embedded data as src with
CP DMA.

Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Marek Olšák <maraeo@gmail.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
11 years agoi965/gen7: fix encoding of (huge) surface size for BRW_SURFACE_BUFFER 86/6186/1
Chia-I Wu [Wed, 10 Apr 2013 13:32:30 +0000 (21:32 +0800)]
i965/gen7: fix encoding of (huge) surface size for BRW_SURFACE_BUFFER

Unlike GEN6, the bits of entry count are distributed like this

  width  = (entry_count & 0x0000007f);       /* bits [6:0] */
  height = (entry_count & 0x001fff80) >> 7;  /* bits [20:7] */
  depth  = (entry_count & 0x7fe00000) >> 21; /* bits [30:21] */

The maximum entry count is still limited to 2^27.

This was noted while going over the PRM.  No test is impacted, because
1<<20 (the bit that moved) is much larger than GL_UNIFORM_BLOCK_MAX_SIZE,
GL_MAX_TEXTURE_BUFFER_SIZE, or MAX_*_UNIFORM_COMPONENTS.

v2: Explain more in the commit message (by anholt)

Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agoi965/gen7: fix 3DSTATE_LINE_STIPPLE_PATTERN 85/6185/1
Chia-I Wu [Wed, 10 Apr 2013 13:32:13 +0000 (21:32 +0800)]
i965/gen7: fix 3DSTATE_LINE_STIPPLE_PATTERN

The inverse repeat count should taks up bits 31:15 and is in U1.16.  Fixes
the "Restarting lines within a single Begin/End block" subtest of piglit
linestipple, and gets the other failing subtests much closer to passing.

v2: Rewrite commit message with more detailed piglit info (by anholt)

Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agoi965: fix SURFACE_STATE dumping 84/6184/1
Chia-I Wu [Wed, 10 Apr 2013 13:31:56 +0000 (21:31 +0800)]
i965: fix SURFACE_STATE dumping

Wrong fields were used when dumping width and height.

Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agoi965: Remove strange comments about math functions. 83/6183/1
Matt Turner [Tue, 23 Apr 2013 04:43:34 +0000 (21:43 -0700)]
i965: Remove strange comments about math functions.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agoi965: Remove traces of nonexistent TAN math function. 82/6182/1
Matt Turner [Tue, 23 Apr 2013 04:33:38 +0000 (21:33 -0700)]
i965: Remove traces of nonexistent TAN math function.

Never existed? At least never supported. Doesn't appear in 965, G45,
or ILK documentation.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agoglsl: Teach basic block analysis about break/continue/discard. 81/6181/1
Paul Berry [Mon, 22 Apr 2013 21:02:00 +0000 (14:02 -0700)]
glsl: Teach basic block analysis about break/continue/discard.

Previously, the only kind of ir_jump that would terminate a basic
block was "return".  However, the other possible types of ir_jump
("break", "continue", and "discard") should terminate a basic block
too.  This patch modifies basic block analysis so that it terminates a
basic block on any type of ir_jump, not just ir_return.

Fixes piglit test dead-code-break-interaction.shader_test.

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agoglsl: Add virtual function ir_instruction::as_jump() 80/6180/1
Paul Berry [Mon, 22 Apr 2013 20:59:17 +0000 (13:59 -0700)]
glsl: Add virtual function ir_instruction::as_jump()

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agor600g/llvm: Pass struct r600_bytecode to r600_llvm_compile 79/6179/1
Tom Stellard [Wed, 24 Apr 2013 03:08:57 +0000 (20:08 -0700)]
r600g/llvm: Pass struct r600_bytecode to r600_llvm_compile

This way we don't need to update the function signature everytime we
emit a new config value.  This also fixes the build with
--enable-opencl.

11 years agowinsys/sw/xlib: Prevent shared memory segment leakage. 78/6178/1
José Fonseca [Wed, 24 Apr 2013 09:49:57 +0000 (10:49 +0100)]
winsys/sw/xlib: Prevent shared memory segment leakage.

