aco: Remove use of deprecated Operand constructors in aco_builder.h

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11653>

aco: Remove use of deprecated Operand constructors in test_to_hw_instr.cpp

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11653>

aco: Replace Operand literal constructors with factory member functions

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11653>

ci: Make sure that we build the piglit dmabuf tests.

Force the option rather than relying on autodetection -- ARM runners were
apparently finding the necessary deps, but the x86 rootfs (radeonsi, iris)
and x86_test-gl container (i915g) were not.

Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11834>

zink: simplify modifier ifdefs

these are the only two defines referenced, so they can be defined to 0
for platforms that don't support modifiers in order to remove a ton of
ifdefs and make the code more readable

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11847>

freedreno/ir3: Add float immed "FLUT" support

We can encode a limited set of float immeds into cat2 instructions,
using hw's float lookup table (FLUT) feature.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/36
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8705>

freedreno/computerator: Add script to probe FLUT values

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8705>

turnip: Split tu6_emit_xs()

Emit all the state layout config (such as push-const CONSTLEN) first,
before emitting anything that depends on that state.  This fixes an
issue that was showing up when FLUT is enabled in ir3 (which results
in higher probability of not having any immediats lowered to push-
consts).

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8705>

turnip: avoid some UB

Reduce a bit of extra noise that makes diffing cmdstream traces more
annoying.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8705>

glsl: Delete lower_texture_projection

This is only used by i965 and we've been getting it through
nir_lower_tex since forever.  Get rid of the GLSL IR pass.

Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11827>

ci: only trigger gallium_core_file_list jobs from dri and glx frontend changes

these are the only frontends which may be used by gallium drivers in ci,
so stop triggering all driver jobs when other frontends are changed since
those changes can never affect ci

<MrCooper> Not that simple unfortunately. E.g. the llvmpipe-piglit-cl job hits
           src/gallium/frontends/clover & possibly src/gallium/targets/opencl,
           many jobs hit src/gallium/{frontends,targets}/dri and probably
           src/gallium/targets/pipe-loader, lavapipe jobs hit src/gallium/{frontends,targets}/lavapipe.

Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11832>

ci: add vulkan files to lavapipe rules

Acked-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11833>

pan/bi: Create a nop clause when the shader starts with ATEST

Otherwise there would be no clause with the dependencies needed for
ATEST set, so the GPU would get stuck.

Not needed on v7, as there shader_wait_dependency in the RSD will wait
for the dependencies before the shader starts.

Explicitly create a NOP instruction, as it is assumed that clauses
have a non-zero count of instructions in various places.

Fixes GPU timeouts in many applications, such as SuperTuxKart and
GZDoom.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11842>

panfrost: Fix full_threads calculation on v6

Fixes: 8ba2f9f6985 ("panfrost: Create a blitter library to replace the existing preload helpers")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11842>

gbm/dri: Fix leaking bo memory on failure path

In gbm_dri_bo_create, when modifiers are requested but not supported, do
not return NULL immediately, but first go to cleanup section to free
already allocated buffer object.

Fixes: cb9ae4273d68 ("dri: add loader_dri_create_image helper")
Signed-off-by: Heinrich Fink <hfink@snap.com>
Reviewed-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Simon Ser <contact@emersion.fr>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11844>

pps: Panfrost documentation

Add documentation for the Panfrost Perfetto datasource.

Signed-off-by: Antonio Caggiano <antonio.caggiano@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10215>

pps: Panfrost pps driver

Add the Panfrost pps driver.

v2: Human readable names for counter blocks and use `unreachable`.
v3: Use libpanfrost_perf to collect counter values.

Signed-off-by: Antonio Caggiano <antonio.caggiano@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10215>

mesa: clear shader_info::is_lowered in prog_to_nir

This needs to be resetted each time prog_to_nir is called because it
turns st_nir_assign_vs_in_locations into a no-op when set.

