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.
[dev-20021023, Sun JDK 1.4.1_01, MDI] 1. mounted a local filesystem and jar into the repository. 2. was about to add another jar file 3. right-clicked on the Filesystems node, selected Mount 4. moved mouse over the Mount submenu 5. before I could select the Jar File item, IDE deadlocked Thread dump will be attached shortly. Two suspicious threads seem to be doing something with VCS. Note that I didn't have any VCS filesystems mounted and I didn't want to mount one. Feel free to re-assign if not in VCS.
Created attachment 7750 [details] Tread dump
I would argue that this is P2. According to http://qa.netbeans.org/processes/bug_priority_guidelines.html it should happen fairly often, which is not the case right ? Would you mind Honzo to downgrade it ? Isn't it rather performance issue ? What do you think Martin ?
Yes, I would mind downgrading it to P3. According to the guidelines, a reliably reproducible deadlock is a P1. Intermittent deadlock is a P2. I have no evidence how frequently it occurs, but the facts that it occured once and a thread dump is available should be good reasons to get if fixed quickly. After all, every deadlock is a potential data loss.
This is basically a duplicate of issue #21316. The problem is, that FS commands are stored in Nodes hierarchy. The performance problem could be, that somebody is trying to instantiate my filesystem. I would say, that this is a problem of core. Should we file a PERFORMANCE defect for this? *** This issue has been marked as a duplicate of 21316 ***
*** Issue 28207 has been marked as a duplicate of this issue. ***
*** Issue 28205 has been marked as a duplicate of this issue. ***
I have seen this yesterday too, which means that fixing #21316 didn't help. If you could Martin file something against core, I would appreciate it. Thanks ! Verified in development build #200211080100 of NetBeans 4.0.
> I have seen this yesterday too, which means that fixing #21316 > didn't help. What do you mean by 'this'? The exactly same stack trace? I don't believe. Can you attach the full thread dump please? > If you could Martin file something against core, I would appreciate > it. Something?? It did not hapen to me! ;-) So I don't know where the problem is.
By "this" I mean a deadlock while opening "Filesystems|Mount ..." popup menu. I am sorry, I didn't manage to capture the thread dump. I'll do this next time I see it. :)