virCopyLastError is intended to be used after last error is set.
However due to virLastErrorObject failures (very unlikely though
as thread local error is allocated on first use) we can have zero
fields in a copy as a result. In particular code field can be set
to VIR_ERR_OK.
In some places (qemu monitor, qemu agent and qemu migaration code
for example) we use copy result as a flag and this leads to bugs.
Let's set OOM-like error in copy in case of virLastErrorObject failures.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Commit [1] dropped support for synchronous block job cancel.
This patch erases remnants from comments.
[1] commit 2350d101 "qemu: Remove support for legacy block jobs"
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Current script confuses on lines like this:
static virHypervisorDriver parallelsHypervisorDriver;
It interprets next lines as if there is open brace.
Let's filter this case from matches.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Broken by [1] commit - trailing comma instead of semicolon. Fortunately
the issue did not get sneak in released 4.2 version. Note that uriSchemes
for parallelsConnectDriver should not be allocated on stack.
[1] 8e4f9a27: "driver: declare supported URI schemes in virConnectDriver struct"
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
On Core i5 650 x86_64 kvm guest fail to start with error [1] for next cpu config:
<cpu mode='host-model' check='partial'>
<model fallback='allow'/>
<feature policy='require' name='x2apic'/>
</cpu>
The problem is in full CPU calculation in virQEMUCapsInitHostCPUModel.
It is supposed to include features emulated by qemu and missed on host. Some of
such features may be not included however.
For Core i5 650 host CPU is detected as Westmere and reported CPU as
SandyBridge. x2apic is missed on host and provided by installed qemu. The
feature is not mentioned in reported CPU features explicitly because SandyBridge
model include it. As a result full CPU does not include x2apic too.
Solution is to expand guest cpu features before updating fullCPU features.
[1] error: the CPU is incompatible with host CPU: \
Host CPU does not provide required features: x2apic
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
Ensure all enum cases are listed in switch statements, or cast away
enum type in places where we don't wish to cover all cases.
Build is broken after 67966ad51 [1].
[1] m4: enforce that all enum cases are listed in switch statements
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Build is broken by 5529b057 [1].
[1] cfg: forbid includes of headers in network and storage drivers again
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
This flag is only used for tests. Let's instead overload bind syscall
in mocks where it is not done yet.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Range check in virPortAllocatorSetUsed is not useful anymore
when we manage ports for entire unsigned short range values.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Range check in virPortAllocatorSetUsed is not useful anymore
when we manage ports for entire unsigned short range values.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Host tcp4/tcp6 ports is a global resource thus we need to make
port accounting also global or we have issues described in [1] when
port allocator ranges of different instances are overlapped (which
is by default for qemu for example).
Let's have only one global port allocator object that take care
of the entire ports range (0 - 65535) and introduce port range object
for clients to specify desired auto allocation band.
[1] https://www.redhat.com/archives/libvir-list/2017-December/msg00600.html
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
This patch pass event error up to the place where we can
use it. Error is passed only for sync blockjob event mode
as we can't use the error in async mode. In async mode we
just pass the event details to the client thru event API
but current blockjob event API can not carry extra parameter.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
After the virNetDaemonAddServerPostExec call in virtlogd we should have
netserver refcount set to 2. One goes to netdaemon servers hashtable
and one goes to virt{logd,lock} own reference to netserver. Let's add
the missing increment in virNetDaemonAddServerPostExec itself while
holding the daemon lock.
Since lockd defers management of the @srv object by the presence
in the hash table, virLockDaemonNewPostExecRestart must Unref the
alloc'd Ref on the @srv object done as part of virNetDaemonAddServerPostExec
and virNetServerNewPostExecRestart processing. The virNetDaemonGetServer
in lock_daemon main will also take a reference which is Unref'd during
main cleanup.
One of the usecases of iohelper is to read from pipe and write
to file with O_DIRECT. As we read from pipe we can have partial
read and then we fail to write this data because output file
is open with O_DIRECT and buffer size is not aligned.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
If same boot order is specified twice (or more) in domain xml
we call free for uninitiaziled loadparm on cleanup in virDomainDeviceBootParseXML
and SIGABRT (or similar) as a result.
In case of real migration (not migrating to file on save, dump etc)
migration info is not complete at time qemu finishes migration
in normal (non postcopy) mode. We need to update disks stats,
downtime info etc. Thus let's not expose this job status as
completed.