Fixes: 81d106d6eca ("radeonsi: lower IO intrinsics - complete rewrite of input/output scanning")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5001
Reviewed-by: Isaac Bosompem <mrisaacb@google.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11636>

v3dv: use defines for push constant offsets used by texel buffer copy shaders

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11843>

v3dv: allow batching texel buffer copies for 3D images

For these we only need to check that the depth extent we are
copying is the same across regions in the batch, since we use
that to specify the number of layers in the framebuffer used
for the copy.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11843>

v3dv: implement layered texel buffer copies using a geometry shader

Instead of specifying a separate framebuffer per layer which is expected
to be much slower.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11843>

v3dv: fix push constant range for texel buffer copy pipelines

As per get_texel_buffer_copy_fs(), we load 24 bytes of data.

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11843>

zink: remove duplicate format-mapping on little-endian

Doing *both* of thse ends up rewriting the previous mapping. Since this
doesn't seem to have lead to issues, it seems like the new mapping works
just as well.

Fixes: a22a1c0324e ("zink: Fix VK_FORMAT_A8B8G8R8_SRGB_PACK32 mapping on big-endian")
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11417>

loader: Handle failure to load DRI driver library

I factored out the chunk of loader code that dlopen()s
libraries from the rest of the DRI driver loader function
in this commit:

  commit bc343154f816b637b200933e1c1a35db19f478ef
  Author: James Jones <jajones@nvidia.com>
  Date:   Thu Apr 22 23:17:08 2021 -0700

  loader: Factor out driver library loading code

However, I failed to adjust the DRI loader function that
now uses the new helper function to handle the case where
the requested DRI library is not found.

This change restores the prior behavior, and also ensures
loader_open_driver() consistently returns NULL in the
out_driver_handle parameter on failure.

Fixes: bc343154f816 ("loader: Factor out driver library loading code")
Signed-off-by: James Jones <jajones@nvidia.com>
Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11807>

aco: fix extract_vector optimization

If the allocated_vec map contains a different RegType
for the elements, ensure that the size matches exactly.

Otherwise, it could happen that extracting a dword
element matched with a subdword element.

No fossil-db changes.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11823>

aco: fix self-intersecting register swaps

Splitting self-intersecting register swaps into
3 sections was unnecessary and only worked because
the middle section was always empty for full dword
swaps.

No fossil-db changes.

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11823>

lavapipe: add the separate depth/stencil layout enable.

This was missing.

Fixes: ac07ad8414cd ("lavapipe: implement KHR_separate_depth_stencil_layouts")
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11840>

panfrost: Fix format swizzles on G72

Careless typo. Cleaner fix suggested by Icecream95.

Reported-by: macc24
Tested-by: macc24
Fixes: 531d80de7c7 ("panfrost: Specialize blendable_formats for v6")
Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11837>

relnotes: add some missing zink/lavapipe updates

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11838>

features: mark off line rasterization for lavapipe

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11838>

zink: add a pipe_screen::resource_create_with_modifiers hook

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: add fallback for linear modifier use

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: don't pass modifier count to first image create

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: explicitly disallow using the modifier image create for non-linear images

the first call is just for the base image, the second call is the one that needs
modifiers

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: use VkImageDrmFormatModifierListCreateInfoEXT for creating from modifier array

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: add a pipe_screen::resource_get_param hook

this is used for querying image properties

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: store modifier aspect to resource

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: start storing modifiers to the base resource struct

need these for rebinds

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: populate modifier props onto screen object during init

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

zink: implement support for non-planar DRM modifiers

Planar DRM modifiers are a bit more tricky, so let's kick that ball down
the road. For now, this should help a bit.

Acked-by: Dave Airlie <airlied@redhat.com>
Reviewed-by: Emil Velikov <emil.l.velikov@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10782>

mailmap: Add two more lines for Alyssa Rosenzweig

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11788>

Convert a few files to UTF-8

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11788>

mailmap: Update for Emma's new e-mail address

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Acked-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11788>

pan/bi: Add a bundling heuristic

Pick instructions locally that are likely to minimize nop's globally.
Note the obvious greedy solution is ineffective. Instead the principle
is to choose instructions which will increase _future_ scheduler
freedom, on the assumption that means fewer nop's overall will be
needed.

