[prev in list] [next in list] [prev in thread] [next in thread] 

List:       linux-kernel
Subject:    Re: 2.1.93..
From:       Cyrille Chepelov <chepelov () rip ! ens-cachan ! fr>
Date:       1998-04-18 23:13:22
[Download RAW message or body]

On Fri, 17 Apr 1998, Gerard Roudier wrote:

> If you are willing to be fine with the user's view of disks, you can 
> allow users to affect identifiers or names to disks and register this
> information somewhere.
> The right place for this information is, IMO, the disk itself.

This is what NT does, by the way : they put a 32-bit integer in the first
sector, next to the partition map (at offset 0x1b8, precisely). 
Unfortunately, LILO trashes it (that was quite a unilateral move from MS),
making the Disk Administrator complain that it cannot find its babies, and
asking for permission to put a "harmless" signature on the disks. If you
allow LILO to touch a disk which is a part of a stripe set (that happened
to me), you're going to have tough times... NT won't mount the partition
again.  I've read somewhere that NT5 was moving away from this system --
and it would be darn right : the standard for partition tables has been
broken. Perhaps it could be wiser to tag the *partitions* themselves (in
the superblock, probably).. 

 > 
> You will be fine with driver guys at the same time since they will not 
> have to worry about this issue anymore.
> 
> I am thinking that if it was possible to read some information from the 
> disks handled by BIOS using BIOS itself prior to attaching resources, we
> could implement smart things in this regard.
> 
> BTW, my knowledge of BIOS things is near zero, so the above is just 
> an idea and I donnot know how it can be implemented nor I'm sure 
> it is feasible.
> I donnot use M$ O/Ses or just very rarely in order not to be fired by my
> employer :) and never try to understand how those things are working
> during the short period of time they seem to be so :), but I am feeling
> that M$ O/Ses apply something similar.
> Do they ? 
> 
> 
> Gerard.
> 
> 
> 
> 
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.rutgers.edu
> 

------------------------------
Zog Zog



-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

[prev in list] [next in list] [prev in thread] [next in thread]