To archive this let's set status to 'qemu completed' after
qemu reports migration is finished. It is not visible as complete
job to clients. Cookie code on confirm phase will finally turn
job into completed. As we don't need more things to do when
migrating to file status is set to 'completed' as before
in this case.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
When getting job info in case mirror does not reach ready phase
fetch mirror stats from qemu. Otherwise mirror stats are already
saved in current job.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Looks like it is more simple to drop this optimization as we are
going to add getting disks stats during migration via quering qemu
process and checking if we have to acquire job condition becomes
more complicate.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Instead of checking stat.status let's set status to migrating
as soon as migrate command is send (waiting for completion
is a good place too).
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Setting status to none has little value - getting job status
will not return even elapsed time.
After this patch getting job stats stays correct in a sence
it will not fetch migration stats because it consults
stats.status before doing the fetch.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Querying destination migration statistics may result in getting
a failure or getting a elapsed time value depending on stats.status
value which is odd. Instead let's always fail. Clients should
be ready to handle this as currently getting failure period
can be considerable.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
qemuMigrationFetchJobStatus is rather inconvinient. Some of its
callers don't need status to be updated, some don't need to update
elapsed time right away. So let's update status or elapsed time
in callers instead.
This patch drops updating job status on getting job stats by
client. This way we will not provide status 'completed' while
it is not yet updated by migration routine.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
This way we get stats only in one place. The former code waits for
complete/postcopy status basically and don't need to mess with stats.
The patch drops raising an error on stats updates failure. This
does not make much sense anyway.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Let's introduce QEMU_DOMAIN_JOB_STATUS_POSTCOPY state for job.current->status
instead of checking job.current->stats.status. The latter can be changed
when fetching migration statistics. Moving state function from the variable
and leave only store function seems more managable.
This patch removes all state checking usage of stats except for
qemuDomainGetJobStatsInternal. This place will be handled separately.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
This patch simply switches code from using VIR_DOMAIN_JOB_* to
introduced QEMU_DOMAIN_JOB_STATUS_*. Later this gives us freedom
to introduce states for postcopy and mirroring phases.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
We call qemuDomainGetMachineName on domain start. On first
start (after daemon start) pid is 0 and virSystemdGetMachineNameByPID
don't get called. But after domain shutting down pid became -1 so
on next start virSystemdGetMachineNameByPID is called and returned an error.
Error is ignored so it is not critical. But at least on my system
(systemd-219 with extra patches) systemd-machined is crashed on
this request.
This behaviour is triggered by eaf2c9f89.
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
At the time the check was written virtuozzo did not use disabled items in boot
order configuration. Boot items were always enabled. Now they can be disabled
as well. Supporting such items is easy - they just should be ignored.
Disk serial schema has extra '.+' allowed characters in comparison
with check in code. Looks like there is no reason for that as qemu
allows any character AFAIK for serial. This discrepancy is originated
in commit id '85d15b51' where the ability to add serial was added.
Alter the disk-serial test to add a disk with all the possible
characters listed as the serial value.
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>
Current code for example can call unsubscribe if connection
succeeds but subscribing fails. This will probabaly lead
only to spurious error messages without any actual inconsistencies
but nevertheless.
Because this is invalid xml for containers. This patch almost
reverts 7eda8369, but still skips converting vz sdk bootorder
for containers to libvirt bootorder because we use boot order
in containers for quite different purpurse.
qemuAgentNotifyEvent accesses monitor structure and is called on qemu
reset/shutdown/suspend events under domain lock. Other monitor
functions on the other hand take monitor lock and don't hold domain lock.
Thus it is possible to have risky simultaneous access to the structure
from 2 threads. Let's take monitor lock here to make access exclusive.
In case of 0 filesystems *info is not set while according
to virDomainGetFSInfo contract user should call free on it even
in case of 0 filesystems. Thus we need to properly set
it. NULL will be enough as free eats NULLs ok.
When save/migrate a domain and we autogenerated a port, then if we
print the inactive domain config, write out a -1 for the socket value;
otherwise, it's possible that the subsequent start will fail if the
autogenerated websocket used conflicts with an existing running config
that also used autogenerated websockets.
Examples:
== A. Can not restore domain with autoconfigured websocket.
domain 1 and 2 have autoconfigured websocket.
