Dan Williams <dcbw@xxxxxxxxxx> writes: > On Wed, 2013-05-01 at 07:53 +0200, Teppo Kotilainen wrote: >> I've been testing this device for some time and it seems to work fine. >> >> some information about the device: >> https://telewell.fi/en/product/3g4glte-products/TW-LTE4G/tw-lte-4g3g-modem >> >> Acked-By: Matthias Urlichs <matthias@xxxxxxxxxx> > > While you're at it, can you post full lsusb -v output for this device? > Given it's an LTE device, there's a good chance it's got a Qualcomm > chipset, and thus there's a good chance it speaks QMI too. > > Does it provide a network interface, eg wwan0 or usb0? The lsusb -v output will be useful, but I'm afraid we'll need more. I don't know what the idea is, but ZTE hide as much information as they can. All functions are ff/ff/ff and there are usually no associated string descriptors or functional descriptors which can help us guess. And just to top it all off: We've registered ZTE devices with the QMI function on interface number 1, 2, 3, 4, 5 and 6. There just is no system... So we either need probing or Windows *.inf files to be able to guess where the (possible) QMI function is. If you have a Windows installation then it is easies to let the modem install its drivers there and then fetch the *.inf files. You might also find these files on the embedded driver CD, but ZTE are often making things annoyingly difficult here as well, hiding the driver files inside an installer application. If you want to try probing, a recent ModemManager should do most of the job automatically. Run it in debug mode with the option driver bound to all interfaces first (like your patch will do). MM will likely find one or more AT ports, a QCDM port and possibly more serial functions. Note which interfaces MM were unable to detect, and unbind the option driver from them, using echo x-y:z.w >/sys/bus/usb/drivers/option/unbind (replacing "x-y:z.w" with the actual interface device name for each interface). After that, add the device dynamically to qmi_wwan, using modprobe qmi_wwan echo 19d2 0412 >/sys/bus/usb/drivers/qmi_wwan/new_id Hopefully qmi_wwan will now bind to one or more of the interfaces you unbound above. This is still no definite sign of success. Rerun ModemManager (with QMI support enabled) in debug mode to probe them. Take note of which interface was successfully probed as QMI and send the result here. Or just prepare and submit a patch for qmi_wwan and a matching blacklist patch for option. BTW, Greg will need a Signed-off-by for your option patch. Please see https://www.kernel.org/doc/Documentation/SubmittingPatches Bjørn -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
