[clang] [clang-tools-extra] [llvm] [CLANG] Add warning when INF or NAN are used in a binary operation or as function argument in fast math mode. (PR #76873)

[Andy Kaylor via cfe-commits]

================
@@ -2835,6 +2835,9 @@ class Preprocessor {
     if (Identifier.getIdentifierInfo()->isRestrictExpansion() &&
         !SourceMgr.isInMainFile(Identifier.getLocation()))
       emitRestrictExpansionWarning(Identifier);
+
+    if (Identifier.getIdentifierInfo()->getName() == "INFINITY")
----------------
andykaylor wrote:

> Since NAN is defined as (-(float)(INFINITY * 0.0F)) when used in the program it will trigger the "use of infinity via a macro results in undefined behavior due to the currently enabled floating-point options [-Wnan-and-infinity-disabled]". Wouldn't that be enough? If I add NaN here we would get 2 warnings.

This is actually a bit unfortunate. If I use "-fhonor-nans -fno-honor-infinities" this definition of NAN will trigger a warning, even though I expect that we'll constant-fold it as expected without regard to the "no-honor-infinities" setting.

https://github.com/llvm/llvm-project/pull/76873