Re: AW: Problems with Sierra MC 8790 with older firmware revision

[Aleksander Morgado <aleksander lanedo com>]

On 04/04/2013 11:08 AM, Aleksander Morgado wrote:
> On 04/03/2013 08:18 PM, Dan Williams wrote:
> > On Wed, 2013-04-03 at 18:59 +0200, Aleksander Morgado wrote:
> > > On 03/04/2013 03:33 PM, Dan Williams wrote:
> > > > On Fri, 2013-03-01 at 15:08 +0100, Gerald Richter - ECOS wrote:
> > > > > The following patch solves the problem for us. It works with all revisions we have available for testing
> > > >
> > > > Pushed to git master and 0.6, thanks!
> > > >
> > > > Dan
> > > >
> > > > > Gerald
> > > > >
> > > > > --- mm-plugin-sierra.c~   2012-08-29 17:02:18.000000000 +0200
> > > > > +++ mm-plugin-sierra.c    2013-03-01 13:22:09.000000000 +0100
> > > > > @@ -78,6 +78,9 @@
> > > > >          if (strstr (response, "C885"))
> > > > >              g_object_set_data (G_OBJECT (task), TAG_SIERRA_APP_PPP_OK, GUINT_TO_POINTER (TRUE));
> > > > >  
> > > > > +        if (strstr (response, "MC8790"))
> > > > > +            g_object_set_data (G_OBJECT (task), TAG_SIERRA_APP_PPP_OK, GUINT_TO_POINTER (TRUE));
> > > > > +
> > > > >          /* For debugging: let users figure out if their device supports it or not */
> > > > >          if (getenv ("MM_SIERRA_APP1_PPP_OK")) {
> > > > >              mm_dbg ("Sierra: APP1 PPP OK '%s'", response);
> > >
> > >
> > > Ah, the fun... :)
> > >
> > > So I've got a MC8790 here which does *not* like the APP1 port for PPP.
> > > It has the following revision:
> > > 'K2_0_7_35AP C:/WS/FW/K2_0_7_35AP/MSM6290/SRC 2010/03/04 17:37:08'
> > >
> > > This is what we get from the probing:
> > >
> > > log_port(): (/sys/devices/pci0000:00/0000:00:1d.0/usb2/2-1/2-1.7)
> > > tty/ttyUSB3 at (primary)
> > > log_port(): (/sys/devices/pci0000:00/0000:00:1d.0/usb2/2-1/2-1.7)
> > > tty/ttyUSB4 data (primary)
> > > log_port(): (/sys/devices/pci0000:00/0000:00:1d.0/usb2/2-1/2-1.7)
> > > tty/ttyUSB1 qcdm
> > >
> > > And it just stalls when we try to launch the ATD call in ttyUSB4:
> > >
> > > [1365007565.579948] [mm-broadband-bearer.c:200]
> > > common_get_at_data_port(): Connection through a plain serial AT port
> > > (ttyUSB4)
> > > [1365007565.579997] [mm-serial-port.c:958] mm_serial_port_open():
> > > (ttyUSB4) device open count is 2 (open)
> > > [1365007565.580036] [mm-serial-port.c:1003] mm_serial_port_close():
> > > (ttyUSB3) device open count is 1 (close)
> > > [1365007565.580080] [mm-at-serial-port.c:408] debug_log(): (ttyUSB4):
> > > --> 'ATD*99***2#<CR>'
> > > ...
> > > And that's it, won't reply.
> > >
> > > Running the ATD call in the primary ttyUSB3 port (without the patch
> > > above) kind of works; in that case ttyUSB4 is marked as secondary, but
> > > again ttyUSB4 doesn't know how to properly work not even as secondary
> > > and either reports error or times out most messages...
> > >
> > > Some additional info for Dan, surely not very useful :)
> >
> > Can you send:
> >
> > AT!ENTERCND="A710"
> >
> > and then enter all these again?  I get a response to NVMUXMODE when I do
> > this at least, though its 0 (not set).  After you've done this, what do
> > you get for "AT!CUSTOM?" ?
> >
> > If you're feeling really, really lucky, you could try AT!
> > CUSTOM="MUXMODE",2 to disable MUX mode; the default is "don't override"
> > whatever is currently set, and of course we don't know what that is.
> > This might brick the card though, so beware.
> >
> > FWIW, my USB306 says:
> >
> > !CUSTOM: 
> >             PUKPRMPT            0x01
> >             MEPCODE             0x01
> >             ISVOICEN            0x01
> >             PRLREGION           0x01
> >             PCSCDISABLE         0x03
> >             GPSENABLE           0x01
> >             SWOCENABLE          0x03
> >             USBMSENABLE         0x01
> >
> > so my MUXMODE pref hasn't been modified at all.
>
> There you go:
>
> AT!ENTERCND="A710"
>   OK
> AT!NVPORTSET?
>   ERROR
> AT!MXPORTMAP?
>   !MXPORTMAP:
>   00
>   OK
> AT!NVMUXMODE?
>   !NVMUXMODE:
>   00
>   OK
> AT!NVMUXMODE=?
>   !NVMUXMODE:
>   (0-2)
>   0 - MUX mode not set
>   1 - MUX mode on
>   2 - MUX mode off
>   OK
> AT!MAPUART=?
>   ERROR
> AT!MAPUART?
>   ERROR
> AT!NVPORTMAP?
>   ERROR
> AT!CUSTOM?
>   !CUSTOM:
>             PUKPRMPT            0x01
>             MEPCODE             0x01
>             ISVOICEN            0x02
>             PRLREGION           0x01
>             PCSCDISABLE         0x03
>             GPSENABLE           0x01
>   OK
> AT!CUSTOM="MUXMODE",2
>   OK
> AT!CUSTOM?
>   !CUSTOM:
>             PUKPRMPT            0x01
>             MEPCODE             0x01
>             ISVOICEN            0x02
>             PRLREGION           0x01
>             PCSCDISABLE         0x03
>             GPSENABLE           0x01
>             MUXMODE             0x02
>   OK
>
> Disabling the muxmode didn't really help it seems.

Some additional information about this device, forgetting about the APP1
port issue...

The MC8790 is really a Direct-IP capable device, but without DHCP. The
sierra_net kernel driver filters the wwan port in the following check:

/* test whether firmware supports DHCP */
if (!(status == sizeof(fwattr) &&
     (fwattr & SWI_GET_FW_ATTR_MASK))) {
    /* found incompatible firmware version */
    dev_err(&dev->udev->dev, "Incompatible driver and firmware"
            " versions\n");
    kfree(priv);
    return -ENODEV;
}

If that check is removed, we get a wwan0 which can get connected with
AT!SCACT, but DHCP won't work on it. I could get the static IP to set
with AT!SCPADDR, though, but I couldn't get proper gateway or DNS
addresses. After using the x.x.x.1 address as default route, I could
actually ping the Internet...


-- 
Aleksander