The @meminfo allocated in qemuMonitorGetMemoryDeviceInfo() may be
lost when qemuDomainObjExitMonitor() failed.
Signed-off-by: Yi Wang <wang.yi59@zte.com.cn>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Setting the 'group_name' for a disk would falsely trigger a error path
as in commit 4b57f76502 we did not properly check the return value of
VIR_STRDUP.
This reverts commit 2841e675.
It turns out that adding the host_mtu field to the PCI capabilities in
the guest bumps the length of PCI capabilities beyond the 32 byte
boundary, so the virtio-net device gets 64 bytes of ioport space
instead of 32, which offsets the address of all the other following
devices. Migration doesn't work very well when the location and length
of PCI capabilities of devices is changed between source and
destination.
This means that we need to make sure that the absence/presence of
host_mtu on the qemu commandline always matches between source and
destination, which means that we need to make setting of host_mtu an
opt-in thing (it can't happen automatically when the bridge being used
has a non-default MTU, which is what commit 2841e675 implemented).
I do want to re-implement this feature with an <mtu auto='on'/>
setting, but probably won't backport that to any stable branches, so
I'm first reverting the original commit, and that revert can be pushed
to the few releases that have been made since the original (3.1.0 -
3.3.0)
Resolves: https://bugzilla.redhat.com/1449346
The error message would contain first vcpu id after the list of vcpus
selected for modification. To print the proper vcpu id remember the
first vcpu selected to be modified.
The code causes the 'offset' variable to be overwritten (possibly with
NULL if neither of the vCPUs is halted) which causes a crash since the
variable is still used after that part.
Additionally there's a bug, since strstr() would look up the '(halted)'
string in the whole string rather than just the currently processed line
the returned data is completely bogus.
Rather than switching to single line parsing let's remove the code
altogether since it has a commonly used JSON monitor alternative and
the data itself is not very useful to report.
The code was introduced in commit cc5e695bde
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1452106
Namely, this patch is about virMediatedDeviceGetIOMMUGroup{Dev,Num}
functions. There's no compelling reason why these functions should take
an object, on the contrary, having to create an object every time one
needs to query the IOMMU group number, discarding the object afterwards,
seems odd.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Since we allow active layer block commit the users are allowed to commit
the top of the chain (e.g. vda) into the backing image. The API would
not accept that parameter, as it tried to look up the image in the
backing chain.
Add the ability to use the top level image target name explicitly as the
top image of the block commit operation.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1451394
The QEMU default is GICv2, and some of the code in libvirt
relies on the exact value. Stop pretending that's not the
case and use GICv2 explicitly where needed.
Signed-off-by: Andrea Bolognani <abologna@redhat.com>
There are currently some limitations in the emulated GICv3
that make it unsuitable as a default. Use GICv2 instead.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1450433
Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Currently we consider all UNIX paths with specific prefix as generated
by libvirt, but that's a wrong assumption. Let's make the detection
better by actually checking whether the whole path matches one of the
paths that we generate or generated in the past.
The UNIX path isn't stored in config XML since libvirt-1.3.1.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1446980
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Add kernel_irqchip=split/on to the QEMU command line
and a capability that looks for it in query-command-line-options
output. For the 'split' option, use a version check
since it cannot be reasonably probed.
https://bugzilla.redhat.com/show_bug.cgi?id=1427005
If a shutdown is expected because it was triggered via libvirt we can
also expect the monitor to close. In those cases do not report an
internal error like:
"internal error: End of file from qemu monitor"
Signed-off-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
Adjust the current message to make it clear, that it is the hotplug
operation that is unsupported with the given host device type.
https://bugzilla.redhat.com/show_bug.cgi?id=1450072
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Added only in drivers that were already calling
virCapabilitiesInitNUMA(). Instead of refactoring all the callers to
behave the same way in case of error, just follow what the callers are
doing for all the functions.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Even though there are several checks before calling this function
and for some scenarios we don't call it at all (e.g. on disk hot
unplug), it may be possible to sneak in some weird files (e.g. if
domain would have RNG with /dev/shm/some_file as its backend). No
matter how improbable, we shouldn't unlink it as we would be
unlinking a file from the host which we haven't created in the
first place.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
Just like in previous commit, this fixes the same issue for
hotplug.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
While the code allows devices to already be there (by some
miracle), we shouldn't try to create devices that don't belong to
us. For instance, we shouldn't try to create /dev/shm/file
because /dev/shm is a mount point that is preserved. Therefore if
a file is created there from an outside (e.g. by mgmt application
or some other daemon running on the system like vhostmd), it
exists in the qemu namespace too as the mount point is the same.
It's only /dev and /dev only that is different. The same
reasoning applies to all other preserved mount points.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
Currently, all we need to do in qemuDomainCreateDeviceRecursive() is to
take given @device, get all kinds of info on it (major & minor numbers,
owner, seclabels) and create its copy at a temporary location @path
(usually /var/run/libvirt/qemu/$domName.dev), if @device live under
/dev. This is, however, very loose condition, as it also means
/dev/shm/* is created too. Therefor, we will need to pass more arguments
into the function for better decision making (e.g. list of mount points
under /dev). Instead of adding more arguments to all the functions (not
easily reachable because some functions are callback with strictly
defined type), lets just turn this one 'const char *' into a 'struct *'.
New "arguments" can be then added at no cost.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
When setting up mount namespace for a qemu domain the following
steps are executed:
1) get list of mountpoints under /dev/
2) move them to /var/run/libvirt/qemu/$domName.ext
3) start constructing new device tree under /var/run/libvirt/qemu/$domName.dev
4) move the mountpoint of the new device tree to /dev
5) restore original mountpoints from step 2)
Note the problem with this approach is that if some device in step
3) requires access to a mountpoint from step 2) it will fail as
the mountpoint is not there anymore. For instance consider the
following domain disk configuration:
<disk type='file' device='disk'>
<driver name='qemu' type='raw'/>
<source file='/dev/shm/vhostmd0'/>
<target dev='vdb' bus='virtio'/>
<address type='pci' domain='0x0000' bus='0x00' slot='0x0a' function='0x0'/>
</disk>
In this case operation fails as we are unable to create vhostmd0
in the new device tree because after step 2) there is no /dev/shm
anymore. Leave aside fact that we shouldn't try to create devices
living in other mountpoints. That's a separate bug that will be
addressed later.
Currently, the order described above is rearranged to:
1) get list of mountpoints under /dev/
2) start constructing new device tree under /var/run/libvirt/qemu/$domName.dev
3) move them to /var/run/libvirt/qemu/$domName.ext
4) move the mountpoint of the new device tree to /dev
5) restore original mountpoints from step 3)
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Cedric Bosdonnat <cbosdonnat@suse.com>
While fixing a bug with incorrectly freed memory in commit
v3.1.0-399-g5498aa29a, I accidentally broke persistent migration of
transient domains. Before adding qemuDomainDefCopy in the path, the code
just took NULL from vm->newDef and used it as the persistent def, which
resulted in no persistent XML being sent in the migration cookie. This
scenario is perfectly valid and the destination correctly handles it by
using the incoming live definition and storing it as the persistent one.
After the mentioned commit libvirtd would just segfault in the described
scenario.
https://bugzilla.redhat.com/show_bug.cgi?id=1446205
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
When creating v3.2.0-77-g8be3ccd04 commit, I completely forgot that one
migration capability is very special. It's the "events" capability which
tells QEMU to report "MIGRATION" events. Since libvirt always wants the
events, it is enabled in qemuConnectMonitor and the rest of the code
should not touch it.
https://bugzilla.redhat.com/show_bug.cgi?id=1439841https://bugzilla.redhat.com/show_bug.cgi?id=1441165
Messed-up-by: Jiri Denemark <jdenemar@redhat.com>
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
... with VIR_NET_GENERATED_MACV???_PREFIX, which is defined in
util/virnetdevmacvlan.h.
Since VIR_NET_GENERATED_PREFIX is used for plain tap devices, it is
renamed to VIR_NET_GENERATED_TAP_PREFIX and moved to virnetdev.h
Nothing that could happen during networkNotifyActualDevice() could
justify unceremoniously killing the qemu process, but that's what we
were doing.
