mesa.git
11 years agor600g/llvm: get use_kill from compiler shader
Vincent Lejeune [Sat, 27 Apr 2013 22:01:00 +0000 (00:01 +0200)]
r600g/llvm: get use_kill from compiler shader

11 years agoi965/fs: Print out the estimated cycle count in INTEL_DEBUG=wm
Eric Anholt [Mon, 8 Apr 2013 23:38:57 +0000 (16:38 -0700)]
i965/fs: Print out the estimated cycle count in INTEL_DEBUG=wm

This could be used by shader-db for hopefully more accurate regression
testing.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agoi965/fs: Allow LRPs with uniform registers.
Eric Anholt [Fri, 26 Apr 2013 03:20:05 +0000 (20:20 -0700)]
i965/fs: Allow LRPs with uniform registers.

Improves GLB2.7 performance on my HSW by 0.671455% +/- 0.225037% (n=62).

v2: Make is_valid_3src() a method of the fs_reg. (recommended by Ken)

Reviewed-by: Matt Turner <mattst88@gmail.com> (v1)
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> (v1)
11 years agointel: Be more conservative in disabling tiling to save memory.
Eric Anholt [Thu, 25 Apr 2013 21:41:36 +0000 (14:41 -0700)]
intel: Be more conservative in disabling tiling to save memory.

Improves GLB2.7 trex performance 1.01985% +/- 0.721366% on my IVB (n=10)
and by 3.38771% +/- 0.584241% (n=15) on my HSW, due to a 32x32 ARGB8888
cubemap going from untiled to tiled.

Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
11 years agoi965: Disable Z16 on contexts that don't require it.
Eric Anholt [Thu, 25 Apr 2013 19:34:07 +0000 (12:34 -0700)]
i965: Disable Z16 on contexts that don't require it.

It appears that Z16 on Intel hardware is in fact slower than Z24, so
people are getting surprisingly hurt when trying to use Z16 as a
performance-versus-precision tradeoff, or when they're targeting GLES2 and
that's all you get.

GL 3.0+ have Z16 on the list of required exact format sizes, but GLES
doesn't, so choose the better-performing layout in that case.  Improves
GLB 2.7 trex performance at 1920x1080 by 10.7% +/- 1.1% (n=3) on my IVB
system.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agointel: Report FBO incompleteness causes through GL_ARB_debug_output.
Eric Anholt [Tue, 23 Apr 2013 02:41:40 +0000 (19:41 -0700)]
intel: Report FBO incompleteness causes through GL_ARB_debug_output.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agointel: Fold the one last function intel_tex_format.c into the caller.
Eric Anholt [Mon, 22 Apr 2013 23:04:25 +0000 (16:04 -0700)]
intel: Fold the one last function intel_tex_format.c into the caller.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
11 years agomesa: Fix error checking for GS UBO getters.
Eric Anholt [Wed, 10 Apr 2013 17:04:11 +0000 (10:04 -0700)]
mesa: Fix error checking for GS UBO getters.

These are supposed to be present if both things are available, but we were
enabling them if either one was.

11 years agomesa: Add a clarifying comment about EXTRA_ error checking.
Eric Anholt [Wed, 10 Apr 2013 16:59:41 +0000 (09:59 -0700)]
mesa: Add a clarifying comment about EXTRA_ error checking.

11 years agomesa: Add an extra clarifying set of braces to getter checking.
Eric Anholt [Wed, 10 Apr 2013 16:53:11 +0000 (09:53 -0700)]
mesa: Add an extra clarifying set of braces to getter checking.

For this multi-page single statement, my thought the end was to that the
next block was mis-indented, rather than that the dropped indentation
actually indicated the end of the loop.

11 years agomesa: Fix error checking for getters consisting of only API versions.
Eric Anholt [Wed, 10 Apr 2013 16:49:37 +0000 (09:49 -0700)]
mesa: Fix error checking for getters consisting of only API versions.

In almost all of our cases, getters that are turned on for only some API
variants will have an extension listed as one of the things that can
enable it, and thus api_check gets set.  For extra_gl30_es3 (used for
NUM_EXTENSIONS, MAJOR_VERSION, MINOR_VERSION) on a GL 2.1 context, though,
we would check twice, not find either one, but never actually throw the
error.

