Re: [Qemu-devel] insmod virtio-blk is broken in qemu 1.0

[Anthony Liguori]

On 12/17/2011 09:13 AM, Anthony Liguori wrote:
> On 12/17/2011 02:33 AM, Richard W.M. Jones wrote:
> > On Fri, Dec 16, 2011 at 07:44:10PM -0600, Anthony Liguori wrote:
> > > On 12/16/2011 06:53 PM, Max Filippov wrote:
> > > > > > git bisect says this. I didn't believe it first time, so I ran it
> > > > > > twice with a few modifications, and it pointed to the same commit both
> > > > > > times ...
> > > > >
> > > > > Richard,
> > > > > could you please elaborate on your testcase and configuration
> > > > > (host/target architecture, command lines, etc).
> > > >
> > > > Ok, I've found most of details, what's not clear to me is how you
> > > > decide whether the build is good or bad.
> > > >
> > > > I mean, you need to rebuild qemu on every bisection step, but neither
> > > > this commit nor the previous or the next one
> > > > change anything that would compile for x86 targets.
> > >
> > > Fairly certain this bisect is a red herring.
> > >
> > > tglx reported this the other day in IRC. He narrowed it down to
> > > virtio-serial. He was able to reproduce it both with kvm tools and
> > > QEMU.
> >
> > Yes, we do use virtio-serial. The command line is:
> >
> > /home/rjones/d/qemu/qemu.wrapper \
> > -drive file=/tmp/libguestfs-test-tool-sda-b4hesH,cache=off,format=raw,if=virtio 
> > \
> > -nodefconfig \
> > -machine accel=kvm:tcg \
> > -nodefaults \
> > -nographic \
> > -m 500 \
> > -no-reboot \
> > -no-hpet \
> > -device virtio-serial \
> > -serial stdio \
>
> Okay, I can reproduce this now with a F15 guest.
>
> I've narrowed it down to '-nodefaults -serial stdio'. Can you confirm that if
> you remove those options it works for you?

I've even further narrowed it down to the presents or lack of '-vga cirrus'.  If 
you add '-vga cirrus' to the above command line, the guest will boot successfully.

Regards,

Anthony Liguori

> Regards,
>
> Anthony Liguori
>
>
> > -chardev socket,path=/tmp/libguestfsQQ187c/guestfsd.sock,id=channel0 \
> > -device virtserialport,chardev=channel0,name=org.libguestfs.channel.0 \
> > -kernel /var/tmp/.guestfs-500/kernel.24171 \
> > -initrd /var/tmp/.guestfs-500/initrd.24171 \
> > -append 'panic=1 console=ttyS0 udevtimeout=300 no_timer_check acpi=off
> > printk.time=1 cgroup_disable=memory selinux=0 guestfs_verbose=1 TERM=xterm ' \
> > -drive file=/var/tmp/.guestfs-500/root.24171,snapshot=on,if=virtio,cache=unsafe
> >
> > which comes from this libguestfs test case:
> >
> > LIBGUESTFS_QEMU=~/d/qemu/qemu.wrapper libguestfs-test-tool -t 60
> >
> > where qemu.wrapper is:
> >
> > #!/bin/sh -
> > qemudir=/home/rjones/d/qemu
> > exec $qemudir/x86_64-softmmu/qemu-system-x86_64 -L $qemudir/pc-bios "$@"
> >
> > I'll try it with/without virtio-serial.
> >
> > git bisect red herring is pretty strange?
> >
> > Rich.