radeonsi: pass at most 3 images and/or shader buffers via user SGPRs for compute

This should slightly decrease shader lifetime.

Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5209>

radeonsi: remove const_buffers_declared hacks

This was a bug that was uncovered by 4553fc66a5f23607c2e872d8ac8755c747bd0bd2.

Piglit: spec@arb_uniform_buffer_object@maxblocks

Fixes: 4553fc66a5f23607c2e872d8ac8755c747bd0bd2
Acked-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5209>

radeonsi: remove unused leftover code for INDIRECT_BUFFER inside IBs

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5209>

nir: gather which images are MSAA

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5209>

nir: gather which images are buffers

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5209>

nir: don't count samplers and images in interface blocks

Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5209>

ac/nir: support v2f16 derivatives

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac/nir: set the second v_cvt_pkrtz argument to undef if it's unused

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac/nir: select v_cvt_pkrtz for all conversions from f32 to f16 for radeonsi

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac/nir: handle nir_op_[fiu]2[fiu]mp opcodes

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac/nir: support 16-bit data in image opcodes

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac/nir: support 16-bit data in buffer_load_format opcodes

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac/nir: remove type and num_channels args from ac_build_buffer_store_common

They were only used for type overloading where we can just use
the type of data.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac/nir: support vector types in the type suffix of overloaded intrinsics

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac/nir: use more types from ac_llvm_context

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

ac: rename has_double_rate_fp16 -> has_packed_math_16bit

Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5003>

gallium: add shader caps INT16 and FP16_DERIVATIVES

Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

glsl: lower samplers with highp coordinates correctly

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

glsl: lower the precision of imageLoad

Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

glsl: lower mediump partial derivatives

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

glsl: lower mediump integer types to int16 and uint16

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

glsl: handle int16 and uint16 types and add instructions for mediump

v2: add more changes to ir_validate.cpp

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

glsl: treat lowp as mediump when lowering builtins

This seems to have been missed.

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

nir: add options::vectorize_vec2_16bit to limit vectorization to vec2 16

for hardware that is scalar but can do 2 16-bit operations on low and high
16 bits of registers at once.

Acked-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

nir: fix lower_wpos for 16-bit fddy

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

nir: lower int16 and uint16 in nir_lower_mediump_outputs

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

nir: add int16 and uint16 type helpers

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

nir: add i2imp and u2ump opcodes for conversions to mediump

Reviewed-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

nir: Fold f2f16(b2f32(x)) to b2f16(x)

By definition.

This reduces register pressure on freedreno so that the noubo expected
failure goes away.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Reviewed-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5002>

meson: remove "empty array"/"array of an empty string" confusion

Until Meson 0.47, setting `-D arrayoption=` was not the same as setting
`-D arrayoption=[]`; the latter cleared the array, while the former
filled it with an empty string option.

Since Meson 0.47 [1], the former maps to the latter, so empty items can
only be set by explicitly giving an array containing an empty string,
ie. `-D arrayoption="['']"`; however note that this is *not* what we
want in any of the current Mesa code anyway.

This makes the code handling array options a bit more complicated, and
a lot more error-prone, so let's get rid of the confusion by removing
the empty-string option.

[1] https://github.com/mesonbuild/meson/commit/f3a8f9c34d95d862fb4d12869a7b31cea592561f

Signed-off-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Dylan Baker <dylan@pnwbakers.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/386>

turnip: fix RENDER_COMPONENTS value

This fixes render_components being 0 when mrt_count=8, because shift by 32
is UB and in arm64 it ends up shifting by 0. This fixes tests with 8 MRTs.

Fixes the 3d path sysmem CmdClearAttachments to set RENDER_COMPONENTS, as
it was previously relying on tu6_emit_mrt setting it, but it is now part of
the pipeline state.

Also switch back to the previous behavior of not setting render components
for VK_ATTACHMENT_UNUSED attachments: we don't update the MRT state for
such attachments so we definitely don't want to be trying writing to those.

Fixes: 078aa9df8daff60e52a66d8f ("tu: Move RENDER_COMPONENTS setting to pipeline state")
Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5292>

CI: Disable Panfrost T720/T760

We're reconfiguring our Cambridge office lab networking and physical
setup for more scalability amongst other things. We can still run jobs
on one RK3399 at the peak outage, but we'll lose the T7x0 this morning,
so disable it until it's all back.

T820 is still disabled due to an unrelated BayLibre internal outage.

Signed-off-by: Daniel Stone <daniels@collabora.com>
Reviewed-by: Michel Dänzer <mdaenzer@redhat.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5296>

gitlab-ci: Use separate docker images for cross builds

Using x86_build-base as a separate base layer as well for storage &
network bandwidth efficiency.

