>Is there any reason that the kernel version of printf has return type
>void? The return value (number of characters printed) is readily
>available from kprintf() (the internal version). Changing it to
>return an int would also make it have matching semantics with
>userspace, and with FreeBSD, amongst others.
>
>Originally, the kernel printf was indeed void, but we've had an
>int-returning kprintf since subr_prf.c rev 1.45:
>
>date: 1997/10/24 18:14:25;  author: chuck;  state: Exp;  lines: +841 -467
>import chris torek's vfprintf() from libc [renamed to kprintf and all
>floating point stuff removed].
>
>If noone objects, I'll make the trivial change in a couple of days.

i'm not objecting...just thinking that the lint fallout from such a
change will be staggering.  just a thought.

-- 
|-----< "CODE WARRIOR" >-----|
codewarrior@daemon.org             * "ah!  i see you have the internet
twofsonet@graffiti.com (Andrew Brown)                that goes *ping*!"
werdna@squooshy.com       * "information is power -- share the wealth."