Fwd: Re: Better default for ParallelGCThreads and ConcGCThreads by using number of physical cores and CPU mask.

Jon Masamitsu jon.masamitsu at oracle.com
Fri Nov 22 23:51:36 UTC 2013


Jungwoo,

I forwarded the webrev to the runtime mailing list and got this
back.  In case you are not on that mailing list, here it is.

Jon


-------- Original Message --------
Subject: 	Re: Better default for ParallelGCThreads and ConcGCThreads by 
using number of physical cores and CPU mask.
Date: 	Fri, 22 Nov 2013 09:58:18 -0800
From: 	Brandon Mitchell <brandon at twitter.com>
To: 	Jon Masamitsu <jon.masamitsu at oracle.com>
CC: 	hotspot-runtime-dev at openjdk.java.net



On Fri, Nov 22, 2013 at 9:24 AM, Jon Masamitsu <jon.masamitsu at oracle.com 
<mailto:jon.masamitsu at oracle.com>> wrote:

    This is a contribution regarding the number of GC worker threads to
    use.  Part of the change queries /proc on linux to get the number of
    active cores on the platform.  The changes are in

    http://cr.openjdk.java.net/~jmasa/8028554/webrev.00/hotspot/src/os_cpu/linux_x86/vm/os_linux_x86.cpp.frames.html
    <http://cr.openjdk.java.net/%7Ejmasa/8028554/webrev.00/hotspot/src/os_cpu/linux_x86/vm/os_linux_x86.cpp.frames.html>

    Can someone familiar with this code take a look to see
    if it is reasonable and done in a way that is consistent
    with other /proc queries.


/sys/devices/system/cpu/cpu<N>/topology/thread_siblings_list might be 
somewhat easier to parse for this information.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.java.net/pipermail/hotspot-gc-dev/attachments/20131122/e156e6bc/attachment.htm>


More information about the hotspot-gc-dev mailing list