net: Another duplicate function to get pkt address

[pfalcon]

Following #4421 , this is 3rd function with almost the same functionality (previous duplication case fixed). Well, this time, it has signature:

static void pkt_get_sockaddr(sa_family_t family, struct net_pkt *pkt,
                             struct sockaddr_ptr *addr)

i.e., it takes struct sockaddr_ptr*, whereas net_pkt_get_dst_addr() takes struct sockaddr*. Still, I guess it's worth looking into to avoid duplication.

[pfalcon]

@askawu : Let me know if you're interested to look into this, based on the previous experience. If not, this is in my queue.

[askawu]

@pfalcon , please assign this to me. :)

[pfalcon]

@askawu : Done, thanks!

[pfalcon]

@askawu : Ping, didn't hear from you for a while, is this still on your plate?

[nashif]

you should try to assign bugs to one person only, otherwise everyone will be relying on the others

[laperie]

it's actually a feature, not a bug ;)

[pfalcon]

it's actually a feature, not a bug ;)

Well, it's definitely not a "bug", but:

1. Objectively, a missed optimization opportunity. I didn't hear about code size comparison of Zephyr's IP stack with other IP stack, but given such cases, it's hard to shake off a hunch that we don't compare favorably.
2. It shows that we have some inconsistency/technical debt. Actually, somewhere around 1.10 I got an idea that we have a pretty vast and featureful stack, and no single party (except for hopefully @jukkar) knows it well enough, so there's a chance that a particular developer when adding new feature might reinvent/duplicate some functionality (even if partially).

I don't think that we're ready for full-fledged optimization (that's tracked in #3547), but IMHO, p.2 is a real concern, especially is we're talking about things like MISRA C (what's the point of MISRA C, which tells us how to write expressions and statements, if we have problems with how code is being written?).

Anyway, in the current situation, this is definitely prio:low.