[libvirt] [PATCH v3 01/30] qemu: Explicitly add/remove /dev/vfio/vfio to/from NS/CGroups

[Cole Robinson]

On 12/2/19 9:26 AM, Michal Privoznik wrote:
> In near future, the decision what to do with /dev/vfio/vfio with
> respect to domain namespace and CGroup is going to be moved out
> of qemuDomainGetHostdevPath() because there will be some other
> types of devices than hostdevs that need access to VFIO.
> 
> All functions that I'm changing (except qemuSetupHostdevCgroup())
> assume that hostdev we are adding/removing to VM is not in the
> definition yet (because of how qemuDomainNeedsVFIO() is written).
> Fortunately, this assumption is true.
> 
> For qemuSetupHostdevCgroup(), the worst thing that may happen is
> that we allow /dev/vfio/vfio which was already allowed.
> 
> Signed-off-by: Michal Privoznik <mprivozn at redhat.com>
> ---
>  src/qemu/qemu_cgroup.c | 47 +++++++++++++++++++++++++++++++++++++++++-
>  src/qemu/qemu_domain.c | 36 ++++++++++++++++++++++++++++++++
>  2 files changed, 82 insertions(+), 1 deletion(-)
> 
> diff --git a/src/qemu/qemu_cgroup.c b/src/qemu/qemu_cgroup.c
> index 31cf71146b..65b148cd83 100644
> --- a/src/qemu/qemu_cgroup.c
> +++ b/src/qemu/qemu_cgroup.c
> @@ -25,6 +25,7 @@
>  #include "qemu_domain.h"
>  #include "qemu_process.h"
>  #include "qemu_extdevice.h"
> +#include "qemu_hostdev.h"
>  #include "virlog.h"
>  #include "viralloc.h"
>  #include "virerror.h"
> @@ -359,6 +360,17 @@ qemuTeardownInputCgroup(virDomainObjPtr vm,
>  }
>  
>  
> +/**
> + * qemuSetupHostdevCgroup:
> + * vm: domain object
> + * @dev: device to allow
> + *
> + * For given host device @dev allow access to in Cgroups.
> + * Note, @dev must not be in @vm's definition.
> + *

This last line should be dropped, it is still called like that via
qemuSetupDevicesCgroup

> + * Returns: 0 on success,
> + *         -1 otherwise.
> + */
>  int
>  qemuSetupHostdevCgroup(virDomainObjPtr vm,
>                         virDomainHostdevDefPtr dev)
> @@ -385,6 +397,16 @@ qemuSetupHostdevCgroup(virDomainObjPtr vm,
>              goto cleanup;
>      }
>  

Maybe just add a comment here explaining that, at VM startup, this can
lead to /dev/vfio/vfio being added to the cgroup multiple times, but
that has never shown to be a problem.

Reviewed-by: Cole Robinson <crobinso at redhat.com>

- Cole