https://bugzilla.redhat.com/show_bug.cgi?id=1198645
Once upon a time, there was a little domain. And the domain was pinned
onto a NUMA node and hasn't fully allocated its memory:
<memory unit='KiB'>2355200</memory>
<currentMemory unit='KiB'>1048576</currentMemory>
<numatune>
<memory mode='strict' nodeset='0'/>
</numatune>
Oh little me, said the domain, what will I do with so little memory.
If I only had a few megabytes more. But the old admin noticed the
whimpering, barely audible to untrained human ear. And good admin he
was, he gave the domain yet more memory. But the old NUMA topology
witch forbade to allocate more memory on the node zero. So he
decided to allocate it on a different node:
virsh # numatune little_domain --nodeset 0-1
virsh # setmem little_domain 2355200
The little domain was happy. For a while. Until bad, sharp teeth
shaped creature came. Every process in the system was afraid of him.
The OOM Killer they called him. Oh no, he's after the little domain.
There's no escape.
Do you kids know why? Because when the little domain was born, her
father, Libvirt, called numa_set_membind(). So even if the admin
allowed her to allocate memory from other nodes in the cgroups, the
membind() forbid it.
So what's the lesson? Libvirt should rely on cgroups, whenever
possible and use numa_set_membind() as the last ditch effort.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
The destination libvirt daemon in a migration may segfault if the client
disconnects immediately after the migration has begun:
# virsh -c qemu+tls://remote/system list --all
Id Name State
----------------------------------------------------
...
# timeout --signal KILL 1 \
virsh migrate example qemu+tls://remote/system \
--verbose --compressed --live --auto-converge \
--abort-on-error --unsafe --persistent \
--undefinesource --copy-storage-all --xml example.xml
Killed
# virsh -c qemu+tls://remote/system list --all
error: failed to connect to the hypervisor
error: unable to connect to server at 'remote:16514': Connection refused
The crash is in:
1531 void
1532 qemuDomainObjEndJob(virQEMUDriverPtr driver, virDomainObjPtr obj)
1533 {
1534 qemuDomainObjPrivatePtr priv = obj->privateData;
1535 qemuDomainJob job = priv->job.active;
1536
1537 priv->jobs_queued--;
Backtrace:
#0 at qemuDomainObjEndJob at qemu/qemu_domain.c:1537
#1 in qemuDomainRemoveInactive at qemu/qemu_domain.c:2497
#2 in qemuProcessAutoDestroy at qemu/qemu_process.c:5646
#3 in virCloseCallbacksRun at util/virclosecallbacks.c:350
#4 in qemuConnectClose at qemu/qemu_driver.c:1154
...
qemuDomainRemoveInactive calls virDomainObjListRemove, which in this
case is holding the last remaining reference to the domain.
qemuDomainRemoveInactive then calls qemuDomainObjEndJob, but the domain
object has been freed and poisoned by then.
This patch bumps the domain's refcount until qemuDomainRemoveInactive
has completed. We also ensure qemuProcessAutoDestroy does not return the
domain to virCloseCallbacksRun to be unlocked in this case. There is
similar logic in bhyveProcessAutoDestroy and lxcProcessAutoDestroy
(which call virDomainObjListRemove directly).
Signed-off-by: Michael Chapman <mike@very.puzzling.org>
Instead of always using controller 0 and incrementing port number,
respect the maximum port numbers of controllers and use all of them.
Ports for virtio consoles are quietly reserved, but not formatted
(neither in XML nor on QEMU command line).
Also rejects duplicate virtio-serial addresses.
https://bugzilla.redhat.com/show_bug.cgi?id=890606https://bugzilla.redhat.com/show_bug.cgi?id=1076708
Test changes:
* virtio-auto.args
Filling out the port when just the controller is specified.
switched from using
maxport + 1
to:
first free port on the controller
* virtio-autoassign.args
Filling out the address when no <address> is specified.
Started using all the controllers instead of 0, also discards
the bus value.
* xml -> xml output of virtio-auto
The port assignment is no longer done as a part of XML parsing,
so the unspecified values stay 0.
Two places would call to qemuPrepareCpumap() with priv->autoNodeset to
convert it to a cpuset. Remove the function and use the prepared cpuset
automatically.
When the synchronous pivot option is selected, libvirt would not update
the backing chain until the job was exitted. Some applications then
received invalid data as their job serialized first.
This patch removes polling to wait for the ABORT/PIVOT job completion
and replaces it with a condition. If a synchronous operation is
requested the update of the XML is executed in the job of the caller of
the synchronous request. Otherwise the monitor event callback uses a
separate worker to update the backing chain with a new job.
