freedreno/registers: Add a couple things used on kernel side

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6900>

freedreno/drm: Also clean ring_cache

We also need to release all the entries from the ring_cache when tearing
down the dev.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6900>

freedreno/a6xx: Small cleanup

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6900>

freedreno: Don't leak border_color_buf reference

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6900>

freedreno/drm: drop bo's dev reference

This is a bit over-paranoid, and can cause drm device fd leaks if there
is a bo leak somewhere.  Which is kind of a worse outcome.

This "fixes" a fd leak seen in:

dEQP-EGL.functional.query_context.get_current_display.*
dEQP-EGL.functional.query_context.get_current_context.*
dEQP-EGL.functional.query_context.get_current_display.*

(Still tracking down the root leak)

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6900>

zink: unify all occurrences of waiting on a fence

this was duplicated in a few places

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6922>

zink: combine all surface layout-setting for src/dst into util function

we do a lot of src/dst layout-setting, so ensure that we use the same code
everywhere to avoid cases where src==dst and we aren't setting GENERAL
layout as we should

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6922>

radv: Write correct dispatch size for RGP.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6875>

ci: Enable remaining (non-rotate) mustpass CTS tests

The gles3 and gles31 multisample and 565-no-depth-no-stencil caselists
are also mustpass.  And they don't add a significant number of extra
test cases.

The remaining mustpass caselists all involve rotation, which is not
currently supported in the surfaceless deqp build.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6553>

ci: cherry-pick deqp fix for config choosing

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6553>

ci/deqp-runner: Allow overriding width/height/config

This will allow adding multi-sample caselists, and jobs with larger
surface size.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6553>

intel: Add support for i945g to intel_stub_gpu.

I wanted to check the impact of nir-to-tgsi on the i915g driver, but I
don't have an i915 set up.  Similarly, for driconf refactoring I needed to
make sure that classic i915_dri.so still loaded.

Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6868>

Add driver override to envvars.rst

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6911>

gallium/vl: add chroma_format arg to vl_video_buffer functions

vl_mpeg12_decoder needs to override the chroma_format value to get the
correct size calculated (chroma_format is used by vl_video_buffer_adjust_size).

I'm not sure why it's needed, but this is needed to get correct mpeg decode.

Fixes: 24f2b0a8560 ("gallium/video: remove pipe_video_buffer.chroma_format")
Acked-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6817>

gallium/vl: do not call transfer_unmap if transfer is NULL

CC: mesa-stable
Acked-by: Leo Liu <leo.liu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6817>

tu: Enable multi-position output

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6515>

ir3, tu: Link per-view position correctly

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6515>

ir3, tu: Run optimization loop twice

This call to ir3_optimize_nir() mirrors what st/mesa does for us in
Gallium, and will be necessary for cross-stage linking and the multiview
lowering.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6515>

tu: Refactor shader compilation flow

In order to do cross-stage linking, we'll need to split out SPIR-V->NIR
and NIR finalization, so that we can do a round of linking in between.
The multiview lowering pass also assumes that it sits between two
optimization loops, which in anv are the pre-linking optimizations and
post-linking finalization.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6515>

tu: Write multiview control registers in binning pass

Multiview is never used with binning, but we still need to make sure to
disable it in the binning pass.

Fixes: c0c7dbd ("tu: Implement multiview pipeline state")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6515>

gallium: Make pipe_viewport_state swizzle_x/y/z/w bit-fields 8 bits wide

Previously there were 20 padding bits after them, which would be left
uninitialized and preserved when writing to the swizzle members. This
could result in two equal viewport states spuriously being considered
different (because memcmp compared the padding bits as well).

Noticed while looking for something else with valgrind:

==801624== Conditional jump or move depends on uninitialised value(s)
==801624==    at 0x10B86259: cso_set_viewport (cso_context.c:739)
==801624==    by 0x10B862C7: cso_set_viewport_dims (cso_context.c:764)
==801624==    by 0x1057E3A1: clear_with_quad (st_cb_clear.c:335)
==801624==    by 0x1057E3A1: st_Clear (st_cb_clear.c:545)
==801624==    [...]
==801624==
==801624== Conditional jump or move depends on uninitialised value(s)
==801624==    at 0x10B885DE: cso_restore_viewport (cso_context.c:777)
==801624==    by 0x10B885DE: cso_restore_state (cso_context.c:1710)
==801624==    by 0x1057E4CB: clear_with_quad (st_cb_clear.c:364)
==801624==    by 0x1057E4CB: st_Clear (st_cb_clear.c:545)
==801624==    [...]

