From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1LWrn2-0001Ui-IB for qemu-devel@nongnu.org; Tue, 10 Feb 2009 07:32:36 -0500 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1LWrn0-0001UV-3b for qemu-devel@nongnu.org; Tue, 10 Feb 2009 07:32:35 -0500 Received: from [199.232.76.173] (port=55108 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1LWrmz-0001US-WE for qemu-devel@nongnu.org; Tue, 10 Feb 2009 07:32:34 -0500 Received: from mx20.gnu.org ([199.232.41.8]:31266) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1LWrmz-0001wr-Nh for qemu-devel@nongnu.org; Tue, 10 Feb 2009 07:32:33 -0500 Received: from mail.codesourcery.com ([65.74.133.4]) by mx20.gnu.org with esmtp (Exim 4.60) (envelope-from ) id 1LWrmy-0003hU-SA for qemu-devel@nongnu.org; Tue, 10 Feb 2009 07:32:33 -0500 From: Paul Brook Subject: Re: [Qemu-devel] [patch 2/2] qemu: PCI device, disk and host network hot-add / hot-remove Date: Tue, 10 Feb 2009 12:32:29 +0000 References: <20090206174406.589893548@amt.cnet> <20090209224125.GA11520@amt.cnet> <20090210122037.GG31871@redhat.com> In-Reply-To: <20090210122037.GG31871@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Message-Id: <200902101232.30279.paul@codesourcery.com> Reply-To: qemu-devel@nongnu.org List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org, "Daniel P. Berrange" Cc: Marcelo Tosatti , Markus Armbruster , Avi Kivity > Then info pci shows this: > > =A0 ...snip... > =A0 Bus =A00, device =A0 3, function 0: > =A0 =A0 Ethernet controller: PCI device 10ec:8029 > =A0 =A0 =A0 IRQ 11. > =A0 =A0 =A0 BAR0: I/O at 0xc100 [0xc1ff]. Sounds like the real problem is that info pci and info net don't contain=20 enough information. > Yes, if the device already had a unqiue identifier, it some ways it > would be easier to just be able todo 'nic_del 01:02:03:04:05:02' > Or for a SCSI disk 'drive_del bus=3Dscsi,index=3D3'. QEMU ought to have > enough info internally to be able to map this to the PCI address it > needs to remove. It may be reasonably to assume a 1:1 mapping between network interfaces and= =20 devices (for current devices at least). However the same is not true for=20 disks. Adding/removing host adapters, and adding/removing disks should be=20 separate operaions. The same strategy could also be used for network device= s=20 (you create a PCI NIC, then you connect that NIC to a vlan). We're probably= =20 going to want this separation anyway for machine config files. Paul