mgetty hangs linux system

• Messages sorted by:[ date ][ thread ][ subject ][ author ]

On Tue, Jun 30, 1998 at 09:27:23PM +0200, Gert Doering wrote:
> Hi,
> 
> I haven't heard about such a problem before, and I can assure you that
> it's not mgetty's fault -- and even if it were, it's the system's fault if
> a "bad" process can make the system hang.
> 
> I have quoted your mail completely, and set a CC: to the mgetty mailing
> list.  Maybe someone out there has any tips on this.
> 
> What I find "suspcious" is that ttyS2 and ttyS3 seem to share the IRQ
> number with the ttyS0 and ttyS1 ports.  Unless you have special hardware
> that can do this (PCI cards, and multiport boards that can share IRQs
> between ports ON THE SAME CARD), this won't work.  It might seem to, but
> bad things will happen sooner or later.

Check your setserial calls, too.  Look someplace in /etc/rc.d or its
subdirectories for a bash script that calls setserial.

If the cards are set one way, but setserial has different 
parameters, expect bad things to happen.

The default script in a couple of distributions uses autodetect; I've
found that sometimes this falsely detects shared IRQs.

If that is so, tweak the file so it doesn't use autodetect.

And as Gert said, if your cards are sharing IRQs and if they
aren't intended for this use, also expect bad things.

[ ... ]
> > (root@boulder.denver.com.au) (gcc version 2.7.2.1) #1 Thu Oct 23 15:39:59
> > WST 1997
> > Jun 30 11:40:32 boulder kernel: Serial driver version 4.13 with no serial
> > options enabled
> > Jun 30 11:40:32 boulder kernel: tty00 at 0x03f8 (irq = 4) is a 16550A
> > Jun 30 11:40:32 boulder kernel: tty01 at 0x02f8 (irq = 3) is a 16550A
> > Jun 30 11:40:32 boulder kernel: tty02 at 0x03e8 (irq = 4) is a 16550A
> > Jun 30 11:40:32 boulder kernel: tty03 at 0x02e8 (irq = 3) is a 16550A
[ ... ]

-- 
Dan Wilder		                <dan@gasboy.com>