Problem with Sierra mc7710 LTE modem

[mansson]

Hi,

I have a new Vaio laptop with a built-in 4G LTE modem Sierra mc7710 that I can't get to work in Ubuntu, it works perfect in Windows.

The first problem seams to be that the radio is not enabled, it stays in Low power mode

AT!GSTATUS?
!GSTATUS: 
Current Time:  602              Temperature: 3t 5
Bootup Time:   1                Mode:        LOW POWER MODE 
System mode:   Unknown          PS state:    Not attached 
GMM (PS) state:DEREGISTERED     NO SERVICE     
MM (CS) state: IDLE             NO SERVICE

I have tried with AT+CFUN=1 to enable the radio but there is no difference. In Windows when I have enabled the modem AT!GSTATUS show that the modem is connected and not in Low Power Mode.

I have tried with different version of the kernel even 3.6.0 rc1 but without success. I have tried both via network-manager and with sierra ip direct

[bmork]

Hi,

I have a new Vaio laptop with a built-in 4G LTE modem Sierra mc7710 that I can't get to work in Ubuntu, it works perfect in Windows.

The first problem seams to be that the radio is not enabled, it stays in Low power mode

AT!GSTATUS?
!GSTATUS: 
Current Time:  602              Temperature: 3t 5
Bootup Time:   1                Mode:        LOW POWER MODE 
System mode:   Unknown          PS state:    Not attached 
GMM (PS) state:DEREGISTERED     NO SERVICE     
MM (CS) state: IDLE             NO SERVICE

I have tried with AT+CFUN=1 to enable the radio but there is no difference. In Windows when I have enabled the modem AT!GSTATUS show that the modem is connected and not in Low Power Mode.

What does

rfkill list

display? The MC7710 can be configured to either power off or enter low power mode on an external rfkill signal. Maybe the problem is that Sony has configured it for low power mode?

I have tried with different version of the kernel even 3.6.0 rc1 but without success. I have tried both via network-manager and with sierra ip direct

I don't think the DirectIP drivers will work with the default configuration of the MC7710 in a Vaio, but I am very interested in having that assumption confirmed. Is the device ID 1199:68a2? dmesg or lsusb will tell you.

A full dmesg dump right after boot is interesting in any case. There might be issues with some ACPI or WMI driver on such a new laptop, and this could very well cause problems like unwanted rfkill blocking.

BTW, Did you build the qmi_wwan driver in the 3.6.0 rc1 kernel? Did it load successfully, giving you /dev/cdc-wdmX and wwanY devices?

[mansson]

What does

rfkill list

display? The MC7710 can be configured to either power off or enter low power mode on an external rfkill signal. Maybe the problem is that Sony has configured it for low power mode?

Yes it seams like it's low power mode. I had actually to enable the modem in Windows or I wouldn't see it at all with lsusb or /dev/ttyusb0-2 didn't show up neither. When I have enabled the modem in Windows with the Sony network tool kit then it showed up in Linux as well.

I don't think the DirectIP drivers will work with the default configuration of the MC7710 in a Vaio, but I am very interested in having that assumption confirmed. Is the device ID 1199:68a2? dmesg or lsusb will tell you.

The device ID is correct
Bus 001 Device 006: ID 1199:68a2 Sierra Wireless, Inc.

rfkill list
0: sony-wifi: Wireless LAN
    Soft blocked: no
    Hard blocked: no
1: sony-bluetooth: Bluetooth
    Soft blocked: no
    Hard blocked: no
2: sony-wwan: Wireless WAN
    Soft blocked: no
    Hard blocked: no
3: hci0: Bluetooth
    Soft blocked: no
    Hard blocked: no
4: phy0: Wireless LAN
    Soft blocked: no
    Hard blocked: no

A full dmesg dump right after boot is interesting in any case. There might be issues with some ACPI or WMI driver on such a new laptop, and this could very well cause problems like unwanted rfkill blocking.

BTW, Did you build the qmi_wwan driver in the 3.6.0 rc1 kernel? Did it load successfully, giving you /dev/cdc-wdmX and wwanY devices?

The dmesg dump is here http://pastebin.com/GUdTxykG

/dev/cdc-wd1 are there but not /dev/wwanY

[bmork]

Yes it seams like it's low power mode. I had actually to enable the modem in Windows or I wouldn't see it at all with lsusb or /dev/ttyusb0-2 didn't show up neither. When I have enabled the modem in Windows with the Sony network tool kit then it showed up in Linux as well.

And it was after this it was locked in low power mode in Linux?

The device ID is correct
Bus 001 Device 006: ID 1199:68a2 Sierra Wireless, Inc.