This is a regression since 1a92c71910
When the ABORT job is finished synchronously you get the following call
stack:
#0 qemuBlockJobEventProcess
#1 qemuDomainBlockJobImpl
#2 qemuDomainBlockJobAbort
#3 virDomainBlockJobAbort
While previously or while using the _ASYNC flag you'd get:
#0 qemuBlockJobEventProcess
#1 processBlockJobEvent
#2 qemuProcessEventHandler
#3 virThreadPoolWorker
When using 'dimm' memory devices with qemu, some of the information
like the slot number and base address need to be reloaded from qemu
after process start so that it reflects the actual state. The state then
allows to use memory devices across migrations.
virnetdevopenvswitch.h declares a few functions that can be called to
add ports to and remove them from OVS bridges, and retrieve the
migration data for a port. It does not contain any data definitions
that are used by domain_conf.h. But for some reason, domain_conf.h
virnetdevopenvswitch.h should be directly #including it. This adds a
few lines to the project, but saves all the files that don't need it
from the extra computing, and makes the dependencies more clear cut.
When libvirt is starting a domain, it reports the state as SHUTOFF until
it's RUNNING. This is not ideal because domain startup may take a long
time (usually because of some configuration issues, firewalls blocking
access to network disks, etc.) and domain lists provided by libvirt look
awkward. One can see weird shutoff domains with IDs in a list of active
domains or even shutoff transient domains. In any case, it looks more
like a bug in libvirt than a normal state a domain goes through.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Use the utilities introduced in the previous patches so the qemu
driver is able to create tap devices that are bound (and unbound
on domain destroyal) to Midonet virtual ports.
Signed-off-by: Antoni Segura Puimedon <toni+libvirt@midokura.com>
We're parsing memballoon status period as unsigned int, but when we're
trying to set it, both we and qemu use signed int. That means large
values will get wrapped around to negative one resulting in error.
Basically the same problem as commit e3a7b874 was dealing with when
updating live domain.
QEMU changed the accepted value to int64 in commit 1f9296b5, but even
values as INT_MAX don't make sense since the value passed means seconds.
Hence adding capability flag for this change isn't worth it.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1140958
Signed-off-by: Luyao Huang <lhuang@redhat.com>
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
As pointed out by jtomko in his review of the IOThreads pinning code:
http://www.redhat.com/archives/libvir-list/2015-March/msg00495.html
there are some comments sprinkled in indicating IOThreads were using
the same structure as the VcpuPin code...
This is the first patch of a few that will change the virDomainVcpuPin*
structures and code to just virDomainPin* - starting with the data
structure naming...
As there are two possible approaches to define a domain's memory size -
one used with legacy, non-NUMA VMs configured in the <memory> element
and per-node based approach on NUMA machines - the user needs to make
sure that both are specified correctly in the NUMA case.
To avoid this burden on the user I'd like to replace the NUMA case with
automatic totaling of the memory size. To achieve this I need to replace
direct access to the virDomainMemtune's 'max_balloon' field with
two separate getters depending on the desired size.
The two sizes are needed as:
1) Startup memory size doesn't include memory modules in some
hypervisors.
2) After startup these count as the usable memory size.
Note that the comments for the functions are future aware and document
state that will be present after a few later patches.
Surprisingly we did not grab a VM job when a block job finished and we'd
happily rewrite the backing chain data. This made it possible to crash
libvirt when queueing two backing chains tightly and other badness.
To fix it, add yet another handler to the helper thread that handles
monitor events that require a job.
https://bugzilla.redhat.com/show_bug.cgi?id=1197600
So, libvirt uses pid file to track pid of started qemus. Whenever
a domain is started, its pid is put into corresponding pid file.
The pid file path is generated based on domain name and stored
into domain object internals. However, it's not stored in the
status XML and therefore lost on daemon restarts. Hence, later,
when domain is being shut down, the daemon does not know which
pid file to unlink, and the correct pid file is left behind. To
avoid this, lets generate the pid file path again in
qemuProcessReconnect().
Reported-by: Luyao Huang <lhuang@redhat.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Instead of checking defaultMode for every channel that has no mode
configured, test it only once outside of channel loop. This fixes a bug
that in case all possible channels are fore example set to insecure, but
defaultMode is set to secure, we wouldn't auto-generate TLS port. This
results in failure while starting a guest.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1143832
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
We have two different places that needs to be updated while touching
code for allocation spice ports. Add a bool option to
'qemuProcessSPICEAllocatePorts' function to switch between true and fake
allocation so we can use this function also in qemu_driver to generate
native domain definition.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Since adding the support for scheduler policy settings in commit
8680ea97, there are two enums with the same information. That was
caused by rewriting the patch since first draft.
Find out thanks to clang, but there was no impact whatsoever.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Upon BLOCK_JOB_COMPLETED event delivery, we check if the job has
completed (in qemuMonitorJSONHandleBlockJobImpl()). For better image,
the event looks something like this:
"timestamp": {"seconds": 1423582694, "microseconds": 372666}, "event":
"BLOCK_JOB_COMPLETED", "data": {"device": "drive-virtio-disk0", "len":
8412790784, "offset": 409993216, "speed": 8796093022207, "type":
"mirror", "error": "No space left on device"}}
If "len" does not equal "offset" it's considered an error, and we can
clearly see "error" field filled in. However, later in the event
processing this case was handled no differently to case of job being
aborted via separate API. It's time that we start differentiate these
two because of the future work.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Currently, upon BLOCK_JOB_* event, disk->mirrorState is not updated
each time. The callback code handling the events checks if a blockjob
was started via our public APIs prior to setting the mirrorState.
