Timothy Arceri [Fri, 22 Sep 2017 06:04:17 +0000 (16:04 +1000)]
glsl: mark xfb inputs as always_active_io
We won't split varyings marked as always active because there
is no point in doing so. This means we need to mark both
sides of the interface as always active otherwise we will have
a mismatch and start removing things we shouldn't.
Reviewed-by: Eric Anholt <eric@anholt.net>
Timothy Arceri [Thu, 5 Oct 2017 22:19:35 +0000 (09:19 +1100)]
nir: add variant of lower_io_to_scalar to be called earlier
This is intended to be called before nir_lower_io() so that we
can do some linking optimisations with the results. It can also
be used with drivers that don't use nir_lower_io() at all such
as RADV.
v2: pass mode mask rather than first and last stage integer.
Reviewed-by: Eric Anholt <eric@anholt.net>
Timothy Arceri [Tue, 19 Sep 2017 23:49:57 +0000 (09:49 +1000)]
nir: add glsl_channel_type() helper
Reviewed-by: Eric Anholt <eric@anholt.net>
Timothy Arceri [Sun, 4 Dec 2016 06:21:16 +0000 (17:21 +1100)]
nir: add glsl_type_is_64bit() to nir_types
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Ilia Mirkin [Fri, 25 Aug 2017 04:47:05 +0000 (00:47 -0400)]
a2xx: add support for a few 16-bit color rendering formats
The rest should be possible too, just needs some additional
investigation. Passes fbo-*-formats piglit tests.
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Wladimir J. van der Laan <laanwj@gmail.com>
Wladimir J. van der Laan [Fri, 25 Aug 2017 14:34:08 +0000 (14:34 +0000)]
freedreno/a20x: Enable rendering to RGBA/RGBX
Signed-off-by: Wladimir J. van der Laan <laanwj@gmail.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Wladimir J. van der Laan [Fri, 25 Aug 2017 14:29:53 +0000 (14:29 +0000)]
freedreno/a20x: Fix rendering to BGRX
Make sure that BGRX rendering is swapped the correct way around.
Signed-off-by: Wladimir J. van der Laan <laanwj@gmail.com>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Brian Paul [Fri, 13 Oct 2017 22:16:48 +0000 (16:16 -0600)]
mesa: minor simplification in test_attachment_completeness()
We already have a pointer to the texture object. Use it here.
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Lucas Stach [Thu, 12 Oct 2017 14:07:48 +0000 (16:07 +0200)]
etnaviv: rework TS enable to be a derived state
Draw operations should not use the TS if the TS buffer content is invalid,
as this leads to wrong rendering or even GPU hangs. As the TS valid status
can change between draws (clear operations changing it to valid, blits using
the RS to the color or ZS buffer changing it to invalid), the TS_MEM_CONFIG
must be updated before each draw if the status has changed.
This fixes the remaining TS related piglit failures (regressions of a
standard run against a piglit run with TS completely disabled).
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Wladimir J. van der Laan <laanwj@gmail.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Lucas Stach [Thu, 12 Oct 2017 14:07:47 +0000 (16:07 +0200)]
etnaviv: skip unused vertex attributes when assigning VS inputs
When not all of the vertex attributes are actually used in the shader,
we end up with some inputs without an assigned reg. Those are marked
as invalid and must be skipped when assigning the inputs, as those would
overwrite other valid inputs otherwise.
Fixes piglit drawpixels and a bunch of other tests using the st_draw path.
Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Reviewed-by: Wladimir J. van der Laan <laanwj@gmail.com>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Samuel Pitoiset [Thu, 12 Oct 2017 20:55:32 +0000 (22:55 +0200)]
radv: do not allocate CMASK for non-MSSA images with 128 bit formats
This saves some useless CMASK initializations/eliminations in
the Vulkan SSAO demo.
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Samuel Pitoiset [Fri, 13 Oct 2017 16:01:56 +0000 (18:01 +0200)]
radv: set correct INDEX_TYPE for indexed indirect draws on GFX9
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Samuel Pitoiset [Fri, 13 Oct 2017 16:20:21 +0000 (18:20 +0200)]
radv: add the draw count buffer to the list of buffers
My guess is that the GPU is going to report VM faults if
vkCmdDrawIndirectCountAMD() (and friends) are used.
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Jason Ekstrand [Fri, 13 Oct 2017 17:40:45 +0000 (10:40 -0700)]
blob: Use intptr_t instead of ssize_t
ssize_t is a GNU extension and is not available on Windows or MacOS.
Instead, we use intptr_t which should be effectively equivalent and is
part of the C standard. This should fix the Windows and Mac OS builds.
