mesa.git
5 years agonir/lower_input_attachments: pass on non-uniform access flag
Rhys Perry [Sat, 21 Sep 2019 14:50:25 +0000 (15:50 +0100)]
nir/lower_input_attachments: pass on non-uniform access flag

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
5 years agonir/lower_non_uniform: lower image/texture instructions taking derefs
Rhys Perry [Sat, 21 Sep 2019 14:49:18 +0000 (15:49 +0100)]
nir/lower_non_uniform: lower image/texture instructions taking derefs

v2: always assert on the texture/sampler handle's num_components
v3: replicate the deref inside the loop
v4: remove a case of useless line wrapping

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
5 years agoetnaviv: rework etna_resource_create tiling choice
Jonathan Marek [Wed, 24 Jul 2019 14:35:46 +0000 (10:35 -0400)]
etnaviv: rework etna_resource_create tiling choice

Now that the base resource is allowed to be incompatible with PE, we can
make a smarter choice of tiling mode to avoid allocating a PE compatible
base that is never used for regular textures. This affects GPUs like GC2000
where there is no tiling compatible with both PE and TE.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
5 years agoetnaviv: rework compatible render base
Jonathan Marek [Sat, 10 Aug 2019 21:00:32 +0000 (17:00 -0400)]
etnaviv: rework compatible render base

For PE-incompatible layouts, use a mechanism similar to what texture does
to create a compatible base resource.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
5 years agoetnaviv: get addressing mode from tiling layout
Jonathan Marek [Sat, 10 Aug 2019 20:44:07 +0000 (16:44 -0400)]
etnaviv: get addressing mode from tiling layout

Remove the "addressing_mode" state, which is currently set incorrectly, and
instead deduce the addressing mode from the tiling layout.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
5 years agoetnaviv: clear texture cache and flush ts when texture is modified
Jonathan Marek [Tue, 2 Jul 2019 17:34:02 +0000 (13:34 -0400)]
etnaviv: clear texture cache and flush ts when texture is modified

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
5 years agoetnaviv: output the same shader-db format as freedreno, v3d and intel
Christian Gmeiner [Sat, 5 Oct 2019 12:53:46 +0000 (14:53 +0200)]
etnaviv: output the same shader-db format as freedreno, v3d and intel

This lets us reuse their report.py.

Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Jonathan Marek <jonathan@marek.ca>
5 years agoetnaviv: nir: start to make use of compile_error(..)
Christian Gmeiner [Thu, 3 Oct 2019 20:14:20 +0000 (22:14 +0200)]
etnaviv: nir: start to make use of compile_error(..)

Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Jonathan Marek <jonathan@marek.ca>
5 years agogitlab-ci: Disable meson-mingw32-x86_64 job again for now
Michel Dänzer [Fri, 11 Oct 2019 09:10:01 +0000 (11:10 +0200)]
gitlab-ci: Disable meson-mingw32-x86_64 job again for now

The wrapdb.mesonbuild.com SSL certificate expired, causing the job to
fail: https://gitlab.freedesktop.org/mesa/mesa/-/jobs/731864

Switching to http:// doesn't avoid it:
https://gitlab.freedesktop.org/daenzer/mesa/-/jobs/732043

5 years agogitlab-ci: Add .use-debian-10 template
Michel Dänzer [Thu, 10 Oct 2019 13:27:17 +0000 (15:27 +0200)]
gitlab-ci: Add .use-debian-10 template

It simplifies the definitions of jobs using the Debian 10 image.

The needs: was previously missing from the llvmpipe/softpipe test jobs,
so they could spuriously run if the debian-10 job failed or was
cancelled.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
5 years agogitlab-ci: Remove redundant .meson-cross template script
Michel Dänzer [Tue, 8 Oct 2019 13:58:27 +0000 (15:58 +0200)]
gitlab-ci: Remove redundant .meson-cross template script

It was identical to the one inherited from the .meson-build template.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
5 years agogallivm: fix coroutines on aarch64 with llvm 8
Dave Airlie [Thu, 10 Oct 2019 03:27:30 +0000 (13:27 +1000)]
gallivm: fix coroutines on aarch64 with llvm 8

The coroutine split pass is missing a dependency before LLVM 9.0,
and fails to initialise properly if the CallGraphWrapperPass hasn't
be initialised earlier (x86 does it due to some of it's passes
requiring it).

