[PATCH] iface-modem-3gpp: ignore initial registration check result when enabling

[Dan Williams]

On Fri, 2018-05-18 at 22:11 +0200, Aleksander Morgado wrote:
> On 16/05/18 21:45, Ben Chan wrote:
> > This patch extends commit c9e85b671 ("iface-modem-3gpp: ignore
> > initial
> > registration check result when appropriate") to handle the
> > situation
> > where a registration check update may prematurely transition the
> > modem
> > to the 'enabled' state while the modem is still being enabled.
> > 
> > The issue is identified by Dan Williams <dcbw at redhat.com>. More
> > details
> > can be found in https://bugs.freedesktop.org/show_bug.cgi?id=106468
> 
> Pushed to git master. But looks like this alone didn't solve the bug
> reporter's problems?

In the second set of logs the reporter wasn't getting as far as
enabling; they were stuck at the "double-flash" problem we talked about
last week.  I tried to continue tracking that down but ran out of time.

Dan

> > ---
> >  src/mm-iface-modem-3gpp.c | 14 +++++++-------
> >  1 file changed, 7 insertions(+), 7 deletions(-)
> > 
> > diff --git a/src/mm-iface-modem-3gpp.c b/src/mm-iface-modem-3gpp.c
> > index 4ac32fd5..4e19815d 100644
> > --- a/src/mm-iface-modem-3gpp.c
> > +++ b/src/mm-iface-modem-3gpp.c
> > @@ -1332,17 +1332,17 @@ update_registration_state (MMIfaceModem3gpp
> > *self,
> >          if (ctx->reloading_registration_info)
> >              return;
> >  
> > -        /* If the modem isn't already enabled or being enabled,
> > this
> > -         * registration state update is due to a previously
> > scheduled
> > -         * initial registration check when the modem was being
> > enabled.
> > -         * We need to ignore it as otherwise it may cause an
> > incorrect
> > -         * transition of the registration state and modem state
> > when the
> > -         * modem is being disabled. */
> > +        /* If the modem isn't already enabled, this registration
> > state update
> > +         * could be due to a previously scheduled initial
> > registration check
> > +         * when the modem was being enabled.  We need to ignore it
> > as otherwise
> > +         * it may cause an incorrect transition of the
> > registration state and
> > +         * modem state when the modem is being disabled or still
> > going through
> > +         * enable steps */
> >          modem_state = MM_MODEM_STATE_UNKNOWN;
> >          g_object_get (self,
> >                        MM_IFACE_MODEM_STATE, &modem_state,
> >                        NULL);
> > -        if (modem_state < MM_MODEM_STATE_ENABLING) {
> > +        if (modem_state < MM_MODEM_STATE_ENABLED) {
> >              mm_dbg ("Modem %s: 3GPP Registration state change
> > ignored as modem isn't enabled",
> >                      g_dbus_object_get_object_path (G_DBUS_OBJECT
> > (self)));
> >              return;
> > 
> 
>