Most of our errors complaining about an inability to support a
particular action due to qemu limitations used CONFIG_UNSUPPORTED,
but we had a few outliers. Reported by Jiri Denemark.
* src/qemu/qemu_command.c (qemuBuildDriveDevStr): Prefer
CONFIG_UNSUPPORTED.
* src/qemu/qemu_driver.c (qemuDomainReboot)
(qemuDomainBlockJobImpl): Likewise.
* src/qemu/qemu_hotplug.c (qemuDomainAttachPciControllerDevice):
Likewise.
* src/qemu/qemu_monitor.c (qemuMonitorTransaction)
(qemuMonitorBlockJob, qemuMonitorSystemWakeup): Likewise.
So that a domain xml which doesn't have "placement" specified, but
"cpuset" is specified, could be parsed. And in this case, the
"placement" mode will be set as "static".
The objects (domain, pool, network, etc) for testing are defined/
started each time when opening a connect to test driver, and thus
the UUID for the objects will be generated each time, with different
values. e.g.
% for i in {1..3}; do ./tools/virsh --connect \
test:///default dumpxml test | grep uuid; done
<uuid>a1b6ee1f-97de-f0ee-617a-0cdb74947df5</uuid>
<uuid>ee68d7d2-3eb9-593e-2769-797ce1f4c4aa</uuid>
<uuid>fecb1d3a-918a-8412-e534-76192cf32b18</uuid>
It's the potential bug which can cause operations like below to fail:
$ virsh -c test:///default dumpxml test > test.xml
[ Some modificatons, though it's not supported, but it should work ]
$ virsh -c test:///default define test.xml
This patch set fixed UUID for objects which support it. (domain,
pool, network).
A "ide-drive" device can be either a hard disk or a CD-ROM,
if there is ",media=cdrom" specified for the backend, it's
a CD-ROM, otherwise it's a hard disk.
Upstream qemu splitted "ide-drive" into "ide-hd" and "ide-cd"
since commit 1f56e32, and ",media=cdrom" is not required for
ide-cd anymore. "ide-drive" is still supported for backwards
compatibility, but no doubt we should go foward.
A "scsi-disk" device can be either a hard disk or a CD-ROM,
if there is ",media=cdrom" specified for the backend, it's
a CD-ROM, otherwise it's a hard disk.
But upstream qemu splitted "scsi-disk" into "scsi-hd" and
"scsi-cd" since commit b443ae, and ",media=cdrom" is not
required for scsi-cd anymore. "scsi-disk" is still supported
for backwards compatibility, but no doubt we should go
foward.
If console[0] is an alias for serial[0], do not enforce the former to
have a PTY source type. This breaks serial consoles on stdio and makes
no sense.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
When using the xm/xend stack to manage instances there is a bug
that causes the emulated interfaces to be unusable when the vif
config contains type=ioemu.
The current code already has a special quirk to not use this
keyword if no specific model is given for the emulated NIC
(defaulting to rtl8139).
Essentially it works because regardless of the type argument,i
the Xen stack always creates emulated and paravirt interfaces and
lets the guest decide which one to use. So neither xl nor xm stack
actually require the type keyword for emulated NICs.
Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
Currently, we have 3 boolean arguments we have to pass
to qemuProcessStart(). As libvirt grows it is harder and harder
to remember them and their position. Therefore we should
switch to flags instead.
lvcreate want's the parent pool's name, not the pool path
lvchange and lvremove want lv specified as $vgname/$lvname
This largely worked before because these commands strip off a
starting /dev. But https://bugzilla.redhat.com/show_bug.cgi?id=714986
is from a user using a 'nested VG' that was having problems.
I couldn't find any info on nested LVM and the reporter never responded,
but I reproduced with XML that specified a valid source name, and
set target path to a symlink.
As explained in previous patch, numad will balance the affinity
dynamically, so reflecting the cpuset from numad at the first
time doesn't make much case, and may just could cause confusion.
