| commit | 822227740b4fabf62d618cec49e135e4b1239cac | [log] [tgz] |
|---|---|---|
| author | Craig Topper <craig.topper@intel.com> | Mon Feb 19 22:07:31 2018 +0000 |
| committer | Craig Topper <craig.topper@intel.com> | Mon Feb 19 22:07:31 2018 +0000 |
| tree | 24cddb71851e9a03fc36007c617f7cb2fc846d63 | |
| parent | c6e8ed569c30b7e6402e8a2fef618e7a35e7cf5b [diff] |
[X86] Use vpmovq2m/vpmovd2m for truncate to vXi1 when possible. Previously we used vptestmd, but the scheduling data for SKX says vpmovq2m/vpmovd2m is lower latency. We already used vpmovb2m/vpmovw2m for byte/word truncates. So this is more consistent anyway. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@325534 91177308-0d34-0410-b5e6-96231b3b80d8