Re: [PATCH] qmi: sim: Implement query_facility_lock(LockedPins property)

Hi Denis,

On 18.11.24 г. 18:43 ч., Denis Kenzior wrote:
> Hi Ivo,
> 
> On 11/17/24 5:04 AM, Ivaylo Dimitrov wrote:
>> ---
>>   drivers/qmimodem/sim.c | 93 
>> ++++++++++++++++++++++++++++++++++++++++++++++++++
>>   1 file changed, 93 insertions(+)
>>
>> diff --git a/drivers/qmimodem/sim.c b/drivers/qmimodem/sim.c
>> index 1fffae2..b1d8f22 100644
>> --- a/drivers/qmimodem/sim.c
>> +++ b/drivers/qmimodem/sim.c
>> @@ -41,6 +41,8 @@ struct sim_status {
>>       uint8_t app_type;
>>       uint8_t passwd_state;
>>       int retries[OFONO_SIM_PASSWORD_INVALID];
>> +    uint8_t pin1_state;
>> +    uint8_t pin2_state;
>>   };
>>   struct sim_data {
>> @@ -52,6 +54,20 @@ struct sim_data {
>>       struct l_timeout *retry_timer;
>>   };
>> +struct query_locked_data {
>> +    enum ofono_sim_password_type passwd_type;
>> +};
> 
> Why is this structure needed?  Can't you simply stuff this into 
> cb_data::user using L_UINT_TO_PTR and L_PTR_TO_UINT?
> 

I was wondering what is GUINT_TO_POINTER() replacement in ell, however...

>> +
>> +static inline void cb_user_data_unref(void *user_data)
>> +{
>> +    struct cb_data *cbd = user_data;
>> +
>> +    if (cbd->ref == 1 && cbd->user)
>> +        l_free(cbd->user);
>> +
>> +    cb_data_unref(user_data);
>> +}
>> +
> 
> That way this could also be dropped
> 

I introduced that for the future (sim pin change functionality), 
however, looking now at the code, I wonder why I decided it will be needed.

Will drop and resend.

Thanks and regards,
Ivo