yes I discovered the same behavior yesterday with esteban.

thanks very much for the details!

On 2013-06-14, at 23:53, Sven Van Caekenberghe <[email protected]> wrote:
> On 07 Jun 2013, at 17:15, Camillo Bruni <[email protected]> wrote:
> 
>> Can you try on the command line only?
> 
> This is getting _very_ weird:
> 
> This is on Ubuntu 12.10 x64 running on a non-virtualised Intel(R) Core(TM) 
> i5-3470 CPU @ 3.20GHz, 8GB RAM.
> 
> $ ./pharo Pharo.image printVersion
> [version] 3.0 #30199
> 
> -1-
> 
> $ curl get.pharo.org/vm| bash
> 
> $ ./pharo -version
> 3.9-7 #1 Wed Mar 13 18:22:44 CET 2013 gcc 4.4.3
> NBCoInterpreter NativeBoost-CogPlugin-EstebanLorenzano.18 uuid: 
> a53445f9-c0c0-4015-97a3-be7db8d9ed6b Mar 13 2013
> NBCogit NativeBoost-CogPlugin-EstebanLorenzano.18 uuid: 
> a53445f9-c0c0-4015-97a3-be7db8d9ed6b Mar 13 2013
> git://gitorious.org/cogvm/blessed.git Commit: 
> 412abef33cbed05cf1d75329e451d71c0c6aa5a7 Date: 2013-03-13 17:48:50 +0100 By: 
> Esteban Lorenzano <[email protected]> Jenkins build #14535
> Linux linux-ubuntu-10 2.6.32-38-server #83-Ubuntu SMP Wed Jan 4 11:26:59 UTC 
> 2012 x86_64 GNU/Linux
> plugin path: /home/bt/pharo/pharo-vm/ [default: /home/bt/pharo/pharo-vm/]
> 
> $ ./pharo Pharo.image eval '1 tinyBenchmarks'
> '1276807980 bytecodes/sec; 172715872 sends/sec'
> 
> $ ./pharo Pharo.image eval '[ SystemNavigation new allCallsOn: #ifTrue: ] 
> timeToRun'
> 59
> 
> $ ./pharo Pharo.image eval '[ SystemNavigation new allCallsOn: #ifTrue: ] 
> timeToRun'
> 59
> 
> -2-
> 
> $ curl get.pharo.org/vmLatest | bash
> 
> $ ./pharo --version
> 3.9-7 #1 Fri May 31 13:43:31 CEST 2013 gcc 4.6.3
> NBCoInterpreter NativeBoost-CogPlugin-GuillermoPolito.19 uuid: 
> acc98e51-2fba-4841-a965-2975997bba66 May 31 2013
> NBCogit NativeBoost-CogPlugin-GuillermoPolito.19 uuid: 
> acc98e51-2fba-4841-a965-2975997bba66 May 31 2013
> git://gitorious.org/cogvm/blessed.git Commit: 
> 0f562e0afe14e533470278943b5d65d42ea133c2 Date: 2013-05-30 16:38:08 +0200 By: 
> Igor Stasenko <[email protected]> Jenkins build #14595
> Linux pharo-linux 3.2.0-31-generic-pae #50-Ubuntu SMP Fri Sep 7 16:39:45 UTC 
> 2012 i686 i686 i386 GNU/Linux
> plugin path: /home/bt/pharo/pharo-vm/ [default: /home/bt/pharo/pharo-vm/]
> 
> $ ./pharo Pharo.image printVersion
> [version] 3.0 #30199
> 
> $ ./pharo Pharo.image eval '1 tinyBenchmarks'
> '1335071707 bytecodes/sec; 179533604 sends/sec'
> 
> $ ./pharo Pharo.image eval '[ SystemNavigation new allCallsOn: #ifTrue: ] 
> timeToRun'
> 705
> 
> $ ./pharo Pharo.image eval '[ SystemNavigation new allCallsOn: #ifTrue: ] 
> timeToRun'
> 697
> 
> 
> So #tinyBenchmarks reports approx. the same result, but #allCallsOn: differs 
> by a factor 10. Why ?
> 
> Sven
> 
> PS: We really need to start tracking benchmarks across platforms and versions 
> - so much to do.

actually https://ci.inria.fr/pharo/view/VM/job/PharoVM-tests/ succesfully 
detected the degeneration and thus failed.
And yes, we will add real benchmarks at some point. But for that we first need 
dedicated machines, as running them on a virtualized environment doesn't make 
much sense :/.


Reply via email to