This is a workaround for llvm 8 (coroutines are only supported in 8
and higher). It adds another pass that has a dependency on the pass
the coroutines split requires. This pass shouldn't have any raal
effects.

Fixes: d32690b43c9 (gallivm: add coroutine pass manager support)
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
5 years agollvmpipe: add support for tg4 component selection.
Dave Airlie [Wed, 9 Oct 2019 03:12:31 +0000 (13:12 +1000)]
llvmpipe: add support for tg4 component selection.

This is needed as part of GLES3.1 and helps for ARB_gpu_shader5.

Fixes: KHR-GLES31.core.texture_gather.* cases
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
5 years agost/glsl: add support for alternate TG4 encoding.
Dave Airlie [Wed, 9 Oct 2019 03:09:00 +0000 (13:09 +1000)]
st/glsl: add support for alternate TG4 encoding.

This will encode the component selection value (0, 1, 2, 3) into
the X swizzle of the sampler, if the driver requests it.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
5 years agogallium: add a a new cap for changing the TGSI TG4 instruction encoding
Dave Airlie [Wed, 9 Oct 2019 03:07:04 +0000 (13:07 +1000)]
gallium: add a a new cap for changing the TGSI TG4 instruction encoding

Accessing the TG4 component via immediates in the llvmpipe backend is quite
messy (like really messy). Roland suggested we change the instruction encoding,
so introduce a cap to allow the component to be selected to be store in the
sampler swizzle, which should be otherwise unused.

I could probably switch all drivers over, but virgl would need some work that
I'd prefer not to rush it.

Signed-off-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
5 years agogallivm/sample: add gather component selection to the key.
Dave Airlie [Fri, 30 Aug 2019 00:30:00 +0000 (10:30 +1000)]
gallivm/sample: add gather component selection to the key.

This allows for component selection to work as per ARB_gpu_shader5/GLES3.1

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
5 years agollvmpipe: increase max texture size to 2GB
Roland Scheidegger [Thu, 10 Oct 2019 18:10:34 +0000 (20:10 +0200)]
llvmpipe: increase max texture size to 2GB

The 1GB limit was arbitrary, increase this to 2GB (which is the max
possible without code changes).

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
5 years agogitlab-ci: Add a mingw x86_64 job
Dylan Baker [Thu, 19 Sep 2019 17:21:51 +0000 (10:21 -0700)]
gitlab-ci: Add a mingw x86_64 job

Acked-by: Eric Engestrom <eric.engestrom@intel.com> (v1)
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agoappveyor: Add support for meson as well as scons on windows
Dylan Baker [Wed, 23 Jan 2019 22:21:26 +0000 (14:21 -0800)]
appveyor: Add support for meson as well as scons on windows

This job uses the vs2017 backend of meson (msbuild) as opposed to the
ninja backend used on MacOS and Linux.

v7: - rebase on master
    - remove llvm (we'll add that back later)
    - remove cygwin (we'll add that back later too)
v6: - rebase on master, including the addition of cygwin
    - consolidate 3 appveyor patches into this one patch
v5  - use the new b_vscrt option instead of manually specifying the crt
v4: - rebase on python3 generators
    - cache meson wraps
    - Build x86 instead of x86_64, since that's what the pre-built LLVM
      is
    - update to vs2017 from vs2015
    - set the default-library to static
    - use the new vscrt override
    - add the /m switch to msbuild to make the build somewhat faster

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agodocs: update meson docs for windows
Dylan Baker [Tue, 11 Jun 2019 18:25:31 +0000 (11:25 -0700)]
docs: update meson docs for windows

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agoglsl/tests: Handle no-exec errors
Dylan Baker [Mon, 7 Oct 2019 20:03:58 +0000 (13:03 -0700)]
glsl/tests: Handle no-exec errors

Currently meson doesn't correctly handle passing compiled binaries to
scripts in tests. This patch looks to the future (0.53) when meson will
have this functionality, but also immediately it fixes these tests in
cross compiles by causing them to return 77, which meson interprets as
skip.