However, some block jobs may be started internally (e.g. during
storage migration), in which case we don't bother with setting
disk->mirror (there's nothing we can set it to anyway), or other
fields. But it will come handy if we update the mirrorState in these
cases too. The event wasn't delivered just for fun - we've started the
job after all.
So, in this commit, the mirrorState is set to whatever job status
we've obtained. Of course, there are some actions on some statuses
that we want to perform. But instead of if {} else if {} else {} ...
enumeration, let's move to switch().
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
We do have a check for valid per-domain security model, however we still
do permit an invalid security model for a domain's device (those which
are specified with <source> element).
This patch introduces a new function virSecurityManagerCheckAllLabel
which compares user specified security model against currently
registered security drivers. That being said, it also permits 'none'
being specified as a device security model.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1165485
Signed-off-by: Ján Tomko <jtomko@redhat.com>
If a previous commit I fixed the incorrect handling of vcpu pids
for TCG mode QEMU:
commit b07f3d821d
Author: Daniel P. Berrange <berrange@redhat.com>
Date: Thu Dec 18 16:34:39 2014 +0000
Don't setup fake CPU pids for old QEMU
The code assumes that def->vcpus == nvcpupids, so when we setup
fake CPU pids for old QEMU with nvcpupids == 1, we cause the
later code to read off the end of the array. This has fun results
like sche_setaffinity(0, ...) which changes libvirtd's own CPU
affinity, or even better sched_setaffinity($RANDOM, ...) which
changes the affinity of a random OS process.
The intent was that this would merely disable the ability to set
per-vCPU affinity. It should still have been possible to set VM
level host CPU affinity.
Unfortunately, when you set <vcpu cpuset='0-1'>4</vcpu>, the XML
parser will internally take this & initialize an entry in the
def->cputune.vcpupin array for every VCPU. IOW this is implicitly
being treated as
<cputune>
<vcpupin cpuset='0-1' vcpu='0'/>
<vcpupin cpuset='0-1' vcpu='1'/>
<vcpupin cpuset='0-1' vcpu='2'/>
<vcpupin cpuset='0-1' vcpu='3'/>
</cputune>
Even more fun, the faked cputune elements are hidden from view when
querying the live XML, because their cpuset mask is the same as the
VM default cpumask.
The upshot was that it was impossible to set VM level CPU affinity.
To fix this we must update qemuProcessSetVcpuAffinities so that it
only reports a fatal error if the per-VCPU cpu mask is different
from the VM level cpu mask.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
It is often helpful to know which version of libvirt and QEMU
was present when a guest was first launched. Ensure this info
is written into the QEMU log file for each guest.
Depending on the context, either error out if the domain
has disappeared in the meantime, or just ignore the value
to allow marking the function as ATTRIBUTE_RETURN_CHECK.
https://bugzilla.redhat.com/show_bug.cgi?id=1161024
In the device type-specific functions, exit early
if the domain has disappeared, because the cleanup
should have been done by qemuProcessStop.
Check the return value in processDeviceDeletedEvent
and qemuProcessUpdateDevices.
Skip audit and removing the device from live def because
it has already been cleaned up.
Make local copy of the disk alias in qemuProcessInitPasswords,
instead of referencing the one in domain definition, which
might get freed if the domain crashes while we're in monitor.
Also copy the memballoon period value.
QEMU internally updates the size of video memory if the domain XML had
provided too low memory size or there are some dependencies for a QXL
devices 'vgamem' and 'ram' size. We need to know about the changes and
store them into the status XML to not break migration or managedsave
through different libvirt versions.
The values would be loaded only if the "vgamem_mb" property exists for
the device. The presence of the "vgamem_mb" also tells that the
"ram_size" and "vram_size" exists for QXL devices.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
There is one problem that causes various errors in the daemon. When
domain is waiting for a job, it is unlocked while waiting on the
condition. However, if that domain is for example transient and being
removed in another API (e.g. cancelling incoming migration), it get's
unref'd. If the first call, that was waiting, fails to get the job, it
unref's the domain object, and because it was the last reference, it
causes clearing of the whole domain object. However, when finishing the
call, the domain must be unlocked, but there is no way for the API to
know whether it was cleaned or not (unless there is some ugly temporary
variable, but let's scratch that).
The root cause is that our APIs don't ref the objects they are using and
all use the implicit reference that the object has when it is in the
domain list. That reference can be removed when the API is waiting for
a job. And because each domain doesn't do its ref'ing, it results in
the ugly checking of the return value of virObjectUnref() that we have
everywhere.
This patch changes qemuDomObjFromDomain() to ref the domain (using
virDomainObjListFindByUUIDRef()) and adds qemuDomObjEndAPI() which
should be the only function in which the return value of
virObjectUnref() is checked. This makes all reference counting
deterministic and makes the code a bit clearer.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Although QMP returns info about vCPU threads in TCG mode, the
data it returns is mostly lies. Only the first vCPU has a valid
thread_id returned. The thread_id given for the other vCPUs is
in fact the main emulator thread. All vCPUs actually run under
the same thread in TCG mode.
