[PATCH] D98435: [LoopVectorize] Add strict in-order reduction support for fixed-width vectorization

[Peter Waller via Phabricator via llvm-commits]

peterwaller-arm added a comment.

Some minor nits. Ran out of time for a more thorough review at this moment.



================
Comment at: llvm/lib/Transforms/Utils/LoopUtils.cpp:1081
+
+  std::function<Value *(Value *, Value *)> CreateRdx;
+  // Predication is done by masking out the inactive elements of the vector
----------------
Nit: this variable appears to be unused?


================
Comment at: llvm/lib/Transforms/Vectorize/LoopVectorize.cpp:329
+cl::opt<bool> EnableStrictReductions(
+    "enable-strict-reducs", cl::init(false), cl::Hidden,
+    cl::desc("Enable the vectorisation of loops with in-order (strict) "
----------------
nit: As 'reduction' is spelt out on the adjacent arguments I would prefer consistency over brevity unless there is a reason to differ here.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D98435/new/

https://reviews.llvm.org/D98435