In particular, new code added in commit 85bcc022 (first appearred in
libvirt-3.2.0) attempts to reattach tap devices to their assigned
bridge devices when libvirtd restarts (to make it easier to recover
from a restart of a libvirt network). But if the network has been
stopped and *not* restarted, the bridge device won't exist and
networkNotifyActualDevice() will fail.
This patch changes networkNotifyActualDevice() and
qemuProcessNotifyNets() to return void, so that qemuProcessReconnect()
will soldier on regardless of what happens (any errors will still be
logged though).
Partially resolves: https://bugzilla.redhat.com/1442700
This is a USB3 controller and it's a better choice than piix3-uhci.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Acked-by: Andrea Bolognani <abologna@redhat.com>
The new logic will set the piix3-uhci if available regardless of
any architecture and it will be updated to better model based on
architecture and device existence.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Acked-by: Andrea Bolognani <abologna@redhat.com>
Since commit c5f6151390 qemuDomainBlockInfo tries to update the
"physical" storage size for all network storage and not only block
devices.
Since the storage driver APIs to do this are not implemented for certain
storage types (RBD, iSCSI, ...) the code would fail to retrieve any data
since the failure of qemuDomainStorageUpdatePhysical is fatal.
Since it's desired to return data even if the total size can't be
updated we need to ignore errors from that function and return plausible
data.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1442344
Since the private data structure is not freed upon stopping a VM, the
usbaddrs pointer would be leaked:
==15388== 136 (16 direct, 120 indirect) bytes in 1 blocks are definitely lost in loss record 893 of 1,019
==15388== at 0x4C2CF55: calloc (vg_replace_malloc.c:711)
==15388== by 0x54BF64A: virAlloc (viralloc.c:144)
==15388== by 0x5547588: virDomainUSBAddressSetCreate (domain_addr.c:1608)
==15388== by 0x144D38A2: qemuDomainAssignUSBAddresses (qemu_domain_address.c:2458)
==15388== by 0x144D38A2: qemuDomainAssignAddresses (qemu_domain_address.c:2515)
==15388== by 0x144ED1E3: qemuProcessPrepareDomain (qemu_process.c:5398)
==15388== by 0x144F51FF: qemuProcessStart (qemu_process.c:5979)
[...]
Clean the stale data after shutting down the VM. Otherwise the data
would be leaked on next VM start. This happens due to the fact that the
private data object is not freed on destroy of the VM.
This patch maps /domain/cpu/cache element into -cpu parameters:
- <cache mode='passthrough'/> is translated to host-cache-info=on
- <cache level='3' mode='emulate'/> is transformed into l3-cache=on
- <cache mode='disable'/> is turned in host-cache-info=off,l3-cache=off
Any other <cache> element is forbidden.
The tricky part is detecting whether QEMU supports the CPU properties.
The 'host-cache-info' property is introduced in v2.4.0-1389-ge265e3e480,
earlier QEMU releases enabled host-cache-info by default and had no way
to disable it. If the property is present, it defaults to 'off' for any
QEMU until at least 2.9.0.
The 'l3-cache' property was introduced later by v2.7.0-200-g14c985cffa.
Earlier versions worked as if l3-cache=off was passed. For any QEMU
until at least 2.9.0 l3-cache is 'off' by default.
QEMU 2.9.0 was the first release which supports probing both properties
by running device-list-properties with typename=host-x86_64-cpu. Older
QEMU releases did not support device-list-properties command for CPU
devices. Thus we can't really rely on probing them and we can just use
query-cpu-model-expansion QMP command as a witness.
Because the cache property probing is only reliable for QEMU >= 2.9.0
when both are already supported for quite a few releases, we let QEMU
report an error if a specific cache mode is explicitly requested. The
other mode (or both if a user requested CPU cache to be disabled) is
explicitly turned off for QEMU >= 2.9.0 to avoid any surprises in case
the QEMU defaults change. Any older QEMU already turns them off so not
doing so explicitly does not make any harm.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Not all async jobs are visible via virDomainGetJobStats (either they are
too fast or getting the stats is not allowed during the job), but
forcing all of them to advertise the operation is easier than hunting
the jobs for which fetching statistics is allowed. And we won't need to
think about this when we add support for getting stats for more jobs.
https://bugzilla.redhat.com/show_bug.cgi?id=1441563
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
As with virtio-scsi, the "internal error" messages after
preparing a vhost-scsi hostdev overwrites more meaningful
error messages deeper in the callchain. Remove it too.
Signed-off-by: Eric Farman <farman@linux.vnet.ibm.com>
I tried to attach a SCSI LUN to two different guests, and forgot
to specify "shareable" in the hostdev XML. Attaching the device
to the second guest failed, but the message was not helpful in
telling me what I was doing wrong:
$ cat scsi_scratch_disk.xml
<hostdev mode='subsystem' type='scsi'>
<source>
<adapter name='scsi_host3'/>
<address bus='0' target='15' unit='1074151456'/>
</source>
</hostdev>
$ virsh attach-device dasd_sles_d99c scsi_scratch_disk.xml
Device attached successfully
$ virsh attach-device dasd_fedora_0e1e scsi_scratch_disk.xml
error: Failed to attach device from scsi_scratch_disk.xml
error: internal error: Unable to prepare scsi hostdev: scsi_host3:0:15:1074151456
I eventually discovered my error, but thought it was weird that
Libvirt doesn't provide something more helpful in this case.
Looking over the code we had just gone through, I commented out
the "internal error" message, and got something more useful:
$ virsh attach-device dasd_fedora_0e1e scsi_scratch_disk.xml
error: Failed to attach device from scsi_scratch_disk.xml
error: Requested operation is not valid: SCSI device 3:0:15:1074151456 is already in use by other domain(s) as 'non-shareable'
Looking over the error paths here, we seem to issue better
messages deeper in the callchain so these "internal error"
messages overwrite any of them. Remove them, so that the
more detailed errors are seen.
Signed-off-by: Eric Farman <farman@linux.vnet.ibm.com>
0feebab2 adds calling qemuBlockNodeNamesDetect for completed job
on updating block jobs. This affects cancelling drive mirror logic as
this function drops vm lock. Now we have to recheck all disks
before the disk with the completed block job before going
to wait for block job events.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
qemuDomainGetNumaParameters would return the automatic nodeset even for
the persistent config if the domain was running. This is incorrect since
the automatic nodeset will be re-queried upon starting the vm.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1445325
While peer-to-peer migration enters the Confirm phase even if the
Perform phase fails, the client which initiated a non-p2p migration will
never call virDomainMigrateConfirm* API if the Perform phase failed.
Thus we need to explicitly reset migration before reporting a failure
from the Perform phase API.
https://bugzilla.redhat.com/show_bug.cgi?id=1425003
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Migration with old QEMU which does not support query-migrate-parameters
would fail because the QMP command is called unconditionally since the
introduction of TLS migration. Previously it was only called if the user
explicitly requested a feature which uses QEMU migration parameters. And
even then the situation was not ideal, instead of reporting an
unsupported feature we'd just complain about missing QMP command.
Trivially no migration parameters are supported when
query-migrate-parameters QMP command is missing. There's no need to
report an error if it is missing, the callers will report better error
if needed.
https://bugzilla.redhat.com/show_bug.cgi?id=1441934
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
it should be a comparison of modes between new and old devices. So
the argument of the second virDomainNetGetActualDirectMode should be
newdev.
Signed-off-by: ZhiPeng Lu <lu.zhipeng@zte.com.cn>
This patch makes use of the virNetDevSetCoalesce() function to make
appropriate settings effective for devices that support them.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1414627
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
We are currently parsing only rx/frames/max because that's the only
value that makes sense for us. The tun device just added support for
this one and the others are only supported by hardware devices which
we don't need to worry about as the only way we'd pass those to the
domain is using <hostdev/> or <interface type='hostdev'/>. And in
those cases the guest can modify the settings itself.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
In the vcpu hotplug code if exit from the monitor failed we would still
attempt to save the status XML. When the daemon is terminated the
monitor socket is closed. In such case, the written status XML would not
contain the monitor path and thus be invalid.
Avoid this issue by only saving status XML on success of the monitor
command.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1439452
The history of USB controller for ppc64 guest is complex and goes
back to libvirt 1.3.1 where the fun started.
Prior Libvirt 1.3.1 if no model for USB controller was specified
we've simply passed "-usb" on QEMU command line.