Running piglit with this was causing all sort of weird stuff happening
to my desktop (Chromium webpages become blank, Qt Creator flickered,
etc).  I tracked this down to shared memory segment leakage when GL is
not shutdown properly. The segments can be seen running `ipcs` and
looking for nattch==0.

This changes fixes this by calling shmctl(IPC_RMID) soon after creation
(which does not remove the segment immediately, but simply marks it for
removal when no more processes are attached).

This matches src/mesa/drivers/x11/xm_buffer.c behaviour.

v2:
- move shmctl(IPC_RMID) after XShmAttach() for *BSD, per Chris Wilson
- remove stray debug printfs, spotted by Ian Romanick

NOTE: This is a candidate for stable branches.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agodraw/gs: preserve leading vertex info for gs 77/6177/1
Zack Rusin [Tue, 23 Apr 2013 00:44:21 +0000 (20:44 -0400)]
draw/gs: preserve leading vertex info for gs

We need to handle the leading vertex information when
assembling primitives for the geometry shader otherwise
the resulting triangles will have vertices at incorrect
input locations.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agor200: fix build regression introduced with 9a32203e1618486e87c7baf494134e05f0e38cf3 76/6176/1
Laurent Carlier [Wed, 24 Apr 2013 10:47:18 +0000 (12:47 +0200)]
r200: fix build regression introduced with 9a32203e1618486e87c7baf494134e05f0e38cf3

Signed-off-by: Laurent Carlier <lordheavym@gmail.com>
Signed-off-by: Marek Olšák <maraeo@gmail.com>
11 years agoradeonsi: cleanup disabling tiling for UVD v3 75/6175/1
Christian König [Sat, 20 Apr 2013 11:19:33 +0000 (13:19 +0200)]
radeonsi: cleanup disabling tiling for UVD v3

Should fix: https://bugs.freedesktop.org/show_bug.cgi?id=63702

v2: add a comment that this is just a workaround
v3: fix typo in comment

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
11 years agoegl/dri2: Fix min/max swap interval of configs 74/6174/1
Chad Versace [Tue, 23 Apr 2013 02:17:48 +0000 (04:17 +0200)]
egl/dri2: Fix min/max swap interval of configs

The commit below exposed a bug in dri2_add_config.

    commit 3998f8c6b5da1a223926249755e54d8f701f81ab
    Author: Ralf Jung <post@ralfj.de>
    Date:   Tue Apr 9 14:09:50 2013 +0200

egl/x11: Fix initialisation of swap_interval

This little code snippet near the bottom of dri2_add_config,

    if (double_buffer) {
       ...
       conf->base.MinSwapInterval = dri2_dpy->min_swap_interval;
       conf->base.MaxSwapInterval = dri2_dpy->max_swap_interval;
    }

it never did what it claimed to do. The assignment never changed the value
of conf->base.MaxSwapInterval, because dri2_dpy->max_swap_interval was,
until the above exposing commit, unitialized here. That is,
conf->base.MaxSwapInterval was 0 before and after assignment. Ditto for
the min swap interval.

Above the troublesome code snippet, the call to _eglFilterArray rejects
the config as unmatching if its swap interval bounds differ from the base
config's.  Before the exposing commit, at the call to _eglFilterArray, the
swap interval bounds were always [0,0], and hence no config was rejected
due to swap interval.

After the exposing commit, _eglFilterArray incorrectly rejected some
configs, which prevented dri2_egl_config::dri_double_config from getting
set for the rejected config, which resulted in a NULL pointer getting
passed into dri2CreateNewDrawable, and then segfault.

The solution: set the swap interval bounds before _eglFilterArray.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=63447
Tested-by: Lu Hua <huax.lu@intel.com>
Signed-off-by: Chad Versace <chad.versace@linux.intel.com>
11 years agomesa: Add unpack functions for A/I/L/LA [U]INT8/16/32 formats. 73/6173/1
Kenneth Graunke [Tue, 23 Apr 2013 06:52:22 +0000 (23:52 -0700)]
mesa: Add unpack functions for A/I/L/LA [U]INT8/16/32 formats.