1. domain 1 is started then, saved
2. domain 2 is started
3. domain 1 restoration is failed:
error: internal error: qemu unexpectedly closed the monitor: 2016-11-21T10:23:11.356687Z
qemu-kvm: -vnc 0.0.0.0:2,websocket=5700: Failed to start VNC server on `(null)':
Failed to bind socket: Address already in use
== B. Can not migrate domain with autoconfigured websocket.
domain 1 on host A, domain 2 on host B, both have autoconfigured websocket
1. domain 1 started, domain 2 started
2. domain 1 migration to host B is failed with the above error.
qemuDomainObjExitAgent is unsafe.
First it accesses domain object without domain lock.
Second it uses outdated logic that goes back to commit 79533da1 of
year 2009 when code was quite different. (unref function
instead of unreferencing only unlocked and disposed object
in case of last reference and leaved unlocking to the caller otherwise).
Nowadays this logic may lead to disposing locked object
i guess.
Another problem is that the callers of qemuDomainObjEnterAgent
use domain object again (namely priv->agent) without domain lock.
This patch address these two problems.
qemuDomainGetAgent is dropped as unused.
Sometimes after domain restart agent is unavailabe even
if it is up and running in guest. Diagnostic message is
"QEMU guest agent is not available due to an error"
that is 'priv->agentError' is set. Investiagion shows that
'priv->agent' is not NULL, so error flag is set probably
during domain shutdown process and not cleaned up eventually.
The patch is quite simple - just clean up error flag unconditionally
upon domain stop.
Other hunks address other cases when error flag is not cleaned up.
1. processSerialChangedEvent. We need to clean error flag
unconditionally here too. For example if upon first 'connected' event we
fail to connect and set error flag and then connect on second
'connected' event then error flag will remain set erroneously
and make agent unavailable.
2. qemuProcessHandleAgentEOF. If error flag is set and we get
EOF we need to change state (and diagnostic) from 'error' to
'not connected'.
qemuConnectAgent return -1 or -2 in case of different errors.
A. -1 is a case of unsuccessuful connection to guest agent.
B. -2 is a case of destoyed domain during connection attempt.
All qemuConnectAgent callers handle the first error the same way
so let's move this logic into qemuConnectAgent itself. Patched
function returns 0 in case A and -1 in case B.
It is already discussed in "[RFC] daemon: remove hardcode dep on libvirt-guests" [1].
Mgmt can use means to save/restore domains on system shutdown/boot other than
libvirt-guests.service. Thus we need to specify appropriate ordering dependency between
libvirtd, domains and save/restore service. This patch takes approach suggested
in RFC and introduces a systemd target, so that ordering can be built next way:
libvirtd -> domain -> virt-guest-shutdown.target -> save-restore.service.
This way domains are decoupled from specific shutdown service via intermediate
target.
[1] https://www.redhat.com/archives/libvir-list/2016-September/msg01353.html
Rather than waiting until we've free'd up all the resources, cause the
'workerPool' thread pool to flush as soon as possible during stateCleanup.
Otherwise, it's possible something waiting to run will SEGV such as is the
case during race conditions of simultaneous exiting libvirtd and qemu process.
Resolves the following crash:
[1] crash backtrace: (bt is shortened a bit):
0 0x00007ffff7282f2b in virClassIsDerivedFrom
(klass=0xdeadbeef, parent=0x55555581d650) at util/virobject.c:169
1 0x00007ffff72835fd in virObjectIsClass
(anyobj=0x7fffd024f580, klass=0x55555581d650) at util/virobject.c:365
2 0x00007ffff7283498 in virObjectLock
(anyobj=0x7fffd024f580) at util/virobject.c:317
3 0x00007ffff722f0a3 in virCloseCallbacksUnset
(closeCallbacks=0x7fffd024f580, vm=0x7fffd0194db0,
cb=0x7fffdf1af765 <qemuProcessAutoDestroy>)
at util/virclosecallbacks.c:164
4 0x00007fffdf1afa7b in qemuProcessAutoDestroyRemove
(driver=0x7fffd00f3a60, vm=0x7fffd0194db0) at qemu/qemu_process.c:6365
5 0x00007fffdf1adff1 in qemuProcessStop
(driver=0x7fffd00f3a60, vm=0x7fffd0194db0, reason=VIR_DOMAIN_SHUTOFF_CRASHED,
asyncJob=QEMU_ASYNC_JOB_NONE, flags=0)
at qemu/qemu_process.c:5877
6 0x00007fffdf1f711c in processMonitorEOFEvent
(driver=0x7fffd00f3a60, vm=0x7fffd0194db0) at qemu/qemu_driver.c:4545
7 0x00007fffdf1f7313 in qemuProcessEventHandler
(data=0x555555832710, opaque=0x7fffd00f3a60) at qemu/qemu_driver.c:4589
8 0x00007ffff72a84c4 in virThreadPoolWorker
(opaque=0x555555805da0) at util/virthreadpool.c:167
Thread 1 (Thread 0x7ffff7fb1880 (LWP 494472)):
1 0x00007ffff72a7898 in virCondWait
(c=0x7fffd01c21f8, m=0x7fffd01c21a0) at util/virthread.c:154
2 0x00007ffff72a8a22 in virThreadPoolFree
(pool=0x7fffd01c2160) at util/virthreadpool.c:290
3 0x00007fffdf1edd44 in qemuStateCleanup ()
at qemu/qemu_driver.c:1102
4 0x00007ffff736570a in virStateCleanup ()
at libvirt.c:807
5 0x000055555556f991 in main (argc=1, argv=0x7fffffffe458) at libvirtd.c:1660
Do not dereference the 'dmn' until after the virStateCleanup is completed.
