This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
This bug was originally marked as duplicate of bug 186348, that is already resolved. This bug is still valid, so this seems to be another bug, but it might be related. Build: NetBeans IDE Dev (Build 20111011-dfcb3e627c5f) VM: OpenJDK 64-Bit Server VM, 20.0-b11, OpenJDK Runtime Environment, 1.6.0_23-b23 OS: Linux User Comments: yardus: Editing build.xml in a freeform project Maximum slowness yet reported was 4018 ms, average is 4018
Created attachment 111851 [details] nps snapshot
There is an open input stream blocking getOutputStream. Probably somebody forget to close it and it waits for finalization (but I don't see the close of inputstream in the snapshot at all).
This problem looks rather serious (I mean the aspect that I can't save the file) - I could not find any workaround, not even restarting the IDE. Basically, I can not edit this build.xml in the IDE.
There is an input stream open on build.xml but there is no way from the log file to tell where that is, so there is no way to know what code is actually responsible. (Does masterfs record this information in a way that the diagnostics could be better?) Is there a particular build.xml that this always happens with? Any other information to reproduce?
More meaningful summary. The GUI freeze is just a side effect of there being a read lock (and of SaveAction being called in EQ).
I think MutualExclusionSupport should hold a stack trace when assertions are enabled, adding the lock owner(s) as a root cause to any IOException it throws.
Finally found a (temporarily) reproducible case, and managed to add useful diagnostics.
f3a8eecb1dbd seems to have been responsible.
core-main #2183ba37266d
Integrated into 'main-golden' Changeset: http://hg.netbeans.org/main-golden/rev/2183ba37266d User: Jesse Glick <jglick@netbeans.org> Log: #203508: Read lock held open on build.xml
*** Bug 201007 has been marked as a duplicate of this bug. ***