Commit Graph

4953 Commits

Author SHA1 Message Date
Eric Blake
f9a9214629 maint: add gnulib gettimeofday module
* bootstrap.conf (gnulib_modules): Add gettimeofday.
2010-06-25 07:46:28 -06:00
Daniel P. Berrange
39f8af3880 Fix test case failure due to missing -nodefconfig
The previous change which split -nodefconfig probing off
from -device broke the test case because it missed adding
the QEMUD_CMD_FLAG_NODEFCONFIG to the test

* src/util/bridge.c: Set QEMUD_CMD_FLAG_NODEFCONFIG for all
  configs with QEMUD_CMD_FLAG_DEVICE set
2010-06-25 14:40:24 +01:00
Daniel P. Berrange
28e96d720b Fix migration in text mode and shared storage migration in json mode
The patches for shared storage migration were not correctly written
for json mode. Thus the 'blk' and 'inc' parameters were never being
set. In addition they didn't set the QEMU_MONITOR_MIGRATE_BACKGROUND
so migration was synchronous. Due to multiple bugs in QEMU's JSON
impl this wasn't noticed because it treated the sync migration requst
as asynchronous anyway. Finally 'background' parameter was converted
to take arbitrary flags but not renamed, and not all uses were changed
to unsigned int.

* src/qemu/qemu_driver.c: Set QEMU_MONITOR_MIGRATE_BACKGROUND in
  doNativeMigrate
* src/qemu/qemu_monitor_json.c: Process QEMU_MONITOR_MIGRATE_NON_SHARED_DISK
  and QEMU_MONITOR_MIGRATE_NON_SHARED_INC flags
* src/qemu/qemu_monitor.c, src/qemu/qemu_monitor.h,
  src/qemu/qemu_monitor_json.h, src/qemu/qemu_monitor_text.c,
  src/qemu/qemu_monitor_text.h: change 'int background' to
  'unsigned int flags' in migration APIs. Add logging of flags
  parameter
2010-06-25 14:40:06 +01:00
Daniel P. Berrange
755b53f946 Avoid blocking all APIs during incoming migration
During incoming migration the QEMU monitor is not able to be
used. The incoming migration code did not keep hold of the
job lock because migration is split across multiple API calls.
This meant that further monitor commands on the guest would
hang until migration finished with no timeout.

In this change the qemuDomainMigratePrepare method sets the
job flag just before it returns. The qemuDomainMigrateFinish
method checks for this job flag & clears it once done. This
prevents any use of the monitor between prepare+finish steps.

The qemuDomainGetJobInfo method is also updated to refresh
the job elapsed time. This means that virsh domjobinfo can
return time data during incoming migration

* src/qemu/qemu_driver.c: Keep a job active during incoming
  migration. Refresh job elapsed time when returning job info
2010-06-25 14:39:58 +01:00
Daniel P. Berrange
2bad82f71e Set labelling for character devices in security drivers
When configuring serial, parallel, console or channel devices
with a file, dev or pipe backend type, it is necessary to label
the file path in the security drivers. For char devices of type
file, it is neccessary to pre-create (touch) the file if it does
not already exist since QEMU won't be allowed todo so itself.
dev/pipe configs already require the admin to pre-create before
starting the guest.

* src/qemu/qemu_security_dac.c: set file ownership for character
  devices
* src/security/security_selinux.c: Set file labeling for character
  devices
* src/qemu/qemu_driver.c: Add character devices to cgroup ACL
2010-06-25 14:39:54 +01:00
Daniel P. Berrange
bf1f3f7a36 Add API for iterating over all character devices
The parallel, serial, console and channel devices are all just
character devices. A lot of code needs todo the same thing to
all these devices. This provides an convenient API for iterating
over all of them.

