[Users] [Bug 4385] Fedora reporting kernel crashes related to Claws Mail

[edwardp at gmx.com]

On Tue, 22 Sep 2020 06:58:16 +0600
Dustin Miller <dustbiz at gmail.com> wrote:

> On Mon, 21 Sep 2020 08:40:45 -0400
> <edwardp at gmx.com> wrote:
>
> > On Mon, 21 Sep 2020 12:12:15 -0000
> > Paul <claws at thewildbeast.co.uk> wrote:
> >
> > > On Mon, 21 Sep 2020 08:05:20 -0400
> > > <edwardp at gmx.com> wrote:
> > >
> > > > Thread 1 "claws-mail" received signal SIGSEGV, Segmentation
> > > > fault. 0x00007ffff72cb660 in g_type_check_instance_is_a ()
> > > >    from /lib64/libgobject-2.0.so.0
> > >
> > > That's not claws-mail --debug output.
> >
> > ... What part of the very end of the 'gdb' output would you like?
> > I'll be happy to copy/paste/upload it and the new 'bt full' output
> > to the bug report.
> >
> DM: Edward, I think you're misunderstanding what Paul is asking for
> here. I haven't taken the time to do so, but you may find it helpful
> to check one of his earliest posts in regards to the process he
> wanted you to go through. As part of that I think he wanted you to
> use the command 'claws-mail --debug' to start Claws rather than just
> 'claws-mail' which is the norm. When you start Claws from a terminal,
> this allows you to see all the terminal output and when you add the
> --debug option, this allows for additional things to be added to this
> output that are helpful for debugging. Normally, to get what Paul is
> asking for here, you would just go to the terminal and copy what you
> need from there. I'm not exactly sure how it works if you are pairing
> it with running gdb. However, the main point to be aware of is that
> Paul is looking for information from two different sources -- one is
> the relevant output from gdb, and the other is the relevant output
> from running Claws with the --debug option. These are two separate
> things. That's my understanding; Paul can correct me if I've
> misunderstood. Hope that helps, ---Dustin

The info above showing the library in question, is at the very end of
the output from the gdb process and before the backtrace command 'bt
full' was run. That is the crash, that is what I initially thought Paul
was requesting.

Yesterday, I uploaded to my bug report, the last ~20 seconds of that
same output - but from a new crash - and the same information is at the
very end, except the memory address is slightly different.