[PATCH] D13161: [PATCH, PR24373] Combine shifts for x86

[Andrea Di Biagio via llvm-commits]

andreadb added a subscriber: andreadb.
andreadb added a comment.

In http://reviews.llvm.org/D13161#266882, @evstupac wrote:

> During final check 2 new tests failed on x86:
>
> 1. sar_fold64.ll failed on Windows. Fixed by using regular expression for parameter register: "movswq %{{[cd][xi]}}, %rax" as parameter could be "cx" or "di" and "movsbq %{{[cdi]*l}}, %rax" as parameter could be "cl" or "dil"


The easier way to fix those problems is to force a target triple to the tests.
What if you explicitly force -mtriple=x86_64-unknown-unknown to test sar_fold64.ll and -mtriple=i686-unknown-unknown to sar_fold.ll?

About the vector-sext.ll failures,
are those failures only related to the last RUN line in the file (the i686 run line)? Does the problem disappear if you replace -mcpu=i686 with -mcpu=generic ?


Repository:
  rL LLVM

http://reviews.llvm.org/D13161