[llvm] PreISelIntrinsicLowering: Lower llvm.exp to a loop if scalable vec arg (PR #117568)

[Matt Arsenault via llvm-commits]

================
@@ -456,7 +456,11 @@ bool PreISelIntrinsicLowering::lowerIntrinsics(Module &M) const {
       break;
     case Intrinsic::exp:
       Changed |= forEachCall(F, [&](CallInst *CI) {
-        return lowerUnaryMathIntrinsicWithScalableVecArgAsLoop(M, CI);
+        // TODO: Check legality
+        if (!CI->getArgOperand(0)->getType()->isVectorTy()) {
----------------
arsenm wrote:

Without a legality check, doing this for all vectors is more likely to regress something in the fixed vector case 

https://github.com/llvm/llvm-project/pull/117568