[libvirt] [PATCH v2] qemu: Reject unsupported tuning in session mode

[Martin Kletzander]

On Tue, Mar 04, 2014 at 02:04:05PM -0700, Eric Blake wrote:
> On 03/04/2014 07:15 AM, Martin Kletzander wrote:
> > When domain is started with setting that cannot be done, i.e. those
> > that require cgroups, there is no error reported and it succeeds
> > without any message whatsoever.
> >
> > When setting with API, virsh, an error is reported, but only due to
> > the fact that no cgroups are mounted (priv->cgroup == NULL).
> >
> > Given the above it seems reasonable to reject such unsupported
> > settings.
> >
> > Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1023366
> >
> > Signed-off-by: Martin Kletzander <mkletzan at redhat.com>
> > ---
> >
> > @@ -7443,6 +7445,13 @@ static char *qemuDomainGetSchedulerType(virDomainPtr dom,
> >      if (virDomainGetSchedulerTypeEnsureACL(dom->conn, vm->def) < 0)
> >          goto cleanup;
> >
> > +    cfg = virQEMUDriverGetConfig(driver);
> > +    if (!cfg->privileged) {
> > +        virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s",
> > +                       _("CPU tuning is not available in session mode"));
> > +        goto cleanup;
> > +    }
> > +
> >      /* Domain not running, thus no cgroups - return defaults */
> >      if (!virDomainObjIsActive(vm)) {
>
> I can understand failing the Set commands if we can't change things; but
> since we have a fallback for the Get command even for an offline domain,
> shouldn't we stick to returning the defaults rather than erroring out?
>

Particularly for SchedulerType, we can allow the Get method, but the
reading will still fail as there are no cgroups set up.  If we rework
our cgroups code for non-privileged daemons, we will have to check the
cgroup in which the domain belongs to, which will be the same as other
user processes, and doe to that fact the settings won't apply only to
that domain, but for everything in that cgroup; that will make the
values less useful (e.g. maximum used memory cannot be guaranteed
since the cgroup has more tasks inside than only qemu).  Our cgroup
path assembly functions will then need to differ based on whether it
is non-privileged or not.

If that's really needed, I'll rework it that way (even though it'll
probably need to get changed back when user cgroups will be
user-editable), but until then, could we make the error reported now:

$ virsh -c qemu:///session schedinfo dummy
Scheduler      : Unknown
error: Requested operation is not valid: cgroup CPU controller is not mounted

changed into:

$ virsh -c qemu:///session schedinfo dummy
Scheduler      : Unknown
error: Operation not supported: CPU tuning is not available in session mode

???

// I should've added that info into the commit message, I just forgot.

Martin

> Weak ACK; I'd still wait for Dan to weigh in.
>
> --
> Eric Blake   eblake redhat com    +1-919-301-3266
> Libvirt virtualization library http://libvirt.org
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://listman.redhat.com/archives/libvir-list/attachments/20140305/b1e78754/attachment-0001.sig>