Backporting 8020983: OutOfMemoryError caused by non garbage collected JPEGImageWriter Instances
Omair Majid
omajid at redhat.com
Tue Aug 27 09:06:42 PDT 2013
On 08/27/2013 12:01 PM, Andrew Hughes wrote:
> ----- Original Message -----
>> Hi Andrew,
>>
>> On 08/27/2013 11:22 AM, Andrew Hughes wrote:
>>> 0. What problem does this solve?
>>
>> The current code has memory leaks. With some usage patterns, Java will
>> run heap space. For more details, please see the bug:
>> http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=8020983
>>
>
> "This image writer object will never be garbage collected. This is reproducible with java 6u45. It works fine for us until version 6_u43."
>
> It sounds like a recent regression. Have you been able to reproduce the issue
> on OpenJDK 6?
Yes, it was originally reported against icedtea6-1.11:
https://bugzilla.redhat.com/show_bug.cgi?id=976897
> Ok, it should probably also go in 1.11.x and 1.12.x if the issue is present there.
Will do.
>>> 3. Where has the bug ID information gone from the backported changeset?
>>
>> webrevs can be generated against patches as well as changesets. This one
>> is not generated against a changeset (yet) since a simple hg
>> export/import did not work.
>
> I'm well aware of that, but I can't approve it unless it's in its final form.
> I want to make sure the original information isn't being lost.
I see. I will ask Mario to create a final webrev, then.
Thanks,
Omair
--
PGP Key: 66484681 (http://pgp.mit.edu/)
Fingerprint = F072 555B 0A17 3957 4E95 0056 F286 F14F 6648 4681
More information about the jdk6-dev
mailing list