Baslinux mailing list

[sindi keesan <keesan AT sdf.lonestar.org>]

I was in the middle of compiling a kernel for sound and scsi (I thought I 
had used the BL2 kernel with scsi modules but it must have been bare.i 
that worked with scsi_mod and sd_mod and aic7xxx) when my computer started 
giving me pages of reset I/O error and hdb hd not ready for command status 
timeout and ide0: unexpected interrupt, status 0xfe, count=0 (up to 40 or 
so until I hit Ctrl-C).  I rebooted out of BL2 and did an e2fsck with BL3 
on /dev/hdb2 and it fixed errors.  This happened again in BL2, and then in 
BL3, and I could not reboot from BL3. (I managed to save .config to fd). 
Now I can't boot to DOS from HD and if i boot from FD freedos finds no hds 
nor does spinrite or syschk (which did find my two floppy drives) or CMOS 
auto hdd detection.

The HD power down is disabled.
CPU tempeartures 37C/98F, MB temperature 31/87.
CPU Fan Speed: xxxxRPM [Err]  - is this a problem?
VCORE Voltage: 2.2V
+3.3V Voltage 3.4-3.5V - is this a problem?
+12V  is 12.3
-12V is -11.7
+5 is 5.0-5.1

The problem actually started last night but was not severe and went away.

The CD-ROM drive was not found either, suggesting a controller problem.

The floppy controller works.  Is it likely I could fix the problem by 
disabling the onboard IDE controller and putting in a newish controller 
card (with support for disks over 504MB)?

Is there first some test I should be trying to run in linux?  I tried 
(before I gave up and powered off and on again to reboot) badblocks but it 
wants block_count and that is not found in the badblocks man page that I 
found online.  BL3 was not behaving well enough to read the included 
manpage for badblocks.  What is block_count? I could try testing the 
drives using floppy disk BL3.

Is there some way to test the controller itself rather than the hard 
drives?   I doubt they both went bad at once.  The linux filesystem got 
screwed up several times but that could be the controller.

I do not currently have a way to use a browser.  (A separate problem). I 
am dialed into a shell account and lynx is producing 7-bit garbage.  The 
free ISP is running at 0-500 bytes/sec late afternoons so I gave up on it.
I got as far as google's list of sites discussing block_count and this 
seems to be a feature of Redhat 4's badblocks.

This is my super-speedy computer (300MHz) but a friend just offered me his 
old 400MHz 'with lots of bays'.  I could probably just move over the 
drives but I don't want to give up on this one if it is fixable.  (It was 
given us to recycle but so were most of our other computers.)  It was 
working perfectly until last night.


keesan AT sdf.lonestar.org
SDF Public Access UNIX System - http://sdf.lonestar.org