[PATCH] D125235: [LoopVectorize] Add overflow checks when tail-folding with scalable vectors

[Sander de Smalen via Phabricator via llvm-commits]

sdesmalen added inline comments.


================
Comment at: llvm/lib/Transforms/Vectorize/LoopVectorize.cpp:2989
+    Value *LHS = ConstantInt::get(
+        Count->getType(), cast<IntegerType>(Count->getType())->getBitMask());
+    LHS = Builder.CreateSub(LHS, Count);
----------------
The interface of `getBitMask` returns a `uint64_t`, which means this functionality is limited to loops with i64 induction variables. I don't know whether this is a limitation that's already assumed elsewhere, but perhaps you can use `getMask()` instead (which returns an `APInt`).


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D125235/new/

https://reviews.llvm.org/D125235