Fixes: 3af1c829891a4530682bce113fdd512d4f2de3c6
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103253
Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Tested-by: Vinson Lee <vlee@freedesktop.org>
Kenneth Graunke [Fri, 13 Oct 2017 07:35:30 +0000 (00:35 -0700)]
i965: Rename brw->no_batch_wrap to intel_batchbuffer::no_wrap
This really makes more sense in the intel_batchbuffer struct.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Kenneth Graunke [Fri, 13 Oct 2017 07:31:27 +0000 (00:31 -0700)]
i965: Delete dead brw_context fields.
fast_clear_op is leftover from the meta-fast-clear days.
No idea what the other thing was for, but it isn't used now.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Emil Velikov [Fri, 13 Oct 2017 15:35:47 +0000 (16:35 +0100)]
mapi/shared-glapi/test: rework glapitable.h handling
Currently all the build systems but Meson generate the header in
src/mapi/glapi. Meson cannot do that since:
- it does not allow user control over the location of output files
- moving the generation rule(s) causes explosion due to the unusual
structure of glapi and friends
- copying the file into the correct location is a non-trivial task
To workaround the above deficiency in the least invasive way, let's
adjust the #include directive and add a few -I flags to the autotools
build.
Note: both builddir and srcdir, should be used. Otherwise building from
a release tarball fails badly.
Cc: Dylan Baker <dylanx.c.baker@intel.com>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Tested-by: Mark Janes <mark.a.janes@intel.com>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Dylan Baker [Fri, 13 Oct 2017 17:34:07 +0000 (10:34 -0700)]
meson: fix blob test includes
Since blob.h moved up to src/compiler the test should include that
instead of src/compiler/glsl
fixes:
0e3bd56c6ea783dbc ("compiler: Move blob up a level")
Signed-off-by: Dylan Baker <dylanx.c.baker@intel.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Emil Velikov [Fri, 13 Oct 2017 13:54:53 +0000 (14:54 +0100)]
Revert "make: Fix test to be meson compatible"
This reverts commit
fc48ad24272799b154de37f6a944eb391bdbbb82.
There commit reference the previous commit as it justification of
changing behaviour. Although unlike the said commit, there's nothing
obviously wrong there.
I'll take a look close why Meson fails to pick the file, but in the
interim reverting this commit fixes the normal distcheck target.
Mark Thompson [Sun, 1 Oct 2017 17:39:05 +0000 (18:39 +0100)]
st/dri: Add definitions to allow importing 16-bit surfaces
Necessary to support P010/P016 surfaces for video.
Signed-off-by: Mark Thompson <sw@jkqxz.net>
Acked-by: Leo Liu <leo.liu@amd.com>
Mario Kleiner [Fri, 6 Oct 2017 16:11:31 +0000 (18:11 +0200)]
i965: Complete 'expose RGBA visuals only on Android'
Commit
731ba6924a2ed6cdf47a78fd637a91a19ddcf9ed
"expose RGBA visuals only on Android" replaced
ARRAY_SIZE(formats) by num_formats, but there are
3 loops which add configs, and only one was updated
to num_formats.
Also update loops for configs with accumulation buffer
and multisample configs.
Fixes: 731ba6924a2 "i965: expose RGBA visuals only on Android"
Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Emil Velikov <emil.velikov@collabora.com>
Emil Velikov [Thu, 5 Oct 2017 10:19:08 +0000 (11:19 +0100)]
configure.ac: add missing LLVM components for OpenCL
Coverage and LTO seems to be hard requirements for Clang, while
coroutines is needed as of LLVM/Clang 4.0.
Mark the last one as "optional" so we handle every case.
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Tobias Droste <tdroste@gmx.de>
Emil Velikov [Thu, 5 Oct 2017 10:19:07 +0000 (11:19 +0100)]
configure.ac: add llvm_add_optional_component helper
We want to add "optional" components, which have been added with later
LLVM versions.
One such in-tree example is inteljitevents. Others are to follow
shortly.
v2: Use the correct function, add blank line between functions (Tobias)
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Tobias Droste <tdroste@gmx.de>
Emil Velikov [Thu, 5 Oct 2017 10:19:06 +0000 (11:19 +0100)]
Travis: add binutils 2.26 for a few more LLVM 3.9 builds
Otherwise we error out at link stage as follows:
/usr/lib/llvm-3.9/lib/libLLVMAMDGPUCodeGen.a(R600OptimizeVectorRegisters.cpp.o):
unrecognized relocation (0x2a) in section
`.text._ZNK12_GLOBAL__N_119R600VectorRegMerger16getAnalysisUsageERN4llvm13AnalysisUsageE'
/usr/bin/ld: final link failed: Bad value
Cc: mesa-stable@lists.freedesktop.org
Cc: Jan Vesely <jan.vesely@rutgers.edu
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Emil Velikov [Thu, 5 Oct 2017 10:19:05 +0000 (11:19 +0100)]
configure.ac: rework llvm libs handling for 3.9+
Earlier versions need different quirks, but as of LLVM 3.9 llvm-config
provides --link-shared/link-static toggles.
The output of which seems to be reliable - looking at LLVM 3.9, 4.0 and
5.0.
Note that there are earlier code will be used for pre LLVM 3.9 and is
unchanged.
This effectively fixes LLVM static linking, while providing a clearer
and more robust solution for future versions.
Mildly interesting side notes:
- build-mode (introduced with 3.8) was buggy with 3.8
It shows "static" when build with -DLLVM_LINK_LLVM_DYLIB=ON, yet it was
consistent with --libs. The latter shows the static libraries.
- libnames and libfiles are broken with LVM 3.9
The library prefix and extension is printed twice liblibLLVM-3.9.so.so
v2: Invoke llvm-config twice, instead of using sed, to combine the two
lines into one (Tobias)
Cc: mesa-stable@lists.freedesktop.org
Cc: Dieter Nützel <Dieter@nuetzel-hh.de>
Cc: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Tobias Droste <tdroste@gmx.de>
Emil Velikov [Thu, 5 Oct 2017 10:19:04 +0000 (11:19 +0100)]
configure.ac: factor out detection for old and buggy llvm
As of LLVM 3.9 one could use consistent ways to handle the component.
Factor out the current handling, as it will be used for older versions.
Cc: mesa-stable@lists.freedesktop.org
Cc: Michel Dänzer <michel.daenzer@amd.com>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Tobias Droste <tdroste@gmx.de>
Emil Velikov [Thu, 5 Oct 2017 10:19:03 +0000 (11:19 +0100)]
configure.ac: remove no longer necessary llvm-config --libs check
Prior to the refactor/cleanup by Tobias one could add an invalid
component to LLVM_COMPONENTS.
Since that's no longer the case we can drop the current check.
Cc: Tobias Droste <tdroste@gmx.de>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Tobias Droste <tdroste@gmx.de>
Emil Velikov [Thu, 12 Oct 2017 10:39:46 +0000 (11:39 +0100)]
eglmesaext: add forward declaration for struct wl_buffers
The user does not need to know the specifics of the struct, as only a
pointer to it is used.
Just forward declare the struct making the header self-contained.
v2: Remove deprecation warning text/bugzilla - patch does no help there.
Cc: Greg V <greg@unrelenting.technology>
Fixes: 5cddb1ce3c9 ("wayland: Add an extension to create wl_buffers from
EGLImages")
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com> (v1)
Emil Velikov [Tue, 3 Oct 2017 16:51:16 +0000 (17:51 +0100)]
configure.ac: bump Clover LLVM requirement to 3.9
The only driver that utilises Clover already depends on LLVM 3.9.
Close to every supported distribution has said version.
Additionally libclc also requires LLVM 3.9.
With this in mind, we can safely bump the requirement.
There is a handful of dead code that we could remove, which will be
resolved with later commits.
Note: this drops the LLVM 3.6 build from the Travis build. LLVM 3.9 (and
later) are already covered in there.
https://lists.freedesktop.org/archives/mesa-dev/2017-September/170028.html
v2: Add reference to discussion thread (Eric), adjust libclc LLVM req.
(Jan).
Cc: Aaron Watry <awatry@gmail.com>
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Eric Engestrom <eric.engestrom@imgtec.com>
Acked-by: Vedran Miletić <vedran@miletic.net>
Acked-by: Jan Vesely <jan.vesely@rutgers.edu>
Acked-by: Francisco Jerez <currojerez@riseup.net>
Emil Velikov [Wed, 27 Sep 2017 18:49:12 +0000 (19:49 +0100)]
wayland-drm: constify the callbacks struct, take 2
Now that wayland-drm (correctly) keeps a local copy of the callbacks,
this should not longer cause explosions.
After all the symbol is a local, constant data.
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Tested-by: Derek Foreman <derekf@osg.samsung.com>
Emil Velikov [Wed, 27 Sep 2017 18:49:11 +0000 (19:49 +0100)]
wayland-drm: use a copy of the wayland_drm_callbacks struct
The callbacks may be called even when they are no longer valid.
Say, the user is dlclose(ing) libEGL while the buffers are being
destroyed.
Cc: mesa-stable@lists.freedesktop.org
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Tested-by: Derek Foreman <derekf@osg.samsung.com>
Emil Velikov [Tue, 10 Oct 2017 17:50:13 +0000 (18:50 +0100)]
egl/dri: don't crash when createImageFromRenderbuffer2 is NULL
The __DRI_IMAGE version can be 17 or over, while the function pointer is
NULL. Guard for that instead of crashing.
