projects / platform / upstream / llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 67a4f22) | patch
Revert r211771. It was: "[X86] Improve the selection of SSE3/AVX addsub instructions".
authorAndrea Di Biagio <Andrea_DiBiagio@sn.scee.net>
Wed, 23 Jul 2014 11:20:24 +0000 (11:20 +0000)
committerAndrea Di Biagio <Andrea_DiBiagio@sn.scee.net>
Wed, 23 Jul 2014 11:20:24 +0000 (11:20 +0000)
commit842355e900c41814c73b0c0d3083b2783f1bbd6a
tree594b49f44a060fe9dba279f4c2a8e97568ffffd2tree | snapshot
parent67a4f22f12b07c48f1779b3716f6b369c8a1a830commit | diff
Revert r211771. It was: "[X86] Improve the selection of SSE3/AVX addsub instructions".

This chang fully reverts r211771.
That revision added a canonicalization rule which has the potential to causes a
combine-cycle in the target-independent canonicalizing DAG combine.

The plan is to move the logic that forms target specific addsub nodes as part of
the lowering of shuffles.

llvm-svn: 213736
llvm/lib/Target/X86/X86ISelLowering.cpp diff | blob | history
llvm/test/CodeGen/X86/sse3-avx-addsub.ll diff | blob | history
Domain: System / Toolchain;