Re: [lwip-users] Optimistic error handling in lwIP

[Kieran Mansley]

In general I agree that returning error codes from functions is a good
idea (and I much prefer having an non-void function or error code
argument rather than using the errno variable).

However, I'm not sure if I've missed the point of your mbox_fetch
example.  If it gets interrupted, should sys_arch_mbox_fetch not just try
again?  If we checked the result of sys_mbox_fetch to check if it was
NULL, all it would do as a result is call sys_mbox_fetch again and hope it
succeeded the second time.  Therefore, why not keep the majority of code
simple and check for a NULL return in sys_arch_mbox_fetch, and only return
when you have something?  It seems to me that the assumption the lwip code
is making (that sys_mbox_fetch only returns when it has a message) is
valid, or at least *should* be valid if the sys_arch has been properly
implemented.

Kieran