[ENH] MsecClockValueOptimization-nk ( [sm][et] small,
straightforward, recommend approval )
cdshaffer at acm.org
cdshaffer at acm.org
Wed Mar 10 22:01:34 CET 2004
Ned just manually "inlined" a primitive callout. Seems safe enough. I
wouldn't really expect the performance impact of this improvement to be
very easily measured with existing benchmarks but I would feel remiss if
I didn't at least run them. Using the Benchmarks package on SqueakMap:
Slopstone and Smopstone showed no real improvement, I couldn't get the
MacroBenchmarks to run in 5816 and don't have time to sort it out now.
Recommend approval.
More information about the Squeak-harvest
mailing list