Community
Participate
Working Groups
Since Eclipse 2019-09 M2 I regularly (but not always) get a freeze in the staging view. I stage some file. Press "Commit and Push" and then the complete staging view get's inactive (grey) so that once can no longer press buttons. I can only recover by restarting the IDE. After the restart I see that the commit did not happen. The staging view has the same state as it was directly bevor I have pressed "Commit and Push".
is your home directory on a network volume ?
(In reply to Matthias Sohn from comment #1) > is your home directory on a network volume ? No it's not.
Let's wait until M3 comes out. I also set up a 2019-09 Eclipse and got strange things recently with EGit nightly. (Though nothing like comment 0.) Since I don't have any such problems with EGit nightly in 2019-06 or earlier instances I wrote it off to 2019-09 M2 being a development version. I'll check again once M3 is here.
I do all my development in about 10 different installations of M2 and never have seen that (but I use egit nightly). When you write freeze, you do not mean the application is not responding anymore, but just that all controls in the view get disabled, right? That is normal behaviour with every commmit. I assume that some jgit operation then doesn't come back, and egit never re-enables the buttons. Did you ever wait as long as Preferences > Team > Git > Conneciton timeout before restarting? Do you see anything in the error log view?
Create a couple of thread dumps when it hangs, this should reveal what's going on.
Any errors in the log? See also bug 550429.
(In reply to Thomas Wolf from comment #6) > Any errors in the log? See also bug 550429. The symptom sounds exactly what happens. I will look into the error log when it happens next time.
Created attachment 279689 [details] Stacktrace of disabled staging view This is the stacktrace. The staging view stays inactive. The application is still running (so the main thread is not blocked).
(In reply to Matthias Becker from comment #8) > Created attachment 279689 [details] > Stacktrace of disabled staging view Well, the stack trace just show a completely idle Eclipse. What about the error log?
(In reply to Thomas Wolf from comment #9) > (In reply to Matthias Becker from comment #8) > > Created attachment 279689 [details] > > Stacktrace of disabled staging view > > Well, the stack trace just show a completely idle Eclipse. What about the > error log? It says this: java.lang.NullPointerException at org.eclipse.ui.texteditor.AbstractTextEditor$TextEditorSavable.isDirty(AbstractTextEditor.java:7186) at org.eclipse.ui.internal.Workbench.getFilteredSaveables(Workbench.java:3495) at org.eclipse.ui.internal.Workbench.saveAll(Workbench.java:3468) at org.eclipse.egit.ui.UIUtils.saveAllEditors(UIUtils.java:886) at org.eclipse.egit.ui.internal.staging.StagingView.commit(StagingView.java:4236) at org.eclipse.egit.ui.internal.staging.StagingView.access$28(StagingView.java:4218) at org.eclipse.egit.ui.internal.staging.StagingView$22.widgetSelected(StagingView.java:1167) at org.eclipse.swt.widgets.TypedListener.handleEvent(TypedListener.java:252) at org.eclipse.swt.widgets.EventTable.sendEvent(EventTable.java:89) at org.eclipse.swt.widgets.Display.sendEvent(Display.java:4383) at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1512) at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1535) at org.eclipse.swt.widgets.Widget.sendEvent(Widget.java:1520) at org.eclipse.swt.widgets.Widget.notifyListeners(Widget.java:1324) at org.eclipse.swt.widgets.Display.runDeferredEvents(Display.java:4170) at org.eclipse.swt.widgets.Display.readAndDispatch(Display.java:3787) at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine$5.run(PartRenderingEngine.java:1160) at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:338) at org.eclipse.e4.ui.internal.workbench.swt.PartRenderingEngine.run(PartRenderingEngine.java:1049) at org.eclipse.e4.ui.internal.workbench.E4Workbench.createAndRunUI(E4Workbench.java:155) at org.eclipse.ui.internal.Workbench.lambda$3(Workbench.java:635) at org.eclipse.core.databinding.observable.Realm.runWithDefault(Realm.java:338) at org.eclipse.ui.internal.Workbench.createAndRunWorkbench(Workbench.java:559) at org.eclipse.ui.PlatformUI.createAndRunWorkbench(PlatformUI.java:150) at org.eclipse.ui.internal.ide.application.IDEApplication.start(IDEApplication.java:150) at org.eclipse.equinox.internal.app.EclipseAppHandle.run(EclipseAppHandle.java:203) at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.runApplication(EclipseAppLauncher.java:137) at org.eclipse.core.runtime.internal.adaptor.EclipseAppLauncher.start(EclipseAppLauncher.java:107) at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:400) at org.eclipse.core.runtime.adaptor.EclipseStarter.run(EclipseStarter.java:255) at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at java.base/jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) at java.base/jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) at java.base/java.lang.reflect.Method.invoke(Method.java:566) at org.eclipse.equinox.launcher.Main.invokeFramework(Main.java:660) at org.eclipse.equinox.launcher.Main.basicRun(Main.java:597) at org.eclipse.equinox.launcher.Main.run(Main.java:1468)