Instead of returning a CPUs list, numad returns NUMA node
list instead, this patch is to convert the node list to
cpumap before affinity setting. Otherwise, the domain
processes will be pinned only to CPU[$numa_cell_num],
which will cause significiant performance losses.
Also because numad will balance the affinity dynamically,
reflecting the cpuset from numad back doesn't make much
sense then, and it may just could produce confusion for
the users. Thus the better way is not to reflect it back
to XML. And in this case, it's better to ignore the cpuset
when parsing XML.
The codes to update the cpuset is removed in this patch
incidentally, and there will be a follow up patch to ignore
the manually specified "cpuset" if "placement" is "auto",
and document will be updated too.
The linux-2.6.32 kernel header does not yet define IFLA_VF_MAX and others,
which breaks compiling a new libvirt on old systems like Debian Squeeze.
(I also have to add --without-macvtap --disable-werror --without-virtualport to
./configure to get it to compile.)
Signed-off-by: Philipp Hahn <hahn@univention.de>
Although it should be harmless to do:
disk = disk = def->disks[i]
some not-so-wise compilers may fool around.
Besides, such assignment is useless here.
On newer xend (v3.x and after) there is no state and domid reported
for inactive domains. When initially creating connections this is
handled in various places by assigning domain->id = -1.
But once an instance has been running, the id is set to the current
domain id. And it does not change when the instance is shut down.
So when querying the domain info, the hypervisor driver, which gets
asked first will indicate it cannot find information, then the
xend driver is asked and will set the status to NOSTATE because it
checks for the -1 domain id.
Checking domain/status for 0 seems to be more reliable for that.
One note: I am not sure whether the domain->id also should get set
back to -1 whenever any sub-driver thinks the instance is no longer
running.
BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=746007
BugLink: http://bugs.launchpad.net/bugs/929626
Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
This patch adds a netlink callback when migrating a VEPA enabled
virtual machine. It fixes a Bug where a VM would not request a port
association when it was cleared by lldpad.
This patch requires the latest git version of lldpad to work.
Signed-off-by: D. Herrendoerfer <d.herrendoerfer@herrendoerfer.name>
If dynamic_ownership is off and we are creating a file on NFS
we force chown. This will fail as chown/chmod are not supported
on NFS. However, with no dynamic_ownership we are not required
to do any chown.
In my testing, I was able to provoke an odd block pull failure:
$ virsh blockpull dom vda --bandwidth 10000
error: Requested operation is not valid: No active operation on device: drive-virtio-disk0
merely by using gdb to artifically wait to do the block job set speed
until after the pull had already finished. But in reality, that should
be a success, since the pull finished before we had a chance to set
speed. Furthermore, using a double job lock is not only annoying, but
a bug in itself - if you do parallel virDomainBlockRebase, and hit
the race window just right, the first call grabs the VM job to start
a fast block job, then the second call grabs the VM job to start
a long-running job with unspecified speed, then the first call finally
regrabs the VM job and sets the speed, which ends up running the
second job under the speed from the first call. By consolidating
things into a single job, we avoid opening that race, as well as reduce
the time between starting the job and changing the speed, for less
likelihood of the speed change happening after block job completion
in the first place.
* src/qemu/qemu_monitor.h (BLOCK_JOB_CMD): Add new mode.
* src/qemu/qemu_driver.c (qemuDomainBlockRebase): Move secondary
job call...
(qemuDomainBlockJobImpl): ...here, for fewer locks.
* src/qemu/qemu_monitor_json.c (qemuMonitorJSONBlockJob): Change
return value on new internal mode.
Without the VIR_DOMAIN_BLOCK_JOB_ABORT_ASYNC flag, libvirt will internally
poll using qemu's "query-block-jobs" API and will not return until the
operation has been completed. API users are advised that this operation
is unbounded and further interaction with the domain during this period
may block. Future patches may refactor things to allow other queries in
parallel with this polling. For older qemu, we synthesize the cancellation
event, since qemu won't generate it.
