From: king@rtsg.mot.com (Steven King, Software Archaeologist) Subject: Re: HD-Problem Date: Mon, 15 Feb 1993 20:13:37 GMT
fbrueng@gwdu03.gwdg.de (Frank Bruenger) publicly declared:
>I'm working with linux 0.99p2 extended file system.
>There seemd to be a HD- problem. message is:
>HD: read_intr: error = 0x40
>HD: read_intr: status= 0x59
>hardisk I/O error
>dev 0342, sektor 311
>Kernel panic: unable to read i-node block
I see the same thing on my aging Seagate MFM drives. I *know* there
are bad sectors on these drives, but I don't think Linux ignores them
properly.
I used my machine's BIOS to do a bad sector scan and low-level format
of the drive. It reported bad sectors (no surprise) and I'm pretty
sure it marked them in whatever magical area is reserved for such
things. I then used fdisk to make the 40 Meg drive one large primary
partition, and 'mkfs -c' to make it a minix file system. As I
understand the '-c' option, it's supposed to honor the bad block list
on the drive. No luck, I still get the read_intr and write_intr errors
periodically.
This is using an SLS install, 0.99pl2 from the SLS distribution
immediately prior to the switch to gzip. Am I doing something wrong?
Is mkfs hosed?
--
=======================================================+=======================
Elephants don't step on you because they don't like | Steven King
squishy things between their toes. | Motorola Cellular
| king@rtsg.mot.com