Activation Codes and Methods, Hardware Details, Sniffing
Post Reply
dzlabing
Posts: 2
Joined: 09 Dec 2007, 01:13

Successfully switched the Option Icon72

Post by dzlabing » 09 Dec 2007, 01:19

Hi,

I just managed to switch my Option Icon 72 to modem-mode. The problem was that the command sequence in the configuration of "usb_modeswitch" did not work, and the commands I found sniffing the Windows driver did not work either. Finally I succeeded by sending the whole sequence of commands.

You can find a modified version of icon_switch in my blog:
https://www.zlabinger.at/blog/2007/12/0 ... der-linux/

I did not try to modify usb_modeswitch, but it should be straight forward to implement the same changes to this program.

Dietmar
dietmar.zlabinger@gmail.com

Poodle
Posts: 6
Joined: 07 Dec 2007, 15:45

Post by Poodle » 12 Dec 2007, 16:24

The newer 7.2 devices from Option use an improved scheme for ZeroCD. The VID/PID 056Ch/1000h is dropped. Instead, the device uses the same VID/PID (0AF0h/xxxxh) for both cd-rom and 3G modem.

On MacOSX and Windows, initially the device is recognised as a mass storage device (class code 08h/06h/50h). The installation software runs off the cd-rom and associates the VID/PID with the 3G driver, overriding the mass storage driver association.

The next time the device is plugged in, the OS automatically loads the 3G driver. The 3G driver sends the mode-switch command.

Josh
Site Admin
Posts: 6570
Joined: 03 Nov 2007, 00:30

Post by Josh » 13 Dec 2007, 01:25

Regarding the new Option devices, the obvious consequence of all this is that we need a sniffing log of a plug-in with the Windows driver already installed (only the first seconds).

Dzlabing, do you think you could provide one?

Also, according to what poodle says, the switching should be happening right after the init phase of the device.

Poodle
Posts: 6
Joined: 07 Dec 2007, 15:45

Post by Poodle » 14 Dec 2007, 12:15

Not really. You still have to send the REZERO UNIT SCSI command as with older Option devices.

The main difference between older and newer devices is the exposed USB interface when you plug in the data card:
* Old: VID/PID 056Ch/1000h, class 08h/06h/50h
* New: VID/PID 0AF0h/xxxxh, class 08h/06h/50h

So the newer devices come up with an Option ID but with mass storage class. The switch program should check for this. When USB interface #0 is mass storage, then the switch command must be send (unless the 3G driver already does this ofcourse).

Josh
Site Admin
Posts: 6570
Joined: 03 Nov 2007, 00:30

Post by Josh » 14 Dec 2007, 20:16

O.K., I finally got it.

I will update the docs and the config file after I'm back from a little journey. And I will streamline the Option switching in the prog.

Thanks for all your expert input, poodle :!:

dzlabing
Posts: 2
Joined: 09 Dec 2007, 01:13

Post by dzlabing » 18 Dec 2007, 01:55

Hi,

sorry for replying late.
I've just created a new log, its a bit messy, but it starts with a normal login (at-commands), then I uninstall the windows driver, the nI get the
"55 53 xx" sequences in the middle of the log.

https://www.zlabinger.at/blog//wp-conte ... /snoop.zip
(will be deleted by 31-dec-07)

My device identifies itself originally as vendor 05c6 product 1000 (Qualcom), after switching it reappears as vendor 0af0 product 6901 (Option).

I can confirm that the initial id disapperars, but under Windows it is no longer accessible as zero-CD after switching (at least on my machine).

Dietmar

Poodle
Posts: 6
Joined: 07 Dec 2007, 15:45

Post by Poodle » 02 Jan 2008, 12:19

dzlabing wrote: My device identifies itself originally as vendor 05c6 product 1000 (Qualcom), after switching it reappears as vendor 0af0 product 6901 (Option).
In this case, your device uses the old style zero-CD. There should be a service running named GtFlashSwitch. This service sends the mode-switch command.

When you disable this service and then plug in the device, then the device isn't switched over at all.

Post Reply