mesa.git
6 years agoradeonsi/gfx9: update bin sizes
Marek Olšák [Fri, 1 Jun 2018 03:48:03 +0000 (23:48 -0400)]
radeonsi/gfx9: update bin sizes

This is based on our docs (recently updated), not amdvlk.

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi/gfx9: update primitive binning code for EQAA
Marek Olšák [Fri, 1 Jun 2018 03:47:17 +0000 (23:47 -0400)]
radeonsi/gfx9: update primitive binning code for EQAA

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi: assume that rasterizer state is non-NULL in draw_vbo
Marek Olšák [Thu, 31 May 2018 03:21:28 +0000 (23:21 -0400)]
radeonsi: assume that rasterizer state is non-NULL in draw_vbo

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi: micro-optimize prim checking and fix guardband with lines+adjacency
Marek Olšák [Thu, 31 May 2018 02:59:41 +0000 (22:59 -0400)]
radeonsi: micro-optimize prim checking and fix guardband with lines+adjacency

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi: move the guardband registers into a separate state atom
Marek Olšák [Thu, 31 May 2018 02:38:05 +0000 (22:38 -0400)]
radeonsi: move the guardband registers into a separate state atom

They have a different frequency of updates and don't change when scissors
change.

I think this even fixes something in si_update_vs_viewport_state.

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi/gfx9: implement the scissor bug workaround without performance drop
Marek Olšák [Thu, 31 May 2018 02:02:00 +0000 (22:02 -0400)]
radeonsi/gfx9: implement the scissor bug workaround without performance drop

This might improve performance on Vega10 and Raven.

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi: don't set VGT_LS_HS_CONFIG if it doesn't change
Marek Olšák [Thu, 31 May 2018 01:24:06 +0000 (21:24 -0400)]
radeonsi: don't set VGT_LS_HS_CONFIG if it doesn't change

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi: move VGT_GS_OUT_PRIM_TYPE into si_shader_gs
Marek Olšák [Thu, 31 May 2018 01:16:14 +0000 (21:16 -0400)]
radeonsi: move VGT_GS_OUT_PRIM_TYPE into si_shader_gs

same as amdvlk.

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi: record CLIPVERTEX output usage properly for compatibility profiles
Marek Olšák [Fri, 25 May 2018 21:37:51 +0000 (17:37 -0400)]
radeonsi: record CLIPVERTEX output usage properly for compatibility profiles

This was missed when adding CLIPVERTEX support into GS & tess.

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi: fix FBFETCH with 2D MSAA arrays
Marek Olšák [Fri, 18 May 2018 03:26:56 +0000 (23:26 -0400)]
radeonsi: fix FBFETCH with 2D MSAA arrays

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoac: handle undefined EQAA samples in ac_apply_fmask_to_sample
Marek Olšák [Fri, 18 May 2018 03:23:24 +0000 (23:23 -0400)]
ac: handle undefined EQAA samples in ac_apply_fmask_to_sample

RADV might wanna use this helper too.

Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
6 years agoradeonsi: return real memory usage instead of per-process usage
Marek Olšák [Wed, 13 Jun 2018 01:13:44 +0000 (21:13 -0400)]
radeonsi: return real memory usage instead of per-process usage

Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoac/gpu_info: report real total memory sizes
Marek Olšák [Wed, 13 Jun 2018 00:59:42 +0000 (20:59 -0400)]
ac/gpu_info: report real total memory sizes

The change from MIN2 to MAX2 is intentional.

Cc: 18.1 <mesa-stable@lists.freedesktop.org>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agodocs: mark virgl GL 4.0 features as complete.
Dave Airlie [Thu, 14 Jun 2018 00:37:28 +0000 (10:37 +1000)]
docs: mark virgl GL 4.0 features as complete.

virgl should now expose GL4.1 where it can.

6 years agovirgl: add ARB_tessellation_shader support. (v2)
Dave Airlie [Fri, 8 Jun 2018 04:38:14 +0000 (14:38 +1000)]
virgl: add ARB_tessellation_shader support. (v2)

This should add all the pieces to enable tess shaders on virgl.

v2: fixup transform to handle tess and strip out precise.
set default for max patch varyings to work around issue when
tess gets enabled from v1 caps but v2 caps aren't in place. (Elie)

Reviewed-by: Elie Tournier <elie.tournier@collabora.com>
6 years agoglsl: allow standalone semicolons outside main()
Dave Airlie [Wed, 13 Jun 2018 23:52:21 +0000 (09:52 +1000)]
glsl: allow standalone semicolons outside main()

GLSL 4.60 offically added this but games and older CTS suites actually
had shaders that did this, we may as well enable it everywhere.

Adding stable because it appears apps in the wild do this.

Acked-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Cc: <mesa-stable@lists.freedesktop.org>
6 years agoradv: don't fast clear HTILE for 16-bit depth surfaces on GFX8
Samuel Pitoiset [Wed, 13 Jun 2018 18:19:23 +0000 (20:19 +0200)]
radv: don't fast clear HTILE for 16-bit depth surfaces on GFX8

This causes rendering issues in Shadow Warrior 2 with DXVK.

