Message-ID: <45641422.5090601@goop.org>
Date: Wed, 22 Nov 2006 01:10:58 -0800
From: Jeremy Fitzhardinge <jeremy@goop.org>
User-Agent: Thunderbird 1.5.0.8 (X11/20061107)
MIME-Version: 1.0
To: Avi Kivity <avi@qumranet.com>
CC: "H. Peter Anvin" <hpa@zytor.com>, Arnd Bergmann <arnd@arndb.de>,
       kvm-devel@lists.sourceforge.net, akpm@osdl.org,
       linux-kernel@vger.kernel.org
Subject: Re: [kvm-devel] [PATCH] KVM: Avoid using vmx instruction directly
References: <4563667B.2060209@goop.org> <4563F158.3060209@qumranet.com>
In-Reply-To: <4563F158.3060209@qumranet.com>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 863
Lines: 23

Avi Kivity wrote:
> Very interesting.
>
> Will it work on load/store architectures?  Since all memory access is
> through a register, won't the constraint generate a useless register
> load (and a use of the variable)?

Don't know; interesting question.  It might be worth lobbying the gcc
folks for an asm() constraint which means "pretend this is being
read/written, but don't generate any code, and raise an error if the asm
actually tries to use it".  Or perhaps there's some way to do that already.

On the other hand, load/store archs tend to have lots of registers
anyway, so maybe it isn't a big deal.

    J


-
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/