NOTE: This is a candidate for stable branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=63569
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Add unpack functions for R/RG/RGB [U]INT8/16/32 formats. 72/6172/1
Kenneth Graunke [Tue, 23 Apr 2013 06:37:06 +0000 (23:37 -0700)]
mesa: Add unpack functions for R/RG/RGB [U]INT8/16/32 formats.

NOTE: This is a candidate for stable branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=63569
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Add an unpack function for ARGB2101010_UINT. 71/6171/1
Kenneth Graunke [Tue, 23 Apr 2013 06:13:47 +0000 (23:13 -0700)]
mesa: Add an unpack function for ARGB2101010_UINT.

v2: Remove extra parenthesis (suggested by Brian).

NOTE: This is a candidate for stable branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=63569
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Fix unpack function for ETC2_SRGB8_PUNCHTHROUGH_ALPHA1. 70/6170/1
Kenneth Graunke [Tue, 23 Apr 2013 05:58:49 +0000 (22:58 -0700)]
mesa: Fix unpack function for ETC2_SRGB8_PUNCHTHROUGH_ALPHA1.

We accidentally set MESA_FORMAT_ETC2_RGB8_PUNCHTHROUGH_ALPHA1 twice,
rather than setting the RGB8 and SRGB8 formats.

NOTE: This is a candidate for stable branches.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=63569
Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Fix up some final license word wrapping issues by hand. 69/6169/1
Kenneth Graunke [Sun, 21 Apr 2013 20:55:37 +0000 (13:55 -0700)]
mesa: Fix up some final license word wrapping issues by hand.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Restore 78-column wrapping of license text in C++-style comments. 68/6168/1
Kenneth Graunke [Sun, 21 Apr 2013 20:52:08 +0000 (13:52 -0700)]
mesa: Restore 78-column wrapping of license text in C++-style comments.

The previous commit introduced extra words, breaking the formatting.

This text transformation was done automatically via the following shell
command:
$ git grep 'THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY' | sed 's/:.*$//' | xargs -I {} sh -c 'vim -e -s {} < vimscript2

where 'vimscript2' is a file containing:
/THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY/;/^ *$/ !fmt -w 78 -p '// '
:wq

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Restore 78-column wrapping of license text in C-style comments. 67/6167/1
Kenneth Graunke [Sun, 21 Apr 2013 20:46:48 +0000 (13:46 -0700)]
mesa: Restore 78-column wrapping of license text in C-style comments.

The previous commit introduced extra words, breaking the formatting.

This text transformation was done automatically via the following shell
command:
$ git grep 'THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY' | sed 's/:.*$//' | xargs -I {} sh -c 'vim -e -s {} < vimscript

where 'vimscript' is a file containing:
/THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY/;/\*\// !fmt -w 78 -p ' * '
:wq

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Add "OR COPYRIGHT HOLDERS" to license text disclaiming liability. 66/6166/1
Kenneth Graunke [Sun, 21 Apr 2013 20:09:39 +0000 (13:09 -0700)]
mesa: Add "OR COPYRIGHT HOLDERS" to license text disclaiming liability.

This brings the license text in line with the MIT License as published
on the Open Source Initiative website:

http://opensource.org/licenses/mit-license.php

Generated automatically be the following shell command:
$ git grep 'THE AUTHORS BE LIABLE' | sed 's/:.*$//g' | xargs -I '{}' \
  sed -i 's/THE AUTHORS/THE AUTHORS OR COPYRIGHT HOLDERS/' {}

This introduces some wrapping issues, to be fixed in the next commit.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Change "BRIAN PAUL OR IBM" to "THE AUTHORS" in license text. 65/6165/1
Kenneth Graunke [Sun, 21 Apr 2013 20:01:45 +0000 (13:01 -0700)]
mesa: Change "BRIAN PAUL OR IBM" to "THE AUTHORS" in license text.

