Re: Modem Manager stuck in a loop, ppp can't connect.

[Dan Williams <dcbw redhat com>]

On Wed, 2017-05-10 at 14:21 -0400, A. F. Cano wrote:
> On Mon, May 08, 2017 at 09:27:23AM -0500, Dan Williams wrote:
> > ...
> > Does the modem reply to AT+CSS with valid SID/NID values now, or
> > still
> > "?, 0"?
>
> Interesting...
>
> +CSS: ?, 22
>
> OK

The device is using an early form of AT+CSS, which reports only the
CDMA System ID, which is 22.  This means the modem has registered with
the network correctly and we can most likely do packet data.  If it
reports '0' then it's assumed that it hasn't successfully gotten
registration far enough for data.

> When connecting to the modem (with kermit) echo was off so my command
> isn't visible, but it did return the above.  What does the 22 mean or
> imply?
>
> I have also noticed these warnings:
>
> May  6 19:12:40 fbx ModemManager[277]: <warn>  Could not get service
> status: No AT port available to run command
> May  6 19:13:10 fbx ModemManager[277]: <warn>  Could not get service
> status: No AT port available to run command
> May  6 19:13:40 fbx ModemManager[277]: <warn>  Could not get service
> status: No AT port available to run command
> May  6 19:14:10 fbx ModemManager[277]: <warn>  Could not get service
> status: No AT port available to run command

The modem only has one serial port, and that's taken by PPP when the
data connection is active.  This means there's no port available to run
the status commands.  But it's a bug this is printed at a <warn> log
level rather than <debug> or something like that.  Anyway, the message
is expected and harmless.

> And they keep repeating every 30 seconds.
>
> I have encountered this the last time I connected.  The ppp
> connection
> was working fine but this message kept repeating the whole time the
> connection was up:
>
> May  9 21:02:06 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:07 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer

This would be a bug in NetworkManager.  Any chance that, when this
happens, you could 'gdb attach <pid of NetworkManager>', then 'break
g_log', then 'continue'.  Then next time it breaks out to the '(gdb)'
prompt, run "t a a bt" to get a backtrace.  That will likely tell us
where the issue is.

Dan

> May  9 21:02:07 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:08 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:08 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:09 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:09 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:11 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:11 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:12 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:12 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:13 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:13 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:14 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:14 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:15 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:15 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:16 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:16 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:17 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:17 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:18 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:18 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:19 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:19 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:20 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:20 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G_TYPE_CHECK_INSTANCE (instance)' failed
> May  9 21:02:23 fbx NetworkManager[379]: instance with invalid (NULL)
> class pointer
> May  9 21:02:23 fbx NetworkManager[379]: g_signal_emit_valist:
> assertion 'G
>
> At about 2 per second, the log file quickly becomes pretty huge.
>
> Thanks again for replying.
>
>
> Augustine
>
> _______________________________________________
> networkmanager-list mailing list
> networkmanager-list gnome org
> https://mail.gnome.org/mailman/listinfo/networkmanager-list