Fixes: bad24395d91 ("egl/dri: use createImageFromRenderbuffer2 when
available")
Signed-off-by: Emil Velikov <emil.velikov@collabora.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Ville Syrjälä [Mon, 9 Oct 2017 22:15:42 +0000 (01:15 +0300)]
meson: Build i915
Build i915 with meson. More or less copied from i965, with all
the unneeded cruft removed, and the libdrm_intel dependency added.
Cc: Dylan Baker <dylan@pnwbakers.com>
Cc: Eric Anholt <eric@anholt.net>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Ville Syrjälä [Mon, 9 Oct 2017 22:34:18 +0000 (01:34 +0300)]
meson: Fix xf86vm dep
The pkg-config file is called xxf86vm.pc not xf86vm.pc.
Cc: Dylan Baker <dylan@pnwbakers.com>
Cc: Eric Anholt <eric@anholt.net>
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Jason Ekstrand [Fri, 29 Sep 2017 19:22:48 +0000 (12:22 -0700)]
intel/cs: Make thread_local_id a regular builtin param
This is a lot more natural than special casing it all over the place.
We still have to do a bit of special-casing in assign_constant_locations
but it's not special-cased quite as bad as it was before.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 18:30:25 +0000 (11:30 -0700)]
intel/compiler: Allocate pull_param in assign_constant_locations
Now that everything is nicely ralloc'd, we can allocate the pull_param
array in assign_constant_locations instead of higher up. We can also
re-allocate the param array so that it's exactly the needed size. This
should save us some memory because we're not allocating the total needed
param space for both push and pull.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 04:45:41 +0000 (21:45 -0700)]
intel: Allocate prog_data::[pull_]param deeper inside the compiler
Now that we're always growing the param array as-needed, we can
allocate the param array in common code and stop repeating the
allocation everywere. In order to keep things sane, we ralloc the
[pull_]param array off of the compile context and then steal it back
to a NULL context later. This doesn't get us all the way to where
prog_data::[pull_]param is purely an out parameter of the back-end
compiler but it gets us a lot closer.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Wed, 16 Aug 2017 23:50:46 +0000 (16:50 -0700)]
ralloc: Allow reparenting to a NULL context
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Ian Romanick <idr@freedesktop.org>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 18:18:04 +0000 (11:18 -0700)]
anv/pipeline: Refactor setup of the prog_data::param array
Now that the only thing we put in the array up-front are client push
constants, we can simplify anv_pipeline_compile a bit.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 18:10:22 +0000 (11:10 -0700)]
anv/pipeline: Grow the param array for images
Before, we were calculating up-front and then filling in later. Now we
just grow as needed in anv_nir_apply_pipeline_layout.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 18:09:04 +0000 (11:09 -0700)]
anv/pipeline: Whack nir->num_uniforms to MAX_PUSH_CONSTANT_SIZE
This way any image uniforms end up having locations higher than
MAX_PUSH_CONSTANT_SIZE. There's no bug here at the moment, but this
consistency will make the next commit easier. Also, because
nir_apply_pipeline_layout properly increments nir->num_uniforms when
it expands the param array, we no longer need to stomp it to match
prog_data::nr_params because it already does.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 17:50:07 +0000 (10:50 -0700)]
intel/vs: Grow the param array for clip planes
Instead of requiring the caller of brw_compile_vs to figure it out, just
grow the param array on-demand.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 17:37:40 +0000 (10:37 -0700)]
intel/cs: Grow prog_data::param on-demand for thread_local_id_index
Instead of making the caller of brw_compile_cs add something to the
param array for thread_local_id_index, just add it on-demand in
brw_nir_intrinsics and grow the array. This is now safe to do because
everyone is now using ralloc for prog_data::param.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 6 Oct 2017 17:08:11 +0000 (10:08 -0700)]
intel/compiler: Make brw_nir_lower_intrinsics compute-specific
It's already only ever called from brw_compile_cs and only handles
compute intrinsics. Let's just make it CS-specific. We can always
make it handle other stages again later if we want.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 18:05:55 +0000 (11:05 -0700)]
intel/compiler: Add a helper for growing the prog_data::param array
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 17:12:12 +0000 (10:12 -0700)]
intel/compiler: Stop adding params for texture sizes
We haven't needed this ever since we started using NIR for lowering
rectangle textures.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 17:16:43 +0000 (10:16 -0700)]
i965: Only add the wpos state reference if we lowered something
Otherwise, in the ARB program case _mesa_add_state_reference may grow
the parameter array which will cause brw_nir_setup_arb_uniforms to write
past the end of the param array because it only looks at the parameter
list length but the parma array is allocated based on nir->num_uniforms.