Since Libvirt 1.3.1 there is a patch (8156493d8d) that fixes this
issue by using "-device pci-ohci,..." but it breaks migration with
older Libvirts which was agreed that's acceptable. However this
patch didn't reflect this change in the domain XML and the model
was still missing.
Since Libvirt 2.2.0 there is a patch (f55eaccb0c) that fixes the
issue with not setting the USB model into domain XML which we need
to know about to not break the migration and since the default
model was *pci-ohci* it was used as default in this patch as well.
This patch tries to take all the previous changes into account and
also change the default for newly defined domains that don't specify
any model for USB controller.
The VIR_DOMAIN_DEF_PARSE_ABI_UPDATE is set only if new domain is
defined or new device is added into a domain which means that in
all other cases we will use the old *pci-ohci* model instead of the
better and not broken *nec-usb-xhci* model.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1373184
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
So far there is probably no change that is allowed to be done
by the VIR_DOMAIN_DEF_PARSE_ABI_UPDATE flag that would break
guest ABI but this may change in the future.
This introduces new VIR_DOMAIN_DEF_PARSE_ABI_UPDATE_MIGRATION
which should be used only for ABI updates that are "safe" for
persistent migration.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
With QEMU older than 2.9.0 libvirt uses CPUID instruction to determine
what CPU features are supported on the host. This was later used when
checking compatibility of guest CPUs. Since QEMU 2.9.0 we ask QEMU for
the host CPU data. But the two methods we use usually provide disjoint
sets of CPU features because QEMU/KVM does not support all features
provided by the host CPU and on the other hand it can enable some
feature even if the host CPU does not support them.
So if there is a domain which requires a CPU features disabled by
QEMU/KVM, libvirt will refuse to start it with QEMU > 2.9.0 as its guest
CPU is incompatible with the host CPU data we got from QEMU. But such
domain would happily start on older QEMU (of course, the features would
be missing the guest CPU). To fix this regression, we need to combine
both CPU feature sets when checking guest CPU compatibility.
https://bugzilla.redhat.com/show_bug.cgi?id=1439933
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
We already know from QEMU which CPU features will block migration. Let's
use this information to make a migratable copy of the host CPU model and
use it for updating guest CPU specification. This will allow us to drop
feature filtering from virCPUUpdate where it was just a hack.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Soon we will need to store multiple host CPU definitions in
virQEMUCapsHostCPUData and qemuCaps users will want to request the one
they need. This patch introduces virQEMUCapsHostCPUType enum which will
be used for specifying the requested CPU definition.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
We need to store several CPU related data structure for both KVM and
TCG. So instead of keeping two different copies of everything let's
make a virQEMUCapsHostCPUData struct and use it twice.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
This introduces virQEMUCapsHostCPUDataCopy which will later be
refactored a bit and called twice from virQEMUCapsNewCopy.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
CLang's optimizer is more aggressive at inlining functions than
gcc and so will often inline functions that our tests want to
mock-override. This causes the test to fail in bizarre ways.
We don't want to disable inlining completely, but we must at
least prevent inlining of mocked functions. Fortunately there
is a 'noinline' attribute that lets us control this per function.
A syntax check rule is added that parses tests/*mock.c to extract
the list of functions that are mocked (restricted to names starting
with 'vir' prefix). It then checks that src/*.h header file to
ensure it has a 'ATTRIBUTE_NOINLINE' annotation. This should prevent
use from bit-rotting in future.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Introduce new wrapper functions without *Machine* in the function
name that take the whole virDomainDef structure as argument and
call the existing functions with *Machine* in the function name.
Change the arguments of existing functions to *machine* and *arch*
because they don't need the whole virDomainDef structure and they
could be used in places where we don't have virDomainDef.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Since the disks are copied by qemu, there's no need to enforce
cache=none. Thankfully the code that added qemuMigrateDisk did not break
existing configs, since if you don't select any disk to migrate
explicitly the code behaves sanely.
The logic for determining whether a disk should be migrated is
open-coded since using qemuMigrateDisk twice would be semantically
incorrect.
The code that validates whether an internal snapshot is possible would
reject an empty but not-readonly drive. Since floppies can have this
property, add a check for emptiness.
==20406== 8 bytes in 1 blocks are definitely lost in loss record 24 of 1,059
==20406== at 0x4C2CF55: calloc (vg_replace_malloc.c:711)
==20406== by 0x54BF530: virAllocN (viralloc.c:191)
==20406== by 0x54D37C4: virConfGetValueStringList (virconf.c:1001)
==20406== by 0x144E4E8E: virQEMUDriverConfigLoadFile (qemu_conf.c:835)
==20406== by 0x1452A744: qemuStateInitialize (qemu_driver.c:664)
==20406== by 0x55DB585: virStateInitialize (libvirt.c:770)
==20406== by 0x124570: daemonRunStateInit (libvirtd.c:881)
==20406== by 0x5532990: virThreadHelper (virthread.c:206)
==20406== by 0x8C82493: start_thread (in /lib64/libpthread-2.24.so)
==20406== by 0x8F7FA1E: clone (in /lib64/libc-2.24.so)
==20406== 4 bytes in 1 blocks are definitely lost in loss record 6 of 1,059
==20406== at 0x4C2AF3F: malloc (vg_replace_malloc.c:299)
==20406== by 0x8F17D39: strdup (in /lib64/libc-2.24.so)
==20406== by 0x552C0E0: virStrdup (virstring.c:784)
==20406== by 0x54D3622: virConfGetValueString (virconf.c:945)
==20406== by 0x144E4692: virQEMUDriverConfigLoadFile (qemu_conf.c:687)
==20406== by 0x1452A744: qemuStateInitialize (qemu_driver.c:664)
==20406== by 0x55DB585: virStateInitialize (libvirt.c:770)
==20406== by 0x124570: daemonRunStateInit (libvirtd.c:881)
==20406== by 0x5532990: virThreadHelper (virthread.c:206)
==20406== by 0x8C82493: start_thread (in /lib64/libpthread-2.24.so)
==20406== by 0x8F7FA1E: clone (in /lib64/libc-2.24.so)
Commit a4a39d90 added a check that checks for VFIO support with mediated
devices. The problem is that the hostdev preparing functions behave like
a fallthrough if device of that specific type doesn't exist. However,
the check for VFIO support was independent of the existence of a mdev
device which caused the guest to fail to start with any device to be
directly assigned if VFIO was disabled/unavailable in the kernel.
The proposed change first ensures that it makes sense to check for VFIO
support in the first place, and only then performs the VFIO support check
itself.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1441291
Signed-off-by: Erik Skultety <eskultet@redhat.com>
This removes the hacky extern global variable and modifies the
test code to properly create QEMU capabilities cache for QEMU
binaries used in our tests.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
This attribute is not needed here, since @mon is in use.
Signed-off-by: Marc Hartmayer <mhartmay@linux.vnet.ibm.com>
Reviewed-by: Bjoern Walk <bwalk@linux.vnet.ibm.com>
Implement qemuMonitorRegister() as there is already a
qemuMonitorUnregister() function. This way it may be easier to
understand the code paths.
Signed-off-by: Marc Hartmayer <mhartmay@linux.vnet.ibm.com>
Reviewed-by: Bjoern Walk <bwalk@linux.vnet.ibm.com>
This way qemuDomainLogContextRef() and qemuDomainLogContextFree() is
no longer needed. The naming qemuDomainLogContextFree() was also
somewhat misleading. Additionally, it's easier to turn
qemuDomainLogContext in a self-locking object.
Signed-off-by: Marc Hartmayer <mhartmay@linux.vnet.ibm.com>
Reviewed-by: Bjoern Walk <bwalk@linux.vnet.ibm.com>
There were multiple race conditions that could lead to segmentation
faults. The first precondition for this is qemuProcessLaunch must fail
sometime shortly after starting the new QEMU process. The second
precondition for the segmentation faults is that the new QEMU process
dies - or to be more precise the QEMU monitor has to be closed
irregularly. If both happens during qemuProcessStart (starting a
domain) there are race windows between the thread with the event
loop (T1) and the thread that is starting the domain (T2).
