Re: [libvirt] [PATCH v4 11/14] qemu: Bump the memory locking limit for mdevs as well

On Wed, 2017-03-22 at 16:27 +0100, Erik Skultety wrote: > Since mdevs are just another type of VFIO devices, we should increase > the memory locking limit the same way we do for VFIO PCI devices. > > Signed-off-by: Erik Skultety <eskultet(a)redhat.com&gt; > --- >  src/qemu/qemu_domain.c | 9 +++++---- >  1 file changed, 5 insertions(+), 4 deletions(-) > > diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c > index d1ac1d641b..04e64b47ea 100644 > --- a/src/qemu/qemu_domain.c > +++ b/src/qemu/qemu_domain.c > @@ -6347,11 +6347,12 @@ qemuDomainRequiresMemLock(virDomainDefPtr def) >          return true; > >      for (i = 0; i < def->nhostdevs; i++) { > -        virDomainHostdevDefPtr dev = def->hostdevs[i]; > +        virDomainHostdevSubsysPtr subsys = &def->hostdevs[i]->source.subsys; > > -        if (dev->mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS && > -            dev->source.subsys.type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI && > -            dev->source.subsys.u.pci.backend == VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO) > +        if (def->hostdevs[i]->mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS && > +            (subsys->type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_MDEV || > +             (subsys->type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI && > +              subsys->u.pci.backend == VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO))) >              return true; >      } Now that we have test suite coverage for the calculation of memory locking limits (qemumemlocktest), it would be neat if you could add test cases for mdevs as well.

Maybe the comment above the loop (which has been moved to the qemuDomainGetMemLockLimitBytes() function now) could be updated to mention mdev-specific information, if any?

-- Andrea Bolognani / Red Hat / Virtualization