Vitaly Davidovich wrote:
I figured you did, but wanted to check. :)
So the perf hit was with c2 compilation? Were you able to check the
assembly (or enable inlining printing in hotspot) and see that lack of
inlining (and whatever further opto that enabled) was the difference by
simply adding a local or two? I'm legitimately curious here because if
that's the case and this was on a somewhat recent hotspot build, it sort
of goes against what TomR seemed to have been saying.
The underlying issue is that code size is one of the criteria
that JITs use to decide to compile/inline etc. So long as they do
so, there will be cases here and there where it critically
important to keep sizes small in bottleneck code. Not many,
but still enough for me to object to efforts that might
blindly increase code size for the sake of warnings cleanup.
-Doug