See previous commit for the rationale.  These weren't caught by the
automatic conversion due to the "OR IBM" addition.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Change "BRIAN PAUL" to "THE AUTHORS" in license text. 64/6164/1
Kenneth Graunke [Sun, 21 Apr 2013 19:56:09 +0000 (12:56 -0700)]
mesa: Change "BRIAN PAUL" to "THE AUTHORS" in license text.

Generated automatically be the following shell command:
$ git grep 'BRIAN PAUL BE LIABLE' | sed 's/:.*$//g' | xargs -I '{}' \
  sed -i 's/BRIAN PAUL/THE AUTHORS/' {}

The intention here is to protect all authors, not just Brian Paul.  I
believe that was already the sensible interpretation, but spelling it
out is probably better.

More practically, it also prevents people from accidentally copy &
pasting the license into a new file which says Brian is not liable when
he isn't even one of the authors.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: make _mesa_save_vtxfmt_init() static 63/6163/1
Brian Paul [Tue, 23 Apr 2013 15:19:07 +0000 (09:19 -0600)]
mesa: make _mesa_save_vtxfmt_init() static

It's called from nowhere else.

Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agodocs: document issue with Viewperf proe-05 test 6 62/6162/1
Brian Paul [Tue, 23 Apr 2013 19:18:41 +0000 (13:18 -0600)]
docs: document issue with Viewperf proe-05 test 6

11 years agomesa: use new _mesa_inside_dlist_begin_end() function 61/6161/1
Brian Paul [Tue, 23 Apr 2013 16:57:45 +0000 (10:57 -0600)]
mesa: use new _mesa_inside_dlist_begin_end() function

Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agomesa: use new _mesa_inside_begin_end() function 60/6160/1
Brian Paul [Tue, 23 Apr 2013 16:53:00 +0000 (10:53 -0600)]
mesa: use new _mesa_inside_begin_end() function

Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agomesa: remove unused opcodes AND, DP2A, NOT, NRM3, NRM4, OR, PRINT, XOR 59/6159/1
Marek Olšák [Sun, 14 Apr 2013 17:59:06 +0000 (19:59 +0200)]
mesa: remove unused opcodes AND, DP2A, NOT, NRM3, NRM4, OR, PRINT, XOR

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: don't flush vertices and don't flag _NEW_COLOR in ClearColor, ClearIndex 58/6158/1
Marek Olšák [Mon, 15 Apr 2013 01:41:43 +0000 (03:41 +0200)]
mesa: don't flush vertices and don't flag _NEW_COLOR in ClearColor, ClearIndex

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: don't flush vertices and don't flag _NEW_COLOR for GL_CLAMP_READ_COLOR 57/6157/1
Marek Olšák [Mon, 15 Apr 2013 01:41:43 +0000 (03:41 +0200)]
mesa: don't flush vertices and don't flag _NEW_COLOR for GL_CLAMP_READ_COLOR

There used to be a derived state _ClampReadColor, so setting _NEW_COLOR
made sense. The state is gone now.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: don't flag _NEW_DEPTH in Begin/EndQuery if driver implements the functions 56/6156/1
Marek Olšák [Mon, 15 Apr 2013 01:41:43 +0000 (03:41 +0200)]
mesa: don't flag _NEW_DEPTH in Begin/EndQuery if driver implements the functions

We don't want to set the flag for Gallium.

I think only swrast needs the flag to be set for occlusion queries.

v2: fix stats_wm updates in i965

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: don't flush vertices and don't flag _NEW_DEPTH in ClearDepth 55/6155/1
Marek Olšák [Mon, 15 Apr 2013 01:40:06 +0000 (03:40 +0200)]
mesa: don't flush vertices and don't flag _NEW_DEPTH in ClearDepth

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: don't flush and don't flag _NEW_STENCIL in ClearStencil, ActiveStencilFace 54/6154/1
Marek Olšák [Mon, 15 Apr 2013 00:54:58 +0000 (02:54 +0200)]
mesa: don't flush and don't flag _NEW_STENCIL in ClearStencil, ActiveStencilFace