Right. That is what I expected. It means that the MC7710 card is in QMI mode, and supported by the qmi_wwan and qcserial drivers in recent kernels.

rfkill list
0: sony-wifi: Wireless LAN
    Soft blocked: no
    Hard blocked: no
1: sony-bluetooth: Bluetooth
    Soft blocked: no
    Hard blocked: no
2: sony-wwan: Wireless WAN
    Soft blocked: no
    Hard blocked: no
3: hci0: Bluetooth
    Soft blocked: no
    Hard blocked: no
4: phy0: Wireless LAN
    Soft blocked: no
    Hard blocked: no

Well, that looks good. The sony-wwan device is recognized, and is not blocked.

The dmesg dump is here http://pastebin.com/GUdTxykG

Weird. I cannot see the module mentioned at all there. Which is obviously bogus as it shows up in the lsusb output and drivers are loaded and working (must be as you can issue AT commands to the module).

/dev/cdc-wd1 are there but not /dev/wwanY

The last one should be a netdevice. It will show up as wwan0 if you do
"ifconfig -a" or "ip link show".


I have tried a few settings on my MC7710 module, but I am unable to recreate a problem similar to this. Could you try a few AT show commands and see if there is something obvious in the output (I am including the output from my module for reference):

at!custom?
!CUSTOM: 
            PUKPRMPT            0x01
            MEPCODE             0x01
            ISVOICEN            0x02
            PRLREGION           0x01
            PCSCDISABLE         0x03
            GPSENABLE           0x01
            GPSLPM              0x01


OK
at!pcinfo?
State: 1 (ONLINE)
LPM force flags - W_DISABLE:0, User:0, Temp:0, Volt:0
W_DISABLE: 0
Poweroff mode: 1
LPM Persistent: 0


OK
at+cfun=0
OK
at!pcinfo?
State: 0 (LPM)
LPM force flags - W_DISABLE:0, User:1, Temp:0, Volt:0
W_DISABLE: 0
Poweroff mode: 1
LPM Persistent: 0


OK

The "LPM force flags" should indicate why the module is in LPM. "W_DISABLE" is the PCIe input pin used by the rfkill switch. "User" means that LPM was enabled using AT commands (or another user control protocol like CnS or QMI). "Temp" and "Volt" means the module powered down due to sensor readings outside the acceptable range. And I believe the "LPM Persistent" setting is also interesting here. But even if I tried to enable it, I could not recreate your problem. The module stil worked as expected, starting up in whatever mode it was shut down in, but still allowing me to enter online mode using AT+CFUN=1

[mansson]

And it was after this it was locked in low power mode in Linux?

It's in low power mode in Linux, before I enabled the modem in Windows it was not visible at all in Linux, lsusb didn't show the modem.

Right. That is what I expected. It means that the MC7710 card is in QMI mode, and supported by the qmi_wwan and qcserial drivers in recent kernels.

Well, that looks good. The sony-wwan device is recognized, and is not blocked.
Weird. I cannot see the module mentioned at all there. Which is obviously bogus as it shows up in the lsusb output and drivers are loaded and working (must be as you can issue AT commands to the module).

The last one should be a netdevice. It will show up as wwan0 if you do
"ifconfig -a" or "ip link show".

ifconfig -a show
wwan0     Link encap:Ethernet  HWaddr 1a:a8:9d:3d:e3:ca  
          BROADCAST MULTICAST  MTU:1500  Metric:1
          RX packets:0 errors:0 dropped:0 overruns:0 frame:0
          TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:0 (0.0 B)  TX bytes:0 (0.0 B)

I have tried a few settings on my MC7710 module, but I am unable to recreate a problem similar to this. Could you try a few AT show commands and see if there is something obvious in the output (I am including the output from my module for reference):

at!custom?
!CUSTOM: 
            PUKPRMPT            0x01
            MEPCODE             0x01
            ISVOICEN            0x02
            PRLREGION           0x01
            PCSCDISABLE         0x03
            GPSENABLE           0x01
            GPSLPM              0x01


OK
at!pcinfo?
State: 1 (ONLINE)
LPM force flags - W_DISABLE:0, User:0, Temp:0, Volt:0
W_DISABLE: 0
Poweroff mode: 1
LPM Persistent: 0


OK
at+cfun=0
OK
at!pcinfo?
State: 0 (LPM)
LPM force flags - W_DISABLE:0, User:1, Temp:0, Volt:0
W_DISABLE: 0
Poweroff mode: 1
LPM Persistent: 0


OK

at!custom?
!CUSTOM: 
            PUKPRMPT            0x01
            MEPCODE             0x01
            ISVOICEN            0x02
            PRLREGION           0x01
            PCSCDISABLE         0x03
            GPSENABLE           0x01
            AUTONETWORKMODE     0x02
            GPSLPM              0x01
            GPSSEL              0x01

at!pcinfo?
State: 0 (LPM)
LPM force flags - W_DISABLE:1, User:1, Temp:0, Volt:0
W_DISABLE: 1
Poweroff mode: 0
LPM Persistent: 0

at+cfun=0
OK

at!pcinfo?
State: 0 (LPM)
LPM force flags - W_DISABLE:1, User:1, Temp:0, Volt:0
W_DISABLE: 1
Poweroff mode: 0
LPM Persistent: 0


