Summary: | Kdevelop crashes sometimes on close | ||
---|---|---|---|
Product: | [Applications] kdevelop | Reporter: | André Stein <andre.stein.1985> |
Component: | general | Assignee: | kdevelop-bugs-null |
Status: | RESOLVED UPSTREAM | ||
Severity: | crash | CC: | cfeck, olivier.jg |
Priority: | NOR | ||
Version: | git master | ||
Target Milestone: | 4.3.0 | ||
Platform: | Compiled Sources | ||
OS: | Linux | ||
Latest Commit: | Version Fixed In: | ||
Sentry Crash Report: |
Description
André Stein
2012-11-05 11:52:18 UTC
My Qt Version is 4.8.3 KDE Version: 4.9.2 I'm on Ubuntu. So, KDevelop will crash everytime you close it with your current configuration? With one open Project? It's one open project with some files open. But it doesn't crash always. It's quite regular (I'd say in one of 5 cases). It must be something related to multithreading. I have six cores configured in a virtual machine and 2 parsing threads. Git commit f7fb38ff5c1c6fc091728f744b2e0047f0caf9bb by Milian Wolff. Committed on 09/11/2012 at 21:57. Pushed by mwolff into branch 'master'. Add VcsJob::JobTy::Unknown and use it by default. This is OK for e.g. the "git --version" job, as shown in the valgrind report of the following bug: M +1 -0 vcs/vcsjob.cpp M +1 -0 vcs/vcsjob.h http://commits.kde.org/kdevplatform/f7fb38ff5c1c6fc091728f744b2e0047f0caf9bb please try to reproduce this with a recent qt 4.8.4 that contains these fixes: https://bugreports.qt-project.org/browse/QTBUG-23871 https://bugreports.qt-project.org/browse/QTBUG-27322 if it's still reproducible, please reopen. Milian, I am not sure if the QDeclarative crash is connected to the QtScript bug (tracked at bug 305718), partly because one of them happens on initialization, the other on destruction. You mean the two bugs above? Well, the first one "uncovered" the last one, but yes - they are not really related per se. Defacto you need both fixes though, which is why I listed them. I am unsure, because we get a lot of similar crashes in Plasma (see bug 302931), but also with Qt 4.8.2, so it is unrelated to the QtScript regression in Qt 4.8.3. Putting it differently, do you think bug 302931 is also fixed in Qt 4.8.4? I'd say its fixed until someone reproduces it with Qt 4.8.4. Anything before that has the two known issues which can result in pretty much everything. So I'd say: wait for feedback of someone who uses Qt 4.8.4. Cheers |