Rustem Rafikov wrote: > One possible reason of degradation is increasing size of some HashMap's > methods. So those methods are not inlined, but at the same time inlining is > necessary for scalar replacement. So we perhaps need to tune the inliner's > parameters to make it working. I agree that we need to learn it more. > But let's revert the patch for a while. Hope I can drill into the new > HashMap on the next week.
I've reverted the change at r577715, but I would definitely be iterested to help figure out why it causes such a large degradation in performance. > Another question I have is what VM did you use to run jbb2005 that showed an > improvement? The IBM VME (with the caveat that I have a different version to the usual one). Regards, Tim
