[PATCH] D38611: Fix X86 regression on linpack

[Simon Pilgrim via Phabricator via llvm-commits]

RKSimon added reviewers: craig.topper, RKSimon, andreadb.
RKSimon added inline comments.


================
Comment at: lib/Target/X86/X86TargetTransformInfo.cpp:2538
+    // stores with folded address, treating each 2 such stores as 1
+    // additional instruction.
+    unsigned C1Insns = ((C1.FoldedStoreAddresses >> 1) + C1.Insns);
----------------
This seems incredibly specific to a particular Intel architecture - you said you've tried to make a best average - based on what?


================
Comment at: test/Transforms/LoopStrengthReduce/X86/folded_addresses.ll:8
+; Function Attrs: norecurse nounwind uwtable
+define void @foo(i32, float, float* noalias nocapture readonly, float* noalias nocapture) local_unnamed_addr #0 {
+; CHECK-LABEL: @foo(
----------------
qcolombet wrote:
> Could you use `opt -instnamer` on the input test?
Is there any way that this can be reduced further?


Repository:
  rL LLVM

https://reviews.llvm.org/D38611