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson/util: Don't run string_buffer tests on mingw
Dylan Baker [Tue, 8 Oct 2019 20:38:01 +0000 (13:38 -0700)]
meson/util: Don't run string_buffer tests on mingw

They succeed with MSVC but not with MinGW. I don't understand why they
fail.

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: glcpp tests are expected to fail on windows
Dylan Baker [Fri, 20 Sep 2019 20:09:54 +0000 (13:09 -0700)]
meson: glcpp tests are expected to fail on windows

v2: - Exclude the tests rather than xfail them

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: only build timspec test if timespec is available
Dylan Baker [Fri, 20 Sep 2019 19:19:52 +0000 (12:19 -0700)]
meson: only build timspec test if timespec is available

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: don't error on formaters with mingw
Dylan Baker [Thu, 27 Jun 2019 21:52:40 +0000 (14:52 -0700)]
meson: don't error on formaters with mingw

MSVC is generally happy, but mingw errors. I've spent as much time
(several days) trying to squash all of these warnings and I'm done with
it, just leave them as warnings with MinGW.

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: add msvc compat args to swr
Dylan Baker [Tue, 29 Jan 2019 00:07:20 +0000 (16:07 -0800)]
meson: add msvc compat args to swr

This has always been present in the scons build, so it should be in
the meson build as well.

Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: maintain names of shared API libraries
Dylan Baker [Tue, 10 Jul 2018 18:09:27 +0000 (11:09 -0700)]
meson: maintain names of shared API libraries

Mesa uses the lib prefix, and doesn't use a version for it's dynamic
libraries, which meson defaults to.

v2: - this patch

Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: don't build or run mesa-sha1 test on windows
Dylan Baker [Wed, 23 May 2018 00:32:21 +0000 (17:32 -0700)]
meson: don't build or run mesa-sha1 test on windows

It crashes hard (pop-up window and all).

v2: - Change comment to FIXME

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: disable graw tests on mingw
Dylan Baker [Wed, 23 May 2018 00:22:45 +0000 (17:22 -0700)]
meson: disable graw tests on mingw

I can't figure out why symbols are being exposed that shouldn't.

v2: - change comment to FIXME

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: don't build gallium trivial tests on windows
Dylan Baker [Wed, 23 May 2018 00:06:35 +0000 (17:06 -0700)]
meson: don't build gallium trivial tests on windows

They require the pipe-loaders, which require xmlconfig, which doesn't
build with msvc.

Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: Set visibility and compat args for graw
Dylan Baker [Wed, 23 May 2018 00:06:09 +0000 (17:06 -0700)]
meson: Set visibility and compat args for graw

Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: Add msvc compat args to util/tests
Dylan Baker [Tue, 22 May 2018 22:00:35 +0000 (15:00 -0700)]
meson: Add msvc compat args to util/tests

To keep this building with msvc

Acked-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: Add idep_getopt for tests
Dylan Baker [Tue, 22 May 2018 21:57:14 +0000 (14:57 -0700)]
meson: Add idep_getopt for tests

There are quite a few tests that require getopt, when using MSVC we need
to use the bundled version of getopt since there isn't a system version.

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: don't define USE_ELF_TLS for windows
Dylan Baker [Tue, 22 May 2018 21:08:36 +0000 (14:08 -0700)]
meson: don't define USE_ELF_TLS for windows

Because the macros for exporting dll symbols and using TLS are mutually
exclusive.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: add switches for SWR with MSVC
Dylan Baker [Tue, 22 May 2018 17:54:12 +0000 (10:54 -0700)]
meson: add switches for SWR with MSVC

This makes two changes for SWR,

The first is that it reorders the arguments to try to put the ICL ones
first. This is required to support older versions of meson that don't
add enough "error in this case" switches to ICL, which causes it to
happy accept -mavx (for example) even though it doesn't support them,
resulting in compilation failures.

The second is to fix the names of the libraries, setting the soversion
to '' will result in <lib>.dll, instead of <lib>-0.dll. Since these are
not versioned dll's, but implement an internal API we should communicate
that. It's also what scons does.