Using separate images allows dropping the workarounds from the cross
build job scripts.

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5296>

gitlab-ci: Add x86_build-base docker image

Similar to x86_test-base.

Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5296>

gitlab-ci: Move meson back to x86_test-gl/vk ephemeral packages lists

And python3-distutils back to x86_test-gl's list.

These are not used for building Mesa, only for other components used in
test jobs.

This partially reverts commit c1a290bdd57536d6afcff6a02f1512fba7328729
"meson: Bump required version to 0.52.0".

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5296>

gitlab-ci: Stop using packages from Debian testing

Not needed anymore (for now?).

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5296>

gitlab-ci: Use Debian 10 wine-development packages

They're version 4.2, new enough for the MinGW job tests to pass.

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5296>

gitlab-ci: Move LLVM/clang 6/7 packages to the x86_build_old image

They're available in Debian 9 (stretch) as well.

This will avoid conflicts with packages from Debian testing.

Reviewed-by: Erik Faye-Lund <erik.faye-lund@collabora.com>
Reviewed-by: Eric Anholt <eric@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5296>

docs: add missing "shader_" in VK_KHR_shader_subgroup_extended_types

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5295>

vulkan-overlay/meson: use install_data instead of configure_file

We don't want to copy the file into the build directory, we want to
install it. That's what install_data is for.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2924
Fixes: 56ccea58ae7f6fd56cf4a1697d2cceb68866b552
       ("vulkan/overlay: Add basic overlay control script.")

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

meson: deprecated 'true' and 'false' in combo options for 'enabled' and 'disabled'

To prepare to use meson's builtin feature options in the future, which
are more powerful and provide useful feature for packagers, like the
ability to turn all "automagic" features off, and then explicitly turn
on the ones they want.

This is designed to make the transition softer, since the 'true' and
'false' are still accepted, just with a warning.

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

meson: use 2 space not 3 space indent

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

meson: use gnu_symbol_visibility argument

This uses a meson builtin to handle -fvisibility=hidden. This is nice
because we don't need to track which languages are used, if C++ is
suddenly added meson just does the right thing.

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

drm-shim/meson: Use portable override_options for setting C standard

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

drm-shim/meson: The name of the target is a string not a list

This happens to work, but it's not guaranteed to

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

meson: Use builtins for checking gnu __attributes__

This requires less code, and will fast skip on compilers that are known
to not have these, like MSVC.

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

meson: Use build_always_stale instead of build_always

which was deprecated in 0.47. This doesn't change behavior, just shuts
up a warning.

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

meson: Use the check_header function

Instead of open coding it. This was new in 0.47

Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

meson: Bump required version to 0.52.0

This matches what other graphics space projects require now, and allows
us to simplify a number of cases, as well as make use of new features in
meson.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/2737
Acked-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4740>

pan/mdg: Enable out-of-order execution after texture ops

We don't make great use of it (due to the scheduler not being aware
yet), but we can pack for it regardless and maybe pick up some win.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5286>

pan/mdg: Add quirk for missing out-of-order support

Added in T760, like the other good parts of Midgard.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5286>

pan/mdg: Disassemble out-of-order bits

Optimization for texture instructions, allowing ALU and LD/ST within a
single thread while a texture read is still in flight.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5286>

panfrost: Remove unused nir_lower_framebuffer pass

Superseded.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5284>

panfrost: Don't flush explicitly when mipmapping

The reorder work already takes cares of this nicely.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5284>

panfrost: Use VTX tag for vertex texturing

Fixes BARRIER faults.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5284>

panfrost: Permit AFBC of RGB8

Ugly but hey.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5284>

panfrost: Fix PRESENT flag mix-up

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5284>

pan/mdg: Fuse f2f16 into load_interpolated_input

To become a ld_vary intrinsic.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5283>

pan/mdg: Handle 16-bit ld_vary

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5283>

panfrost: Update fails list

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Use internal_format throughout

Fixes R32F_S8 texturing.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Add separate_stencil BO to batch

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Check for large tilebuffer requirements

Fixes the rest of dEQP-GLES3.functional.fragment_out.array.uint.*, this
situation occurs with MRT and large pixels.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Let Gallium pack colours

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Account for differing types in blend lower

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Conditionally allow fp16 blending

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Switch to pan_lower_framebuffer

It now supports what we need.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack sRGB via NIR

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack R11G11B10

NIR has a helper for it already; we can reuse.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack RGB10_A2_UINT

It's different. Because forget me.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack RGB10_A2_UNORM

It's a funny one.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack RGB565 and RGB5A1

Basically the same as RGBA4

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack UNORM 4

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Flesh out dispatch

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack 8-bit UNORM

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack pure 8-bit

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack pure 16-bit

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Un/pack pure 32-bit

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Stub out lowering boilerplate

Structure ourselves as a NIR pass replacing loads/stores with
unpacked/packed versions as necessary. Not actually functional yet.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Determine classes for stores

Fewer special cases here, thankfully.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Determine load classes for formats

Via quirks.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Add quirks for blend shader types

Every hardware has its own set of what it can and can't do... let's
document it all as quirks so the lowering code is GPU-agnostic.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Determine unpacked type for formats

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

panfrost: Add theory for new framebuffer lowering

We take a somewhat different strategy that should be more flexible.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

pan/mdg: Implement raw colourbuf loads on T720

Uses a similar path to the fp16 cbuf loads on T760. It should make sense
given the symmetry with T860.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

pan/mdg: Drop the u8 from the colorbuf op names

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

pan/mdg: Print 8-bit constants

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

pan/mdg: Handle bitsize for packs

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

pan/mdg: Treat packs "specially"

We maybe would prefer synthetic ops? We'll find out in due time..

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

pan/mdg: Add pack_unorm_4x8 via 8-bit

More efficient than the 32-bit version in NIR.

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

pan/mdg: Handle un/pack opcodes as moves

Signed-off-by: Alyssa Rosenzweig <alyssa.rosenzweig@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5265>

iris: Fixup copy'n'paste mistake in Makefile.sources

In changing iris_seqno.[ch] to iris_fine_fence.[ch] and moving the
lines earlier, the newline escape was forgotten.

Fixes: 034329128b703f8c2e8a ("iris: Rename iris_seqno to iris_fine_fence")
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5264>

intel/dev: Don't consider all TGL SKUs as GT1 only

We should be passing _gt instead of 1 to GEN12_FEATURES or else all TGL
SKUs will be considered as gt1 only.

Fixes: 54996ad4927 ("intel/dev: Split .num_subslices out of GEN12_FEATURES macro")
Signed-off-by: Satyeshwar Singh <satyeshwar.singh@intel.com>
Reviewed-by: Eric Engestrom <eric@engestrom.ch>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5261>

r300g: Remove extra printf format specifiers.

Fix warning reported by Coverity Scan.
Missing argument to printf format specifier (PRINTF_ARGS)
missing_argument: No argument for format specifier %s.

Fixes: 04c1536bf7ab ("r300g: rasterizer debug logging")
Fixes: 85efb2fff0d4 ("r300g: try to use color varyings for texcoords if max texcoord limit is exceeded")
Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5274>

nouveau: allow invalidating coherent/persistent buffer backings

This is needed to support the core's usage of coherent buffers for
glVertex-style input. The reason why this was disallowed is that any
mappings will be invalidated. Let the state tracker worry about that,
and just reallocate when we're told.

Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Cc: mesa-stable@lists.freedesktop.org
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5276>

intel/fs: Emit HALT for discard on Gen4-5

Using HALT to immediately jump to the end of the shader is required to
implement GL_EXT_gpu_shader4 and OpenGL 3.0.  However, vanilla OpenGL
1.2 doesn't forbid it and it likely makes something somewhere faster.
We should be consistent and implement the same discard behavior on all
hardware if we can.

The rules for HALT on Gen4-5 are a bit different from Gen6+.  On the
older hardware, there is no stack for HALT; instead it's up to software
to save and restore mask registers.  However, there's no real saving
needed since we only use HALT to jump to the end of the program where
we're about about to do our FB writes.  All we need to do is reset AMask
to DMask, the value it was initialized to at the start of the thread.

Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5244>

intel/fs: Fix unused texture coordinate zeroing on Gen4-5

We were inserting the right number of MOVs but, thanks to the way we
advanced msg_end earlier in the function, were often writing the zeros
past the end of where we actually read in the register file.

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5243>

intel/vec4: Stomp the return type of RESINFO to UINT32

We already do this in the FS back-end; we just weren't doing it in vec4
so RESINFO messages weren't returning the right data.

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5243>

radv: fix regression with builtin cache

If the ~/.cache dir already exists continue on without failing.

Fixes: cd61f5234d2c ("radv: Handle failing to create .cache dir.")
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5249>

gallium/dri: Remove lowered_yuv tracking for plane mapping.

Just heard that etnaviv is also compatible with it even
in the non-lowered cases, so let us enable it for everyone.

Reviewed-by: Lucas Stach <l.stach@pengutronix.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/5253>