NEW 153797
Implement LLVM's greedy register allocator in Air
https://bugs.webkit.org/show_bug.cgi?id=153797
Summary Implement LLVM's greedy register allocator in Air
Filip Pizlo
Reported 2016-02-02 12:29:16 PST
Current performance data suggests that LLVM's Greedy allocator runs slightly faster than IRC. We can probably make IRC faster, but since everything in LLVM is systematically slower than in B3, it's a pretty good bet than our implementation of Greedy would be faster than LLVM's.
Attachments
Note You need to log in before you can comment on or make changes to this bug.