Skip to content

[HLSL] Add experimental_constrained_tan intrinsic #91421

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Closed
farzonl opened this issue May 8, 2024 · 1 comment · Fixed by #93314 or #94559
Closed

[HLSL] Add experimental_constrained_tan intrinsic #91421

farzonl opened this issue May 8, 2024 · 1 comment · Fixed by #93314 or #94559
Assignees
@farzonl
Labels
clang:codegen IR generation bugs: mangling, exceptions, etc. HLSL HLSL Language Support

Comments

@farzonl
Copy link
Member

farzonl commented May 8, 2024

We want to do the same thing that cos\sin do in the below examples:

  case Builtin::BI__builtin_cos:
  case Builtin::BI__builtin_cosf:
  case Builtin::BI__builtin_cosf16:
  case Builtin::BI__builtin_cosl:
  case Builtin::BI__builtin_cosf128:
    return RValue::get(emitUnaryMaybeConstrainedFPBuiltin(*this, E,
                                 Intrinsic::cos,
                                 Intrinsic::experimental_constrained_cos));
  
  case Builtin::BI__builtin_sin:
  case Builtin::BI__builtin_sinf:
  case Builtin::BI__builtin_sinf16:
  case Builtin::BI__builtin_sinl:
  case Builtin::BI__builtin_sinf128:
    return RValue::get(emitUnaryMaybeConstrainedFPBuiltin(*this, E,
                                 Intrinsic::sin,
                                 Intrinsic::experimental_constrained_sin));

Originally posted by @farzonl in #90276 (comment)
@farzonl farzonl self-assigned this May 8, 2024
@farzonl farzonl added clang Clang issues not falling into any other category HLSL HLSL Language Support and removed new issue labels May 8, 2024
@farzonl farzonl removed this from HLSL Support May 8, 2024
@farzonl farzonl moved this to Planning in HLSL Support May 8, 2024
@farzonl farzonl mentioned this issue May 24, 2024
Merged
@farzonl farzonl moved this from Planning to Active in HLSL Support May 29, 2024
farzonl added a commit that referenced this issue May 29, 2024
@farzonl
[clang] Add tanf16 builtin and support for tan constrained intrinsic (#…
b15a0a3 
…93314)

In LLVM, the `llvm.experimental.constrained.cos` and
`llvm.experimental.constrained.sin` intrinsics are used for performing
cosine and sine calculations with additional constraints on
floating-point operations. This behavior is expected for all
floating-point math intrinsics. This change adds these constraints for
the `tan` intrinsic.

-  `Builtins.td` - replace TanF128 with F16F128MathTemplate
- `CGBuiltin.cpp` - map existing tan builtins to `tan` and
`constrained_tan` intrinsic
-   `ConstrainedOps.def` map tan and constrained_tan  to an ISDOpcode.
-  `ISDOpcodes.h` - define tan and strict tan  opcodes

resolves  #91421
@github-project-automation github-project-automation bot moved this from Active to Done in HLSL Support May 29, 2024
@EugeneZelenko EugeneZelenko added clang:codegen IR generation bugs: mangling, exceptions, etc. and removed clang Clang issues not falling into any other category labels May 29, 2024
@llvmbot
Copy link
Member

llvmbot commented May 29, 2024

@llvm/issue-subscribers-clang-codegen

Author: Farzon Lotfi (farzonl)

We want to do the same thing that cos\sin do in the below examples: 
  case Builtin::BI__builtin_cos:
  case Builtin::BI__builtin_cosf:
  case Builtin::BI__builtin_cosf16:
  case Builtin::BI__builtin_cosl:
  case Builtin::BI__builtin_cosf128:
    return RValue::get(emitUnaryMaybeConstrainedFPBuiltin(*this, E,
                                 Intrinsic::cos,
                                 Intrinsic::experimental_constrained_cos));
  
  case Builtin::BI__builtin_sin:
  case Builtin::BI__builtin_sinf:
  case Builtin::BI__builtin_sinf16:
  case Builtin::BI__builtin_sinl:
  case Builtin::BI__builtin_sinf128:
    return RValue::get(emitUnaryMaybeConstrainedFPBuiltin(*this, E,
                                 Intrinsic::sin,
                                 Intrinsic::experimental_constrained_sin));

Originally posted by @farzonl in #90276 (comment)

@farzonl farzonl mentioned this issue May 29, 2024
Merged
farzonl added a commit that referenced this issue May 29, 2024
@farzonl
Revert "[clang] Add tanf16 builtin and support for tan constrained in…
7348bb2 
…trinsic (#93314)" (#93721)

This reverts commit b15a0a3.

This should undo PR: #93314
will need to re-open #91421

wait for #90503 to land
@farzonl farzonl reopened this May 29, 2024
@farzonl farzonl moved this from Done to Planning in HLSL Support May 29, 2024
@farzonl farzonl mentioned this issue Jun 6, 2024
Merged
@farzonl farzonl moved this from Planning to Active in HLSL Support Jun 6, 2024
farzonl added a commit that referenced this issue Jun 11, 2024
@farzonl
[clang] Reland Add tanf16 builtin and support for tan constrained int…
189d471 
…rinsic (#94559)

Relanding this PR now that
#90503 has merged. with `FTAN`
landing in
[TargetLoweringBase.cpp:L1021](https://github.com/llvm/llvm-project/blob/main/llvm/lib/CodeGen/TargetLoweringBase.cpp#L1020C23-L1021C63
) There is now a llvm tan intrinsic 32\64\128 Expand case for all llvm
backends.

In LLVM, the `llvm.experimental.constrained.cos` and
`llvm.experimental.constrained.sin` intrinsics are used for performing
cosine and sine calculations with additional constraints on
floating-point operations. This behavior is expected for all
floating-point math intrinsics. This change adds these constraints for
the `tan` intrinsic.

-  `Builtins.td` - replace TanF128 with F16F128MathTemplate
- `CGBuiltin.cpp` - map existing tan builtins to `tan` and
`constrained_tan` intrinsic
-   `ConstrainedOps.def` map tan and constrained_tan  to an ISDOpcode.

resolves  #91421

---------

Co-authored-by: Farzon Lotfi <[email protected]>
@github-project-automation github-project-automation bot moved this from Active to Done in HLSL Support Jun 11, 2024
Lukacma pushed a commit to Lukacma/llvm-project that referenced this issue Jun 12, 2024
@farzonl @Lukacma
[clang] Reland Add tanf16 builtin and support for tan constrained int…
16d221c 
…rinsic (llvm#94559)

Relanding this PR now that
llvm#90503 has merged. with `FTAN`
landing in
[TargetLoweringBase.cpp:L1021](https://github.com/llvm/llvm-project/blob/main/llvm/lib/CodeGen/TargetLoweringBase.cpp#L1020C23-L1021C63
) There is now a llvm tan intrinsic 32\64\128 Expand case for all llvm
backends.

In LLVM, the `llvm.experimental.constrained.cos` and
`llvm.experimental.constrained.sin` intrinsics are used for performing
cosine and sine calculations with additional constraints on
floating-point operations. This behavior is expected for all
floating-point math intrinsics. This change adds these constraints for
the `tan` intrinsic.

-  `Builtins.td` - replace TanF128 with F16F128MathTemplate
- `CGBuiltin.cpp` - map existing tan builtins to `tan` and
`constrained_tan` intrinsic
-   `ConstrainedOps.def` map tan and constrained_tan  to an ISDOpcode.

resolves  llvm#91421

---------

Co-authored-by: Farzon Lotfi <[email protected]>
@damyanp damyanp moved this to Closed in HLSL Support Apr 25, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@farzonl farzonl

Labels
clang:codegen IR generation bugs: mangling, exceptions, etc. HLSL HLSL Language Support
Projects
HLSL Support
Status: Closed
Milestone
No milestone
3 participants
@farzonl @EugeneZelenko @llvmbot