* src/conf/domain_conf.c, src/conf/domain_conf.c,
  src/libvirt_private.syms: Add virDomainChrDefForeach
2010-06-25 14:39:45 +01:00
Daniel P. Berrange
a711633909 Add missing parameter in python Disk IO error callback
The IO error callback was forgetting to pass the action
parameter, causing a stack trace when IO errors arrive

* python/libvirt-override-virConnect.py: Add missing action
  parameter in IO error callback
2010-06-25 14:39:37 +01:00
Laine Stump
7d02393b1b Check for presence of qemu -nodefconfig option before using it
We previously assumed that if the -device option existed in qemu, that
-nodefconfig would also exist. It turns out that isn't the case, as
demonstrated by qemu-kvm-0.12.3 in Fedora 13.

*/src/qemu/qemu_conf.[hc] - add a new QEMUD_CMD_FLAG, set it via the
                            help output, and check it before adding
                            -nodefconfig to the qemu commandline.
2010-06-24 18:04:07 -04:00
Eduardo Otubo
edfedfb414 Adding Storage Management driver (style and indentation)
This is just the indentation, style  and cleanup patch.
2010-06-24 14:03:50 -06:00
Matthias Bolte
e214bcf37c esx: Use bool instead of int where appropriated 2010-06-24 14:56:04 +02:00
Matthias Bolte
d72d92565b esx: Add support for the controller element
Also don't abuse the disk driver name to specify the SCSI controller
model anymore:

  <driver name='buslogic'/>

Use the newly added model attribute of the controller element for this:

  <controller type='scsi' index='0' model='buslogic'/>

The disk driver name approach is deprecated now, but still works for
backward compatibility reasons.

Update the documentation and tests accordingly.

Fix usage of the words controller and id in the VMX handling code. Use
controller, bus and unit properly.
2010-06-24 14:56:04 +02:00
Matthias Bolte
c391291244 Add optional model attribute to the controller element
This is a step towards controller support for the ESX driver.
2010-06-24 14:56:04 +02:00
Matthias Bolte
f8f29b1fc2 Add wide SCSI bus disk address generation support
The domain XML parsing code autogenerates disk address and
controller elements when they are not explicitly specified.
The code assumes a narrow SCSI bus (7 units per bus). ESX
uses a wide SCSI bus (16 units per bus).

This is a step towards controller support for the ESX driver.
2010-06-24 14:56:04 +02:00
Matthias Bolte
a73b389d12 Cleanup some LIBADD and CFLAGS
Move libnl to libvirt_util.la, because macvtap.c requires it.

Add GnuTLS to libvirt_driver.la, because libvirt.c calls gcrypt functions.
When built without loadable driver modules, then the remote driver pulls
in GnuTLS.

Move libgnu.la from libvirt_parthelper_CFLAGS to libvirt_parthelper_LDADD.
2010-06-24 14:16:27 +02:00
Ryota Ozaki
4a4eb13e7a cgroup: Enable memory.use_hierarchy of cgroup for domain
Through conversation with Kumar L Srikanth-B22348, I found
that the function of getting memory usage (e.g., virsh dominfo)
doesn't work for lxc with ns subsystem of cgroup enabled.

This is because of features of ns and memory subsystems.
Ns creates child cgroup on every process fork and as a result
processes in a container are not assigned in a cgroup for
domain (e.g., libvirt/lxc/test1/). For example, libvirt_lxc
and init (or somewhat specified in XML) are assigned into
libvirt/lxc/test1/8839/ and libvirt/lxc/test1/8839/8849/,
respectively. On the other hand, memory subsystem accounts
memory usage within a group of processes by default, i.e.,
it does not take any child (and descendant) groups into
account. With the two features, virsh dominfo which just
checks memory usage of a cgroup for domain always returns
zero because the cgroup has no process.

Setting memory.use_hierarchy of a group allows to account
(and limit) memory usage of every descendant groups of the group.
By setting it of a cgroup for domain, we can get proper memory
usage of lxc with ns subsystem enabled. (To be exact, the
setting is required only when memory and ns subsystems are
enabled at the same time, e.g., mount -t cgroup none /cgroup.)
2010-06-23 14:31:38 -06:00
Ryota Ozaki
842b51ff5d cgroup: Change virCgroupRemove to remove all descendant groups at first
As same as normal directories, a cgroup cannot be removed if it
contains sub groups. This patch changes virCgroupRemove to remove
all descendant groups (subdirectories) of a target group before
removing the target group.

