OSDN Git Service
Chia-I Wu [Thu, 12 Feb 2015 19:59:45 +0000 (03:59 +0800)]
ilo: fix rectlist length on Gen8
5 PIPE_CONTROLs, 2 3DSTATE_WM_HZ_OP, and depth buffer setup require 65 DWords.
Chia-I Wu [Thu, 12 Feb 2015 19:54:32 +0000 (03:54 +0800)]
ilo: fix 3DSTATE_VF_TOPOLOGY
The pipe primitive type was wrongly translated twice.
Jose Fonseca [Fri, 13 Feb 2015 13:51:28 +0000 (13:51 +0000)]
os,llvmpipe: Set rasterizer thread names on Linux.
To help identify llvmpipe rasterizer threads -- especially when there
can be so many.
We can eventually generalize this to other OSes, but for that we must
restrict the function to be called from the current thread. See also
http://stackoverflow.com/a/
7989973
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Jose Fonseca [Fri, 13 Feb 2015 09:46:44 +0000 (09:46 +0000)]
uti/u_atomic: Don't test p_atomic_add with booleans.
Add another class of tests.
Fixes https://bugs.freedesktop.org/show_bug.cgi?id=89112
I failed to spot this in my previous change, because bool was a typedef
for char on the system I tested.
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Tapani Pälli [Thu, 12 Feb 2015 12:33:53 +0000 (14:33 +0200)]
mesa: fix OES_texture_float texture render target behavior
Current implementation allowed usage of unsized type texture GL_FLOAT
and GL_HALF_FLOAT as a render target as this was 'expected behavior' by
WEBGL_oes_texture_float and is also allowed by the oes-texture-float
WebGL test. However this broke some ES3 conformance tests that do not
accept such behavior. Patch sets such an fbo incomplete as expected by
the ES3 conformance tests. Textures with sized types like RGBA32F will
still continue to work as render targets.
v2: code style cleanups (Ian Romanick, Matt Turner)
Signed-off-by: Tapani Pälli <tapani.palli@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88905
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Cc: "10.5" <mesa-stable@lists.freedesktop.org>
Eric Anholt [Thu, 12 Feb 2015 22:17:21 +0000 (14:17 -0800)]
vc4: Make SF be a flag on the QIR instructions.
Right now the places that used to emit a mov.sf just put the SF on the
previous instruction when it generated the source of the SF value. Even
without optimization to push the sf up further (and kill thus potentially
kill more MOVs), this gets us:
total uniforms in shared programs: 13455 -> 13457 (0.01%)
uniforms in affected programs: 3 -> 5 (66.67%)
total instructions in shared programs: 40296 -> 40198 (-0.24%)
instructions in affected programs: 12595 -> 12497 (-0.78%)
Eric Anholt [Wed, 11 Feb 2015 22:55:42 +0000 (14:55 -0800)]
r200: Drop unused variable.
Quiets compiler warning since
e7f2f2dea5acdbd1a12ed88914e64a38a97432f0.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Eric Anholt [Wed, 11 Feb 2015 22:54:33 +0000 (14:54 -0800)]
i965: Quiet another compiler warning about uninitialized values.
The compiler can't tell that we're always going to hit the first if block
on the first time through the loop.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Eric Anholt [Wed, 11 Feb 2015 22:53:08 +0000 (14:53 -0800)]
i965: Move some asserts to unreachable.
If execution was supposed to be supported in this case, we'd run into
trouble from completely uninitialized sat_imm values.
v2: Drop the '!' before the string.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Eric Anholt [Wed, 11 Feb 2015 22:50:35 +0000 (14:50 -0800)]
i965: Shut up a compiler warning about uninitialized var.
We always pass this argument, even if it won't be used by the particular
texture op.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Carl Worth [Thu, 12 Feb 2015 18:14:24 +0000 (10:14 -0800)]
Revert use of Mesa IR optimizer for ARB_fragment_programs
Commit
f82f2fb3dc770902f1657ab1c22e6004faa3afab added use of the Mesa
IR optimizer for both ARB_fragment_program and ARB_vertex_program, but
only justified the vertex-program portions with measured performance
improvements.
