[PATCH] D90950: [AArch64]Add memory op cost model for Neon

[Renato Golin via Phabricator via llvm-commits]

rengolin accepted this revision.
rengolin added a comment.
This revision is now accepted and ready to land.

Nice change, simpler to read and easier to get right. LGTM, thanks!



================
Comment at: llvm/test/Analysis/CostModel/AArch64/mem-op-cost-model.ll:51
+; CHECK-NEON: Cost Model: Found an estimated cost of 64 for instruction:
+; CHECK-SVE: Cost Model: Found an estimated cost of 64 for instruction:
+; CHECK-SVE-256: Cost Model: Found an estimated cost of 1 for instruction:
----------------
I'm assuming that the idea is that this vector will be NEON and not SVE, so the cost is the same.


================
Comment at: llvm/test/Analysis/CostModel/AArch64/scalable-mem-op-cost-model.ll:1
+
+; Checks if the memory cost model does not break when using scalable vectors
----------------
Nit: No empty first line and one empty line between RUN line and first piece of code.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D90950/new/

https://reviews.llvm.org/D90950