The handling is required when we run lxc with ns subsystem of cgroup.
Ns subsystem automatically creates child cgroups on every process
forks, but unfortunately the groups are not removed on process exits,
so we have to remove them by ourselves.

With this patch, such child (and descendant) groups are surely removed
at lxc shutdown, i.e., lxcVmCleanup which calls virCgroupRemove.
2010-06-23 14:30:19 -06:00
Eric Blake
4cc2b6d676 virsh: document attach-disk better
http://bugzilla.redhat.com/601143, part 1 - document existing
behavior.  Ever since Mar 2010 (commit ced154cb), the use of
'attach-disk' or 'attach-device' to change cdrom/floppy media has been
documented but deprecated, but the replacement to use 'update-device'
was not documented.

* tools/virsh.c (cmdAttachInterface, cmdAttachDisk): Fix bad error
message.
* tools/virsh.pod (attach-device, attach-disk): Refer to
update-device for cdrom and floppy behavior.
(update-device): Add documentation.
2010-06-23 08:35:40 -06:00
Alan Pevec
4efaf77b19 network: allow tftp port if tftp is defined
add iptables rules to allow TFTP from the virtual network if <tftp>
element is defined in the network definition.

Fedora bz#580215

* src/network/bridge_driver.c: open UDP port 69 for TFTP traffic if
  tftproot is defined
2010-06-23 08:24:00 -06:00
Alan Pevec
0c141c893a bridge_driver.c: fix file description 2010-06-23 08:18:57 -06:00
Daniel P. Berrange
f310b25341 Add '-nodefconfig' command line arg to QEMU
We already use the '-nodefaults' command line arg with QEMU to stop
it adding any default devices to guests. Unfortunately, QEMU will
load global config files from /etc/qemu that may also add default
devices. These aren't blocked by '-nodefaults', so we need to also
add the '-nodefconfig' arg to prevent that.

Unfortunately these global config files are also used to define
custom CPU models. So in blocking global hardware device addition
we also block definitions of new CPU models. Libvirt doesn't know
about these custom CPU models though, so it would never make use
of them anyway. Thus blocking them via -nodefconfig isn't a show
stopping problem. We would need to expand libvirt's own CPU model
XML database to support these instead.

* src/qemu/qemu_conf.c: Add '-nodefconfig' if available
* tests/qemuxml2argvdata/: Add '-nodefconfig' to all data files which
  have '-nodefaults' present
2010-06-23 14:08:05 +01:00
Daniel P. Berrange
c212160260 Fix reference handling leak on qemuMonitor
The current code pattern requires that callers of qemuMonitorClose
check for the return value == 0, and if so, set priv->mon = NULL
and release the reference held on the associated virDomainObjPtr

The change d84bb6d6a3 violated that
requirement, meaning that priv->mon never gets set to NULL, and
a reference count is leaked on virDomainObjPtr.

This design was a bad one, so remove the need to check the return
valueof qemuMonitorClose(). Instead allow registration of a
callback that's invoked just when the last reference on qemuMonitorPtr
is released.

Finally there was a potential reference leak in qemuConnectMonitor
in the failure path.

* src/qemu/qemu_monitor.c, src/qemu/qemu_monitor.h: Add a destroy
  callback invoked from qemuMonitorFree
* src/qemu/qemu_driver.c: Use the destroy callback to release the
  reference on virDomainObjPtr when the monitor is freed. Fix other
  potential reference count leak in connecting to monitor
2010-06-23 14:08:05 +01:00
Daniel P. Berrange
8d616decc5 Make checks for inactive QEMU guest more robust
Before issuing monitor commands it is neccessary to check whether
the guest is still running. Most places use virDomainIsActive()
correctly, but a few relied on 'priv->mon != NULL'. In theory
these should be equivalent, but the release of the last reference
count on priv->mon can be delayed a small amount of time until
the event handler is finally deregistered. A further ref counting
bug also means that priv->mon might be never released. In such a
case, code could mistakenly issue a monitor command and wait for
a response that will never arrive, effectively leaving the QEMU
driver waiting on virCondWait() forever..