Meanwhile, the optimizer was seen to generate hundreds of unused
immediates without discarding them, causing failures.
Discard the use of the optimizer for now to fix the regression. (In
the future, we anticpate things moving from Mesa IR to NIR for better
optimization anyway.)
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=82477
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
CC: "10.3 10.4 10.5" <mesa-stable@lists.freedesktop.org>
Jose Fonseca [Thu, 12 Feb 2015 19:31:20 +0000 (19:31 +0000)]
util/u_atomic: Use lower-case variables in _Interlocked* helpers.
Jose Fonseca [Wed, 11 Feb 2015 15:30:39 +0000 (15:30 +0000)]
util/u_atomic: Add _InterlockedExchangeAdd8/16 for older MSVC.
We need to build certain parts of Mesa (namely gallium, llvmpipe, and
therefore util) with Windows SDK 7.0.7600, which includes MSVC 2008.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Jose Fonseca [Wed, 11 Feb 2015 15:21:25 +0000 (15:21 +0000)]
util/u_atomic: Test p_atomic_add() for 8bit integers.
Reviewed-by: Brian Paul <brianp@vmware.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Ilia Mirkin [Thu, 12 Feb 2015 11:16:27 +0000 (06:16 -0500)]
docs: add ARB_draw_indirect to ES 3.1 list
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Axel Davy [Wed, 14 Jan 2015 19:36:04 +0000 (20:36 +0100)]
egl: Soften several HAVE_DRM_PLATFORM to HAVE_LIBDRM
To fix build when libdrm is not found,
commit
a594cec7e3ef275c386054127a357110a19dd823 did put several
parts of egl code under #ifdef HAVE_DRM_PLATFORM.
HAVE_DRM_PLATFORM means the egl drm platform is being built.
What should have been used instead is HAVE_LIBDRM.
At a few locations, the HAVE_DRM_PLATFORM introduced
have already been replaced by HAVE_LIBDRM, this patch
replaces the remaining occurences.
This patch makes for example EGL_EXT_image_dma_buf_import
be advertised by egl under x11 when the drm egl platform
is not built, whereas previously it required the drm egl
platform to be built.
Signed-off-by: Axel Davy <axel.davy@ens.fr>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Emil Velikov [Tue, 10 Feb 2015 15:11:09 +0000 (15:11 +0000)]
auxiliary/vl: bring back the VL code for the dri targets
With commit
c642e87d9f4(auxiliary/vl: rework the build of the VL code)
we split out the VL code into a separate static library that was meant
to be used by the VL targets alone - va, vdpau, xvmc.
The commit failed to consider the way we handle vdpau-gl interop and
broke it. Bring back the functionality by keeping the vl <> vl_stub
separation as requrested by Christian.
v2: Update the omx target as well. Update mesa-stable email address.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=86837
Cc: "10.5" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Tested-by: Andy Furniss <adf.lists@gmail.com>
Emil Velikov [Tue, 10 Feb 2015 14:14:16 +0000 (14:14 +0000)]
configure: rework wayland_scanner handling(fix make distcheck)
Currently having the wayland-scanner is optional, which causes problems
when autotools parses through the makefiles, and tries to generate all
the BUILT_SOURCES.
As the config option --with-egl-platform=wayland is not the default, we
won't end up setting the WAYLAND_SCANNER variable, which in turn will
cause some files to not get generated.
There has been a wayland-scanner package as of wayland 1.2 which
provides a variable for the scanner binary, so let's use that one and
fall back to manually searching via AC_PATH_PROG when needed.
Cc: "10.5" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Kristian Høgsberg <krh@bitplanet.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Emil Velikov [Tue, 10 Feb 2015 13:10:18 +0000 (13:10 +0000)]
nir: add missing header to the sources list
Cc: "10.5" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Emil Velikov [Sat, 7 Feb 2015 19:20:16 +0000 (19:20 +0000)]
nir: resolve nir.h dependency list (fix make distcheck)
Use nir/nir_opcodes.h as is (w/o the absolute path), as it is the target
name used to generate the actual file. Otherwise the target is missing,
the file won't get generated and the build will fail.