11 years agomesa: Clarify the names of error checking variables for glGet.
Eric Anholt [Wed, 10 Apr 2013 16:47:12 +0000 (09:47 -0700)]
mesa: Clarify the names of error checking variables for glGet.

There's no reason to actually count these things, so the integer ++
behavior was just confusing.

11 years agoi915: Add support for GL_EXT_texture_sRGB and GL_EXT_texture_sRGB_decode.
Eric Anholt [Thu, 18 Apr 2013 02:10:29 +0000 (19:10 -0700)]
i915: Add support for GL_EXT_texture_sRGB and GL_EXT_texture_sRGB_decode.

This brings the driver up to GL 2.1.

11 years agoi915: Always enable GL 2.0 support.
Eric Anholt [Wed, 17 Apr 2013 20:55:08 +0000 (13:55 -0700)]
i915: Always enable GL 2.0 support.

There's no point in shipping a non-GL2 driver today.

11 years agoi915: Correctly set the OQ counter bits.
Eric Anholt [Wed, 17 Apr 2013 20:58:00 +0000 (13:58 -0700)]
i915: Correctly set the OQ counter bits.

While we may provide the extension, we need to tell applications that they
can't actually use it:

            An implementation can either set QUERY_COUNTER_BITS_ARB to the
            value 0, or to some number greater than or equal to n.  If an
            implementation returns 0 for QUERY_COUNTER_BITS_ARB, then the
            occlusion queries will always return that zero samples passed the
            occlusion test, and so an application should not use occlusion
            queries on that implementation.

11 years agoi965: Move is_math/is_tex/is_control_flow() to backend_instruction.
Kenneth Graunke [Sun, 28 Apr 2013 08:35:57 +0000 (01:35 -0700)]
i965: Move is_math/is_tex/is_control_flow() to backend_instruction.

These are entirely based on the opcode, which is available in
backend_instruction.  It makes sense to only implement them in one
place.

This changes the VS implementation of is_tex() slightly, which now
accepts FS_OPCODE_TXB and SHADER_OPCODE_LOD.  However, since those
aren't generated in the VS anyway, it should be fine.

This also makes is_control_flow() available in the VS.

Signed-off-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
11 years agodraw/so: fix overflow calculation
Zack Rusin [Sat, 27 Apr 2013 06:51:26 +0000 (02:51 -0400)]
draw/so: fix overflow calculation

only report overflow for missing targets if they're actually being
used. if the targets are missing but are not being used by any
slot in the stream output declaration we should correctly just
ignore them.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agollvmpipe: Fix queries when screen->num_threads == 0.
José Fonseca [Mon, 29 Apr 2013 14:40:06 +0000 (15:40 +0100)]
llvmpipe: Fix queries when screen->num_threads == 0.

That is, when llvmpipe is run in single-threaded mode.

Trivial.

Tested with

  LP_NUM_THREADS=0 glean --run results --overwrite --quick --tests occluQry

11 years agoRevert "st/mesa: add a simple path to BufferData if it only discards buffer contents"
José Fonseca [Mon, 29 Apr 2013 14:12:26 +0000 (15:12 +0100)]
Revert "st/mesa: add a simple path to BufferData if it only discards buffer contents"

This reverts commit 5649f886f76023532538b8792605a3578cec1ed1.

It causes segfaults when size is zero.

11 years agor600g: force full cache for hyperz
Jerome Glisse [Wed, 24 Apr 2013 23:15:52 +0000 (19:15 -0400)]
r600g: force full cache for hyperz

Seems that in some case allowing half cache usage confuse the gpu
and trigger lockup. Force full cache use.

Should fix :
https://bugs.freedesktop.org/show_bug.cgi?id=59592
https://bugs.freedesktop.org/show_bug.cgi?id=60848
https://bugs.freedesktop.org/show_bug.cgi?id=60969
https://bugs.freedesktop.org/show_bug.cgi?id=61747
https://bugs.freedesktop.org/show_bug.cgi?id=62466
https://bugs.freedesktop.org/show_bug.cgi?id=62669
https://bugs.freedesktop.org/show_bug.cgi?id=62721
https://bugs.freedesktop.org/show_bug.cgi?id=63124

Signed-off-by: Jerome Glisse <jglisse@redhat.com>
11 years agofreedreno: fix rebase screw-up
Rob Clark [Mon, 29 Apr 2013 11:36:27 +0000 (07:36 -0400)]
freedreno: fix rebase screw-up

Add back 2nd arg to emit_vertexbufs() which got lost in rebase.

