[llvm] [KnownFPClass] Fix fadd/fsub input handling and improve fadd_self (PR #225605)

[Eric Ross via llvm-commits]

================
@@ -363,9 +385,9 @@ KnownBits KnownFPClass::toKnownBits(const fltSemantics &FltSemantics) const {
   return Known;
 }
 
-// Handle known sign bit and nan cases for fadd.
-static KnownFPClass fadd_impl(const KnownFPClass &KnownLHS,
-                              const KnownFPClass &KnownRHS, DenormalMode Mode) {
----------------
ZERICO2005 wrote:

I removed `DenormalMode Mode` from `fadd_impl` because it was not used anywhere in this function.

https://github.com/llvm/llvm-project/pull/225605