The choice of polling duration copies from the code in qemu_migration.c.
Signed-off-by: Adam Litke <agl@us.ibm.com>
Cc: Stefan Hajnoczi <stefanha@gmail.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
Probably in the noise, but this will let us scale more efficiently
as we learn to recognize even more qemu events.
* src/qemu/qemu_monitor_json.c (eventHandlers): Sort.
(qemuMonitorEventCompare): New helper function.
(qemuMonitorJSONIOProcessEvent): Optimize event lookup.
Block job cancellation can take a while. Now that upstream qemu 1.1
has asynchronous block cancellation, we want to expose that to the user.
Therefore, the following updates are made to the virDomainBlockJob API:
A new block job event type VIR_DOMAIN_BLOCK_JOB_CANCELED is managed by
libvirt. Regardless of the flags used with virDomainBlockJobAbort, this
event will be raised: 1. when using synchronous block_job_cancel (the
event will be synthesized by libvirt), and 2. whenever it is received
from qemu (via asynchronous block-job-cancel). Note that the event
may be detected by libvirt even before the virDomainBlockJobAbort
completes (always true when it is synthesized, but also possible if
cancellation was fast).
A new extension flag VIR_DOMAIN_BLOCK_JOB_ABORT_ASYNC is added to the
virDomainBlockJobAbort API. When enabled, this function will allow
(but not require) asynchronous operation (ie, it returns as soon as
possible, which might be before the job has actually been canceled).
When the API is used in this mode, it is the responsibility of the
caller to wait for a VIR_DOMAIN_BLOCK_JOB_CANCELED event or poll via
the virDomainGetBlockJobInfo API to check the cancellation status.
This patch also exposes the new flag through virsh, and makes virsh
slightly easier to use (--async implies --abort, and lack of any options
implies --info), although it leaves the qemu implementation for later
patches.
Signed-off-by: Adam Litke <agl@us.ibm.com>
Cc: Stefan Hajnoczi <stefanha@gmail.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
RHEL 6.2 was released with an early version of block jobs, which only
worked on the qed file format, where the commands were spelled with
underscore (contrary to QMP style), and where 'block_job_cancel' was
synchronous and did not trigger an event.
The upcoming qemu 1.1 release has fixed these short-comings [1][2]:
the commands now work on multiple file types, are spelled with dash,
and 'block-job-cancel' is asynchronous and emits an event upon conclusion.
[1]qemu commit 370521a1d6f5537ea7271c119f3fbb7b0fa57063
[2]https://lists.gnu.org/archive/html/qemu-devel/2012-04/msg01248.html
This patch recognizes the new spellings, and fixes virDomainBlockRebase
to give a graceful error when talking to a too-old qemu on a partial
rebase attempt. Fixes for the new semantics will come later. This
patch also removes a bogus ATTRIBUTE_NONNULL mistakenly added in
commit 10ec36e2.
* src/qemu/qemu_capabilities.h (QEMU_CAPS_BLOCKJOB_SYNC)
(QEMU_CAPS_BLOCKJOB_ASYNC): New bits.
* src/qemu/qemu_capabilities.c (qemuCaps): Name them.
* src/qemu/qemu_monitor_json.c (qemuMonitorJSONCheckCommands): Set
them.
(qemuMonitorJSONBlockJob): Manage both command names.
(qemuMonitorJSONDiskSnapshot): Minor formatting fix.
* src/qemu/qemu_monitor.h (qemuMonitorBlockJob): Alter signature.
* src/qemu/qemu_monitor_json.h (qemuMonitorJSONBlockJob): Likewise.
* src/qemu/qemu_monitor.c (qemuMonitorBlockJob): Pass through
capability bit.
* src/qemu/qemu_driver.c (qemuDomainBlockJobImpl): Update callers.
The new safe console handling introduced a possibility to deadlock the
qemu driver when a new console connection forcibly disconnects a
previous console stream that belongs to an already closed connection.