During initialization, virStateInitialize requires/uses the "dmn" as the
argument to/for the daemonInhibitCallback functions. Thus, cleanup cannot
dereference 'dmn' until after calling the virStateCleanup which calls the
the daemonInhibitCallback using 'dmn'; otherwise, the following crash occurs:
backtrace (shortened a bit)
1 0x00007fd3a791b2e6 in virCondWait (c=<optimized out>, m=<optimized out>)
at util/virthread.c:154
2 0x00007fd3a791bcb0 in virThreadPoolFree (pool=0x7fd38024ee00)
at util/virthreadpool.c:266
3 0x00007fd38edaa00e in qemuStateCleanup () at qemu/qemu_driver.c:1116
4 0x00007fd3a79abfeb in virStateCleanup () at libvirt.c:808
5 0x00007fd3a85f2c9e in main (argc=<optimized out>, argv=<optimized out>)
at libvirtd.c:1660
Thread 1 (Thread 0x7fd38722d700 (LWP 32256)):
0 0x00007fd3a7900910 in virClassIsDerivedFrom
(klass=0xdfd36058d4853, parent=0x7fd3a8f394d0) at util/virobject.c:169
1 0x00007fd3a7900c4e in virObjectIsClass
(anyobj=anyobj@entry=0x7fd3a8f2f850, klass=<optimized out>)
at util/virobject.c:365
2 0x00007fd3a7900c74 in virObjectLock (anyobj=0x7fd3a8f2f850)
at util/virobject.c:317
3 0x00007fd3a7a24d5d in virNetDaemonRemoveShutdownInhibition
(dmn=0x7fd3a8f2f850) at rpc/virnetdaemon.c:547
4 0x00007fd38ed722cf in qemuProcessStop
(driver=driver@entry=0x7fd380103810, vm=vm@entry=0x7fd38025b6d0,
reason=reason@entry=VIR_DOMAIN_SHUTOFF_SHUTDOWN,
asyncJob=asyncJob@entry=QEMU_ASYNC_JOB_NONE, flags=flags@entry=0)
at qemu/qemu_process.c:5786
5 0x00007fd38edd9428 in processMonitorEOFEvent
(vm=0x7fd38025b6d0, driver=0x7fd380103810) at qemu/qemu_driver.c:4588
6 qemuProcessEventHandler (data=<optimized out>, opaque=0x7fd380103810)
at qemu/qemu_driver.c:4632
7 0x00007fd3a791bb55 in virThreadPoolWorker
(opaque=opaque@entry=0x7fd3a8f1e4c0) at util/virthreadpool.c:145
Absent driver name attribute is invalid xml. Which in turn makes
unusable 'virsh edit' for example. The value does not make
much sense and ignored on input so nobody will hurt.
vz sdk supports setting serial number only for disk devices.
Getting serial upon cdrom(for example) is error however
setting is just ignored. Let's check for disk device
explicitly for clarity in both cases.
Setting serial number for other devices is ignored
with an info note just as before.
We need usual conversion from "" to NULL in direction
vz sdk -> libvirt, because "" is not valid for libvirt
and "" means unspecifiend in vz sdk which is NULL for libvirt.