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: disable sse4.1 optimizations with msvc
Dylan Baker [Thu, 17 May 2018 21:45:10 +0000 (14:45 -0700)]
meson: disable sse4.1 optimizations with msvc

There isn't an obvious command line switch here, /arch:AVX *might* be
the right thing, but meson doesn't know what to do here either and
leaves the -msse4.1 and -mstackrealign.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: force inclusion of inttypes.h for glcpp with msvc
Dylan Baker [Wed, 23 May 2018 17:05:20 +0000 (10:05 -0700)]
meson: force inclusion of inttypes.h for glcpp with msvc

Because we provide a copy if MSVC doesn't, and we need it to make flex
do what we want.

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: Add support for using win_flex and win_bison on windows
Dylan Baker [Thu, 17 May 2018 20:42:50 +0000 (13:42 -0700)]
meson: Add support for using win_flex and win_bison on windows

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: don't look for rt on windows
Dylan Baker [Thu, 17 May 2018 20:32:12 +0000 (13:32 -0700)]
meson: don't look for rt on windows

v6: - Minor refactor

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: fix pipe-loader compilation for windows
Dylan Baker [Thu, 17 May 2018 18:52:26 +0000 (11:52 -0700)]
meson: fix pipe-loader compilation for windows

v2: - Add missing D to pound define
    - Simply define the variable rather than set it to 1 (mirrors
      android.mk not scons)

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agoutil/xmlconfig: include strndup.h for windows
Dylan Baker [Thu, 17 May 2018 18:33:35 +0000 (11:33 -0700)]
util/xmlconfig: include strndup.h for windows

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: Don't check for posix_memalign on windows
Dylan Baker [Wed, 18 Apr 2018 20:54:12 +0000 (13:54 -0700)]
meson: Don't check for posix_memalign on windows

There's a mingw bug for this, it exports __builtin_posix_memalign but
not posix_memalign, so the check will succeed, but compiling will fail.

Reviewed-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: fix gallium-osmesa to build for windows
Dylan Baker [Wed, 18 Apr 2018 20:29:26 +0000 (13:29 -0700)]
meson: fix gallium-osmesa to build for windows

v2: - set so_version to '' (only affects windows)
    - always set lib prefix to 'lib', even on msvc
v5: - key NO_EXPORTS on shared glapi instead of gles.

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: build graw-gdi target
Dylan Baker [Wed, 18 Apr 2018 20:29:04 +0000 (13:29 -0700)]
meson: build graw-gdi target

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: build libgl-gdi target
Dylan Baker [Wed, 18 Apr 2018 20:28:38 +0000 (13:28 -0700)]
meson: build libgl-gdi target

v4: - Fix check for broken mingw (should be for x86 not x86_64)
    - Add comment about why check is needed

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: build wgl state tracker
Dylan Baker [Wed, 18 Apr 2018 20:28:04 +0000 (13:28 -0700)]
meson: build wgl state tracker

v4: - Handle enable gles properly
    - Add comments about what various #defines do
v5: - key NO_EXPORTS on shared glapi instead of gles.

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: build gallium gdi winsys
Dylan Baker [Wed, 18 Apr 2018 20:27:13 +0000 (13:27 -0700)]
meson: build gallium gdi winsys

v6: - use null_dep instead of []

Reviewed-by: Eric Anholt <eric@anholt.net> (v5)
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: Add necessary defines for mesa_gallium on windows
Dylan Baker [Wed, 18 Apr 2018 20:23:48 +0000 (13:23 -0700)]
meson: Add necessary defines for mesa_gallium on windows

v4: - Retain scons comments for windows specific defines
v5: - key GLAPI_NO_EXPORTS off of shared-glapi instead of gles

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agomeson: Add windows defines to glapi
Dylan Baker [Wed, 18 Apr 2018 20:22:35 +0000 (13:22 -0700)]
meson: Add windows defines to glapi

These are needed to control the export or symbols due to differences
between the way windows and *nix handle symbol exports.

