Commit Graph

5827 Commits

Author SHA1 Message Date
Philipp Hahn
b14e7d2a16 Fix URL-escaping for domainDefine
'+' in strings get translated to ' ' when editing domains.
While xenDaemonDomainCreateXML() did URL-escape the sexpr,
xenDaemonDomainDefineXML() did not.

Remove the explicit urlencode() in xenDaemonDomainCreateXML() and add
the direct encoding calls to xend_op_ext() because it calls xend_post()
which uses "Content-Type: application/x-www-form-urlencoded". According
to <http://www.w3.org/TR/html4/interact/forms.html#h-17.13.4.1> this
requires all parameters to be url-encoded as specified in rfc1738.

Notice: virBufferAsprintf(..., "%s=%s", ...) is again replaced by three
calls to virBufferURIEncodeString() and virBufferAddChar() because '='
is a "reserved" character, which would get escaped by
virBufferURIEncodeString(), which - by the way - escapes anything not
c_isalnum().

Signed-off-by: Philipp Hahn <hahn@univention.de>
2011-09-09 15:18:39 +08:00
Peter Krempa
498d783387 xml: Change virtual file names of xml documents parsed in memory
While parsing XML strings from memory, the previous convention in
libvirt was to set the virtual file name to "domain.xml" or something
similar. This could potentialy trick the user into looking for a file
named domain.xml on the disk in an attempt to fix the error.

This patch changes these filenames to something that can't be as easily
confused for a valid filename.

Examples of error messages:
---------------------------
Error while loading file from disk:

15:07:59.015: 527: error : catchXMLError:709 : /path/to/domain.xml:1: StartTag: invalid element name
<domain type='kvm'><
--------------------^

Error while parsing definition in memory:

15:08:43.581: 525: error : catchXMLError:709 : (domain definition):2: error parsing attribute name
  <name>vm1</name>
--^
2011-09-08 17:20:33 +01:00
Eric Blake
2acd4a1640 snapshot: fix regression with system checkpoints
Regression introduced in commit d6f6b2d194.  Running
'virsh snapshot-create dom' would mistakenly report that
disks can only be specified for disk snapshots.

* src/conf/domain_conf.c (virDomainSnapshotDefParseString): Only
give error about no disk support when <disk> was found.
2011-09-08 14:51:48 +01:00
Eric Blake
6c14439e51 snapshot: new APIs for inspecting snapshot object
These functions access internals of the opaque object, and do
not need any rpc counterpart.  It could be argued that we should
have provided these when snapshot objects were first introduced,
since all the other vir*Ptr objects have at least a GetName accessor.

* include/libvirt/libvirt.h.in (virDomainSnapshotGetName)
(virDomainSnapshotGetDomain, virDomainSnapshotGetConnect): Declare.
* src/libvirt.c (virDomainSnapshotGetName)
(virDomainSnapshotGetDomain, virDomainSnapshotGetConnect): New
functions.
* src/libvirt_public.syms: Export them.
2011-09-08 13:15:13 +01:00
Soren Hansen
9c7283166f Don't treat pci_system_init failure as fatal if no PCI bus is present
Xen PV domU's have no PCI bus. node_device_udev.c calls pci_system_init
which looks for /sys/bus/pci. If it does not find /sys/bus/pci (which it
won't in a Xen PV domU) it returns unsuccesfully (ENOENT), which libvirt
considers fatal. This makes libvirt unusable in this environment, even
though there are plenty of valid virtualisation options that work
there (LXC, UML, and QEmu spring to mind)

https://bugzilla.redhat.com/show_bug.cgi?id=709471

Signed-off-by: Soren Hansen <soren@linux2go.dk>
2011-09-08 11:36:18 +01:00
Alex Jia
db8ffc2dfb rpc: avoid memory leak on virNetTLSContextValidCertificate
* src/rpc/virnettlscontext.c: fix memory leak on
  virNetTLSContextValidCertificate.

