In preparing the report, I discovered that the regression is actually going from 2.10.9 to 2.10.10, with no improvement in 3.0.3. Hopefully that makes it easier to track down, though quite a bit must have been done for 2.10.10. The installer package goes from 85MB to 102MB.
On Sat, Jan 19, 2013 at 9:34 AM, Rodrigo Kumpera <[email protected]> wrote: > Please file a bug report as this definitely a regression. > > > On Saturday, January 19, 2013, Charles Esterbrook wrote: >> >> Hi, >> >> I have a benchmark for my project which takes the following number of >> milliseconds to run based on System.Diagnostics.Stopwatch, expressed >> in ranges: >> >> Mono JIT compiler version 2.10.9 (tarball Mon May 7 20:25:51 EDT 2012) >> 6888 - 6943 - default gc / boehm >> 5850 - 5932 - sgen >> >> Mono JIT compiler version 3.0.3 (master/39c48d5 Tue Jan 8 12:12:24 EST >> 2013) >> 9815 - 9892 - default gc / boehm >> 8454 - 8502 - sgen >> >> My system is Mac OS X 10.6.8, Intel Xeon. I shut down web browsers, >> cloud services and other apps to reduce interference, and I also ran >> multiple times, hence the ranges. >> >> The slowdown is about 40%. My program does a variety of things >> including loading assemblies, reflection, file i/o, string processing, >> data structures, etc. I have not yet investigated to see if some parts >> are slowing down more than others. >> >> What about other people on this list? If you test the performance of >> your apps between Mono 2.10 and 3.0, is there a change? How much and >> in what direction? >> >> For your convenience: >> https://www.google.com/search?q=.net+stopwatch >> >> >> -- >> Charles Esterbrook >> http://charles-esterbrook.com >> _______________________________________________ >> Mono-list maillist - [email protected] >> http://lists.ximian.com/mailman/listinfo/mono-list _______________________________________________ Mono-list maillist - [email protected] http://lists.ximian.com/mailman/listinfo/mono-list