The only reason this hasn't caused us problems is because we are padding
out the param array for fragment programs unnecessarily.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 05:16:55 +0000 (22:16 -0700)]
intel/compiler: Add a flag for pull constant support
The Vulkan driver does not support pull constants. It simply limits
things such that we can always push everything. Previously, we were
determining whether or not to push things based on whether or not the
prog_data::pull_param array is non-null. This is rather hackish and
about to stop working.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 17:06:17 +0000 (10:06 -0700)]
anv/pipeline: Ralloc prog_data::param of the compile mem_ctx
This way we stop leaking it. This is completely safe because, when we
hand it off to anv_shader_bin_create or anv_pipeline_cache_upload_kernel,
they make a copy of the entire param array.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 04:51:48 +0000 (21:51 -0700)]
anv/pipeline: Add a mem_ctx parameter to anv_pipeline_compile
This lets us avoid some of the manual ralloc stealing and prepares for
future commits in which we will want to ralloc prog_data::param.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 02:04:31 +0000 (19:04 -0700)]
i965: Store image_param in brw_context instead of prog_data
This burns an extra 10k of memory or so in the case where you don't have
any images. However, if you have several shaders which use images, this
should be much less memory. It also gets rid of a part of prog_data
that really has nothing to do with the compiler.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 29 Sep 2017 02:03:38 +0000 (19:03 -0700)]
i965: Use prog->info.num_images for needs_dc computation
This should be just as good as looking in prog_data but removes our one
state setup dependency on brw_stage_prog_data::nr_image_param.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Thu, 28 Sep 2017 23:25:31 +0000 (16:25 -0700)]
intel: Rewrite the world of push/pull params
This moves us away to the array of pointers model and onto a model where
each param is represented by a generic uint32_t handle. We reserve 2^16
of these handles for builtins that get generated by somewhere inside the
compiler and have well-defined meanings. Generic params have handles
whose meanings are defined by the driver.
The primary downside to this new approach is that it moves a little bit
of the work that we would normally do at compile time to draw time. On
my laptop this hurts OglBatch6 by no more than 1% and doesn't seem to
have any measurable affect on OglBatch7. So, while this may come back
to bite us, it doesn't look too bad.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Thu, 28 Sep 2017 22:27:46 +0000 (15:27 -0700)]
i965: Get rid of gen7_cs_state.c
The only thing it was handling was push constants. We pull the actual
constant upload code into gen6_constant_state.c and the atoms into
genX_state_upload.c.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Thu, 28 Sep 2017 22:05:51 +0000 (15:05 -0700)]
i965: Add a helper for populating constant buffers
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Thu, 28 Sep 2017 21:39:00 +0000 (14:39 -0700)]
i965: Move brw_upload_pull_constants to gen6_constant_state.c
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Tue, 29 Aug 2017 00:06:24 +0000 (17:06 -0700)]
nir: Get rid of the variable on vote intrinsics
This looks like a copy+paste error. They don't actually write into that
variable as would be implied by putting the return there.
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: mesa-stable@lists.freedesktop.org
Jason Ekstrand [Mon, 28 Aug 2017 22:05:11 +0000 (15:05 -0700)]
nir/opcodes: Fix constant-folding of ufind_msb
We didn't fold correctly in the case of 0x1 because we never let the
loop counter hit 0. Switching it to bit >= 0 solves this problem.
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Cc: mesa-stable@lists.freedesktop.org
Jason Ekstrand [Fri, 13 Oct 2017 05:36:48 +0000 (22:36 -0700)]
meta: Delete the PBO texsubimage path for real
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Wed, 11 Oct 2017 19:13:35 +0000 (12:13 -0700)]
anv/pipeline_cache: Rework to use multialloc and blob
This gets rid of all of our hand-rolled size calculation and
serialization code and replaces it with safe "standards" that are used
elsewhere in anv and mesa. This should be significantly safer than
rolling our own.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Wed, 11 Oct 2017 19:28:20 +0000 (12:28 -0700)]
anv/pipeline: Declare bind maps closer to their use
This is just a trivial cleanup.