* Detected in valgrind run:

==25667==
==25667== 6,085 (44 direct, 6,041 indirect) bytes in 1 blocks are definitely
lost in loss record 326 of 351
==25667==    at 0x4005447: calloc (vg_replace_malloc.c:467)
==25667==    by 0x4F2791F3: _asn1_add_node_only (structure.c:53)
==25667==    by 0x4F27997A: _asn1_copy_structure3 (structure.c:421)
==25667==    by 0x4F276A50: _asn1_append_sequence_set (element.c:144)
==25667==    by 0x4F2743FF: asn1_der_decoding (decoding.c:1194)
==25667==    by 0x4F22B9CC: gnutls_x509_crt_import (x509.c:229)
==25667==    by 0x805274B: virNetTLSContextCheckCertificate
(virnettlscontext.c:1009)
==25667==    by 0x804DE32: testTLSSessionInit (virnettlscontexttest.c:693)
==25667==    by 0x804F14D: virtTestRun (testutils.c:140)
==25667==
==25667== 23,188 (88 direct, 23,100 indirect) bytes in 11 blocks are definitely
lost in loss record 346 of 351
==25667==    at 0x4005447: calloc (vg_replace_malloc.c:467)
==25667==    by 0x4F22B841: gnutls_x509_crt_init (x509.c:50)
==25667==    by 0x805272B: virNetTLSContextCheckCertificate
(virnettlscontext.c:1003)
==25667==    by 0x804DDD1: testTLSSessionInit (virnettlscontexttest.c:673)
==25667==    by 0x804F14D: virtTestRun (testutils.c:140)

* How to reproduce?
% cd libvirt && ./configure && make && make -C tests valgrind
or
% valgrind -v --leak-check=full ./tests/virnettlscontexttest

Signed-off-by: Alex Jia <ajia@redhat.com>
2011-09-08 11:06:39 +01:00
Alex Jia
4b4e4a69a8 libxl: avoid a dereference of a null pointer
Variable 'l_disk' initialized to a null pointer value, control jumps to 'case
VIR_DOMAIN_DISK_DEVICE_DISK and then taking false branch, Within the expansion
of the macro 'libxlError': Field access results in a dereference of a null
pointer (loaded from variable 'l_disk').

* src/libxl/libxl_driver.c: Field access results in a dereference of a null
  pointer (loaded from variable 'l_disk')

Signed-off-by: Alex Jia <ajia@redhat.com>
2011-09-08 11:00:30 +01:00
Eric Blake
6977fd95bf blockinfo: fix qemu regression in handling disk name
Regression introduced in commit 89b6284fd, due to an incorrect
conversion to the new means of converting disk names back to
the correct object.

* src/qemu/qemu_driver.c (qemuDomainGetBlockInfo): Avoid NULL deref.
2011-09-08 10:52:43 +01:00
Matthias Bolte
4d3d3e475f esx: Fix managed object lookup with optional occurrence
Exit early if managed object is not found, instead of dereferencing
a NULL pointer and triggering a segfault.
2011-09-08 10:36:48 +02:00
Michal Privoznik
d1a366be05 Threadpool: Initialize new dynamic workers
Although we were initializing worker threads during pool creating,
we missed this during virThreadPoolSendJob. This bug led to segmenation
fault as worker thread free() given argument.
2011-09-07 14:23:26 +02:00
Peter Krempa
e0a07bb1f2 link-state: qemu: Add net intf modification to virUpdateDeviceFlags
This patch enables modifying network device configuration using the
virUpdateDeviceFlags API method. Matching of devices is accomplished
using MAC addresses.

While updating live configuration of a running domain, the user is
allowed only to change link state of the interface. Additional
modifications may be added later. For now the code checks for
unsupported changes and thereafter changes the link state, if
applicable.

When updating persistent configuration of guest's network interface the
whole configuration (except for the MAC address) may be modified and
is stored for the next startup.

