Dead KDB console
From: Jas Cluff (jascluff_at_cruzio.com)
Date: 02/18/05
- Next message: vicks: "writing to register from user space -> process killed"
- Previous message: Måns Rullgård: "Re: Memory alignment under Linux"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Fri, 18 Feb 2005 12:37:56 -0800
Greetings,
(Hope this isn't a duplicate msg - my mailer blew up on the 1st tx.)
I've just brought up a 2.6.9 kernel with the 4.4 kdb patch from SGI.
Looks great up through hitting the Pause key. I get a KDB> prompt,
but keystrokes are not detected, so it's reset time.
I'm familiar w/ kdb on SVR4; and I've also used kgdb/linux, but this
is my first kdb-on-Linux installation, so pilot error is high on my
list of suspects.
I've read SGI FAQs, where there's a posting that acknowledges such a
bug, with a workaround, namely restart /etc/rc.d/init.d/gpm. The FAQ
supposes that the bug was fixed in v0.6. I've tried the workaround to
no avail, and I would also hope that the bug was indeed fixed way
before v4.4.
I've also tried a boot arg of console=tty0, in partial imitation of
the instructions for using both serial- and console-console. Noting
that the tty command from vt1 yields "/dev/tty1", I've also tried
console=tty1. Since the name field of struct console vt_console_driver
is "tty", I've tried console=tty too.
Suggestions/solutions??? Thanks...
ciao - jas
- Next message: vicks: "writing to register from user space -> process killed"
- Previous message: Måns Rullgård: "Re: Memory alignment under Linux"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]