First segmentation fault scenario:
If qemuProcessLaunch fails during qemuProcessStart the code branches
to the 'stop' path where 'qemuMonitorSetDomainLog(priv->mon, NULL,
NULL, NULL)' will set the log function of the monitor to NULL (done in
T2). In the meantime the event loop of T1 will wake up with an EOF
event for the QEMU monitor because the QEMU process has died. The
crash occurs if T1 has checked 'mon->logFunc != NULL' in qemuMonitorIO
just before the logFunc was set to NULL by T2. If this situation
occurs T1 will try to call mon->logFunc which leads to the
segmentation fault.
Solution:
Require the monitor lock for setting the log function.
Backtrace:
0 0x0000000000000000 in ?? ()
1 0x000003ffe9e45316 in qemuMonitorIO (watch=<optimized out>,
fd=<optimized out>, events=<optimized out>, opaque=0x3ffe08aa860) at
../../src/qemu/qemu_monitor.c:727
2 0x000003fffda2e1a4 in virEventPollDispatchHandles (nfds=<optimized
out>, fds=0x2aa000fd980) at ../../src/util/vireventpoll.c:508
3 0x000003fffda2e398 in virEventPollRunOnce () at
../../src/util/vireventpoll.c:657
4 0x000003fffda2ca10 in virEventRunDefaultImpl () at
../../src/util/virevent.c:314
5 0x000003fffdba9366 in virNetDaemonRun (dmn=0x2aa000cc550) at
../../src/rpc/virnetdaemon.c:818
6 0x000002aa00024668 in main (argc=<optimized out>, argv=<optimized
out>) at ../../daemon/libvirtd.c:1541
Second segmentation fault scenario:
If qemuProcessLaunch fails it will unref the log context and with
invoking qemuMonitorSetDomainLog(priv->mon, NULL, NULL, NULL)
qemuDomainLogContextFree() will be invoked. qemuDomainLogContextFree()
invokes virNetClientClose() to close the client and cleans everything
up (including unref of _virLogManager.client) when virNetClientClose()
returns. When T1 is now trying to report 'qemu unexpectedly closed the
monitor' libvirtd will crash because the client has already been
freed.
Solution:
As the critical section in qemuMonitorIO is protected with the monitor
lock we can use the same solution as proposed for the first
segmentation fault.
Backtrace:
0 virClassIsDerivedFrom (klass=0x3100979797979797,
parent=0x2aa000d92f0) at ../../src/util/virobject.c:169
1 0x000003fffda659e6 in virObjectIsClass (anyobj=<optimized out>,
klass=<optimized out>) at ../../src/util/virobject.c:365
2 0x000003fffda65a24 in virObjectLock (anyobj=0x3ffe08c1db0) at
../../src/util/virobject.c:317
3 0x000003fffdba4688 in
virNetClientIOEventLoop (client=client@entry=0x3ffe08c1db0,
thiscall=thiscall@entry=0x2aa000fbfa0) at
../../src/rpc/virnetclient.c:1668
4 0x000003fffdba4b4c in
virNetClientIO (client=client@entry=0x3ffe08c1db0,
thiscall=0x2aa000fbfa0) at ../../src/rpc/virnetclient.c:1944
5 0x000003fffdba4d42 in
virNetClientSendInternal (client=client@entry=0x3ffe08c1db0,
msg=msg@entry=0x2aa000cc710, expectReply=expectReply@entry=true,
nonBlock=nonBlock@entry=false) at ../../src/rpc/virnetclient.c:2116
6 0x000003fffdba6268 in
virNetClientSendWithReply (client=0x3ffe08c1db0, msg=0x2aa000cc710) at
../../src/rpc/virnetclient.c:2144
7 0x000003fffdba6e8e in virNetClientProgramCall (prog=0x3ffe08c1120,
client=<optimized out>, serial=<optimized out>, proc=<optimized out>,
noutfds=<optimized out>, outfds=0x0, ninfds=0x0, infds=0x0,
args_filter=0x3fffdb64440
<xdr_virLogManagerProtocolDomainReadLogFileArgs>, args=0x3ffffffe010,
ret_filter=0x3fffdb644c0
<xdr_virLogManagerProtocolDomainReadLogFileRet>, ret=0x3ffffffe008) at
../../src/rpc/virnetclientprogram.c:329
8 0x000003fffdb64042 in
virLogManagerDomainReadLogFile (mgr=<optimized out>, path=<optimized
out>, inode=<optimized out>, offset=<optimized out>, maxlen=<optimized
out>, flags=0) at ../../src/logging/log_manager.c:272
9 0x000003ffe9e0315c in qemuDomainLogContextRead (ctxt=0x3ffe08c2980,
msg=0x3ffffffe1c0) at ../../src/qemu/qemu_domain.c:4422
10 0x000003ffe9e280a8 in qemuProcessReadLog (logCtxt=<optimized out>,
msg=msg@entry=0x3ffffffe288) at ../../src/qemu/qemu_process.c:1800
11 0x000003ffe9e28206 in qemuProcessReportLogError (logCtxt=<optimized
out>, msgprefix=0x3ffe9ec276a "qemu unexpectedly closed the monitor")
at ../../src/qemu/qemu_process.c:1836
12 0x000003ffe9e28306 in
qemuProcessMonitorReportLogError (mon=mon@entry=0x3ffe085cf10,
msg=<optimized out>, opaque=<optimized out>) at
../../src/qemu/qemu_process.c:1856
13 0x000003ffe9e452b6 in qemuMonitorIO (watch=<optimized out>,
fd=<optimized out>, events=<optimized out>, opaque=0x3ffe085cf10) at
../../src/qemu/qemu_monitor.c:726
14 0x000003fffda2e1a4 in virEventPollDispatchHandles (nfds=<optimized
out>, fds=0x2aa000fd980) at ../../src/util/vireventpoll.c:508
15 0x000003fffda2e398 in virEventPollRunOnce () at
../../src/util/vireventpoll.c:657
16 0x000003fffda2ca10 in virEventRunDefaultImpl () at
../../src/util/virevent.c:314
17 0x000003fffdba9366 in virNetDaemonRun (dmn=0x2aa000cc550) at
../../src/rpc/virnetdaemon.c:818
18 0x000002aa00024668 in main (argc=<optimized out>, argv=<optimized
out>) at ../../daemon/libvirtd.c:1541
Other code parts where the same problem was possible to occur are
fixed as well (qemuMigrationFinish, qemuProcessStart, and
qemuDomainSaveImageStartVM).
Signed-off-by: Marc Hartmayer <mhartmay@linux.vnet.ibm.com>
Reported-by: Sascha Silbe <silbe@linux.vnet.ibm.com>
So far only QEMU_MONITOR_MIGRATION_CAPS_POSTCOPY was reset, but only in
a single code path leaving post-copy enabled in quite a few cases.
https://bugzilla.redhat.com/show_bug.cgi?id=1425003
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
It's only called from qemuMigrationReset now so it doesn't need to be
exported and {tls,sec}Alias are always NULL.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
This new API is supposed to reset all migration parameters to make sure
future migrations won't accidentally use them. This patch makes the
first step and moves qemuMigrationResetTLS call inside
qemuMigrationReset.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Migration parameters are either reset by the main migration code path or
from qemuProcessRecoverMigration* in case libvirtd is restarted during
migration.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Finished qemuMigrationRun does not mean the migration itself finished
(it might have just switched to post-copy mode). While resetting TLS
parameters is probably OK at this point even if migration is still
running, we want to consolidate the code which resets various migration
parameters. Thus qemuMigrationResetTLS will be called from the Confirm
phase (or at the end of the Perform phase in case of v2 protocol), when
migration is either canceled or finished.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
qemuProcessRecoverMigrationOut doesn't explicitly call
qemuMigrationResetTLS relying on two things:
- qemuMigrationCancel resets TLS parameters
- our migration code resets TLS before entering
QEMU_MIGRATION_PHASE_PERFORM3_DONE phase
But this is not obvious and the assumptions will be broken soon. Let's
explicitly reset TLS parameters on all paths which do not kill the
domain.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
There is no async job running when a freshly started libvirtd is trying
to recover from an interrupted incoming migration. While at it, let's
call qemuMigrationResetTLS every time we don't kill the domain. This is
not strictly necessary since TLS is not supported when v2 migration
protocol is used, but doing so makes more sense.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
We will need to store two more host CPU models and nested structs look
better than separate items with long complicated names.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
qemuProcessVerifyHypervFeatures is supposed to check whether all
requested hyperv features were actually honored by QEMU/KVM. This is
done by checking the corresponding CPUID bits reported by the virtual
CPU. In other words, it doesn't work for string properties, such as
VIR_DOMAIN_HYPERV_VENDOR_ID (there is no CPUID bit we could check). We
could theoretically check all 96 bits corresponding to the vendor
string, but luckily we don't have to check the feature at all. If QEMU
is too old to support hyperv features, the domain won't even start.