There is no concern about register pressure and little concern about
message timing, since this is post-RA and message ordering was fixed.

total tuples in shared programs: 125304 -> 123770 (-1.22%)
tuples in affected programs: 88301 -> 86767 (-1.74%)
helped: 548
HURT: 198
helped stats (abs) min: 1.0 max: 31.0 x̄: 3.43 x̃: 2
helped stats (rel) min: 0.33% max: 16.67% x̄: 3.09% x̃: 2.44%
HURT stats (abs)   min: 1.0 max: 12.0 x̄: 1.73 x̃: 1
HURT stats (rel)   min: 0.27% max: 11.11% x̄: 2.82% x̃: 2.35%
95% mean confidence interval for tuples value: -2.39 -1.72
95% mean confidence interval for tuples %-change: -1.77% -1.27%
Tuples are helped.

total clauses in shared programs: 26059 -> 25655 (-1.55%)
clauses in affected programs: 10821 -> 10417 (-3.73%)
helped: 311
HURT: 42
helped stats (abs) min: 1.0 max: 6.0 x̄: 1.44 x̃: 1
helped stats (rel) min: 0.48% max: 16.00% x̄: 5.26% x̃: 5.00%
HURT stats (abs)   min: 1.0 max: 2.0 x̄: 1.05 x̃: 1
HURT stats (rel)   min: 2.13% max: 33.33% x̄: 7.49% x̃: 6.25%
95% mean confidence interval for clauses value: -1.26 -1.02
95% mean confidence interval for clauses %-change: -4.28% -3.21%
Clauses are helped.

total cycles in shared programs: 12154.79 -> 12114.83 (-0.33%)
cycles in affected programs: 1907.63 -> 1867.67 (-2.09%)
helped: 227
HURT: 99
helped stats (abs) min: 0.041665999999999315 max: 1.2083360000000027 x̄: 0.22 x̃: 0
helped stats (rel) min: 0.36% max: 20.00% x̄: 4.14% x̃: 3.85%
HURT stats (abs)   min: 0.041665999999999315 max: 0.5 x̄: 0.09 x̃: 0
HURT stats (rel)   min: 0.28% max: 9.09% x̄: 2.75% x̃: 2.63%
95% mean confidence interval for cycles value: -0.15 -0.09
95% mean confidence interval for cycles %-change: -2.51% -1.59%
Cycles are helped.

total arith in shared programs: 4658.13 -> 4603.42 (-1.17%)
arith in affected programs: 3449.83 -> 3395.12 (-1.59%)
helped: 509
HURT: 249
helped stats (abs) min: 0.041665999999999315 max: 1.2083360000000027 x̄: 0.14 x̃: 0
helped stats (rel) min: 0.36% max: 20.00% x̄: 3.56% x̃: 2.63%
HURT stats (abs)   min: 0.041665999999999315 max: 0.5 x̄: 0.07 x̃: 0
HURT stats (rel)   min: 0.28% max: 25.00% x̄: 3.32% x̃: 2.27%
95% mean confidence interval for arith value: -0.09 -0.06
95% mean confidence interval for arith %-change: -1.64% -0.97%
Arith are helped.

total quadwords in shared programs: 111394 -> 110114 (-1.15%)
quadwords in affected programs: 78074 -> 76794 (-1.64%)
helped: 503
HURT: 204
helped stats (abs) min: 1.0 max: 43.0 x̄: 3.33 x̃: 2
helped stats (rel) min: 0.41% max: 13.79% x̄: 3.02% x̃: 2.44%
HURT stats (abs)   min: 1.0 max: 26.0 x̄: 1.94 x̃: 1
HURT stats (rel)   min: 0.48% max: 7.55% x̄: 2.44% x̃: 2.22%
95% mean confidence interval for quadwords value: -2.16 -1.46
95% mean confidence interval for quadwords %-change: -1.67% -1.21%
Quadwords are helped.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

pan/bi: Calculate dependency graph when bundling

Code is ported from Midgard, modified to be scalar, post-RA, and to put
the arrays on the worklist instead of the instruction to save memory.
This enables out-of-order scheduling.

total tuples in shared programs: 128691 -> 125304 (-2.63%)
tuples in affected programs: 114091 -> 110704 (-2.97%)
helped: 844
HURT: 377
helped stats (abs) min: 1.0 max: 150.0 x̄: 4.88 x̃: 3
helped stats (rel) min: 0.30% max: 26.42% x̄: 5.56% x̃: 4.35%
HURT stats (abs)   min: 1.0 max: 8.0 x̄: 1.94 x̃: 1
HURT stats (rel)   min: 0.20% max: 33.33% x̄: 6.84% x̃: 3.23%
95% mean confidence interval for tuples value: -3.16 -2.38
95% mean confidence interval for tuples %-change: -2.19% -1.27%
Tuples are helped.

