I don't think we need to complicate the code to improve optimization with older compilers. People who really care about performance will use newer compilers anyway. So let's focus on 0001-dfa-speed-up-at-initial-state.patch.

I confirmed your performance results on my platform: the change makes test case "k" about 60% faster and test case "l" about 14% slower. At this point we're just in bugfix mode so this patch can wait. It'd be nice if it didn't slow down the short-line case so much. Also, I wonder how well this change works for "typical" searches. (I realize that "typical" is hard to define....)



Reply via email to