Re: elevator algorithm bug in ll_rw_blk.c

Matthias Urlichs (smurf@noris.de)
24 Nov 1998 15:07:51 +0100

Myreen Johan <Johan.Myreen@setec.fi> writes:
>
> I'd prefer letting the drive be our friend, and throw requests
> at it in any order. Let the drive sort it all out, instead
> of trying to beat it at its own game.
>
That works only if the request queue in the drive is big enough.

It isn't. In fact, the current queue in the kernel is too short.

You also assume that the drive can store multiple requests and reorder them
in the first place. That only works with reasonably-new SCSI disks which
support tagged queuing (and SCSI host adapters which do).
It does NOT work with IDE anything.

-- 
Matthias Urlichs  |  noris network GmbH   |   smurf@noris.de  |  ICQ: 20193661
The quote was selected randomly. Really.    |      http://www.noris.de/~smurf/

-- 
Any coward can fight a battle when he is sure of winning; but give me
the man who has pluck to fight when he's sure of losing.  That's my way,
sir; and there are many victories worse than a defeat.
                                -- George Eliot


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/