From owner-freebsd-hackers  Wed Apr  3 10:09:56 1996
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.3/8.7.3) id KAA11913
          for hackers-outgoing; Wed, 3 Apr 1996 10:09:56 -0800 (PST)
Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211])
          by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id KAA11887
          Wed, 3 Apr 1996 10:09:50 -0800 (PST)
Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id LAA19519; Wed, 3 Apr 1996 11:02:50 -0700
From: Terry Lambert <terry@lambert.org>
Message-Id: <199604031802.LAA19519@phaeton.artisoft.com>
Subject: Re: Hacked kernel with option to disable "green" mode
To: Brett_Glass@ccgate.infoworld.com (Brett Glass)
Date: Wed, 3 Apr 1996 11:02:50 -0700 (MST)
Cc: terry@lambert.org, bde@zeta.org.au, freebsd-hackers@freebsd.org,
        freebsd-hardware@freebsd.org
In-Reply-To: <9603038285.AA828546177@ccgate.infoworld.com> from "Brett Glass" at Apr 2, 96 03:43:16 pm
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-hackers@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

> > Actually "fixing" the problem requires making the driver aware of drive
> > spin downs and making it deal with it transparently instead of screwing
> > up by passing an error to higher level code.
> 
> This is one way to solve the problem. However, disabling spindowns is also
> a viable solution. Why not offer both?

Uh... because it's impossible to offer the spindown disabling safely
in a drive/controller independent fashion?

The 0xFB for the Seagate is not the same as is used by all other "green"
hard disks, if they even support it in the first place...

> > Disabling the spindown is not what I'd call "making the code 'green
> > aware'"...
> 
> No, but it is both a performance enhancement and a way to ensure that
> things work.

That's the problem... it can't ensure this because you can't send the
command reliably.

But you can reliably recover from a spin-down, if your disk driver
has been written correctly.

So the generic soloution must be to recover from spindown, not to
disable it.

> That's why I'd propose using config flags -- so that the system will be
> able to perform adequately during installation and immediately thereafter
> -- AND a utility that could change the setting later on. Both could be
> based on some of the kernel hacking I've already done here.

This works for Seagates for people who have multiple systems so that
they can build kernels.

This *is* better than the situation before, but it effectively leaves
a lot of wires sticking out of the case.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.