* src/qemu/qemu_driver.c   - Add dispatching of virUpdateDevice for
                             network devices update (live/config)
* src/qemu/qemu_hotplug.c  - add setting of initial link state on live
                             device addition
                           - add function to change network device
                             configuration. By now it supports only
                             changing of link state
* src/qemu/qemu_hotplug.h  - Headers to above functions
* src/qemu/qemu_process.c  - set link states before virtual machine
                             start. Qemu does not support setting of
                             this on the command line.
2011-09-06 16:23:47 +08:00
Peter Krempa
8277c15151 link-state: qemu: Add monitor handling for link state modification
This patch adds handlers for modification of guest's interface
link state. Both HMP and QMP commands are supported, but as the
link state functionality is from the beginning supported in QMP
the HMP code will probably never be used.
2011-09-06 16:18:57 +08:00
Peter Krempa
edd1295e1d link-state: conf: Add element to XML for controling link state
A new element is introduced to XML that allows to control
state of virtual network interfaces in hypervisors.

Live modification of the link state allows networking tools
propagate topology changes to guest OS or testing of
scenarios in complex (virtual) networks.

This patch adds elements to XML grammars and parsing and generating
code.
2011-09-06 16:08:15 +08:00
Peter Krempa
c246b02586 link-state: util: Add equality comparison functions for structures
This patch adds functions to compare structures containing network
device configuration for equality. They serve for the purpose of
disallowing unsupported changes to live network devices.
2011-09-06 16:05:53 +08:00
Peter Krempa
9fd3bb7a88 XML: Improve XML parsing error messages
This patch modifies error handling function for the XML parser provided
by libxml2.

Originaly only a line number and error message were logged. With this
new error handler function, the user is provided with a more complex
description of the parsing error.

Context of the error is printed in libXML2 style and filename of the
file, that caused the error is printed. Example of an parse error:

13:41:36.262: 16032: error : catchXMLError:706 :
/etc/libvirt/qemu/rh_bad.xml:58: Opening and ending tag mismatch: name
line 2 and domain
</domain>
---------^

Context of the error gives the user hints that may help to quickly
locate a corrupt xml file.

fixes BZs:
----------
Bug 708735 - [RFE] Show column and line on XML parsing error
https://bugzilla.redhat.com/show_bug.cgi?id=708735

Bug 726771 - libvirt does not specify problem file if persistent xml is
invalid
https://bugzilla.redhat.com/show_bug.cgi?id=726771
2011-09-06 15:48:22 +08:00
Marc-André Lureau
2e0dbaad9b redirdev: allows to specify device address
It is important to be able to attach USB redirected devices to a
particular controller (one that supports USB2 for instance).
Without this patch, only the default bus was used.

     <redirdev bus='usb' type='spicevmc'>
       <address type='usb' bus='0' port='4'/>
     </redirdev>
2011-09-06 15:12:52 +08:00
Daniel Veillard
59e22b7258 latency: fix make check for remote protocol structs and numbers 2011-09-06 15:11:44 +08:00
Osier Yang
ee0d8c3b5c latency: Implemente internal API for qemu driver 2011-09-06 12:02:51 +08:00
Osier Yang
2f58ba8996 latency: Update monitor functions for new latency fields
The mainly changes are:

1) Update qemuMonitorGetBlockStatsInfo and it's children (Text/JSON)
   functions to return the value of new latency fields.
2) Add new function qemuMonitorGetBlockStatsParamsNumber, which is
   to count how many parameters the underlying QEMU supports.
3) Update virDomainBlockStats in src/qemu/qemu_driver.c to be
   compatible with the changes by 1).
2011-09-06 12:02:51 +08:00
Osier Yang
efa7fc9f75 latency: Wire up the remote protocol 2011-09-06 12:02:51 +08:00
Osier Yang
1f80c3eb86 latency: Implemente the public API 2011-09-06 12:02:51 +08:00
Osier Yang
1c622a3d53 latency: Define the internal driver callback 2011-09-06 12:02:51 +08:00
Osier Yang
c843478ec8 latency: Define new public API and structure 2011-09-06 12:02:51 +08:00
Michal Privoznik
d38897a5d4 qemu: Deal with stucked qemu on daemon startup
If libvirt daemon gets restarted and there is (at least) one
unresponsive qemu, the startup procedure hangs up. This patch creates
one thread per vm in which we try to reconnect to monitor. Therefore,
blocking in one thread will not affect other APIs.
2011-09-05 18:14:08 +02:00
Michal Privoznik
3005cacb69 qemu: Introduce job queue size limit
This patch creates an optional BeginJob queue size limit. When
active, all other attempts above level will fail. To set this
feature assign desired value to max_queued variable in qemu.conf.
Setting it to 0 turns it off.
2011-09-05 18:14:08 +02:00
Michal Privoznik
597fe3cee6 daemon: Create priority workers pool
This patch annotates APIs with low or high priority.
In low set MUST be all APIs which might eventually access monitor
(and thus block indefinitely). Other APIs may be marked as high
priority. However, some must be (e.g. domainDestroy).

For high priority calls (HPC), there are some high priority workers
(HPW) created in the pool. HPW can execute only HPC, although normal
worker can process any call regardless priority. Therefore, only those
APIs which are guaranteed to end in reasonable small amount of time
can be marked as HPC.

The size of this HPC pool is static, because HPC are expected to end
quickly, therefore jobs assigned to this pool will be served quickly.
It can be configured in libvirtd.conf via prio_workers variable.
Default is set to 5.

To mark API with low or high priority, append priority:{low|high} to
it's comment in src/remote/remote_protocol.x. This is similar to
autogen|skipgen. If not marked, the generator assumes low as default.
2011-09-05 18:14:08 +02:00
Eric Blake
6337989089 snapshot: use SELinux and lock manager with external snapshots
With this, it is now possible to create external snapshots even
when SELinux is enforcing, and to protect the new file with a
lock manager.

* src/qemu/qemu_driver.c
(qemuDomainSnapshotCreateSingleDiskActive): Create and register
new file with proper permissions and locks.
(qemuDomainSnapshotCreateDiskActive): Update caller.
2011-09-05 07:03:05 -06:00
Eric Blake
9c21b44131 snapshot: wire up live qemu disk snapshots
Lots of earlier patches led up to this point - the qemu snapshot_blkdev
monitor command can now be controlled by libvirt!  Well, insofar as
SELinux doesn't prevent qemu from open(O_CREAT) on the files.  There's
still some followup work before things work with SELinux enforcing,
but this patch is big enough to post now.

There's still room for other improvements, too (for example, taking a
disk snapshot of an inactive domain, by using qemu-img for both internal
and external snapshots; wiring up delete and revert control, including
additional flags from my RFC; supporting active QED disk snapshots;
supporting per-storage-volume snapshots such as LVM or btrfs snapshots;
etc.).  But this patch is the one that proves the new XML works!

* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML): Wire in
active disk snapshots.
(qemuDomainSnapshotDiskPrepare)
(qemuDomainSnapshotCreateDiskActive)
(qemuDomainSnapshotCreateSingleDiskActive): New functions.
2011-09-05 07:03:05 -06:00
Eric Blake
e702b5bab1 snapshot: wire up new qemu monitor command
No one uses this yet, but it will be important once
virDomainSnapshotCreateXML learns a VIR_DOMAIN_SNAPSHOT_DISK_ONLY
flag, and the xml allows passing in the new file names.

* src/qemu/qemu_monitor.h (qemuMonitorDiskSnapshot): New prototype.
* src/qemu/qemu_monitor_text.h (qemuMonitorTextDiskSnapshot):
Likewise.
* src/qemu/qemu_monitor_json.h (qemuMonitorJSONDiskSnapshot):
Likewise.
* src/qemu/qemu_monitor.c (qemuMonitorDiskSnapshot): New
function.
* src/qemu/qemu_monitor_json.c (qemuMonitorJSONDiskSnapshot):
Likewise.
2011-09-05 07:03:05 -06:00
Eric Blake
c111517a88 snapshot: make it possible to audit external snapshot
Snapshots alter the set of disk image files opened by qemu, so
they must be audited.  But they don't involve a full disk definition
structure, just the new filename.  Make the next patch easier by
refactoring the audit routines to just operate on file name.