The functions don't affect driver state. There is no code that would rely
on vertices being flushed prior to changing the states, and no code that
would check for _NEW_STENCIL before using the states.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: don't set _NEW_BUFFERS in GenerateMipmap and BlitFramebuffer 53/6153/1
Marek Olšák [Sun, 14 Apr 2013 20:31:50 +0000 (22:31 +0200)]
mesa: don't set _NEW_BUFFERS in GenerateMipmap and BlitFramebuffer

both functions don't change the framebuffer in any way
(if mesa_meta is not used)

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: remove _NEW_PACKUNPACK 52/6152/1
Marek Olšák [Mon, 15 Apr 2013 00:23:37 +0000 (02:23 +0200)]
mesa: remove _NEW_PACKUNPACK

No driver checks the flag. Nobody uses it.

I also removed the FLUSH_VERTICES calls, because PixelStorei has no effect
on rendering.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: convert _NEW_RASTERIZER_DISCARD to a driver flag 51/6151/1
Marek Olšák [Mon, 15 Apr 2013 11:00:54 +0000 (13:00 +0200)]
mesa: convert _NEW_RASTERIZER_DISCARD to a driver flag

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa,i965: use NewDriverState to communicate TFB state changes with the driver 50/6150/1
Marek Olšák [Sun, 14 Apr 2013 22:50:20 +0000 (00:50 +0200)]
mesa,i965: use NewDriverState to communicate TFB state changes with the driver

_NEW_TRANSFORM_FEEDBACK is not used by core Mesa, so it can be removed.
Instead, an new private flag is added to i965 to serve the same purpose.

If you're new to this:

* When creating a context. you can set private dirty flags
  in gl_context::DriverFlags, eg.:
    ctx->DriverFlags.NewStateX = BRW_NEW_STATE_X;

* When StateX is changed, core Mesa does:
    ctx->NewDriverState |= ctx->DriverFlags.NewStateX;

* When you have to draw, read and clear ctx->NewDriverState.

* Pros: not touching NewState, the driver decides the mapping between
  GL states and hw state groups, unlimited number of flags in core Mesa
  (still limited number of flags in the driver though)

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agomesa: remove redundant _NEW_BUFFERS setting in ReadBuffer 49/6149/1
Marek Olšák [Sun, 14 Apr 2013 20:24:09 +0000 (22:24 +0200)]
mesa: remove redundant _NEW_BUFFERS setting in ReadBuffer

already set by _mesa_readbuffer

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agost/mesa: add a simple path to BufferData if it only discards buffer contents 48/6148/1
Marek Olšák [Sun, 14 Apr 2013 21:28:36 +0000 (23:28 +0200)]
st/mesa: add a simple path to BufferData if it only discards buffer contents

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agost/mesa: depth-stencil-alpha state also depends on _NEW_BUFFERS 47/6147/1
Marek Olšák [Mon, 15 Apr 2013 12:39:58 +0000 (14:39 +0200)]
st/mesa: depth-stencil-alpha state also depends on _NEW_BUFFERS

because the code looks at the visual if there is a depth or stencil buffer
before enabling depth or stencil, respectively.

NOTE: This is a candidate for the stable branches.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agogallium: Replace gl_rasterization_rules with lower_left_origin and half_pixel_center. 46/6146/1
José Fonseca [Tue, 23 Apr 2013 18:40:05 +0000 (19:40 +0100)]
gallium: Replace gl_rasterization_rules with lower_left_origin and half_pixel_center.

Squashed commit of the following:

commit 04c5fa2cbb8e89d6f2fa5a75af1cca03b1f6b852
Author: José Fonseca <jfonseca@vmware.com>
Date:   Tue Apr 23 17:37:18 2013 +0100

    gallium: s/lower_left_origin/bottom_edge_rule/

commit 4dff4f64fa83b9737def136fffd161d55e4f1722
Author: José Fonseca <jfonseca@vmware.com>
Date:   Tue Apr 23 17:35:04 2013 +0100

    gallium: Move diagram to docs.

commit 442a63012c8c3c3797f45e03f2ca20ad5f399832
Author: James Benton <jbenton@vmware.com>
Date:   Fri May 11 17:50:55 2012 +0100

    gallium: Replace gl_rasterization_rules with lower_left_origin and half_pixel_center.

