comp.lang.ada
 help / color / mirror / Atom feed
From: "Brian Catlin" <brianc@sannas.org>
Subject: Re: AdaOS
Date: Sat, 17 Nov 2001 10:40:48 -0800
Date: 2001-11-17T18:40:41+00:00	[thread overview]
Message-ID: <9t6av9$601$1@slb3.atl.mindspring.net> (raw)
In-Reply-To: 3BF019F4.C9C460E5@mida.se

The HAL is essentially the driver for the motherboard, and controls the
interrupt controller, DMA controller, timer, etc.  The HAL used to contain the
bus drivers through NT4, but starting with Windows 2000, the bus drivers were
migrated out of the HAL and into standalone drivers, reducing the need for OEMs
to create their own HALs (most were created to deal with PCI controllers and
non-Intel core logic).  The kernel is processor architecture specific (there is
a different kernel for each supported architecture, x86, Alpha, PPC, MIPS,
Itanium), while the executive is the same across all platforms.

 -Brian

--
Brian Catlin, Sannas Consulting 310-798-8930
Windows NT/2000 Internals, WDM Device Driver Training & Consulting
See WWW.AZIUS.COM for courses and scheduling

"Mats Karlssohn" <mats@mida.se> wrote in message
news:3BF019F4.C9C460E5@mida.se...
> Ted Dennison wrote:
> >
> > In article <9scke6$12jb14$1@ID-25716.news.dfncis.de>, Nick Roberts says...
> > >
> > >Yes, but Phil, it's not the actual bits of software that are redundant --
> > >I've no doubt they indeed do useful things -- it's the term 'HAL' that's
> > >redundant. The hardware abstraction layer IS the device drivers, and always
> > >has been. Someone's just invented a fancy name, that's all!
> >
> > I think there's a bit more to it than that. I believe its purpose initially
was
> > to shelter the kernel from processor implemenation details, so that only the
HAL
> > would have to be rewritten to support a new processor. In that respect, you
can
> > kind of think of it as a microkernel.
>
> As usual you are quite right Ted. Please correct me if I'm workn, but I've
> been led to belive that in NT the main purpose of the HAL is to isolate the
> device drivers from the architectural differenses of of different processors
> and machines. Like the difference between the separate I/O addresspace of
> the x86 vs. {alpha,mips} memory mapped I/O, the idea is that the same device
> driver can be used for a card, it doesn't matter weather the card is in
> a AlphaStation or an Intel box. There is more to it though (as usual).
>
>
>
> --
> Mats Karlssohn, developer                         mailto:mats@mida.se
> Mida Systemutveckling AB                          http://www.mida.se
> Box 64, S-732 22 ARBOGA, SWEDEN
> Phone: +46-(0)589-89808   Fax: +46-(0)589-89809





  parent reply	other threads:[~2001-11-17 18:40 UTC|newest]

Thread overview: 48+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-10-25  9:10 AdaOS Franz Huber
2001-10-25 11:35 ` AdaOS Alfred Hilscher
2001-10-28  1:44 ` AdaOS Nick Roberts
2001-10-29 16:40 ` AdaOS Tony Gair
2001-10-29 16:50   ` AdaOS Preben Randhol
2001-10-29 17:02     ` AdaOS Franz Huber
2001-10-31  8:54       ` AdaOS Tony Gair
2001-11-04 17:32         ` AdaOS Nick Roberts
2001-11-05  0:59           ` AdaOS David Starner
2001-11-05 18:15             ` AdaOS Nick Roberts
2001-11-05 12:50           ` AdaOS Franz Huber
2001-11-05 18:11             ` AdaOS Nick Roberts
2001-11-05 23:12               ` AdaOS Marin David Condic
2001-11-06 15:38                 ` AdaOS Nick Roberts
2001-11-06 17:17                   ` AdaOS Darren New
2001-11-06 17:26                     ` AdaOS Nick Roberts
2001-11-06 20:37                       ` AdaOS Larry Kilgallen
2001-11-06 22:28                         ` AdaOS Nick Roberts
2001-11-06 22:47                       ` HAL (was AdaOS) Hans-Olof Danielsson
2001-11-07  7:22                         ` tmoran
2001-11-07 22:11                           ` Matthew Woodcraft
2001-11-07  2:28                   ` AdaOS Philip Cummins
2001-11-07 22:30                     ` AdaOS Nick Roberts
2001-11-08 18:49                       ` AdaOS Marin David Condic
2001-11-08 22:02                         ` AdaOS Nick Roberts
2001-11-09 16:06                       ` AdaOS Ted Dennison
2001-11-09 23:30                         ` AdaOS Nick Roberts
2001-11-12 18:50                         ` AdaOS Mats Karlssohn
2001-11-12 20:03                           ` AdaOS Ted Dennison
2001-11-12 20:29                             ` AdaOS Marin David Condic
2001-11-12 22:25                               ` AdaOS Ted Dennison
2001-11-17 18:40                           ` Brian Catlin [this message]
2001-11-17 22:49                             ` AdaOS Nick Roberts
2001-11-19 16:53                               ` AdaOS Marin David Condic
2001-11-19 17:17                               ` AdaOS Ted Dennison
2001-11-19 19:51                                 ` AdaOS Peter Hend�n
2001-11-19 22:41                                   ` AdaOS Ted Dennison
2001-11-08 12:34                     ` AdaOS Georg Bauhaus
2001-11-09 15:58                   ` AdaOS Ted Dennison
2001-11-09 18:56                     ` AdaOS Peter Hend�n
2001-11-10  2:44                     ` AdaOS Eric Merritt
  -- strict thread matches above, loose matches on Subject: below --
2002-06-10 21:39 AdaOS Peter I. Hansen
2002-06-10 22:49 ` AdaOS Preben Randhol
2002-06-10 23:35   ` AdaOS Gerhard Häring
2002-06-11  0:07     ` AdaOS Preben Randhol
2002-06-11 16:42     ` AdaOS Wes Groleau
2002-06-11 17:06       ` AdaOS Preben Randhol
2001-09-10 15:09 AdaOS Peter Hermann
replies disabled

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox