When setting up namespace for QEMU we look at mount points under
/dev (like /dev/pts, /dev/mqueue/, etc.) because we want to
preserve those (which is done by moving them to a temp location,
unshare(), and then moving them back). We have a convenience
helper - qemuDomainGetPreservedMounts() - that processes the
mount table and (optionally) moves the other filesystems too.
This helper is also used when attempting to create a path in NS,
because the path, while starting with "/dev/" prefix, may
actually lead to one of those filesystems that we preserved.
And here comes the corner case: while we require the parent mount
table to be in shared mode (equivalent of `mount --make-rshared /'),
these mount events propagate iff the target path exist inside the
slave mount table (= QEMU's private namespace). And since we
create only a subset of /dev nodes, well, that assumption is not
always the case.
For instance, assume that a domain is already running, no
hugepages were configured for it nor any hugetlbfs is mounted.
Now, when a hugetlbfs is mounted into '/dev/hugepages', this is
propagated into the QEMU's namespace, but since the target dir
does not exist in the private /dev, the FS is not mounted in the
namespace.
Fortunately, this difference between namespaces is visible when
comparing /proc/mounts and /proc/$PID/mounts (where PID is the
QEMU's PID). Therefore, if possible we should look at the latter.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
When creating a path in a domain's mount namespace we try to set
ACLs on it, so that it's a verbatim copy of the path in parent's
namespace. The ACLs are queried upfront (by
qemuNamespaceMknodItemInit()) but this is fault tolerant so the
pointer to ACLs might be NULL (meaning no ACLs were queried, for
instance because the underlying filesystem does not support
them). But then we take this NULL and pass it to virFileSetACLs()
which immediately returns an error because NULL is invalid value.
Mimic what we do with SELinux label - only set ACLs if they are
non-NULL which includes symlinks.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
When retiring QEMU_CAPS_BLOCKDEV_HOSTDEV_SCSI capability the
commit removed a bit too much. Previously, all other devices than
VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_SCSI were ignored in
qemuDomainDeviceHostdevDefPostParseRestoreBackendAlias(). But the
commit in question removed not only the capability check but also
this return early statement. Restore it back.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2129239
Fixes: dc8dbb27d4
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
Upcoming patch which is fixing the opening of drivers in monolithic mode
needs to know whether we are inside 'libvirtd' but the code where the
decision needs to happen is not re-compiled per daemon. Thus we need to
pass this information to the stateful driver init function so that it
can be remebered.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The aim of qemuProcessNeedHugepagesPath() is to determine whether
a hugetlbfs mount point is required for given domain (as in
whether qemuBuildMemoryBackendProps() picks up
memory-backend-file pointing to a hugetlbfs mount point). Well,
when domain is configured to use memfd backend then that
condition can never be true. Therefore, skip creating domain's
private path under hugetlbfs mount points.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
Now that all preceding flags were deleted we can fix the enum value.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
'qemuMonitorJSONMigrate' is called from:
- qemuMonitorMigrateToHost
- qemuMonitorMigrateToSocket
Both of the above function are called only from
qemuMigrationSrcStart.
- qemuMonitorMigrateToFd
- called from:
- qemuMigrationSrcToFile
Both instances here pass QEMU_MONITOR_MIGRATE_BACKGROUND
directly.
- qemuMigrationSrcStart
qemuMigrationSrcStart is then called from qemuMigrationSrcRun and
qemuMigrationSrcResume, both of which always add QEMU_MONITOR_MIGRATE_BACKGROUND
to the flags.
Thus any caller always passes the flag so that we can remove the flag
altogether.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Remove the support for enabling the 'blk' and 'inc' parameters of the
'migrate' command as there are no users any more.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
QEMU supported the NBD server required for the new-style migration for a
long time already and when coupled with -blockdev the old style
migration doesn't even work, thus remove support for it.
This patch modifies the code to check that the destination returned data
for the NBD migration and returns an error if it did not and deletes the
fallback code paths which would not work.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The NBD server (detected via 'nbd-server-start' qmp command) was added
to qemu in v1.3 and can't be compiled out.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
In commit 6111b23522 removing pre-blockdev code paths I've
improperly refactored the setup of non-shared storage migration.
Specifically the code checking that there are disks and setting up the
NBD data in the migration cookie was originally outside of the loop
checking the user provided list of specific disks to migrate, but became
part of the block as it was not un-indented when a higher level block
was being removed.
