[llvm] [HLSL] [DirectX] translate llvm fast math flags to llvm 3.7 fast math flags (PR #122025)

[Chris B via llvm-commits]

================
@@ -759,10 +759,6 @@ uint64_t DXILBitcodeWriter::getOptimizationFlags(const Value *V) {
       Flags |= bitc::NoSignedZeros;
     if (FPMO->hasAllowReciprocal())
       Flags |= bitc::AllowReciprocal;
-    if (FPMO->hasAllowContract())
-      Flags |= bitc::AllowContract;
-    if (FPMO->hasApproxFunc())
----------------
llvm-beanz wrote:

Also, in LLVM 3.7 UnsafeAlgebra was a magic catch all flag for "all things fast", it was separated out into granular flags in this commit - https://github.com/llvm/llvm-project/commit/629c41153876b.

It had the impact in the LLVM 3.7 optimizer of controlling both contraction and reassociation (to the extent that LLVM 3.7 had optimizations).

https://github.com/llvm/llvm-project/pull/122025