OK

OK

The "LPM force flags" should indicate why the module is in LPM. "W_DISABLE" is the PCIe input pin used by the rfkill switch. "User" means that LPM was enabled using AT commands (or another user control protocol like CnS or QMI). "Temp" and "Volt" means the module powered down due to sensor readings outside the acceptable range. And I believe the "LPM Persistent" setting is also interesting here. But even if I tried to enable it, I could not recreate your problem. The module stil worked as expected, starting up in whatever mode it was shut down in, but still allowing me to enter online mode using AT+CFUN=1

I tried CFUN=1 but it's the same output

OK
at+cfun=1
OK
at!pcinfo?
State: 0 (LPM)
LPM force flags - W_DISABLE:1, User:1, Temp:0, Volt:0
W_DISABLE: 1
Poweroff mode: 0
LPM Persistent: 0

OK

[bmork]

at!pcinfo?
State: 0 (LPM)
LPM force flags - W_DISABLE:1, User:1, Temp:0, Volt:0
W_DISABLE: 1
Poweroff mode: 0
LPM Persistent: 0


OK

Right. That was very useful. It confirms my first thought: The W_DISABLE input is asserted, and the card is configured to go to LPM instead of powering down on this signal.

So that means the the rfkill output was wrong. The rfkill signal is in fact active, and that is why you cannot disable LPM. I have no idea how to fix that, but it could be a simple matter of updating the sony-laptop driver with details of this laptop. I suggest you contact the maintainer of that driver with as much info as possible about your laptop and the problem. According to the get_maintainer script, that would be

Mattia Dongili <malattia@linux.it> (maintainer:SONY VAIO CONTROL...)
Matthew Garrett <mjg@redhat.com> (maintainer:X86 PLATFORM DRIVERS)
platform-driver-x86@vger.kernel.org (open list:SONY VAIO CONTROL...)

[mansson]

Thank youBjørn, I will do that

[bmork]

Just for reference, this bug now seems to be resolved upstream:
https://bugzilla.kernel.org/show_bug.cgi?id=47751


Bjørn

[iklymov]

Well, actually no (I'm the author of that bugreport). I can connect to modem, it registers in network, but never connects for actual data transfer:

AT!GSTATUS?
!GSTATUS: 
Current Time:  1350             Temperature: 32
Bootup Time:   1                Mode:        ONLINE         
System mode:   WCDMA            PS state:    Attached     
WCDMA band:    WCDMA 2100 
WCDMA channel: 10662
GMM (PS) state:REGISTERED       NORMAL SERVICE 
MM (CS) state: IDLE             NORMAL SERVICE 

WCDMA L1 State:L1M_PCH_SLEEP    RRC State:   DISCONNECTED   
RX level C0:   -68              LAC:         1389 (5001)
RX level C1:   -106             Cell ID:     01F5C481 (32883841)


OK
AT+CGDCONT?
+CGDCONT: 1,"IP","3g.utel.ua","0.0.0.0",0,0
+CGDCONT: 2,"IP","3g.utel.ua","0.0.0.0",0,0

OK
AT!SCACT=1,1
+CME ERROR: no network service

Any suggestions? (PPP mode do not work also, i can connect with this APN on my windows)

[bmork]

Well, actually no (I'm the author of that bugreport). I can connect to modem, it registers in network, but never connects for actual data transfer:

AT!SCACT=1,1
+CME ERROR: no network service

Any suggestions? (PPP mode do not work also, i can connect with this APN on my windows)

PPP should work, so there are probably more issues here...

But the AT!SCACT error is expected. That will only work if the modem is in DirectIP mode, and AFAIK the Vaio module is running in QMI mode by default. I would leave it that way too, if I were you. Sony may whitelist these modules, and changing the mode will change the USB PID. If they are as difficult as Lenovo then the laptop won't even boot if the whitelist check fails.

To connect the wwan0 interface in QMI mode you will need to use QMI. ModemManager support for that is in progress, but not quite ready for production yet. But you can use the cli tools coming with libqmi (which the MM support will be based on). See
http://lists.freedesktop.org/mailman...o/libqmi-devel
http://cgit.freedesktop.org/libqmi

I typically connect manually like this:

 qmicli -d /dev/cdc-wdm0  --dms-uim-verify-pin=PIN,xxxx
 qmi-network /dev/cdc-wdm0 start
 dhclient -d -4 wwan0

This will of course be a lot simpler once the MM support is ready and NetworkManager supports the new MM version (which is not quite there yet, either)

Note that all this is about wwan0. Using ppp over the ttyUSBx supporting AT commands (most likely ttyUSB2) should just work as with any 3G modem. So if you can't get that working then there is something else we need to sort out.


Bjørn