| commit | b619dd5d5b69ba9f4571a96e1a96d09d8aed03a7 | [log] [tgz] |
|---|---|---|
| author | Benjamin Kramer <benny.kra@googlemail.com> | Fri Jul 12 18:06:44 2013 +0000 |
| committer | Benjamin Kramer <benny.kra@googlemail.com> | Fri Jul 12 18:06:44 2013 +0000 |
| tree | 79411ae481a896f857d5a454961479a75289eac9 | |
| parent | 2494dfcf1795a9bd28030d04cf2fb15e5c3b0f33 [diff] |
X86: Shrink certain forms of movsx. In particular: movsbw %al, %ax --> cbtw movswl %ax, %eax --> cwtl movslq %eax, %rax --> cltq According to Intel's manual those have the same performance characteristics but come with a smaller encoding. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@186174 91177308-0d34-0410-b5e6-96231b3b80d8