Subject: Re: fsck and others die with Floating Point exception
To: None <>
From: Ken Nakata <>
List: macbsd-general
Date: 03/09/1995 14:45:27
   Date: Thu, 9 Mar 1995 21:02:55 +0900
   From: (Yoshihisa Sugimoto)
   Cc: macbsd-general@NetBSD.ORG

   At  1:06 PM 95.3.8 -0500, Ken Nakata wrote:
   >   # ./fsck -d
   >   ** /dev/rsd1a (NO WRITE)
   >   &sblock 2e000, &altsblock 31000, sblock.fs_sbsize, 800
   >Are these block addresses same to those MacOS side MKFS reported?  My
   >Quantum 540M drive looks like having different geometries to MKFS and
   >the kernel, and that *might* make the kernel look other places for
   >alternate superblocks than where MKFS placed.

   I didn't record the block addresses MKFS reported. So I cannot compare

I think you can run MKFS without really formatting the partition and
see what it says, can't you?

   In addition, I found more odd things.
   The record of fsck I reported was made by "# ./fsck -d > fsck_result".
   When I command "# ./fsck -d | more ", the result is _different_ from
   that with the redirection.
   # ./fsck -d | more
   ** /dev/rsd1a (NO WRITE)
   &sblock 2d000, &altsblock 30000, sblock.fs_sbsize, 800

   The addresses of superblock and alternate are 0x1000 smaller than
   those by redirection. The data are identical except for them. Why?

I have no idea why it is so, but have you tried to mkfs from MacBSD?
I mkfs'ed my Quantum drive with BSD mkfs (not MacOS MKFS) because the
geometry MacBSD reported looked more like a real thing than the one
MacOS MKFS reported.  I can't mount that partition from the Installer
(due to a non-default mkfs parameter set I gave), but otherwise I have
no problem.

Hope this helps,