To protect against these possibilities, make sure all code uses
virDomainIsActive(), not 'priv->mon != NULL'

* src/qemu/qemu_driver.c: Replace 'priv->mon != NULL' with
  calls to 'priv->mon != NULL'()
2010-06-23 14:08:05 +01:00
Daniel P. Berrange
9b0244ae38 Improve some error messages about unsupported APIs/URIs
If there is no driver for a URI we report

  "no hypervisor driver available"

This is bad because not all virt drivers are hypervisors (ie container
based virt).

If there is no driver support for an API we report

  "this function is not supported by the hypervisor"

This is bad for the same reason, and additionally because it is
also used for the network, interface & storage drivers.

* src/util/virterror.c: Improve error messages
2010-06-23 14:07:39 +01:00
Jiri Denemark
6c267f01b1 Don't leak open fd to virsh in libvirt-guests init script
Running virsh while having /var/lib/libvirt/libvirt-guests file open
makes SELinux emit messages about preventing virsh from reading that
file. Since virsh doesn't really want to read anything, it's better to
run it with /dev/null on stdin to prevent those messages.
2010-06-22 17:48:14 +02:00
Stefan Berger
cab5a52aa2 nwfilter: fix loadable module support
Following Daniel Berrange's multiple helpful suggestions for improving
this patch and introducing another driver interface, I now wrote the
below patch where the nwfilter driver registers the functions to
instantiate and teardown the nwfilters with a function in
conf/domain_nwfilter.c called virDomainConfNWFilterRegister. Previous
helper functions that were called from qemu_driver.c and qemu_conf.c
were move into conf/domain_nwfilter.h with slight renaming done for
consistency. Those functions now call the function expored by
domain_nwfilter.c, which in turn call the functions of the new driver
interface, if available.
2010-06-21 14:18:31 -04:00
Justin Clift
c7a33939bc virsh: remove a doubled up include for errno.h 2010-06-21 10:11:34 +02:00
Jiri Denemark
d147b18de2 Misc cleanups
- Fix documentation for virGetStorageVol: it has 'key' argument instead
  of 'uuid'.
- Remove TODO comment from virReleaseStorageVol: we use volume key as an
  identifier instead of UUID.
- Print human-readable UUID string in debug message in virReleaseSecret.
2010-06-21 10:11:34 +02:00
Jiri Denemark
72a7f8b2a9 Do not free static buffer with UUID
As anywhere else, uuid is defined as a fixed size array inside
_virSecret structure; we shouldn't try to free it.
2010-06-21 10:11:34 +02:00
Jiri Denemark
c5ec45a3a4 Remove unnecessary check for non-NULL uuid
The first thing we do in all these functions is to check uuid != NULL
and fail if it isn't.
2010-06-21 10:11:34 +02:00
Jiri Denemark
30ec755ecb Index hashes by UUID instead of name
Per-connection hashes for domains, networks, storage pools and network
filter pools were indexed by names which was not the best choice. UUIDs
are better identifiers, so lets use them.
2010-06-21 10:11:34 +02:00
Philipp Hahn
ff5f7d7204 Allow one-or-more <boot dev="..."/> entries
According to docs/formatdomain.html.in, "The boot element can be
repeated multiple times to setup a priority list of boot devices to try
in turn." The Relax-NG schema required / allowed exactly one entry.

Signed-off-by: Philipp Hahn <hahn@univention.de>
2010-06-21 10:11:33 +02:00
Stefan Berger
7057f39c51 nwfilter: extensions of docs with
As requested, here a couple of paragraphs about the recently added statematch attribute and some advanced (and tricky) traffic filtering topics.
2010-06-18 13:44:17 -04:00
Cole Robinson
9edceb3233 Add ACK'd v2 changes for previous commit 2010-06-18 11:56:04 -04:00
Cole Robinson
58406dd54e qemu: Fix crash on failed VM startup
If VM startup fails early enough (can't find a referenced USB device),
libvirtd will crash trying to clear the VNC port bit, since port = 0,
which overflows us out of the bitmap bounds.