Our vCPU pinning code is not at all able to cope with this
so if you try to set CPU affinity per-vCPU you end up with
wierd errors
error: Failed to start domain instance-00000007
error: cannot set CPU affinity on process 24365: Invalid argument
Since few people will care about the performance of TCG with
strict CPU pinning, lets just disable that for now, so we get
a clear error message
error: Failed to start domain instance-00000007
error: Requested operation is not valid: cpu affinity is not supported
The code assumes that def->vcpus == nvcpupids, so when we setup
fake CPU pids for old QEMU with nvcpupids == 1, we cause the
later code to read off the end of the array. This has fun results
like sche_setaffinity(0, ...) which changes libvirtd's own CPU
affinity, or even better sched_setaffinity($RANDOM, ...) which
changes the affinity of a random OS process.
We now have a qemuInterfaceStartDevices() which does the final
activation needed for the host-side tap/macvtap devices that are used
for qemu network connections. It will soon make sense to have the
converse qemuInterfaceStopDevices() which will undo whatever was done
during qemuInterfaceStartDevices().
A function to "stop" a single device has also been added, and is
called from the appropriate place in qemuDomainDetachNetDevice(),
although this is currently unnecessary - the device is going to
immediately be deleted anyway, so any extra "deactivation" will be for
naught. The call is included for completeness, though, in anticipation
that in the future there may be some required action that *isn't*
nullified by deleting the device.
This patch is a part of a more complete fix for:
https://bugzilla.redhat.com/show_bug.cgi?id=1081461
The patch that added qemuInterfaceStartDevices() (upstream commit
82977058f5) had an extra conditional to
prevent calling it if the reason for starting the CPUs was
VIR_DOMAIN_RUNNING_UNPAUSED or VIR_DOMAIN_RUNNING_SAVE_CANCELED. This
was put in by the author as the result of a reviewer asking if it was
necessary to ifup the interfaces in *all* occasions (because these
were the two cases where the CPU would have already been started (and
stopped) once, so the interface would already be ifup'ed).
It turns out that, as long as there is no corresponding
qemuInterfaceStopDevices() to ifdown the interfaces anytime the CPUs
are stopped, neglecting to ifup when reason is RUNNING_UNPAUSED or
RUNNING_SAVE_CANCELED doesn't cause any problems (because it just
happens that the interface will have already been ifup'ed by a prior
call when the CPU was previously started for some other reason).
However, it also doesn't *help*, and there will soon be a
qemuInterfaceStopDevices() function which *will* ifdown these
interfaces when the guest CPUs are stopped, and once that is done, the
interfaces will be left down in some cases when they should be up (for
example, if a domain is paused and then unpaused).
So, this patch is removing the condition in favor of always calling
qemuInterfaeStartDevices() when the guest CPUs are started.
This patch (and the aforementioned patch) resolve:
https://bugzilla.redhat.com/show_bug.cgi?id=1081461
Currently, MAC registration occurs during device creation, which is
early enough that, during live migration, you end up with duplicate
MAC addresses on still-running source and target devices, even though
the target device isn't actually being used yet.
This patch proposes to defer MAC registration until right before
the guest can actually use the device -- In other words, right
before starting guest CPUs.
Signed-off-by: Matthew Rosato <mjrosato@linux.vnet.ibm.com>
Signed-off-by: Laine Stump <laine@laine.org>
Move entering the job into the thread to simplify the program flow. Also
as the code holds a separate reference to the domain object some
conditions can be simplified.
After this patch qemuDomainObjTransferJob is no longer needed so this
patch removes it.
There are some small issue in qemuProcessAttach:
1.Fix virSecurityManagerGetProcessLabel always get pid = 0,
move 'vm->pid = pid' before call virSecurityManagerGetProcessLabel.
2.Use virSecurityManagerGenLabel to get image label.
3.Fix always set selinux label for other security driver label.
Signed-off-by: Luyao Huang <lhuang@redhat.com>
When a block{commit,copy} job was aborted on a domain, block job handler
did not process it correctly, leaving a phantom job in the background.
Any further calls to any blockjob causes "block <jobtype> still active"
error. This patch fixes the blockjob handler so that it checks not only
for VIR_DOMAIN_BLOCK_JOB_FAILED status, but VIR_DOMAIN_BLOCK_JOB_CANCELED
status as well, followed by our existing cleanup routine.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1135169
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1160084
As of b6d4dad11b (1.2.5) we are trying to keep the status of FSFreeze
in the guest. Even though I've tried to fixed couple of corner cases
(6ea54769ba), it occurred to me just recently, that the approach is
broken by design. Firstly, there are many other ways to talk to
qemu-ga (even through libvirt) that filesystems can be thawed (e.g.
qemu-agent-command) without libvirt noticing. Moreover, there are
plenty of ways to thaw filesystems without even qemu-ga noticing (yes,
qemu-ga keeps internal track of FSFreeze status). So, instead of
keeping the track ourselves, or asking qemu-ga for stale state, it's
the best to let qemu-ga deal with that (and possibly let guest kernel
propagate an error).