*** Bug 550429 has been marked as a duplicate of this bug. ***
(In reply to Matthias Becker from comment #10) > (In reply to Thomas Wolf from comment #9) > > (In reply to Matthias Becker from comment #8) > > > Created attachment 279689 [details] > > > Stacktrace of disabled staging view > > > > Well, the stack trace just show a completely idle Eclipse. What about the > > error log? > > It says this: > > java.lang.NullPointerException > at > org.eclipse.ui.texteditor.AbstractTextEditor$TextEditorSavable. > isDirty(AbstractTextEditor.java:7186) > at > org.eclipse.ui.internal.Workbench.getFilteredSaveables(Workbench.java:3495) > at org.eclipse.ui.internal.Workbench.saveAll(Workbench.java:3468) > at org.eclipse.egit.ui.UIUtils.saveAllEditors(UIUtils.java:886) > at > org.eclipse.egit.ui.internal.staging.StagingView.commit(StagingView.java: > 4236) Was this error reported in the UI (via a dialog)? What Java version? In bug 550429 it was Java 12.0.1, and also a 4.13 Eclipse. Move this to platform? If AbstractTextEditor$TextEditorSavable.isDirty() throws NPEs something is utterly broken in the text editor framework. Follow-up bug 550513 for EGit created, but note that that will only mask this problem.
See also bug 237680. :-(
And also bug 173250 and various bugs linked from there. Looks like something in 4.13 has unfixed that.
Moving to Platform/UI.
This is a severy issue. I would say a showstopper for 4.13 and 2019-09
(In reply to Matthias Becker from comment #16) > This is a severy issue. I would say a showstopper for 4.13 and 2019-09 Can you provide exact steps to reproduce the NPE?
(In reply to Dani Megert from comment #17) > (In reply to Matthias Becker from comment #16) > > This is a severy issue. I would say a showstopper for 4.13 and 2019-09 > Can you provide exact steps to reproduce the NPE? "Exact"? No Just do Java Coding and every now and then you will get this.
(In reply to Matthias Becker from comment #18) > (In reply to Dani Megert from comment #17) > > (In reply to Matthias Becker from comment #16) > > > This is a severy issue. I would say a showstopper for 4.13 and 2019-09 > > Can you provide exact steps to reproduce the NPE? > > "Exact"? No > > Just do Java Coding and every now and then you will get this. Presumably you also stage and commit some changes from time to time. Or can it also be reproduced by having dirty editors (maybe many tabs, so that there are hidden dirty editors) and then choosing "Close all" from the context menu? And what Java version do you get this with? (In the duplicate bug 550429 the user reported Java 12.0.1.) I don't see any 2019-09 M3 EPP packages yet, but... does this also occur with M3?
(In reply to Thomas Wolf from comment #19) > Presumably you also stage and commit some changes from time to time. Yes. >Or can > it also be reproduced by having dirty editors (maybe many tabs, so that > there are hidden dirty editors) and then choosing "Close all" from the > context menu? Never tried. As I don't get an Error dialog but only this error log entry I may not notice this NPE. > > And what Java version do you get this with? (In the duplicate bug 550429 the > user reported Java 12.0.1.) It's Java 11.0.2 > I don't see any 2019-09 M3 EPP packages yet, but... does this also occur > with M3? Don't know.
(In reply to Matthias Becker from comment #18) > (In reply to Dani Megert from comment #17) > > (In reply to Matthias Becker from comment #16) > > > This is a severy issue. I would say a showstopper for 4.13 and 2019-09 > > Can you provide exact steps to reproduce the NPE? > > "Exact"? No > > Just do Java Coding and every now and then you will get this. Believe me I do a lot of that and have never seen this (including using EGit).
I have seen this problem but cannot provide exact steps to reproduce nor say what exact 4.13 build had/started this problem. I also tried to intentional trigger this error but failed.
(In reply to Paul Pazderski from comment #22) > I have seen this problem but cannot provide exact steps to reproduce nor say > what exact 4.13 build had/started this problem. Did you see this only in combination with Staging view?