The virStreamFree function calls subsequently a the virReleaseConnect
function that tries to lock the driver while discarding the connection,
but the driver was already locked in qemuDomainOpenConsole.
Backtrace of the deadlocked thread:
0 0x00007f66e5aa7f14 in __lll_lock_wait () from /lib64/libpthread.so.0
1 0x00007f66e5aa3411 in _L_lock_500 () from /lib64/libpthread.so.0
2 0x00007f66e5aa322a in pthread_mutex_lock () from/lib64/libpthread.so.0
3 0x0000000000462bbd in qemudClose ()
4 0x00007f66e6e178eb in virReleaseConnect () from/usr/lib64/libvirt.so.0
5 0x00007f66e6e19c8c in virUnrefStream () from /usr/lib64/libvirt.so.0
6 0x00007f66e6e3d1de in virStreamFree () from /usr/lib64/libvirt.so.0
7 0x00007f66e6e09a5d in virConsoleHashEntryFree () from/usr/lib64/libvirt.so.0
8 0x00007f66e6db7282 in virHashRemoveEntry () from/usr/lib64/libvirt.so.0
9 0x00007f66e6e09c4e in virConsoleOpen () from /usr/lib64/libvirt.so.0
10 0x00000000004526e9 in qemuDomainOpenConsole ()
11 0x00007f66e6e421f1 in virDomainOpenConsole () from/usr/lib64/libvirt.so.0
12 0x00000000004361e4 in remoteDispatchDomainOpenConsoleHelper ()
13 0x00007f66e6e80375 in virNetServerProgramDispatch () from/usr/lib64/libvirt.so.0
14 0x00007f66e6e7ae11 in virNetServerHandleJob () from/usr/lib64/libvirt.so.0
15 0x00007f66e6da897d in virThreadPoolWorker () from/usr/lib64/libvirt.so.0
16 0x00007f66e6da7ff6 in virThreadHelper () from/usr/lib64/libvirt.so.0
17 0x00007f66e5aa0c5c in start_thread () from /lib64/libpthread.so.0
18 0x00007f66e57e7fcd in clone () from /lib64/libc.so.6
* src/qemu/qemu_driver.c: qemuDomainOpenConsole()
-- unlock the qemu driver right after acquiring the domain
object
In case an API fails with "cannot acquire state change lock", searching
for the API that possibly forgot to end its job is not always easy.
Let's keep track of the job owner and print it out for easier
identification.
As reported by Daniel Berrangé, we have a huge performance regression
for virDomainGetInfo() due to the change which makes virDomainEndJob()
save the XML status file every time it is called. Previous to that
change, 2000 calls to virDomainGetInfo() took ~2.5 seconds. After that
change, 2000 calls to virDomainGetInfo() take 2 *minutes* 45 secs.
We made the change to be able to recover from libvirtd restart in the
middle of a job. However, only destroy and async jobs are taken care of.
Thus it makes more sense to only save domain state XML when these jobs
are started/stopped.
I noticed these compiler warnings when building for the s390 architecture.
* src/node_device/node_device_udev.c (udevDeviceMonitorStartup):
Mark unused variable.
* src/nodeinfo.c (linuxNodeInfoCPUPopulate): Avoid unused variable.
* src/qemu/qemu_command.c: Wire up -bios with <loader>
* tests/qemuxml2argvdata/qemuxml2argv-bios.args,
tests/qemuxml2argvdata/qemuxml2argv-bios.xml: Expand
existing BIOS test case to cover <loader>
Leak introduced in commit 0436d32. If we allocate an actions array,
but fail early enough to never consume it with the qemu monitor
transaction call, we leaked memory.
But our semantics of making the transaction command free the caller's
memory is awkward; avoiding the memory leak requires making every
intermediate function in the call chain check for error. It is much
easier to fix things so that the function that allocates also frees,
while the call chain leaves the caller's data intact. To do that,
I had to hack our JSON data structure to make it easy to protect a
portion of an arbitrary JSON tree from being freed.
