From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:55062) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XZmit-0006Ze-Jm for qemu-devel@nongnu.org; Thu, 02 Oct 2014 16:11:41 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1XZmin-0002Ap-KG for qemu-devel@nongnu.org; Thu, 02 Oct 2014 16:11:35 -0400 Received: from mail-wg0-x231.google.com ([2a00:1450:400c:c00::231]:49858) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XZmin-0002Al-DN for qemu-devel@nongnu.org; Thu, 02 Oct 2014 16:11:29 -0400 Received: by mail-wg0-f49.google.com with SMTP id x12so4233039wgg.32 for ; Thu, 02 Oct 2014 13:11:28 -0700 (PDT) From: =?UTF-8?B?0JLQsNGB0LjQuyDQoNCw0L3Qs9C10LvQvtCy?= References: <000101cfde75$27a55400$76effc00$@gmail.com> In-Reply-To: Date: Thu, 2 Oct 2014 23:10:53 +0300 Message-ID: <000201cfde7d$01d04b60$0570e220$@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Content-Language: bg Subject: Re: [Qemu-devel] WiFi inside VMs? List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: 'Nikolay Nikolaev' , 'qemu-devel' Hi. A USB/PCI device that is passed through to the VM would get a real WiFi = inside a VM, but doing THAT would place the VM in a *real* wireless = network. The problem is that's not the goal. The goal is to introduce a WiFi into a *virtual* wireless network, i.e. = a WiFi network between this VM and other VMs that have virtual WiFi = adapters. The network itself would be managed outside of Qemu (by GNS3), = but Qemu needs to provide support for a virtual device that would = send/receive data (to/from GNS3, which would then forward it over the = UDP tunnel to the appropriate peers). To quote the documentation: ```` There are two parts to networking within QEMU:=20 ## the virtual network device that is provided to the guest (e.g. a PCI = network card).=20 ## the network backend that interacts with the emulated NIC (e.g. puts = packets onto the host's network). ```` When using USB/PCI passthrough, you're using the real device as BOTH the = backend and the device being provided to the guest. But the goal is for the UDP tunnel backend to be used, so that frames = can be "trapped" in the virtual network. Currently, only Ethernet cards can use a UDP tunnel. What I'm asking for is for at least one model of WiFi network card to be = provided as a virtual device, and be able to use the UDP tunnel backend, = even if it can't use any other backend. -----Original Message----- From: Nikolay Nikolaev [mailto:n.nikolaev@virtualopensystems.com]=20 Sent: Thursday, October 2, 2014 10:42 PM To: =D0=92=D0=B0=D1=81=D0=B8=D0=BB = =D0=A0=D0=B0=D0=BD=D0=B3=D0=B5=D0=BB=D0=BE=D0=B2 Cc: qemu-devel Subject: Re: [Qemu-devel] WiFi inside VMs? Hello, On Thu, Oct 2, 2014 at 10:14 PM, =D0=92=D0=B0=D1=81=D0=B8=D0=BB = =D0=A0=D0=B0=D0=BD=D0=B3=D0=B5=D0=BB=D0=BE=D0=B2 = wrote: > Hello. > > I'd like to ask if it would be possible to add support for at least=20 > one model of WiFi device on guest machines? USB and/or PCI passthorugh might help here, or am I getting something = wrong? > > > I realize the feature request has been discussed previously (here: > http://lists.nongnu.org/archive/html/qemu-devel/2007-03/msg00575.html=20 > ), but my reason for wanting it is slightly different than the reason = back then... > > There's this program for simulating networks you've definitely heard=20 > of (I mean... It's been mentioned on this list...) - GNS3. It=20 > currently can't emulate Wireless networks, exactly because Qemu (nor=20 > any other emulator for that matter) supports WiFi devices inside VMs. > > The old discussion above points out how it would be problematic to=20 > bridge a real WiFi adapter to the virtual one, but as far as GNS3 is=20 > concerned, that's not really a problem, since the backend in its case=20 > needs to be a UDP tunnel. As long as the WiFi frames can be=20 > encapsulated into a UDP tunnel, > GNS3 could theoretically provide the connection between two WiFi=20 > devices inside a simulated topology. > > > Even if such a feature's usefulness is only limited to GNS3, that's=20 > still IMHO a great deal of usefulness, and therefore it should be a=20 > worthy goal to pursue. > > So... Pleeeease? Pretty please? :-) > > Thanks, > Regards, > Vasil Rangelov > > regards, Nikolay Nikolaev Virtual Open Systems