The above caused that if non-shared storage migration is requested, but
the user doesn't provide the list of disks to migrate (thus implying to
migrate every appropriate disk) the code doesn't actually setup the
migration and then later on falls back to the old-style migration which
no longer works with blockdev.
Move the check that there's anything to migrate out of the
'nmigrate_disks' block.
Fixes: 6111b23522
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2125111
Resolves: https://gitlab.com/libvirt/libvirt/-/issues/373
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Although these and functions in the following two patches are for
now just being used by the qemu driver, it makes sense to have all
begin job functions in the same file.
Signed-off-by: Kristina Hanicova <khanicov@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
This patch moves qemuDomainObjEndJob() into
src/conf/virdomainjob as universal virDomainObjEndJob().
Signed-off-by: Kristina Hanicova <khanicov@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
This patch moves qemuDomainObjBeginJob() into
src/conf/virdomainjob as universal virDomainObjBeginJob().
Signed-off-by: Kristina Hanicova <khanicov@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
This patch uses the job object directly in the domain object and
removes the job object from private data of all drivers that use
it as well as other relevant code (initializing and freeing the
structure).
Signed-off-by: Kristina Hanicova <khanicov@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
This patch adds the generalized job object into the domain object
so that it can be used by all drivers without the need to extract
it from the private data.
Because of this, the job object needs to be created and set
during the creation of the domain object. This patch also extends
xmlopt with possible job config containing virDomainJobObj
callbacks, its private data callbacks and one variable
(maxQueuedJobs).
This patch includes:
* addition of virDomainJobObj into virDomainObj (used in the
following patches)
* extending xmlopt with job config structure
* new function for freeing the virDomainJobObj
Signed-off-by: Kristina Hanicova <khanicov@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
The following patches move job object as a member into the domain
object. Because of this, domain_conf (where the domain object is
defined) needs to import the file with the job object.
It makes sense to move jobs to the same level as the domain_conf:
into src/conf/
Signed-off-by: Kristina Hanicova <khanicov@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
This patch moves qemuDomainObjBeginJobInternal() as
virDomainObjBeginJobInternal() into hypervisor in order to be
used by other hypervisors in the following patches.
Signed-off-by: Kristina Hanicova <khanicov@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Now that we assume all the virtio capabilities, this function does not
check anything.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Added by QEMU commit:
commit 74b3e46630446568aecb0be1c77c4875d7a52f6d
Author: Jason Wang <jasowang@redhat.com>
CommitDate: 2019-10-25 07:46:22 -0400
virtio: add property to enable packed virtqueue
Signed-off-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: Eugenio Pérez <eperezma@redhat.com>
Reviewed-by: Jens Freimann <jfreimann@redhat.com>
Message-Id: <20191025083527.30803-9-eperezma@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
git describe: v4.1.0-1780-g74b3e46630 contains: v4.2.0-rc0~32^2~17
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
All the supported QEMU versions should have iothread support
on the virtio-scsi controllers if they are compiled in.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Introduced by QEMU commit 0846e6359c407e372f446723b8b7b09ac20d0f03
released in QEMU 1.3.0
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
We have always considered "migrate_cancel" QMP command to return after
successfully cancelling the migration. But this is no longer true (to be
honest I'm not sure it ever was) as it just changes the migration state
to "cancelling". In most cases the migration is canceled pretty quickly
and we don't really notice anything, but sometimes it takes so long we
even get to clearing migration capabilities before the migration is
actually canceled, which fails as capabilities can only be changed when
no migration is running. So to avoid this issue, we can wait for the
migration to be really canceled after sending migrate_cancel. The only
place where we don't need synchronous behavior is when we're cancelling
migration on user's request while it is actively watched by another
thread.
https://bugzilla.redhat.com/show_bug.cgi?id=2114866
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
We will need a little bit more code around qemuMonitorMigrateCancel to
make sure it works as expected. The new qemuMigrationSrcCancel helper
will avoid repeating the code in several places.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Let's call this qemuMigrationSrcCancelUnattended as the function is
supposed to be used when no other thread is watching the migration.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Just before pushing my earlier commit I've switch order of two
arguments of virProcessGetStatInfo() (as suggested in review).
However, I forgot to swap the arguments in
qemuDomainGetStatsCpuProc() which leads to userTime and sysTime
being swapped.
