Summary: | Konqui was in background when it crashed (tab closing) [KonqMainWindow::slotAddClosedUrl] | ||
---|---|---|---|
Product: | [Applications] konqueror | Reporter: | m.wege |
Component: | general | Assignee: | Konqueror Developers <konq-bugs> |
Status: | RESOLVED FIXED | ||
Severity: | crash | CC: | andresbajotierra, benderamp, faure, fdruec1, frank78ac, gatoso, jeff.levasseur, knurz, mtadeunet, ronin3d, tim.eberhardt, xtekhne |
Priority: | NOR | ||
Version: | unspecified | ||
Target Milestone: | --- | ||
Platform: | Unlisted Binaries | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: |
Description
m.wege
2009-10-14 15:51:03 UTC
I see. Notes for myself: if (m_pWorkingTab) should fix the crash at least But this should really pass the tab as argument rather than use a member var. @David: Note that there were quite a few reports with that backtrace already (see bug 150162 and duplicates). *** Bug 210918 has been marked as a duplicate of this bug. *** *** Bug 213517 has been marked as a duplicate of this bug. *** *** Bug 214209 has been marked as a duplicate of this bug. *** *** Bug 219435 has been marked as a duplicate of this bug. *** *** Bug 221321 has been marked as a duplicate of this bug. *** *** Bug 222072 has been marked as a duplicate of this bug. *** *** Bug 223035 has been marked as a duplicate of this bug. *** *** Bug 224724 has been marked as a duplicate of this bug. *** Shouldn't this be merged with bug 150162 ? (or is there some difference?) Regards *** Bug 225357 has been marked as a duplicate of this bug. *** *** Bug 225362 has been marked as a duplicate of this bug. *** (In reply to comment #11) > Shouldn't this be merged with bug 150162 ? (or is there some difference?) It seems there really is a difference: In KonqViewManager::removeTab, the argument currentFrame is 0 here, which is dereferenced later on -> segfault which can be fixed like David suggested in comment 1. In bug 150162, currentFrame is not 0, but does not seem to be a valid pointer either. I have only noticed this now, I'm not sure if I've marked the duplicates correctly... *** Bug 229175 has been marked as a duplicate of this bug. *** Frank: No worries, it's the same bug. Actually I see how m_pWorkingTab can be dangling (if one does Ctrl+W more than once, faster than the singleshot timer), more than how it could be 0 (delete+add very fast? one has to be really really fast). But anyway the fix will fix both. SVN commit 1098373 by dfaure: Fix crash when closing tabs fast: don't keep pointers to tabs/frames across event loop returns, keep ints instead (tab numbers) Fixed for: 4.5 with this commit, to be backported if it proves stable enough. BUG: 210551 BUG: 150162 M +36 -14 konqmainwindow.cpp M +7 -4 konqmainwindow.h M +8 -8 konqtabs.cpp M +8 -1 konqtabs.h M +10 -12 konqviewmanager.cpp M +2 -2 konqviewmanager.h M +19 -1 tests/konqviewmgrtest.cpp M +1 -0 tests/konqviewmgrtest.h WebSVN link: http://websvn.kde.org/?view=rev&revision=1098373 *** Bug 230810 has been marked as a duplicate of this bug. *** *** Bug 231548 has been marked as a duplicate of this bug. *** *** Bug 231940 has been marked as a duplicate of this bug. *** *** Bug 231961 has been marked as a duplicate of this bug. *** *** Bug 235619 has been marked as a duplicate of this bug. *** SVN commit 1120142 by dfaure: Backport fix for bug 210551 (crash when closing tabs fast) to the 4.4 branch, so that the fix is in 4.4.3. CCBUG: 210551 M +36 -14 konqmainwindow.cpp M +7 -4 konqmainwindow.h M +9 -13 konqtabs.cpp M +7 -0 konqtabs.h M +9 -11 konqviewmanager.cpp M +2 -2 konqviewmanager.h M +19 -1 tests/konqviewmgrtest.cpp M +1 -0 tests/konqviewmgrtest.h WebSVN link: http://websvn.kde.org/?view=rev&revision=1120142 *** Bug 230047 has been marked as a duplicate of this bug. *** |