Cc: "10.5" <mesa-stable@lists.freedesktop.org>
Signed-off-by: Emil Velikov <emil.l.velikov@gmail.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Martin Peres [Thu, 12 Feb 2015 09:24:37 +0000 (11:24 +0200)]
docs: update GL3.txt to state my current work on the dsa extension
Signed-off-by: Martin Peres <martin.peres@linux.intel.com>
Ben Widawsky [Wed, 11 Feb 2015 01:22:45 +0000 (17:22 -0800)]
i965/vs/skl: Use vec4 datatypes for message header
We're using a SIMD4x2 sampler message, which has execsize 4, and so the
register width must be <= 4. Use <4,4,1> regioning instead of <8,8,1>
regioning to access the same data but avoid tripping the assert.
Fixes the following piglit tests:
spec/glsl-1.20/compiler/structure-and-array-operations/array-selection.vert
spec/glsl-es-3.00/compiler/uniform_block/interface-name-basic.vert
spec/glsl-es-3.00/compiler/uniform_block/interface-name-field-clashes-with-struct.vert
spec/glsl-es-3.00/compiler/uniform_block/interface-name-field-clashes-with-function.vert
spec/glsl-es-3.00/compiler/uniform_block/interface-name-array.vert
glslparsertest/glsl2/condition-07.vert
spec/glsl-es-3.00/compiler/uniform_block/interface-name-field-clashes-with-variable.vert
v2: Better commit message courtesy of Ken.
I had a discussion with Ken, and we both question how we end up with a mov and
execsize 4. For now though, this fixes the piglit tests, so we can worry about
it later.
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Chia-I Wu [Tue, 27 Jan 2015 11:48:39 +0000 (19:48 +0800)]
ilo: update screen init for Gen8
This is very preliminary and is only tested with glxgears. All information
about Gen8 is derived from i965 and beignet.
Chia-I Wu [Wed, 11 Feb 2015 06:31:54 +0000 (14:31 +0800)]
ilo: update outdated render command emissions for Gen8
Chia-I Wu [Wed, 11 Feb 2015 06:21:42 +0000 (14:21 +0800)]
ilo: update rectlist command emission for Gen8
Chia-I Wu [Tue, 27 Jan 2015 12:11:27 +0000 (20:11 +0800)]
ilo: update draw command emission for Gen8
Chia-I Wu [Tue, 27 Jan 2015 12:10:46 +0000 (20:10 +0800)]
ilo: update surface state emission for Gen8
Chia-I Wu [Tue, 27 Jan 2015 12:10:19 +0000 (20:10 +0800)]
ilo: update dynamic state emission for Gen8
Chia-I Wu [Tue, 27 Jan 2015 12:00:48 +0000 (20:00 +0800)]
ilo: update outdated gen assertions for Gen8
Chia-I Wu [Mon, 26 Jan 2015 09:12:33 +0000 (17:12 +0800)]
ilo: add new WM related helpers for Gen8
Chia-I Wu [Mon, 9 Feb 2015 23:13:11 +0000 (07:13 +0800)]
ilo: update VS related functions for Gen8
Chia-I Wu [Mon, 9 Feb 2015 23:10:20 +0000 (07:10 +0800)]
ilo: update VF related functions for Gen8
Chia-I Wu [Sun, 25 Jan 2015 10:05:04 +0000 (18:05 +0800)]
ilo: update SAMPLER_STATE for Gen8
Chia-I Wu [Mon, 26 Jan 2015 07:52:06 +0000 (15:52 +0800)]
ilo: update SAMPLER_BORDER_COLOR_STATE for Gen8
Chia-I Wu [Mon, 26 Jan 2015 07:35:14 +0000 (15:35 +0800)]
ilo: update depth clear value for Gen8
Chia-I Wu [Tue, 27 Jan 2015 08:34:45 +0000 (16:34 +0800)]