Otherwise, it is always supported.
Without this patch, libvirt refuses to start a domain which contains
<features>
<hyperv>
<vendor_id state='on' value='...'/>
</hyperv>
</features>
reporting internal error: "unknown CPU feature __kvm_hv_vendor_id.
This regression was introduced by commit v3.1.0-186-ge9dbe7011, which
(by fixing the virCPUDataCheckFeature condition in
qemuProcessVerifyHypervFeatures) revealed an old bug in the feature
verification code. It's been there ever since the verification was
implemented by commit v1.3.3-rc1-5-g95bbe4bf5, which effectively did not
check VIR_DOMAIN_HYPERV_VENDOR_ID at all.
https://bugzilla.redhat.com/show_bug.cgi?id=1439424
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
This header file has been created so that we can expose
internal functions to the test suite without making them
public: those in qemu_capabilities.h bearing the comment
/* Only for use by test suite */
are obvious candidates for being moved over.
Buggy condition meant that vcpu0 would not be iterated in the checks.
Since it's not hotpluggable anyways we would not be able to break the
configuration of a live VM.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1437013
Like all devices, add the 'id' option for mdevs as well. Patch also
adjusts the test accordingly.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1438431
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Depending on the architecture, requirements for ACPI and UEFI can
be different; more specifically, while on x86 UEFI requires ACPI,
on aarch64 it's the other way around.
Enforce these requirements when validating the domain, and make
the error message more accurate by mentioning that they're not
necessarily applicable to all architectures.
Several aarch64 test cases had to be tweaked because they would
have failed the validation step otherwise.
The capabilities used in test cases should match those used
during normal operation for the tests to make any sense.
This results in the generated command line for a few test
cases (most notably non-x86 test cases that were wrongly
assuming they could use -no-acpi) changing.
Instead of having a single function that probes the
architecture from the monitor and then sets a bunch of
basic capabilities based on it, have a separate function
for each part: virQEMUCapsInitQMPArch() only sets the
architecture, and virQEMUCapsInitQMPBasicArch() only sets
the capabilities.
This split will be useful later on, when we will want to
set basic capabilities from the test suite without having
to go through the pain of mocking the monitor.
Currently, if we want to zero out disk source (e,g, due to
startupPolicy when starting up a domain) we use
virDomainDiskSetSource(disk, NULL). This works well for file
based storage (storage type file, dir, or block). But it doesn't
work at all for other types like volume and network.
So imagine that you have a domain that has a CDROM configured
which source is a volume from an inactive pool. Because it is
startupPolicy='optional', the CDROM is empty when the domain
starts. However, the source element is not cleared out in the
status XML and thus when the daemon restarts and tries to
reconnect to the domain it refreshes the disks (which fails - the
storage pool is still not running) and thus the domain is killed.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
So far our code is full of the following pattern:
dom = virGetDomain(conn, name, uuid)
if (dom)
dom->id = 42;
There is no reasong why it couldn't be just:
dom = virGetDomain(conn, name, uuid, id);
After all, client domain representation consists of tuple (name,
uuid, id).
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
In 9e2465834 a check that denies internal snapshots when pflash
based loader is configured for the domain. However, if there's
none and an user tries to do an internal snapshot they will
witness daemon crash as in that case vm->def->os.loader is NULL
and we dereference it unconditionally.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
CPU features which change their value from disabled to enabled between
two calls to query-cpu-model-expansion (the first with no extra
properties set and the second with 'migratable' property set to false)
can be marked as enabled and non-migratable in qemuMonitorCPUModelInfo.
Since the code consuming qemuMonitorCPUModelInfo currently ignores the
migratable flag, this change is effectively changing the CPU model
advertised in domain capabilities to contain all features (even those
which block migration). And this matches what we do for QEMU older than
2.9.0, when we detect all CPUID bits ourselves without asking QEMU.
As a result of this change
<cpu mode='host-model'>
<feature name='invtsc' policy='require'/>
</cpu>
will work with all QEMU versions. Such CPU definition would be forbidden
with QEMU >= 2.9.0 without this patch.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
If calling query-cpu-model-expansion on the 'host'/'max' CPU model with
'migratable' property set to false succeeds, we know QEMU is able to
tell us which features would disable migration. Thus we can mark all
enabled features as migratable.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
QEMU is able to tell us whether a CPU feature would block migration or
not. This patch adds support for storing such features in
qemuMonitorCPUModelInfo.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
When idx is 0 virStorageFileChainLookup returns the base (bottom) of the
backing chain rather than the top. This is expected by the callers of
qemuDomainGetStorageSourceByDevstr.
Add a special case for idx == 0
One of the problems with our virGetDomain function is that it
copies just domain name and domain UUID. Therefore it's very
easy to forget aboud domain ID. This can cause some bugs, like
virConnectGetAllDomainStats not reporting proper domain IDs.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
For guests that use <memoryBacking><locked>, our only option
is to remove the memory locking limit altogether.
Partially-resolves: https://bugzilla.redhat.com/1431793
Instead of having a separate function, we can simply return
zero from the existing qemuDomainGetMemLockLimitBytes() to
signal the caller that the memory locking limit doesn't need
to be set for the guest.
Having a single function instead of two makes it less likely
that we will use the wrong value, which is exactly what
happened when we started applying the limit that was meant
for VFIO-using guests to <memoryBacking><locked>-using
guests.
This reverts commit c2e60ad0e5.
Turns out this check is excessively strict: there are ways
other than <memtune><hard_limit> to raise the memory locking
limit for QEMU processes, one prominent example being
tweaking /etc/security/limits.conf.
Partially-resolves: https://bugzilla.redhat.com/1431793
Creating a copy of the definition we want to add in a migration cookie
makes the code cleaner and less prone to memory leaks or double free
errors.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
QEMU allows for TSC frequency to be explicitly set to enable migration
with invtsc (migration fails if the destination QEMU cannot set the
exact same frequency used when starting the domain on the source host).
Libvirt already supports setting the TSC frequency in the XML using
<clock>
<timer name='tsc' frequency='1234567890'/>
</clock>
which will be transformed into
-cpu Model,tsc-frequency=1234567890
QEMU command line.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The hyperv panic notifier reports additional data in form of 5 registers
that are reported in the crash event from qemu. Log them into the VM log
file and report them as a warning so that admins can see the cause of
crash of their windows VMs.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1426176
For certain kinds of panic notifiers (notably hyper-v) qemu is able to
report some data regarding the crash passed from the guest.
Make the data accessible to the callback in qemu so that it can be
processed further.
Format the mediated devices on the qemu command line as
-device vfio-pci,sysfsdev='/path/to/device/in/syfs'.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
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@redhat.com>
As goes for all the other hostdev device types, grant the qemu process
access to /dev/vfio/<mediated_device_iommu_group>.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Keep track of the assigned mediated devices the same way we do it for
the rest of hostdevs. Methods like 'Prepare', 'Update', and 'ReAttach'
are introduced by this patch.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
So far, the official support is for x86_64 arch guests so unless a
different device API than vfio-pci is available let's only turn on
support for PCI address assignment. Once a different device API is
introduced, we can enable another address type easily.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
A mediated device will be identified by a UUID (with 'model' now being
a mandatory <hostdev> attribute to represent the mediated device API) of
the user pre-created mediated device. We also need to make sure that if
user explicitly provides a guest address for a mdev device, the address
type will be matching the device API supported on that specific mediated
device and error out with an incorrect XML message.
The resulting device XML:
<devices>
<hostdev mode='subsystem' type='mdev' model='vfio-pci'>
<source>
<address uuid='c2177883-f1bb-47f0-914d-32a22e3a8804'>
</source>
</hostdev>
</devices>
Signed-off-by: Erik Skultety <eskultet@redhat.com>
This way more drivers can utilize the functionality without copying
the code. And we can therefore test it in one place for all of them.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
That file has only two exported files and each one of them has
different naming. virNode is what all the other files use, so let's
use it. It wasn't used before because the clash with public API
naming, so let's fix that by shortening the name (there is no other
private variant of it anyway).
