lima: Fix glFrontFace handling

Bit 12 of render->aux1 is GL_CCW/GL_CW. For GL_CCW (default of glFrontFace) we have
to set that bit active.

This is not what the blob does and what the original reverse engineering documentation
says. The blob sets this value inverted and does some bogus negation of the fragment
shaders gl_FrontFacing variable instead.

Anyway, doing it this way does not cause regressions but fixes
dEQP-GLES2.functional.shaders.builtin_variable.frontfacing and 4 piglit tests.

Reviewed-by: Vasily Khoruzhick <anarsoul@gmail.com>
Signed-off-by: Andreas Baierl <ichgeh@imkreisrum.de>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7690>

lavapipe: move to new shared features/properties

Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12953>

lavapipe: enable vulkan 1.2 support.

The remaining extensions are optional features, just turn on vk 1.2
with them reporting as off.

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12953>

gallivm/nir: fix subgroup invocation read.

Again if you get passed an invoc but the exec mask has the
active lane somewhere other than at 0, then if we have an
invoc we should find the active lane and extract the value
from invoc rather than using the idx.

This fixes a bunch of VK 1.2 subgroup tests once 1.2 is enabled:
dEQP-VK.subgroups.ballot_broadcast.compute.subgroupbroadcast_nonconst*

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12953>

lavapipe: move to 1.2 features/properties structs.

This just adds all the wrappers in the right places hopefully

Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12953>

lavapipe: add support for KHR_shader_float_controls

These pass all the CTS tests, though not sure how useful they are.

[airlied: these may need some work in the future depending on app expectations]

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12953>

features: mark off bindless texture for zink

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12855>

zink: export PIPE_CAP_BINDLESS_TEXTURE

this is the 6th descriptor set being bound, so don't even advertise it
if 6 sets can't be bound

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12855>

zink: implement bindless textures

this works by tracking 1024-member arrays of images and textures using idalloc
for indexing. each idalloc id is an index into the array that is returned as a handle,
and this handle is used to index into the array in shaders.

in the driver, VK_EXT_descriptor_indexing features are used to enable updates on the live
bindless descriptor set and leave unused members of the arrays unbound, which works as
long as no member is updated while it is in use. to avoid this, idalloc ids must cycle through
a batch once the image/texture handle is destroyed before being returned to the available pool

in shaders, bindless ops come in one of two types:
- i/o variables
- bindless instructions

for i/o, the image/texture variables have to be rewritten back to the integer
handles which represent them so that the successive shader stage utilizing them
can perform the indexing

for instructions, the src representing the image/texture has to be rewritten as a deref
into the bindless image/texture array

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12855>

zink: hook up VK_EXT_descriptor_indexing

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12855>

zink: handle bindless images and samplers in ntv

these are going to come through as direct variable derefs, so it's simple
to handle the functionality by reusing the same codepath to generate image
types

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12855>

zink: refactor some shader image code to make it reusable

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12855>

zink: don't add resource to pending barrier set if no barrier will be generated

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12855>

zink: split out bvci creation from object creation

make this reusable like the others

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12855>

zink: force lazy descriptor set rebinds if pipeline compatibility changes

if the pipeline layout changes, it's technically illegal to not rebind the descriptor
set even if it hasn't changed

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12931>

zink: remove duplicated struct member set

this should only happen in zink_descriptors_update_lazy_masked

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12931>

ac/surface: correct a comment about DCC image stores

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13013>

ac/surface: use DCC compatible with image stores for < 4K resolutions

We don't have to use the special DCC settings for lower resolutions.

This will cause corruption if X and an windowed app use different Mesa
versions. The fix is to restart the X server. I expect to get false bug
reports due to this.

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13013>

radeonsi: remove the unused cs parameter from remaining packet functions

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13015>

radeonsi: remove the unused cs parameter from radeon_set_uconfig_reg

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13015>

radeonsi: remove the unused cs parameter from radeon_set_sh_reg

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13015>

radeonsi: remove the unused cs parameter from radeon_set_(config|context)_reg

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13015>

radeonsi: remove the unused cs parameter from radeon_emit_array

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13015>

radeonsi: remove the unused cs parameter from radeon_emit

Reviewed-by: Pierre-Eric Pelloux-Prayer <pierre-eric.pelloux-prayer@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13015>

docs: mark off GL_EXT_multisampled_render_to_texture for zink

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: implement GL_EXT_multisampled_render_to_texture

this requires allocating and using a lazily-allocated msaa surface as a transient
attachment for the base render operation, resolving it

...except vulkan has no "replicate" renderpass attachment mechanism, so for now we're
just gonna allocate a transient surface and hang on to it. sorry tilers!