total clauses in shared programs: 27579 -> 26059 (-5.51%)
clauses in affected programs: 20606 -> 19086 (-7.38%)
helped: 941
HURT: 39
helped stats (abs) min: 1.0 max: 21.0 x̄: 1.66 x̃: 1
helped stats (rel) min: 0.69% max: 44.44% x̄: 10.48% x̃: 9.09%
HURT stats (abs)   min: 1.0 max: 2.0 x̄: 1.15 x̃: 1
HURT stats (rel)   min: 1.89% max: 10.00% x̄: 4.73% x̃: 4.55%
95% mean confidence interval for clauses value: -1.63 -1.47
95% mean confidence interval for clauses %-change: -10.27% -9.48%
Clauses are helped.

total cycles in shared programs: 12262.54 -> 12154.79 (-0.88%)
cycles in affected programs: 2210.54 -> 2102.79 (-4.87%)
helped: 374
HURT: 56
helped stats (abs) min: 0.041665999999999315 max: 6.25 x̄: 0.30 x̃: 0
helped stats (rel) min: 0.42% max: 26.00% x̄: 6.90% x̃: 7.14%
HURT stats (abs)   min: 0.041665999999999315 max: 0.5833319999999986 x̄: 0.11 x̃: 0
HURT stats (rel)   min: 0.16% max: 100.00% x̄: 55.17% x̃: 50.00%
95% mean confidence interval for cycles value: -0.29 -0.21
95% mean confidence interval for cycles %-change: -1.37% 3.73%
Inconclusive result (%-change mean confidence interval includes 0).

total arith in shared programs: 4852.29 -> 4658.13 (-4.00%)
arith in affected programs: 4525.17 -> 4331 (-4.29%)
helped: 1112
HURT: 166
helped stats (abs) min: 0.041665999999999315 max: 6.25 x̄: 0.19 x̃: 0
helped stats (rel) min: 0.42% max: 33.33% x̄: 6.59% x̃: 5.36%
HURT stats (abs)   min: 0.041665999999999315 max: 0.5833319999999986 x̄: 0.07 x̃: 0
HURT stats (rel)   min: 0.16% max: 100.00% x̄: 25.05% x̃: 2.40%
95% mean confidence interval for arith value: -0.17 -0.14
95% mean confidence interval for arith %-change: -3.44% -1.51%
Arith are helped.

total quadwords in shared programs: 117141 -> 111394 (-4.91%)
quadwords in affected programs: 104390 -> 98643 (-5.51%)
helped: 1245
HURT: 76
helped stats (abs) min: 1.0 max: 69.0 x̄: 4.74 x̃: 4
helped stats (rel) min: 0.28% max: 35.00% x̄: 7.88% x̃: 6.45%
HURT stats (abs)   min: 1.0 max: 8.0 x̄: 2.01 x̃: 1
HURT stats (rel)   min: 0.20% max: 10.00% x̄: 3.52% x̃: 4.25%
95% mean confidence interval for quadwords value: -4.61 -4.09
95% mean confidence interval for quadwords %-change: -7.56% -6.88%
Quadwords are helped.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

pan/bi: Handle 4-src instructions in scheduler

Spill to a move. This allows us to emit SHADDX and general CSEL.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

pan/bi: Add bi_before_tuple convenience method

For constructing instructions during scheduling.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

pan/bi: Handle multiple destinations in scheduler

Needed for correct handling of pseudo ops like SHADDX and CUBEFACE

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

pan/bi: Do helper termination analysis on clauses

Unlike the dependency analysis for the skip bits which is a function of
the data flow graph, the thread termination analysis is dependent on the
actual sequence of instructions. As such, it must be done after
scheduling to be correct in the presence of out-of-order scheduling.

Furthermore it's specified in terms of clauses, not instructions.
Reflecting this in our code gets a nice simplification. As a side effect
this puts extra td flags on subsequent clauses, which matches the DDK's
behaviour. (Maybe td is just a hint?)

