diff options
| author | Elena Demikhovsky <elena.demikhovsky@intel.com> | 2015-06-16 06:07:24 +0000 |
|---|---|---|
| committer | Elena Demikhovsky <elena.demikhovsky@intel.com> | 2015-06-16 06:07:24 +0000 |
| commit | 77f0e9f662c3eef7d742ad9c7f60b31aa9037c96 (patch) | |
| tree | 76d405c264cfca551e48429bd26a2d348f50c833 /llvm/lib/CodeGen/MachineSink.cpp | |
| parent | e5409783d1e13f8d7ac487aa790774b5b5a95196 (diff) | |
| download | bcm5719-llvm-77f0e9f662c3eef7d742ad9c7f60b31aa9037c96.tar.gz bcm5719-llvm-77f0e9f662c3eef7d742ad9c7f60b31aa9037c96.zip | |
X86: optimized i64 vector multiply with constant
When we multiply two 64-bit vectors, we extract lower and upper part and use the PMULUDQ instruction.
When one of the operands is a constant, the upper part may be zero, we know this at compile time.
Example: %a = mul <4 x i64> %b, <4 x i64> < i64 5, i64 5, i64 5, i64 5>.
I'm checking the value of the upper part and prevent redundant "multiply", "shift" and "add" operations.
llvm-svn: 239802
Diffstat (limited to 'llvm/lib/CodeGen/MachineSink.cpp')
0 files changed, 0 insertions, 0 deletions
