SCIENTIFIC-LINUX-USERS Archives

November 2004

SCIENTIFIC-LINUX-USERS@LISTSERV.FNAL.GOV

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
"Alan J. Flavell" <[log in to unmask]>
Reply To:
Alan J. Flavell
Date:
Mon, 1 Nov 2004 18:49:54 +0000
Content-Type:
TEXT/PLAIN
Parts/Attachments:
TEXT/PLAIN (47 lines)
On Sun, 31 Oct 2004, John A. Goebel wrote:

> ++ 29/10/04 19:31 +0100 - <Alan J. Flavell>:
>
> > Perhaps I need a "magic sysreq".  These are areas I haven't really
> > explored before.  Haven't done kernel hacking / interrupt handling
> > stuff since IBM VM/370 or so  ;-)
>
> Definately you'll want magic sysrq keys. They are very handy when it
> comes to getting a fairly clean filesystem when you need to force a
> reboot (sysrq-s, sysrq-u, sysrq-b).

Oh yes, I've been there, back with RH7.2 and some problems with
a recalcitrant SCSI.  But I've never done anything more-exciting
with it than that before.

> Also, if you have a frozen system, you can get a task list dump.
> Redirecting that to serial sometimes helps narrow down the offending
> program/function.

I got that same answer from others.  A pity if there's no way to
collect that in some kind of console buffer and display it on the
affected machine, seeing that the problem I'm investigating is on a
laptop...

One type of laptop only has a serial port when coupled to its docking
station.  However, I do seem to be also getting these hangs on a
different type of laptop.  Oh, and I have to report that booting with
acpi=off didn't seem to make any substantive difference, so it seems
it might not have been relevant to the problem.

> Also memory dumps via sysrq is handy to see if memory exhaustion
> exists and Linux is not killing off the process in time. Stuff like
> that.

The status at the moment is that the wireless cards are working fine -
once they've been started up.  Just sometimes, while trying to start
them up, everything goes unresponsive, and then we get the morse-code
saying FATAL EXCEPTION.

I'm still worried that there seem to be "unseen forces" which are
moving the scenery around when the card is detected, in addition to
the overt commands (modprobe ndiswrapper, ifconfig eth1 etc.) that I'm
issuing.  I wish someone who understood these aspects better would
contact me off the list and hopefully optimise the effort I'm putting
into it, then I'd be happy to share any recipes developed.

ATOM RSS1 RSS2