[PATCH] D82756: Port some floating point options to new option marshalling infrastructure

[Anastasia Stulova via Phabricator via cfe-commits]

Anastasia added inline comments.


================
Comment at: clang/include/clang/Driver/Options.td:1176
+defm reciprocal_math : OptInFFlag< "reciprocal-math", "Allow division operations to be reassociated", "", "", [], "LangOpts->AllowRecip">;
+def fapprox_func : Flag<["-"], "fapprox-func">, Group<f_Group>, Flags<[CC1Option, NoDriverOption]>,
+  MarshallingInfoFlag<"LangOpts->ApproxFunc", "false">;
----------------
could this also be OptInFFlag?


================
Comment at: clang/lib/Driver/ToolChains/Clang.cpp:2805
     CmdArgs.push_back("-menable-unsafe-fp-math");
+    ApproxFunc = true;
+  }
----------------
Is this a bug fix ?


================
Comment at: clang/test/CodeGen/fp-function-attrs.cpp:2
+// RUN: %clang_cc1 -triple x86_64-linux-gnu -ffast-math -ffinite-math-only -menable-unsafe-fp-math \
+// RUN:   -menable-no-infs -menable-no-nans -fno-signed-zeros -freciprocal-math \
+// RUN:   -fapprox-func -mreassociate -ffp-contract=fast -emit-llvm -o - %s | FileCheck %s
----------------
Not clear why do you need to pass these extra flags now?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D82756/new/

https://reviews.llvm.org/D82756