Signed-off-by: Rob Clark <robclark@freedesktop.org>
11 years agoi965/fs: Don't try to use bogus interpolation modes pre-Gen6.
Chris Forbes [Fri, 26 Apr 2013 23:00:46 +0000 (11:00 +1200)]
i965/fs: Don't try to use bogus interpolation modes pre-Gen6.

Interpolation modes other than perspective-barycentric-pixel-center (and
their associated coefficients in the WM payload) only exist in Gen6 and
later.

Unfortunately, if a varying was declared as `centroid`, we would blindly
read the nonexistant values, and so produce all manner of bad behavior
-- texture swimming, snow, etc.

Fixes rendering in Counter-Strike Source and Team Fortress 2 on
Ironlake.

NOTE: This is a candidate for the 9.1 branch.

Signed-off-by: Chris Forbes <chrisf@ijw.co.nz>
Reviewed-by: Paul Berry <stereotype441@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Tested-by: Jordan Justen <jordan.l.justen@intel.com>
11 years agoi965/vs: Fix order of source arguments to LRP.
Matt Turner [Sun, 28 Apr 2013 21:35:01 +0000 (14:35 -0700)]
i965/vs: Fix order of source arguments to LRP.

The order or arguments matches DirectX, and is backwards from GLSL's
mix() built-in.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=63983

11 years agollvmpipe: stop crashing when one of the so targets is null
Zack Rusin [Sat, 27 Apr 2013 04:52:49 +0000 (00:52 -0400)]
llvmpipe: stop crashing when one of the so targets is null

Fixes a crash when one of the so targets is null.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agodraw/so: indicate overflow when buffer is missing
Zack Rusin [Sat, 27 Apr 2013 04:49:23 +0000 (00:49 -0400)]
draw/so: indicate overflow when buffer is missing

We were crashing if one of the buffers wasn't set, we should
just treat it as an overflow. It's useful when using so
statistics because it allows one to figure out how much data
would be generated by so without actually writing any of it.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agogallivm: fix indirect addressing of temps in soa mode
Zack Rusin [Sat, 27 Apr 2013 02:53:07 +0000 (22:53 -0400)]
gallivm: fix indirect addressing of temps in soa mode

we weren't adding the soa offsets when constructing the indices
for the gather functions. That meant that we were always returning
the data in the first vertex/primitive/pixel in the SoA structure
and not correctly fetching from all structures.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agotgsi/ureg: Add a function to return the number of outputs
Zack Rusin [Wed, 24 Apr 2013 03:36:40 +0000 (23:36 -0400)]
tgsi/ureg: Add a function to return the number of outputs

We already hold the variable, just weren't providing access
to it.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
11 years agodraw/so: Fix overflow calculations
Zack Rusin [Tue, 23 Apr 2013 22:56:47 +0000 (18:56 -0400)]
draw/so: Fix overflow calculations

We weren't taking the buffer offset, destination offset or the
stride into consideration so we were frequently writing into
an overflown buffer.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
11 years agodraw/llvm: fix viewport transformations
Zack Rusin [Tue, 23 Apr 2013 22:47:08 +0000 (18:47 -0400)]
draw/llvm: fix viewport transformations

This was a very serious bug. We were always doing the viewport
transformations on the first output of the vertex shader. That means
that every application that was storing position in anything but
OUT[0] was outputing untransformed vertices and had broken output
for whatever it was storing at OUT[0]. Correctly take into
consideration where the vertex position is actually stored.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
11 years agogallium: increase the number of available stream output decls
Zack Rusin [Sat, 27 Apr 2013 03:00:38 +0000 (23:00 -0400)]
gallium: increase the number of available stream output decls

There can be more stream output decls than shader outputs because
individual components from them can be split and distributed
among different so buffers.

