From: Robin G. <rob...@gm...> - 2013-06-25 04:35:54
|
There are two strategies that generally work: - Google the name of the source package - Go to the download site for the relevant project and look in their archives. cheers On 22/06/13 03:23, Zhijia Zhao wrote: > Steve, > > I see. Your explanation is helpful! > > As we talked about the older version, I have a naive question. > For "dacapo-2006-10", when I tried to download the benchmark sources, > I found some problems, which probably caused by the out of date links. > > So may I get the source codes at that time (probably Oct 2006) ? > I mean the source codes that were used in the released jar file. > > > Zhijia > > > On Fri, Jun 21, 2013 at 1:16 PM, Stephen Blackburn > <ste...@an...> wrote: >> Hi Zhijia, >> >> The two versions use the same .cnf file mechanism (though I think we made enhancements in the later version, so they may not be identical). >> >> However, of course, the particular settings are generally very different between the different versions of the suite (if you take a benchmark that appears in both versions, you should expect that the two versions of that benchmark will have different command line arguments). Generally we try to keep the wall clock time more or less constant across versions. However, we change the test hardware to use "contemporary" hardware w.r.t. the release date, and of course we change the versions of the software. We then re-calibrate the inputs for the new version to try to roughly match the target running times. >> >> Cheers, >> >> --Steve >> >> ________________________________________ >> From: Zhijia Zhao [zz...@cs...] >> Sent: Saturday, June 22, 2013 3:09 AM >> To: dac...@li... >> Subject: Re: [dacapobench-researchers] Input differences between three inputs (small, default and large) >> >> Thanks, Steve! >> >> I didn't look into the "9.12" version, but an older version "dacapo-2006-10". >> But they should be quite similar in the way of setting the input sizes. >> >> What you said is very correct! Some of them even have "xsmall" or >> "huge" configurations. Interesting. >> >> Zhijia >> >> On Fri, Jun 21, 2013 at 12:40 PM, Stephen Blackburn >> <ste...@an...> wrote: >>> Hi, >>> >>> The definitive answer is always to check the .cnf file for the relevant >>> benchmark. >>> >>> You can do this trivially by pulling the relevant cnf file out of the dacapo >>> jar: >>> >>> % unzip dacapo-9.12-bach.jar cnf/avrora.cnf >>> Archive: [...]/dacapo-9.12-bach.jar >>> inflating: cnf/antlr.cnf >>> >>> or grab cnf files for all benchmarks: >>> >>> % unzip dacapo-9.12-bach.jar "*.cnf" >>> >>> Then you can see precisely what command line arguments are passed to the >>> benchmark for each of the size settings (as well as other info about the >>> benchmarks). For example, if you look at sunflow.cnf, you'll see that in >>> each case the benchmark gets just an integer argument, which is 32, 256 and >>> 512 for the small, default and large cases respectively. If you look at >>> fop.cnf you'll see that there are only small and default cases, and these >>> respectively get fop to process the "readme" and "test" test files. And so >>> on... >>> >>> --Steve >>> >>> ________________________________ >>> From: Zhijia Zhao [zz...@cs...] >>> Sent: Saturday, June 22, 2013 12:55 AM >>> To: dac...@li... >>> Subject: [dacapobench-researchers] Input differences between three inputs >>> (small, default and large) >>> >>> Hi All, >>> >>> I would like to know the how the three inputs are different from each other. >>> >>> After reading the Dacapo website, I found this: >>> "For some benchmarks default and large workloads are identical" >>> >>> But how different they are between the small input and the default one? >>> >>> If you have any reference related, that would be great! >>> >>> Thanks, >>> Dragonfly >>> >>> ------------------------------------------------------------------------------ >>> This SF.net email is sponsored by Windows: >>> >>> Build for Windows Store. >>> >>> http://p.sf.net/sfu/windows-dev2dev >>> _______________________________________________ >>> dacapobench-researchers mailing list >>> dac...@li... >>> https://lists.sourceforge.net/lists/listinfo/dacapobench-researchers >>> >> >> >> -- >> Zhijia Zhao >> PhD Student >> Computer Science Dept., The College of William & Mary, USA >> web: http://www.cs.wm.edu/~zzhao >> >> ------------------------------------------------------------------------------ >> This SF.net email is sponsored by Windows: >> >> Build for Windows Store. >> >> http://p.sf.net/sfu/windows-dev2dev >> _______________________________________________ >> dacapobench-researchers mailing list >> dac...@li... >> https://lists.sourceforge.net/lists/listinfo/dacapobench-researchers >> >> >> >> ------------------------------------------------------------------------------ >> This SF.net email is sponsored by Windows: >> >> Build for Windows Store. >> >> http://p.sf.net/sfu/windows-dev2dev >> _______________________________________________ >> dacapobench-researchers mailing list >> dac...@li... >> https://lists.sourceforge.net/lists/listinfo/dacapobench-researchers > > |