(In reply to Dani Megert from comment #23) > (In reply to Paul Pazderski from comment #22) > > I have seen this problem but cannot provide exact steps to reproduce nor say > > what exact 4.13 build had/started this problem. > Did you see this only in combination with Staging view? Yes, same situation as described in the initial description with the minor difference that I used "Commit" instead of "Commit and Push"
(In reply to Paul Pazderski from comment #24) > (In reply to Dani Megert from comment #23) > > (In reply to Paul Pazderski from comment #22) > > > I have seen this problem but cannot provide exact steps to reproduce nor say > > > what exact 4.13 build had/started this problem. > > Did you see this only in combination with Staging view? > > Yes, same situation as described in the initial description with the minor > difference that I used "Commit" instead of "Commit and Push" OK, so, could also be caused by new version of EGit instead of new version of Eclipse?
(In reply to Dani Megert from comment #25) > (In reply to Paul Pazderski from comment #24) > > (In reply to Dani Megert from comment #23) > > > (In reply to Paul Pazderski from comment #22) > > > > I have seen this problem but cannot provide exact steps to reproduce nor say > > > > what exact 4.13 build had/started this problem. > > > Did you see this only in combination with Staging view? > > > > Yes, same situation as described in the initial description with the minor > > difference that I used "Commit" instead of "Commit and Push" > OK, so, could also be caused by new version of EGit instead of new version > of Eclipse? Not sure about that. My used Egit version is 5.4.0.201906121030-r.
(In reply to Dani Megert from comment #25) > (In reply to Paul Pazderski from comment #24) > > (In reply to Dani Megert from comment #23) > > > (In reply to Paul Pazderski from comment #22) > > > > I have seen this problem but cannot provide exact steps to reproduce nor say > > > > what exact 4.13 build had/started this problem. > > > Did you see this only in combination with Staging view? > > > > Yes, same situation as described in the initial description with the minor > > difference that I used "Commit" instead of "Commit and Push" > OK, so, could also be caused by new version of EGit instead of new version > of Eclipse? Highly unlikely; there've been no changes to the relevant code in EGit as far as I see. It also doesn't occur in a git-related operation; it occurs when EGit tries to save all open editors before a commit. (If any are saved, it'll then ask the user whether he wants to include those changes in the commit.) In any case, EGit calls IWorkbench.saveAll() and then one of the editors throws that NPE. Hardly the staging view's fault, isn't it? Also, in bug 237680 somebody reported this in 2015 with a stack trace that shows no trace of EGit or of any other team provider. Were there any changes to the handling of the Saveables in platform lately?
Kalyan, please have a look.
(In reply to Dani Megert from comment #28) > Kalyan, please have a look. I have not been able to reproduce this issue. I am using Java 11.0.2 with Egit 5.4.0.201906121030-r on Windows 10. From the stack trace it appears that Matthias is facing this issue on Mac. @Paul Are you also facing this issue on Mac?
FYI: I have reported the same problem in Bug 550429, running eclipse in Java 12 under Linux, it is thus unlikely that it is a Mac-only problem.
(In reply to Kalyan Prasad Tatavarthi from comment #29) > @Paul Are you also facing this issue on Mac? No. Windows 7 or 10, not sure on which one it actually happened maybe both.
The only way AbstractTextEditor$TextEditorSavable.isDirty() call can cause an NPE is if AbstractTextEditor$TextEditorSavable.disconnectEditor() call has occurred before this call and therefore the TextEditor associated with the saveable is null. As a last resort we can add a null pointer check in the isDirty() call (Because the associated TextEditor is disconnected, its safe to return isDirty as false by this saveable.) Also, 4.13 M3 will be available tomorrow. you can try using this and provide your feedback. I will create a NPE check gerrit patch and deliver the changes as part of 4.13 RC2 on 3rd September based on your feedback of using 4.13 M3 if that is ok with everyone.
I can reproduce it without egit. I "think" my IDE is 1909 M2 of SimRel. Where can I finde a "M2"-Version of Platform? I get the NPE also when closing a project while a editor is open for a file inside this project. The good news is with M3 from: https://www.eclipse.org/downloads/download.php?file=/eclipse/downloads/drops4/S-4.13M3-201908211800/eclipse-SDK-4.13M3-macosx-cocoa-x86_64.dmg&mirror_id=1039 this does not happen. With M1 from: https://www.eclipse.org/downloads/download.php?file=/eclipse/downloads/drops4/S-4.13M1-201907111805/eclipse-SDK-4.13M1-macosx-cocoa-x86_64.dmg&mirror_id=1039 it also does not happen. So it seems to be fixed in M3.
(In reply to Matthias Becker from comment #33) > Where can I finde a "M2"-Version of Platform? For M2 we contribute only an I-Build. See Bug 546902. I would not spend time testing in M2, testing with M3 is more important.
(In reply to Lars Vogel from comment #34) > For M2 we contribute only an I-Build. See Bug 546902. Btw.: Why is this? Why do we only contribute an I-Build?
(In reply to Matthias Becker from comment #35) > Btw.: Why is this? Why do we only contribute an I-Build? IIRC we decided that two tested M-builds are sufficient for a 3 months release cycle.
(In reply to Matthias Becker from comment #33) > So it seems to be fixed in M3. No. Just happened again in I20190901-1800 and EGit 5.5.0.201908280940-m3.
(In reply to Paul Pazderski from comment #37) > (In reply to Matthias Becker from comment #33) > > So it seems to be fixed in M3. > > No. Just happened again in I20190901-1800 and EGit 5.5.0.201908280940-m3. In the next EGit 5.5.0 version (I guess that'll be -rc1; otherwise in our final release version -r) you'll get an error dialog for this and the staging view will re-enable. (In reply to Kalyan Prasad Tatavarthi from comment #32) > > As a last resort we can add a null pointer check in the isDirty() call That would just completely hide the problem. Perhaps the right thing to do for the 4.13 release if the root cause cannot be determined, but maybe one could remove such a null check early in 4.14 again with the intent of tracking this down for good in 4.14.
New Gerrit change created: https://git.eclipse.org/r/148745
(In reply to Eclipse Genie from comment #39) > New Gerrit change created: https://git.eclipse.org/r/148745 Abandoning this gerrit patch
(In reply to Kalyan Prasad Tatavarthi from comment #40) > (In reply to Eclipse Genie from comment #39) > > New Gerrit change created: https://git.eclipse.org/r/148745 > > Abandoning this gerrit patch New Gerrit change created: https://git.eclipse.org/r/#/c/148746/ This change catches the NPE in AbstractTextEditor$TextEditorSavable.isDirty and logs it.
Gerrit change https://git.eclipse.org/r/148746 was merged to [master]. Commit: http://git.eclipse.org/c/platform/eclipse.platform.text.git/commit/?id=20f86e123c7e6fa0950be7a8f87a1a61f2650395
(In reply to Eclipse Genie from comment #42) > Gerrit change https://git.eclipse.org/r/148746 was merged to [master]. > Commit: > http://git.eclipse.org/c/platform/eclipse.platform.text.git/commit/ > ?id=20f86e123c7e6fa0950be7a8f87a1a61f2650395 Moving this bug to 4.14 to check for the exact reason for the NPE.
(In reply to Kalyan Prasad Tatavarthi from comment #43) > (In reply to Eclipse Genie from comment #42) > > Gerrit change https://git.eclipse.org/r/148746 was merged to [master]. > > Commit: > > http://git.eclipse.org/c/platform/eclipse.platform.text.git/commit/ > > ?id=20f86e123c7e6fa0950be7a8f87a1a61f2650395 > > Moving this bug to 4.14 to check for the exact reason for the NPE. May be we should close this bug to track the 4.13 RC2 change and clone this bug for 4.14.
(In reply to Lakshmi Shanmugam from comment #44) > (In reply to Kalyan Prasad Tatavarthi from comment #43) > > (In reply to Eclipse Genie from comment #42) > > > Gerrit change https://git.eclipse.org/r/148746 was merged to [master]. > > > Commit: > > > http://git.eclipse.org/c/platform/eclipse.platform.text.git/commit/ > > > ?id=20f86e123c7e6fa0950be7a8f87a1a61f2650395 > > > > Moving this bug to 4.14 to check for the exact reason for the NPE. > > May be we should close this bug to track the 4.13 RC2 change and clone this > bug for 4.14. Cloned a bug 550691 for 4.14 M1 and resolving this bug as fixed in 4.13 RC2 to track the changes.
@Paul Can you please verify this fix in tonight's build, tomorrow for 4.13 RC2?
(In reply to Kalyan Prasad Tatavarthi from comment #46) > @Paul Can you please verify this fix in tonight's build, tomorrow for 4.13 > RC2? No. I have still no clue what triggers this bug and the last time I tried to reproduce it everything worked fine. But I will anyway use this build soon and can tell if the new error message pop up in log.
Still happens on latest I build. See bug 569286.
Maybe the reason for this is that > org.eclipse.ui.texteditor.AbstractTextEditor.dispose() does not inform the listener about the close/dispose in contrast to > org.eclipse.ui.texteditor.AbstractTextEditor.doSetInput(IEditorInput) May this happens magically somewhere in between but if not the savable remains in the SaveablesList and Workbench.saveAll will try to call the (invalid) savable... can someone verify this observation?
Moving this bug to Resolved as there is
Resolved this bug by mistake. So, Reopening
*** This bug has been marked as a duplicate of bug 569286 ***