Cc: mesa-stable@lists.freedesktop.org
Fixes: ccc64f3133 ("radv: enable TC-compat HTILE for 16-bit depth surfaces on GFX8")
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106912
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoconfigure.ac: Test for __atomic_add_fetch in atomic checks
Andrew Galante [Mon, 11 Jun 2018 22:05:25 +0000 (15:05 -0700)]
configure.ac: Test for __atomic_add_fetch in atomic checks

Some platforms have 64-bit __atomic_load_n but not 64-bit
__atomic_add_fetch, so test for both of them.

Bug: https://bugs.gentoo.org/655616
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
6 years agomeson: Test for __atomic_add_fetch in atomic checks
Andrew Galante [Mon, 11 Jun 2018 22:03:36 +0000 (15:03 -0700)]
meson: Test for __atomic_add_fetch in atomic checks

Some platforms have 64-bit __atomic_load_n but not 64-bit
__atomic_add_fetch, so test for both of them.

Bug: https://bugs.gentoo.org/655616
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
6 years agomeson: Fix -latomic check
Matt Turner [Mon, 11 Jun 2018 21:56:26 +0000 (14:56 -0700)]
meson: Fix -latomic check

Commit 54ba73ef102f (configure.ac/meson.build: Fix -latomic test) fixed
some checks for -latomic, and then commit 54bbe600ec26 (configure.ac:
rework -latomic check) further extended the fixes in configure.ac but
not in Meson. This commit extends those fixes to the Meson tests.

Fixes: 54bbe600ec26 (configure.ac: rework -latomic check)
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
6 years agomeson: Remove various completed todos
Dylan Baker [Tue, 12 Jun 2018 16:03:28 +0000 (09:03 -0700)]
meson: Remove various completed todos

v3: - Remove "won't do" todos, so only completed todo's are now removed.

Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com> (v2)
6 years agomeson: Make use of optional modules
Dylan Baker [Thu, 7 Jun 2018 18:22:48 +0000 (11:22 -0700)]
meson: Make use of optional modules

meson 0.43 gained support for optional modules, which clover wold like
to use. Since we require 0.44.1 now we can rely on them being available
for clover.

compile tested only.

Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
6 years agomeson: Add support for ppc assembly/optimizations
Dylan Baker [Thu, 7 Jun 2018 18:13:34 +0000 (11:13 -0700)]
meson: Add support for ppc assembly/optimizations

v2: - Use -mpower8-vector in compiler test for altivec
    - rename altivec option to power8
    - reword power8 option description to be more clear, originally I
      had made it a boolean, but replaced it with an auto option.

Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
6 years agomeson: Add support for SPARC assembly
Dylan Baker [Thu, 7 Jun 2018 17:58:06 +0000 (10:58 -0700)]
meson: Add support for SPARC assembly

This was blindly copied from autotools and tested by a helpful gentoo
user.

Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
6 years agomeson: Set include dirs for asm
Dylan Baker [Mon, 11 Jun 2018 17:12:46 +0000 (10:12 -0700)]
meson: Set include dirs for asm

v2: - split this from the next patch
    - Only include x86-64 and not x86 when buiding x86_64

Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
6 years agomeson: move cc and cpp definitions to top of main meson.build
Dylan Baker [Mon, 11 Jun 2018 17:15:15 +0000 (10:15 -0700)]
meson: move cc and cpp definitions to top of main meson.build

This just makes using cc and cpp easier.

v2: - Add this patch to fix altivec

Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
6 years agoRevert "intel/compiler: Properly consider UBO loads that cross 32B boundaries."
Jason Ekstrand [Wed, 13 Jun 2018 16:23:28 +0000 (09:23 -0700)]
Revert "intel/compiler: Properly consider UBO loads that cross 32B boundaries."

This reverts commit b8fa847c2ed9c7c743f31e57560a09fae3992f46.

This broke about 30k Vulkan CTS tests.

6 years agointel/compiler: Properly consider UBO loads that cross 32B boundaries.
Kenneth Graunke [Fri, 8 Jun 2018 21:24:16 +0000 (14:24 -0700)]
intel/compiler: Properly consider UBO loads that cross 32B boundaries.

The UBO push analysis pass incorrectly assumed that all values would fit
within a 32B chunk, and only recorded a bit for the 32B chunk containing
the starting offset.

For example, if a UBO contained the following, tightly packed:

   vec4 a;  // [0, 16)
   float b; // [16, 20)
   vec4 c;  // [20, 36)

then, c would start at offset 20 / 32 = 0 and end at 36 / 32 = 1,
which means that we ought to record two 32B chunks in the bitfield.

Similarly, dvec4s would suffer from the same problem.

Reviewed-by: Rafael Antognolli <rafael.antognolli@intel.com>
6 years agodrivers/dri/i965: add missing #include
Ross Burton [Tue, 12 Jun 2018 10:59:01 +0000 (11:59 +0100)]
drivers/dri/i965: add missing #include

brw_bufmgr.h uses time_t without include time.h, so the build fails under musl.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agoanv/android: Use an address for each anv_image plane
Mauro Rossi [Sun, 3 Jun 2018 18:41:59 +0000 (20:41 +0200)]
anv/android: Use an address for each anv_image plane

Fixes to avoid building error after change in image->planes[] structure,
{bo,bo_offset} has to be replaced by address.{bo,offset}
and update is needed also in the assert() for debug builds.