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6867>

radv: move lowering of FS outputs outside of ACO

This enables lowering of FS outputs for RADV/LLVM.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6865>

radv: call nir_io_add_const_offset_to_base for FS outputs

The store_output of RADV/LLVM expects the const offset to be 0.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6865>

radv/aco: Set I/O variable locations outside ACO.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Acked-by: Daniel Schürmann <daniel@schuermann.dev>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6865>

radeonsi: enable support for GL_NV_shader_atomic_int64

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6708>

mesa,glsl: add support for GL_NV_shader_atomic_int64

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6708>

gallium: prepare framework for supporting GL_NV_shader_atomic_int64

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6708>

turnip: initial implementation of VK_KHR_push_descriptor

Add missing descriptor sets code for push descriptors, and a simple initial
implementation to enable the extension and pass dEQP tests.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6755>

turnip: delete unused/broken pipeline layout hashing code

Note: immutable samplers hash was wrong since we have an array of
tu_sampler and not 4 dwords like radv.

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6755>

turnip: remove unused cmd_buffer/device arguments in descriptor sets

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6755>

panfrost: Add support for rbg16 formats

Fixes: 171e94fe629 ("mesa/st: enable EXT_color_buffer_half_float when formats supported")
Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6887>

CI: Re-enable Panfrost T7xx

Lab maintenance is complete.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6910>

nir: Handle per-view io in nir_io_add_const_offset_to_base()

This isn't strictly necessary for freedreno, since we aren't using it
yet, but I wanted to avoid any problems if we do. If we wanted to handle
this "properly", and handle matrix and array per-view variables, we'd
probably want to encode the "view stride" (number of views per user
location) and base view in the intrinsic, but for now we just don't do
any offsetting and assume the indirect offset is the view.

Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6514>

nir: Add per_view to IO semantics

Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6514>

nir/lower_io_arrays: Fix xfb_offset bug

I noticed this once I started gathering xfb_info after
nir_lower_io_arrays_to_elements_no_indirect.

Fixes: b2bbd978d0b ("nir: fix lowering arrays to elements for XFB outputs")
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6514>

nir: Count i/o slots correctly for per-view variables

This function wasn't counting driver slots correctly, resulting in
incorrect driver_location's and input_count. It seems intel doesn't use
this yet.

Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6514>

anv: Use nir_lower_multiview pass

Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6514>

nir: Add nir_lower_multiview pass

Taken mostly directly from the anv pass. A few anv-specific things that
I could leave in anv aren't included. Specifically on turnip we don't
need to set gl_Layer to 0, and we can handle the case where the FS reads
gl_ViewIndex, so that check is moved into anv.

Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6514>

gallium/util: allow scaling blits for stencil-fallback

We also need to be able to scale the stencil-blits, so let's pass the
full destination box instead of just the position.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6898>

gallium/util: fix texture-coordinates for stencil-fallback

Set up texture coordinates for the stencil-fallback blit code. This
worked in the orignal NIR code, but accidentally broke when rewriting to
use TGSI, and my test-case had a constant colored stencil buffer.

Fixes: e8a40715a8b ("gallium/util: add blitter-support for stencil-fallback")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6898>

gallium/util: set right dst-dimensions

This should have been the destination surface size, not the dimensions
of the source box. These were the same in the test-case I used while
developing this, but this matters for the GTF framebuffer-blit
functional CTS tests.

