Error reading from istream: Error reading from file descriptor: No such device

[Nicolas RENAULT]

Hi,

Le 06/03/2020 à 09:38, Aleksander Morgado a écrit :
> Hey,
>
>> I have a problem :
>>
>> * Debian stretch  / kernel 4.9.0.12
>> * Sierra Wireless, Incorporated  MC7304
>> * libqmi 1.22.6
>>
>> when launching connections :
>>
>> Loading profile at /etc/qmi-network.conf...
>>      APN: free
>>      APN user: free
>>      APN password: free
>>      qmi-proxy: no
>> Checking data format with 'qmicli -d /dev/cdc-wdm1 --wda-get-data-format '...
>> [06 Mar 2020, 07:53:15] -Warning ** [/dev/cdc-wdm1] requested auto mode but no MBIM QMUX support available
>> Device link layer protocol retrieved: 802-3
>> Getting expected data format with 'qmicli -d /dev/cdc-wdm1 --get-expected-data-format'...
>> [06 Mar 2020, 07:53:15] -Warning ** [/dev/cdc-wdm1] requested auto mode but no MBIM QMUX support available
>> Expected link layer protocol retrieved: 802-3
>> Device and kernel link layer protocol match: 802-3
>> Starting network with 'qmicli -d /dev/cdc-wdm1 --wds-start-network=apn='free',username='free',password='free',ip-type=4  --client-no-release-cid '...
>> [06 Mar 2020, 07:53:15] -Warning ** [/dev/cdc-wdm1] requested auto mode but no MBIM QMUX support available
>> [06 Mar 2020, 07:53:22] -Warning ** Error reading from istream: Error reading from file descriptor: No such device
>> error: operation failed: operation failed and couldn't be aborted: Cannot write message: Error writing to file descriptor: No such device
>> Saving state at /tmp/qmi-network-state-cdc-wdm1... (CID: 9)
>> error: network start failed, no packet data handle
>> [06 Mar 2020, 07:54:00] -Warning ** [/dev/cdc-wdm1] requested auto mode but no MBIM QMUX support available
>> error: couldn't release client: QMI protocol error (3): 'Internal'
>> Clearing state at /tmp/qmi-network-state-cdc-wdm1...
>>
>>
>> I deploy libqmi on lot of the same hardware but got problem on the two last installation.
>> I think the main events is "Error reading from istream: Error reading from file descriptor: No such device"
>>
> Yes that is the main event. This totally looks like the device reboots
> itself for some reason? Can you run "dmesg -w" while you run the test
> to see if the device goes away and comes back?
>
during the ifup , on dmesg

[ 5470.976648] qmi_wwan 1-4.4:1.10: nonzero urb status received: -71
[ 5470.976938] qmi_wwan 1-4.4:1.10: wdm_int_callback - 0 bytes
[ 5471.072679] qmi_wwan 1-4.4:1.10: nonzero urb status received: -71
[ 5471.072963] qmi_wwan 1-4.4:1.10: wdm_int_callback - 0 bytes
[ 5471.168718] qmi_wwan 1-4.4:1.10: nonzero urb status received: -71
[ 5471.169004] qmi_wwan 1-4.4:1.10: wdm_int_callback - 0 bytes
[ 5471.170556] usb 1-4.4: USB disconnect, device number 11
[ 5471.171488] qcserial ttyUSB0: Qualcomm USB modem converter now 
disconnected from ttyUSB0
[ 5471.171598] qcserial 1-4.4:1.0: device disconnected
[ 5471.172970] qcserial ttyUSB1: Qualcomm USB modem converter now 
disconnected from ttyUSB1
[ 5471.173078] qcserial 1-4.4:1.2: device disconnected
[ 5471.174313] qcserial ttyUSB2: Qualcomm USB modem converter now 
disconnected from ttyUSB2
[ 5471.174422] qcserial 1-4.4:1.3: device disconnected
[ 5471.174908] qmi_wwan 1-4.4:1.8 wwan0: unregister 'qmi_wwan' 
usb-0000:00:14.0-4.4, WWAN/QMI device
[ 5471.202621] qmi_wwan 1-4.4:1.10 wwan1: unregister 'qmi_wwan' 
usb-0000:00:14.0-4.4, WWAN/QMI device

error Warning ** Error reading from istream: Error reading from file 
descriptor: No such device

[ 5482.373961] usb 1-4.4: new high-speed USB device number 12 using xhci_hcd
[ 5482.474860] usb 1-4.4: config 1 has an invalid interface number: 8 
but max is 5
[ 5482.474875] usb 1-4.4: config 1 has an invalid interface number: 10 
but max is 5
[ 5482.474885] usb 1-4.4: config 1 has an invalid interface number: 11 
but max is 5
[ 5482.474895] usb 1-4.4: config 1 has no interface number 1
[ 5482.474905] usb 1-4.4: config 1 has no interface number 4
[ 5482.474913] usb 1-4.4: config 1 has no interface number 5
[ 5482.475729] usb 1-4.4: New USB device found, idVendor=1199, 
idProduct=68c0, bcdDevice= 0.06
[ 5482.475744] usb 1-4.4: New USB device strings: Mfr=1, Product=2, 
SerialNumber=3
[ 5482.475753] usb 1-4.4: Product: MC7304
[ 5482.475762] usb 1-4.4: Manufacturer: Sierra Wireless, Incorporated
[ 5482.484238] qcserial 1-4.4:1.0: Qualcomm USB modem converter detected
[ 5482.485152] usb 1-4.4: Qualcomm USB modem converter now attached to 
ttyUSB0
[ 5482.487081] qcserial 1-4.4:1.2: Qualcomm USB modem converter detected
[ 5482.487251] usb 1-4.4: Qualcomm USB modem converter now attached to 
ttyUSB1
[ 5482.488371] qcserial 1-4.4:1.3: Qualcomm USB modem converter detected
[ 5482.488529] usb 1-4.4: Qualcomm USB modem converter now attached to 
ttyUSB2
[ 5482.490115] qmi_wwan 1-4.4:1.8: cdc-wdm0: USB WDM device
[ 5482.490566] qmi_wwan 1-4.4:1.8 wwan0: register 'qmi_wwan' at 
usb-0000:00:14.0-4.4, WWAN/QMI device, b6:a2:88:6d:70:90
[ 5482.491955] qmi_wwan 1-4.4:1.10: cdc-wdm1: USB WDM device
[ 5482.492544] qmi_wwan 1-4.4:1.10 wwan1: register 'qmi_wwan' at 
usb-0000:00:14.0-4.4, WWAN/QMI device, b6:a2:88:6d:70:90

effective modem reboot I thinks.

So now , what is "qmi_wwan 1-4.4:1.10: nonzero urb status received: -71"



>> Do I have the take care of "Warning ** [/dev/cdc-wdm1] requested auto mode but no MBIM QMUX support available".
>>
> No, you can ignore that. I think we should avoid printing that warning
> if MBIM support was not built in libqmi, it's very misleading. Again,
> just ignore that specific message.
>
>
regards,

Nicolas.