freedreno/ir3: Improve error msg for block level validation

Printing whatever happened to be the last instruction is misleading.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12742>

freedreno/ir3: Validate physical successors

Each block's physical successors should be a superset of the logical
successors.

Also validate that the successors are sane (ie. we shouldn't have the
2nd one if we don't have the first)

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12742>

freedreno/ir3: Fix double printing of branch suffix

print_instr_name() already prints the suffix.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12742>

freedreno/ir3: Fix physical successors for break out of loop

We can't just clobber the existing successor[0] which is based on
logical successor[0]

Suggested-by: Connor Abbott <cwabbott0@gmail.com>
Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12742>

freedreno/ir3: Update physical_successors after retargetting jumps

Signed-off-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12742>

glsl: break out early if compound assignment's operand errored out

Fixes compiler crashes on:

struct Foo
{
  float does_exist_member;
};

in vec2 tex;
out vec4 color;

void
main(void)
{
  Foo foo;

  foo.does_not_exist_member %= 3; /* or any of: <<=, >>=, &=, |=, ^= */
  color = vec4(tex.xy, tex.xy);
}

Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
CC: mesa-stable
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12717>

glsl: propagate errors from *=, /=, +=, -= operators

Fixes compiler crash on:

void main()
{
    gl_FragColor = a += 1;
}

(a is not declared anywhere)

Found with AFL++.

Fixes: d1fa69ed61d ("glsl: do not attempt assignment if operand type not parsed correctly")
Signed-off-by: Marcin Ślusarz <marcin.slusarz@intel.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Reviewed-by: Tapani Pälli <tapani.palli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12717>

zink: fix some pipe caps for max instructions

all shaders are supported now

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12760>

zink: use dispatch table for (almost) all vulkan calls

try to avoid hitting the loader when possible

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12757>

lima: Improve error messages for unsupported GP operations

Reviewed-by: Erico Nunes <nunes.erico@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12738>

lima: Fix crashes for GPUs with more than four cores

Reviewed-by: Erico Nunes <nunes.erico@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12738>

lima: Enable PIPE_CAP_VERTEX_COLOR_UNCLAMPED

Fixes lighting being too bright in Neverball.

Reviewed-by: Erico Nunes <nunes.erico@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12738>

glsl: fix variable scope for do-while loops

Without this the following code was successfully compiling.

   void main()
   {
      do
         int var_1 = 0;
      while (false);

      var_1++;
   }

Fixes: a87ac255cf7e ("Initial commit.  lol")

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12465>

glsl: handle scope correctly when inlining loop expression

We need to clone the previously evaluated loop expression when
inlining otherwise we will have conflicts with shadow variables
defined in deeper scopes.

Fixes: 5c02e2e2de75 ("glsl: Generate IR for switch statements")

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5255

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12465>

glsl: fix variable scope for loop-expression

We need to evaluate the loop-expression of a for loop before
we evaluate the loop body otherwise we will find the wrong
variable for the following loop.

   int var_0 = 0;
   for(; var_0 < 10; var_0++) {
      const float var_0 = 1.0;
      gl_FragColor = vec4(0.0, var_0, 0.0, 0.0);
   }

Fixes: a87ac255cf7e ("Initial commit.  lol")

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5262

Reviewed-by: Ian Romanick <ian.d.romanick@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12465>

mesa: fix mesa_problem() call in _mesa_program_state_flags()

Previously STATE_INTERNAL returned 0 for the unhandled states
but the outer switch throws the error for unhandled states.

Fixes: b4f3497786ef ("mesa: remove STATE_INTERNAL")

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5316

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12746>

zink: use EXT_primitive_topology_list_restart where available

this also fixes an error where PATCHES was reported as supported

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12626>

zink: hook up VK_EXT_primitive_topology_list_restart

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12626>

lavapipe: fix primitive restart with indexed indirect draws

restart index needs to be set

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12761>

util/primconvert: force restart rewrites if original primtype wasn't supported

if we're doing a primtype change, the restart needs to be eliminated to
avoid losing restart data after the rewrite

Fixes: 583070748cb ("util/primconvert: handle rewriting of prim-restart draws with unsupported primtype")

Acked-by: Rob Clark <robdclark@chromium.org>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12625>

microsoft/spirv_to_dxil: Add `install : true` to spirv_to_dxil library.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12758>

ci: Add osmesa to Windows GitLab CI

osmesa CI coverage on Windows was lost when Windows CI with
Meson build was migrated from AppVeyor to GitLab CI

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8672>

main: don't always clamp pixels read from snorm buffers

From the OpenGL 4.6 spec, section 18.2.8:

"For a signed normalized fixed-point color buffer, each component is
clamped to [0,1] if read color clamping is enabled, or if type
represents un- signed integer components; otherwise type represents
signed integer components, and each component is clamped to [−1,1]."

