llvm-project.git/llvm/lib/Target/DirectX/DirectXTargetTransformInfo.cpp, branch users/nico/python-2

[DirectX] Make firstbithigh intrinsic use first arg as overload type (#145401)

2025-06-24T18:00:21+00:00

Sibling to https://github.com/llvm/llvm-project/pull/145350 part of
https://github.com/llvm/llvm-project/issues/144966

firstbituhigh and firstbitshigh should use the first arg for overloads
instead of the return type since the return is always i32

[DirectX] make firstbitlow intrinsic use first argument instead of return for overload type (#145350)

2025-06-23T18:04:06+00:00

fixes #144966
Easy fix just add `dx_firstbitlow` to
`DirectXTTIImpl::isTargetIntrinsicWithOverloadTypeAtArg`

[DirectX] scalarize the dx.isinf intrinsic (#140638)

2025-05-20T17:26:58+00:00

The DXIL IsInf op only takes scalars.
Closes #140577

[TTI] Make all interface methods const (NFCI) (#136598)

2025-04-22T03:27:29+00:00

Making `TargetTransformInfo::Model::Impl` `const` makes sure all
interface methods are `const`, in `BasicTTIImpl`, its bases, and in all
derived classes.

Pull Request: https://github.com/llvm/llvm-project/pull/136598

[DirectX] Scalarize the dx.saturate intrinsic (#134381)

2025-04-07T23:54:10+00:00

The DXIL Saturate op only takes scalars.

Fixes #134378.

[DirectX][TTI] Sort switch statements. NFC (#134379)

2025-04-04T13:26:46+00:00

[HLSL][SPIRV][DXIL] Implement `WaveActiveMax` intrinsic (#123428)

2025-01-28T07:26:56+00:00

```    - add clang builtin to Builtins.td
      - link builtin in hlsl_intrinsics
      - add codegen for spirv intrinsic and two directx intrinsics to retain
        signedness information of the operands in CGBuiltin.cpp
      - add semantic analysis in SemaHLSL.cpp
      - add lowering of spirv intrinsic to spirv backend in
        SPIRVInstructionSelector.cpp
      - add lowering of directx intrinsics to WaveActiveOp dxil op in
    DXIL.td

      - add test cases to illustrate passespendent pr merges.
```
Resolves #99170

[HLSL][SPIRV][DXIL] Implement `WaveActiveSum` intrinsic (#118580)

2025-01-16T18:35:23+00:00

```    - add clang builtin to Builtins.td
      - link builtin in hlsl_intrinsics
      - add codegen for spirv intrinsic and two directx intrinsics to retain
        signedness information of the operands in CGBuiltin.cpp
      - add semantic analysis in SemaHLSL.cpp
      - add lowering of spirv intrinsic to spirv backend in
        SPIRVInstructionSelector.cpp
      - add lowering of directx intrinsics to WaveActiveOp dxil op in
    DXIL.td

      - add test cases to illustrate passespendent pr merges.
```
Resolves #70106

---------

Co-authored-by: Finn Plummer

[HLSL] Implement elementwise firstbitlow builtin (#116858)

2025-01-15T22:36:50+00:00

Closes https://github.com/llvm/llvm-project/issues/99116

Implements `firstbitlow` by extracting common functionality from
`firstbithigh` into a shared function while also fixing a bug for an edge
case where `u64x3` and larger vectors will attempt to create vectors
larger than the SPRIV max of 4.
---------

Co-authored-by: Steven Perron

[NFC][TargetTransformInfo][VectorUtils] Consolidate `isVectorIntrinsic...` api (#117635)

2024-12-19T19:54:26+00:00

- update `VectorUtils:isVectorIntrinsicWithScalarOpAtArg` to use TTI for
all uses, to allow specifiction of target specific intrinsics
- add TTI to the `isVectorIntrinsicWithStructReturnOverloadAtField` api
- update TTI api to provide `isTargetIntrinsicWith...` functions and
  consistently name them
- move `isTriviallyScalarizable` to VectorUtils
  
- update all uses of the api and provide the TTI parameter

Resolves #117030