[clang] [llvm] [HexFloat] add HexFloat to APFloat (PR #179771)

[via cfe-commits]

================
@@ -257,7 +259,13 @@ class APFloatBase {
     S_Float4E2M1FN,
     // TODO: Documentation is missing.
     S_x87DoubleExtended,
-    S_MaxSemantics = S_x87DoubleExtended,
+
+    // HexFloat
+    S_HexFP32,
+    S_HexFP64,
+    S_HexFP128,
----------------
Ariel-Burton wrote:

The original RFC gave an overview.  See also:

- the wiki page for HexFloat (https://en.wikipedia.org/wiki/IBM_hexadecimal_floating-point)
- chapters 9 and 18 of the z/Architecture Principles of Operation
- the 1999 paper by Schwartz et al (http://eece.cu.edu.eg/~hfahmy/arith_class/hex_bin_FP.pdf)

https://github.com/llvm/llvm-project/pull/179771