[PATCH] D46273: [InstCombine, ARM] Convert vld1 to llvm load

[Eli Friedman via Phabricator via llvm-commits]

efriedma added a comment.

Sorry about the delay.  I ran some benchmarks internally, looks like it's perfomance-neutral, so we're fine there.



================
Comment at: test/Transforms/InstCombine/ARM/vld1.ll:24
+; CHECK-NEXT:    [[TMP1:%.*]] = bitcast i8* [[PTR:%.*]] to <8 x i8>*
+; CHECK-NEXT:    [[TMP2:%.*]] = load <8 x i8>, <8 x i8>* [[TMP1]], align 8
+; CHECK-NEXT:    ret <8 x i8> [[TMP2]]
----------------
Alignment on an LLVM IR load instruction is in bytes, not bits.


https://reviews.llvm.org/D46273