external/mesa/src/intel/vulkan/anv_android.c:188:21:
error: no member named 'bo' in 'struct anv_image::(anonymous at external/mesa/src/intel/vulkan/anv_private.h:2647:4)'
   image->planes[0].bo = bo;
   ~~~~~~~~~~~~~~~~ ^
1 error generated.

Fixes: bf34ef16ac ("anv: Use an address for each anv_image plane")
Signed-off-by: Mauro Rossi <issor.oruam@gmail.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agoanv/android: Set the BO flags in bo_cache_import (v2)
Mauro Rossi [Mon, 4 Jun 2018 00:48:09 +0000 (02:48 +0200)]
anv/android: Set the BO flags in bo_cache_import (v2)

Changes to avoid building error:

external/mesa/src/intel/vulkan/anv_android.c:131:72:
error: too few arguments to function call, expected 5, have 4
   result = anv_bo_cache_import(device, &device->bo_cache, dma_buf, &bo);
            ~~~~~~~~~~~~~~~~~~~                                        ^
1 error generated.

(v2) Set the correct bo_flags based on support of 48bit addresses and soft-pin

Fixes: b0d50247a7 ("anv/allocator: Set the BO flags in bo_cache_alloc/import")
Fixes: e7d0378bd9 ("anv: Soft-pin client-allocated memory")
Signed-off-by: Mauro Rossi <issor.oruam@gmail.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agoanv: Disable __gen_validate_value if NDEBUG is set.
Kenneth Graunke [Thu, 7 Jun 2018 22:44:43 +0000 (15:44 -0700)]
anv: Disable __gen_validate_value if NDEBUG is set.

We were enabling undefined memory checking for genxml values based on
Valgrind being installed at build time, even for release builds.  This
generates piles and piles of assembly whenever you touch genxml.

With gcc 7.3.1 and -O3 and -march=native on a Kabylake with Valgrind
installed at build time:

      text    data    bss     dec    hex filename
   5978385  262884  13488 6254757 5f70a5 libvulkan_intel.so
   3799377  262884  13488 4075749 3e30e5 libvulkan_intel.so

That's a 36% reduction in text size.

Fixes: 047ed02723071d7eccbed3210b5be6ae73603a53 (vk/emit: Use valgrind to validate every packed field)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agoREADME: wording fix for previous commit
Eric Engestrom [Mon, 11 Jun 2018 17:34:42 +0000 (18:34 +0100)]
README: wording fix for previous commit

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agoREADME: add link to WhosWho for IRC nicks
Eric Engestrom [Mon, 11 Jun 2018 17:31:06 +0000 (18:31 +0100)]
README: add link to WhosWho for IRC nicks

Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agoadd project README
Eric Engestrom [Fri, 8 Jun 2018 15:56:03 +0000 (16:56 +0100)]
add project README

Now that we're using GitLab, let's take advantage of the "landing page"
README feature with some minimal information, mostly to point people to
the right resources.

Acked-by: Dylan Baker <dylan@pnwbakers.com>
Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agoi965: fix resource leak
Eric Engestrom [Sun, 10 Jun 2018 17:42:45 +0000 (18:42 +0100)]
i965: fix resource leak

v2: intel_miptree_release() already takes care of the planes, no need
    to hand-code the loop (Lionel)

Coverity ID: 1436909
Fixes: 3352f2d746d3959b22ca4 "i965: Create multiple miptrees for planar YUV images"
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Signed-off-by: Eric Engestrom <eric@engestrom.ch>
6 years agofreedreno/ir3: use pipe_image_view's cpp
Rob Clark [Thu, 7 Jun 2018 19:44:28 +0000 (15:44 -0400)]
freedreno/ir3: use pipe_image_view's cpp

At least for PIPE_BUFFER, we could get the resource used as (for
example) R32F imageBuffer.  So using cpp=1 from the rsc is wrong.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/ir3: fix image dimensions offset
Rob Clark [Thu, 7 Jun 2018 17:54:15 +0000 (13:54 -0400)]
freedreno/ir3: fix image dimensions offset

copy-pasta fail from how SSBO sizes are handled.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/a5xx: correct image/ssbo offset
Rob Clark [Tue, 5 Jun 2018 13:50:26 +0000 (09:50 -0400)]
freedreno/a5xx: correct image/ssbo offset

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/ir3: use saml always if we have lod
Rob Clark [Mon, 4 Jun 2018 20:40:52 +0000 (16:40 -0400)]
freedreno/ir3: use saml always if we have lod

In some cases we get plain tex opcodes (but w/ a lod argument).. in this
case always use the saml instruction.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/ir3: don't cp absneg into meta:fi
Rob Clark [Mon, 4 Jun 2018 17:31:47 +0000 (13:31 -0400)]
freedreno/ir3: don't cp absneg into meta:fi

If using a fanin (collect) to collect of consecutive registers together,
we can CP mov's into the fanin, but not (abs) or (neg).  No places that
allow those modifiers are consuming a fanin anyways.  But this caused an
absneg to be lost between a ldgb and stgb for shaders like:

  outputs[n] = abs(input[n])

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/ir3: rework size/type conversion instructions
Rob Clark [Thu, 26 Apr 2018 18:54:24 +0000 (14:54 -0400)]
freedreno/ir3: rework size/type conversion instructions

With 8b and 16b, there are a lot more to handle.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/ir3: propagate HALF flag across fanout
Rob Clark [Thu, 26 Apr 2018 18:52:09 +0000 (14:52 -0400)]