Moreover, there's one bug with the following approach, if fsfreeze
command failed, we've executed fsthaw subsequently. So issuing
domfsfreeze in virsh gave the following result:
virsh # domfsfreeze gentoo
Froze 1 filesystem(s)
virsh # domfsfreeze gentoo
error: Unable to freeze filesystems
error: internal error: unable to execute QEMU agent command 'guest-fsfreeze-freeze': The command guest-fsfreeze-freeze has been disabled for this instance
virsh # domfsfreeze gentoo
Froze 1 filesystem(s)
virsh # domfsfreeze gentoo
error: Unable to freeze filesystems
error: internal error: unable to execute QEMU agent command 'guest-fsfreeze-freeze': The command guest-fsfreeze-freeze has been disabled for this instance
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Improve the monitor function to also retrieve the guest state of
character device (if provided) so that we can refresh the state of
virtio-serial channels and perhaps react to changes in the state in
future patches.
This patch changes the returned data from qemuMonitorGetChardevInfo to
return a structure containing the pty path and the state for all the
character devices.
The change to the testsuite makes sure that the data is parsed
correctly.
New qemu added a new event that is emitted when a virtio serial channel
is opened in the guest OS. This allows us to update the state of the
port in the output-only XML element.
This patch implements the monitor callbacks and necessary handlers to
update the state in the definition.
To unify future additions that require information from "query-chardev"
rename qemuMonitorGetPtyPaths and friends to qemuMonitorGetChardevInfo
and move the allocation of the returned hash into the top level
function.
When creating a disk image snapshot the libvirt code would blindly copy
the parents label to the newly created image. This runs into problems
when you start a VM from an image hosted on NFS (or other storage system
that doesn't support selinux labels) and the snapshot destination is on
a storage system that does support selinux labels. Libvirt's code in
that case generates a different security label for the image hosted on
NFS. This label is valid only for NFS images and doesn't allow access in
case of a locally stored image.
To fix this issue libvirt needs to refrain from copying security
information in cases where the default domain seclabel is a better
choice.
This patch repurposes the now unused @force argument of
virStorageSourceInitChainElement to denote whether a copy of the
security labelling stuff should be attempted or not. This allows to
fine-control the copy operation for cases where we need to keep the
label of the old disk vs. the cases where we need to keep the label
unset to use the default domain imagelabel.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1151718
Ethernet interfaces in libvirt currently do not support bandwidth setting.
For example, following xml file for an interface will not apply these
settings to corresponding qdiscs.
<interface type="ethernet">
<mac address="02:36:1d:18:2a:e4"/>
<model type="virtio"/>
<script path=""/>
<target dev="tap361d182a-e4"/>
<bandwidth>
<inbound average="984" peak="1024" burst="64"/>
<outbound average="2000" peak="2048" burst="128"/>
</bandwidth>
</interface>
Signed-off-by: Anirban Chakraborty <abchak@juniper.net>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Commit 6e5c79a1 tried to fix deadlock between nwfilter{Define,Undefine}
and starting of guest, but this same deadlock exists for
updating/attaching network device to domain.
The deadlock was introduced by removing global QEMU driver lock because
nwfilter was counting on this lock and ensure that all driver locks are
locked inside of nwfilter{Define,Undefine}.
This patch extends usage of virNWFilterReadLockFilterUpdates to prevent
the deadlock for all possible paths in QEMU driver. LXC and UML drivers
still have global lock.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1143780
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
In one of my previous patches (3a3c3780b) I've tried to fix the
problem of nvram path disappearing on a domain that's been
started and shut down again. I fixed this by explicitly saving
domain's config file. However, I did a bit of clumsy without
realizing we have a transient domains for which we don't save the
config file. Hence, any domain using UEFI became persistent.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=1160084
As of b6d4dad1 (1.2.5) libvirt keeps track if domain disks have been
frozen. However, this falls into that set of information which don't
survive domain restart. Therefore, we need to clear the flag upon some
state transitions. Moreover, once we clear the flag we must update the
status file too.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
This is a reaction to Michal's fix [1] for non-NUMA systems that also
splits out conf/ out of util/ because libvirt_util shouldn't require
libvirt_conf if it is the other way around. This particular use case
worked, but we're trying to avoid it as mentioned [2], many times.
The only functions from virnuma.c that needed numatune_conf were
virDomainNumatuneNodesetIsAvailable() and virNumaSetupMemoryPolicy().
The first one should be in numatune_conf as it works with
virDomainNumatune, the second one just needs nodeset and mode, both of
which can be passed without the need of numatune_conf.
Apart from fixing that, this patch also fixes recently added
code (between commits d2460f85^..5c8515620) that doesn't support
non-contiguous nodesets. It uses new function
virNumaNodesetIsAvailable(), which doesn't need a stub as it doesn't use
any libnuma functions, to check if every specified nodeset is available.
