re: patch: debug instrumentation for dev/raidframe/rf_netbsdkintf.c

[matthew green <mrg%eterna.com.au@localhost>]

> >>> @@ -472,6 +472,9 @@
> >>>  	const char *bootname = device_xname(bdv);
> >>>  	size_t len = strlen(bootname);
> >>>
> >>> +	if (bdv == NULL)
> >>> +		return 0;
> >>> +
> >>
> >> This looked suspicious, even before I read the code.
> >>
> >> The question is if it is ever legitimate for bdv to be NULL.
> >
> > infact, bdv has already been dereferenced at this point:
> > the assignment to bootname 3 lines up.
> 
> So, if we're going to insert a KASSERT() we would need to separate
> the assignment from bootname's declaration, and do the assignment
> after the KASSERT()

i am not a huge fan of KASSERT(ptr != NULL).  page fault is
pretty much just as clear as the assert, and does not need
special code to handle it.

that said, i do see some utility in it as a guide to others
reading this code, and won't object to this usage.


.mrg.