From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43)
	id 1NezdE-0008Bm-P2
	for qemu-devel@nongnu.org; Tue, 09 Feb 2010 18:36:36 -0500
Received: from [199.232.76.173] (port=50350 helo=monty-python.gnu.org)
	by lists.gnu.org with esmtp (Exim 4.43) id 1NezdE-0008Bd-Ed
	for qemu-devel@nongnu.org; Tue, 09 Feb 2010 18:36:36 -0500
Received: from Debian-exim by monty-python.gnu.org with spam-scanned (Exim
	4.60) (envelope-from <av1474@comtv.ru>) id 1NezdD-0006kF-0N
	for qemu-devel@nongnu.org; Tue, 09 Feb 2010 18:36:36 -0500
Received: from fe02x03-cgp.akado.ru ([77.232.31.165]:59220 helo=akado.ru)
	by monty-python.gnu.org with esmtp (Exim 4.60)
	(envelope-from <av1474@comtv.ru>) id 1NezdC-0006k3-Ld
	for qemu-devel@nongnu.org; Tue, 09 Feb 2010 18:36:34 -0500
Date: Wed, 10 Feb 2010 02:36:33 +0300 (MSK)
From: malc <av1474@comtv.ru>
Subject: Re: [Qemu-devel] [PATCH 07/15] ac97: convert to new PCI API
In-Reply-To: <4B71EB73.608@codemonkey.ws>
Message-ID: <alpine.LNX.2.00.1002100231200.17472@linmac>
References: <1265752899-26980-1-git-send-email-aliguori@us.ibm.com>
	<1265752899-26980-8-git-send-email-aliguori@us.ibm.com>
	<alpine.LNX.2.00.1002100143480.14423@linmac>
	<4B71E72C.8050702@codemonkey.ws>
	<alpine.LNX.2.00.1002100205500.24185@linmac>
	<4B71EB73.608@codemonkey.ws>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
List-Id: qemu-devel.nongnu.org
List-Unsubscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.gnu.org/pipermail/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <http://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Anthony Liguori <anthony@codemonkey.ws>
Cc: Michael Tsirkin <mstirkin@redhat.com>, qemu-devel@nongnu.org, Alex Graf <agraf@suse.de>

On Tue, 9 Feb 2010, Anthony Liguori wrote:

> On 02/09/2010 05:06 PM, malc wrote:
> > 
> > > We already have this problem with the current interface.
> > >      
> > Uh, i've meant the registration of one function to rule them all, instead
> > of how it's done currently - separate accessors for b/w/l/whatever.
> >    
> 
> How does that make any difference?  Both the ioport and memory registrations
> interface take the same function pointer regardless of access size.
> 
> If you wanted to introduce a quad-word specific accessor, you would need to
> introduce a different registration mechanism or you would have to change the
> signature for all of the functions.

Let's see:

Currently we have this


readb(...):
  dostuff
  return stuff

readw(...):
  dostuff
  return stuff

You are replacing it with

read(size...):
  if (size == 1): do1
  elif (size == 2): do2
  else: # and here your code assumes that everything is handy dandy
        # and size is 4
    do4

The interface being implicit rather than explicit about the sizes
makes this possible, so i'm against it. The code was written the
way it was written for a purpose.

-- 
mailto:av1474@comtv.ru