[1] https://www.redhat.com/archives/libvir-list/2014-November/msg00118.html
[2] http://www.redhat.com/archives/libvir-list/2011-June/msg01040.html
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Particularly in qemuBuildNumaArgStr(), there was a need for the advice
due to memory backing, which needs to know the nodeset it will be pinned
to. With newer qemu this caused the following error when starting
domain:
error: internal error: Advice from numad is needed in case of
automatic numa placement
even when starting perfectly valid domain, e.g.:
...
<vcpu placement='auto'>4</vcpu>
<numatune>
<memory mode='strict' placement='auto'/>
</numatune>
<cpu>
<numa>
<cell id='0' cpus='0' memory='524288'/>
<cell id='1' cpus='1' memory='524288'/>
</numa>
</cpu>
...
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1138545
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
commit 3e1e16aa8d (Use a port from the
migration range for NBD as well) changed ndb port allocation from
remotePorts to migrationPorts, but did not change the port releasing
process, which makes an error when migrating several times (above 64):
error: internal error: Unable to find an unused port in range
'migration' (49152-49215)
https://bugzilla.redhat.com/show_bug.cgi?id=1159245
Signed-off-by: Weiwei Li <nuonuoli@tencent.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
If VM is configured with many devices(including passthrough devices)
and large memory, libvirtd will take seconds(in the worst case) to
wait for monitor. In this period the qemu process may run on any
PCPU though I intend to pin emulator to the specified PCPU in xml
configuration.
Actually qemu process takes high cpu usage during vm startup.
So this is not the strict CPU isolation in this case.
Signed-off-by: Zhou yimin <zhouyimin@huawei.com>
NIC_RX_FILTER_CHANGED is sent by qemu any time a NIC driver in the
guest modified the NIC's RX Filter (for example, if the MAC address of
the NIC is changed by the guest).
This patch doesn't do anything useful with that event; it just sets up
all the plumbing to get news of the event into a worker thread with
all proper locking/reference counting, and provide an easy place to
add in desired functionality.
See src/qemu/EVENTHANDLERS.txt for information/instructions on adding
a libvirt-internal handler for a qemu event (using
NIC_RX_FILTER_CHANGED as an example).
If we don't properly clean up all processes in the
machine-<vmname>.scope systemd won't remove the cgroup and subsequent vm
starts fail with
'CreateMachine: File exists'
Additional processes can e.g. be added via
echo $PID > /sys/fs/cgroup/systemd/machine.slice/machine-${VMNAME}.scope/tasks
but there are other cases like
http://bugs.debian.org/761521
Invoke TerminateMachine to be on the safe side since systemd tracks the
cgroup anyway. This is a noop if all processes have terminated already.
Commit fba6bc4 introduced the non-migratable invtsc feature,
breaking save/migration with host-model and host-passthrough.
On hosts with this feature present it was automatically included
in the CPU definition, regardless of QEMU support.
Commit de0aeaf stopped including it by default for host-model,
but failed to fix host-passthrough.
This commit ignores checking of CPU features with host-passthrough,
since we don't pass them to QEMU (only -cpu host is passed),
allowing domains using host-passthrough that were saved with
the broken version of libvirtd to be restored.
https://bugzilla.redhat.com/show_bug.cgi?id=1147584
On a domain startup, the variable store path is generated if needed.
The path is intended to be generated only once. However, the updated
domain definition is not saved into config dir rather than state XML
only. So later, whenever the domain is destroyed and the daemon is
restarted, the generated path is forgotten and the file may be left
behind on virDomainUndefine() call.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Request erroring out from the backing chain traveller and drop qemu's
internal backing chain integrity tester.
The backing chain traveller reports errors by itself with possibly more
detail than qemuDiskChainCheckBroken ever could.
We also need to make sure that we reconnect to existing qemu instances
even at the cost of losing the backing chain info (this really should be
stored in the XML rather than reloaded from disk, but that needs some
work).
Commit id '9a2f36ec' added a build conditional of CAP_SYS_RAWIO
in order to determine whether or not a disk definition using rawio
should be allowed on platforms without CAP_SYS_RAWIO. If one was
found, virReportError was used but the code didn't goto cleanup.
This patch adds the goto.
We are not detecting the presence of FIPS from QEMU, but from procfs and
that means it's not QEMU capability. It was decided that we will pass
this flag to QEMU even if it's not supported by old QEMU binaries.
This patch also reverts changes done by commit a21cfb0f to
qemucapabilitestest and implements a new test case in qemuxml2argvtest.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1135431
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
If there are no iothreads, then return from qemuProcessDetectIOThreadPIDs
without error; otherwise, the following occurs:
error: Failed to start domain $dom
error: An error occurred, but the cause is unknown
Modify qemuProcessStart() in order to allowing setting affinity to
specific CPU's for IOThreads. The process followed is similar to
that for the vCPU's.
This involves adding a function to fetch the IOThread id's via
qemuMonitorGetIOThreads() and adding them to iothreadpids[] list.
Then making sure all the cgroup data has been properly set up and
finally assigning affinity.
In qemuProcessInitPCIAddresses() if qemuMonitorGetAllPCIAddresses()
returns a negative (or zero) value, then no need to call the
qemuProcessDetectPCIAddresses().
