Summary: | KWin crashed and restarted on "Move [window] to Desktop" | ||
---|---|---|---|
Product: | [Plasma] kwin | Reporter: | Thomas Michalka <Thomas.Michalka> |
Component: | general | Assignee: | KWin default assignee <kwin-bugs-null> |
Status: | RESOLVED DUPLICATE | ||
Severity: | crash | ||
Priority: | NOR | ||
Version: | 4.11.9 | ||
Target Milestone: | --- | ||
Platform: | openSUSE | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: |
Description
Thomas Michalka
2014-09-10 17:03:17 UTC
*** This bug has been marked as a duplicate of bug 334152 *** How can I see recognize that "my bug" is a duplicate of an already filed one *before* I file the bug? Particularly the action on which the crash happened (see my summary) is completely different to bug 334152. I think "my crash" happened under other circumstances. So, why is it a duplicate? The condition is exactly the same - you're opening the context menu (and because the amount of activities changed at some point, you get a crash in return) Therefore it's also the the exact same backtrace. Bugzilla should have suggested similar bugs based upon the backtrace (before you finally submitted the bug), but there's some recent issue in bugzilla (see bug #337742, I have no further knowlege in that area, sorry) which lead to the 410 error and broke the DrKonqui bug submission. The bug is a dupe for sure - you may just trust me on this ;-) Hi Thomas, Am 10.09.2014 19:59, schrieb Thomas Lübking : > https://bugs.kde.org/show_bug.cgi?id=338984 > > --- Comment #3 from Thomas Lübking <thomas.luebking@gmail.com> --- > The condition is exactly the same - you're opening the context menu I agree so far. > (and because the amount of activities changed at some point, I never changed the "amount of activities" consciously. Are there conditions for changing this automatically without knowledge of the user? > you get a crash in return) Therefore it's also the the exact same > backtrace. Could there be other, may be unwanted reasons to reach the same code position and therefore getting exactly the same backtrace (under similar but not the same conditions)? > Bugzilla should have suggested similar bugs based upon the backtrace > (before you finally submitted the bug), It did, but it's really hard to recognize this. Visually there seemed to be some similarities but they looked not like identical traces. And: be aware that normal users don't have any clue how they should handle stuff like backtraces and duplicate bug messages. Thus I consider the benefit of comparison backtraces doubtful if reporting the error is the main issue respectively the real goal. > but there's some recent > issue in bugzilla (see bug #337742, I have no further knowlege in > that area, sorry) which lead to the 410 error and broke the DrKonqui > bug submission. Ok, let's hope for a soon resolution. > The bug is a dupe for sure - you may just trust me on this ;-) I willingly trust you as the expert of course :-) But the condition that lead to the bug seems to be partially another one. Happy fixing! Tom (In reply to Thomas Michalka from comment #4) > I never changed the "amount of activities" consciously. Are there > conditions for changing this automatically without knowledge of the user? Since it's more about "running" activities than "all" activities - "potentially" (read: "I frankly don't know, but could be - yes") > Could there be other, may be unwanted reasons to reach the same code > position and therefore getting exactly the same backtrace (under similar > but not the same conditions)? "Unlikely" - the menu action gets deleted, but not NULLed, what's (afaics) because it's deleted by moc what happens when its parent is deleted what happens when the activity submenu gets removed because there's only one activity left. That does however not matter, because the patch catches whatever deletes the menuaction with a guarded pointer. > Happy fixing! It is fixed since 4.11.10 |