Reviewed-by: Eric Anholt <eric@anholt.net> (v2)
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
v5: - key NO_EXPORT off of shared-glapi instead of gles

5 years agomeson: add windows compiler checks and libraries
Dylan Baker [Wed, 18 Apr 2018 20:19:54 +0000 (13:19 -0700)]
meson: add windows compiler checks and libraries

v4: - Fix typo in warning code (4246 -> 4267)
    - Copy comments from scons for what MSVC warnings codes do
    - Merge linker argument changes into this commit
v5: - Add /GR- on windows if LLVM is build without rtti (equivalent to
      GCc's -fno-rtti')
    - Add /wd4291, which is catching the same hting that
    -Wno-non-virtual-dtor is on GCC/Clang

Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agoutil: use _WIN32 instead of WIN32
Dylan Baker [Thu, 19 Sep 2019 18:04:21 +0000 (11:04 -0700)]
util: use _WIN32 instead of WIN32

MinGW defines only _WIN32, but doesn't have fcntl, so we need to use the
windows path.

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Acked-by: Kristian H. Kristensen <hoegsberg@google.com>
5 years agofreedreno/ir3: handle multi component alu src when propagating shifts
Rob Clark [Wed, 9 Oct 2019 16:47:13 +0000 (09:47 -0700)]
freedreno/ir3: handle multi component alu src when propagating shifts

Signed-off-by: Rob Clark <robdclark@chromium.org>
5 years agofreedreno/ir3: drop unused param
Rob Clark [Wed, 9 Oct 2019 16:15:45 +0000 (09:15 -0700)]
freedreno/ir3: drop unused param

Signed-off-by: Rob Clark <robdclark@chromium.org>
5 years agoclover: fix the nir_serialize build failure
Marek Olšák [Thu, 10 Oct 2019 22:43:47 +0000 (18:43 -0400)]
clover: fix the nir_serialize build failure

Fixes: dd4cc56ebd0 "nir: add a strip parameter to nir_serialize"
5 years agollvmpipe/draw: handle UBOs that are < 16 bytes.
Dave Airlie [Wed, 9 Oct 2019 05:11:14 +0000 (15:11 +1000)]
llvmpipe/draw: handle UBOs that are < 16 bytes.

Not sure if this is a bug in the user or not, but some CTS
tests fail due to using an 8 byte constant buffer.

Fixes: KHR-GLES31.core.layout_binding.block_layout_binding_block_VertexShader
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
5 years agollvmpipe/draw: fix image sizes for vertex/geometry shaders.
Dave Airlie [Wed, 9 Oct 2019 04:13:19 +0000 (14:13 +1000)]
llvmpipe/draw: fix image sizes for vertex/geometry shaders.

since images are a single level, minify before passing the w/h
to draw.

Fixes: KHR-GLES31.core.shader_image_size.basic-nonMS-vs-*
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
5 years agollvmpipe: make texture buffer offset alignment == 16
Dave Airlie [Sun, 1 Sep 2019 23:52:55 +0000 (09:52 +1000)]
llvmpipe: make texture buffer offset alignment == 16

Due to use vmovdqa instructions in the asm, which require 16-byte
aligned buffers.

This fixes a crash in
KHR-GLES31.core.texture_buffer.texture_buffer_texture_buffer_range

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
5 years agomeson: skip installation of GLVND-provided headers
Eric Engestrom [Mon, 30 Sep 2019 13:08:29 +0000 (14:08 +0100)]
meson: skip installation of GLVND-provided headers

Fixes: 93df862b6affb6b8507e ("meson: re-add incorrect pkg-config files with GLVND for backward compatibility")
Closes: https://gitlab.freedesktop.org/mesa/mesa/issues/1846
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
5 years agomeson: split Mesa headers as a separate installation
Eric Engestrom [Mon, 30 Sep 2019 13:06:07 +0000 (14:06 +0100)]
meson: split Mesa headers as a separate installation

Fixes: 93df862b6affb6b8507e ("meson: re-add incorrect pkg-config files with GLVND for backward compatibility")
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
5 years agomeson: split headers one per line
Eric Engestrom [Mon, 30 Sep 2019 13:00:51 +0000 (14:00 +0100)]
meson: split headers one per line

Fixes: 93df862b6affb6b8507e ("meson: re-add incorrect pkg-config files with GLVND for backward compatibility")
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
5 years agomeson: move a couple of include installs around
Eric Engestrom [Mon, 30 Sep 2019 12:40:22 +0000 (13:40 +0100)]
meson: move a couple of include installs around

Preparation for a later commit.

