Re: Linux 2.4.19 patch for Suse compatibility

[Andi Kleen <ak@suse.de>]

On Tue, Nov 19, 2002 at 11:32:45AM -0200, Rik van Riel wrote:
> On 19 Nov 2002, Andi Kleen wrote:
> > Alan Cox <alan@lxorguk.ukuu.org.uk> writes:
> 
> > > This was discussed on the kernel list about four to six weeks ago and
> > > rejected then as well. See the previous discussion
> >
> > Actually I don't remember it being rejected, just the discussion dropped
> > off and there was no suggestion on how to solve the problem this ioctl
> > solves in a better way.
> 
> So, what problem does it try to solve ?

The output of all the startup scripts is logged by a special daemon to 
a log file (blogd). It does this by redirecting /dev/console.

For it to be able to still output something to the screen and later 
resetting the console it needs to know the output device that
/dev/console maps to. The ioctl returns it. 

There is actually a fallback path that tries to discover the underlying
device when the ioctl is not there, but it is far too ugly to describe
and also has some problems.

-Andi