[llvm] [ADT] Add fshl/fshr operations to APInt (PR #153790)

[Craig Topper via llvm-commits]

================
@@ -3169,3 +3169,21 @@ APInt APIntOps::pow(const APInt &X, int64_t N) {
   }
   return Acc;
 }
+
+APInt llvm::APIntOps::fshl(const APInt &Hi, const APInt &Lo,
+                           const APInt &Shift) {
+  assert(Hi.getBitWidth() == Lo.getBitWidth());
+  unsigned ShiftAmt = rotateModulo(Hi.getBitWidth(), Shift);
+  if (ShiftAmt == 0)
+    return Hi;
+  return APInt(Hi.shl(ShiftAmt) | Lo.lshr(Hi.getBitWidth() - ShiftAmt));
----------------
topperc wrote:

Why do we need the outer `APInt` constructor call?

https://github.com/llvm/llvm-project/pull/153790