Fixes: 93df862b6affb6b8507e ("meson: re-add incorrect pkg-config files with GLVND for backward compatibility")
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
5 years agomeson: rename `glvnd_missing_pc_files` to `not glvnd_has_headers_and_pc_files`
Eric Engestrom [Mon, 30 Sep 2019 09:26:43 +0000 (10:26 +0100)]
meson: rename `glvnd_missing_pc_files` to `not glvnd_has_headers_and_pc_files`

This reflects better what is provided by glvnd or not.

Fixes: 93df862b6affb6b8507e ("meson: re-add incorrect pkg-config files with GLVND for backward compatibility")
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
5 years agoGL: drop symbols mangling support
Eric Engestrom [Thu, 10 Oct 2019 15:26:52 +0000 (16:26 +0100)]
GL: drop symbols mangling support

SCons and Meson have never supported that feature, and Autotools was
deleted over 6 months ago and no-one complained yet, so it's pretty
obvious nobody cares about it.

Fixes: 95aefc94a941701616fd ("Delete autotools")
Signed-off-by: Eric Engestrom <eric.engestrom@intel.com>
Acked-by: Dylan Baker <dylan@pnwbakers.com>
5 years agoaco: update print_ir
Rhys Perry [Thu, 10 Oct 2019 11:17:04 +0000 (12:17 +0100)]
aco: update print_ir

Mostly adds GFX10 stuff.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-By: Timur Kristóf <timur.kristof@gmail.com>
5 years agoaco: rework scratch resource code
Rhys Perry [Thu, 10 Oct 2019 11:15:49 +0000 (12:15 +0100)]
aco: rework scratch resource code

Fix compute, cleanup and add GFX10 support.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-By: Timur Kristóf <timur.kristof@gmail.com>
5 years agoaco/gfx10: disable GFX9 1D texture workarounds
Rhys Perry [Thu, 10 Oct 2019 11:13:28 +0000 (12:13 +0100)]
aco/gfx10: disable GFX9 1D texture workarounds

Navi added back support for 1D textures.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-By: Timur Kristóf <timur.kristof@gmail.com>
5 years agoaco/gfx10: fix inline uniform blocks
Rhys Perry [Thu, 10 Oct 2019 11:12:29 +0000 (12:12 +0100)]
aco/gfx10: fix inline uniform blocks

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-By: Timur Kristóf <timur.kristof@gmail.com>
5 years agoradv/aco: disable NGG when ACO is used
Rhys Perry [Thu, 10 Oct 2019 11:10:15 +0000 (12:10 +0100)]
radv/aco: disable NGG when ACO is used

Note that radv_device.c still has to be modified to use ACO with Navi.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-By: Timur Kristóf <timur.kristof@gmail.com>
5 years agoac/nir: add back nir_op_fmod
Marek Olšák [Tue, 8 Oct 2019 04:08:19 +0000 (00:08 -0400)]
ac/nir: add back nir_op_fmod

radeonsi doesn't lower it for doubles.

This partially reverts commit d861401554b52b2c2fc6721c69bdfe1697ee608f.

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
5 years agogallium: remove PIPE_SHADER_CAP_SCALAR_ISA
Marek Olšák [Tue, 8 Oct 2019 02:50:36 +0000 (22:50 -0400)]
gallium: remove PIPE_SHADER_CAP_SCALAR_ISA

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
5 years agotgsi_to_nir: use nir_shader_compiler_options::lower_to_scalar
Marek Olšák [Tue, 8 Oct 2019 02:48:41 +0000 (22:48 -0400)]
tgsi_to_nir: use nir_shader_compiler_options::lower_to_scalar

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
5 years agost/mesa: use nir_shader_compiler_options::lower_to_scalar
Marek Olšák [Tue, 8 Oct 2019 02:47:23 +0000 (22:47 -0400)]
st/mesa: use nir_shader_compiler_options::lower_to_scalar

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
5 years agonir: add nir_shader_compiler_options::lower_to_scalar
Marek Olšák [Tue, 8 Oct 2019 02:46:00 +0000 (22:46 -0400)]
nir: add nir_shader_compiler_options::lower_to_scalar

This will replace PIPE_SHADER_CAP_SCALAR_ISA.

