Enable native vector DXIL intrinsic overload for vector load/store

Add a new native vector overload type to DXIL intrinsics and the corresponding generation.
Add new raw buffer vector load/store intrinsics that use that overload type.

Generate native vector raw buffers load/stores

When the loaded/stored type is a vector of more than 1 element, the
shader model is 6.9 or higher, and the operation is on a raw buffer,
enable the generation of a native vector raw buffer load or store.

Add validation of vector load stores

actually allow the given ops to take vectors
add vector overload type and apply to the relevant builtins

Build lowering functions to allow vector supporting intrinsics through

Preliminary groupshared support. keep groupshared as vectors for 6.9. They are no longer represented as inidivual groupshared scalars.

adds groupshared to the test and performs the switch to CS to allow it.

Support dot product on long vecs by expanding the inrinsic into
mul/mad ops like is done with integer dot products

Since the or() and and() intrinsics did their own scalarization, the or/and operators would never be applied to full vectors. This leaves the scalarization for the scalarization pass, which will skip it for 6.9

…soft#7309)

This PR modifies WaveSizeRange test which depends on shader model 6.8.
The compiler needs -select-validator internal.
This will allow the tests to be run in different testing environments
when an external validator that isn't sufficient is available.

The name matches for azure pipelines to run excluded staging branches.
This adds them in.

Not sure this is desired, but it's here if we want it.

Conflicts:

        both modified:   include/dxc/DXIL/DxilConstants.h
        both modified:   include/dxc/DXIL/DxilOperations.h
        both modified:   lib/DXIL/DxilOperations.cpp
        both modified:   lib/DxilValidation/DxilValidation.cpp
        both modified:   lib/HLSL/DxilLinker.cpp
        both added:      lib/HLSL/DxilScalarizeVectorLoadStores.cpp
        both modified:   lib/HLSL/HLOperationLower.cpp
        both modified:   tools/clang/lib/Sema/SemaHLSL.cpp
both added:
tools/clang/test/CodeGenDXIL/hlsl/intrinsics/buffer-load-stores-sm69.hlsl
both added:
tools/clang/test/CodeGenDXIL/hlsl/types/longvec-operators-cs.hlsl
both added: tools/clang/test/DXILValidation/vector-validation.hlsl
        both modified:   utils/hct/hctdb.py
        both modified:   utils/hct/hctdb_instrhelp.py

Merge latest main into staging-sm6.9.

Update staging-sm6.9 with latest main.

This merges main into staging-sm6.9.  There were no conflicts.

…ics (microsoft#7290)

Implements
HLSL:
__builtin_MatVecMul
__builtin_MatVecMulAdd
__builtin_OuterProductAccumulate
__builtin_VectorAccumulate

Lowered to
DXIL:
@dx.op.matVecMul
@dx.op.matVecMulAdd
 @dx.op.outerProductAccumulate
 @dx.op.vectorAccumulate

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Damyan Pepper <[email protected]>
Co-authored-by: Simon Moll <[email protected]>
Co-authored-by: Tex Riddell <[email protected]>
Co-authored-by: Chris B <[email protected]>

…crosoft#7366)

DXC seems to be building inocrrectly with GCC-13 and later, which is
causing our pre-merge testing on 24.04 to fail. This will take some time
to sort out, so in the meantime I'm reverting to 22.04 on our pipelines.

(cherry picked from commit b4a3076)

Co-authored-by: Chris B <[email protected]>

)

This PR is a basic implementation of the priority long vector execution
tests microsoft#7260.

Merge main into staging-sm6.9

…rosoft#7375)

Moving the long vector test utility functionality to its own file to
help make subsequent reviews easier. No new code or logic updates.

- All trivial scalar/vector/matrix getters
- HitObject::FromRayQuery with procedural hit
- HitObject::GetAttributes<T> with custom attributes and procedural hit

SER implementation tracker: microsoft#7214

This PR introduces the linear algebra header file, and places it in a
location that is by default included in all HLSL compilation.
The builtins in the API aren't yet defined, and depend on the microsoft#7290 PR
merging first.
The tests that have been added have temporary diagnostic messages while
7290 is in progress. They will need to be updated.
Open to feedback on better / suggested error messages, or whether there
shouldn't be any sema-level validation for these errors.

Fixes
[microsoft#7304](microsoft#7304)

---------

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

…ocal constants / simplifications

… simplified tests/ added SERInvokeNoSBTTest

…iler into staging-sm6.9

Conflict in HLOperations.h

Merge was clean apart from one conflict in HLOperations.h where
HitObject::FromRayQuery wasn't in the staging version and got added.

…atch

- Array of 8 HitObjects, each with different ray flags
- Samples at 'random' positions to block SROA from breaking down the
  array

…rProductAccumulate (microsoft#7424)

ExecutionTest::CoopVec_MulAdd:

Functional verification for the Mul() and MulAdd() HLSL APIs. The driver
matrix conversion API is tested as well. These tests should be
considered
as work-in-progress as this point. They include coverage primarily for
SINT8,
FLOAT16, FLOAT_E4M3, and FLOAT_E5M2. The test queries the driver for all
supported configurations and runs each one, with a filtering mechanism
to
limit the set of tests to the minimal feature set. The set of tests can
be
further filtered by the following TE parameters:

  CoopVecMatrixInterp: SINT8, FLOAT16, FLOAT_E4M3, ...
CoopVecMatrixLayout: ROW_MAJOR, COLUMN_MAJOR, MUL_OPTIMAL,
OUTER_PRODUCT_OPTIMAL
  CoopVecBiasInterp: SINT32, FLOAT16, FLOAT_E4M3, ...
  CoopVecInputInterp: SINT8, FLOAT16, FLOAT_E4M3, ...
CoopVecInputType: SINT8, UINT8, SINT16, UINT16, SINT32, UINT32, FLOAT16,
FLOAT32, ...
  CoopVecOutputType: SINT32, UINT32, FLOAT16, FLOAT32, ...

Filter example:
$ TE.exe ... -p:CoopVecMatrixInterp=FLOAT16
-p:CoopVecMatrixLayout=MUL_OPTIMAL

Precision coverage is minimal at this point, using an all-ones input
matrix and
test vector with ones in the first two components. This is enough to
test basic
functionality, but more comprehensive tests are needed.

ExecutionTest::CoopVec_OuterProduct:

Functional verification for the OuterProductAccumulate() HLSL API. This
test queries
the driver for all supported configurations and runs each one. No
filtering is
currently implemented.

… arith for hitKind computation

…/tri hit order

…atch

Issue: microsoft/hlsl-specs#613

Changes:
- Added permutation testing for getters and intersection attributes.
  Combinations of:
  * Producing HitObjects (HitObject::TraceRay, HitObject::FromRayQuery,
    HitObject::MakeMiss)
  * Querying property from HitObject (Direct HitObject getter or
    HitObject::Invoke+classic getter)
  * HitObject tested in raygen,closesthit or miss (adds recursion)
  * with and without reorder

- Fused all Scalar/VectorMatrixGetter tests into one SERGetterPermutationTest
- Added SERAttributesPermutationTest: Testing procedural and triangle
  intersection attributes

- Added SERNOPValuesTest: Explicit test for default return values in NOP
  HitObjects.

- Added SERMultiPayloadTest: Testing multiple live HitObjects with
  differing payload types, SBT indices and loop control flow.

Issue: microsoft/hlsl-specs#612

Adds code paths with new HitObject::GetAttributes API (guarded by #if NEW_GETATTRIBUTES_API to help transition).