Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 13 Oct 2002 15:56:53 -0400
From:      Donn Miller <dmmiller@cvzoom.net>
To:        Ian Dowse <iedowse@maths.tcd.ie>
Cc:        stable@freebsd.org
Subject:   Re: 4,7 Kernel panic
Message-ID:  <3DA9D005.1A0EBB3E@cvzoom.net>
References:  <aocgn0$1qf7$1@FreeBSD.csie.NCTU.edu.tw>

next in thread | previous in thread | raw e-mail | index | archive | help


Ian Dowse wrote:
> 
> In message <200210131410.30799.Unix@henoc.com>, Roger Savard writes:
> >Even with the latest :
> 
> >nexus_print_all_ressources(...
> 
> That is the other 4.7 problem that people have been reporting
> recently, but only on some hardware. You could try reverting John
> Baldwin's latest change to src/sys/i386/i386/nexus.c by grabbing
> version 1.26.2.6 of that file from
> 
>         http://www.freebsd.org/cgi/cvsweb.cgi/~checkout~/src/sys/i386/i386/nexus.c?rev=1.26.2.6

So the error is in the probe/attach routines.  Correct in that the fatal
trap only occurs on certain machines.  On my laptop, the kernel bombs
immediately after the EISA bus is probed.  For example:

eisa0:  <EISA bus> *BOMB*

Backtrace reveals:

nexus_print_all_resources(c0e62280,c0e4a680,c0e62280,c0e62280,0) at 
nexus_print_all_resources+0x14
nexus_print_child(c0e4a680,c0e62280,c0535f68,c01eb838,c0e4a680) at 
nexus_print_child+0x19
BUS_PRINT_CHILD(c0e4a680,c0e62280,c0e62280,c0e4a680,c0535f68) at 
BUS_PRINT_CHILD+0x31
device_print_child(c0e4a680,c0e62280) at device_print_child+0x20
device_probe_and_attach(c0e62280) at device_probe_and_attach+0x5a
nexus_attach(c0e4a680,c0535f68,c01ec003,c0e4a680,c0e4a680) at 
nexus_attach+0x4e
DEVICE_ATTACH(c0e4a680,c0e4a680,c0429cd0,53a000,1) at DEVICE_ATTACH+0x2e
device_probe_and_attach(c0e4a680) at device_probe_and_attach+0x63
root_bus_configure(c0b30800,c03e7f0c,0) at root_bus_configure+0x16
configure(0,532c00,53a000,0,c012bb70) at configure+0x33
mi_startup(0,0,0,0,0) at mi_startup+0x68
begin() at begin+0x47


When I did "ps" at the DDB prompt, it said swapper was the last procss
running.  The message said something about a virtual address of 0x0.  I
think it's odd that swapper is starting before the probing is finished
(during the probing of the EISA bus).  Sounds like something in the Oct
10 nexus.c commit is causing certain things to be started out of order
on certain HW.  In a nutshell, it looks as though swapper is being
started even before /etc/fstab is read.

I should point out that my laptop is an HP Pavilion N5440, a Pentium
III.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3DA9D005.1A0EBB3E>