openjdk team mailing list archive

Thread
Date

Bug#853758: openjdk-8: Please cherry-pick 8dfbb002197a / fix for JDK-8164293 [memory leak in hotspot]

To: Debian Bug Tracking System <submit@xxxxxxxxxxxxxxx>
From: Niels Thykier <niels@xxxxxxxxxxx>
Date: Tue, 31 Jan 2017 17:31:51 +0100
Delivered-to: submit@xxxxxxxxxxxxxxx
Reply-to: Niels Thykier <niels@xxxxxxxxxxx>, 853758@xxxxxxxxxxxxxxx
Resent-cc: niels@xxxxxxxxxxx, OpenJDK Team <openjdk@xxxxxxxxxxxxxxxxxxx>
Resent-date: Tue, 31 Jan 2017 16:33:02 +0000
Resent-from: Niels Thykier <niels@xxxxxxxxxxx>
Resent-message-id: <handler.853758.B.148588032024626@xxxxxxxxxxxxxxx>
Resent-to: debian-bugs-dist@xxxxxxxxxxxxxxxx

Package: openjdk-8
Version: 8u102-b14.1-1
Severity: serious
Tags: upstream patch
Control: forwarded -1 https://bugs.openjdk.java.net/browse/JDK-8164293

Hi,

Please consider cherry picking the fix for JDK-8164293, which is a
memory leak in the hotspot compiler.  The effect is that the Java
process silently leaks memory on an OS level (but Java itself does not
see it, so it is invisible in APM-based/agent-based monitoring).

My understanding is that it effectively breaks Java-based web
applications and related containers (e.g. tomcat, jetty, jenkins).

I have confirmed via sources.debian.net that the unpatched code is in
the hotspot.tar.xz inside the source package in 8u121-b13-2 and the
version in backports.

Upstream bug: https://bugs.openjdk.java.net/browse/JDK-8164293
Upstream fix: http://hg.openjdk.java.net/jdk8u/jdk8u/hotspot/rev/8dfbb002197a