diff options
| author | Arnold Schwaighofer <aschwaighofer@apple.com> | 2013-04-18 17:22:34 +0000 |
|---|---|---|
| committer | Arnold Schwaighofer <aschwaighofer@apple.com> | 2013-04-18 17:22:34 +0000 |
| commit | 4cd6aa110c3c96b6c8aadb4fbeeb0b80a0403586 (patch) | |
| tree | 8936885d99432e165caab4b7917490102d6a2a91 /lldb/source/Commands/CommandObjectBreakpointCommand.cpp | |
| parent | 9624985b1a0843454377355f1f34056baa23ceff (diff) | |
| download | llvm-4cd6aa110c3c96b6c8aadb4fbeeb0b80a0403586.zip llvm-4cd6aa110c3c96b6c8aadb4fbeeb0b80a0403586.tar.gz llvm-4cd6aa110c3c96b6c8aadb4fbeeb0b80a0403586.tar.bz2 | |
LoopVectorizer: Recognize min/max reductions
A min/max operation is represented by a select(cmp(lt/le/gt/ge, X, Y), X, Y)
sequence in LLVM. If we see such a sequence we can treat it just as any other
commutative binary instruction and reduce it.
This appears to help bzip2 by about 1.5% on an imac12,2.
radar://12960601
llvm-svn: 179773
Diffstat (limited to 'lldb/source/Commands/CommandObjectBreakpointCommand.cpp')
0 files changed, 0 insertions, 0 deletions