ilo: update ilo_zs_surface for Gen8
Chia-I Wu [Tue, 27 Jan 2015 06:58:32 +0000 (14:58 +0800)]
ilo: update ilo_view_surface for Gen8
Chia-I Wu [Sun, 25 Jan 2015 07:12:52 +0000 (15:12 +0800)]
ilo: update texture layout for Gen8
Chia-I Wu [Sun, 25 Jan 2015 10:20:43 +0000 (18:20 +0800)]
ilo: update ilo_blend_state and related functions for Gen8
Chia-I Wu [Mon, 26 Jan 2015 07:28:15 +0000 (15:28 +0800)]
ilo: update ilo_dsa_state and related functions for Gen8
Chia-I Wu [Mon, 26 Jan 2015 05:34:51 +0000 (13:34 +0800)]
ilo: update multisample related states for Gen8
Chia-I Wu [Mon, 26 Jan 2015 10:06:00 +0000 (18:06 +0800)]
ilo: update WM and PS related functions for Gen8
Chia-I Wu [Mon, 26 Jan 2015 16:08:48 +0000 (00:08 +0800)]
ilo: update SBE related functions for Gen8
Chia-I Wu [Mon, 26 Jan 2015 15:12:55 +0000 (23:12 +0800)]
ilo: update SF related functions for Gen8
Chia-I Wu [Tue, 27 Jan 2015 11:57:40 +0000 (19:57 +0800)]
ilo: update CLIP related functions for Gen8
Chia-I Wu [Mon, 26 Jan 2015 08:23:29 +0000 (16:23 +0800)]
ilo: update SF_CLIP_VIEWPORT for Gen8
Chia-I Wu [Mon, 26 Jan 2015 08:37:38 +0000 (16:37 +0800)]
ilo: update streamout related functions for Gen8
Chia-I Wu [Mon, 26 Jan 2015 08:08:23 +0000 (16:08 +0800)]
ilo: update 3DSTATE_{DS,HS,GS} for Gen8
Chia-I Wu [Mon, 26 Jan 2015 07:51:41 +0000 (15:51 +0800)]
ilo: update 3DSTATE_CONSTANT_x for Gen8
Chia-I Wu [Tue, 27 Jan 2015 16:20:09 +0000 (00:20 +0800)]
ilo: update 3DSTATE_URB_x for Gen8
Chia-I Wu [Tue, 27 Jan 2015 16:07:15 +0000 (00:07 +0800)]
ilo: update 3DSTATE_PUSH_CONSTANT_ALLOC_x for Gen8
Chia-I Wu [Mon, 26 Jan 2015 07:34:34 +0000 (15:34 +0800)]
ilo: update render engine common helpers for Gen8
Chia-I Wu [Sun, 25 Jan 2015 08:19:14 +0000 (16:19 +0800)]
ilo: update BLT helpers for Gen8
Chia-I Wu [Sun, 25 Jan 2015 07:25:33 +0000 (15:25 +0800)]
ilo: update MI helpers for Gen8
Chia-I Wu [Sun, 25 Jan 2015 10:04:25 +0000 (18:04 +0800)]
ilo: add functions for Gen8 relocs
Extend ilo_builder_writer_reloc() for Gen8 memory addressing. Add new
wrappers, ilo_builder_surface_reloc64(() and ilo_builder_batch_reloc64().
Chia-I Wu [Fri, 23 Jan 2015 17:12:04 +0000 (01:12 +0800)]
ilo: update the toy compiler for Gen8
Based on what we know from the classic driver.
Chia-I Wu [Fri, 23 Jan 2015 07:44:53 +0000 (15:44 +0800)]
ilo: update genhw headers
Accumulated changes for various renames and additions, including Gen8
definitions. Some of the dynamic state __SIZE no longer means the size of an
element, but the size of an array of elements. The changes can be seen in
ilo_render_dynamic.c.
Chia-I Wu [Mon, 26 Jan 2015 06:58:16 +0000 (14:58 +0800)]
ilo: clean up ilo_gpe_init_dsa()
Add dsa_get_stencil_enable_gen6(), dsa_get_depth_enable_gen6(), and
dsa_get_alpha_enable_gen6() to be called from ilo_gpe_init_dsa().
Chia-I Wu [Sun, 25 Jan 2015 17:08:31 +0000 (01:08 +0800)]
ilo: clean up ilo_gpe_init_blend()
Make ilo_blend_state more space efficient and forward-looking.