freedreno/ir3: propagate HALF flag across fanout

If we have a fanout (split) meta instruction to split the result of a
vector instruction, propagate the HALF flag back to the original
instruction.  Otherwise result ends up in a full precision register
while instruction(s) that use the result look in a half-precision
register.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/a5xx: add sample-id/sample-mask-in
Rob Clark [Fri, 1 Jun 2018 18:56:57 +0000 (14:56 -0400)]
freedreno/a5xx: add sample-id/sample-mask-in

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/ir3: add sample-id/sample-mask-in
Rob Clark [Fri, 1 Jun 2018 18:56:38 +0000 (14:56 -0400)]
freedreno/ir3: add sample-id/sample-mask-in

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno: update generated headers
Rob Clark [Fri, 1 Jun 2018 15:59:04 +0000 (11:59 -0400)]
freedreno: update generated headers

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agofreedreno/ir3: image atomics use image-store path
Rob Clark [Thu, 31 May 2018 23:17:30 +0000 (19:17 -0400)]
freedreno/ir3: image atomics use image-store path

image reads are handled via tex state, whereas image writes and atomics
are handled via SSBO state block.  Previously we were only considering
image write, and not image atomics which also uses the SSBO state block.

Signed-off-by: Rob Clark <robdclark@gmail.com>
6 years agoegl/glvnd: Fix a segfault in eglGetProcAddress.
Kyle Brenneman [Wed, 6 Jun 2018 15:08:47 +0000 (09:08 -0600)]
egl/glvnd: Fix a segfault in eglGetProcAddress.

If FindProcIndex in egldispatchstubs.c is called with a name that's less than
the first entry in the array, it would end up trying to store an index of -1 in
an unsigned integer, wrap around to 2^32, and then crash when it tries to look
that up.

Change FindProcIndex so that it uses bsearch(3) instead of implementing its own
binary search, like the GLX equivalent FindGLXFunction does.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agomesa/program_binary: add implicit UseProgram after successful ProgramBinary
Jordan Justen [Wed, 6 Jun 2018 08:57:15 +0000 (01:57 -0700)]
mesa/program_binary: add implicit UseProgram after successful ProgramBinary

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106810
Fixes: b4c37ce2140 "i965: Add ARB_get_program_binary support using nir_serialization"
Ref: 3fe8d04a6d6 "mesa: don't always set _NEW_PROGRAM when linking"
Ref: c505d6d8522 "mesa: use gl_program for CurrentProgram rather than gl_shader_program"
Signed-off-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Plamena Manolova <plamena.manolova@intel.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
6 years agofeatures.txt: update virgl GL4.1 status.
Dave Airlie [Mon, 11 Jun 2018 00:49:14 +0000 (10:49 +1000)]
features.txt: update virgl GL4.1 status.

All the features for GL4.1 are done (64-bit attribs were part of
the fp64 enable).

Once tessellation shaders land this will be advertised

6 years agovirgl: enable ARB_gpu_shader_fp64
Dave Airlie [Mon, 21 May 2018 04:00:37 +0000 (14:00 +1000)]
virgl: enable ARB_gpu_shader_fp64

This enables ARB_gpu_shader_fp64 if the host provides it.

Tested-by: Gurchetan Singh <gurchetansingh@chromium.org>
Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
6 years agoradv: add a workaround for DXVK hangs by setting amdgpu-skip-threshold
Samuel Pitoiset [Fri, 8 Jun 2018 09:38:01 +0000 (11:38 +0200)]
radv: add a workaround for DXVK hangs by setting amdgpu-skip-threshold

Workaround for bug in llvm that causes the GPU to hang in presence
of nested loops because there is an exec mask issue. The proper
solution is to fix LLVM but this might require a bunch of work.

This fixes a bunch of GPU hangs that happen with DXVK.

Vega10:
Totals from affected shaders:
SGPRS: 110456 -> 110456 (0.00 %)
VGPRS: 122800 -> 122800 (0.00 %)
Spilled SGPRs: 7478 -> 7478 (0.00 %)
Spilled VGPRs: 36 -> 36 (0.00 %)
Code Size: 9901104 -> 9922928 (0.22 %) bytes
Max Waves: 7143 -> 7143 (0.00 %)

Code size slightly increases because it inserts more branch
instructions but that's expected. I don't see any real performance
changes.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105613
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoradv: fix missing ZRANGE_PRECISION(1) for GFX9+
Samuel Pitoiset [Fri, 8 Jun 2018 15:59:49 +0000 (17:59 +0200)]
radv: fix missing ZRANGE_PRECISION(1) for GFX9+

ZRANGE_PRECISION(1) seems to be the default optimal value, but
it was only set for VI and older chips.

This fixes a rendering issue with Banished through DXVK, and
might fix more than that.

There is still the ZRANGE_PRECISION bug that we need to handle
but that can be fixed later.

Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoanv: enable VK_EXT_shader_stencil_export
Gustavo Lima Chaves [Tue, 20 Mar 2018 06:06:45 +0000 (23:06 -0700)]
anv: enable VK_EXT_shader_stencil_export

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agospirv: add/hookup SpvCapabilityStencilExportEXT
Gustavo Lima Chaves [Mon, 12 Feb 2018 02:26:40 +0000 (18:26 -0800)]
spirv: add/hookup SpvCapabilityStencilExportEXT

v2:
An attempt to support SpvExecutionModeStencilRefReplacingEXT's behavior
