Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 4 Apr 1996 12:43:29 -0800 (PST)
From:      "Rodney W. Grimes" <rgrimes@GndRsh.aac.dev.com>
To:        nate@sri.MT.net (Nate Williams)
Cc:        root@deadline.snafu.de, jgreco@brasil.moneng.mei.com, current@FreeBSD.org
Subject:   Re: tty-level buffer overflows - what to do?
Message-ID:  <199604042043.MAA07188@GndRsh.aac.dev.com>
In-Reply-To: <199604041914.MAA17253@rocky.sri.MT.net> from Nate Williams at "Apr 4, 96 12:14:09 pm"

next in thread | previous in thread | raw e-mail | index | archive | help
> > But another problem is still going on on that machine. During boot there
> > are several occurences of "stray irq 7" messages until syslog says it would
> > not log them anymore. I have no idea where this stray irq's should happen.
> > Physically there is no adaptor card installed which could ever generate a
> > IRQ 7 ? Possible that this has to do with the other thing?
> 
> This is a pretty good indication that something is mis-confugred.  IRQ 7
> is the 'junk' interrupt, which means it gets all of the interrupts not
> otherwise assigned to a particular piece of hardware.  Something is
> generating interrupts on your system bogusly and you need to find out
> what.

This is not quite correct, IRQ 7 is signal when someone asserts an IRQ singal
then removes that signal _before_ the CPU runs the interrupt acknowledge
cycle to the 8259 PIC.

Devices that fully assert and hold it asserted do _not_ cause IRQ7, they
cause there respective interrupt.

Often the cause of stray IRQ7's is noisy or floating IRQ signals from boards
that are not recognized by FreeBSD.

In any case a system having this problem is going to be marginal at best,
you should be able to see the rate of these events with a vmstat -i,
watch the irq 7 counter... if you only get a burst of these during boot
then you are probably okay and it is caused by some strange piece of
hardware that is tickled into bad behavior by the invasive FreeBSD probe
code, one way to cut that down is to eliminate all devices from the kernel
that you do not really have.

-- 
Rod Grimes                                      rgrimes@gndrsh.aac.dev.com
Accurate Automation Company                 Reliable computers for FreeBSD



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199604042043.MAA07188>