Description of problem: This and a few other crashes seems to happen when loading video to be streamed. Version-Release number of selected component: webkitgtk4-2.14.2-1.fc25 Additional info: reporter: libreport-2.8.0 backtrace_rating: 4 cmdline: /usr/libexec/webkit2gtk-4.0/WebKitWebProcess 33 crash_function: WTF::RefCountedBase::derefBase() const executable: /usr/libexec/webkit2gtk-4.0/WebKitWebProcess global_pid: 4034 kernel: 4.8.15-300.fc25.x86_64 pkg_fingerprint: 4089 D8F2 FDB1 9C98 pkg_vendor: Fedora Project runlevel: N 5 type: CCpp uid: 1000
Created attachment 1237632 [details] File: backtrace
Created attachment 1237633 [details] File: cgroup
Created attachment 1237634 [details] File: core_backtrace
Created attachment 1237636 [details] File: dso_list
Created attachment 1237637 [details] File: environ
Created attachment 1237638 [details] File: exploitable
Created attachment 1237639 [details] File: limits
Created attachment 1237640 [details] File: maps
Created attachment 1237641 [details] File: mountinfo
Created attachment 1237642 [details] File: namespaces
Created attachment 1237643 [details] File: open_fds
Created attachment 1237644 [details] File: proc_pid_status
Created attachment 1237645 [details] File: var_log_messages
(In reply to Victor Toso from comment #0) > Description of problem: > This and a few other crashes seems to happen when loading video to be > streamed. OK, the crash is at a fairly high level in the GTK API code, which is good, but due to another bug the missing plugin installer really shouldn't be running at all, so I'm confused how you were able to get this far. Is there a particular website that you can reproduce this on? Also, I need to mark the bug non-private to report it upstream, OK? (I don't see anything sensitive in the backtrace.)
> Is there a particular website that you can reproduce this on? Not really. It has happened twice in different websites but trying to reproduce them it does not crash. > Also, I need to mark the bug non-private to report it upstream, OK? (I don't > see anything sensitive in the backtrace.) That's fine :) If I get a reliable way to make it crash, I'll let you know.
OK, reported upstream. (In reply to Victor Toso from comment #15) > If I get a reliable way to make it crash, I'll let you know. That'd be great, thanks.
*** Bug 1419275 has been marked as a duplicate of this bug. ***