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
There is no "node driver" as there was before, drivers have to do
their own ACL checking anyway, so they all specify their functions and
nodeinfo is basically just extending conf/capablities. Hence moving
the code to src/conf/ is the right way to go.
Also that way we can de-duplicate some code that is in virsysfs and/or
virhostcpu that got duplicated during the virhostcpu.c split. And
Some cleanup is done throughout the changes, like adding the vir*
prefix etc.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
There is no reason for it not to be in the utils, all global symbols
under that file already have prefix vir* and there is no reason for it
to be part of DRIVER_SOURCES because that is just a leftover from
older days (pre-driver modules era, I believe).
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Both QEMU and bhyve are using the same function for setting up the CPU
in virCapabilities, so de-duplicate it, save code and time, and help
other drivers adopt it.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Management tools may want to check whether the threshold is still set if
they missed an event. Add the data to the bulk stats API where they can
also query the current backing size at the same time.
To allow updating stats based on the node name, add a helper function
that will fetch the required data from 'query-named-block-nodes' and
return it in hash table for easy lookup.
Detect the node names when setting block threshold and when reconnecting
or when they are cleared when a block job finishes. This operation will
become a no-op once we fully support node names.
To allow matching the node names gathered via 'query-named-block-nodes'
we need to query and then use the top level nodes from 'query-block'.
Add the data to the structure returned by qemuMonitorGetBlockInfo.
qemu for some time already sets node names automatically for the block
nodes. This patch adds code that attempts a best-effort detection of the
node names for the backing chain from the output of
'query-named-block-nodes'. The only drawback is that the data provided
by qemu needs to be matched by the filename as seen by qemu and thus
if two disks share a single backing store file the detection won't work.
This will allow us to use qemu commands such as
'block-set-write-threshold' which only accepts node names.
In this patch only the detection code is added, it will be used later.
Add monitor tooling for calling query-named-block-nodes. The monitor
returns the data as the raw JSON array that is returned from the
monitor.
Unfortunately the logic to extract the node names for a complete backing
chain will be so complex that I won't be able to extract any meaningful
subset of the data in the monitor code.
The code is currently simple, but if we later add node names, it will be
necessary to generate the names based on the node name. Add a helper so
that there's a central point to fix once we add self-generated node
names.
The event is fired when a given block backend node (identified by the
node name) experiences a write beyond the bound set via
block-set-write-threshold QMP command. This wires up the monitor code to
extract the data and allow us receiving the events and the capability.
qemuMigrationResetTLS() does not initialize 'ret' by default,
so when it jumps to 'cleanup' on error, the 'ret' variable will be
uninitialized, which clang complains about.
Set it to '-1' by default.
https://bugzilla.redhat.com/show_bug.cgi?id=1300769
If the migration flags indicate this migration will be using TLS,
then while we have connection in the Begin phase check and setup the
TLS environment that will be used by virMigrationRun during the Perform
phase for the source to configure TLS.
Processing adds an "-object tls-creds-x509,endpoint=client,..." and
possibly an "-object secret,..." to handle the passphrase response.
Then it sets the 'tls-creds' and possibly 'tls-hostname' migration
parameters.
The qemuMigrateCancel will clean up and reset the environment as it
was originally found.
Signed-off-by: John Ferlan <jferlan@redhat.com>
If the migration flags indicate this migration will be using TLS,
then set up the destination during the prepare phase once the target
domain has been started to add the TLS objects to perform the migration.
This will create at least an "-object tls-creds-x509,endpoint=server,..."
for TLS credentials and potentially an "-object secret,..." to handle the
passphrase response to access the TLS credentials. The alias/id used for
the TLS objects will contain "libvirt_migrate".
Once the objects are created, the code will set the "tls-creds" and
"tls-hostname" migration parameters to signify usage of TLS.
During the Finish phase we'll be sure to attempt to clear the
migration parameters and delete those objects (whether or not they
were created). We'll also perform the same reset during recovery
if we've reached FINISH3.
If the migration isn't using TLS, then be sure to check if the
migration parameters exist and clear them if so.
Add an asyncJob argument for add/delete TLS Objects. A future patch will
add/delete TLS objects from a migration which may have a job to join.
Signed-off-by: John Ferlan <jferlan@redhat.com>
Add the fields to support setting tls-creds and tls-hostname during
a migration (either source or target). Modify the query migration
function to check for the presence and set the field for future
consumers to determine which of 3 conditions is being met (NULL,
present and set to "", or present and sent to something). These
correspond to qemu commit id '4af245dc3' which added support to
default the value to "" and allow setting (or resetting) to ""
in order to disable. This reset option allows libvirt to properly
use the tls-creds and tls-hostname parameters.
Modify code paths that either allocate or use stack space in order
to call qemuMigrationParamsClear or qemuMigrationParamsFree for cleanup.
Signed-off-by: John Ferlan <jferlan@redhat.com>
Add a new TLS X.509 certificate type - "migrate". This will handle the
creation of a TLS certificate capability (and possibly repository) to
be used for migrations. Similar to chardev's, credentials will be handled
via a libvirt secrets; however, unlike chardev's enablement and usage
will be via a CLI flag instead of a conf flag and a domain XML attribute.
The migrations using the *x509_verify flag require the client-cert.pem
and client-key.pem files to be present in the TLS directory - so let's
also be sure to note that in the qemu.conf file.
Signed-off-by: John Ferlan <jferlan@redhat.com>
If the variable store (<nvram>) file is raw qemu can't do a snapshot of
it and thus the snapshot fails. QEMU rejects such snapshot by a message
which would not be properly interpreted as an error by libvirt.
Additionally allowing to use a qcow2 variable store backing file would
solve this issue but then it would become eligible to become target of
the memory dump.
Offline internal snapshot would be incomplete too with either storage
format since libvirt does not handle the pflash file in this case.
Forbid such snapshot so that we can avoid problems.
The 'mon' argument validity is checked in the QEMU_CHECK_MONITOR for the
following functions, so they don't need the NONNULL on their prototype:
qemuMonitorUpdateVideoMemorySize
qemuMonitorUpdateVideoVram64Size
qemuMonitorGetAllBlockStatsInfo
qemuMonitorBlockStatsUpdateCapacity
Signed-off-by: John Ferlan <jferlan@redhat.com>
The prototype requires not passing a NULL in the parameter and the callers
all would fail far before this code would fail if 'vm' was NULL, so just
remove the check.
Signed-off-by: John Ferlan <jferlan@redhat.com>
Commit id '85af0b8' added a 'timeout' as the 4th parameter to
qemuMonitorOpen, but neglected to update the ATTRIBUTE_NONNULL(4)
to be (5) for the cb parameter.
We reported error in caller virQEMUCapsCacheLookupByArch.
So the same error messages in qemuConnectGetDomainCapabilities
is useless.
Signed-off-by: Chen Hanxiao <chenhanxiao@gmail.com>
Calling virCPUUpdateLive on a domain with no guest CPU configuration
does not make sense. Especially when doing so would crash libvirtd.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
When starting a domain with custom guest CPU specification QEMU may add
or remove some CPU features. There are several reasons for this, e.g.,
QEMU/KVM does not support some requested features or the definition of
the requested CPU model in libvirt's cpu_map.xml differs from the one
QEMU is using. We can't really avoid this because CPU models are allowed
to change with machine types and libvirt doesn't know (and probably
doesn't even want to know) about such changes.
Thus when we want to make sure guest ABI doesn't change when a domain
gets migrated to another host, we need to update our live CPU definition
according to the CPU QEMU created. Once updated, we will change CPU
checking to VIR_CPU_CHECK_FULL to make sure the virtual CPU created
after migration exactly matches the one on the source.
https://bugzilla.redhat.com/show_bug.cgi?id=822148https://bugzilla.redhat.com/show_bug.cgi?id=824989
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
qemuMonitorGetGuestCPU can now optionally create CPU data from
filtered-features in addition to feature-words.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The checks are now in a dedicated qemuProcessVerifyHypervFeatures
function.
In addition to moving the code this patch also fixes a few bugs: the
original code was leaking cpuFeature and the return value of
virCPUDataCheckFeature was not checked properly.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The disk tuning group parameter is ignored by qemu if no other
throttling options are set. Reject such configuration, since the name
would not be honored after setting parameters via the live tuning API.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1433180
When checking capabilities for qemu we need to check whether subsets of
the disk throttling settings are supported. Extract the checks into a
separate functions as they will be reused in next patch.