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Wed, 11 Oct 2017 19:13:19 +0000 (12:13 -0700)]
anv/multialloc: Add new add_size helper
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Wed, 11 Oct 2017 19:10:08 +0000 (12:10 -0700)]
compiler/blob: Make some parameters void instead of uint8_t
There are certain advantages to using uint8_t internally such as
well-defined arithmetic on all platforms. However, interfaces that
work in terms of raw data should use a void* type.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Wed, 11 Oct 2017 19:09:02 +0000 (12:09 -0700)]
compiler/blob: Constify the reader
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Fri, 13 Oct 2017 03:58:43 +0000 (20:58 -0700)]
compiler/blob: Add (reserve|overwrite)_(uint32|intptr) helpers
These helpers not only call blob_reserve_bytes but also make sure that
the blob is properly aligned as if blob_write_* were called.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Connor Abbott [Fri, 15 Sep 2017 04:29:46 +0000 (00:29 -0400)]
compiler/blob: make blob_reserve_bytes() more useful
Despite the name, it could only be used if you immediately wrote to the
pointer. Noboby was using it outside of one test, so clearly this
behavior wasn't that useful. Instead, make it return an offset into the
data buffer so that the result isn't invalidated if you later write to
the blob. In conjunction with blob_overwrite_bytes(), this will be
useful for leaving a placeholder and then filling it in later, which
we'll need to do for handling phi nodes when serializing NIR.
v2 (Jason Ekstrand):
- Detect overflow in the offset + to_write computation
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Fri, 13 Oct 2017 04:02:48 +0000 (21:02 -0700)]
compiler/blob: Allow for fixed-size blobs with a NULL data pointer
These can be used to easily count up the number of bytes that will be
required by "writing" it into the NULL blob.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Wed, 11 Oct 2017 16:52:07 +0000 (09:52 -0700)]
compiler/blob: Add a concept of a fixed-allocation blob
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Wed, 11 Oct 2017 16:44:33 +0000 (09:44 -0700)]
compiler/blob: Switch to init/finish instead of create/destroy
There's no reason why that tiny bit of memory needs to be on the heap.
We always put blob_reader on the stack, so why not do the same with the
writable blob.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Wed, 11 Oct 2017 16:54:55 +0000 (09:54 -0700)]
compiler: Move blob up a level
We're going to want to use the blob for Vulkan pipeline caching so it
makes sense to have it in libcompiler not libglsl.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Jason Ekstrand [Fri, 13 Oct 2017 04:19:32 +0000 (21:19 -0700)]
meson: Add inc_compiler to the libglsl includes
Jason Ekstrand [Wed, 11 Oct 2017 20:32:45 +0000 (13:32 -0700)]
glsl/blob: Return false from grow_to_fit if we've ever failed
Otherwise we could have a failure followed by a smaller write that
succeeds and get a corrupted blob. If we ever OOM, we should stop.
v2 (Jason Ekstrand):
- Initialize the new boolean member in create_blob
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Cc: mesa-stable@lists.freedesktop.org
Jason Ekstrand [Wed, 11 Oct 2017 17:56:48 +0000 (10:56 -0700)]
glsl/blob: Return false from ensure_can_read on overrun
Otherwise, if you have a large read fail and then try to do a small
read, the small read may succeed even though it's at the wrong offset.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Cc: mesa-stable@lists.freedesktop.org
Chris Wilson [Wed, 11 Oct 2017 20:43:45 +0000 (21:43 +0100)]
i965: Share the flush for brw_blorp_miptree_download into a pbo
As all users of brw_blorp_miptree_download() must emit a full pipeline
and cache flush when targetting a user PBO (as that PBO may then be
subsequently bound or *be* bound anywhere and outside of the driver
dirty tracking) move that flush into brw_blorp_miptree_download()
itself.
v2 (Ken): Rebase without userptr stuff so it can land sooner.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Fri, 2 Jun 2017 03:38:19 +0000 (20:38 -0700)]
meta: Delete the PBO texture upload/download path
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Tue, 6 Jun 2017 16:59:01 +0000 (09:59 -0700)]
i965: Use blorp instead of meta for PBO pixel reads
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Tue, 6 Jun 2017 16:58:23 +0000 (09:58 -0700)]
i965: Use blorp instead of meta for PBO texture downloads
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Thu, 1 Jun 2017 00:53:34 +0000 (17:53 -0700)]
i965/tex: Use blorp texture upload for all CCS_E textures
This improves the FillTex benchmark in GLBench 2.7 by 30% on my Broxton.
On Ken's Broxton which only has single-channel ram, it improves by 210%.
v2 (Ken): Check mt->aux_usage == ISL_AUX_USAGE_CCS_E rather than using
intel_miptree_is_lossless_compressed().
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Tue, 6 Jun 2017 16:58:07 +0000 (09:58 -0700)]
i965: Use blorp instead of meta for PBO texture uploads
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Sun, 27 Nov 2016 15:48:05 +0000 (17:48 +0200)]
i965: Add blorp-based texture upload and download paths
v1 (Topi Pohjolainen): original patch.
v2 (Topi Pohjolainen):
- Fix return value (s/MESA_FORMAT_NONE/false/) (Anuj)
- Move _mesa_tex_format_from_format_and_type() just
in the end avoiding additional if-block (Anuj)
- Explain better the array alignment restriction (Anuj)
- Do not bail out in case of gl_pixelstore_attrib::ImageHeight,
it is handled by _mesa_image_offset() automatically (Ken).
