Re: badblocks blocksize setting

[Oleg Drokin <green@namesys.com>]

Hello!

On Sat, Mar 15, 2003 at 12:38:23PM +0100, Marius Reiner wrote:

> > Can you please show the command line and the resulting output with
> badblocks?
> > Also what is in dmesg output after such a run?
> # badblocks /dev/hdc8 -b 4096
> # dmesg -c
> invalidate: busy buffer
> # badblocks /dev/hdc8
> 979932
> # dmesg
> invalidate: busy buffer
> attempt to access beyond end of device
> 16:08: rw=0, want=979936, limit=979933
> attempt to access beyond end of device
> 16:08: rw=0, want=979936, limit=979933

Hm, is 979936 looks like size of your device in kilobytes (+-3)?
If so, this is known kernel bug/problem, where it cannot access parts
of device at the end sometimes. This is somehow related to setblocksize.
This can be ignored, then.

> However, what I'm missing in the FAQ, is a short note about _why_ one should
> use the -b option, so no worries arise when getting bad blocks without it.

-b is needed so block output by bad blocks have the same bad blocks as
in reiserfs. (4096 bytes in size, without -b, badblocks
defaults to 1024 bytes blocks, so you need to divide its output by 4).

Bye,
    Oleg