We can receive NULL as sync reply in two situations. First
is garbage sync reply and this situation is handled by
resending sync message. Second is different cases
of rebooting guest, destroing domain etc and we can
give more meaningful error message. Actually we have
this error message in qemuAgentCommand already which checks
for the same sitatuion. AFAIK case with mon->running
is just to be safe on adding some future(?) cases of
returning NULL reply.
We can easily handle receiving garbage on sync. We don't
have to make client deal with this situation. We just
need to resend sync command but this time garbage is
not be possible.
When we wait for sync reply we can receive delayed
reply to syncs or commands that were sent erlier. We can
safely skip them until we receive sync reply with correct id.
There is no much sense report this situation to client.
Actually with a bit of "luck" if we involve client into
this the play can go on forever: send sync 0, receive
sync reply -1, send sync 1, receive reply 0 ...
After sync is sent we can receive garbare and this is not error.
Consider next regular case:
1. libvirtd sends sync
2. qga sends partial sync reply and die
3. libvirtd sends sync
4. qga sends sync reply
5. libvirtd receives garbage
(half of first reply and second reply together)
We should handle this situation as it is recoverable.
Next sync can succeed. Let's report reply is NULL,
it will be converted to the VIR_ERR_AGENT_UNSYNCED
which signals client to retry.
Errors in qemuAgentIOProcessLine stop agent IO processing just
like any regular IO error, however some of current errors
that this functions spawns are false positives. Consider
next case for example:
1. send sync (unsynced state)
2. receive sync reply (sync established)
3. command send, but timeout occured (unsynced state)
4. receive command reply
Last IO triggers error because current code ignores
only delayed syncs when unsynced
We should not treat any delayed reply as error in unsynced
state. Until client and qga are not in sync delayed reply to any
command is possible. msg == NULL is the exact criterion
that we are not in sync.
When a new filter is being defined, the return code is not handled properly,
thus triggering OOM error reporting routine (bug introduced by 51b2606f).
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Since the domain lock is not held during preparation of an external XML
config, it is possible that the value can change resulting in unexpected
failures during ABI consistency checking for some save and migrate
operations.
This patch adds a new flag to skip the checking of the cur_balloon value
and then sets the destination value to the source value to ensure
subsequent checks without the skip flag will succeed.
This way it is protected from forges and is keeped up to date too.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
vzDomainMigrateConfirm3Params is whitelisted. Otherwise we need to
move removing domain from domain list from perform to confirm
step. This would further imply adding a flag and check that migration
is in progress to prohibit mistakenly (maliciously) removing domains
on confirm step. vz version of p2p also need to be fixed to include confirm step.
One would also need to add means to cleanup pending migration
on client disconnect as now is has state across several API
calls.
On the other hand current version of confirm step is totaly
harmless thus it is easier to whitelist it at the moment.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Signed-off-by: Maxim Nestratov <mnestratov@virtuozzo.com>
This way we make naming consistent to API calls and make subsequent
ACL checks possible (otherwise ACL check would discover name
discrepancies).
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
ACL check on perform step should be in API call itself to make ACL
checking script pass. Thus we need to reorganize code to obtain
domain object in perform API itself. Most of this is straight
forward, the only nuance is dropping locks on lengthy remote
operations.
The other motivation is to have only perform step ACL checks for
p2p migration instead of both begin in perform if we can leave
ACL check in vzDomainMigratePerformStep.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
We need it to prepare the calls for ACL checks otherwise ACL checking
script will fail.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
This action deserves its own function and makes main API call
structure much cleaner.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
The original motivation is to expand API calls like start/stop etc so that
the ACL checks could be added. But this patch has its own befenits.
1. functions like prlsdkStart/Stop use common routine to wait for
job without domain lock. They become more self contained and do
not return intermediate PRL_RESULT.
2. vzDomainManagedSave do not update cache twice.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Signed-off-by: Maxim Nestratov <mnestratov@virtuozzo.com>
Unfortunately vz sdk do not provide detail information on migration
progress, only progress percentage. Thus vz driver provides percents
instead of bytes in data fields of virDomainJobInfoPtr.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
As commit id 'e2b86f580' notes, when mode=agent possibly setting the
fake reboot flag to true wouldn't be necessary; however, it doesn't
"force" the issue by just ensuring the fake reboot is false, so this
patch adds the explicit setting for the reboot path.
More investigation and details can be found in commit id '8be502fd'
as well as in the archives at:
https://www.redhat.com/archives/libvir-list/2015-April/msg00715.html