* src/conf/domain_audit.h (virDomainAuditDisk): Update prototype.
* src/conf/domain_audit.c (virDomainAuditDisk): Act on strings,
not definition structures.
(virDomainAuditStart): Update caller.
* src/qemu/qemu_hotplug.c (qemuDomainChangeEjectableMedia)
(qemuDomainAttachPciDiskDevice, qemuDomainAttachSCSIDisk)
(qemuDomainAttachUsbMassstorageDevice)
(qemuDomainDetachPciDiskDevice, qemuDomainDetachDiskDevice):
Likewise.
2011-09-05 07:03:05 -06:00
Eric Blake
7807e05d43 snapshot: reject unimplemented disk snapshot features
My RFC for snapshot support [1] proposes several rules for when it is
safe to delete or revert to an external snapshot, predicated on
the existence of new API flags.  These will be incrementally added
in future patches, but until then, blindly mishandling a disk
snapshot risks corrupting internal state, so it is better to
outright reject the attempts until the other pieces are in place,
thus incrementally relaxing the restrictions added in this patch.

[1] https://www.redhat.com/archives/libvir-list/2011-August/msg00361.html

* src/qemu/qemu_driver.c (qemuDomainSnapshotCountExternal): New
function.
(qemuDomainUndefineFlags, qemuDomainSnapshotDelete): Use it to add
safety valve.
(qemuDomainRevertToSnapshot, qemuDomainSnapshotCreateXML): Add safety
valve.
2011-09-05 07:03:04 -06:00
Eric Blake
e03a62b456 snapshot: add flag for requesting disk snapshot
Prior to this patch, <domainsnapshot>/<disks> was ignored.  This
changes it to be an error unless an explicit disk snapshot is
requested (a future patch may relax things if it turns out to
be useful to have a <disks> specification alongside a system
checkpoint).

* include/libvirt/libvirt.h.in
(VIR_DOMAIN_SNAPSHOT_CREATE_DISK_ONLY): New flag.
* src/libvirt.c (virDomainSnapshotCreateXML): Document it.
* src/esx/esx_driver.c (esxDomainSnapshotCreateXML): Disk
snapshots not supported yet.
* src/vbox/vbox_tmpl.c (vboxDomainSnapshotCreateXML): Likewise.
* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML): Likewise.
2011-09-05 07:03:04 -06:00
Eric Blake
89b6284fd9 snapshot: also support disks by path
I got confused when 'virsh domblkinfo dom disk' required the
path to a disk (which can be ambiguous, since a single file
can back multiple disks), rather than the unambiguous target
device name that I was using in disk snapshots.  So, in true
developer fashion, I went for the best of both worlds - all
interfaces that operate on a disk (aka block) now accept
either the target name or the unambiguous path to the backing
file used by the disk.

* src/conf/domain_conf.h (virDomainDiskIndexByName): Add
parameter.
(virDomainDiskPathByName): New prototype.
* src/libvirt_private.syms (domain_conf.h): Export it.
* src/conf/domain_conf.c (virDomainDiskIndexByName): Also allow
searching by path, and decide whether ambiguity is okay.
(virDomainDiskPathByName): New function.
(virDomainDiskRemoveByName, virDomainSnapshotAlignDisks): Update
callers.
* src/qemu/qemu_driver.c (qemudDomainBlockPeek)
(qemuDomainAttachDeviceConfig, qemuDomainUpdateDeviceConfig)
(qemuDomainGetBlockInfo, qemuDiskPathToAlias): Likewise.
* src/qemu/qemu_process.c (qemuProcessFindDomainDiskByPath):
Likewise.
* src/libxl/libxl_driver.c (libxlDomainAttachDeviceDiskLive)
(libxlDomainDetachDeviceDiskLive, libxlDomainAttachDeviceConfig)
(libxlDomainUpdateDeviceConfig): Likewise.
* src/uml/uml_driver.c (umlDomainBlockPeek): Likewise.
* src/xen/xend_internal.c (xenDaemonDomainBlockPeek): Likewise.
* docs/formatsnapshot.html.in: Update documentation.
* tools/virsh.pod (domblkstat, domblkinfo): Likewise.
* docs/schemas/domaincommon.rng (diskTarget): Tighten pattern on
disk targets.
* docs/schemas/domainsnapshot.rng (disksnapshot): Update to match.
* tests/domainsnapshotxml2xmlin/disk_snapshot.xml: Update test.
2011-09-05 07:03:04 -06:00
Eric Blake
d6f6b2d194 snapshot: add <disks> to snapshot xml
Adds an optional element to <domainsnapshot>, which will be used
to give user control over external snapshot filenames on input,
and specify generated filenames on output.