While the code path that queries the monitor allocates a separate copy
of the 'group_name' string the path querying the config would not copy
it. The call to virTypedParameterAssign would then steal the pointer
(without clearing it) and the RPC layer freed it. Any subsequent call
resulted into a crash.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1433183
ioh3420 is emulated Intel hardware, so it always looked
quite out of place in aarch64/virt guests. Even for x86/q35
guests, the recently-introduced pcie-root-port is a better
choice because, unlike ioh3420, it doesn't require IO space
(a fairly constrained resource) to work.
If pcie-root-port is available in QEMU, use it; ioh3420 is
still used as fallback for when pcie-root-port is not
available.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1408808
QEMU 2.9 introduces the pcie-root-port device, which is
a generic version of the existing ioh3420 device.
Make the new device available to libvirt users.
There were couple of reports on the list (e.g. [1]) that guests
with huge amounts of RAM are unable to start because libvirt
kills qemu in the initialization phase. The problem is that if
guest is configured to use hugepages kernel has to zero them all
out before handing over to qemu process. For instance, 402GiB
worth of 1GiB pages took around 105 seconds (~3.8GiB/s). Since we
do not want to make the timeout for connecting to monitor
configurable, we have to teach libvirt to count with this
fact. This commit implements "1s per each 1GiB of RAM" approach
as suggested here [2].
1: https://www.redhat.com/archives/libvir-list/2017-March/msg00373.html
2: https://www.redhat.com/archives/libvir-list/2017-March/msg00405.html
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Some users might want to pass a blockdev or a chardev as a
backend for NVDIMM. In fact, this is expected to be the mostly
used configuration. Therefore libvirt should allow the device in
devices CGroup then.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Now that we have APIs for relabel memdevs on hotplug, fill in the
missing implementation in qemu hotplug code.
The qemuSecurity wrappers might look like overkill for now,
because qemu namespace code does not deal with the nvdimms yet.
Nor does our cgroup code. But hey, there's cgroup_device_acl
variable in qemu.conf. If users add their /dev/pmem* device in
there, the device is allowed in cgroups and created in the
namespace so they can successfully passthrough it to the domain.
It doesn't look like overkill after all, does it?
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
For NVDIMM devices it is optionally possible to specify the size
of internal storage for namespaces. Namespaces are a feature that
allows users to partition the NVDIMM for different uses.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
So, majority of the code is just ready as-is. Well, with one
slight change: differentiate between dimm and nvdimm in places
like device alias generation, generating the command line and so
on.
Speaking of the command line, we also need to append 'nvdimm=on'
to the '-machine' argument so that the nvdimm feature is
advertised in the ACPI tables properly.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
NVDIMM is new type of memory introduced into QEMU 2.6. The idea
is that we have a Non-Volatile memory module that keeps the data
persistent across domain reboots.
At the domain XML level, we already have some representation of
'dimm' modules. Long story short, NVDIMM will utilize the
existing <memory/> element that lives under <devices/> by adding
a new attribute 'nvdimm' to the existing @model and introduce a
new <path/> element for <source/> while reusing other fields. The
resulting XML would appear as:
<memory model='nvdimm'>
<source>
<path>/tmp/nvdimm</path>
</source>
<target>
<size unit='KiB'>523264</size>
<node>0</node>
</target>
<address type='dimm' slot='0'/>
</memory>
So far, this is just a XML parser/formatter extension. QEMU
driver implementation is in the next commit.
For more info on NVDIMM visit the following web page:
http://pmem.io/
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Frankly, this function is one big mess. A lot of arguments,
complicated behaviour. It's really surprising that arguments were
in random order (input and output arguments were mixed together),
the documentation was outdated, the description of return values
was bogus.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Even though this variable contains just values from an enum where
zero has the usual meaning, it's enum after all and we should
check it as such.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
One of the main reasons for introducing host-model CPU definition in a
domain capabilities XML was the inability to express disabled features
in a host capabilities XML. That is, when a host CPU is, e.g., Haswell
without x2apic support, host capabilities XML will have to report it as
Westmere + a bunch of additional features., but we really want to use
Haswell - x2apic when creating a host-model CPU.
Unfortunately, I somehow forgot to do the last step and the code would
just copy the CPU definition found in the host capabilities XML. This
changed recently for new QEMU versions which allow us to query host CPU,
but any slightly older QEMU will not benefit from any change I did. This
patch makes sure the right CPU model is filled in the domain
capabilities even with old QEMU.
The issue was reported in
https://bugzilla.redhat.com/show_bug.cgi?id=1426456
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The function is now called virQEMUCapsProbeHostCPU. Both the refactoring
and the change of the name is done for consistency with a new function
which will be introduced in the following commit.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
When creating host CPU definition usable with a given emulator, the CPU
should not be defined using an unsupported CPU model. The new @models
and @nmodels parameters can be used to limit CPU models which can be
used in the result.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The parameter can be used to request either VIR_CPU_TYPE_HOST (which has
been assumed so far) or VIR_CPU_TYPE_GUEST definition.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
cpuNodeData has always been followed by cpuDecode as no hypervisor
driver is really interested in raw CPUID data for a host CPU. Let's
create a new CPU driver API which returns virCPUDefPtr directly.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1431112
Yeah, that's right. A mount point doesn't have to be a directory.
It can be a file too. However, the code that tries to preserve
mount points under /dev for new namespace for qemu does not count
with that option.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
bhyve supports 'gop' video device that allows clients to connect
to VMs using VNC clients. This commit adds support for that to
the bhyve driver:
- Introducr 'gop' video device type
- Add capabilities probing for the 'fbuf' device that's
responsible for graphics
- Update command builder routines to let users configure
domain's VNC via gop graphics.
Signed-off-by: Roman Bogorodskiy <bogorodskiy@gmail.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1430634
If a qemu process has died, we get EOF on its monitor. At this
point, since qemu process was the only one running in the
namespace kernel has already cleaned the namespace up. Any
attempt of ours to enter it has to fail.
This really happened in the bug linked above. We've tried to
attach a disk to qemu and while we were in the monitor talking to
qemu it just died. Therefore our code tried to do some roll back
(e.g. deny the device in cgroups again, restore labels, etc.).
However, during the roll back (esp. when restoring labels) we
still thought that domain has a namespace. So we used secdriver's
transactions. This failed as there is no namespace to enter.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
If the delivery of the DEVICE_DELETED event for the vCPU being deleted
would time out, the code would not call 'qemuDomainResetDeviceRemoval'.
Since the waiting thread did not unregister itself prior to stopping the
waiting the monitor code would try to wake it up instead of dispatching
it to the event worker. As a result the unplug process would not be
completed and the definition would not be updated.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1428893https://bugzilla.redhat.com/show_bug.cgi?id=1427801
This reverts commit c96bd78e4e.
So our code is one big mess and we modify domain definition while
building qemu_command line and our hotplug code share only part
of the parsing and command line building code. Let's revert
that change because to fix it properly would require refactor and
move a lot of things.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1430275
When libvirtd is started we call qemuDomainRecheckInternalPaths
to detect whether a domain has VNC socket path generated by libvirt
based on option from qemu.conf. However if we are parsing status XML
for running domain the existing socket path can be generated also if
the config XML uses the new <listen type='socket'/> element without
specifying any socket.
The current code doesn't make difference how the socket was generated
and always marks it as "fromConfig". We need to store the
"autoGenerated" value in the status XML in order to preserve that
information.
The difference between "fromConfig" and "autoGenerated" is important
for migration, because if the socket is based on "fromConfig" we don't
print it into the migratable XML and we assume that user has properly
configured qemu.conf on both hosts. However if the socket is based
on "autoGenerated" it means that a new feature was used and therefore
we need to leave the socket in migratable XML to make sure that if
this feature is not supported on destination the migration will fail.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Split apart and rename qemuDomainGetChardevTLSObjects in order to make a
more generic API that can create the TLS JSON prop objects (secret and
tls-creds-x509) to be used to create the objects
Signed-off-by: John Ferlan <jferlan@redhat.com>
Create a qemuDomainAddChardevTLSObjects which will encapsulate the
qemuDomainGetChardevTLSObjects and qemuDomainAddTLSObjects so that
the callers don't need to worry about the props.
