On Fri, 2011-01-07 at 17:03 +0100, Jan Kiszka wrote: > Am 07.01.2011 16:58, Alex Williamson wrote: > > If migration is done in the background with the -d option, mon is > > NULL and this error message is lost. Instead use error_report(). > > > > Signed-off-by: Alex Williamson <alex.william...@redhat.com> > > Reviewed-by: Jan Kiszka <jan.kis...@siemens.com> > > As already at it: Is this the only error during migration that can get > lost this way?
On the migration source side, I think so. All the save callbacks are return void, so device saves aren't allow to fail. The unmigratable device registration is the only way I know to trigger an error on this end. Thanks, Alex > > --- > > > > savevm.c | 2 +- > > 1 files changed, 1 insertions(+), 1 deletions(-) > > > > diff --git a/savevm.c b/savevm.c > > index 90aa237..148871d 100644 > > --- a/savevm.c > > +++ b/savevm.c > > @@ -1543,7 +1543,7 @@ int qemu_savevm_state_complete(Monitor *mon, QEMUFile > > *f) > > > > r = vmstate_save(f, se); > > if (r < 0) { > > - monitor_printf(mon, "cannot migrate with device '%s'\n", > > se->idstr); > > + error_report("cannot migrate with device '%s'\n", se->idstr); > > return r; > > } > > } > > > >