Signed-off-by: Italo Nicola <italonicola@collabora.com>
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12749>

zink: clamp query results to 500 per qbo on 32bit

this makes 4000byte qbos instead of 40000, which avoids bypassing slab
allocation and makes the buffers more reusable to avoid exploding 32bit
address space

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12752>

llvmpipe: memcpy user_buffers at set_constant_buffer time.

The data in the user buffer is only valid for a short period of time, and
we could use-after-free it if rendering hadn't been flushed by shader
deletion time.

Fixes: #5254
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12724>

llvmpipe: Fix leak of CS local memory with 0 threads.

Fixes: bde08ce4d7bf ("llvmpipe: handle compute shader launch with 0 threads")
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12724>

llvmpipe: Free CS shader images on context destroy.

Cc: mesa-stable
Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12724>

iris: signal the syncobj after a failed batch

When a batch fails we either recreate our context (in case we got -EIO
or -ENOMEM) or we abort() (every other error). If we don't abort and a
later batch has a dependency on the batch that failed, then this newer
batch will fail with -EINVAL since it requires a syncobj that was
never submitted, which means we'll abort().

To avoid this problem, in this patch we simply signal syncobjs of
failed batches. This means we may be breaking our dependency tracking,
but IMHO it's better than simply letting it abort() later.

In other words, this moves the situation for some apps from "app
causes a GPU hang and then aborts" to "app causes a GPU hang but keeps
running".

Note: on some older Kernels (like today's Debian 5.10 Kernel) I see X
simply freezing after the GPU hang when the app doesn't decide to
abort(). Switching to a more recent Kernel fixes this issue for me, so
in case it happens to you make sure you have the most recent stable
trees.

v2:
  - Fix coding style (Ken).
  - Use the big comment block provided by Ken (Ken).
  - Adjust the commit message so avoid saying we retry (Ken).
  - Rebase after the syncobj ownership changes.
  - Drive-by add a missing white space in the header.

Acked-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Kenneth Graunke <kenneth@whitecape.org>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12657>

ci/freedreno: Skip slow SizedDeclarationsPrimitive in CI.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12753>

ci/freedreno: Mark a630 basic-glsl-misc-fs as flaky.

My branch mostly fixed it, but apparently it's not stable yet.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12753>

ci/freedreno: Fix typo in glx-tfp flake annotation.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12753>

freedreno/a6xx: Emit a WFI after event writes flushing CCU.

Following on from !12475, do the same CCU workaround for freedreno as
turnip is doing, so that we flush CCU correctly for when color/depth is
next read under a different cache domain from where it was written.

Fixes instability (particularly if the GPU is clocked up) on
dEQP-GLES3.functional.fbo.blit.depth_stencil.depth32f_stencil8* in bypass
mode on a630.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/11481>

ac,radv: remove unused inputs array and VS input code

The previous commit stopped using "inputs".

"load_layer_id" has always been broken and it was probably unused anyway.

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12570>

ac,radeonsi: load VS inputs at the call site of nir_intrinsic_load_input

to match ACO

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12570>

ac/llvm: implement nir_intrinsic_elect

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12570>

ac/llvm: implement nir_intrinsic_overwrite_*_arguments_amd

This should work if the intrinsics are not called from conditional blocks.

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12570>

ac: make ac_shader_abi::inputs an array instead of a pointer

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12570>

ac: remove needless parameters from ac_shader_abi::emit_outputs

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12570>

ac/llvm: implement a bunch of NIR AMD intrinsics for NGG

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12570>

ac/llvm: remove load_tess_coord callback

Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12570>

ci/iris: Switch GL/GLES testing to suites.

Better board utilization, less fussy load-balancing of the boards, and
being able to add KHR-GLES* and multisample testing without having to add
more reboots.

Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12437>

ci/iris: Consistently use .test-manual-mr for our unstable hardware.

... and place it in the board's definition, instead of the consuming jobs.

GLK and APL are the boards that are enabled by default so far, and the
rest are off.  But, if you want to check on some other board, you should
be able to click the button.

Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12437>

ci/iris: Unmark dma_buf_import_export tests as failing.

We now get the binaries built in our rootfs.

Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12437>

ci/iris: Mark create_context-no_error as failing.

Consistent in CI, and other drivers fail at it too.

Reviewed-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Acked-by: Daniel Stone <daniels@collabora.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12437>

ir3/lower_pcopy: Set entry->done in the swap loop

We process all the remaining copy entries one-at-a-time when emitting
swaps, so normally we wouldn't need to care about whether the entry is
done, except that we need to look at the other entries to split them up
and we need to skip this when the copy to be split has already been
processed.

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5307
Fixes: 0ffcb19b9d9 ("ir3: Rewrite register allocation")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12747>

ir3/lower_pcopy: Use right flags for src const/immed

At some point I split up the flags into overall/source flags and made
copies from immed/const only set IR3_REG_IMMED/IR3_REG_CONST on the
source flags, but I forgot to update this. Noticed by inspection.

Fixes: 0ffcb19b9d9 ("ir3: Rewrite register allocation")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12747>

llvmpipe: Don't wait for already-terminated threads on Windows

In the case of an app returning from main(), Windows will apparently terminate
other threads before invoking final cleanup on the main thread.

llvmpipe can't wait for threads to signal a semaphore if the thread is already
gone. Since we're already in a WIN32 special case, just call the Win32 API
to check if the thread is terminated or STILL_ALIVE.

Reviewed-by: Jose Fonseca <jfonseca@vmware.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12731>

v3d: Enable PIPE_CAP_TEXTURE_MIRROR_CLAMP_TO_EDGE

Reviewed-by: Juan A. Suarez <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12744>

winsys/radeonsi: Set vce_encode = true when VCE found

Allows VA-API encoding on AMD PITCAIRN

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5305
Fixes: 43c04ab2b43 ("radeonsi: separate video hw info based on HW engine individually")

Reviewed-by: James Zhu <James.Zhu@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12679>

radeon/vcn: Add FW header flag to enable VP9 header parsing

set flags to enable VP9 uncompressed header parsing in FW and extract
quantization and loop filter parameters.

Signed-off-by: Veerabadhran Gopalakrishnan <veerabadhran.gopalakrishnan@amd.com>
Reviewed-by: Boyuan Zhang <boyuan.zhang@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12703>

zink: fix PIPE_SHADER_CAP_FP16_DERIVATIVES handling

this requires 16bit io, not just alus

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12634>

zink: handle bo struct types that are just a runtime array

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12634>

zink: handle 8/16bit ssbo storage

this is a bit gross, but basically just add an array of extra spvids
so that each bitsize can have its own variables to keep the types in sync

glsl can't do this, but (future) internal mesa shaders can

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12634>

zink: implement nir_op_pack_half_2x16_split

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12634>

zink: lower 32_2x16_split pack/unpack instructions

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12634>

zink: hook up 8/16bit storage exts

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12634>

zink: add 8bit alu handling

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12634>

zink: fix PIPE_CAP_DRAW_PARAMETERS export

before 1.2, assume this is available if the extension is there, otherwise
require the feature

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12710>

zink: always emit sample id 0 for non-msaa texel pointers in ntv

in practice this shouldn't matter since the sample id must always be zero,
but it's still a spec violation

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12710>

zink: move drawid_broken to unhashed pipeline state

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12660>

zink: move point sprite rasterizer bits to unhashed pipeline state

avoid reading values that aren't part of the pipeline state

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12660>

zink: add pipeline state flag for determining if output type is points

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12660>

zink: move sample part of fs key to renderpass

this is more accurate

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12660>

zink: remove tcs shader keys

these never did anything and need a rework to become useful, assuming
I even find a case where such a thing would be needed

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12660>

zink: hash blend state pointers on creation

Reviewed-by: Dave Airlie <airlied@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12660>

turnip: re-emit vertex params after they are invalidated

Constants could be invalidated via HLSQ_INVALIDATE_CMD which is
emitted when new pipeline is bound and in CmdClearAttachments.
Also they become invalid after secondary cmd execution.

Fixes geometry flickering in Genshin Impact
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5313

Fixes: 815a85dd7c44f4be4cbf5b351836c921a7a923fd "turnip: do not re-emit same vs params"

Signed-off-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12684>

freedreno/registers: add a6xx media formats

Define hardware formats which correspond to media formats.