Fixes: e8a40715a8b ("gallium/util: add blitter-support for stencil-fallback")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6898>

radv/llvm: call nir_lower_io_to_vector with FS to fix array tests

Fixes dEQP-VK.glsl.440.linkage.varying.component.frag_out.*.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6883>

docs: add INTEL_SHADER_ASM_READ_PATH description

It was implemented in
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/955

Signed-off-by: Danylo Piliaiev <danylo.piliaiev@globallogic.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Sagar Ghuge <sagar.ghuge@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6895>

radeonsi: call nir_io_add_const_offset_to_base only once per shader

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6890>

nir/lower_io: change nir_io_add_const_offset_to_base to use bitfield modes

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6890>

st/mesa: use roundf instead of floorf for lod-bias rounding

There's no good reason not to use a symmetric rounding mode here. This
fixes the following GL CTS case for me:

GTF-GL33.gtf21.GL3Tests.texture_lod_bias.texture_lod_bias_all

Fixes: 132b69c4edb ("st/mesa: round lod_bias to a multiple of 1/256")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6892>

docs: gallium -> Gallium

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6894>

docs: docker -> Docker

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6894>

docs: debian -> Debian

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6894>

docs: flavours -> flavors

We generally prefer US English over UK English in the docs.

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6894>

docs: stabilisation -> stabilization

We generally prefer US English in the docs over UK English.

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6894>

docs: consistantly -> consistently

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6894>

docs: clippping -> clipping

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6894>

docs: drop outdated gallium-docs comment

The separate readthedocs documentation is quite pointless these days, as
it's been moved to docs.mesa3d.org, where all other documentation
already is. There's nothing special about this documentation any longer.

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6893>

kmsro: Add mediatek entry point

Add an entry for the mediatek modesetting driver.

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6885>

panfrost: Add preliminary support for Mali G72

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6885>

panfrost: Adjust quirks for bifrost v6

The G72, which is a bifrost v6, supports fast operations. The selection
should probably be done on a per-GPU basis instead of forcing this quirk
for all v6 GPUs.

Signed-off-by: Boris Brezillon <boris.brezillon@collabora.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6885>

mesa: bptc fixes for decompress rgba_unorm and rgb_float

Save bit_offset between iterations fixes for piglit:
* LIBGL_ALWAYS_SOFTWARE=true GALLIUM_DRIVER=softpipe piglit/bin/bptc-float-modes
* LIBGL_ALWAYS_SOFTWARE=true GALLIUM_DRIVER=llvmpipe piglit/bin/bptc-float-modes

Memset to zero in reserved mode for rgba_unorm fixes for VK-GL-CTS with libvulkan_val:
* dEQP-VK.texture.compressed.bc7_unorm_block_2d_pot
* dEQP-VK.texture.compressed.bc7_srgb_block_2d_pot
* dEQP-VK.texture.compressed.bc7_unorm_block_2d_npot
* dEQP-VK.texture.compressed.bc7_srgb_block_2d_npot

Signed-off-by: Denis Pauk <pauk.denis@gmail.com>
Reviewed-by: Dave Airlie <airlied@redhat.com>
Acked-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6809>

ac/llvm: fix invalid IR if image stores are shrinked using the format

It's not always v4f32 (or v4f16 for 16-bit) when image stores are
shrinked using the format.

This fixes a ton of crashes with RADV_DEBUG=checkir,llvm.

Fixes: e4d75c22beb ("nir/opt_shrink_vectors: shrink image stores using the format")
Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6882>

radv: lower deref operations for global memory for both backends

To match ACO.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5316>

ac/nir: implement nir_intrinsic_global_atomic_*

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5316>

ac/nir: implement nir_intrinsic_{load,store}_global

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5316>

CI: Temporarily disable Panfrost T7xx

Due to more work in the on-site lab, we need to take the T7xx machines
out of the rotation for a couple of hours.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6896>

pan/mdg: Fix memory leak on error path.

Fix defect reported by Coverity Scan.

Resource leak (RESOURCE_LEAK)
leaked_storage: Variable l going out of scope leaks the storage it
points to.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6858>

gallium/swr: Remove unreachable code.

pContextMem has already been checked and cannot be NULL.

Fix defect reported by Coverity Scan.

Logically dead code (DEADCODE)
dead_error_line: Execution cannot reach this statement: return tsCtx;

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Jan Zielinski <jan.zielinski@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6857>

aco: Initialize mad_info member literal_idx.

Fix defect reported by Coverity Scan.