Signed-off-by: John Ferlan <jferlan@redhat.com>
When using split UEFI image, it may come handy if libvirt manages per
domain _VARS file automatically. While the _CODE file is RO and can be
shared among multiple domains, you certainly don't want to do that on
the _VARS file. This latter one needs to be per domain. So at the
domain startup process, if it's determined that domain needs _VARS
file it's copied from this master _VARS file. The location of the
master file is configurable in qemu.conf.
Temporary, on per domain basis the location of master NVRAM file can
be overridden by this @template attribute I'm inventing to the
<nvram/> element. All it does is holding path to the master NVRAM file
from which local copy is created. If that's the case, the map in
qemu.conf is not consulted.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
Total time of a migration and total downtime transfered from a source to
a destination host do not count with the transfer time to the
destination host and with the time elapsed before guest CPUs are
resumed. Thus, source libvirtd remembers when migration started and when
guest CPUs were paused. Both timestamps are transferred to destination
libvirtd which uses them to compute total migration time and total
downtime. Obviously, this requires the time to be synchronized between
the two hosts. The reported times are useless otherwise but they would
be equally useless if we didn't do this recomputation so don't lose
anything by doing it.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
When QEMU fails during incoming migration after we successfully started
it (i.e., during Perform or Finish phase), we report a rather unhelpful
message
Unable to read from monitor: Connection reset by peer
We already have a code that takes error messages from QEMU's error
output but we disable it once QEMU successfully starts. This patch
postpones this until the end of Finish phase during incoming migration
so that we can report a much better error message:
internal error: early end of file from monitor: possible problem:
Unknown savevm section or instance '0000:00:05.0/virtio-balloon' 0
load of migration failed
https://bugzilla.redhat.com/show_bug.cgi?id=1090093
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Add umask to _virCommand, allow user to set umask to command.
Set umask(002) to qemu process to overwrite the default umask
of 022 set by many distros, so that unix sockets created for
virtio-serial has expected permissions.
Fix problem reported here:
https://sourceware.org/bugzilla/show_bug.cgi?id=13078#c11https://bugzilla.novell.com/show_bug.cgi?id=888166
To use virtio-serial device, unix socket created for chardev with
default umask(022) has insufficient permissions.
e.g.:
-device virtio-serial \
-chardev socket,path=/tmp/foo,server,nowait,id=foo \
-device virtserialport,chardev=foo,name=org.fedoraproject.port.0
srwxr-xr-x 1 qemu qemu 0 21. Jul 14:19 /tmp/somefile.sock
Other users in the same group (like real user, test engines, etc)
cannot write to this socket.
Signed-off-by: Chunyan Liu <cyliu@suse.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
The 'min_guarantee' is used by VMware ESX and OpenVZ drivers,
with qemu however, libvirt should report error when starting a domain,
because this element is not used.
Resolves https://bugzilla.redhat.com/show_bug.cgi?id=1122455
Currently, qemu driver uses qemuTranslateDiskSourcePool()
to translate disk volume information. This function is
general enough and could be used for other drivers as well,
so move it to conf/domain_conf.c along with its helpers.
- qemuTranslateDiskSourcePool: move to storage/storage_driver.c
and rename to virStorageTranslateDiskSourcePool,
- qemuAddISCSIPoolSourceHost: move to storage/storage_driver.c
and rename to virStorageAddISCSIPoolSourceHost,
- qemuTranslateDiskSourcePoolAuth: move to storage/storage_driver.c
and rename to virStorageTranslateDiskSourcePoolAuth,
- Update users of qemuTranslateDiskSourcePool to use a
new name.
Pin existing vcpus rather than existing vcpu pinning infos. This
increases the complexity of the lookup, but avoids pinning cpus that are
not enabled actually.
When editing guest's XML (on QEMU), it was possible to add multiple
listen elements into graphics parent element. However QEMU does not
support listening on multiple addresses. Configuration is tested for
multiple 'listen address' and if positive, an error is raised.
https://bugzilla.redhat.com/show_bug.cgi?id=1119212
During a QEMU live migration several warning messages about job
handling could be written to syslog on the destination host:
"entering monitor without asking for a nested job is dangerous"
The messages are written because the job handling during migration
uses hard coded asyncJob values in several places that are incorrect.
This patch passes the required asyncJob value around and prevents
the warnings as well as any issues that the warnings may be referring
to.
https://bugzilla.redhat.com/show_bug.cgi?id=1130089
Signed-off-by: Sam Bobroff <sam.bobroff@au1.ibm.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
A future patch is going to wire up qemu active block commit jobs;
but as they have similar events and are canceled/pivoted in the
same way as block copy jobs, it is easiest to track all bookkeeping
for the commit job by reusing the <mirror> element. This patch
adds domain XML to track which job was responsible for creating a
mirroring situation, and adds a job='copy' attribute to all
existing uses of <mirror>. Along the way, it also massages the
qemu monitor backend to read the new field in order to generate
the correct type of libvirt job (even though it requires a
future patch to actually cause a qemu event that can be reported
as an active commit). It also prepares to update persistent XML
to match changes made to live XML when a copy completes.
