Re: Consoles and d-i

2011-04-07 Thread Jurij Smakov
On Sun, Apr 03, 2011 at 03:05:38PM -0300, Otavio Salvador wrote: > On Sun, Apr 3, 2011 at 11:03, Jurij Smakov wrote: > > I have now come up with a set of changes which implements it, see the > > attached patch. I've tested it on my sparc box with a 2.6.38 netboot > > image, with both regular and s

Re: Consoles and d-i

2011-04-03 Thread Jurij Smakov
On Sun, Apr 03, 2011 at 03:05:38PM -0300, Otavio Salvador wrote: > On Sun, Apr 3, 2011 at 11:03, Jurij Smakov wrote: > > I have now come up with a set of changes which implements it, see the > > attached patch. I've tested it on my sparc box with a 2.6.38 netboot > > image, with both regular and s

Re: Consoles and d-i

2011-04-03 Thread Jurij Smakov
On Sun, Apr 03, 2011 at 04:34:05PM +0200, Julien Cristau wrote: > On Sun, Apr 3, 2011 at 15:03:32 +0100, Jurij Smakov wrote: [...] > > + * Licensed under GPLv2 > > + * > > + * Print out major:minor number of the real console device, > > + * using the TIOCGDEV ioctl (only works on kernels >= 2.6.38

Re: Consoles and d-i

2011-04-03 Thread Otavio Salvador
On Sun, Apr 3, 2011 at 11:03, Jurij Smakov wrote: > I have now come up with a set of changes which implements it, see the > attached patch. I've tested it on my sparc box with a 2.6.38 netboot > image, with both regular and serial console. Why do you want to provide compatibility with older kerne

Re: Consoles and d-i

2011-04-03 Thread Julien Cristau
On Sun, Apr 3, 2011 at 15:03:32 +0100, Jurij Smakov wrote: > >From d3f29283db9d58ad5b1702f66f9da15df6881526 Mon Sep 17 00:00:00 2001 > From: Jurij Smakov > Date: Sun, 27 Mar 2011 20:14:56 +0100 > Subject: [PATCH] Use TIOCGDEV ioctl to find real serial console device on > Linux. > > Kernels >=

Re: Consoles and d-i

2011-04-03 Thread Jurij Smakov
On Thu, Mar 24, 2011 at 11:56:10PM +, Jurij Smakov wrote: > Hello, > > I've noticed that 2.6.38 kernel supports TIOCGDEV ioctl, which can be > used to unambigously determine the real tty device corresponding to > /dev/console [0]. I believe that by using this we can significantly > simplify