Re: [Xen-devel] Debootstrapping a rootfs for Xen on FastModels A15

[Stefano Stabellini <stefano.stabellini@xxxxxxxxxxxxx>]

On Sun, 31 Mar 2013, Sander Bogaert wrote:
> Hi,
> 
> I've tried several different approaches to debootstrap a working
> rootfs for with a Cortex A15 model ( FastModels ) and Xen but still
> face some issues.
> 
> Some information on my setup.
> For a Dom0 kernel I'm using the latest version from
> git://github.com/torvalds/linux.git, I attached my config as it might
> be of some use for others. I'm using the latest Xen unstable branch
> git://xenbits.xen.org/xen.git and my device tree comes from:
> git://xenbits.xen.org/people/sstabellini/device-trees.git .
> 
> These are my xen and kernel arguments:
> bootargs = "dom0_mem=128M dom0_max_vcpus=1 console=com1,vga";
>  xen,dom0-bootargs = "earlyprintk=xenboot console=ttyAMA0
> root=/dev/mmcblk0 init=/bin/bash rw ip=dhcp";
> 
> This is how I launch the model ( where the variables of course contain
> the right paths ):
> ./model_shell ../resources/Cortex-A15x1/cadi_system_Linux-Release-GCC-4.4.so
> $1 -C motherboard.flashloader0.fname=$2 -C
> motherboard.smsc_91c111.enabled=1 -C
> motherboard.hostbridge.userNetworking=1 -C
> motherboard.pl011_uart0.out_file=./log.txt -C
> motherboard.mmc.p_mmc_file="$3"
> 
> To get a basic filesystem I:
> - created a empty file ( with fallocate or dd )
> - formated it ext4
> - mounted the file and used HOSTNAME=armhf sudo debootstrap --foreign
> --arch armhf wheezy $path
> - did the some configuration and the second stage debootstrapping
> 
> The problem now is Xen hangs on/after booting dom0, I included a log (
> log.txt ). This filesystem boots fine with the same kernel, without
> Xen on the same FastModel model.
> 
> My question: did I miss something in the debootstrapping? Are there
> some steps I don't know of? Does anyway have an idea on what goes
> wrong ( see log.txt )?

I don't see any mistakes in your setup. Unless you are running an
outdated version of the FastModel (they were known to have a broken
virt_timer implementation), my guess is that you are seeing an
issue that we actually have already solved.

At the moment Ian is on vacation (the Xen on ARM committer), as a
consequence we have several outstanding patches, many of which are
actually bug fixes. Maybe the very latest trees would solve your issue.
Could you please try:

git://xenbits.xen.org/people/sstabellini/linux-pvhvm.git 3.9-rc3-smp-dom0-hack-2
git://xenbits.xen.org/people/sstabellini/xen-unstable.git guest-smp-3

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel