Commit Graph

4 Commits

Author SHA1 Message Date
Craig Topper cada9f2275 [AVX-512] Add support for commuting VPERMT2(B/W/D/Q/PS/PD) to/from VPERMI2(B/W/D/Q/PS/PD).
Summary:
The index and one of the table operands can be swapped by changing the opcode to the other version. Neither of these operands are the one that can load from memory so this can't be used to increase memory folding opportunities.

We need to handle the unmasked forms and the kz forms. Since the load operand isn't being commuted we can commute the load and broadcast instructions too.

Reviewers: igorb, delena, Ayal, Farhana, RKSimon

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D25652

llvm-svn: 287621
2016-11-22 04:57:34 +00:00
Simon Pilgrim 5401bae523 [X86][AVX512] Add support for VBMI VPERMV3 target shuffle combines
llvm-svn: 287496
2016-11-20 15:24:38 +00:00
Simon Pilgrim 3f40412e0f [X86][AVX512] Add support for VBMI VPERMV target shuffle combines
llvm-svn: 287495
2016-11-20 15:05:45 +00:00
Simon Pilgrim 9e3f5cc015 [X86][AVX512] Add some initial VBMI target shuffle combine tests
llvm-svn: 287494
2016-11-20 14:45:46 +00:00