[libvirt PATCH v3 3/4] qemu_migration: get migration blockers before hardcoded checks

[Eugenio Perez Martin]

On Wed, Jul 20, 2022 at 4:48 PM Jiri Denemark <jdenemar at redhat.com> wrote:
>
> On Wed, Jul 20, 2022 at 16:29:45 +0200, Eugenio Perez Martin wrote:
> > On Wed, Jul 20, 2022 at 4:01 PM Jiri Denemark <jdenemar at redhat.com> wrote:
> > >
> > > On Wed, Jul 20, 2022 at 14:15:57 +0200, Eugenio Pérez wrote:
> > > > since qemu 6.0, if migration is blocked for some reason, 'query-migrate'
> > > > will return an array of error strings describing the migration blockers.
> > > > This can be used to check whether there are any devices blocking
> > > > migration, etc.
> > > >
> > > > Enable qemuMigrationSrcIsAllowed to query it.
> > > >
> > > > Signed-off-by: Eugenio Pérez <eperezma at redhat.com>
> > > > ---
> > > > v3:
> > > > * Report message with a colon.
> > > > * Report all blockers instead of only the first.
> > > > ---
> > > >  src/qemu/qemu_migration.c | 34 ++++++++++++++++++++++++++++++++++
> > > >  1 file changed, 34 insertions(+)
> > > >
> > > > diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c
> > > > index b12cb518ee..6ac4ef150b 100644
> > > > --- a/src/qemu/qemu_migration.c
> > > > +++ b/src/qemu/qemu_migration.c
> > > > @@ -1414,6 +1414,20 @@ qemuMigrationSrcIsAllowedHostdev(const virDomainDef *def)
> > > >      return true;
> > > >  }
> > > >
> > > > +static int
> > > > +qemuDomainGetMigrationBlockers(virQEMUDriver *driver,
> > > > +                               virDomainObj *vm,
> > > > +                               char ***blockers)
> > > > +{
> > > > +    qemuDomainObjPrivate *priv = vm->privateData;
> > > > +    int rc;
> > > > +
> > > > +    qemuDomainObjEnterMonitor(driver, vm);
> > > > +    rc = qemuMonitorGetMigrationBlockers(priv->mon, blockers);
> > > > +    qemuDomainObjExitMonitor(vm);
> > > > +
> > > > +    return rc;
> > > > +}
> > > >
> > > >  /**
> > > >   * qemuMigrationSrcIsAllowed:
> > > > @@ -1439,6 +1453,26 @@ qemuMigrationSrcIsAllowed(virQEMUDriver *driver,
> > > >      int nsnapshots;
> > > >      int pauseReason;
> > > >      size_t i;
> > > > +    int r;
> > > > +
> > > > +    /* Ask qemu if it have a migration blocker */
> > > > +    if (virQEMUCapsGet(priv->qemuCaps, QEMU_CAPS_MIGRATION_BLOCKED_REASONS)) {
> > > > +        g_auto(GStrv) blockers = NULL;
> > > > +        r = qemuDomainGetMigrationBlockers(driver, vm, &blockers);
> > > > +        if (r != 0) {
> > > > +            virReportError(VIR_ERR_OPERATION_INVALID,
> > > > +                           _("cannot migrate domain: %s"),
> > > > +                           _("error getting blockers"));
> > > > +            return false;
> > > > +        }
> > >
> > > As mentioned in v2 review the virReportError call should be dropped as
> > > it overwrites the error reported by qemuDomainGetMigrationBlockers. That
> > > is, you can just
> > >
> > >     if (qemuDomainGetMigrationBlockers(driver, vm, &blockers) < 0)
> > >         return false;
> > >
> >
> > But there are a few conditions that don't report an error, like a bad
> > JSON answer. For example, if "blockers" is not an array parsing the
> > response JSON, libvirt would not print any error, isn't it?
>
> Well in qemuDomainGetMigrationBlockers you now have
>
>     if (!(jblockers = virJSONValueObjectGetArray(data, "blocked-reasons")))
>         return 0;
>
> so you would not get pass r != 0 in such case anyway. If you wanted to
> distinguish missing blocked-reasons (that is no blockers) and
> blocked-reasons which is not an array (invalid response), you would need
> to do something else:
>
>     if (!(jblockers = virJSONValueObjectGet(data, "blocked-reasons")))
>         return 0;
>
>     if (!virJSONValueIsArray(jblockers)) {
>         virReportError(VIR_ERR_INTERNAL_ERROR, "%s",
>                        _("blocker-reasons is not an array"));
>         return -1;
>     }
>
> All this inside qemuDomainGetMigrationBlockers() to make sure the
> function reports an error when returning -1. The caller would not report
> any error anyway.
>

That's right.

Deleting error report then.

Thanks!