Alex, as I told you, why not try to still boot the way you used to (mmc/nfs/...) and then simply mount the virtio device from user space? Then you can just put a printk in whatever code you are not sure if executiong and verify using dmesg. (come to think of it you can do that anyway just looking at the guest console output). Once you have virtio working, so you can mount that device and interact with whatever file system is on there, you can fix booting on that. The virtio device(s) is going to be either populated by the devtmps from the kernel or populated by udev. You can set rootfs=/dev/MYVIRTIODEV on the kernel boot command line to something that the kernel can recognize to your virtio device (I forget how this is mapped up exactly, but take a look at some other driver/configuration and see how it's mapped, or simply look at the kernel command line parsing). But again, I suggest you split up mounting and acutally booting. -Christoffer On Wed, Mar 21, 2012 at 1:09 PM, Alexander Golec <thejfasi at gmail.com> wrote: > I've successfully managed to boot qemu with the virtio-mmio changes, but I'm not sure what to tell it to use as a root device. The boot just stops at the 'Wait for root device?' stage, and I'm not sure what to give it as a command line param. > > On a possibly unrelated note, I'm not sure how to check that the virtio-mmio init code is even running? > > Alex > _______________________________________________ > Android-virt mailing list > Android-virt at lists.cs.columbia.edu > https://lists.cs.columbia.edu/cucslists/listinfo/android-virt
