Subject: Re: fsck hangs up on error (is it normal ?)
To: Vincent <10.50@free.fr>
From: Manuel Bouyer <bouyer@antioche.eu.org>
List: tech-kern
Date: 09/30/2004 20:16:09
On Wed, Sep 29, 2004 at 12:28:50PM +0200, Vincent wrote:
> Hello,
> a friend of mine gave me a defectuous hard drive just to see what I 
> could make of it under Unix. The nasty thing is that I suspect the 
> controler, not the disk, to be damaged. Anyhow.
> 
> I make a disklabel+newfs okay, try to copy a large file to pin the 
> problem, and the cp (or the fsck I did after) hangs up on :
> 
> wd1: transfer error, downgrading to Ultra-DMA mode 1
> wd1(piixide0:1:0): using PIO mode 4, Ultra-DMA mode 1 (using DMA data 
> transfers)
> wd1e: error reading fsbn 756960 of 756960-757087 (wd1 bn 757023; cn 751 
> tn 0 sn 15), retrying
> wd1: (uncorrectable data error)
> 
> Thereafter nothing happens, fsck_ffs seems to be blocked on physio, 
> probably in the driver.
> 
> Is that a normal behaviour ?

Clearly not.
Can you see on what the kernel thread matching piixide0:1 (atabus1 I guess,
unless there are others IDE controllers in the box) is blocked ?

-- 
Manuel Bouyer <bouyer@antioche.eu.org>
     NetBSD: 26 ans d'experience feront toujours la difference
--