On Sat, Oct 12, 2013 at 10:01:42AM -0400, Sasha Levin wrote: > On 10/12/2013 05:49 AM, Pekka Enberg wrote: > >Hi Peter, > > > >(Adding bunch of CCs.) > > > >On 10/12/13 11:05 AM, P?ter Szab? wrote: > >>First, thank you very much for writing lkvm, it's awesome, and very > >>easy to set up. > >> > >>If sending an e-mail to you is not the right way to report lkvm > >>issues, please tell me how I should do it. > > > >Sure, you can report problems to me and I'll do my best to > >try to get them sorted out. > > > >>I'm using afdf92030c7c43b0f9b32b7edbe07ac3b13780f1 from > >>git://github.com/penberg/linux-kvm.git and Linux kernel 3.2.51: > >> > >>config-3.2.51 (42 KB) > >>https://mega.co.nz/#!hgxB1TDJ!SdbX-jp_yr8E6EUJl7t7Tzrh1p4qKxkTHieoss8yu_Y > >>bzImage-3.2.51 (2.0 MB) > >>https://mega.co.nz/#!5lBVmDZL!WpPRWA7ZflevBIPPNGNM_FkkY-ErBNQMoEbw0XePi5I > >> > >>Config: > >> > >>$ grep -E 'VIRTIO|MMIO' .config > >>CONFIG_NET_9P_VIRTIO=y > >>CONFIG_VIRTIO_BLK=y > >>CONFIG_VIRTIO_NET=y > >>CONFIG_VIRTIO_CONSOLE=y > >>CONFIG_HW_RANDOM_VIRTIO=y > >>CONFIG_VIRTIO=y > >>CONFIG_VIRTIO_RING=y > >>CONFIG_VIRTIO_PCI=y > >>CONFIG_VIRTIO_BALLOON=y > >>CONFIG_VIRTIO_MMIO=y > >>CONFIG_HAVE_MMIOTRACE_SUPPORT=y > >> > >>I'm running `lkvm run --network mode=user,trans=mmio', and it fails to > >>create the eth0 device (`cat /proc/net/dev' doesn't show the device). > >>With `lkvm run --network mode=user' everything just works. > >> > >>I have also tried kernels 3.0.99, 3.4.65 an 3.10.15, but they seem to > >>fail the same way. > >> > >>Am I using it correctly? > > > >Asias, Sasha, is mmio transport supported by virtio networking? > > It should be. The ARM folks use it all the time (I suppose). > > Did you properly define the network device within the guest? MMIO > doesn't have automatic detection like PCI does. >
See these commits, we need a kernel parameter to tell guest the device address. commit 7dc9579466c55f45648e8c06321c6a21cd19fe8b Author: Asias He <asias.he...@gmail.com> Date: Sat Apr 7 19:51:21 2012 +0800 kvm tools: Introduce trans=mmio parameter for virtio-net To select virtio-mmio instead of virito-pci transport: lkvm run --network trans=mmio Signed-off-by: Asias He <asias.he...@gmail.com> Signed-off-by: Pekka Enberg <penb...@kernel.org> commit 5c301a39b05bd4c4693f2c9d854dc8ba6bbe821f Author: Asias He <asias.he...@gmail.com> Date: Sat Apr 7 19:44:12 2012 +0800 kvm tools: Add virtio-mmio support This patch is based on Sasha's 'kvm tools: Add support for virtio-mmio' patch. ioeventfds support is added which was missing in the previous one. VQ size/align is still not supported. It adds support for the new virtio-mmio transport layer added in 3.2-rc1. The purpose of this new layer is to allow virtio to work on systems which don't necessarily support PCI, such as embedded systems. To apply the patch on top of the KVM tools tree, you must first pull Linus' tree on top. Also, CONFIG_VIRTIO_MMIO=y should be set in the guest kernel. To easily test it it's recommended to apply Pawel Moll's patch named 'virtio-mmio: Devices parameter parsing' on top, and define the virtio-mmio device using kernel command line. LKVM will print a message to help user to figure out how to add kernel command line to support virtio-mmio. To instantiate guest virtio-mmio devices using kernel command line (or module) parameter, e.g. virtio_mmio.devices=0x200@0xd2000000:5,0x200@0xd2000200:6 > > Thanks, > Sasha > -- Asias -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html