dev_t / udev_t confusion ?

[Poul-Henning Kamp]

In message <40C742A0.5090704 at freebsd.org>, Scott Long writes:

>> 	cdevsw->ioctl() takes a caddr_t pointer argument which
>> 	really should be a void *, do we want to change that
>> 	as well (since it is all the same files we'll have to
>> 	change).
>> 
>
>Is this going to have any consequences on COMPAT_LINUX code or anything
>else that calls ioctl() through obscure means?

Well, it has more impact for the places which implement the ioctl,
in particular if they do pointer arithmetic on the pointer (legal
on caddr_t but illegal on void *).

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.