Hi On 30 August 2011 14:05, Stefan Kamphausen <ska2...@googlemail.com> wrote: > Hi, > > using a current git checkout of master (git describe says > 'clojure-1.3.0-beta2-2-g0689c47') I ran the test-code from > http://clojure.org/refs like this: > > (time (run 100 10 10 1000000)) > "Elapsed time: 18413.314789 msecs" > > Actually I ran that code several times until I would see no improvement > anymore. > > After that I removed (hopefully) all references to Ref.TVal.msec, removed > the calls to System.currentTimeMillis() changed the constructors and the > instantiations. The result seems to be a 4-5% speedup: > > "Elapsed time: 17395.048357 msecs" > > (Again, running it several times til no optimizations seem to kick in > anymore). > > Is this interesting for the developers? Should I try to come up with more > test-scenarios for the various conflicting situations that can arise in > STM? Shall I fork on github and add my changes there?
As far as the github question is concerned, I believe the answer is "No". See: http://clojure.org/contributing http://clojure.org/patches Otherwise it seems interesting :) > Cheers, > Stefan > > PS: Ubuntu 10.04 on a "Pentium(R) Dual-Core CPU E6300 @ 2.80GHz" (from > /proc/cpuinfo) with "MemTotal: 8127764 kB" (/proc/meminfo) using 'java > version "1.6.0_26"' -- Michael Wood <esiot...@gmail.com> -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
