| commit | 6f1a63bcdf5096fcc59cf2b4bc8f2e706865c2e4 | [log] [tgz] |
|---|---|---|
| author | Simon Pilgrim <llvm-dev@redking.me.uk> | Sat Feb 23 17:10:47 2019 +0000 |
| committer | Simon Pilgrim <llvm-dev@redking.me.uk> | Sat Feb 23 17:10:47 2019 +0000 |
| tree | b91eaed77f6e1f86b7cf9ec2f3d89ace46817fac | |
| parent | 713bb48902d77ec8f333a8992bad15e38ec922c7 [diff] |
[X86][AVX] Shuffle->Permute+Blend if we have one v4f64/v4i64 shuffle input in place Even on AVX1 we can pretty cheaply (VPERM2F128+VSHUFPD) permute a single v4f64/v4i64 input (on AVX2 its just a single VPERMPD), followed by a BLENDPD. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@354729 91177308-0d34-0410-b5e6-96231b3b80d8