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
5 years agotgsi_to_nir: add #ifdef header guards
Marek Olšák [Thu, 26 Sep 2019 23:34:31 +0000 (19:34 -0400)]
tgsi_to_nir: add #ifdef header guards

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
5 years agonir/drawpixels: fix what appears to be a copy-paste bug in get_texcoord_const
Marek Olšák [Tue, 1 Oct 2019 02:16:55 +0000 (22:16 -0400)]
nir/drawpixels: fix what appears to be a copy-paste bug in get_texcoord_const

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
5 years agonir/drawpixels: handle load_color0, load_input, load_interpolated_input
Marek Olšák [Tue, 1 Oct 2019 01:29:11 +0000 (21:29 -0400)]
nir/drawpixels: handle load_color0, load_input, load_interpolated_input

for radeonsi

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
5 years agonir: move gl_nir_opt_access from glsl directory
Marek Olšák [Fri, 27 Sep 2019 20:43:31 +0000 (16:43 -0400)]
nir: move gl_nir_opt_access from glsl directory

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
5 years agonir: add a strip parameter to nir_serialize
Marek Olšák [Wed, 9 Oct 2019 17:27:07 +0000 (13:27 -0400)]
nir: add a strip parameter to nir_serialize

so that drivers don't have to call nir_strip manually.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: Rob Clark <robdclark@gmail.com>
5 years agoradv: Enable VK_ANDROID_external_memory_android_hardware_buffer.
Bas Nieuwenhuizen [Tue, 24 Sep 2019 17:44:01 +0000 (19:44 +0200)]
radv: Enable VK_ANDROID_external_memory_android_hardware_buffer.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Check the size of the imported buffer.
Bas Nieuwenhuizen [Tue, 8 Oct 2019 21:24:02 +0000 (23:24 +0200)]
radv: Check the size of the imported buffer.

This is a security feature to disallow malicious apps from passing
a buffer that is too small.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Expose image handle compat types for Android handles.
Bas Nieuwenhuizen [Tue, 8 Oct 2019 20:15:43 +0000 (22:15 +0200)]
radv: Expose image handle compat types for Android handles.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Allow Android image binding.
Bas Nieuwenhuizen [Tue, 24 Sep 2019 17:42:49 +0000 (19:42 +0200)]
radv: Allow Android image binding.

Using delayed layout of images.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv/android: Add android hardware buffer import/export.
Bas Nieuwenhuizen [Wed, 10 Jul 2019 12:45:01 +0000 (14:45 +0200)]
radv/android: Add android hardware buffer import/export.

Support does not include images yet.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Deal with Android external formats.
Bas Nieuwenhuizen [Tue, 24 Sep 2019 14:33:39 +0000 (16:33 +0200)]
radv: Deal with Android external formats.

To abstract things a bit, this adds a helper function in radv_android.c.
However, this means we have to link in radv_android.c on non-android as
well, which means some scaffolding changes.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Derive android usage from create flags.
Bas Nieuwenhuizen [Mon, 8 Jul 2019 13:33:06 +0000 (15:33 +0200)]
radv: Derive android usage from create flags.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Disallow sparse shared images.
Bas Nieuwenhuizen [Tue, 8 Oct 2019 20:14:44 +0000 (22:14 +0200)]
radv: Disallow sparse shared images.

Since we really cannot share them ever.

Also remove an unused switch.

Fixes: b70829708ac "radv: Implement VK_KHR_external_memory"
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv/android: Add android hardware buffer queries.
Bas Nieuwenhuizen [Tue, 25 Jun 2019 09:42:03 +0000 (11:42 +0200)]
radv/android: Add android hardware buffer queries.

Derived from the Intel code.

For the internal format we just use the internal Vulkan format,
as we have Vulkan formats for all android formats we care about.

For the ycbcr properties we just do something. I do not have a real
clue what would be recommended.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv/android: Add android hardware buffer field to device memory.
Bas Nieuwenhuizen [Wed, 10 Jul 2019 12:27:37 +0000 (14:27 +0200)]
radv/android: Add android hardware buffer field to device memory.

You cannot go from BO to Android hardware buffer, so for export we
have to remember it.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Add VK_ANDROID_external_memory_android_hardware_buffer.
Bas Nieuwenhuizen [Tue, 24 Sep 2019 11:56:05 +0000 (13:56 +0200)]
radv: Add VK_ANDROID_external_memory_android_hardware_buffer.

