From: "David Schwartz" <davids@webmaster.com>
To: "Linux-Kernel@Vger. Kernel. Org" <linux-kernel@vger.kernel.org>
Subject: RE: [patch 2.6.19-rc6] Stop gcc 4.1.0 optimizing wait_hpet_tick away
Date: Wed, 29 Nov 2006 17:04:28 -0800
Message-ID: <MDEHLPKNGKAHNMBLJOLKOEGKAAAC.davids@webmaster.com>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="us-ascii"
Content-Transfer-Encoding: 7bit
Importance: Normal
In-Reply-To: <20061128.200453.104036587.davem@davemloft.net>
Reply-To: davids@webmaster.com
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1341
Lines: 32


Ask yourself this question: Can an assignment to a non-volatile variable be
optimized out? Then ask yourself this question: Does casting away volatile
make it not volatile any more?

> The volatile'ness does not simply disappear the moment you
> assign the result to some local variable which is not volatile.

Yes, it does. That's what a cast does, it tells the compiler to, in all
respects, pretend that a variable is of a different type than it 'actually
is', such that it actually isn't anymore.

> Half of our drivers would break if this were true.

On the contrary, they'd break if it was true. If casting away volatile
didn't make it go away, then casting in volatile wouldn't have to make it
appear. A cast causes the compiler to act as if a variable really was the
type you cast it to. If you cast volatile away, that has the reverse of the
same affect casting to volatile has.

The 'readl' function should actually assign the value to a volatile
variable. Assignments to volatiles cannot be cast away, but casts can and
assignments to non-volatile variables can be optimized out.

DS


-
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/