Le 09 Mar 2003 10:57:59 -0500
Perry E. Metzger a écrit :
> 
> Martin Husemann <martin@duskware.de> writes:
> > On Sun, Mar 09, 2003 at 05:01:10AM -0600, Peter Seebach wrote:
> > > Unfortunately, the 4521 only has one physical serial port.
> > 
> > I didn't try but shouldn't it be a no brainer to add a new /dev/null
> > cnattach to sys/arch/x86/x86/consinit.c, enabled by a config option?
> 
> Not even any need to do that -- we should simply create an option that
> replaces the consinit() in init_main.c:main() with a call that
> initializes the function pointers to dummys. Then it will be MI and
> all ports can have the functionality.

Well, in theory. There's a call to consinit() in i386/machdep.c:init386(),
which happens before the one in main().

> Shouldn't take more than ten minutes to write.

Maybe other ports do the same, so it will take a little more. Say 15
minutes.

-- 
Quentin Garnier - cube@cubidou.net
"Feels like I'm fiddling while Rome is burning down.
Should I lay my fiddle down and take a rifle from the ground ?"
Leigh Nash/Sixpence None The Richer, Paralyzed, Divine Discontents, 2002.