Fixes: 044b8744d6
Reported-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
For domains started under session URI, we don't set up CGroups
(well, how could we since we're not running as root anyways).
Nevertheless, fetching CPU statistics exits early because of
lacking cpuacct controller. But with recent extension to
virProcessGetStatInfo() we can get the values we need from the
proc filesystem. Implement the fallback for the session URI as
some of virt tools rely on cpu.* stats to be reported (virt-top,
virt-manager).
Resolves: https://gitlab.com/libvirt/libvirt/-/issues/353
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1693707
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The virProcessGetStatInfo() helper parses /proc stat file for
given PID and/or TID and reports cumulative cpuTime which is just
a sum of user and sys times. But in near future, we'll need those
times separately, so make the function return them too (if caller
desires).
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Split up the condition and report a different error message when the
host or host config results in S390 PV launch security being
unavailable.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=2122534
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Marc Hartmayer <mhartmay@linux.ibm.com>
The replacement is 'serial' and 'parallel' respectively introduced at
least in qemu-2.9 and the old versions are deprecated since qemu-6.0
(qemu commit 5965243641d797b22 ).
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The virtio-*-(non-)-transitional device models which replace the use of
'disable-legacy'/'disable-modern' features were introduced in qemu-4.0.
This means we can remove the specific parts of the code for formatting
the old-style device options and replace all other code to solely depend
on the QEMU_CAPS_VIRTIO_PCI_TRANSITIONAL flag.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
QEMU_CAPS_VIRTIO_PCI_TRANSITIONAL is the evolution of
QEMU_CAPS_VIRTIO_PCI_DISABLE_LEGACY from qemu's point of view. Make sure
that we consider both when assesing whether a device belongs on PCIe.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
When commit bac6b266fb added this "functionality" this was the only
naming I could think of, but after discussion with Dan we found the name
'null' fits a bit better, so change it before we make a release with the
old name.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Just like the socket, remove the pidfile when TPM emulator is being stopped. In
order to make this a bit cleaner, try to remove it even if swtpm_ioctl does not
exist.
https://bugzilla.redhat.com/show_bug.cgi?id=2111301
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
After converting virNetworkDef * to g_autoptr(virNetworkDef) the
cleanup codepath was empty, so it has been removed.
Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The virCommand module is specifically designed so that no caller
has to check for retval of individual virCommand*() APIs except
for virCommandRun() where the actual error is reported. Moreover,
virCommandNew*() use g_new0() to allocate memory and thus it's
not really possible for those APIs to return NULL. Which is why
they are even marked as ATTRIBUTE_NONNULL. But there are few
places where we do check the retval which is a dead code
effectively. Drop those checks.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
When multiple VFIO or VDPA devices are assigned to a guest, the guest
can fail to start because the guest fails to map enough memory. For
example, the case mentioned in
https://bugzilla.redhat.com/show_bug.cgi?id=2111317 results in this
failure:
2021-08-05T09:51:47.692578Z qemu-kvm: failed to write, fd=31, errno=14 (Bad address)
2021-08-05T09:51:47.692590Z qemu-kvm: vhost vdpa map fail!
2021-08-05T09:51:47.692594Z qemu-kvm: vhost-vdpa: DMA mapping failed, unable to continue
The current memlock limit calculation does not work for scenarios where
there are multiple such devices assigned to a guest. The root causes are
a little bit different between VFIO and VDPA devices.
For VFIO devices, the issue only occurs when a vIOMMU is present. In
this scenario, each vfio device is assigned a separate AddressSpace
fully mapping guest RAM. When there is no vIOMMU, the devices are all
within the same AddressSpace so no additional memory limit is needed.
For VDPA devices, each device requires the full memory to be mapped
regardless of whether there is a vIOMMU or not.
In order to enable these scenarios, we need to multiply memlock limit
by the number of VDPA devices plus the number of VFIO devices for guests
with a vIOMMU. This has the potential for pushing the memlock limit
above the host physical memory and negating any protection that these
locked memory limits are providing, but there is no other short-term
solution.
In the future, there should be have a revised userspace iommu interface
(iommufd) that the VFIO and VDPA backends can make use of. This will be
able to share locked memory limits between both vfio and vdpa use cases
and address spaces and then we can disable these short term hacks. But
this is still in development upstream.
Resolves: https://bugzilla.redhat.com/2111317
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