For now, no driver accepts this element; that will come later.

<domainsnapshot>
  ...
  <disks>
    <disk name='vda' snapshot='no'/>
    <disk name='vdb' snapshot='internal'/>
    <disk name='vdc' snapshot='external'>
      <driver type='qcow2'/>
      <source file='/path/to/new'/>
    </disk>
  </disks>
  <domain>
    ...
    <devices>
      <disk ...>
        <driver name='qemu' type='raw'/>
        <target dev='vdc'/>
        <source file='/path/to/old'/>
      </disk>
    </devices>
  </domain>
</domainsnapshot>

* src/conf/domain_conf.h (_virDomainSnapshotDiskDef): New type.
(_virDomainSnapshotDef): Add new elements.
(virDomainSnapshotAlignDisks): New prototype.
* src/conf/domain_conf.c (virDomainSnapshotDiskDefClear)
(virDomainSnapshotDiskDefParseXML, disksorter)
(virDomainSnapshotAlignDisks): New functions.
(virDomainSnapshotDefParseString): Parse new fields.
(virDomainSnapshotDefFree): Clean them up.
(virDomainSnapshotDefFormat): Output them.
* src/libvirt_private.syms (domain_conf.h): Export new function.
* docs/schemas/domainsnapshot.rng (domainsnapshot, disksnapshot):
Add more xml.
* docs/formatsnapshot.html.in: Document it.
* tests/domainsnapshotxml2xmlin/disk_snapshot.xml: New test.
* tests/domainsnapshotxml2xmlout/disk_snapshot.xml: Update.
2011-09-05 07:03:04 -06:00
Eric Blake
5b30b08d66 snapshot: support extra state in snapshots
In order to distinguish disk snapshots from system checkpoints, a
new state value that is only valid for snapshots is helpful.

* include/libvirt/libvirt.h.in (VIR_DOMAIN_LAST): New placeholder.
* src/conf/domain_conf.h (virDomainSnapshotState): New enum mapping.
(VIR_DOMAIN_DISK_SNAPSHOT): New internal enum value.
* src/conf/domain_conf.c (virDomainState): Use placeholder.
(virDomainSnapshotState): Extend mapping by one for use in snapshot.
(virDomainSnapshotDefParseString, virDomainSnapshotDefFormat):
Handle new state.
(virDomainObjSetState, virDomainStateReasonToString)
(virDomainStateReasonFromString): Avoid compiler warnings.
* tools/virsh.c (vshDomainState, vshDomainStateReasonToString):
Likewise.
* src/libvirt_private.syms (domain_conf.h): Export new functions.
* docs/schemas/domainsnapshot.rng: Tighten state definition.
* docs/formatsnapshot.html.in: Document it.
* tests/domainsnapshotxml2xmlout/disk_snapshot.xml: New test.
2011-09-05 07:03:04 -06:00
Eric Blake
6f66423e17 snapshot: allow halting after snapshot
Since a snapshot is fully recoverable, it is useful to have a
snapshot as a means of hibernating a guest, then reverting to
the snapshot to wake the guest up.  This mode of usage is
similar to 'virsh save/virsh restore', except that virsh
save uses an external file while virsh snapshot keeps the
vm state internal to a qcow2 file.  However, it only works on
persistent domains.

In the usage pattern of snapshot/revert for hibernating a guest,
there is no need to keep the guest running between the two points
in time, especially since that would generate runtime state that
would just be discarded.  Add a flag to make it possible to
stop the domain after the snapshot has completed.

