Potential bug / problematic code in gsm411_rx_rl_data()
In gsm411_sms.c the function gsm411_rx_rl_data receives "struct gsm48_hdr *gh" as input then in very first line typecasts the pointer to "struct gsm411_rp_hdr *rp_data" to access its "data" field. struct gsm411_rp_hdr *rp_data = (struct gsm411_rp_hdr*)&gh->data; But the two header structures have their "data" fields offset by one byte as in: struct gsm411_rp_hdr {
uint8_t len; uint8_t msg_type; uint8_t msg_ref; uint8_t data[0]; } __attribute__ ((packed));
struct gsm48_hdr { uint8_t proto_discr; uint8_t msg_type; uint8_t data[0]; } __attribute__((packed));
Obviously this displacement has been compensated for elsewhere in the code as the application works. But this seems to be inadvertent. And if it is deliberate, it is risky programming practice and could create problems later on. Request you to correct and update suitably. Thanks. B.
In gsm411_sms.c the function gsm411_rx_rl_data receives "struct gsm48_hdr *gh" as input then in very first line typecasts the pointer to "struct gsm411_rp_hdr *rp_data" to access its "data" field.
struct gsm411_rp_hdr *rp_data = (struct gsm411_rp_hdr*)&gh->data;
That's not what it does .... you need to either re-read that line or re-read a book on C programming. What is type casted is the pointer to gsm48_hdr data field. And that's because the data field of the gsm48_hdr contains a gsm411_rp_hdr in this case.
Request you to correct and update suitably.
Well, I request you check your claims before making them ... Cheers, Sylvain
participants (2)
-
Bhaskar11 -
Sylvain Munaut