RFR (XS) 8215221: Serial GC misreports young GC time

Man Cao manc at google.com
Wed Mar 6 18:30:21 UTC 2019


Thanks, Aleksey. The new webrev looks good.

-Man

On Wed, Mar 6, 2019 at 4:40 AM Aleksey Shipilev <shade at redhat.com> wrote:

> Hi Man,
>
> Thanks for taking a look. New webrev:
>   http://cr.openjdk.java.net/~shade/8215221/webrev.03/
>
> Still passes hotspot tier1, eyeballing Dacapo logs, jdk-submit is running.
>
>
> On 2/22/19 4:43 AM, Man Cao wrote:
> > I found the following could fix the problem (yes, it's ugly, but I
> couldn't find cleaner way that
> > works in all cases):
> >
> > if (!do_young_collection) {
> >   gc_prologue(complete);
> >   increment_total_collections(complete);
> > }
> > if (!complete) {
> >   increment_total_full_collections();
> > }
>
> I see, dang. Rewritten like this:
>
>     if (!do_young_collection) {
>       gc_prologue(complete);
>       increment_total_collections(complete);
>     }
>
>     // Accounting quirk: total full collections would be incremented when
> "complete"
>     // is set, by calling increment_total_collections above. However, we
> also need to
>     // account Full collections that had "complete" unset.
>     if (!complete) {
>       increment_total_full_collections();
>     }
>
> > I also found that CollectedHeap._total_collections seems wrong with or
> without this change, for
> > SerialGC and CMS. It is smaller than the actual number of collections
> according to the GC ID.
> > Reproduction recipe is the same as above.
> Yeah, that minor thing needs to be fixed too, in a separate change.
>
> -Aleksey
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.java.net/pipermail/hotspot-gc-dev/attachments/20190306/b9cf175e/attachment.html>


More information about the hotspot-gc-dev mailing list