* include/libvirt/libvirt.h.in (VIR_DOMAIN_SNAPSHOT_CREATE_HALT):
New flag.
* src/libvirt.c (virDomainSnapshotCreateXML): Document it.
* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML)
(qemuDomainSnapshotCreateActive): Implement it.
2011-09-05 07:03:04 -06:00
Eric Blake
3d77d0a644 snapshot: introduce new deletion flag
Reverting to a state prior to an external snapshot risks
corrupting any other branches in the snapshot hierarchy that
were using the snapshot as a read-only backing file.  So
disk snapshot code will default to preventing reverting to
a snapshot that has any children, meaning that deleting just
the children of a snapshot becomes a useful operation in
preparing that snapshot for being a future reversion target.
The code for the new flag is simple - it's one less deletion,
plus a tweak to keep the current snapshot correct.

* include/libvirt/libvirt.h.in
(VIR_DOMAIN_SNAPSHOT_DELETE_CHILDREN_ONLY): New flag.
* src/libvirt.c (virDomainSnapshotDelete): Document it, and
enforce mutual exclusion.
* src/qemu/qemu_driver.c (qemuDomainSnapshotDelete): Implement
it.
2011-09-05 07:03:04 -06:00
Eric Blake
e91d27ee45 snapshot: reject transient disks where code is not ready
The previous patch introduced new config, but if a hypervisor does
not support that new config, someone can write XML that does not
behave as documented.  This prevents some of those cases by
explicitly rejecting transient disks for several hypervisors.

Disk snapshots will require a new flag to actually affect a snapshot
creation, so there's not much to reject there.

* src/qemu/qemu_command.c (qemuBuildDriveStr): Reject transient
disks for now.
* src/libxl/libxl_conf.c (libxlMakeDisk): Likewise.
* src/xenxs/xen_sxpr.c (xenFormatSxprDisk): Likewise.
* src/xenxs/xen_xm.c (xenFormatXMDisk): Likewise.
2011-09-05 07:03:04 -06:00
Eric Blake
471235307f snapshot: additions to domain xml for disks
As discussed here:
https://www.redhat.com/archives/libvir-list/2011-August/msg00361.html
https://www.redhat.com/archives/libvir-list/2011-August/msg00552.html

Adds snapshot attribute and transient sub-element:

<devices>
  <disk type=... snapshot='no|internal|external'>
    ...
    <transient/>
  </disk>
</devices>

* docs/schemas/domaincommon.rng (snapshot): New define.
(disk): Add snapshot and persistent attributes.
* docs/formatdomain.html.in: Document them.
* src/conf/domain_conf.h (virDomainDiskSnapshot): New enum.
(_virDomainDiskDef): New fields.
* tests/qemuxml2argvdata/qemuxml2argv-disk-transient.xml: New
test of rng, no args counterpart until qemu support is complete.
* tests/qemuxml2argvdata/qemuxml2argv-disk-snapshot.args: New
file, snapshot attribute does not affect args.
* tests/qemuxml2argvdata/qemuxml2argv-disk-snapshot.xml: Likewise.
* tests/qemuxml2argvtest.c (mymain): Run new test.
2011-09-05 07:03:04 -06:00
Lei Li
5a1f272875 Check for source conflicts in storage pools
Fix bug #611823 storage driver should prohibit pools with duplicate
underlying storage.

Add internal API virStoragePoolSourceFindDuplicate() to do uniqueness
check based on source location infomation for pool type.

* AUTHORS: add Lei Li
2011-09-05 15:52:03 +08:00
Marc-André Lureau
360aaafc63 Default USB device is on slot 1 function 2
Fix qemuAssignDevicePCISlots() and the associated regression tests
2011-09-05 15:03:27 +08:00
Marc-André Lureau
07901bf235 PIIX3 USB controller is on function 2
Current code reserves slot 1 function 2 even if there is a user
defined PIIX3 USB controller there.
2011-09-05 14:48:56 +08:00
Philipp Hahn
8be115ff80 Fix localtime handling for Xen-PV domains
At least Xen-3.4.3 translates the /vm/localtime SXPR value to
/domain/platform/localtime and /domain/image/{linux,hvm}/localtime when
the domain is defined.  When reading back that information libvirt only
handles HVM domains, but not PV domains: This results in libvirtd always
returning
    <clock offset="utc"/>
while Xend used (localtime 1).

