Comment 1 Szymon Ptaszkiewicz 2014-01-23 12:05:41 EST

And I keep getting such errors every 5 minutes.

Comment 2 Szymon Ptaszkiewicz 2014-01-23 12:24:28 EST

When I finally decided to restart my Eclipse to get rid of the annoying periodic popup dialog I received one more NPE:

java.lang.NullPointerException
	at org.eclipse.ui.texteditor.AbstractTextEditor.saveState(AbstractTextEditor.java:7080)
	at org.eclipse.ui.internal.EditorReference.getEditorState(EditorReference.java:177)
	at org.eclipse.ui.internal.EditorReference.persist(EditorReference.java:99)
	at org.eclipse.ui.internal.Workbench$15.run(Workbench.java:1146)
	at org.eclipse.core.runtime.SafeRunner.run(SafeRunner.java:42)
	at org.eclipse.ui.internal.Workbench.persist(Workbench.java:1135)
	at org.eclipse.ui.internal.Workbench.busyClose(Workbench.java:1078)
	at org.eclipse.ui.internal.Workbench.access$19(Workbench.java:1029)
	at org.eclipse.ui.internal.Workbench$19.run(Workbench.java:1332)
	at org.eclipse.swt.custom.BusyIndicator.showWhile(BusyIndicator.java:70)
	at org.eclipse.ui.internal.Workbench.close(Workbench.java:1330)
	at org.eclipse.ui.internal.Workbench.restart(Workbench.java:2508)
	at org.eclipse.ui.internal.handlers.RestartWorkbenchHandler.execute(RestartWorkbenchHandler.java:31)
	at org.eclipse.ui.internal.handlers.HandlerProxy.execute(HandlerProxy.java:290)
	at org.eclipse.ui.internal.handlers.E4HandlerProxy.execute(E4HandlerProxy.java:90)
	at sun.reflect.GeneratedMethodAccessor36.invoke(Unknown Source)
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:37)
	at java.lang.reflect.Method.invoke(Method.java:611)
	at org.eclipse.e4.core.internal.di.MethodRequestor.execute(MethodRequestor.java:56)
	at org.eclipse.e4.core.internal.di.InjectorImpl.invokeUsingClass(InjectorImpl.java:243)
	at org.eclipse.e4.core.internal.di.InjectorImpl.invoke(InjectorImpl.java:224)
	at org.eclipse.e4.core.contexts.ContextInjectionFactory.invoke(ContextInjectionFactory.java:132)
	at org.eclipse.e4.core.commands.internal.HandlerServiceHandler.execute(HandlerServiceHandler.java:163)
	at org.eclipse.core.commands.Command.executeWithChecks(Command.java:499)
	at org.eclipse.core.commands.ParameterizedCommand.executeWithChecks(ParameterizedCommand.java:508)
	at org.eclipse.e4.core.commands.internal.HandlerServiceImpl.executeHandler(HandlerServiceImpl.java:222)
	at org.eclipse.e4.ui.workbench.renderers.swt.HandledContributionItem.executeItem(HandledContributionItem.java:776)
	at org.eclipse.e4.ui.workbench.renderers.swt.HandledContributionItem.handleWidgetSelection(HandledContributionItem.java:668)
	at org.eclipse.e4.ui.workbench.renderers.swt.HandledContributionItem.access$6(HandledContributionItem.java:652)
	at org.eclipse.e4.ui.workbench.renderers.swt.HandledContributionItem$4.handleEvent(HandledContributionItem.java:584)
	at org.eclipse.swt.widgets.EventTable.sendEvent(EventTable.java:84)
	at org.eclipse.swt.widgets.Display.sendEvent(Display.java:4353)
	at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1061)
	at org.eclipse.swt.widgets.Display.runDeferredEvents(Display.java:4172)
	at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3761)
	at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine$9.run(PartRenderingEngine.java:1122)
	at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:332)
	at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine.run(PartRenderingEngine.java:1006)
	at org.eclipse.e4.ui.internal.workbench.E4Workbench.createAndRunUI(E4Workbench.java:146)
	at org.eclipse.ui.internal.Workbench$5.run(Workbench.java:612)
	at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:332)
	at org.eclipse.ui.internal.Workbench.createAndRunWorkbench(Workbench.java:565)
	at org.eclipse.ui.PlatformUI.createAndRunWorkbench(PlatformUI.java:150)
	at org.eclipse.ui.internal.ide.application.IDEApplication.start(IDEApplication.java:125)
	at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:196)
	at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:109)
	at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:80)
	at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:372)
	at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:226)
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:60)
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:37)
	at java.lang.reflect.Method.invoke(Method.java:611)
	at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:636)
	at org.eclipse.equinox.launcher.Main.basicRun(Main.java:591)
	at org.eclipse.equinox.launcher.Main.run(Main.java:1450)