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

pan/bi: Update ins->link after scheduling

Otherwise foreach_instr and friends will be subtly wrong. None of our
current passes rely on this but it's a footgun.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

pan/bi: Make bi_foreach_instr_in_tuple safer

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

pan/bi: Refuse to CSE non-SSA sources

They might have changed in between.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10961>

panfrost: Only access blitter from per-gen

Now the blitter may be compiled per-gen legally.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Init/destroy blitter from per-gen file

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Remove pan_blitter integration

This was useful when panvk was being bootstrapped, allowing the blitter
to be tested against a known-working driver. With panvk in-tree, I don't
see a compelling reason to keep pan_blitter support wired in but
off-by-default. Keeping it will complicate the GenXML change we're about
to make.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Avoid GenXML enum dependences

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Remove reference to mali_blend_equation_packed

Genxml dependent.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Use generic delete for ZSA

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Pin an architecture for blending

Need to pick something but it's all the same, even on Valhall.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Give WLS Instances a default

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Use smaller sizes in blend table

Saves a few kb.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Specialize blendable_formats for v6

Let's bake the swizzle in.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Compile format table multiple times

This allows us to unify the midgard and bifrost tables and just #ifdef
the differences. It will soon also allow us to fix a bunch of enums and
specialize blendable formats.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Add GenXML macros

From intel.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Remove panfrost_bifrost_swizzle

Deprecated and now unused.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panvk: Don't use panfrost_bifrost_swizzle

It's not needed. This produces chooses slightly different formats on v7
but no functional change.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Move arch-independent pan_format code

Now pan_format.c is just tables.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Inline panfrost_get_z_internal_format

Trivial, get it out of the way so pan_format can focus on the tables.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Assert that injected jobs are for blits

Remove a level of indentation.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Inline away pan_invocation.c

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Express pack_work_groups more concisely

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Inline away pan_pool.c

Just noise now that it's been abstracted for panvk.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Move panfrost_vertex/instance_id to per-gen

Now the rest of pan_attributes.c is GenXML-independent.

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

panfrost: Inline flip_compare_func into pan_encoder.h

This will become the home for little GenXML-aware helpers, suitable to
be #include'd from pan_cmdstream.c (or panvk equivalent).

Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11785>

lavapipe: expose strict-lines feature

The strictLines-feature requires lines to be rasterized as rectangles
by default instead of using the parallelograms you get from extending
bresenham lines along their minor axis.

Now that we can specify the line mode fully we can actually express
this, so let's do so.

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11782>

lavapipe: re-expose line-rasterization extension

While we're at it, let's also enable rectangular line support, now that
we can properly control it.

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11782>

llvmpipe: respect rectangular_lines

With the new rectangular_lines state, we can now support rasterizing
wide lines correctly according to the vulkan spec, where this can be
specified independently of the rest of the state.

Because rectangular lines are orthogonal to multi-sampling, we now need
to also adjust with the pixel-offset in the rectangle code-path as well.

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11782>

draw: respect line_rectangular state

Now that we have a separate flag to signal rectangular lines, let's use
that instead of the smooth-flag.

This should have the benefit of also drawing rectangular lines when
multisampling stippled, non-smooth lines. But it seems there were no
test for that, so no results to update.

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11782>

gallium: explicitly specify line rasterization mode

Currently, drivers infer the line rasterization mode from the
multisampling and line_smooth rasterization state. This is always
correct for OpenGL, but is subtly incorrect for DirectX 9, 10 and
Vulkan (when VK_EXT_line_rasterization is supported).

So let's allow front-ends to choose freely between rectangle and
paralellogram rendering.

The reason why there's no added cap for this, is that the implicit
selection that drivers currently do will work just as well (or more
correclty, just as subtly wrong) as before. And there's nothing
reasonable the front-ends can do to get the correct behavior, so
there's really no fall-back code to write either.

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11782>

lavapipe: fix disable_multisample condition

There's two of the line-modes that warrants disabling multisampling,
and that's bresenham and smooth lines.

The reason we need this for the smooth lines case, is that multisampling
overrides the smooth-flag.

