Re: [Evolution] Crashing Evolution 3.46.0 ?

[Milan Crha <mcrha redhat com>]

On Fri, 2022-10-07 at 15:41 -0400, Paul Smith wrote:
> Is the 0xaaaaaaaaaaaaaaaa some kind of overwritten memory due to
> being freed, or something?

        Hi,
yes, it can be, but I'm not sure (I've no idea whether there are any
functions enabled in Flatpak to overwrite pointer addresses after
free).

> (gdb) bt full
> #0  0x00007f8002abc8fb in g_mutex_lock
> (mutex=mutex@entry=0xaaaaaaaaaaaaaab2) at ../glib/gthread-
> posix.c:1529
>         gaicae_oldval = 0
> #1  0x00007f7ffc8977d1 in soup_connection_manager_cleanup
> (manager=0xaaaaaaaaaaaaaaaa, cleanup_idle=cleanup_idle@entry=0) at
> ../libsoup/soup-connection-manager.c:516
>         conns = <optimized out>
> #2  0x00007f7ffc8a8b45 in async_run_queue
> (session=session@entry=0x55e754ed96c0) at ../libsoup/soup-
> session.c:1821
>         priv = 0x55e754ed9620
>         items = 0x0
>         i = <optimized out>
> #3  0x00007f7ffc8a8c35 in queue_dispatch (source=<optimized out>,
> callback=<optimized out>, user_data=<optimized out>) at
> ../libsoup/soup-session.c:184
>         session = 0x55e754ed96c0

The backtrace shows a problem in libsoup3, possibly calling
(asynchronously) soup_connection_manager_cleanup() on an object, which
had been already freed. It looks like that from the backtrace at least,
I do not know libsoup internals that well.

I would file this as an issue into their bug tracker:
https://gitlab.gnome.org/GNOME/libsoup/-/issues

Evolution itself doesn't do many libsoup calls on its own, only a few.
Do you have enabled sender's photos, or RSS, or remote content loading,
or... I do not know off head what all can call libsoup in the evolution
process while viewing messages, I'm sorry.

        Bye,
        Milan