Uninitialized scalar field (UNINIT_CTOR)
uninit_member: Non-static class member literal_idx is not initialized in
this constructor nor in any functions that it calls.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Tony Wasserka <tony.wasserka@gmx.de>
Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6832>

anv: Check file descriptor before closing.

Fix defect reported by Coverity Scan.

Argument cannot be negative (NEGATIVE_RETURNS)
negative_returns: fd is passed to a parameter that cannot be negative

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6643>

intel/vec4: Remove everything related to VS_OPCODE_SET_SIMD4X2_HEADER_GEN9

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6826>

intel/vec4: Remove all support for Gen8+ [v2]

v2: Restore the gen == 10 hunk in brw_compile_vs (around line 2940).
This function is also used for scalar VS compiles.  Squash in:

    intel/vec4: Reindent after removing Gen8+ support
    intel/vec4: Silence unused parameter warning in try_immediate_source

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net> [v1]
Reviewed-by: Matt Turner <mattst88@gmail.com> [v1]
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org> [v1]
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6826>

intel/compiler: Remove INTEL_SCALAR_... env variables

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6826>

intel/vec4: Remove inline lowering of LRP

Since dd7135d55d5 ("intel/compiler: Use the flrp lowering pass for all
stages on Gen4 and Gen5"), it's not possible to get to this function on
GPUs that don't have a LRP instruction.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6826>

intel/compiler: Don't fallback to vec4 when scalar GS compile fails [v2]

v2: Add missing error string handling.  Noticed by Jason.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com> [v1]
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6826>

intel/compiler: Silence unused parameter warning in brw_surface_payload_size

src/intel/compiler/brw_eu_emit.c: In function ‘brw_surface_payload_size’:
src/intel/compiler/brw_eu_emit.c:3070:46: warning: unused parameter ‘p’ [-Wunused-parameter]
 3070 | brw_surface_payload_size(struct brw_codegen *p,
      |                          ~~~~~~~~~~~~~~~~~~~~^

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6826>

intel/vec4: Silence unused paramter warnings in brw_vec4_generator.cpp

src/intel/compiler/brw_vec4_generator.cpp: In function ‘void generate_gs_svb_write(brw_codegen*, brw_vue_prog_data*, brw::vec4_instruction*, brw_reg, brw_reg, brw_reg)’:
src/intel/compiler/brw_vec4_generator.cpp:488:49: warning: unused parameter ‘prog_data’ [-Wunused-parameter]
  488 |                       struct brw_vue_prog_data *prog_data,
      |                       ~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~
src/intel/compiler/brw_vec4_generator.cpp: In function ‘void generate_pull_constant_load(brw_codegen*, brw_vue_prog_data*, brw::vec4_instruction*, brw_reg, brw_reg, brw_reg)’:
src/intel/compiler/brw_vec4_generator.cpp:1269:55: warning: unused parameter ‘prog_data’ [-Wunused-parameter]
 1269 |                             struct brw_vue_prog_data *prog_data,
      |                             ~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~
src/intel/compiler/brw_vec4_generator.cpp: In function ‘void generate_get_buffer_size(brw_codegen*, brw_vue_prog_data*, brw::vec4_instruction*, brw_reg, brw_reg, brw_reg)’:
src/intel/compiler/brw_vec4_generator.cpp:1331:52: warning: unused parameter ‘prog_data’ [-Wunused-parameter]
 1331 |                          struct brw_vue_prog_data *prog_data,
      |                          ~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~
src/intel/compiler/brw_vec4_generator.cpp: In function ‘void generate_pull_constant_load_gen7(brw_codegen*, brw_vue_prog_data*, brw::vec4_instruction*, brw_reg, brw_reg, brw_reg)’:
src/intel/compiler/brw_vec4_generator.cpp:1357:60: warning: unused parameter ‘prog_data’ [-Wunused-parameter]
 1357 |                                  struct brw_vue_prog_data *prog_data,
      |                                  ~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6826>

r600/sfn: Add support for helper invocations

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: remove old code to track uniforms as it is no longer needed

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: replace hand-backed literal check by NIR function

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn; go back to not lowering uniforms to UBOs

Lowering uniforms to UBOs results in an aditional iadd for the
UBO buffer id evaluation, and for indirect buffers access that
results in an unnecessary op that can be avoided by not lowering
uniforms. There is some code duplication when reading the uniforms
but it saves a whole instruction group per indirect cont buffer
access.

