RFR(S): 8173335: Improve logging for j.l.ref.reference processing

Aleksey Shipilev shade at redhat.com
Mon Jun 12 16:06:43 UTC 2017


On 06/10/2017 01:57 AM, sangheon wrote:
> CR: https://bugs.openjdk.java.net/browse/JDK-8173335
> webrev: http://cr.openjdk.java.net/~sangheki/8173335/webrev.0

Oh, good! I had to instrument these by hand when optimizing RP paths.

Comments after brief look:

 *) So, the path with NULL executor are also not handling the timer? E.g. CMS:

 5262   if (rp->processing_is_mt()) {
 5263     rp->balance_all_queues();
 5264     CMSRefProcTaskExecutor task_executor(*this);
 5265     rp->enqueue_discovered_references(&task_executor, _gc_timer_cm);
 5266   } else {
 5267     rp->enqueue_discovered_references(NULL);
 5268   }

 *) I would leave "Ref Counts" line as usual for compatibility reasons. Changing
it to "Counts" would force GC log parsers to handle that corner case too.

 *) This may reuse Indents?

   95       out->print("%s", "    ");

 *) Probably makes sense to "hg mv -A" the workerDataArray files to preserve the
Mercurial history -- webrev should say something like "copied from ...", IIRC.

Thanks,
-Aleksey

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <https://mail.openjdk.org/pipermail/hotspot-gc-dev/attachments/20170612/7d842324/signature.asc>


More information about the hotspot-gc-dev mailing list