    This change is necessary to achieve correct results when using OpenGL
    FBOs.

Reviewed-by: Marek Olšák <maraeo@gmail.com>
11 years agor600g: initialize CMASK and HTILE with the GPU using streamout 45/6145/1
Marek Olšák [Sun, 21 Apr 2013 21:26:52 +0000 (23:26 +0200)]
r600g: initialize CMASK and HTILE with the GPU using streamout

This fixes a crash when a resource cannot be mapped to the CPU's address space
because it's too big.

This puts a global pipe_context in r600_screen, which is guarded by a mutex,
so that we can use pipe_context when there isn't one around.
Hopefully our multi-context support is solid.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
NOTE: This is a candidate for the 9.1 branch.

11 years agogallium/u_blitter: implement buffer clearing 44/6144/1
Marek Olšák [Sun, 21 Apr 2013 17:39:03 +0000 (19:39 +0200)]
gallium/u_blitter: implement buffer clearing

Although this might be useful for ARB_clear_buffer_object,
I need it for initializating resources in r600g.

Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
v2: comment cleanups

NOTE: This is a candidate for the 9.1 branch.

11 years agor600/llvm: Read stacksize from config header 43/6143/1
Vincent Lejeune [Fri, 19 Apr 2013 18:10:44 +0000 (20:10 +0200)]
r600/llvm: Read stacksize from config header

11 years ago/bin/bash: q : commande introuvable 42/6142/1
Vincent Lejeune [Sat, 13 Apr 2013 14:35:39 +0000 (16:35 +0200)]
/bin/bash: q : commande introuvable

11 years agoradeon/llvm: Fix build with LLVM >= r180063 41/6141/1
Tom Stellard [Tue, 23 Apr 2013 14:58:38 +0000 (10:58 -0400)]
radeon/llvm:  Fix build with LLVM >= r180063

11 years agogallivm: Fix build with LLVM >= r180063 40/6140/1
Tom Stellard [Tue, 23 Apr 2013 14:57:54 +0000 (10:57 -0400)]
gallivm: Fix build with LLVM >= r180063

11 years agodraw: use the prim count for ia primitives 39/6139/1
Zack Rusin [Mon, 22 Apr 2013 23:07:14 +0000 (19:07 -0400)]
draw: use the prim count for ia primitives

Number of vertices to fetch doesn't always equal the number of input
vertices. To correctly compute the number if IA primitives we need
to use the total number of input vertices, not only those that
need to be fetched.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agotgsi/scan: set correct input limits for geometry shader 38/6138/1
Zack Rusin [Sat, 20 Apr 2013 07:02:00 +0000 (00:02 -0700)]
tgsi/scan: set correct input limits for geometry shader

TGSI geometry shader input declerations are of the IN[][2] format
and the dimensions of the array have to be deduced from the input
primitive property.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agodraw: add code to reset instance dependent data 37/6137/1
Zack Rusin [Fri, 19 Apr 2013 23:51:27 +0000 (16:51 -0700)]
draw: add code to reset instance dependent data

We want to be able to reset certain parts of the pipeline,
in particular the input primitive index, but only either with
seperate invocations of the draw_vbo or new instances. In all
other cases (e.g. new invocations due to primitive restart)
that data needs to be preserved. Add a function through which
we can reset instance dependent data.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agosoftpipe: fix streamout with an emptry geometry shader 36/6136/1
Zack Rusin [Fri, 19 Apr 2013 19:51:07 +0000 (12:51 -0700)]
softpipe: fix streamout with an emptry geometry shader

