Re: serial connections

["Vitja Makarov" <vitja makarov gmail com>]

2008/4/29 Dan Williams <dcbw redhat com>:
>
> On Mon, 2008-04-28 at 22:57 +0400, Vitja Makarov wrote:
>  > I looked now at cdc-acm and found that some devices define cdc acm
>  > proto subclass that could be one of:
>  >
>  > - V25TER (my phone has this)
>  > - PCCA101
>  > - PCCA101-WAKE
>  > - GSM
>  > - 3G
>  > - CDMA
>  >
>  > I think this could be added to hal-info to provide device info.
>  > But my phone is definitely GSM, I don't know anout .V25ter,
>  > but it seems to be not only GSM modem.
>  >
>  > In case of modem devices that could be possible to use dynamic probing:
>  >
>  > my phone says this:
>  > < AT+GCAP
>  > > +GCAP: +CGSM,+DS,+W
>  >
>  > But in case of all other chips such as cp2101, ftdio_sio, cypress_m5 and so on,
>  > that should be bad idea to probe them. I not very familar with mobile phones,
>  > but I have a wild guess that they all CDC_ACM modems, and that should be
>  > very good.
>  >
>  > Btw HAL could get some info based on USB info, and if unsure try to
>  > probe the device.
>  > Isn't that a good idea?
>
>  Right, there should still probably be .fdi files that specify whether to
>  probe the port or not.  Or, invert the check and probe everything
>  _except_ what's listed in the .fdi file.  But probing devices that we
>  want to know about is the way to go.
>
>  Dan
>

I've posted patch to hal and here, that tries to probe CDC modems.
It uses callouts mechanism that it should be easy to specify which one
to probe right in .fdi.
I found one probelem with old discharged ubiq phone. When it boots(PC
is charger) it first appear as usb device,
and only after few(~1.5) minutes gives replies.

vitja.