Fixes: 9fbf6b2abff ("lavapipe: implement VK_EXT_line_rasterization")
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11782>

lavapipe: do not disable multisampling for smooth lines

Smooth lines are also rasterized with rectangles instead of using
bresenham-style lines, so let's make sure we exclude those as well
from this test.

Fixes: 9fbf6b2abff ("lavapipe: implement VK_EXT_line_rasterization")
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11782>

aco/meson: remove inc_gallium from include_directories

Reviewed-by: Rhys Perry <pendingchaos02@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11258>

aco: Format.

Manually adjusted some comments for more intuitive line breaks.

Reviewed-by: Tony Wasserka <tony.wasserka@gmx.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11258>

aco: add .clang-format file

Based on src/amd/.clang-format with following changes:
Language: Cpp
Standard: c++14
PointerAlignment: Left
IndentAccessModifiers: False
SpaceBeforeCtorInitializerColon: True
AllowShortCaseLabelsOnASingleLine: True
AllowShortFunctionsOnASingleLine: InlineOnly
AllowShortBlocksOnASingleLine: Empty
AllowShortLambdasOnASingleLine: All
BraceWrapping:
  SplitEmptyFunction: false
  SplitEmptyRecord: false
  BeforeLambdaBody: true
IncludeBlocks: Regroup (with specified IncludeCategories)

Reviewed-by: Tony Wasserka <tony.wasserka@gmx.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11258>

zink: ci updates

this one no longer flakes

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11829>

zink: zero out sampler/image descriptor surface info for null descriptor updates

this is the surface struct used by the caching manager to generate a ref update template,
which means it must be kept in sync with the current state of descriptors

when the struct is zeroed, it implies that no refs need to be taken for the cached set

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11829>

ir3/print: Manual formatting fixups

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11801>

ir3: Reformat source with clang-format

Generated using:

cd src/freedreno/ir3 && clang-format -i {**,.}/*.c {**,.}/*.h -style=file

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11801>

freedreno: Add some options to .clang-format

In preparation for reformatting ir3.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11801>

ir3: Manually reformat some places

clang-format does a bad job with a few tables and macros, and there were
some places it was doing wonky things because comments were longer than
80 characters and it tries to fix that without reformatting the comment
itself. Add magic comments to tell it to turn itself off and retab those
places manually (well, with a regex!).

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11801>

ir3: Update .editorconfig and .dir-locals.el

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11801>

ir3/lower_parallelcopy: Don't manually set wrmask

It's automatically set. This avoids some weird line wrapping with
clang-format.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11801>

ir3: Add ir3_collect() for fixed-size collects

This avoids having the specify the size, and fixes weird formatting with
clang-format.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11801>

ir3: Add missing include to ir3_parser.y

This prevents build errors in the generated ir3_parser.h when
clang-format reshuffles the header includes.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11801>

vulkan/wsi/wl: add wl_shm support for lavapipe.

This adds swrast support for rendering from lavapipe.

Reviewed-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11819>

Revert "ci/panfrost: Temporarily disable sun50i/RK3288"

Back by popular demand etc

This reverts commit 9868b15164488f3a87d3204ff8788e9ce2f9c132.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11828>

kmsro: Add mali-dp

The NXP LS1028A SoC includes a mali-dp display controller and a vivante
GPU. This allows using the Etnaviv driver on this SoC.

Signed-off-by: Michael Walle <michael@walle.cc>
Reviewed-by: Christian Gmeiner <christian.gmeiner@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11419>

st/xa: Mark default xa_get_pipe_format case unreachable.

Fix defect reported by Coverity Scan.

Uninitialized scalar variable (UNINIT)
uninit_use: Using uninitialized value fdesc. Field fdesc.format is uninitialized.

Fixes: 9f2f5b3d7fd ("st/xa: Initial import of the xa state-tracker and the xa-vmwgfx target.")
Suggested-by: Adam Jackson <ajax@redhat.com>
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9035>

util/meson: include inc_gallium

There are several dependencies on headers from
   /gallium/include/pipe/

which currently mean that dependencies on util
must include gallium to compile.

Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11811>

i915g: Fix backface stencil when front_ccw is set.

If we're doing separate front/back stencil ops, then in the absence of a
hardware bit for setting front_ccw (which you can already see from how we
manage cull face) we need to flip front/back stencil register settings.

Closes: #4974, #4975
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11706>