Re: [PATCH 1/1] Set MNC length to 2 if EFad doesn't specify it

[Denis Kenzior <denkenz@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 1605 bytes --]

Hi Jonas,

On 09/13/2012 03:44 AM, Jonas Bonn wrote:
> Byte 4 of the EFAD structure, which specifies the MNC length, is optional.
> If this byte is not present, then there is no way to determine the length
> of the MNC.  As the, by far, most common length is 2, we might as well
> default to this value.

Depends on the continent.  On some it is not common at all.

>
> This patch changes the ofono_error to a warning as it's not really an
> error that we're encountering, but the user should nonetheless be given
> a heads-up that something might not work as expected.

What spec are you reading? 31.102 says it is mandatory ;)

>
> Signed-off-by: Jonas Bonn<jonas@southpole.se>

We do not use Signed-off-by here.

> ---
>   src/sim.c |    4 +++-
>   1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/src/sim.c b/src/sim.c
> index 4384eb0..f6109c6 100644
> --- a/src/sim.c
> +++ b/src/sim.c
> @@ -1760,7 +1760,9 @@ static void sim_ad_read_cb(int ok, int length, int record,
>   		return;
>
>   	if (length<  4) {
> -		ofono_error("EFad should contain at least four bytes");
> +		ofono_warn("EFad does not specify (optional) MNC length"
> +			   ", defaulting to 2");
> +		sim->mnc_length = 2;

And I'd rather not do this.  If mnc_length is unable to be determined, 
then oFono does not guess, but simply does not provide this information. 
  This hack might work for really old 2G devices which did not 
understand length 3 MNCs and always assumed length 2, but that is not 
really oFono's target.

>   		return;
>   	}
>

Regards,
-Denis