Same approach as in the llvmpipe, if the geometry shader is
null and we have stream output then attach it to the vertex
shader right before executing the draw pipeline.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agoconfigure.ac: Allow OpenGL ES1 and ES2 only with enabled OpenGL
Andreas Boll [Fri, 1 Feb 2013 12:39:42 +0000 (13:39 +0100)]
configure.ac: Allow OpenGL ES1 and ES2 only with enabled OpenGL

Building OpenGL ES1 and/or ES2 without OpenGL is not supported on mesa
9.0.x

Reviewed-by: Chad Versace <chad.versace@linux.intel.com>
11 years agoi965/fs: Don't save value returned by emit() if it's not used.
Matt Turner [Wed, 10 Apr 2013 02:12:39 +0000 (19:12 -0700)]
i965/fs: Don't save value returned by emit() if it's not used.

Probably a copy-n-paste mistake.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agomesa: Remove extra MapBufferRange in create_beginend_table()
Brian Paul [Mon, 22 Apr 2013 16:42:43 +0000 (10:42 -0600)]
mesa: Remove extra MapBufferRange in create_beginend_table()

Looks like a copy&paste typo.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
11 years agogallium: Add a new clip_halfz rasterizer state.
José Fonseca [Sat, 20 Apr 2013 11:24:44 +0000 (12:24 +0100)]
gallium: Add a new clip_halfz rasterizer state.

gl_rasterization_rules lumps too many different flags.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agoi965: Fix a mistake in the comments for software counters.
Kenneth Graunke [Thu, 28 Feb 2013 03:06:32 +0000 (19:06 -0800)]
i965: Fix a mistake in the comments for software counters.

The code doesn't set brw->query.obj to NULL, it sets query->bo to NULL.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agogallivm: Fix assignment of unsigned values to OUT register.
José Fonseca [Mon, 22 Apr 2013 14:28:32 +0000 (15:28 +0100)]
gallivm: Fix assignment of unsigned values to OUT register.

TEMP is not the only register file that accept unsigned. OUT too.

Actually, what determines the appropriate type of the destination value is
not the opcode, but rather the register.

Also cleanup/simplify code.  Add a few more asserts, but also make
code more robust by handling graceful if assert fails.

This fixes segfault / assertion in the included vert-uadd.sh graw shader.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
11 years agoi965: Apply CMP NULL {Switch} work-around to other Gen7s.
Matt Turner [Sun, 21 Apr 2013 07:18:11 +0000 (00:18 -0700)]
i965: Apply CMP NULL {Switch} work-around to other Gen7s.

Listed in the restrictions section of CMP, but not on the work-arounds
page.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agost/mesa: minor indentation fixes
Brian Paul [Mon, 22 Apr 2013 16:08:06 +0000 (10:08 -0600)]
st/mesa: minor indentation fixes

11 years agomesa: Introduce a globally-available minify() macro.
Eric Anholt [Sat, 13 Apr 2013 08:46:09 +0000 (01:46 -0700)]
mesa: Introduce a globally-available minify() macro.

This matches u_minify()'s behavior, for consistency.

Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agomesa: Generalize TexStorage allocator between swrast and intel.
Eric Anholt [Sat, 13 Apr 2013 08:37:35 +0000 (01:37 -0700)]
mesa: Generalize TexStorage allocator between swrast and intel.

This should be reusable for other non-gallium drivers, so we can make the
extension always be available.

v2: Add a more detailed comment than the old function had (recommended
    by Brian).

Reviewed-by: Brian Paul <brianp@vmware.com> (v1)
11 years agomesa: Add performance debug for meta code.
Eric Anholt [Sat, 13 Apr 2013 03:16:41 +0000 (20:16 -0700)]
mesa: Add performance debug for meta code.

I noticed a fallback in regnum through sysprof, and wanted a nicer way to
get information about it.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agointel: Mention how much data we're trying to subdata in perf debug.
Eric Anholt [Fri, 12 Apr 2013 00:58:09 +0000 (17:58 -0700)]
intel: Mention how much data we're trying to subdata in perf debug.

Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>