This reverts commit 98eb00face93b9af5aac19008ecff5a2bf376745 with
some additional fixes.

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: Fix indirect const buffer access

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: Fix interpolate at sample

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: use fine gradient evaluation for interpolate_at_offset

Also pass in the interpolator x component (because the TGSI does this too)

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: save some instructions when doing multisample on sample 0

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: Support group memory barrier

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: lower to scalar for some optimizations and vectortize later

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: Handle nir_op_b2b32

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

r600/sfn: Fix emitting shared atomics with constant sources

Signed-off-by: Gert Wollny <gert.wollny@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6879>

nir/lower_clip: Add i/o semantics for load/store intrinsics.

ir3 looks at the .location on its inputs for handling
non-VARYING_SLOT_POS, but our uninitialized semantics matched that and
threw a compiler assertion failure.

Fixes: 502abfce7f5d ("nir: save IO semantics in lowered IO intrinsics")
Reviewed-by: Connor Abbott <cwabbott0@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6716>

ci/freedreno: Add trace tests for glxgears, 0 A.D., and xonotic.

glxgears is now trimmed to not be tons of frames, and we've picked up a
couple more traces in the public repo.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6716>

ci/freedreno: Sort the traces in the .yml of expectations

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6716>

turnip: Replace tu_log*() with mesa_log*()

This gets us logging on Android.

Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6806>

mesa: Promote Intel's simple logging façade for Android to util/

I'm bringing up freedreno Vulkan on an Android phone, and my pains are
exactly what Chad said when working on Intel's vulkan for Android in
aa716db0f64d ("intel: Add simple logging façade for Android (v2)"):

    On Android, stdio goes to /dev/null. On Android, remote gdb is even
    more painful than the usual remote gdb. On Android, nothing works like
    you expect and debugging is hell. I need logging.

    This patch introduces a small, simple logging API that can easily wrap
    Android's API. On non-Android platforms, this logger does nothing
    fancy.  It follows the time-honored Unix tradition of spewing
    everything to stderr with minimal fuss.

    My goal here is not perfection. My goal is to make a minimal, clean API,
    that people hate merely a little instead of a lot, and that's good
    enough to let me bring up Android Vulkan.  And it needs to be fast,
    which means it must be small. No one wants to their game to miss frames
    while aiming a flaming bow into the jaws of an angry robot t-rex, and
    thus become t-rex breakfast, because some fool had too much fun desiging
    a bloated, ideal logging API.

Compared to trusty fprintf, _mesa_log[ewi]() is actually usable on
Android.  Compared to os_log_message(), this has different error levels
and supports format arguments.

The only code change in the move is wrapping flockfile/funlockfile in
!DETECT_OS_WINDOWS, since mingw32 doesn't have it.  Windows likely wants
different logging code, anyway.

Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Reviewed-by: Kristian H. Kristensen <hoegsberg@google.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6806>

radv: Record cache flushes for RGP.

Not doing the EOP TS cacheflush event because that break wave counting
in RGP for some reason. But the rest looks to be all there.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6550>

radv: Include flushes in the barrier.

Since the flushes really happen on the next draw delay the barrier
end to include the flushes.

This fixes the barrier duration in RGP.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6550>

radv: Fix RGP Asic CU info for GFX10+.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6550>

radv: Use correct alignment for SQTT buffer sizes.

Otherwise we can miss the start of the buffer, which
results in RGP missing the initial (and only) timestamp
packet and results in RGP completely ignoring the second
SE, causing messy issues and missed events.

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6550>

radv: Fix emitting SQTT userdata.

Otherwise some marker packets were missing ...

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6550>

zink: remove old extension infrastructure

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6789>

zink: use the new, generated extension infrastructure

this is just a find-and-replace from screen->... to screen->info.

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6789>

zink: use the new extension infrastructure in device creation

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6789>

zink: hook zink_device_info.py to build system

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/6789>