Description of problem: Version-Release number of selected component (if applicable): 0.7.997 How reproducible: Plug in Modem, need to plug in many times wait for the CD Drive and Micro SD slot to load. Sometimes NetworkManager show the device. Needs at least 2 or three attempts. Steps to Reproduce: 1. Plug in modem 2. Wait for CD and SD to load 3. Check that dmesg has loaded ttyUSB0 1 2 4. Look at Networkmanager 5. Device not shown Actual results: Device not detected and can only connect after many retries. Expected results: Modem is shown and can connect Additional info: This problem only appeared after recent update.
Also to be noted sometimes show that the modem is in a state of flux. Sometimes it loaded then it disconnects and reconnects (looking at dmesg). I am luckly if it works at all now.
Are you using usb_modeswitch to flip the device to modem mode? That needs to be done first before the modem can really be used. This should usually happen in the kernel, but if the modems USB IDs are yet in the kernel drivers then it may not happen for your device. usb_modeswitch should always be able to flip the device to modem mode for you though.
It started at first when network-manager started to refuse to detect the modem - remember it has been working before. Then I installed usb_modeswitch. The program seemed to have no effect. And in fact seemed to make it worse. To be clear this is Vodafone k3715 which is the same as the E620 USB ID: 12d1:1001 The overview. 1. I installed F12 x86 - all ok no problems modem connected and worked fine 2. After about 1 to 2 months modem stops working. 3. Tried with usb_modeswitch but it now seemed to flip flop modem connects and disconnects. I fixed it by reverting to F11. If you note you have been having problems with this even in beta. Additional: The same problem did arise in F11 but only after I installed yum delta rpm's. Using F11 with no delta RPM's the has not problem occurred. Steven
So you're saying usb_modeswitch makes the device angry? You have to configure it's config file to switch the device you're trying to use; make sure you get the right one :) If that's not working for some reason, can we get some logs from 'dmesg' for when the device has a fit and refuses to play with others? Note that *lots* of Huawei devices use the 12d1:1001 IDs; like my E160G and the E169. Yeah, it's wrong, but nothing we can do about it. Try enabling the E169 section in usb_modeswitch.conf.
No luck with that but in searching the web I cam across this bug report at ubuntu https://bugs.launchpad.net/ubuntu/+source/linux/+bug/446146 Has this kernel been released?
Unsure... 2.6.31.5 was supposed to have the patch that fixed this, but perhaps not completely or for all devices? What kernel version are you currently running? The command "uname -a" in a terminal will tell you.
sorry for the delay, have been using F11 to fix the problem but now have a anther drive that I have just installed F12 USBMode switch Found this code in Fedora forums # Huawei E620 - works for E122 DefaultVendor = 0x12d1 DefaultProduct = 0x1446 MessageEndPoint = "0x01" MessageContent = "55534243000000000000000000000011060000000000000000000000000000" but F11 does not need usbmode switch. It appears that the device is reseting with and error option_instat_callback error -108 Will give you the info for above once I try the above. (Need to restart each time) Steven
The above did not work, Like I said in F11 I did not need usb_modeswitch it was automatic. still get the option driver error. kernel version (is the latest as I have just done an upgrade) 2.6.31.12-174.2.22.fc12.x86_64
Just updated to the latest kernel Problem still persists, please advise what is happening Will test out other distros and see if is a common bug
Please report on what is happening Steven
Can you grab 'lsusb -v' for the device right after you plug it in? Also, what is the contents of both of these files? /lib/udev/rules.d/61-mobile-action.rules /lib/udev/rules.d/61-option-modem-modeswitch.rules ? If either of these files has a line that begins with ATTRS{idVendor}=="12d1" then we may have a modeswitching conflict that should already be fixed upstream but may not have found its way to Fedora yet.
Created attachment 402244 [details] lsusb attached lsusb as requested
Created attachment 402245 [details] moble-action.rules
Created attachment 402246 [details] option modem rules
As you can neither of these files have ATTRS{idVendor}=="12d1" but neither did F11. Any thing else I can do?
Please advise what is happening
Next thing to do is to grab the output of 'dmesg' after you notice the device not showing up about 20 seconds after you plug it in. If we get that we can see what might be going on. You're right, I don't think it's udev-related right now, though we may still need a modeswitch at some point.
Created attachment 405724 [details] dmesg dmesg is attached, thanks for your help. I note the following interesting errors. option: option_instat_callback: error -108 and modem-manager[1729]: segfault at 20 ip 0000003a1800d8da sp 00007fff882066a0 error 4 in libgobject-2.0.so.0.2200.5[3a18000000+43000]
Dan what is happening, this should be sorted before F13
Any Progress?
Update: My modem died completely, after repair I was able to connect under F13 beta. Why it was working under F11 and not F12 and F13 is still a mystery, but it does look like a hardware problem. I am sorry to waste your time, but I did think it was a software regression issue as F11 was working. There is another issue but i'll open another bug report if it has not been reported. (Mobile internet switch problem) I'll mark this bug as not a bug