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: stop setting nr_samples for shader image surface creation

that's not what this should be used for

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: move fb attachment init to new function

this is going to get more complicated

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: remove fbfetch layout thingy from zs renderpass init

zs doesn't do fbfetch

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: reorder draw state updates

starting a renderpass can trigger recursion and other draws, so do all
state updates/binds after the renderpass has begun

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: improve handling of buffer rebinds using tc info

tc rebind info can't distinguish between buffers bound multiple times
in a given bind point (e.g., for multiple vertex buffers), so ensure all
rebinds are processed and the global rebind isn't triggered if multiple rebinds
are processed for a given buffer

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: add ZINK_BIND_TRANSIENT

this is used to create lazily allocated images for use as transient
renderpass attachments

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: add ZINK_HEAP_DEVICE_LOCAL_LAZY

this is for lazily allocated resources

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

zink: repack zink_render_pass_state

saves a couple wasted bits

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12934>

spirv: don't fail on CapabilitySubgroupDispatch if supported

Since only Anv uses the value, I'm only enabling this on anv.

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Fixes: 518693c3eca195 ("spirv: Handle the SubgroupSize execution mode")
Reviewed-by: Caio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13034>

zink: split vk debug logging into separate functions

being able to breakpoint these easily is crucial to debugging, and it sucks to
always have to go back and check the line numbers, so now there's easily-rememberable
functions for it

Reviewed-by: Hoe Hao Cheng <haochengho12907@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13042>

zink/codegen: do not enable extensions based on vulkan version

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13004>

turnip: Swizzle in 0, 1 for D24S8 STENCIL_ASPECT sampling.

This seems to be what is expected by the CTS, but I haven't found a
citation.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13009>

zink: enable timeline ext features

this is required by spec even though nobody cares

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13033>

nir/opt_if: add opt_if_rewrite_uniform_uses

Turns:
if (a == (b=readfirstlane(a)))
   use(a)
into:
if (a == (b=readfirstlane(a)))
   use(b)

Improves divergence analysis and lets us scalarize use(a). Improves
Cyberpunk 2077 performance.

fossil-db (Sienna Cichlid, Cyberpunk 2077):
Totals from 57 (10.56% of 540) affected shaders:
VGPRs: 4904 -> 4040 (-17.62%)
CodeSize: 624360 -> 626828 (+0.40%); split: -0.06%, +0.46%
MaxWaves: 656 -> 824 (+25.61%)
Instrs: 119770 -> 119447 (-0.27%); split: -0.49%, +0.22%
Latency: 1950256 -> 1633110 (-16.26%); split: -16.26%, +0.00%
InvThroughput: 364852 -> 292089 (-19.94%)
VClause: 1512 -> 1008 (-33.33%)
SClause: 2693 -> 3196 (+18.68%)
Copies: 10050 -> 9955 (-0.95%); split: -3.34%, +2.40%
Branches: 3476 -> 3547 (+2.04%)
PreSGPRs: 4003 -> 5076 (+26.80%)
PreVGPRs: 4709 -> 3810 (-19.09%)

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12472>

nir: add nir_src_components_read()

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12472>

radv: Reduce NGG culling small draw threshold to 128.

Many modern games use draw calls with only a few vertices.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12859>

radv: Select PC oversubscription rate based on number of PS params.

Decide the oversubscription rate similarly to how RadeonSI does it:
Oversubscribe a smaller amount of PC (parameter cache) when there are
fewer params.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12859>

turnip: Set the VK_DRIVER_ID to our new enum.

This hasn't been exposed yet, but would be with vulkan 1.2 or
VK_KHR_driver_properties

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13007>

vulkan: Update the XML and headers to 1.2.193

Acked-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13007>

ac/nir/nggc: Don't reuse uniform values from divergent control flow.

