Can you reproduce this? If so, it would be good to call folks_debug_emit_print_status(folks_debug_dup()) (from gdb) just before causing the crash. This crash should only happen if a Persona is removed from an Individual it was never in — which should never happen.

Yeah, it happens pretty often when I suspend my laptop while being online. I'll try to reproduce and provide the info.

I just sent you the output of this command once I hit the crash. Not sure how to catch it just before it.

*** Bug 671148 has been marked as a duplicate of this bug. ***

duplicated bug has also the stacktrace i wish that would help 
https://launchpadlibrarian.net/94067284/Stacktrace.txt

Guillaume, would it be possible for you to reproduce the crash again and get a *correlated* set of debug info? I need:
 • the logs from FOLKS_DEBUG=all G_MESSAGES_DEBUG=all;
 • the folks_debug_emit_print_status() stuff in gdb; and
 • the backtrace from gdb.

(Basically, I need the memory addresses to match between them so I can figure out what's going on.)

From the two stacktraces so far, I've got a theory that the IndividualAggregator's link map is pointing to Individual A, which contains Persona 1 in its set of personas. However, somehow Individual B has claimed ownership of Persona 1, and also has Persona 1 in its set of personas. Persona 1's individual pointer is pointing at Individual B — so when Individual A iterates over its set of personas, it triggers the assertion when it reaches Persona 1.

Created attachment 208967 [details] [review]
core: Don’t modify a HashSet while iterating over it

Alternatively, here’s a patch which might fix the problem. (It certainly fixes _a_ bug.) We were iterating over a set, removing elements of the set by calling iter.remove() as we go. This should’ve been fine, but bug #671327 meant it wasn’t, and we were potentially skipping some elements of the set. This patch works around bug #671327.

Guillaume, could you please give this a go and see if it fixes your crash? If it doesn’t, I’ll need the debug information listed in comment #6.

https://www.gitorious.org/folks/folks/commits/670523-individual-persona-assertion

(In reply to comment #7)
> Guillaume, could you please give this a go and see if it fixes your crash? If
> it doesn’t, I’ll need the debug information listed in comment #6.

Sure, I'm now running it. It uses to happen quiet a lot so if it didn't crash during the next few days I think we'll be good.

Reproduced. I sent you all the info by mail.

*** Bug 671641 has been marked as a duplicate of this bug. ***

with attachment #208967 [details], notify::is-quiescent is never emitted it seems. Patch applied on top of current folks master.

also present in downstream
https://bugs.launchpad.net/ubuntu/+source/empathy/+bug/944418

*** Bug 672352 has been marked as a duplicate of this bug. ***

Review of attachment 208967 [details] [review]:

++

Comment on attachment 208967 [details] [review]
core: Don’t modify a HashSet while iterating over it

commit 46d4506d346199f53039d8d31275b8ce61146528
Author: Philip Withnall <philip@tecnocode.co.uk>
Date:   Mon Mar 5 00:11:09 2012 +0000

    core: Don’t modify a HashSet while iterating over it
    
    Even though we were modifying it from the Iterator (which should’ve been OK),
    we got hit by https://bugzilla.gnome.org/show_bug.cgi?id=671327 which meant
    that some Personas weren't getting disconnected from Individuals. This is a
    likely cause of bug #670523.
    
    This also includes a related (but irrelevant to the bug itself) tidy-up to
    Persona._individual.
    
    Closes: https://bugzilla.gnome.org/show_bug.cgi?id=670523

 NEWS                  |    2 ++
 folks/individual.vala |   11 ++++++-----
 folks/persona.vala    |    4 ++--
 3 files changed, 10 insertions(+), 7 deletions(-)

Guillaume can no longer reproduce the bug with this patch and the one from bug #672381, so let's close this.

*** Bug 673358 has been marked as a duplicate of this bug. ***

*** Bug 675120 has been marked as a duplicate of this bug. ***