From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1M9JoH-0007uU-5Z for qemu-devel@nongnu.org; Wed, 27 May 2009 10:08:49 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1M9JoC-0007u2-K6 for qemu-devel@nongnu.org; Wed, 27 May 2009 10:08:48 -0400 Received: from [199.232.76.173] (port=41045 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1M9JoC-0007tz-GR for qemu-devel@nongnu.org; Wed, 27 May 2009 10:08:44 -0400 Received: from e7.ny.us.ibm.com ([32.97.182.137]:33916) by monty-python.gnu.org with esmtps (TLS-1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1M9JoC-0007S2-2X for qemu-devel@nongnu.org; Wed, 27 May 2009 10:08:44 -0400 Received: from d01relay04.pok.ibm.com (d01relay04.pok.ibm.com [9.56.227.236]) by e7.ny.us.ibm.com (8.13.1/8.13.1) with ESMTP id n4RDv8Iq015007 for ; Wed, 27 May 2009 09:57:08 -0400 Received: from d01av01.pok.ibm.com (d01av01.pok.ibm.com [9.56.224.215]) by d01relay04.pok.ibm.com (8.13.8/8.13.8/NCO v9.2) with ESMTP id n4RE8gP2253784 for ; Wed, 27 May 2009 10:08:42 -0400 Received: from d01av01.pok.ibm.com (loopback [127.0.0.1]) by d01av01.pok.ibm.com (8.12.11.20060308/8.13.3) with ESMTP id n4RE8geZ026681 for ; Wed, 27 May 2009 10:08:42 -0400 Message-ID: <4A1D4965.2020309@us.ibm.com> Date: Wed, 27 May 2009 09:08:37 -0500 From: Anthony Liguori MIME-Version: 1.0 Subject: Re: [Qemu-devel] [PATCH 0/3] Object-based physical memory management References: <1243157375-14329-1-git-send-email-avi@redhat.com> In-Reply-To: <1243157375-14329-1-git-send-email-avi@redhat.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Avi Kivity Cc: qemu-devel@nongnu.org Avi Kivity wrote: > Currently, qemu manages physical memory at a page granularity (and with > subpage, at a byte granularity). This is expensive in memory and runtime > (requiring 8 bytes per page, and likely a cache miss on access), and is > difficult to manage: devices must remember their ranges so they can free > them, and the memory system has no insight into how memory is laid out. > > This patchset introduces object-based memory management. The first patch > adds objects and functions for memory regions; the second patch wraps this > for PCI; and the third converts RTL8139 as an example. > I like the basic notion here. Instead of introducing the API and converting one user, I'd rather do a flag day and convert everything. Once we agree on the API of course. Regards, Anthony Liguori