With NGG culling, the shaders are split into two parts:
the top part that computes just the position output,
and the bottom part which produces the other outputs.

To reduce redundancy between the two, I added some code
to reuse uniform variables between them. However, there is
an edge case I didn't think about: because of vertex repacking,
it is possible for the bottom part to process a different vertex.
Therefore it can take a different divergent code path
(though it must still take the same uniform code path).
Due to this, when a uniform value comes from divergent control
flow, this may be undefined in the bottom part.

This commit stops reusing uniform variables from
divergent control flow, to fix issues that arise from this.

Fossil DB stats on Sienna Cichlid with NGGC on:

Totals from 1723 (1.34% of 128647) affected shaders:
VGPRs: 89312 -> 89184 (-0.14%); split: -0.15%, +0.01%
SpillSGPRs: 4575 -> 120 (-97.38%)
CodeSize: 10846424 -> 10873836 (+0.25%); split: -0.68%, +0.93%
MaxWaves: 34582 -> 34602 (+0.06%); split: +0.06%, -0.01%
Instrs: 2124471 -> 2128835 (+0.21%); split: -0.51%, +0.72%
Latency: 7274569 -> 7293899 (+0.27%); split: -0.22%, +0.48%
InvThroughput: 1637130 -> 1635490 (-0.10%); split: -0.17%, +0.07%
VClause: 25141 -> 25414 (+1.09%); split: -0.02%, +1.10%
SClause: 56367 -> 59503 (+5.56%); split: -1.36%, +6.93%
Copies: 230704 -> 219313 (-4.94%); split: -5.49%, +0.55%
Branches: 72781 -> 72681 (-0.14%); split: -0.21%, +0.07%
PreSGPRs: 118766 -> 100176 (-15.65%); split: -15.70%, +0.05%
PreVGPRs: 76876 -> 76833 (-0.06%)

Fixes: 0bb543bb60f93bea5b1c0ed6382ced49e659273e
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13001>

ac/nir/nggc: Refactor save_reusable_variables.

This makes the code more elegant and also fixes the mistake of
skipping the blocks that come before loops.

Fossil DB changes on Sienna Cichlid with NGGC on:

Totals from 1026 (0.80% of 128647) affected shaders:
SpillSGPRs: 3817 -> 4035 (+5.71%)
CodeSize: 5582856 -> 5538732 (-0.79%); split: -0.89%, +0.10%
Instrs: 1106907 -> 1100180 (-0.61%); split: -0.68%, +0.07%
Latency: 10084948 -> 10052197 (-0.32%); split: -0.37%, +0.05%
InvThroughput: 1567012 -> 1564949 (-0.13%); split: -0.16%, +0.03%
SClause: 39789 -> 39075 (-1.79%); split: -2.33%, +0.54%
Copies: 95184 -> 96456 (+1.34%); split: -0.19%, +1.53%
Branches: 44087 -> 44093 (+0.01%); split: -0.01%, +0.02%
PreSGPRs: 47584 -> 51009 (+7.20%); split: -0.61%, +7.80%

Fixes: 0bb543bb60f93bea5b1c0ed6382ced49e659273e
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Daniel Schürmann <daniel@schuermann.dev>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13001>

radv/llvm: stop using vs_common_out.export_prim_id

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13028>

radv,aco: stop using vs_common_out.export_clip_dists

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13028>

radv: add export_clip_dists for VS and TES to radv_shader_info

To be able to remove radv_vs_out_key entirely.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13028>

radv: Fix gs_vgpr_comp_cnt for NGG culling in vertex shaders.

Previously these shaders always took the path to gs_vgpr_comp_cnt=3,
but now they are 0 when they don't use primitive id.

Fixes: 7ad69e2f7ee10c0e7afc302b9324e7a320424dcb
Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13031>

nir/lower_io_to_vector: Allow Task/Mesh to load from outputs

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12951>

radv/llvm: stop using vs_common_out.as_ngg_passthrough

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13025>

radv: remove useless as_ngg_passthrough init when lowering NGG in NIR

Nothing reads the value after that, except LLVM but we don't lower it
in NIR (yet).

