[PATCH] Code Generator for Gather and Scatter Intrinsics.
Renato Golin
renato.golin at linaro.org
Wed Apr 29 03:55:01 PDT 2015
================
Comment at: test/CodeGen/X86/masked_memop.ll:10-11
@@ -9,4 +9,4 @@
; AVX2-LABEL: test1
-; AVX2: vpmaskmovd 32(%rdi)
-; AVX2: vpmaskmovd (%rdi)
+; AVX2: vpmaskmovd {{.*}}(%rdi)
+; AVX2: vpmaskmovd {{.*}}(%rdi)
; AVX2-NOT: blend
----------------
ab wrote:
> This is surprising, do we know why this happens?
I'm guessing the order has changed. Using AVX2-DAG on both and keeping the 32 as before should work, too, and be more explicit.
; AVX2-DAG: vpmaskmovd 32(%rdi)
; AVX2-DAG: vpmaskmovd (%rdi)
Then, the 32 can come in any order.
http://reviews.llvm.org/D7665
EMAIL PREFERENCES
http://reviews.llvm.org/settings/panel/emailpreferences/
More information about the llvm-commits
mailing list