also follows, with the interpretation to said mode being we prevent
writes to the built-in FragStencilRefEXT variable when the execution
mode isn't set.

v3:
A more cautious reading of 1db44252d01bf7539452ccc2b5210c74b8dcd573 led
me to a missing change that would stop (what I later discovered were)
GPU hangs on the CTS test written to exercise this.

v4:
Turn FragStencilRefEXT decoration usage without StencilRefReplacingEXT
mode into a warning, instead of trying to make the variable read-only.
If we are to follow the originating extension on GL, the built-in
variable in question should never be readable anyway.

v5/v6: rebases.

v7:
Fix check for gen9 lost in rebase. (Ilia)
Reduce the scope of the bool used to track whether
SpvExecutionModeStencilRefReplacingEXT was used. Was in shader_info,
moved to vtn_builder. (Jason)

v8:
Assert for fragment shader handling StencilRefReplacingEXT execution
mode. (Caio)
Remove warning logic, since an entry point might not have
StencilRefReplacingEXT execution mode, but the global output variable
might still exist for another entry point in the module. (Jason)

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agotravis: Add the v3d driver to the automake build.
Eric Anholt [Mon, 4 Jun 2018 22:32:14 +0000 (15:32 -0700)]
travis: Add the v3d driver to the automake build.

Hopefully this reduces the number of fixup commits we need for the
automake build.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
6 years agotravis: Do our automake build tests with srcdir != builddir.
Eric Anholt [Tue, 5 Jun 2018 15:18:51 +0000 (08:18 -0700)]
travis: Do our automake build tests with srcdir != builddir.

This will catch many automake bugs that end-users get to experience first,
otherwise.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
6 years agoautotools/meson: compile against wayland-egl-*backend*
Eric Engestrom [Thu, 7 Jun 2018 14:45:01 +0000 (15:45 +0100)]
autotools/meson: compile against wayland-egl-*backend*

Bug: https://bugs.freedesktop.org/show_bug.cgi?id=106861
Fixes: 1db4ec05462914096b1f "egl: rewire the build systems to use libwayland-egl"
Suggested-by: Emil Velikov <emil.l.velikov@gmail.com>
Tested-by: Andreas Hartmetz <ahartmetz@gmail.com>
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agovulkan/wsi: Destroy swapchain images after terminating FIFO queues
Cameron Kumar [Fri, 1 Jun 2018 11:16:19 +0000 (12:16 +0100)]
vulkan/wsi: Destroy swapchain images after terminating FIFO queues

The queue_manager thread can access the images from x11_present_to_x11,
hence this reorder prevents dereferencing of dangling pointers.

Cc: "18.1" <mesa-stable@lists.freedesktop.org>
Fixes: e73d136a023080 ("vulkan/wsi/x11: Implement FIFO mode.")
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
6 years agoradeonsi: emit_dpbb_state packets optimization
Sonny Jiang [Thu, 7 Jun 2018 16:13:53 +0000 (12:13 -0400)]
radeonsi: emit_dpbb_state packets optimization

Remembering latest states of registers to eliminate redunant SET_CONTEXT_REG packets

Signed-off-by: Sonny Jiang <sonny.jiang@amd.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
6 years agoradeonsi: emit_clip_state packets optimization
Sonny Jiang [Thu, 7 Jun 2018 16:13:52 +0000 (12:13 -0400)]
radeonsi: emit_clip_state packets optimization

Remembering latest states of registers to eliminate redunant SET_CONTEXT_REG packets

Signed-off-by: Sonny Jiang <sonny.jiang@amd.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
6 years agoradeonsi: emit_msaa_sample_locs packets optimization
Sonny Jiang [Thu, 7 Jun 2018 16:13:51 +0000 (12:13 -0400)]
radeonsi: emit_msaa_sample_locs packets optimization

Remembering latest states of registers to eliminate redunant SET_CONTEXT_REG packets

Signed-off-by: Sonny Jiang <sonny.jiang@amd.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
6 years agoradeonsi: emit_msaa_config packets optimization
Sonny Jiang [Thu, 7 Jun 2018 16:13:50 +0000 (12:13 -0400)]
radeonsi: emit_msaa_config packets optimization

Remembering latest states of registers to eliminate redunant SET_CONTEXT_REG packets

Signed-off-by: Sonny Jiang <sonny.jiang@amd.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
6 years agoradeonsi: emit_cb_render_state packets optimization
Sonny Jiang [Thu, 7 Jun 2018 16:13:49 +0000 (12:13 -0400)]
radeonsi: emit_cb_render_state packets optimization

Remembering latest states of registers to eliminate redunant SET_CONTEXT_REG packets

Signed-off-by: Sonny Jiang <sonny.jiang@amd.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
6 years agoradeonsi: emit_db_render_state packets optimization
Sonny Jiang [Thu, 7 Jun 2018 16:13:48 +0000 (12:13 -0400)]
radeonsi: emit_db_render_state packets optimization

Remembering latest states of registers to eliminate redunant SET_CONTEXT_REG packets

Signed-off-by: Sonny Jiang <sonny.jiang@amd.com>
Signed-off-by: Marek Olšák <marek.olsak@amd.com>
6 years agodrisw: Fix invalid pointer arithmetic
Jan Vesely [Thu, 7 Jun 2018 21:24:59 +0000 (17:24 -0400)]
drisw: Fix invalid pointer arithmetic

Use of void * in pointer arithmetic is illegal, use char * instead.
Fixes: cf54bd5e8381dba18d52fe438acda20cc1685bf3 ("drisw: use shared memory when possible")
