[PATCH] D24288: Don't reduce mul if the target doesn't support SSE2

Michael Kuperstein via llvm-commits llvm-commits at lists.llvm.org
Tue Sep 6 21:46:50 PDT 2016


mkuper added a comment.

Thanks, Wei.
The change LGTM, but can you fix the test up a bit?


================
Comment at: test/CodeGen/X86/pr30298.ll:1
@@ +1,2 @@
+; REQUIRES: asserts
+; RUN: llc -mtriple=i386-pc-linux-gnu -mattr=+sse < %s | FileCheck %s
----------------
I don't see why this test would require asserts. A copy/paste bug?

================
Comment at: test/CodeGen/X86/pr30298.ll:3
@@ +2,3 @@
+; RUN: llc -mtriple=i386-pc-linux-gnu -mattr=+sse < %s | FileCheck %s
+; CHECK-NOT: pmullw
+ at c = external global i32*, align 8
----------------
Please add a positive check for the code we do expect.


Repository:
  rL LLVM

https://reviews.llvm.org/D24288





More information about the llvm-commits mailing list