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
11 years agollvmpipe: implement so_overflow query
Zack Rusin [Tue, 23 Apr 2013 10:19:14 +0000 (06:19 -0400)]
llvmpipe: implement so_overflow query

Signed-off-by: Zack Rusin <zackr@vmware.com>
Reviewed-by: José Fonseca <jfonseca@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
11 years agomesa: fix the compressed TexSubImage size checking code
Brian Paul [Fri, 26 Apr 2013 19:49:35 +0000 (13:49 -0600)]
mesa: fix the compressed TexSubImage size checking code

Before, we'd incorrectly generate an error if we we tried to
replace a non-4x4 block near the edge of a NPOT compressed texture.
For example, if the dest image was 15 texels wide and xoffset=12
and width=3 we'd incorrectly generate GL_INVALID_OPERATION.

Verified with new tests added to piglit s3tc-errors test.

Note: This is a candidate for the stable branches.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
11 years agollvmpipe: replace LP_MAX_THREADS with screen->num_threads in query code
Brian Paul [Fri, 26 Apr 2013 13:31:49 +0000 (07:31 -0600)]
llvmpipe: replace LP_MAX_THREADS with screen->num_threads in query code

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
11 years agollvmpipe: bump LP_MAX_THREADS to 16
Brian Paul [Fri, 26 Apr 2013 13:26:46 +0000 (07:26 -0600)]
llvmpipe: bump LP_MAX_THREADS to 16

On the mesa-users list, Burlen Loring reported a speed-up with 16 cores
and his test/app.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
11 years agomesa: updated read_buffer_enum_to_index() comment
Brian Paul [Fri, 26 Apr 2013 13:26:06 +0000 (07:26 -0600)]
mesa: updated read_buffer_enum_to_index() comment

Remove the part about the value of gl_framebuffer::Name.

11 years agor600/uvd: stop advertising MPEG4 on UVD 2.x chips v2
Christian König [Fri, 26 Apr 2013 09:49:55 +0000 (11:49 +0200)]
r600/uvd: stop advertising MPEG4 on UVD 2.x chips v2

That is just not supported by the hardware.

v2: fix compare

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
11 years agoradeon/uvd: stop using anonymous unions
Christian König [Fri, 26 Apr 2013 09:16:19 +0000 (11:16 +0200)]
radeon/uvd: stop using anonymous unions

Signed-off-by: Christian König <christian.koenig@amd.com>
11 years agomesa: fix type comparison errors in sub-texture error checking code
Tapani Pälli [Thu, 18 Apr 2013 06:21:27 +0000 (09:21 +0300)]
mesa: fix type comparison errors in sub-texture error checking code

patch fixes a crash that happens if glTexSubImage2D is called with a
negative xoffset.

NOTE: This is a candidate for stable branches.

Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Brian Paul <brianp@vmware.com>
11 years agoRevert "draw: Yield zeros for LLVM fetches of non-existing vertex elements."
José Fonseca [Thu, 25 Apr 2013 17:51:17 +0000 (18:51 +0100)]
Revert "draw: Yield zeros for LLVM fetches of non-existing vertex elements."

After more thought/discussion, it seems it is better to handle this sort
of stuff in the state tracker.

So this reverts commit 12096f334b82340dc165ed15e6f8f44d4cf94df4, except the
variant->key -> key shorthands.

11 years agoilo: add the driver to the build system
Chia-I Wu [Wed, 12 Dec 2012 22:01:23 +0000 (06:01 +0800)]
ilo: add the driver to the build system

Add ilo to targets/egl-static and add a new target dri-ilo.  Update autoconf
and automake rules.

11 years agoilo: compile VS/GS/FS with the toy compiler
Chia-I Wu [Wed, 12 Dec 2012 21:48:46 +0000 (05:48 +0800)]
ilo: compile VS/GS/FS with the toy compiler

11 years agoilo: add a toy shader compiler
Chia-I Wu [Wed, 12 Dec 2012 21:48:28 +0000 (05:48 +0800)]
ilo: add a toy shader compiler

This is a simple shader compiler that performs almost zero optimizations.  The
generated code is usually much larger comparing to that generated by i965.
The generated code also requires many more registers.

Function-wise, it lacks register spilling and does not support most TGSI
indirections.  Other than those, it works alright.

11 years agoilo: hook up pipe context GPGPU functions
Chia-I Wu [Wed, 12 Dec 2012 21:44:41 +0000 (05:44 +0800)]
ilo: hook up pipe context GPGPU functions

This just adds a stub.