Fix this by being more defensive in the bitmap operations, and only
clearing a previously set VNC port.

Signed-off-by: Cole Robinson <crobinso@redhat.com>
2010-06-18 10:14:34 -04:00
Philipp Hahn
08fee18cc2 Fix description of virStorageVolGetInfo()
Probably a copy-paste-bug in python/libvirt-override-api.xml:
virStorageVolGetInfo() extracts information about a "storage volume",
not the "storage pool" as virStoragePoolGetInfo() does.

Signed-off-by: Philipp Hahn <hahn@univention.de>
2010-06-18 12:57:54 +02:00
Justin Clift
2b39cd355a virsh: add --uuid option to vol-pool
Adds an optional switch, --uuid, for telling the virsh vol-pool command
to return the pool UUID rather than pool name.
2010-06-17 14:45:34 -06:00
Eric Blake
322b1fd44b qemu: reduce file padding requirements
Followup to https://bugzilla.redhat.com/show_bug.cgi?id=599091,
commit 20206a4b, to reduce disk waste in padding.

* src/qemu/qemu_monitor.h (QEMU_MONITOR_MIGRATE_TO_FILE_BS): Drop
back to 4k.
(QEMU_MONITOR_MIGRATE_TO_FILE_TRANSFER_SIZE): New macro.
* src/qemu/qemu_driver.c (qemudDomainSaveFlag): Update comment.
* src/qemu/qemu_monitor_text.c (qemuMonitorTextMigrateToFile): Use
two invocations of dd to output non-aligned large blocks.
* src/qemu/qemu_monitor_json.c (qemuMonitorJSONMigrateToFile):
Likewise.
2010-06-17 13:29:16 -06:00
Stefan Berger
51d3fb0276 nwfilter: add XML attribute to control iptables state match
This patch adds an optional XML attribute to a nwfilter rule to give the user control over whether the rule is supposed to be using the iptables state match or not. A rule may now look like shown in the XML below with the statematch attribute either having value '0' or 'false' (case-insensitive).

[...]
<rule action='accept' direction='in' statematch='false'>
<tcp srcmacaddr='1:2:3:4:5:6'
           srcipaddr='10.1.2.3' srcipmask='32'
           dscp='33'
           srcportstart='20' srcportend='21'
           dstportstart='100' dstportend='1111'/>
</rule>
[...]

I am also extending the nwfilter schema and add this attribute to a test case.
2010-06-17 14:12:34 -04:00
Justin Clift
c2160b137d virsh: ensure persistence and autostart are shown for dominfo and pool-info
This patch adds the persistence status (yes/no) to the output of the virsh
dominfo and pool-info commands.  This patch also adds the autostart status
to the output of the virsh pool-info command.

Red Hat BZ for this:

  https://bugzilla.redhat.com/show_bug.cgi?id=603696
2010-06-17 11:57:54 -06:00
Eduardo Otubo
788269e970 phyp: adding support for IVM
Use virBuffer* API to conditionally keep the portion of the command
line specific to HMC, so that IVM can work.

Signed-off-by: Eric Blake <eblake@redhat.com>
2010-06-17 11:49:09 -06:00
Stefan Berger
2dce970162 nwfilter: use match target on incoming traffic
The following patch enables the iptables match target to be used by
default for incoming traffic. So far it has only be used for outgoing
traffic.
2010-06-17 07:15:20 -04:00
Stefan Berger
045a5722ab macvtap: work-around for 2.6.32 and older kernels
This patch works around a recent extension of the netlink driver I had made use of when building the netlink messages. Unfortunately older kernels don't accept IFLA_IFNAME + name of interface as a replacement for the interface's index, so this patch now gets the interface index ifindex if it's not provided (ifindex <= 0).
2010-06-17 07:05:38 -04:00
Justin Clift
07be2403b6 virsh: change printf() calls to vshPrint()
Trivial fix changing printf() calls to vshPrint() where the ctl
variable is available.
2010-06-16 17:17:08 -06:00
Justin Clift
7d38d7b49b virsh: improve help text for vol query commands
Improves the help text for vol-path, vol-name, and vol-key, which
previously referred to volume UUIDs.

