Any special reason set cid=0 as invalid CID

[Dan Williams]

On Wed, 2019-02-27 at 14:55 +0800, 王道之 wrote:
> Hey
> When I read the MM code, the bearer treat the cid=0 as a invalid CID
> confused me.  But I think there may be some special reasons.
> In the MM Ver1.10.0 source code, src/mm-broadband-bearer.c

Older versions of the ETSI 27.007 spec specified the minimum PDP
context number as 1. But at least v11+ says:

"The range of permitted values (minimum value = 1 or if the
initial PDP context is supported (see subclause 10.1.0), minimum value
= 0) is returned by the test form of the
command."

MM didn't get updated to account for that. So there is no special
reason except history, and that can change :)

Dan

> *static guint*
> *cid_selection_3gpp_finish (MMBroadbandBearer  *self,*
> *                           GAsyncResult       *res,*
> *                           GError            **error)*
> *{*
> *    gssize cid;*
> 
> *    /* We return 0 as an invalid CID, not -1 */*
> *    cid = g_task_propagate_int (G_TASK (res), error);*
> *    return (guint) (cid < 0 ? 0 : cid);*
> *}*
> 
> As my Intel 7560 module(not in Mbim on Linux),  after set AT+COPS=0 ,
> it
> would auto get IP on the cid=0 PDP.
> Then the MM logic would treat the cid=0 as invalid, after that MM
> would
> creat a new PDP context with +CGDCONT=1,*,*.
> Though it works at the end. But I think it is better to use the auto
> connect one that cid=0.
> So what are the special reasons for the cid=0 invalid? Or maybe can
> we use
> *#define INVALID_CID G_MAXUINT*
> as a invalid cid?
> 
> Quincy
> _______________________________________________
> ModemManager-devel mailing list
> ModemManager-devel at lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/modemmanager-devel