harmony-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Alexey Varlamov (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HARMONY-3907) [drlvm][gc_gen] noticeable perf degradation on drlvm tests
Date Fri, 18 May 2007 10:45:16 GMT

    [ https://issues.apache.org/jira/browse/HARMONY-3907?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#action_12496837
] 

Alexey Varlamov commented on HARMONY-3907:
------------------------------------------

Xiao-Feng, which category would you suggest then? 
As for the priority, this issue really affects most people developing DRLVM, and I suspect
it may affect Harmony users' experience in unwanted way. If we keep gc_gen the default GC,
it should be outstanding one among neighbour impls in all important aspects, do you agree?
And performance is very important aspect for GC.

> [drlvm][gc_gen] noticeable perf degradation on drlvm tests
> ----------------------------------------------------------
>
>                 Key: HARMONY-3907
>                 URL: https://issues.apache.org/jira/browse/HARMONY-3907
>             Project: Harmony
>          Issue Type: Bug
>          Components: DRLVM
>            Reporter: Alexey Varlamov
>         Assigned To: Xiao-Feng Li
>
> This is a placeholder for tracking general problem:
> I checked my perceived changes in pre-commit testing time after
> Harmony switched to gcv5, comparing with old gc_cc on the following
> test:
> > sh build.sh -DBUILD_CFG=<cfg>  -Dtest.mode=jet test -Dtest.vmargs="-XX:vm.dlls=<gc>"
> Some smoke tests (outofmemory.*, perf.SeveralThreads, etc)
> intermittently fail/hang on gc_gen, so I've run each configuration 2-3
> times. All machines are multicore (>= 2 CPUs) server boxes.
> Ant-reported build time records are listed below:
> avg time in sec, gc_cc / gc_gen
> --------------------------------------------
> Win32 release:    340 / 520
> Win32 debug:      620 / 1400
> Linux32 release:  300 /  1200
> Linux32 debug:    440 / 1000
> Linux64 debug:    550 / 1300
> Indeed the time for testing is increased at least twice with gc_gen!
> The gc_gen timing is probably somewhat better since HARMONY-3875 commit, but still noticeably
worse.
> The perceived degradation is most evident in stressing smoke tests like perf.* and outofmemory.*

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message