Addresses BZ # 598365.
2010-06-16 16:32:05 -06:00
Justin Clift
31495ef6a1 virsh: add pool support to vol-key command
Presently the vol-key command only supports being provided with
a volume path.

This patch adds support for providing it with a pool and volume
identifier pair as well.

    virsh # vol-key --pool <pool-name-or-uuid> <vol-name-or-path>
2010-06-16 16:32:05 -06:00
Matthias Bolte
08d42b52f1 Add several missing vir*Free calls in libvirtd's remote code
Justin Clift reported a problem with adding virStoragePoolIsPersistent
to virsh's pool-info command, resulting in a strange problem. Here's
an example:

    virsh # pool-create-as images_dir3 dir - - - - "/home/images2"
    Pool images_dir3 created

    virsh # pool-info images_dir3
    Name:           images_dir3
    UUID:           90301885-94eb-4ca7-14c2-f30b25a29a36
    State:          running
    Capacity:       395.20 GB
    Allocation:     30.88 GB
    Available:      364.33 GB

    virsh # pool-destroy images_dir3
    Pool images_dir3 destroyed

At this point the images_dir3 pool should be gone (because it was
transient) and we should be able to create a new pool with the same name:

    virsh # pool-create-as images_dir3 dir - - - - "/home/images2"
    Pool images_dir3 created

    virsh # pool-info images_dir3
    Name:           images_dir3
    UUID:           90301885-94eb-4ca7-14c2-f30b25a29a36
    error: Storage pool not found

The new pool got the same UUID as the first one, but we didn't specify
one. libvirt should have picked a random UUID, but it didn't.

It turned out that virStoragePoolIsPersistent leaks a reference to the
storage pool object (actually remoteDispatchStoragePoolIsPersistent does).
As a result, pool-destroy doesn't remove the virStoragePool for the
"images_dir3" pool from the virConnectPtr's storagePools hash on libvirtd's
side. Then the second pool-create-as get's the stale virStoragePool object
associated with the "images_dir3" name. But this object has the old UUID.

This commit ensures that all get_nonnull_* and make_nonnull_* calls for
libvirt objects are matched properly with vir*Free calls. This fixes the
reference leaks and the reported problem.

All remoteDispatch*IsActive and remoteDispatch*IsPersistent functions were
affected. But also remoteDispatchDomainMigrateFinish2 was affected in the
success path. I wonder why that didn't surface earlier. Probably because
domainMigrateFinish2 is executed on the destination host and in the common
case this connection is opened especially for the migration and gets closed
after the migration is done. So there was no chance to run into a problem
because of the leaked reference.
2010-06-16 23:06:12 +02:00
Matthias Bolte
02d57c2bce esx: Update case insensitive .vmx tests
Commit b9efc7dc3b made virFileHasSuffix
case insensitive. Honor this in the tests by switching vmdk to VMDK.
2010-06-16 23:06:12 +02:00
Matthias Bolte
b0f414c67b esx: Accept 'disk' as harddisk device type in .vmx files 2010-06-16 23:06:12 +02:00
Eric Blake
02988742b7 phyp: sed cleanups
* src/phyp/phyp_driver.c (phypNumDomainsGeneric): Avoid glob
collision by quoting sed argument.
(phypDomainSetCPU): Avoid non-portable \+ in sed.
(phypGetVIOSPartitionID, phypDiskType, phypListDomainsGeneric)
(phypListDefinedDomains): Micro-optimize anchored substitutions.
2010-06-16 14:08:44 -06:00
Justin Clift
043f05fda4 virsh: mark autostart answers for translation
This is a trivial fix for several autostart yes/no strings that
weren't correctly marked for translation.
2010-06-15 15:36:27 -06:00