* src/util/json.h (virJSONType): Name the enum.
(_virJSONValue): New field.
* src/util/json.c (virJSONValueFree): Use it to protect a portion
of an array.
* src/qemu/qemu_monitor_json.c (qemuMonitorJSONTransaction): Avoid
freeing caller's data.
* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateDiskActive):
Free actions array on failure.
We can tell qemuDomainSnapshotFSThaw if we want it to report errors or
not. However, if we don't want to and an error has been already set by
previous qemuReportError() we must keep copy of that error not just a
pointer to it. Otherwise, it get overwritten if FSThaw reports an error.
Detected by valgrind. Leaks are introduced in commit b22eaa7.
* src/conf/domain_conf.c (virDomainDiskDefParseXML): fix memory leaks.
How to reproduce?
% make && make -C tests check TESTS=qemuxml2argvtest
% cd tests && valgrind -v --leak-check=full ./qemuxml2argvtest
actual result:
==2143== 12 bytes in 2 blocks are definitely lost in loss record 74 of 179
==2143== at 0x4A05FDE: malloc (vg_replace_malloc.c:236)
==2143== by 0x39D90A67DD: xmlStrndup (xmlstring.c:45)
==2143== by 0x4F5EC0: virDomainDiskDefParseXML (domain_conf.c:3438)
==2143== by 0x502F00: virDomainDefParseXML (domain_conf.c:8304)
==2143== by 0x505FE3: virDomainDefParseNode (domain_conf.c:9080)
==2143== by 0x5069AE: virDomainDefParse (domain_conf.c:9030)
==2143== by 0x41CBF4: testCompareXMLToArgvHelper (qemuxml2argvtest.c:105)
==2143== by 0x41E5DD: virtTestRun (testutils.c:145)
==2143== by 0x416FA3: mymain (qemuxml2argvtest.c:399)
==2143== by 0x41DCB7: virtTestMain (testutils.c:700)
==2143== by 0x39CF01ECDC: (below main) (libc-start.c:226)
Signed-off-by: Alex Jia <ajia@redhat.com>
If the daemon is restarted it will lose list of active
USB devices assigned to active domains. Therefore we need
to rebuild this list on qemuProcessReconnect().
To prevent assigning one USB device to two domains,
we keep a list of assigned USB devices. On domain
startup - qemuProcessStart() - we insert devices
used by domain into the list but remove them only
on detach-device. Devices are, however, released
on qemuProcessStop() as well.
Originally, qemuDomainCheckEjectableMedia was entering monitor with qemu
driver lock. Commit 2067e31bf9, which I
made to fix that, revealed another issue we had (but didn't notice it
since the driver was locked): we didn't set nested job when
qemuDomainCheckEjectableMedia is called during migration. Thus the
original fix I made was wrong.
XenD-3.1 introduced managed domains. HV-domains have rtc_timeoffset
(hgd24f37b31030 from 2007-04-03), which tracks the offset between the
hypervisors clock and the domains RTC, and is persisted by XenD.
In combination with localtime=1 this had a bug until XenD-3.4
(hg5d701be7c37b from 2009-04-01) (I'm not 100% sure how that bug
manifests, but at least for me in TZ=Europe/Berlin I see the previous
offset relative to utc being applied to localtime again, which manifests
in an extra hour being added)
XenD implements the following variants for clock/@offset:
- PV domains don't have a RTC → 'localtime' | 'utc'
- <3.1: no managed domains → 'localtime' | 'utc'
- ≥3.1: the offset is tracked for HV → 'variable'
due to the localtime=1 bug → 'localtime' | 'utc'
- ≥3.4: the offset is tracked for HV → 'variable'
Current libvirtd still thinks XenD only implements <clock offset='utc'/>
and <clock offset='localtime'/>, which is wrong, since the semantic of
'utc' and 'localtime' specifies, that the offset will be reset on
domain-restart, while with 'variable' the offset is kept. (keeping the
offset over "virsh edit" is important, since otherwise the clock might
jump, which confuses certain guest OSs)
xendConfigVersion was last incremented to 4 by the xen-folks for
xen-3.1.0. I know of no way to reliably detect the version of XenD
(user space tools), which may be different from the version of the
hypervisor (kernel) version! Because of this only the change from
'utc'/'localtime' to 'variable' in XenD-3.1 is handled, not the buggy
behaviour of XenD-3.1 until XenD-3.4.
