From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1M4u4e-0001VO-Uc for qemu-devel@nongnu.org; Fri, 15 May 2009 05:51:28 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1M4u4a-0001Tx-4v for qemu-devel@nongnu.org; Fri, 15 May 2009 05:51:28 -0400 Received: from [199.232.76.173] (port=46960 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1M4u4Z-0001Tl-SG for qemu-devel@nongnu.org; Fri, 15 May 2009 05:51:23 -0400 Received: from mx20.gnu.org ([199.232.41.8]:65042) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1M4u4Z-0005SG-H1 for qemu-devel@nongnu.org; Fri, 15 May 2009 05:51:23 -0400 Received: from mail.codesourcery.com ([65.74.133.4]) by mx20.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1M4u4Y-0003WT-5u for qemu-devel@nongnu.org; Fri, 15 May 2009 05:51:22 -0400 From: Paul Brook Subject: Re: [Qemu-devel] [PATCH] Revived GUEST_BASE support for usermode emulation targets [v4] Date: Fri, 15 May 2009 10:50:59 +0100 References: <1241028203-19687-1-git-send-email-riku.voipio@iki.fi> <200905150325.31918.paul@codesourcery.com> <4A0D2AA5.7000000@opensuse.org> In-Reply-To: <4A0D2AA5.7000000@opensuse.org> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200905151051.00783.paul@codesourcery.com> List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: Martin Mohring , mika.westerberg@iki.fi, Riku Voipio > > I'm pretty sure it's still broken for large offsets. > > > Is big offsets the intention of this patch? As far as I understood its > for small offsets (typical 64k - on debian, ubuntu, suse, fedora and all > the others - i checked it because it tested this), so address 0 will not > be used in typical linux configs today to grab null pointer exceptions. The intent of the patch is to move the guest virtual address space to a void conflicts with the host VM. For a 32-bit guest n a 64-bit host it makes a lot of sense to pick a completely unused 4G region (and probably even reserve it so the host doesn't start using it), which is almost certainly going to be at a high address. Paul