DPU/CAMSS/VENUS/CDSP can consume/produce buffers in these formats with UBWC

Signed-off-by: Jonathan Marek <jonathan@marek.ca>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12740>

i915g: Reject non-unrolled loops or non-flattend IFs at link time.

i915 has no support for control flow, so we need to link-fail shaders that
fail to unrooll loops.  This is valid for GLES2, and non-conformant but
the best we can do for GL2.

Note that we still have some dEQP failures where loops that should be
unrollable and thus are required to be supported don't get unrolled
(#4979).

Closes: #4978
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12218>

mesa/st: Add an optional GLSL link fail msg to finalize_nir.

GLES2 drivers are allowed to reject some GLSL constructs, like dynamic
loop bounds (which neither i915g nor vc4 can fully support), but gallium
hasn't had any way to trigger a link failure.  Add a return msg to the
finalize_nir hook, which is called at the end of GLSL linking, and use
that.  This means that some other callers of finalize need to do something
with the msg, and we (for now) just throw it away.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12218>

i915g: Add finalize_nir.

This allows mesa/st to do some more optimization of state variables, but
more importantly it will be what we use to do GLSL link-time errors for
loops.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12218>

i915g: Expose PIPE_CAP_TGSI_TEXCOORD.

We want to be able to use finalize_nir, but doing so requires this flag to
be set because not having it is incompatible with being able to finalize
twice (since the texcoord workaround varyings shift happens at variant
time rather than link time).

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12218>

i915g: Simplify the process of texcoord mapping to TGSI semantics.

Custom index-as-semantic defines and extra bool flags at state update time
are unnecessary if we just store the semantics and index that each
texcoord should be.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12218>

mesa/st: Add an assertion for finalize_nir versus PIPE_CAP_TEXCOORD.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12218>

cso: Revert using FS sampler count for other stages at context unbind.

You shouldn't be asking drivers without VS texturing to unbind VS
textures, which was breaking i915g.

Fixes: 802e43a6b587 ("gallium/cso: add unbind mask for cso restore")
Reviewed-by: Marek Olšák <marek.olsak@amd.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12218>

nir/lower_amul: Fix usage of nir_foreach_src()

nir_foreach_src() bails after cb returns false for any src.  Which isn't
the behavior we were looking for.  Move progress flag to state struct
instead, so we don't skip visiting some sources.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12732>

nir/lower_amul: Handle load/store_global

These need more than 24b.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Danylo Piliaiev <dpiliaiev@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12732>

aco/ra: don't use ds_write_b8_d16_hi/ds_write_b16_d16_hi on GFX8

GFX8 doesn't support these opcodes.

Signed-off-by: Rhys Perry <pendingchaos02@gmail.com>
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Fixes: c75138ed640 ("aco/ra: refactor subdword definition info")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12721>

broadcom/compiler: check instruction belongs to current block

Check in the ldunif optimization if the current instruction belongs to
current block.

These avoids again searching the instruction when current block is not
correctly set, as it happened in
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12339 and in
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12221.

v2:
 - Remove extra blank line (Iago)

Reviewed-by: Iago Toral Quiroga <itoral@igalia.com>
Signed-off-by: Juan A. Suarez Romero <jasuarez@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12358>

broadcom/compiler: don't enable early fragment tests if shader writes Z

We had an optimization to auto-enable early fragment tests when a shader
didn't have side effects, but of course, we cannot do that this if the
shader writes Z, as in that case the fragment tests need to use the
value written from the shader.

Also, if the shader enables early fragment tests, then any shader Z
writes should be ignored.

Fixes:
dEQP-VK.spirv_assembly.instruction.graphics.early_fragment.*

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12736>

v3dv: honor VkPhysicalDeviceFeatures2 in pNext chain of VkDeviceCreateInfo

Fixes:
dEQP-VK.robustness.buffer_access.through_pointers.*.reads.*

Reviewed-by: Alejandro Piñeiro <apinheiro@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12718>

nv50/ir: Add DeadCodeElim constructor.

Fix defect reported by Coverity Scan.

Uninitialized scalar field (UNINIT_CTOR)
member_not_init_in_gen_ctor: The compiler-generated constructor
for this class does not initialize deadCount.

Signed-off-by: Vinson Lee <vlee@freedesktop.org>
Reviewed-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12697>

freedreno: use OUT_WFI for emit_marker

This is only used pre-a5xx, where we want OUT_WFI. Tested on FD420.

Fixes: 8651cfbbf0 (freedreno: emit_marker() cleanup)
Signed-off-by: Ilia Mirkin <imirkin@alum.mit.edu>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12734>

clover: Don't remove sampler/image uniforms

Otherwise we fool nir_shader_gather_info() into telling us there are no
samplers/images.

Signed-off-by: Rob Clark <robdclark@chromium.org>
Reviewed-by: Karol Herbst <kherbst@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12733>

ci/windows: Build spirv-to-dxil

Reviewed-by: Enrico Galli <enrico.galli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12728>

spirv2dxil: Fix build after spirv_to_dxil signature change

Fixes: ada05759 ("spirv_to_dxil: Convert out parameters to a single object")
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5322
Reviewed-by: Enrico Galli <enrico.galli@intel.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12728>

util/fossilize_db: Don't corrupt keys during entry read.

We change the hash that is still used to set entry->key.

Fixes: d2d642cc014 "util/fossilize_db: Only allocate entries after full read."
Reviewed-by: Timothy Arceri <tarceri@itsqueeze.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12725>

osmesa: Add a unit test for resizing buffers.

This should cover both the former use-after-free of the old pixel buffer
when moving the context to a new pixel buffer, and also successful moving
to a different-sized pixel buffer.

Reviewed-by: Adam Jackson <ajax@redhat.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12685>

gallium/osmesa: fix buffer resizing

Generate a new buffer if OSMesaMakeCurrent is called with a new size
Fix a use-after-free error when the old buffer is destroyed

Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/5276
Fixes: 26c6f6cfbb6 ("gallium/osmesa: Remove the broken buffer-reuse scheme.")
Reviewed-by: Emma Anholt <emma@anholt.net>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12685>

tgsi_to_nir: force int type for LAYER output

this otherwise gets confused as a vec4 most of the time

Cc: mesa-stable@lists.freedesktop.org
Reviewed-by: Timur Kristóf <timur.kristof@gmail.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12709>

freedreno/ir3: Use LDIB for coherent image loads on a5xx.

If the coherent flag is present, then we need to not have an incoherent
cache between us and previous stores to the image that were also decorated
as coherent.  isam apparently (unsurprisingly) goes through a texture
cache.  Use ldib instead, so that we don't get the wrong result.

We would need a similar fix for a4xx, but that uses ldgb and I don't
have hardware to test on.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12704>

freedreno/ir3: Add encode/decode support for a5xx's LDIB.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12704>

freedreno/ir3: Refactor a3xx ibo/ssbo load/store instruction XML.

There are fields common to both loads and stores, but not resinfo.  Move
them to a common bitset to reduce duplication.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12704>

freedreno/ir3: Clarify what's going on in a4xx SSBO atomics.

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12704>

freedreno/ir3: Don't use isam for coherent image loads on a6xx.

If the coherent flag is present, then we need to not have an incoherent
cache between us and previous stores to the image that were also decorated
as coherent.  isam apparently (unsurprisingly) goes through a texture
cache.  Use ldib instead, so that we don't get the wrong result.

We need a similar fix for pre-a6xx, but we don't have
emit_intrinsic_load_image for those (yet).

Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12704>

ci/freedreno: Fix xfail update for arb_draw_indirect.

I saw UnexpectedPass in my patch, and removed the Crash, but it was
actually Crash -> Fail that had happened.

Fixes: 83e9a7fbcf53 ("freedreno/ir3: Align driver param upload size/offset for indirect uploads.")
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12704>

ci/v3dv: generalize the buffer_access.through_pointers flakes.

I saw an unrelated marge pipeline fail on
dEQP-VK.robustness.buffer_access.through_pointers.graphics.reads.fragment.4B_out_of_memory_with_vec4_f32,
and it sure looks like all of these are flaky.

Reviewed-by: Jose Maria Casanova Crespo <jmcasanova@igalia.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12681>

spirv_to_dxil: Add support for non-zero vertex and instance indices

Since DXIL does not have a way to get the base/first vertex and base
instance as well as using a zero-based vertex index, these values need
to be passed in via a constant buffer.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12707>

spirv_to_dxil: Add support for nir_intrinsic_load_num_workgroups

Since DXIL does not have a way to get the number of workgroups used
during dispatch with a system-value, these numbers need to be passed in
via a constant buffer.

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12707>

nir: Add CAN_REORDER to load_ubo_dxil

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12707>

spirv_to_dxil: Convert out parameters to a single object

Reviewed-by: Jesse Natalie <jenatali@microsoft.com>
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12707>