Reviewed-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Jan Vesely <jan.vesely@rutgers.edu>
6 years agoradeonsi: fix possible truncation on renderer string
Timothy Arceri [Thu, 7 Jun 2018 01:03:10 +0000 (11:03 +1000)]
radeonsi: fix possible truncation on renderer string

Fixes truncation warning in gcc 8.1

Fixes: 8539c9bf3158 ("gallium/radeon: add the kernel version into the renderer string")
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
6 years agoac: fix possible truncation of intrinsic name
Timothy Arceri [Thu, 7 Jun 2018 00:30:01 +0000 (10:30 +1000)]
ac: fix possible truncation of intrinsic name

Fixes the gcc warning:
snprintf’ output between 26 and 33 bytes into a destination of size 32

Fixes: d5f7ebda3ec0 ("ac: add LLVM build functions for subgroup instrinsics")
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agoamd/common: Fix number of coords for getlod.
Bas Nieuwenhuizen [Tue, 5 Jun 2018 23:42:17 +0000 (01:42 +0200)]
amd/common: Fix number of coords for getlod.

The LLVM 6 code reduced it to a non-array call. We need to do that
with the new code too.

This fixes dEQP-VK.glsl.texture_functions.query.texturequerylod.*array* for radv.

Fixes: a9a79934412 "amd/common: use the dimension-aware image intrinsics on LLVM 7+"
Reviewed-by: Dave Airlie <airlied@redhat.com>
6 years agofeatures: add virgl to the GL features list
Dave Airlie [Thu, 7 Jun 2018 00:41:53 +0000 (10:41 +1000)]
features: add virgl to the GL features list

This hopefully adds virgl to the correct places and current statuses
of various extensions.

virgl of course relies on two external things
a) host driver that can support the features
b) up to date host virglrenderer library that can support the features.

This list will be maintained as latest (a) + (b) + mesa.

Reviewed-by: Gurchetan Singh <gurchetansingh@chromium.org>
6 years agomeson: Add support for read-only text segment on x86
Matt Turner [Thu, 7 Jun 2018 03:25:09 +0000 (20:25 -0700)]
meson: Add support for read-only text segment on x86

Port of 6dfc5e28f7d0 (configure.ac: Add support to enable read-only text
segment on x86.) to Meson.

Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
6 years agomeson: work around gentoo applying -m32 to host compiler in cross builds
Dylan Baker [Thu, 7 Jun 2018 17:48:38 +0000 (10:48 -0700)]
meson: work around gentoo applying -m32 to host compiler in cross builds

Gentoo's ebuild system always adds -m32 to the compiler for doing x86_64
-> x86 cross builds, while meson expects it not to do that. This
results in an x86 -> x86 cross build, and assembly gets disabled.

Fixes: 2d62fc06465281d3d45b8a7c7fd2b17ef718448c
       ("meson: disable x86 asm in fewer cases.")
Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
6 years agoi965/screen: Sanity check that all formats we advertise are useable
Jason Ekstrand [Wed, 6 Jun 2018 17:36:08 +0000 (10:36 -0700)]
i965/screen: Sanity check that all formats we advertise are useable

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
6 years agoi965/screen: Use RGBA non-sRGB formats for images
Jason Ekstrand [Tue, 5 Jun 2018 18:42:37 +0000 (11:42 -0700)]
i965/screen: Use RGBA non-sRGB formats for images

Not all of the MESA_FORMAT and ISL_FORMAT helpers we use can properly
handle RGBX formats.  Also, we don't want to make decisions based on
those in the first place because we can't render to RGBA and we use the
non-sRGB version to determine whether or not to allow CCS_E.

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
6 years agoi965/screen: Return false for unsupported formats in query_modifiers
Jason Ekstrand [Wed, 6 Jun 2018 17:24:01 +0000 (10:24 -0700)]
i965/screen: Return false for unsupported formats in query_modifiers

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
6 years agoi965/screen: Refactor query_dma_buf_formats
Jason Ekstrand [Tue, 5 Jun 2018 18:13:40 +0000 (11:13 -0700)]
i965/screen: Refactor query_dma_buf_formats

This reworks it to work like query_dma_buf_modifiers and, in particular,
makes it more flexible so that we can disallow a non-static set of
formats.

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
6 years agointel/isl: Add bounds-checking assertions for the format_info table
Jason Ekstrand [Wed, 6 Jun 2018 16:59:25 +0000 (09:59 -0700)]
intel/isl: Add bounds-checking assertions for the format_info table

We follow the same convention as isl_format_get_layout in having two
assertions to ensure that only valid formats are passed in.  We also
check against the array size of the table because some valid formats
such as CCS formats will may be past the end of the table.  This fixes
some potential out-of-bounds array access even in valid cases.

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
6 years agointel/isl: Add bounds-checking assertions in isl_format_get_layout
Jason Ekstrand [Wed, 6 Jun 2018 16:54:17 +0000 (09:54 -0700)]
intel/isl: Add bounds-checking assertions in isl_format_get_layout

We add two assertions instead of one because the first assertion that
format != ISL_FORMAT_UNSUPPORTED is more descriptive and checks for a
real but unsupported enumerant while the second ensures that they don't