Still disabled but now we can add entrypoints.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Unset vk_info in radv_image_create_layout.
Bas Nieuwenhuizen [Tue, 24 Sep 2019 11:42:31 +0000 (13:42 +0200)]
radv: Unset vk_info in radv_image_create_layout.

For better test coverage of this corner case.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Handle slightly different image dimensions.
Bas Nieuwenhuizen [Tue, 24 Sep 2019 11:23:36 +0000 (13:23 +0200)]
radv: Handle slightly different image dimensions.

The minigbm comment really says it all. We should
fix minigbm as well, but for now this is the more
robust solution.

Note that this only changes width and height for
the surface creation, not for the image and hence
also not for the sampler, where it would wreak
havoc due to the normalized coords.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Delay patching for imported images until layout time.
Bas Nieuwenhuizen [Mon, 23 Sep 2019 14:42:39 +0000 (16:42 +0200)]
radv: Delay patching for imported images until layout time.

We want this flexibility because in GFX10 we lose any stride fields,
so we have to make sure our width/height are in alignment with
the external image we import.

Furthermore, we need the ability to inject tiling modifiers on import
time which is strictly after create time for Android. So, with the
layout & patch functions being fully independent of pCreateInfo, we
can delay it until import/bind time.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Split out layout code from image creation.
Bas Nieuwenhuizen [Mon, 23 Sep 2019 14:14:50 +0000 (16:14 +0200)]
radv: Split out layout code from image creation.

So we can delay the layout until later in some import cases.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Handle device memory alloc failure with normal free.
Bas Nieuwenhuizen [Wed, 10 Jul 2019 12:21:11 +0000 (14:21 +0200)]
radv: Handle device memory alloc failure with normal free.

Less duplication/complexity.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agoradv: Cleanup buffer_from_fd.
Bas Nieuwenhuizen [Tue, 24 Sep 2019 15:26:41 +0000 (17:26 +0200)]
radv: Cleanup buffer_from_fd.

Unused stride/offset args.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
5 years agogitlab-ci/lava: Test Lima driver with dEQP
Tomeu Vizoso [Sun, 6 Oct 2019 15:49:56 +0000 (08:49 -0700)]
gitlab-ci/lava: Test Lima driver with dEQP

Run dEQP on boards with Mali 400 and 450 in Baylibre's lab.

There's lots of skipped tests because of crashes and undetermined
behavior. May be a good idea to run the tests with valgrind and fix any
issues found.

Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
5 years agogitlab-ci/lava: Use files to list tests to skip
Tomeu Vizoso [Sun, 6 Oct 2019 22:21:39 +0000 (15:21 -0700)]
gitlab-ci/lava: Use files to list tests to skip

As the non-LAVA runner script does, have per-GPU version files listing
the tests that are to be skipped, due to being very slow, unstable, etc.

Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Reviewed-by: Neil Armstrong <narmstrong@baylibre.com>
5 years agointel/tools: Support multiple contexts in intel_dump_gpu.
Rafael Antognolli [Mon, 16 Sep 2019 16:07:26 +0000 (09:07 -0700)]
intel/tools: Support multiple contexts in intel_dump_gpu.

Create basic aub_context on GEM_CONTEXT_CREATE.

Set it up and submit a context + ring + pphwsp during execbuf
submission, if it has not been initialized yet.

v2: Write the HWSP only once per engine (Lionel).

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
5 years agointel/tools: Add basic aub_context code and helpers.
Rafael Antognolli [Fri, 13 Sep 2019 22:13:31 +0000 (15:13 -0700)]
intel/tools: Add basic aub_context code and helpers.

v2:
 - Only dump context if there were no erros (Lionel).
 - Store counter for context handles in aub_file (Lionel).
v3:
 - Add a comment about aub_context -> GEM context (Lionel).

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
5 years agointel/tools: Use common code for GGTT address allocation.
Rafael Antognolli [Tue, 17 Sep 2019 01:12:02 +0000 (18:12 -0700)]
intel/tools: Use common code for GGTT address allocation.

We want to be able to create contexts on demand, and increase the GGTT
as needed for that. Use the aub_map_ggtt() function for that.

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>