- Support 1D_ARRAY by flipping depth, width and y, z (Ken).
v3 (Topi Pohjolainen):
- Contrary to v2, do not try to handle
gl_pixelstore_attrib::ImageHeight. Currently there are no
tests in piglit or cts for it. One could possibly copy or
modify tests/texturing/texsubimage.c. There, however, seems
to be number of corner cases to consider. Moreover, current
meta path applies the packing height for both source and
targets when determining the offset. This would probably
require re-visiting also.
v4 (Topi Pohjolainen): Rebased on top of merged drm-bacon
v5 (Jason Ekstrand):
- Move to brw_blorp.c
- Significant refactoring
- Fixed 1-D array textures
- Simplified handling of PBOs vs. CPU data.
- Handle gl_pixelstore_attrib::ImageHeight. It turns out there are
piglit tests that cover this. The original version was failing them
because of an error in the way it handled 1-D array textures.
- Add support for texture download
v6 (Kenneth Graunke): Rebase fixes:
- Use intel_miptree_check_level_layer instead of deleted fields
- Update for mesa_format_supports_render[] rename.
- Pass 'false' (read-only) to intel_bufferobj_buffer
v7 (Kenneth Graunke):
- Fix brw_blorp_download_miptree to pass 'false' (not read only) for
the destination buffer (caught by Chris Wilson).
- Fix blorp_get_client_bo to pass intel_bufferobj_buffer !read_only
for the 'writable' parameter instead of 'false' (caught by Jason).
- Support GL_BGR, GL_BGRA, GL_BGRA_INTEGER, GL_BGR_INTEGER, allowing
us to use this for ReadPixels on the window system buffer (caught
by Chris Wilson).
- Fix y-flipping bugs in download path (exposed by BGRA support).
- Fix false vs. NULL return value in blorp_get_client_bo.
Signed-off-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Kenneth Graunke [Thu, 12 Oct 2017 03:33:25 +0000 (20:33 -0700)]
i965: Refactor y-flipping coordinate transform.
I want to reuse it for the BLORP download path.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Jason Ekstrand [Thu, 1 Jun 2017 00:04:13 +0000 (17:04 -0700)]
i965/tex: Check if there is data to upload up-front
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Thu, 15 Jun 2017 05:28:25 +0000 (22:28 -0700)]
i965/barrier: Do the correct flushes for framebuffer access
Framebuffer access includes framebuffer reads so we need to invalidate
the texture cache. We do not, however, need to flush the depth cache
because you cannot do bind a depth texture as an image.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Jason Ekstrand [Thu, 15 Jun 2017 05:27:20 +0000 (22:27 -0700)]
i965/barrier: Do the correct flushes for texture updates
Texture uploads and downloads may go through the render pipe which may
result in texturing from or rendering to the texture or the PBO. We
need to flush accordingly.
Reviewed-by: Topi Pohjolainen <topi.pohjolainen@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Thu, 12 Oct 2017 22:45:11 +0000 (15:45 -0700)]
include: Revert out the update of the Khronos GLX extension header.
They made a mistake in the MESA_swap_control XML, which I'm pursuing in
their github. Until then, we can just back this piece out.
Tested-by: Mark Janes <mark.a.janes@intel.com>
Reviewed-by: Mark Janes <mark.a.janes@intel.com>
Kenneth Graunke [Sat, 9 Sep 2017 07:20:26 +0000 (00:20 -0700)]
i965: Ignore GL_SKIP_DECODE_EXT for textures accessed via texelFetch().
The GL_EXT_texture_sRGB_decode spec says:
"The conversion of sRGB color space components to linear color space is
always performed if the texel lookup function is one of the texelFetch
builtin functions.
Otherwise, if the texel lookup function is one of the texture builtin
functions or one of the texture gather functions, the conversion of sRGB
color space components to linear color space is controlled by the
TEXTURE_SRGB_DECODE_EXT parameter.
If the TEXTURE_SRGB_DECODE_EXT parameter is DECODE_EXT, the conversion
of sRGB color space components to linear color space is performed.
If the TEXTURE_SRGB_DECODE_EXT parameter is SKIP_DECODE_EXT, the value
is returned without decoding. However, if the texture is also accessed
with a texelFetch function, then the result of texture builtin functions
and/or texture gather functions may be returned with decoding or without
decoding."