11 years agoilo: hook up pipe context video functions
Chia-I Wu [Wed, 12 Dec 2012 21:43:04 +0000 (05:43 +0800)]
ilo: hook up pipe context video functions

This just hooks them up with auxiliary/vl layer.

11 years agoilo: add support for time/occlusion/primitive queries
Chia-I Wu [Wed, 12 Dec 2012 21:35:37 +0000 (05:35 +0800)]
ilo: add support for time/occlusion/primitive queries

11 years agoilo: hook up pipe context 3D functions
Chia-I Wu [Tue, 16 Apr 2013 08:36:03 +0000 (16:36 +0800)]
ilo: hook up pipe context 3D functions

11 years agoilo: add GEN7 support for 3D pipeline
Chia-I Wu [Tue, 16 Apr 2013 10:09:35 +0000 (18:09 +0800)]
ilo: add GEN7 support for 3D pipeline

11 years agoilo: add 3D pipeline for GEN6
Chia-I Wu [Wed, 12 Dec 2012 21:28:42 +0000 (05:28 +0800)]
ilo: add 3D pipeline for GEN6

The 3D pipeline is a high-level interface to emit 3D commands and states.  It
uses GEN6 GPE to do the real work.

11 years agoilo: add GEN7 GPE
Chia-I Wu [Tue, 16 Apr 2013 10:09:01 +0000 (18:09 +0800)]
ilo: add GEN7 GPE

11 years agoilo: add GEN6 GPE
Chia-I Wu [Wed, 12 Dec 2012 21:23:34 +0000 (05:23 +0800)]
ilo: add GEN6 GPE

GEN6 GPE (Graphics Processing Engine) is a low-level interface to emit 3D
commands and states.

11 years agoilo: hook up pipe context query functions
Chia-I Wu [Wed, 12 Dec 2012 21:20:40 +0000 (05:20 +0800)]
ilo: hook up pipe context query functions

None of the query types are supported yet.

11 years agoilo: hook up pipe context transfer functions
Chia-I Wu [Wed, 12 Dec 2012 21:18:25 +0000 (05:18 +0800)]
ilo: hook up pipe context transfer functions

11 years agoilo: hook up pipe context blit functions
Chia-I Wu [Wed, 12 Dec 2012 21:15:10 +0000 (05:15 +0800)]
ilo: hook up pipe context blit functions

11 years agoilo: hook up pipe context state functions
Chia-I Wu [Tue, 16 Apr 2013 08:27:50 +0000 (16:27 +0800)]
ilo: hook up pipe context state functions

11 years agoilo: add functions to manage shaders
Chia-I Wu [Wed, 12 Dec 2012 21:05:01 +0000 (05:05 +0800)]
ilo: add functions to manage shaders

This commits add shader cache, shader state, shader variant, and etc.  It does
not add the shader compiler though.

11 years agoilo: hook up pipe context flush function
Chia-I Wu [Tue, 16 Apr 2013 08:24:40 +0000 (16:24 +0800)]
ilo: hook up pipe context flush function

11 years agoilo: add command parser
Chia-I Wu [Wed, 12 Dec 2012 20:36:41 +0000 (04:36 +0800)]
ilo: add command parser

The command parser manages batch buffers and command submissions.

11 years agoilo: hook up pipe screen resource functions
Chia-I Wu [Wed, 12 Dec 2012 20:44:21 +0000 (04:44 +0800)]
ilo: hook up pipe screen resource functions

11 years agoilo: hook up pipe screen format functions
Chia-I Wu [Wed, 12 Dec 2012 20:43:01 +0000 (04:43 +0800)]
ilo: hook up pipe screen format functions

11 years agoilo: hook up pipe_screen param and fence functions
Chia-I Wu [Wed, 12 Dec 2012 20:26:23 +0000 (04:26 +0800)]
ilo: hook up pipe_screen param and fence functions

11 years agoilo: add debug flags settable through ILO_DEBUG
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+
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
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().
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.
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.
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
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
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
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
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
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
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)
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
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
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.
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
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
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.
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.
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.
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.
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.
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.
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.
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
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
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
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
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
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
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
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
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()
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
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
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+
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
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
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
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>