pass in garbage values.  We also update some other helpers to use
isl_format_get_layout instead of using the table directly so that they
get bounds checking too.

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
6 years agomeson: Clarify why asm cannot be used in cross compile
Dylan Baker [Thu, 7 Jun 2018 15:38:07 +0000 (08:38 -0700)]
meson: Clarify why asm cannot be used in cross compile

This makes the reasoning for why a cross compile is not using asm
clearer (hopefully).

v2: - fix typos

Signed-off-by: Dylan Baker <dylan.c.baker@intel.com>
Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agodocs: talk about Wayland instead of libwayland
Eric Engestrom [Tue, 29 May 2018 14:24:53 +0000 (15:24 +0100)]
docs: talk about Wayland instead of libwayland

Reviewed-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agoanv: Set fence/semaphore types to NONE in impl_cleanup
Jason Ekstrand [Sat, 26 May 2018 19:08:22 +0000 (12:08 -0700)]
anv: Set fence/semaphore types to NONE in impl_cleanup

There were some places that were calling anv_semaphore_impl_cleanup and
neither deleting the semaphore nor setting the type back to NONE.  Just
set it to NONE in impl_cleanup to avoid these issues.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106643
Fixes: 031f57eba "anv: Add a basic implementation of VK_KHX_external..."
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
6 years agonir: Add global invocation id intrinsic.
Plamena Manolova [Fri, 25 May 2018 21:49:47 +0000 (22:49 +0100)]
nir: Add global invocation id intrinsic.

Add the missing nir intrinsic for the gl_GlobalInvocationID
compute shader variable.

Signed-off-by: Plamena Manolova <plamena.manolova@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
6 years agotravis: bump libwayland to the first version with libwayland-egl
Eric Engestrom [Tue, 29 May 2018 15:19:59 +0000 (16:19 +0100)]
travis: bump libwayland to the first version with libwayland-egl

Reviewed-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
6 years agoi965: Require softpin support for Cannonlake and later.
Kenneth Graunke [Tue, 10 Apr 2018 08:27:56 +0000 (01:27 -0700)]
i965: Require softpin support for Cannonlake and later.

This isn't strictly necessary, but anyone running Cannonlake will
already have Kernel 4.5 or later, so there's no reason to support
the relocation model on Gen10+.

This will let us avoid dealing with them for new features.

Reviewed-by: Scott D Phillips <scott.d.phillips@intel.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
6 years agoi965: Allocate VMA in userspace for full-PPGTT systems.
Kenneth Graunke [Tue, 10 Apr 2018 08:18:25 +0000 (01:18 -0700)]
i965: Allocate VMA in userspace for full-PPGTT systems.

This patch enables soft-pinning of all buffers, allowing us to skip
relocation processing entirely.  All systems with full PPGTT and > 4GB
of VMA should gain these benefits.  This should be most Gen8+.

Unfortunately, this excludes a few systems:
- Cherryview (only has 32-bit addressing, despite 48-bit pointers)
- Broadwell with a 32-bit kernel
- Anybody running pre-4.5 kernel.

We may enable it for Cherryview in the future, but it would require
some tweaks to the memory zone.

Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
6 years agointel/blorp: Emit VF cache invalidates for 48-bit bugs with softpin.
Kenneth Graunke [Thu, 31 May 2018 23:39:16 +0000 (16:39 -0700)]
intel/blorp: Emit VF cache invalidates for 48-bit bugs with softpin.

commit 92f01fc5f914fd500497d0c3aed75f3ac8dc054d made i965 start emitting
VF cache invalidates when the high bits of vertex buffers change.  But
we were not tracking vertex buffers emitted by BLORP.  This was papered
over by a mistake where I emitted VF cache invalidates all the time,
which Chris fixed in commit 3ac5fbadfd8644d30fce9ff267cb811ad157996a.

This patch adds a new hook which allows the driver to track addresses
and request a VF cache invalidate as appropriate.

v2: Make the driver do the PIPE_CONTROL so it can apply workarounds
    (caught by Jason Ekstrand).  Rebase on anv bug fix.
v3: Don't screw up the boolean (caught by Jason Ekstrand).

Fixes: 92f01fc5f914 ("i965: Emit VF cache invalidates for 48-bit addressing bugs with softpin.")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agonir: add opt_if_loop_terminator()
Timothy Arceri [Fri, 1 Jun 2018 05:37:28 +0000 (15:37 +1000)]
nir: add opt_if_loop_terminator()

This pass detects potential loop terminators and moves intructions
from the non breaking branch after the if-statement.

This enables both the new opt_if_simplification() pass and loop
unrolling to potentially progress further.

Unexpectedly this change speed up shader-db run times by ~3%

Ivy Bridge shader-db results (all changes in dolphin/ubershaders):

total instructions in shared programs: 9995662 -> 9995338 (-0.00%)
instructions in affected programs: 87845 -> 87521 (-0.37%)
helped: 27
HURT: 0

total cycles in shared programs: 230931495 -> 230925015 (-0.00%)
cycles in affected programs: 56391385 -> 56384905 (-0.01%)
helped: 27
HURT: 0

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agonir: move ends_in_break() helper to nir_loop_analyze.h
Timothy Arceri [Fri, 1 Jun 2018 05:37:27 +0000 (15:37 +1000)]
