projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 9e13cbc) | patch
Use movaps instead of movups to spill 16-byte vector values when default alignment...
authorEvan Cheng <evan.cheng@apple.com>
Mon, 21 Jul 2008 06:34:17 +0000 (06:34 +0000)
committerEvan Cheng <evan.cheng@apple.com>
Mon, 21 Jul 2008 06:34:17 +0000 (06:34 +0000)
commit41c08405839c49fc6d981c024fc03efb0309acc6
treef77dac4640fe4aa7a14b08d1fc70c4e5a7b1b7eatree | snapshot
parent9e13cbc171f4c5c3f74b097243eff7cb0b3b942bcommit | diff
Use movaps instead of movups to spill 16-byte vector values when default alignment is >= 16. This fixes some massive performance regressions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@53844 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86InstrInfo.cpp diff | blob | history
C/C++ LLVM-based compilers that forbids OOTA behaviors