Date: Wed, 21 Dec 2005 22:21:31 +0000
From: Russell King <rmk+lkml@arm.linux.org.uk>
To: Adrian Bunk <bunk@stusta.de>
Cc: Andrew Morton <akpm@osdl.org>, torvalds@osdl.org,
       linux-kernel@vger.kernel.org, uclinux-v850@lsi.nec.co.jp
Subject: Re: [RFC: 2.6 patch] include/linux/irq.h: #include <linux/smp.h>
Message-ID: <20051221222131.GL1736@flint.arm.linux.org.uk>
Mail-Followup-To: Adrian Bunk <bunk@stusta.de>,
	Andrew Morton <akpm@osdl.org>, torvalds@osdl.org,
	linux-kernel@vger.kernel.org, uclinux-v850@lsi.nec.co.jp
References: <20051221012750.GD5359@stusta.de> <20051221024133.93b75576.akpm@osdl.org> <20051221110421.GA26630@flint.arm.linux.org.uk> <20051221213321.GC3888@stusta.de> <20051221214806.GJ1736@flint.arm.linux.org.uk> <20051221221114.GA3917@stusta.de>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20051221221114.GA3917@stusta.de>
User-Agent: Mutt/1.4.1i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1907
Lines: 50

On Wed, Dec 21, 2005 at 11:11:14PM +0100, Adrian Bunk wrote:
> On Wed, Dec 21, 2005 at 09:48:06PM +0000, Russell King wrote:
> > On Wed, Dec 21, 2005 at 10:33:21PM +0100, Adrian Bunk wrote:
> > > On Wed, Dec 21, 2005 at 11:04:22AM +0000, Russell King wrote:
> > > > On Wed, Dec 21, 2005 at 02:41:33AM -0800, Andrew Morton wrote:
> > > > > Yes, it's basically always wrong to include asm/foo.h when linux/foo.h
> > > > > exists. 
> > > > 
> > > > There's always an exception to every rule.  linux/irq.h is that
> > > > exception for the above rule.
> > > 
> > > Why?
> > 
> > /*
> >  * Please do not include this file in generic code.  There is currently
> >  * no requirement for any architecture to implement anything held
> >  * within this file.
> >  *
> >  * Thanks. --rmk
> >  */
> > 
> > Using linux/irq.h instead of asm/irq.h _breaks_ architectures
> > which do not use the generic irq code.
> > 
> > Basically, linux/irq.h should have been called asm-generic/irq.h.
> 
> I'm not getting your point.

The point is _exactly_ as the above quotation between Andrew Morton
and myself.  I'm sure it's not me being thick because it's absolutely
damned obvious from the above.

Andrew said: "Yes, it's basically always wrong to include asm/foo.h
when linux/foo.h exists."

That statement is a rule.  I assert that this is an incorrect statement
and I assert that there is a proven case where this statement is incorrect.

Hence, to avoid people reading Andrew's misleading statement, I followed
up on precisely _that_ point and _that_ point alone.

-- 
Russell King
 Linux kernel    2.6 ARM Linux   - http://www.arm.linux.org.uk/
 maintainer of:  2.6 Serial core
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/