[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

https://github.com/echesakov updated https://github.com/llvm/llvm-project/pull/105912 >From 1231a5658bf7dde633e2d84967c6e540b3259e4b Mon Sep 17 00:00:00 2001 From: Egor Chesakov <5292656+echesa...@users.noreply.github.com> Date: Sat, 5 Oct 2024 12:22:27 -0700 Subject: [PATCH] [clang] Allow `Cond

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

echesakov wrote: > Not very clear to me what the issue is. Can you please explain this in the > description. I think here the `#pragma float_control(precise, on)` should be > taking precedence over the `-ffast-math` flag and it's not. Right? @zahiraam Correct, this is not happening right now a

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

https://github.com/echesakov updated https://github.com/llvm/llvm-project/pull/105912 >From 124a1c4c7c102421b74eaaa1133ea58a9017e07b Mon Sep 17 00:00:00 2001 From: Egor Chesakov <5292656+echesa...@users.noreply.github.com> Date: Fri, 23 Aug 2024 18:44:14 -0700 Subject: [PATCH 1/2] [clang] Allow

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

@@ -0,0 +1,36 @@ +// RUN: %clang_cc1 -disable-llvm-passes -emit-llvm %s -o - | FileCheck %s +// RUN: %clang_cc1 -disable-llvm-passes -emit-llvm -menable-no-infs -fapprox-func\ +// RUN: -funsafe-math-optimizations -fno-signed-zeros -mreassociate -freciprocal-math\ +// RUN: -ffp-c

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

@@ -0,0 +1,36 @@ +// RUN: %clang_cc1 -disable-llvm-passes -emit-llvm %s -o - | FileCheck %s +// RUN: %clang_cc1 -disable-llvm-passes -emit-llvm -menable-no-infs -fapprox-func\ +// RUN: -funsafe-math-optimizations -fno-signed-zeros -mreassociate -freciprocal-math\ +// RUN: -ffp-c

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

@@ -4210,26 +4210,45 @@ class AbstractConditionalOperator : public Expr { /// ConditionalOperator - The ?: ternary operator. The GNU "missing /// middle" extension is a BinaryConditionalOperator. -class ConditionalOperator : public AbstractConditionalOperator { +class Conditi

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

zahiraam wrote: Not very clear to me what the issue is. Can you please explain this in the description. I think here the `#pragma float_control(precise, on)` should be taking precedence over the `-ffast-math` flag and it's not. Right? https://github.com/llvm/llvm-project/pull/105912 _

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

echesakov wrote: Ping https://github.com/llvm/llvm-project/pull/105912 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

llvmbot wrote: @llvm/pr-subscribers-clang-codegen @llvm/pr-subscribers-clang-modules Author: Egor Chesakov (echesakov) Changes ...control `FastMathFlags` during code-generation of `ConditionalOperator`. This involves storing `FPOptionsOverride` in trailing storage of `ConditionalOperator

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

github-actions[bot] wrote: Thank you for submitting a Pull Request (PR) to the LLVM Project! This PR will be automatically labeled and the relevant teams will be notified. If you wish to, you can add reviewers by using the "Reviewers" section on this page. If this is not working for you, it

[clang] [clang] Allow `pragma float_control(precise, *)` to... (PR #105912)

https://github.com/echesakov created https://github.com/llvm/llvm-project/pull/105912 ...control `FastMathFlags` during code-generation of `ConditionalOperator`. This involves storing `FPOptionsOverride` in trailing storage of `ConditionalOperator` and storing `CurFPOptionsOverride` when creat