Chia-I Wu [Mon, 26 Jan 2015 06:00:25 +0000 (14:00 +0800)]
ilo: clean up sample patterns
Use signed int for sample positions and add helpers to access them. Call them
patterns instead of positions.
Matt Turner [Fri, 30 Jan 2015 21:42:19 +0000 (13:42 -0800)]
glsl: Optimize (f2i(trunc x)) into (f2i x).
total instructions in shared programs:
5950326 ->
5949286 (-0.02%)
instructions in affected programs: 88264 -> 87224 (-1.18%)
helped: 692
Matt Turner [Wed, 28 Jan 2015 18:09:14 +0000 (10:09 -0800)]
glsl: Optimize round-half-up pattern.
Hurts some Psychonauts shaders, but after the next patch (which this
enables) they're fewer instructions than before this patch.
Matt Turner [Fri, 30 Jan 2015 21:50:28 +0000 (13:50 -0800)]
glsl: Add trunc() to ir_builder.
Matt Turner [Sat, 24 Jan 2015 05:58:51 +0000 (21:58 -0800)]
i965: Add LINTERP/CINTERP to can_do_cmod().
LINTERP is implemented as a PLN instruction or a LINE+MAC. PLN and MAC
can do conditional mod. CINTERP is just a MOV.
total instructions in shared programs:
5952103 ->
5950284 (-0.03%)
instructions in affected programs: 324573 -> 322754 (-0.56%)
helped: 1819
We lose the SIMD16 in one Unigine Heaven shader which appears six times
in shader-db.
Matt Turner [Wed, 11 Feb 2015 02:45:18 +0000 (18:45 -0800)]
program: Remove _mesa_nop_vertex_program/_mesa_nop_fragment_program.
Dead since
commit
284ce20901b0c2cfab1d952cc129b8f3cd068f12
Author: Eric Anholt <eric@anholt.net>
Date: Fri Aug 20 10:52:14 2010 -0700
Remove remnants of the old glsl compiler.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Matt Turner [Wed, 11 Feb 2015 05:43:32 +0000 (21:43 -0800)]
nir: Recognize open-coded fmin/fmax.
And unfortunately other shaders do the same thing but with >=/<= which
we can't apply this optimization to because of NaNs.
instructions in affected programs: 23309 -> 22938 (-1.59%)
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Sat, 7 Feb 2015 01:12:59 +0000 (17:12 -0800)]
nir: Add algebraic opt for int comparisons with identical operands.
No change on shader-db on i965.
v2: Reword the comment due to feedback from Erik Faye-Lund
Reviewed-by: Connor Abbott <cwabbott0@gmail.com> (v1)
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com> (v1)
Eric Anholt [Sat, 7 Feb 2015 01:16:29 +0000 (17:16 -0800)]
nir: Fix load_const comparisons for CSE.
We want the size of a float per component, not the size of a whole vec4.
NIR instructions on i965:
total instructions in shared programs:
1261937 ->
1261929 (-0.00%)
instructions in affected programs: 114 -> 106 (-7.02%)
Looking at one of these examples (tesseract), it's from vec4 load_consts
for a MRT solid fill, which do get CSEed now that we don't memcmp off the
end of the const value and into the SSA def. For the 1-component loads
that are common in i965, we were only memcmping off into the rest of the
usually zero-filled const_value.
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Matt Turner [Wed, 11 Feb 2015 05:36:26 +0000 (21:36 -0800)]
i965/fs: Remove conditional mod when optimizing a SEL into a MOV.
Missed in commit
ca675b73, but got right in the companion commit
3c28b2c0.
Jeremy Huddleston Sequoia [Wed, 11 Feb 2015 06:21:47 +0000 (22:21 -0800)]
darwin: build fix
xfont.c:237:14: error: implicit declaration of function 'GetGLXDRIDrawable' is invalid in C99 [-Werror,-Wimplicit-function-declaration]
glxdraw = GetGLXDRIDrawable(CC->currentDpy, CC->currentDrawable);
^
Fixes regression from
291be28476ea60c6fb1eb2a882e2e25def5d3735
Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com>
Jeremy Huddleston Sequoia [Wed, 11 Feb 2015 04:32:02 +0000 (20:32 -0800)]
darwin: build fix
../../../src/mesa/main/compiler.h:47:10: fatal error: 'util/macros.h' file not found
Signed-off-by: Jeremy Huddleston Sequoia <jeremyhu@apple.com>
Matt Turner [Tue, 10 Feb 2015 20:31:37 +0000 (12:31 -0800)]
glsl: Optimize 1/exp(x) into exp(-x).
