Subject: flaky 'df' output
To: NetBSD/mac68k Mailing List <port-mac68k@netbsd.org>
From: Ryan Ordway <rimsky@teleport.com>
List: port-mac68k
Date: 09/07/1998 03:46:51
disks:
Filesystem  1K-blocks     Used    Avail Capacity  Mounted on
/dev/sd1a      837450   509451   244254    67%    /
/dev/sd1g     1155239   963371    76344    92%    /home
/dev/sd0g      495599        9   446030     0%    /home2
/dev/sd0a      743390   495600   173451    74%    /httpd
/dev/sd0b     1000563   743587   156919    82%    /ftp

	Now, sd1a and sd0g are actually correct. sd1 is an external 1.22
GB SCSI disk on my Quadra 630 at SCSI 1, sd0 is an external 1.09 GB SCSI
disk at SCSI 0. / is a ~800MB partition, and /home takes up the remainder 
of the 1.22 Gigs, minus 50MB for swap. /home2 is a ~500MB partition, and
/httpd and /ftp are approximately 250MB partitions on sd0.

	Obviously, if sd0 == 1.09 GB, then sd0g + sd0a + sd0b should be
1.09 GB, correct? Well, /home2 is right... /httpd it looks like is the sum
of sd0g and sd0a, and sd0b is the size of the whole disk. Similar
situation with sd1... sd1a being correct, and sd1g being the total disk
size.

	Of course, this is from a snapshot around 1.3E/F... I think
mid-may of this year. I'm wondering if this has been fixed in -current.

	Also, does anybody know if the "reaper" code has been added to the
main tree? I'm hoping to upgrade to -current soon, but that would be silly
if I don't have a bootable kernel ;-) [1]

	Thanks!

	Ryan

	[1] See my previous post about the pool_get panic.

--
HELO... my name is rewt... you have SIGKILLed my father... prepare to vi!