On Tue, 24 Apr 2018 16:22:26 +0200 Thomas Huth <th...@redhat.com> wrote:
> On 24.04.2018 16:09, Cornelia Huck wrote: > > On Tue, 24 Apr 2018 16:02:59 +0200 > > Thomas Huth <th...@redhat.com> wrote: > > > >> On 24.04.2018 13:49, Christian Borntraeger wrote: > >>> > >>> > >>> On 04/24/2018 01:44 PM, Thomas Huth wrote: > >>>> The consoles ("sclpconsole" and "sclplmconsole") can only be configured > >>>> with "-device" and "-chardev" so far. Other machines use the convenience > >>>> option "-serial" to configure the default consoles, too, even for virtual > >>>> consoles like spapr-vty on the pseries machine. So let's support this > >>>> option on s390x, too, so we can easily enable the serial console here > >>>> again with "-nodefaults", for example. Also map the second -serial > >>>> option to the "sclplmconsole", so that there is now an easy way to > >>>> configure this second console on s390x, too. > >>>> Additionally, the new code is also smaller than the old one and we have > >>>> less s390x-specific code in vl.c :-) > [...] > >>> I read that as "the current command lines continue to work". Correct? > >> > >> Right. I was a little bit afraid that this might break migration, but I > >> gave it a quick check and it still seems to work fine here. > >> "info qom-tree" and "info qtree" at the HMP monitor show slightly > >> different output, though ... not sure whether that's critical or not? > > > > What looks different? If we still send/expect the same kind of > > information, it should not have any impact, I guess. > > $ diff -u sclp-qtree-before.txt sclp-qtree-after.txt > --- sclp-qtree-before.txt 2018-04-24 16:17:14.527921927 +0200 > +++ sclp-qtree-after.txt 2018-04-24 16:17:35.582877042 +0200 > @@ -62,6 +62,6 @@ > bus: s390-sclp-events-bus.0 > type s390-sclp-events-bus > dev: sclpconsole, id "" > - chardev = "sclpcon0" > + chardev = "serial0" > dev: sclp-cpu-hotplug, id "" > dev: sclpquiesce, id "" > > ... i.e. just the label of the chardev changed. I think that's ok, and > migration likely does not care about that. Nod. > > $ diff -u sclp-qom-tree-before.txt sclp-qom-tree-after.txt > --- sclp-qom-tree-before.txt 2018-04-24 16:17:48.462849585 +0200 > +++ sclp-qom-tree-after.txt 2018-04-24 16:18:02.222820256 +0200 > @@ -1,13 +1,14 @@ > -(qemu) info qom-tree > +(qemu) info qom-tree > /machine (s390-ccw-virtio-2.12-machine) > /unattached (container) > - /system[0] (qemu:memory-region) > /sysbus (System) > - /device[1] (virtio-net-ccw) > + /device[0] (qemu-s390x-cpu) > + /s390.ram[0] (qemu:memory-region) > + /device[1] (sclpconsole) > + /device[2] (virtio-net-ccw) > /virtio-backend (virtio-net-device) > /virtio-bus (virtio-ccw-bus) > - /s390.ram[0] (qemu:memory-region) > - /device[0] (qemu-s390x-cpu) > + /system[0] (qemu:memory-region) > /io[0] (qemu:memory-region) > /sclp (sclp) > /s390-sclp-event-facility (s390-sclp-event-facility) > @@ -17,7 +18,6 @@ > /s390-flic-qemu (s390-flic-qemu) > /s390-skeys (s390-skeys-qemu) > /peripheral-anon (container) > - /device[0] (sclpconsole) > /peripheral (container) > /s390-ipl (s390-ipl) > /virtual-css-bridge (virtual-css-bridge) > > ... the sclpconsole now shows up under /unattached instead of > /peripheral-anon ... does that matter? (I don't have a clue) That's because the device is no longer created through qdev_device_add(), which attaches it to peripheral-anon. Not sure what peripheral{,-anon} are used for?