Community
Participate
Working Groups
Created attachment 252767 [details] Error log I can use eclipse just fine for a while, for hours sometimes, but after a while it freezes up and starts to use more and more memory and then it just crashes. I reviewed all other bugs suggested by the tracker but none of them seemed to be the same. This one produces a similar dump but it's for another component entirely, it seems: https://bugs.eclipse.org/bugs/show_bug.cgi?id=25386 I set the compiler to target java 6 instead of 7 and the crash hasn't happened again but I doubt that's really the fix for this. I will update this report if it happens again.
I classified it as 'major' because in the report it says "A fatal error", I didn't see a 'fatal' classification. Feel free to re-classify it as you see fit.
I see you are using the max heap of 512m. Have you tried a higher value and see if that has any effect?
There's a number of SWT bugs showing the same # SIGSEGV [...] # Problematic frame: # C [libglib-2.0.so.0+0x38fe0] g_str_hash+0x0 I'm marking as duplicate of the one with most details. If you are running Ubuntu, please search the other bug for remarks regarding UBUNTU_MENUPROXY *** This bug has been marked as a duplicate of bug 461311 ***
moving to SWT
I can reproduce bug 461311 but that is not how my crashes happened at all. I never use the 'new window' menu or two eclipse windows for that matter. They occurred randomly, the last time I was toggling a breakpoint and it crashed. Underneath, eclipse may be invoking the same behavior although it doesn't seem likely, you can make that call but if it's not then my crash seems to be triggered by something else and this shouldn't be marked a duplicate.
Marking as duplicate is just a way to group related issues for more efficient investigation. If you can't provide steps for reproducing, it's quite impossible to investigate this issue on its own. OTOH, I suggest you try the workaround mentioned in the other bug and observe if the crash still happens. If it is resolved by the same workaround/solution this should suffice as indication for a common cause indeed.
Bug 419729 has fixed the general problem, no matter who created a radio menu item, i.e. all traces like this: j org.eclipse.swt.internal.gtk.OS._gtk_widget_show(J)V+0 j org.eclipse.swt.internal.gtk.OS.gtk_widget_show(J)V+8 j org.eclipse.swt.widgets.MenuItem.createHandle(I)V+248 j org.eclipse.swt.widgets.Widget.createWidget(I)V+2 j org.eclipse.swt.widgets.MenuItem.<init>(Lorg/eclipse/swt/widgets/Menu;II)V+38 *** This bug has been marked as a duplicate of bug 419729 ***
