>Which suggests return continuation pmc driven memory management costs >(gc and allocation) are currently a major, perhaps even dominant, >factor in method invocation speed.
Yow. Okay, thanks. That means it's time to dive into the think tank and see what we can do about that. Nothing like finally thinking of caveats _after_ hitting send. oofib as a program is generating atypically(?) little garbage. So much of the garbage is continuation pmcs, which thus determine the frequency of costly gc sweeps. In a dirtier, more typical environment, where perhaps other factors are determining the frequency of gc sweeps, the gc cost would look more like the pmc's _incremental_ cost of collection. Maybe. Which would be less "yow". How much less isn't clear. Though given that allocating these pmc's seems to be itself say 1/3 of method call cost without considering gc, they would retain at least that level of yow. Mitchell (btw, regards ICU progress... awesome!!:)