diff options
| author | Shuxin Yang <shuxin.llvm@gmail.com> | 2013-05-09 18:34:27 +0000 |
|---|---|---|
| committer | Shuxin Yang <shuxin.llvm@gmail.com> | 2013-05-09 18:34:27 +0000 |
| commit | 1d8d7e4d38da4105ae8ca44486f9971bb8e59eda (patch) | |
| tree | 33d45a90e1200a4263dc8a14ed52a7529f77162d /llvm/lib/Transforms/Vectorize | |
| parent | 8797d39fd9e60d06b72ea3e28e3d12fa2e709bce (diff) | |
| download | bcm5719-llvm-1d8d7e4d38da4105ae8ca44486f9971bb8e59eda.tar.gz bcm5719-llvm-1d8d7e4d38da4105ae8ca44486f9971bb8e59eda.zip | |
[GVN] Split critical-edge on the fly, instead of postpone edge-splitting to next
iteration.
This on step toward non-iterative GVN. My local hack suggests that getting rid
of iteration will speedup GVN by 30%+ on a medium sized input (2k LOC, C++).
I cannot explain why not 2x or more at this moment.
llvm-svn: 181532
Diffstat (limited to 'llvm/lib/Transforms/Vectorize')
0 files changed, 0 insertions, 0 deletions
