projects / oota-llvm.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 9e5329d) | patch
Some x86 instructions can load/store one of the operands to memory. On SSE, this...
authorNadav Rotem <nrotem@apple.com>
Mon, 24 Dec 2012 09:40:33 +0000 (09:40 +0000)
committerNadav Rotem <nrotem@apple.com>
Mon, 24 Dec 2012 09:40:33 +0000 (09:40 +0000)
commitace0c2fad7c581367cc2519e1d773bca37fc9fec
tree74ec5cfe661003e97716f8c9ef848ef45ff58ed9tree | snapshot
parent9e5329d77e590f757dbd8384f418e44df9dbf91acommit | diff
Some x86 instructions can load/store one of the operands to memory. On SSE, this memory needs to be aligned.
When these instructions are encoded in VEX (on AVX) there is no such requirement. This changes the folding
tables and removes the alignment restrictions from VEX-encoded instructions.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@171024 91177308-0d34-0410-b5e6-96231b3b80d8
lib/Target/X86/X86InstrInfo.cpp diff | blob | history
test/CodeGen/X86/fold-vex.ll [new file with mode: 0644] blob
C/C++ LLVM-based compilers that forbids OOTA behaviors