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13025>

clc: use the defined version for the parser

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13018>

microsoft/clc: small compile fix on Linux

Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13018>

radv: stop using the shader keys for as_ls/as_es/as_ngg when possible

radv_shader_info should be initialized correctly at this point.
There is still few occurences earlier in the pipeline but they aren't
easy to remove right now.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13023>

panfrost: Fix incorrect test condition

Oh, the irony. I linked to this file in a software design class as an
example of how unit testing works under the hood, since it's not using
any particular testing framework. In doing so I realized the test is
totally broken. It doesn't matter -- the tests passed before, pass now,
and the code being tested is correct. But let's still fix the
embarassing logic error for future Panfrost hackers.

Fixes: d42e53c77a2 ("panfrost: Unit test clear colour packing")
Signed-off-by: Alyssa Rosenzweig <alyssa@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13006>

aco/tests: add idep_amdgfxregs_h

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Fixes: 9bf30c4a5cb ("aco/tests: add tests for form_hard_clauses()")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12017>

radv: remove unused radv_tcs_variant_key:primitive_mode

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13022>

broadcom/compiler: remove qpu_acc helper

It is really small, and used just twice, so we just call qpu_magic.

We also update how it is used:

  * QFILE_NULL is an undef so we can just load anything. Previously we
    were using accumulator 0, but there isn't any real reason to use
    an accumulator for this. Using reg 0.

  * QFILE_LOAD_IMM: it seems that we don't use at all right now, so
    let's add an assert

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13008>

broadcom/compiler: remove commented out vir_LOAD_IMM methods

It has been commented several years now. Let's remove it to reduce the
noise.

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13008>

radv: stop loading invocation ID for NGG vertex shaders

This is unnecessary and was copied from RadeonSI which needs it for
edge flags. RADV should never to read the invocation ID for VS.

Signed-off-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12997>

radv: Remove dead min waves code.

Signed-off-by: Georg Lehmann <dadschoorse@gmail.com>
Reviewed-by: Bas Nieuwenhuizen <bas@basnieuwenhuizen.nl>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13005>

lavapipe/ci: drop some fails I fixed recently

Acked-by: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13011>

anv: Switch to new debug message helpers

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10318>

vulkan: Add convenience debug message helpers

Forwards the message to both VK_EXT_debug_utils and
VK_EXT_debug_report provided callbacks.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10318>

vulkan: Add vk_asprintf and vk_vasprintf helpers

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10318>

vulkan/enum_to_str: Add generator for VkObjectType to Vulkan Handle

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10318>

vulkan: Implement VK_EXT_debug_utils

This implements all the necessary features of VK_EXT_debug_utils in
common code.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10318>

gallium/format: move two vertex formats into the proper place.

These two can be vertex formats with vulkan so lavapipe can
expose them.

Fixes
dEQP-VK.pipeline.vertex_input.single_attribute.uvec4.as_a2r10g10b10_uint_pack32_rate_vertex

Fixes: e002f5a086c7 ("gallium: change pipe_vertex_element::src_format to uint8_t")

Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12987>

radv: Use the shared now-in-core feature/prop extension helper functions.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12967>

anv: Use the shared now-in-core feature/prop extension helper functions.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12967>

turnip: Use the shared now-in-core feature/prop extension helper functions.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12967>

vulkan: Support PHYSICAL_DEVICE_1_n_ features/properties in the helpers.

Since we have a filled out struct available, just memcpy it over so the
driver doesn't have to fill it out again.

Suggested by Jason, this is code that isn't copy-and-paste from anv.

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12967>

vulkan: Add helpers for filling exts for core features and properties.

This is boilerplate that drivers have to replicate, so make a nice helper.
The feature/property macro lists are are copy and pasted from anv, though
the helper functions has been split between 1.1 and 1.2 (and thus the macros
refactored).

Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12967>

lavapipe: Use a common vk_queue structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

v3dv: Use a common vk_queue structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

turnip: Use a common vk_queue structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Reviewed-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

radv: Use a common vk_queue structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

anv: Use a common vk_queue structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

vulkan: Add a common vk_queue structure

This can be used e.g. for storing debug labels for the common
implementation of VK_EXT_debug_utils.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

lavapipe: Use a common vk_command_buffer structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-By: Mike Blumenkrantz <michael.blumenkrantz@gmail.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

v3dv: Use a common vk_command_buffer structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

turnip: Use a common vk_command_buffer structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

radv: Use a common vk_command_buffer structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

anv: Use a common vk_command_buffer structure

Switch to using common structure.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

vulkan: Add a common vk_command_buffer structure

This can be used e.g. for storing debug labels for the common
implementation of VK_EXT_debug_utils.

Signed-off-by: Yevhenii Kolesnikov <yevhenii.kolesnikov@globallogic.com>
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/13000>

radv/llvm: Don't read edge flags anymore.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12998>

ac/nir: Emit edge flag instructions conditionally.

They are not needed by RADV but will be needed by RadeonSI.

Fossil DB results on Sienna Cichlid (with NGGC on):
Totals from 56917 (44.24% of 128647) affected shaders:
VGPRs: 1982664 -> 1975936 (-0.34%); split: -0.43%, +0.09%
CodeSize: 152790880 -> 149510316 (-2.15%); split: -2.15%, +0.00%
MaxWaves: 1617984 -> 1621900 (+0.24%)
Instrs: 29272825 -> 28907038 (-1.25%); split: -1.26%, +0.01%
Latency: 128744182 -> 127565678 (-0.92%); split: -1.14%, +0.22%
InvThroughput: 20125915 -> 19805168 (-1.59%); split: -1.63%, +0.03%
VClause: 521312 -> 519804 (-0.29%); split: -0.77%, +0.48%
SClause: 688861 -> 688897 (+0.01%); split: -0.04%, +0.05%
Copies: 3205421 -> 3177799 (-0.86%); split: -1.68%, +0.82%
Branches: 1181457 -> 1183147 (+0.14%); split: -0.03%, +0.17%
PreVGPRs: 1626681 -> 1595406 (-1.92%)

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12998>

radv: Disable HW generated edge flags for NGG shaders.

Vulkan doesn't support user edge flags, and also doesn't have
any decomposed primitives where drawing internal edges should
be disallowed. Hence, we don't need this.

Signed-off-by: Timur Kristóf <timur.kristof@gmail.com>
Reviewed-by: Samuel Pitoiset <samuel.pitoiset@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12998>

microsoft/clc: Add a test for specializing via SPIRV-Tools

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>

microsoft/clc: Add API to independently specialize SPIR-V

We need the ability to specialize unlinked SPIR-V, so use SPIR-V tools
to specialize prior to linking.

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>

microsoft/clc: Support passing specialization consts to spirv_to_nir

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>

microsoft/clc: Parse SPIR-V specialization consts into metadata

We need to be able to validate at the API that set specialization consts
have a valid ID and the value is the correct size.

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>

microsoft/clc: Support SPIR intermediates in the compilation APIs

We can now export SPIR (mainly just for testing) or import SPIR and
convert to SPIR-V.

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>

microsoft/clc: Split clc_object and rename entrypoints

clc_object was overloaded, containing SPIR-V binary and metadata,
and it would only sometimes contain metadata (after linking). Split
it into a more generic clc_binary class which holds some type of data
(the kind depends on where it came from), and clc_metadata which can
be independently parsed on compiled or linked data.

Rename a couple entrypoints to be more explicit about what they're
actually transforming (c_to_spirv, link_spirv, spirv_to_dxil).

Add a logger to SPIR-V binary parsing so it can report errors on app-
provided SPIR-V.

Re-order helper function parameters to be more consistent (out params last).

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>

microsoft/clc: Stop heap-allocating tiny fixed-size transparent structs

The caller can allocate these however they want. They don't need
independent allocations. Removes some unnecessary failure handling.

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>

microsoft/clc: Clean up clc_context

1. Rename it to libclc to match what it is
2. Make it opaque externally
3. Remove it from non-essential entrypoints (compile/link)

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>

microsoft/clc: Rename compiler DLL to clon12compiler

I no longer think there's value in trying to share code on disk here between
CL and GL at runtime, especially since we're now just building GL without LLVM

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/10322>