Comment 3 Dani Megert 2014-01-28 05:42:50 EST

Szymon I cannot reproduce this. Can you provide steps?

Comment 4 Szymon Ptaszkiewicz 2014-01-28 06:19:58 EST

(In reply to Dani Megert from comment #3)
> Szymon I cannot reproduce this. Can you provide steps?

Unfortunately not. It happened just once I don't know what caused this. I guess that for some reason the UI model was in a bad state because it looks like it kept a reference to an editor after all editors were closed and when running periodic workbench save it tried to persist this editor. Restarting Eclipse solved the problem so it's not a big problem.

I'm fine with closing this bug as WORKSFORME, but I it's just worth to know that such NPEs happened (at least on my machine).

Comment 5 Dani Megert 2014-01-28 06:23:39 EST

(In reply to Szymon Ptaszkiewicz from comment #4)
> I'm fine with closing this bug as WORKSFORME, but I it's just worth to know
> that such NPEs happened (at least on my machine).

Right. I think we're closing this for now. As you pointed out, the problem would most likely be in the UI model and not in Text.

Comment 6 Martin Mathew 2014-05-26 04:28:16 EDT

Created attachment 243474 [details]
Error Log to check the time stamp of the exception

Issue was reproducible in 4.4 RC2 Build id: I20140522-1330

Steps to reproduce:
Follow the steps in bug 435740 which results in IAE. 
Just wait, every 5 minutes after the IAE is thrown the NPE was consistently thrown:
java.lang.NullPointerException
	at org.eclipse.ui.texteditor.AbstractTextEditor.saveState(AbstractTextEditor.java:7082)
	at org.eclipse.ui.internal.EditorReference.getEditorState(EditorReference.java:177)
	at org.eclipse.ui.internal.EditorReference.persist(EditorReference.java:99)
	at org.eclipse.ui.internal.Workbench$15.run(Workbench.java:1201)
	at org.eclipse.core.runtime.SafeRunner.run(SafeRunner.java:42)
	at org.eclipse.ui.internal.Workbench.persist(Workbench.java:1189)
	at org.eclipse.ui.internal.Workbench.access$46(Workbench.java:1187)
	at org.eclipse.ui.internal.Workbench$61.runInUIThread(Workbench.java:2951)
	at org.eclipse.ui.progress.UIJob$1.run(UIJob.java:97)
	at org.eclipse.swt.widgets.RunnableLock.run(RunnableLock.java:35)
	at org.eclipse.swt.widgets.Synchronizer.runAsyncMessages(Synchronizer.java:136)
	at org.eclipse.swt.widgets.Display.runAsyncMessages(Display.java:4147)
	at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3764)
	at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine$9.run(PartRenderingEngine.java:1152)
	at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:332)
	at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine.run(PartRenderingEngine.java:1033)
	at org.eclipse.e4.ui.internal.workbench.E4Workbench.createAndRunUI(E4Workbench.java:148)
	at org.eclipse.ui.internal.Workbench$5.run(Workbench.java:636)
	at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:332)
	at org.eclipse.ui.internal.Workbench.createAndRunWorkbench(Workbench.java:579)
	at org.eclipse.ui.PlatformUI.createAndRunWorkbench(PlatformUI.java:150)
	at org.eclipse.ui.internal.ide.application.IDEApplication.start(IDEApplication.java:135)
	at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:196)
	at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:134)
	at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:104)
	at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:379)
	at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:233)
	at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
	at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62)
	at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43)
	at java.lang.reflect.Method.invoke(Method.java:483)
	at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:648)
	at org.eclipse.equinox.launcher.Main.basicRun(Main.java:603)
	at org.eclipse.equinox.launcher.Main.run(Main.java:1465)

Comment 7 Martin Mathew 2014-05-26 04:30:36 EDT

It could be related to bug 435740.

