Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 4 Dec 1999 21:14:50 -0500 (EST)
From:      "Mark W. Krentel" <krentel@dreamscape.com>
To:        freebsd-emulation@FreeBSD.ORG
Subject:   Re: running linux binaries from ext2fs partition
Message-ID:  <199912050214.VAA04654@dreamscape.com>

next in thread | raw e-mail | index | archive | help
A couple days ago, I reported getting panics when I tried running
linux programs from their ext2fs partition.

> My machine dual boots between Freebsd 3.3-stable (as of Nov 7) and Red
> Hat 6.0.  I have the linux_base-6.0 port installed and the linux.ko
> module installed.  I can run linux's ls by copying it onto a freebsd
> partition, but it panicked on ext2fs.

I've been trying to investigate this further, but unfortunately, I've
been unable to get a kernel dump (see below).  I've produced several
panics, but when it crashes, I don't get the usual "Auto reboot in 15
seconds" message.  Instead, I get several panic messages which fly off
the console.  Then it freezes and I can only read the last message.
The instruction pointers are mostly very low, 0x8:350, or very high,
0x8:c45b1153 (with a stack pointer 0x10:0xc45b1234).  I'm guessing
that it panics, tries to dump core and panics again with a less useful
message.

One time it panicked just once, printed just one message and gave the
the "Auto reboot in 15 seconds" message.  This was my first test,
before I had dumpon turned on.  It had a more useful instruction
pointer: 0x8:0xc01f1916.  Looking through the kernel name list:

   % nm kernel | sort
   c0116c40 T btext
   c0116c96 t begin
   ...
   c01f1850 T generic_bzero
   c01f186c T i686_pagezero
   c01f18ac T fillw
   c01f18c0 T bcopyb
   c01f18ec T bcopy
   c01f18f4 T ovbcopy
   c01f18fc T generic_bcopy
   c01f1944 T memcpy
   c01f1968 T copyout
   c01f1970 T generic_copyout
   c01f19ae t done_copyout
   c01f19c0 t copyout_fault
   c01f19dc T copyin
   c01f19e4 T generic_copyin
   c01f1a30 t copyin_fault
   ...
   c02527ec B vga_sub_configure
   c02527f0 B swbuf
   c02527f4 A _end
   c02527f4 A end

So, maybe generic_bcopy?  I'm sorry I don't have more to report.
Anything else I should try?

Now, on to why I can't get a kernel dump.  I have dumpon turned on
in /etc/rc.conf.local:

   dumpdev="/dev/da0b"

And sysctl reports that it is turned on:

   % sysctl -a | grep -i dump
   kern.dumpdev: { major = 4, minor = 1 }
   kern.sugid_coredump: 0
   machdep.do_dump: 1

Memory is 64 Meg, swap is 160 Meg, minfree is 2048.  /var has 75 Meg
free, so that should be enough.  The swap partition, /dev/da0s2b, is
also used as MFS for /tmp.

But I'm not getting a kernel dump.  On reboot, savecore just reports
"no crash dump."  When it panics, the disk and scsi controller are
quiet, so it must not be writing out the dump.

Am I missing a step?  Or is the panic so bad that it can't dump core?
Any suggestions on how to get a dump, or what to try without one?

P.S. I didn't understand this comment in the man page for dumpon under
BUGS: "Dumpon currently allows only devices with minor number 1 to be
used as dump devices."  In /dev:

   brw-r-----  1 root  operator    4,   1 Nov  7 19:14 /dev/da0b
   brw-r-----  1 root  operator    4, 0x00030001 Jul  1 18:51 /dev/da0s2b

Does it matter if I set dumpdev to /dev/da0b or /dev/da0s2b?  Is minor
number 0x00030001 close enough to 1 for dumps?  I tried both and
neither worked.

--Mark Krentel


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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199912050214.VAA04654>