For backward compatibility with previous versions of libvirt Xen-HV
still accepts 'utc' and 'localtime', but they are returned as 'variable'
on the next read-back from Xend to libvirt, since this is what XenD
implements: The RTC is NOT reset back to the specified time on next
restart, but the previous offset is kept.
This behaviour can be turned off by adding the additional attribute
adjustment='reset', in which case libvirt will report an error instead
of doing the conversion. The attribute can also be used as a shortcut to
offset='variable' with basis='...'.
With these changes, it is also necessary to adjust the xen tests:
"localtime = 0" is always inserted, because otherwise on updates the
value is not changed within XenD.
adjustment='reset' is inserted for all cases, since they're all <
XEND_CONFIG_VERSION_3_1_0, only 3.1 introduced persistent
rtc_timeoffset.
Some statements change their order because code was moved around.
Signed-off-by: Philipp Hahn <hahn@univention.de>
Since Xen 3.1 the clock=variable semantic is supported. In addition to
qemu/kvm Xen also knows about a variant where the offset is relative to
'localtime' instead of 'utc'.
Extends the libvirt structure with a flag 'basis' to specify, if the
offset is relative to 'localtime' or 'utc'.
Extends the libvirt structure with a flag 'reset' to force the reset
behaviour of 'localtime' and 'utc'; this is needed for backward
compatibility with previous versions of libvirt, since they report
incorrect XML.
Adapt the only user 'qemu' to the new name.
Extend the RelaxNG schema accordingly.
Document the new 'basis' attribute in the HTML documentation.
Adapt test for the new attribute.
Signed-off-by: Philipp Hahn <hahn@univention.de>
https://bugzilla.redhat.com/show_bug.cgi?id=808979
The leak is really in virProcessInfoGetAffinity, as shown in the
valgrind output given in the above bug report - it calls CPU_ALLOC(),
but then fails to call CPU_FREE().
This leak has existed in every version of libvirt since 0.7.5.
Commit 1b1402b introduced a regression. Since older libvirt versions
would silently round memory up (until the previous patch), but populated
current memory based on querying the guest, it was possible to have
dumpxml show cur > max by the amount of the rounding. For example, if
a user requested 1048570 KiB memory (just shy of 1GiB), the qemu
driver would actually run with 1048576 KiB, and libvirt 0.9.10 would
output a current that was 6KiB larger than the maximum. Situations
where this could have an impact include, but are not limited to,
migration from old to new libvirt, managedsave in old libvirt and
start in new libvirt, snapshot creation in old libvirt and revert in
new libvirt - without this patch, the new libvirt would reject the
VM because of the rounding discrepancy.
Fix things by adding a fuzz factor, and silently clamp current down to
maximum in that case, rather than failing to reparse XML for an existing
VM. From a practical standpoint, this has no user impact: 'virsh
dumpxml' will continue to query the running guest rather than rely on
the incoming xml, which will see the currect current value, and even if
clamping down occurs during parsing, it will be by at most the fuzz
factor of a megabyte alignment, and rounded back up when passed back to
the hypervisor.
Meanwhile, we continue to reject cur > max if the difference is beyond
the fuzz factor of nearest megabyte. But this is not a real change in
behavior, since with 0.9.10, even though the parser allowed it, later
in the processing stream we would reject it at the qemu layer; so
rejecting it in the parser just moves error detection to a nicer place.
* src/conf/domain_conf.c (virDomainDefParseXML): Don't reject
existing XML.
Based on a report by Zhou Peng.
