diff options
| author |  Justin Lebar <jlebar@google.com> | 2017-02-07 10:56:37 -0800 |
|---|---|---|
| committer |  TensorFlower Gardener <gardener@tensorflow.org> | 2017-02-07 11:17:29 -0800 |
| commit | 7135d08d4e6067865d7b5f2907013c960a12ae4f (patch) | |
| tree | 1bb78bc46cb823fe40121f7b77da10020520042a /BUILD | |
| parent | 73899707ae5f1b35c88bcb5898d8bb05e7b69715 (diff) | |
[XLA] Speed up ReferenceUtil::ConvArray4DGeneralDimensionsDilated.
Avoid 64-bit divides where possible. Per Agner Fog's manuals, 64-bit
idiv is at least 4x slower than 32-bit idiv. perf confirms that the
idivs are on the critical path.
http://www.agner.org/optimize/instruction_tables.pdf
Change: 146806085
Diffstat (limited to 'BUILD')
0 files changed, 0 insertions, 0 deletions