Re: [PATCH 3/3] migration-test: Add COLO migration unit test

[Peter Xu <peterx@redhat.com>]

On Thu, Jan 15, 2026 at 11:35:00PM +0100, Lukas Straub wrote:
> > > diff --git a/tests/qtest/migration/framework.c b/tests/qtest/migration/framework.c
> > > index 8c1fc6e009f16dc05a47e917167f62e0250ca992..08bca49a8980f9988be9447acf54b17acd56da94 100644
> > > --- a/tests/qtest/migration/framework.c
> > > +++ b/tests/qtest/migration/framework.c
> > > @@ -315,7 +315,7 @@ int migrate_args(char **from, char **to, const char *uri, MigrateStart *args)
> > >      if (strcmp(arch, "i386") == 0 || strcmp(arch, "x86_64") == 0) {
> > >          memory_size = "150M";
> > >  
> > > -        if (g_str_equal(arch, "i386")) {
> > > +        if (g_str_equal(arch, "i386") || args->is_colo) {  
> > 
> > OK, one more reference..
> > 
> > I'm curious, why Q35 is not supported?
> 
> There is a bug in the emulated ahci disk controller which crashes when
> it's vmstate is loaded more than once.

I wonder what happens if one does a loadvm on top of a migrated QEMU, then.

Is that the only device that is broken?  Could we fix the device instead?

Thanks,

-- 
Peter Xu