Lots of shaders divide by exp2(...) which we turn into a multiplication
by the reciprocal. We can avoid the reciprocal by simply negating exp2's
argument.
total instructions in shared programs:
5947154 ->
5946695 (-0.01%)
instructions in affected programs: 118661 -> 118202 (-0.39%)
helped: 380
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Matt Turner [Tue, 10 Feb 2015 19:23:55 +0000 (11:23 -0800)]
nir: Remove casts from void*.
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Matt Turner [Tue, 10 Feb 2015 19:21:47 +0000 (11:21 -0800)]
nir: Replace assert(0) with unreachable().
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Matt Turner [Tue, 10 Feb 2015 19:20:13 +0000 (11:20 -0800)]
nir: Remove unused has_indirect variable.
Reviewed-by: Jason Ekstrand <jason.ekstrand@intel.com>
Matt Turner [Sat, 31 Jan 2015 02:24:20 +0000 (18:24 -0800)]
i965/vec4: Emit MADs from (x + abs(y * z)).
Same as commit
3654b6d4 to the fs backend.
total instructions in shared programs:
5945788 ->
5945787 (-0.00%)
instructions in affected programs: 36 -> 35 (-2.78%)
helped: 1
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Matt Turner [Sat, 20 Dec 2014 05:35:56 +0000 (21:35 -0800)]
i965/vec4: Emit MADs from (x + -(y * z)).
Same as commit
c4fab711 to the fs backend.
total instructions in shared programs:
5945998 ->
5945788 (-0.00%)
instructions in affected programs: 74665 -> 74455 (-0.28%)
helped: 399
HURT: 180
It hurts some programs because we make no attempts in the vec4 backend
to avoid MADs if they have constant (or vector uniform) arguments.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Neil Roberts [Mon, 2 Feb 2015 17:47:23 +0000 (17:47 +0000)]
i965/skl: Implement WaDisable1DDepthStencil
Skylake+ doesn't support setting a depth buffer to a 1D surface but it
does allow pretending it's a 2D texture with a height of 1 instead.
This fixes the GL_DEPTH_COMPONENT_* tests of the copyteximage piglit
test (and also seems to avoid a subsequent GPU hang).
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89037
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Sat, 23 Nov 2013 02:35:46 +0000 (18:35 -0800)]
i965/gen7-8: Implement glMemoryBarrier().
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Tue, 10 Feb 2015 13:53:14 +0000 (15:53 +0200)]
i965: Generalize the update_null_renderbuffer_surface vtbl hook to non-renderbuffers.
Null surfaces are going to be useful to have something to point
unbound image units to, as the ARB_shader_image_load_store extension
requires us to behave deterministically in cases where some shader
tries to access an unbound image unit: Invalid stores and atomics are
supposed to be discarded and invalid loads are supposed to return
zero, which is precisely what the null surface does.
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Sat, 23 Nov 2013 00:08:12 +0000 (16:08 -0800)]
i965: Allocate binding table space for shader images.
v2: Bump the number of supported image uniforms to 32 (Ken).
Reviewed-by: Paul Berry <stereotype441@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Fri, 14 Nov 2014 18:30:46 +0000 (20:30 +0200)]
i965: Don't tile 1D miptrees.
It doesn't really improve locality of texture fetches, quite the
opposite it's a waste of memory bandwidth and space due to tile
alignment.
v2: Check mt->logical_height0 instead of mt->target (Ken). Add short
comment explaining why they shouldn't be tiled.
Reviewed-by: Neil Roberts <neil@linux.intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Francisco Jerez [Wed, 4 Feb 2015 16:37:46 +0000 (18:37 +0200)]
i965/vec4: Don't set any dependency control bits for F32TO16 on Gen8.
It's expanded to several instructions.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Fri, 6 Feb 2015 12:38:20 +0000 (14:38 +0200)]
i965: Handle negated unsigned immediate values in constant propagation.
Negation of UD/UW sources behaves the same as for D/W sources, taking
the two's complement of the source, except for bitwise logical
operations on Gen8 and up which take the one's complement. Fixes
crash in a GLSL shader with subtraction of two unsigned values.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Tue, 3 Feb 2015 20:50:06 +0000 (22:50 +0200)]
i965/vec4: Take into account non-zero reg_offset during register allocation.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Tue, 3 Feb 2015 18:34:39 +0000 (20:34 +0200)]
i965/vec4: Add register classes up to MAX_VGRF_SIZE.
In preparation for some send from GRF instructions that will require
larger payloads.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Tue, 3 Feb 2015 20:52:37 +0000 (22:52 +0200)]
i965/vec4: Init mlen for several send from GRF instructions.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Tue, 3 Feb 2015 20:42:23 +0000 (22:42 +0200)]
i965/vec4: Don't infer MRF dependencies for send from GRF instructions.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Thu, 5 Feb 2015 20:39:33 +0000 (22:39 +0200)]
i965/vec4: Fix the scheduler to take into account reads and writes of multiple registers.
v2: Avoid nested ternary operators in vec4_instruction::regs_read(). (Matt)
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Thu, 5 Feb 2015 20:58:03 +0000 (22:58 +0200)]
i965/vec4: Make vec4_visitor::implied_mrf_writes() return zero for sends from GRF.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Tue, 3 Feb 2015 20:53:17 +0000 (22:53 +0200)]
i965/vec4: Pass dst register to the vec4_instruction constructor.
So regs_written gets initialized with a sensible value.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Tue, 3 Feb 2015 13:31:46 +0000 (15:31 +0200)]
i965/vec4: Initialize vec4_instruction::predicate and ::predicate_inverse.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Thu, 5 Feb 2015 20:40:07 +0000 (22:40 +0200)]
i965/vec4: Implement equals() method for dst_reg too.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Tue, 3 Feb 2015 19:57:38 +0000 (21:57 +0200)]
i965/fs: Fix fs_inst::regs_written calculation for instructions with scalar dst.
Scalar registers are required to have zero stride, fix the
regs_written calculation not to assume that the instruction writes
zero registers in that case.
v2: Rename CEILING() to DIV_ROUND_UP(). (Matt, Ken)
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Thu, 5 Feb 2015 23:14:51 +0000 (01:14 +0200)]
i965/fs: Fix stack allocation of fs_inst and stop stealing src array provided on construction.
Using 'ralloc*(this, ...)' is wrong if the object has automatic
storage or was allocated through any other means. Use normal dynamic
memory instead.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Tue, 3 Feb 2015 13:52:37 +0000 (15:52 +0200)]
i965/fs: Remove duplicate include of brw_shader.h
The second one was inside an extern "C" block, luckily it was being
discarded by the preprocessor.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Thu, 5 Feb 2015 23:28:12 +0000 (01:28 +0200)]
i965: Move up fs_inst::flag_subreg to backend_instruction.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Thu, 5 Feb 2015 23:27:40 +0000 (01:27 +0200)]
i965: Move up fs_inst::regs_written to backend_instruction.
It will also be useful in the VEC4 back-end.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Thu, 5 Feb 2015 23:25:50 +0000 (01:25 +0200)]
i965/vec4: Remove dependency of vec4_instruction on the visitor class.
The only reason why you need a vec4_visitor to construct a
vec4_instruction is to initialize vec4_instruction::ir and
::annotation. Instead set them from vec4_visitor::emit() just like
fs_visitor does.
Reviewed-by: Matt Turner <mattst88@gmail.com>
Francisco Jerez [Thu, 5 Feb 2015 23:24:17 +0000 (01:24 +0200)]
i965/fs: Remove dependency of fs_inst on the visitor class.
The fs_visitor argument of fs_inst::regs_read() wasn't used at all.
Reviewed-by: Matt Turner <mattst88@gmail.com>