Comment 8 Dani Megert 2014-05-26 09:54:57 EDT

(In reply to Manju Mathew from comment #7)
> It could be related to bug 435740.

This just makes a big leak visible: each time an editor is split (vertically or horizontally), a new editor instance is created. However, once that editor is closed it will only dispose two editors if still split or one editor if no longer in split state.

Those "dead" editors will be updated in the background e.g. when persisting the state every 5 minutes.

If the perspective is then closed, lots of errors will happen, like e.g. the reported NPE for each "dead" editor every 5 minutes.

Comment 9 Eric Moffatt 2014-05-26 15:14:38 EDT

Yoiks ! It does appear that unsplitting doesn't tear down the 'cloned' editor...I'm on it !

Comment 10 Eric Moffatt 2014-05-26 15:21:20 EDT

OK, I now understand this one...I wasn't explicitly setting the 'cloned' editor's TBR to false while it was still connected to the model. By the time it becomes unrendered through the removal the code can't find its way back to the relevant workbench window so we don't remeve the reference...

I have a patch ready for this but I'm experiencing some Gerrit related issues right now (not related to Luna...;-).

If I don't get the patch in today it'll be there early tomorrow once Paul straightens me out.

Comment 11 Eric Moffatt 2014-05-27 09:15:12 EDT

Here's the patch that sets the TBR of the clone to false before removing it...

  https://git.eclipse.org/r/27369

Comment 12 Paul Webster 2014-05-27 16:16:21 EDT

(In reply to Eric Moffatt from comment #10)
> OK, I now understand this one...I wasn't explicitly setting the 'cloned'
> editor's TBR to false while it was still connected to the model. By the time
> it becomes unrendered through the removal the code can't find its way back
> to the relevant workbench window so we don't remeve the reference...

The patch looks straight forward.  I have a question.  This doesn't apply to a split editor that's  still split and then the entire editor is closed, right?  They were always being cleaned up?  It's just an editor that was split/unsplit that was leaking?

PW

Comment 13 Dani Megert 2014-05-28 03:33:25 EDT

(In reply to Paul Webster from comment #12)
> The patch looks straight forward.  I have a question.  This doesn't apply to
> a split editor that's  still split and then the entire editor is closed,
> right?  They were always being cleaned up?  It's just an editor that was
> split/unsplit that was leaking?

Almost. Currently, each split creates a new editor. Closing unsplit disposes one editor and closing in split state disposes two. Any further splitting creates another leak.

Comment 14 Dani Megert 2014-05-28 03:35:27 EDT

I spent some more time this morning to fully test & debug the NPE scenario (comment 0) and found out, that the fix does not clean up the editor references. It does call dipsose() with the patch though.

Test Case:
1. Go to org.eclipse.ui.internal.Workbench.persist(boolean)
2. Set a breakpoint after init of 'editorReferences'
3. Reduce Workbench.getAutoSaveJobTime() to something smaller, otherwise you 
   have to wait minutes
4. Start the target workbench, add a project and a new file
5. Split and unsplit the file
6. close the file
7. wait for the breakpoint to be hit
==> 'editorReferences' will contain one reference but shouldn't (normal open/close does not leave a reference)

NOTE: if you split/unsplit many times it will contain many entries of the same editor reference (see attached picture).

Comment 15 Dani Megert 2014-05-28 03:36:00 EDT

Created attachment 243604 [details]
Picture with the unexpected refrences

Comment 16 Eric Moffatt 2014-05-28 11:23:37 EDT

Dani, I've amended the original commit to prevent adding a reference if it already is in the list. I missed this the first time because I was focused on 'removes' rather than 'adds'...;-(.

My testing for this one is a but different from yours. I opened an editor and did a lot of splitting, unsplitting, changing split orientations...finally I split the editor and closed it.

Then I added a BP at 'openEditor', opened a new editor and inspected the 'editorReferences' which was empty (as expected).