Move the dev->type and haveTLS checks in to the Add function to avoid
an unnecessary call to qemuDomainAddTLSObjects
Signed-off-by: John Ferlan <jferlan@redhat.com>
Refactor the TLS object adding code to make two separate API's that will
handle the add/remove of the "secret" and "tls-creds-x509" objects including
the Enter/Exit monitor commands.
Signed-off-by: John Ferlan <jferlan@redhat.com>
Since qemuDomainObjExitMonitor can also generate error messages,
let's move it inside any error message saving code on error paths
for various hotplug add activities.
Signed-off-by: John Ferlan <jferlan@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1379200
When we are restoring a domain from a saved image, or just
updating its XML in the saved image - we have to make sure that
the ABI guests sees will not change. We have a function for that
which reports errors. But for some reason if this function fails,
we call it again with slightly different argument. Therefore it
might happen that we overwrite the original error and leave user
with less helpful one.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Now that we have some qemuSecurity wrappers over
virSecurityManager APIs, lets make sure everybody sticks with
them. We have them for a reason and calling virSecurityManager
API directly instead of wrapper may lead into accidentally
labelling a file on the host instead of namespace.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
The implementation matches virStringListFreeCount. The only difference
between the two functions is the ordering of their parameters.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The static CPU model expansion is designed to return only canonical
names of all CPU properties. To maintain backwards compatibility libvirt
is stuck with different spelling of some of the features, but we need to
use the full expansion to get the additional spellings. In addition to
returning all spelling variants for all properties the full expansion
will contain properties which are not guaranteed to be migration
compatible. Thus, we need to combine both expansions. First we need to
call the static expansion to limit the result to migratable properties.
Then we can use the result of the static expansion as an input to the
full expansion to get both canonical names and their aliases.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Querying "host" CPU model expansion only makes sense for KVM. QEMU 2.9.0
introduces a new "max" CPU model which can be used to ask QEMU what the
best CPU it can provide to a TCG domain is.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
While query-cpu-model-expansion returns only boolean features on s390,
but x86_64 reports some integer and string properties which we are
interested in.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
virQEMUCapsHasPCIMultiBus() performs a version check on
the QEMU binary to figure out whether multiple buses are
supported, so to get the correct aliases assigned when
dealing with pSeries guests we need to spoof the version
accordingly in the test suite.
Due to the extra architecture-specific logic, it's already
necessary for users to call virQEMUCapsHasPCIMultiBus(),
so the capability itself is just a pointless distraction.
Our documentation states that the chardev logging file is truncated
unless append='on' is specified. QEMU also behaves the same way and
truncates the file unless we provide the argument. The new virlogd
implementation did not honor if the argument was missing and continued
to append to the file.
Truncate the file even when the 'append' attribute is not present to
behave the same with both implementations and adhere to the docs.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1420205
After eca76884ea in case of error in qemuDomainSetPrivatePaths()
in pretended start we jump to stop. I've changed this during
review from 'cleanup' which turned out to be correct. Well, sort
of. We can't call qemuProcessStop() as it decrements
driver->nactive and we did not increment it. However, it calls
virDomainObjRemoveTransientDef() which is basically the only
function we need to call. So call that function and goto cleanup;
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
The CPU driver provides APIs to create and free virCPUDataPtr. Thus all
APIs exported from the driver should work with that rather than
requiring the caller to pass a pointer to an internal part of the
structure.
In other words
virCPUx86DataAddCPUID(cpudata, &cpuid)
is much better than the original
virCPUx86DataAddCPUID(&cpudata->data.x86, &cpuid)
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The new API is called virCPUDataFree. Individual CPU drivers are no
longer required to implement their own freeing function unless they need
to free architecture specific data from virCPUData.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Our documentation of the domain capabilities XML says that the fallback
attribute of a CPU model is used to indicate whether the CPU model was
detected by libvirt itself (fallback="allow") or by asking the
hypervisor (fallback="forbid"). We need to properly set
fallback="forbid" when CPU model comes from QEMU to match the
documentation.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The port is stored in graphics configuration and it will
also get released in qemuProcessStop in case of error.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1397440
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Previously the code called virStorageSourceUpdateBlockPhysicalSize which
did not do anything on empty drives since it worked only on block
devices. After the refactor in c5f6151390 it's called for all devices
and thus attempts to deref the NULL path of empty drives.
Add a check that skips the update of the physical size if the storage
source is empty.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1420718
Fix incorrect jump labels in error paths as the stop jump is only
needed if the driver has already changed the state. For example
'virAtomicIntInc(&driver->nactive)' will be 'reverted' in the
qemuProcessStop call.
Signed-off-by: Marc Hartmayer <mhartmay@linux.vnet.ibm.com>
Reviewed-by: Boris Fiuczynski <fiuczy@linux.vnet.ibm.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
When a domain needs an access to some device (be it a disk, RNG,
chardev, whatever), we have to allow it in the devices CGroup (if
it is available), because by default we disallow all the devices.
But some of the functions that are responsible for setting up
devices CGroup are lacking check whether there is any CGroup
available. Thus users might be unable to hotplug some devices:
virsh # attach-device fedora rng.xml
error: Failed to attach device from rng.xml
error: internal error: Controller 'devices' is not mounted
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
One of the conditions in qemuDomainDeviceCalculatePCIConnectFlags
was missing a break that could result it in falling through to
an incorrect codepath.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
qemuDomainAssignPCIAddresses() hardcoded the assumption
that the only way to support devices on a non-zero bus is
to add one or more pci-bridges; however, since we now
support a large selection of PCI controllers that can be
used instead, the assumption is no longer true.
Moreover, this check was always redundant, because the
only sensible time to check for the availability of
pci-bridge is when building the QEMU command line, and
such a check is of course already in place.
In fact, there were *two* such checks, but since one of
the two was relying on the incorrect assumption explained
above, and it was redundant anyway, it has been dropped.
When switching over the values in the virDomainControllerModelPCI
enumeration, make sure the proper cast is in place so that the
compiler can warn us when the coverage is not exaustive.
For the same reason, fold some unstructured checks (performed by
comparing directly against some values in the enumeration) inside
an existing switch statement.
The CPU model info formating code in virQEMUCapsFormatCache will get
more complicated soon. Separating the code in
virQEMUCapsFormatHostCPUModelInfo will make the result easier to read.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
All features the function is currently supposed to filter out are
specific to x86_64. We should avoid removing them on other
architectures. It seems to be quite unlikely other achitectures would
use the same names, but one can never be sure.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
The functions in virCommand() after fork() must be careful with regard
to accessing any mutexes that may have been locked by other threads in
the parent process. It is possible that another thread in the parent
process holds the lock for the virQEMUDriver while fork() is called.
This leads to a deadlock in the child process when
'virQEMUDriverGetConfig(driver)' is called and therefore the handshake
never completes between the child and the parent process. Ultimately
the virDomainObjectPtr will never be unlocked.
It gets much worse if the other thread of the parent process, that
holds the lock for the virQEMUDriver, tries to lock the already locked
virDomainObject. This leads to a completely unresponsive libvirtd.
It's possible to reproduce this case with calling 'virsh start XXX'
and 'virsh managedsave XXX' in a tight loop for multiple domains.
This commit fixes the deadlock in the same way as it is described in
commit 61b52d2e38.
Signed-off-by: Marc Hartmayer <mhartmay@linux.vnet.ibm.com>
Reviewed-by: Boris Fiuczynski <fiuczy@linux.vnet.ibm.com>
With that users could access files outside /dev/shm. That itself
isn't a security problem, but might cause some errors we want to
avoid. So let's forbid slashes as we do with domain and volume names
and also mention that in the schema.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1395496
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
This follows the same check for disk, because we cannot remove iothread
if it's used by disk or by controller. It could lead to crashing QEMU.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
If virDomainDelIOThread API was called with VIR_DOMAIN_AFFECT_LIVE
and VIR_DOMAIN_AFFECT_CONFIG and both XML were already a different
it could result in removing iothread from config XML even if there
was a disk using that iothread.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
The situation covered by the removed code will not ever happen.
This code is called only while starting a new QEMU process where
the capabilities where already checked and while attaching to
existing QEMU process where we don't even detect the iothreads.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