* docs/schemas/domaincommon.rng: Enhance schema.
* docs/formatdomain.html.in: Document it.
* src/conf/domain_conf.h (_virDomainDiskDef): Add a field.
* src/conf/domain_conf.c (virDomainBlockJobType): String conversion.
(virDomainDiskDefParseXML): Parse job type.
(virDomainDiskDefFormat): Output job type.
* src/qemu/qemu_process.c (qemuProcessHandleBlockJob): Distinguish
active from regular commit.
* src/qemu/qemu_driver.c (qemuDomainBlockCopy): Set job type.
(qemuDomainBlockPivot, qemuDomainBlockJobImpl): Clean up job type
on completion.
* tests/qemuxml2xmloutdata/qemuxml2xmlout-disk-mirror-old.xml:
Update tests.
* tests/qemuxml2argvdata/qemuxml2argv-disk-mirror.xml: Likewise.
* tests/qemuxml2argvdata/qemuxml2argv-disk-active-commit.xml: New
file.
* tests/qemuxml2xmltest.c (mymain): Drive new test.
Signed-off-by: Eric Blake <eblake@redhat.com>
We were not directly saving the domain XML to file after starting
or finishing a blockcopy. Without the startup write, a libvirtd
restart in the middle of a copy job would forget that the job was
underway. Then at pivot, we were indirectly writing new XML in
reaction to events that occur as we stop and restart the guest CPUs.
But there was a race: since pivot is an async action, it is possible
that libvirtd is restarted before the pivot completes, so if XML
changes during the event, that change was not written. The original
blockcopy code cleared out the <mirror> element prior to restarting
the CPUs, but this is also a race, observed if a user does an async
pivot and a dumpxml before the event occurs. Furthermore, this race
will interfere with active commit in a future patch, because that
code will rely on the <mirror> element at the time of the qemu event
to determine whether to inform the user of a normal commit or an
active commit.
Fix things by saving state any time we modify live XML, while
delaying XML disk modifications until after the event completes. We
still need a to teach libvirtd restarts to examine all existing
<mirror> elements to see if the job completed in the meantime (that
is, if libvirtd misses the event, the updated state still needs to be
updated in live XML), but that will be a later patch, in part because
we also need to to start taking advantage of newer qemu's ability to
keep the job around after completion rather than the current usage
where the job disappears both on error and on success.
* src/qemu/qemu_driver.c (qemuDomainBlockCopy): Track XML change
on disk.
(qemuDomainBlockJobImpl, qemuDomainBlockPivot): Move job-end XML
rewrites...
* src/qemu/qemu_process.c (qemuProcessHandleBlockJob): ...here.
Signed-off-by: Eric Blake <eblake@redhat.com>
Doing a blockcopy operation across a libvirtd restart is not very
robust at the moment. In particular, we are clearing the <mirror>
element prior to telling qemu to finish the job. Also, thanks to the
ability to request async completion, the user can easily regain
control prior to qemu actually finishing the effort, and they should
be able to poll the domain XML to see if the job is still going.
A future patch will fix things to actually wait until qemu is done
before modifying the XML to reflect the job completion. But since
qemu issues identical BLOCK_JOB_COMPLETE events regardless of whether
the job was cancelled (kept the original disk) or completed (pivoted
to the new disk), we have to track which of the two operations were
used to end the job. Furthermore, we'd like to avoid attempts to
end a job where we are already waiting on an earlier request to qemu
to end the job. Likewise, if we miss the qemu event (perhaps because
it arrived during a libvirtd restart), we still need enough state
recorded to be able to determine how to modify the domain XML once
we reconnect to qemu and manually learn whether the job still exists.
Although this patch doesn't actually fix the problem, it is a
preliminary step that makes it possible to track whether a job
has already begun steps towards completion.
* src/conf/domain_conf.h (virDomainDiskMirrorState): New enum.
(_virDomainDiskDef): Convert bool mirroring to new enum.
* src/conf/domain_conf.c (virDomainDiskDefParseXML)
(virDomainDiskDefFormat): Handle new values.
* src/qemu/qemu_process.c (qemuProcessHandleBlockJob): Adjust
client.
* src/qemu/qemu_driver.c (qemuDomainBlockPivot)
(qemuDomainBlockJobImpl): Likewise.
* docs/schemas/domaincommon.rng (diskMirror): Expose new values.
* docs/formatdomain.html.in (elementsDisks): Document it.
* tests/qemuxml2argvdata/qemuxml2argv-disk-mirror.xml: Test it.
Signed-off-by: Eric Blake <eblake@redhat.com>
Use better detection of hugetlbfs mount points. Yes, there can be
multiple mount points each serving different huge page size.
Since we already have ability to override the mount point in the
qemu.conf file, this crazy backward compatibility code is brought in.
Now we allow multiple mount points, so the "hugetlbfs_mount" option
must take an list of strings (mount points). But previously, it was
just a string, so we must accept both types now.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
