15.07.2013, 07:31, "John Regehr" <[email protected]>: >> There's nothing really evil in pass_peep::b, it just should never be run on >> files >> larger than 30K. I have not measured performance of rm-toks-1 on such files, >> but I guess it scales much better. > > I looked at pass_peep today. > > Much of its functionality can be replaced by rm-toks-1 and rm-toks-2. > > However, pass_peep also tries to replace identifiers and constants with > 0 and 1, and there currently does not exist a token-based pass for doing > that. As soon as I write that pass, I'll move both peep::a and peep::b > to the --slow mode.
peep::a is not harmful for speed, but it does much les work then b ususally. -- Regards, Konstantin