This patch makes i965 force sRGB decoding for any textures accessed via
texelFetch(). If textures are accessed via texelFetch() and a regular
texture access function, this will affect the other ones too - which is
fine - it's undefined according to the last paragraph quoted.
We could make both work, but we'd have to emit multiple SURFACE_STATEs,
and have two binding table sections, like we do for texture gather hacks
on older platforms.
Fixes the following Android O CTS test:
dEQP-GLES31.functional.srgb_texture_decode.skip_decode.srgba8.texel_fetch
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Kenneth Graunke [Wed, 11 Oct 2017 07:18:38 +0000 (00:18 -0700)]
meta: Unset the textures_used_by_txf bitfield.
Drivers that use Meta are happily using blitting data using texelFetch
and GL_SKIP_DECODE_EXT, but the GL_EXT_texture_sRGB spec unfortunately
makes GL_SKIP_DECODE_EXT not necessarily work with texelFetch.
As a hack, just unset the texture_used_by_txf bitfield so we can
continue with the old desired behavior.
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Kenneth Graunke [Sat, 9 Sep 2017 07:19:57 +0000 (00:19 -0700)]
nir: Make nir_shader_gather_info() track texelFetch texture accesses.
For TGSI-based drivers, st_glsl_to_tgsi records this information.
For NIR-based drivers, nir_shader_gather_info() will do so.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Kenneth Graunke [Sat, 9 Sep 2017 07:19:57 +0000 (00:19 -0700)]
compiler: Move gl_program::TexelFetchSamplers to shader_info.
I'd like to put this sort of metadata in the shader_info structure,
rather than adding more things to gl_program.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Dave Airlie [Thu, 12 Oct 2017 04:10:53 +0000 (05:10 +0100)]
radv: take unsafe_math and sisched into account when hashing shaders.
We want to generate different variants for sisched and unsafe_math
shader variants, so add them to the hash key.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Dave Airlie [Wed, 11 Oct 2017 23:32:14 +0000 (09:32 +1000)]
mesa/bufferobj: fix atomic offset/size get
When I realigned the bufferobj code, I didn't see the getters
were different, realign the getters to work the same as ssbo.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103214
Fixes: 65d3ef7cd (mesa: align atomic buffer handling code with ubo/ssbo (v1.1))
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Tested-by: Mark Janes <mark.a.janes@intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Marek Olšák [Thu, 12 Oct 2017 20:27:20 +0000 (22:27 +0200)]
relnotes: document EGL_ANDROID_native_fence_sync on radeonsi
Eric Anholt [Wed, 11 Oct 2017 16:54:47 +0000 (09:54 -0700)]
include: Update GL headers from khronos opengl registry.
Taken from their
c6a99aff31874697741a08cbc8a3488606ce59c7, keeping the
BUILDING_MESA hunk in place.
Reviewed-by: Daniel Stone <daniels@collabora.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Wed, 11 Oct 2017 16:50:13 +0000 (09:50 -0700)]
mapi: Update extension number of MESA_tile_raster_order.
Reviewed-by: Daniel Stone <daniels@collabora.com>
Acked-by: Kenneth Graunke <kenneth@whitecape.org>
Eric Anholt [Tue, 10 Oct 2017 20:56:55 +0000 (13:56 -0700)]
broadcom/vc5: Remove the u_resource_vtbl usage.
Like for vc4, this was just a wasted indirection.
Eric Anholt [Wed, 11 Oct 2017 17:32:25 +0000 (10:32 -0700)]
mesa: Disallow GL_RED/GL_RG with half-floats on GLES2.
Sure, you'd think that the combination of GL_OES_texture_half_float and
GL_EXT_texture_rg would mean that GL_RG16F exists, but it doesn't.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103227
Fixes: c16a7443e999 ("mesa: Expose GL_OES_required_internalformat on GLES contexts.")
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 12 Sep 2017 18:17:31 +0000 (20:17 +0200)]
radeonsi: implement sync_file import/export
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 12 Sep 2017 18:13:06 +0000 (20:13 +0200)]
winsys/amdgpu: implement sync_file import/export
syncobj is used internally for interactions with command submission.
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Marek Olšák [Tue, 12 Sep 2017 18:06:37 +0000 (20:06 +0200)]
ac: add radeon_info::has_sync_file
Reviewed-by: Nicolai Hähnle <nicolai.haehnle@amd.com>
Eric Anholt [Thu, 12 Oct 2017 00:40:35 +0000 (17:40 -0700)]
broadcom/vc5: Don't pair VPMSETUP with other peripheral access.
The specs don't say you can't, but pairing it with an SFU write on the
7268 breaks all our simple shader tests using gl_MVP * gl_Vertex.