Comment 17 Eric Moffatt 2014-05-28 13:31:40 EDT

Dani, I saw your comment about null checking 'clonedEditor'. This is unnecessary because the 'findElements' call *cannot* return nulls. If there are 3 values in the returned list they'll all be non-null (actually instances of MPart).

I think we should go with the current patch...

Comment 18 Paul Webster 2014-05-28 14:24:41 EDT

+1 component lead

PW

Comment 19 Dani Megert 2014-05-28 15:03:27 EDT

(In reply to Eric Moffatt from comment #17)
> Dani, I saw your comment about null checking 'clonedEditor'. This is
> unnecessary because the 'findElements' call *cannot* return nulls. If there
> are 3 values in the returned list they'll all be non-null (actually
> instances of MPart).
> 
> I think we should go with the current patch...

Makes sense.

We need a third reviewer.

Comment 20 Paul Webster 2014-05-28 15:26:48 EDT

Released as http://git.eclipse.org/c/platform/eclipse.platform.ui.git/commit/?id=45c33454013b9daab6b5c50df6b916346d8c0f66

PW

Comment 21 Dani Megert 2014-05-29 05:36:17 EDT

Verified in I20140528-2000.

Comment 22 David M. Karr 2015-04-09 12:09:55 EDT

I have a question about this bug/fix.

Would this fix have already been integrated into 4.4.2 (or STS 3.6.4)?  That's what I have, and today I saw the following stacktrace:

!ENTRY org.eclipse.ui.workbench 4 2 2015-04-09 08:58:02.997
!MESSAGE Problems occurred when invoking code from plug-in: "org.eclipse.ui.workbench".
!STACK 0
java.lang.NullPointerException
	at org.eclipse.ui.texteditor.AbstractTextEditor.saveState(AbstractTextEditor.java:7083)
	at org.eclipse.ui.internal.EditorReference.getEditorState(EditorReference.java:177)
	at org.eclipse.ui.internal.EditorReference.persist(EditorReference.java:99)
	at org.eclipse.ui.internal.Workbench$15.run(Workbench.java:1201)
	at org.eclipse.core.runtime.SafeRunner.run(SafeRunner.java:42)
	at org.eclipse.ui.internal.Workbench.persist(Workbench.java:1189)
	at org.eclipse.ui.internal.Workbench.access$47(Workbench.java:1187)
	at org.eclipse.ui.internal.Workbench$61.runInUIThread(Workbench.java:2957)
	at org.eclipse.ui.progress.UIJob$1.run(UIJob.java:97)
	at org.eclipse.swt.widgets.RunnableLock.run(RunnableLock.java:35)
	at org.eclipse.swt.widgets.Synchronizer.runAsyncMessages(Synchronizer.java:136)
	at org.eclipse.swt.widgets.Display.runAsyncMessages(Display.java:4147)
	at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3764)
	at org.eclipse.jface.window.Window.runEventLoop(Window.java:832)
	at org.eclipse.jface.window.Window.open(Window.java:808)
	at org.eclipse.ui.internal.dialogs.WorkbenchPreferenceDialog.open(WorkbenchPreferenceDialog.java:221)
        ...

Comment 23 Dani Megert 2015-04-10 04:14:39 EDT

(In reply to David M. Karr from comment #22)
> I have a question about this bug/fix.
> 
> Would this fix have already been integrated into 4.4.2 (or STS 3.6.4)? 

I don't know STS, but Eclipse has the fix since 4.4.

If you can reproduce the bug using 4.5 M6, then please file a new bug with steps to reproduce the NPE. Thanks.

Comment 24 David M. Karr 2015-10-14 16:59:01 EDT

I don't know how to repeat this, but I just started seeing this in my Mars.1 instance (4.5.1).  I haven't restarted it yet.

Comment 25 Dani Megert 2015-10-22 05:53:32 EDT

(In reply to David M. Karr from comment #24)
> I don't know how to repeat this, but I just started seeing this in my Mars.1
> instance (4.5.1).  I haven't restarted it yet.

If you see this with Mars.1 then please file a new bug with steps to reproduce the problem and cc me. Thanks.

Comment 26 David M. Karr 2018-03-09 11:20:02 EST

I'm now seeing this with Oxygen.2.  I use buffer splitting quite often.  I haven't yet restarted to see if it resolves this.