nir: move ends_in_break() helper to nir_loop_analyze.h

We will use the helper while simplifying potential loop terminators
in the following patch.

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agoradv: fix Coverity no effect control flow issue
Timothy Arceri [Wed, 6 Jun 2018 23:49:37 +0000 (09:49 +1000)]
radv: fix Coverity no effect control flow issue

swizzle is unsigned so "desc->swizzle[c] < 0" is never true.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
6 years agointel/blorp: Don't vertex fetch directly from clear values
Jason Ekstrand [Tue, 5 Jun 2018 00:27:53 +0000 (17:27 -0700)]
intel/blorp: Don't vertex fetch directly from clear values

On gen8+, we have to VF cache flush whenever a vertex binding aliases a
previous binding at the same index modulo 4GiB.  We deal with this in
Vulkan by ensuring that vertex buffers and the dynamic state (from which
BLORP pulls its vertex buffers) are in the same 4GiB region of the
address space.  That doesn't work if we're reading clear colors with the
VF unit.  In order to work around this we switch to using MI commands to
copy the clear value into the vertex buffer we allocate for the normal
constant data.

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
6 years agodri: add missing 16bits formats mapping
Lionel Landwerlin [Wed, 6 Jun 2018 11:57:18 +0000 (12:57 +0100)]
dri: add missing 16bits formats mapping

i965 advertises the 16-bit R and RG formats through
eglQueryDmaBufFormatsEXT but falls over when a client tries to use or
asks more information about such a format because
driImageFormatToGLFormat returns MESA_FORMAT_NONE.

Found by Eero Tamminen.

v2: Add G16R16 formats (Lionel)

v3: Fix G16R16 mapping to mesa format (Jason)

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106642
Reviewed-by: Plamena Manolova <plamena.manolova@intel.com> (v2)
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agonir: Look into uniform structs for samplers when counting num_textures.
Eric Anholt [Fri, 30 Mar 2018 23:04:34 +0000 (16:04 -0700)]
nir: Look into uniform structs for samplers when counting num_textures.

mesa/st decides whether to update samplers after a program change based on
whether num_textures is nonzero.  By not counting samplers in a uniform
struct, we would segfault in
KHR-GLES3.shaders.struct.uniform.sampler_vertex if it was run in the same
context after a non-vertex-shader-uniform testcase (as is the case during
a full conformance run).

v2: Implement using two separate pure functions instead of updating
    pointers.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
6 years agov3d: Work around GFXH-1461/GFXH-1689 by using CLEAR_TILE_BUFFERS.
Eric Anholt [Wed, 2 May 2018 18:45:07 +0000 (11:45 -0700)]
v3d: Work around GFXH-1461/GFXH-1689 by using CLEAR_TILE_BUFFERS.

This doesn't seem to have done anything to my test results.  However,
given that we've still got a class of GPU hangs, following the workarounds
that the closed driver does so that we get the same command sequences
seems like a good idea.

6 years agov3d: Enable the new NIR bitfield operation lowering paths.
Eric Anholt [Wed, 2 May 2018 21:17:07 +0000 (14:17 -0700)]
v3d: Enable the new NIR bitfield operation lowering paths.

These together get the GLSL 3.00 unorm/snorm pack functions and
MESA_shader_integer operations working.

v2: Fix commit message typo.

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agonir: Add lowering for nir_op_bit_count.
Eric Anholt [Tue, 8 May 2018 20:04:37 +0000 (13:04 -0700)]
nir: Add lowering for nir_op_bit_count.

This is basically the same as the GLSL lowering path.

v2: Fix typo in the link

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agonir: Add lowering for nir_op_bitfield_reverse.
Eric Anholt [Tue, 8 May 2018 19:47:48 +0000 (12:47 -0700)]
nir: Add lowering for nir_op_bitfield_reverse.

This is basically the same as the GLSL lowering path.

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agonir: Add an ALU lowering pass for mul_high.
Eric Anholt [Tue, 8 May 2018 18:24:40 +0000 (11:24 -0700)]
nir: Add an ALU lowering pass for mul_high.

This is based on the glsl/lower_instructions.cpp implementation, but
should be much more readable.

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agonir: Add lowering for find_lsb.
Eric Anholt [Fri, 4 May 2018 21:02:55 +0000 (14:02 -0700)]
nir: Add lowering for find_lsb.

There is a fairly simple relation to turn this into ufind_msb.

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agonir: Add lowering for ifind_msb to ufind_msb.
Eric Anholt [Fri, 4 May 2018 20:33:47 +0000 (13:33 -0700)]
nir: Add lowering for ifind_msb to ufind_msb.

ufind_msb is easily expressed in terms of clz, and we can reduce ifind_msb
to that.

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
6 years agonir: Add lowering from ibitfield_extract/ubitfield_extract to shifts.
Eric Anholt [Wed, 2 May 2018 20:02:21 +0000 (13:02 -0700)]
nir: Add lowering from ibitfield_extract/ubitfield_extract to shifts.

V3D doesn't have opcodes for ibfe/ubfe, so we need to lower similarly to
glsl/lower_instructions.cpp.

Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>