For PV domains use /domain/image/linux/localtime.
2011-09-05 14:22:36 +08:00
Eric Blake
973fcd8fd3 snapshot: store qemu domain details in xml
When reverting to a snapshot, the inactive domain configuration
has to be rolled back to what it was at the time of the snapshot.
Additionally, if the VM is active and the snapshot was active,
this now adds a failure if the two configurations are ABI
incompatible, rather than risking qemu confusion.

A future patch will add a VIR_DOMAIN_SNAPSHOT_FORCE flag, which
will be required for two risky code paths - reverting to an
older snapshot that lacked full domain information, and reverting
from running to a live snapshot that requires starting a new qemu
process.  Any reverting that stops a running vm is also a form
of data loss (discarding the current running state to go back in
time), but as that is what reversion usually implies, it is
probably not worth requiring a force flag.

* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML): Copy out
domain.
(qemuDomainSnapshotCreateXML, qemuDomainRevertToSnapshot): Perform
ABI compatibility checks.
2011-09-03 08:27:07 -06:00
Eric Blake
8352e04d7e snapshot: correctly escape generated xml
Commit 69278878 fixed one direction of arbitrarily-named snapshots,
but not the round trip path.  While auditing domain_conf, I found
a couple other instances that weren't escaping arbitrary strings.

* src/conf/domain_conf.c (virDomainFSDefFormat)
(virDomainGraphicsListenDefFormat, virDomainSnapshotDefFormat):
Escape arbitrary strings.
2011-09-03 08:10:09 -06:00
Eric Blake
f609cb85ca snapshot: allow full domain xml in snapshot
Just like VM saved state images (virsh save), snapshots MUST
track the inactive domain xml to detect any ABI incompatibilities.

The indentation is not perfect, but functionality comes before form.

Later patches will actually supply a full domain; for now, this
wires up the storage to support one, but doesn't ever generate one
in dumpxml output.

Happily, libvirt.c was already rejecting use of VIR_DOMAIN_XML_SECURE
from read-only connections, even though before this patch, there was
no information to be secured by the use of that flag.

And while we're at it, mark the libvirt snapshot metadata files
as internal-use only.

* src/libvirt.c (virDomainSnapshotGetXMLDesc): Document flag.
* src/conf/domain_conf.h (_virDomainSnapshotDef): Add member.
(virDomainSnapshotDefParseString, virDomainSnapshotDefFormat):
Update signature.
* src/conf/domain_conf.c (virDomainSnapshotDefFree): Clean up.
(virDomainSnapshotDefParseString): Optionally parse domain.
(virDomainSnapshotDefFormat): Output full domain.
* src/esx/esx_driver.c (esxDomainSnapshotCreateXML)
(esxDomainSnapshotGetXMLDesc): Update callers.
* src/vbox/vbox_tmpl.c (vboxDomainSnapshotCreateXML)
(vboxDomainSnapshotGetXMLDesc): Likewise.
* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML)
(qemuDomainSnapshotLoad, qemuDomainSnapshotGetXMLDesc)
(qemuDomainSnapshotWriteMetadata): Likewise.
* docs/formatsnapshot.html.in: Rework doc example.
Based on a patch by Philipp Hahn.
2011-09-03 08:09:35 -06:00
Eric Blake
0ce68c66c7 snapshot: refactor domain xml output
Minor semantic change - allow domain xml to be generated in place
within a larger buffer, rather than having to go through a
temporary string.

* src/conf/domain_conf.c (virDomainDefFormatInternal): Add
parameter.
(virDomainDefFormat, virDomainObjFormat): Update callers.
2011-09-02 21:57:34 -06:00
Eric Blake
e2fb96d92b snapshot: prevent migration from stranding snapshot data
Migration is another case of stranding metadata.  And since
snapshot metadata is arbitrarily large, there's no way to
shoehorn it into the migration cookie of migration v3.

This patch consolidates two existing locations for migration
validation into one helper function, then enhances that function
to also do the new checks.  If we could always trust the source
to validate migration, then the destination would not have to
do anything; but since older servers that did not do checking
can migrate to newer destinations, we have to repeat some of
the same checks on the destination; meanwhile, we want to
detect failures as soon as possible.  With migration v2, this
means that validation will reject things at Prepare on the
destination if the XML exposes the problem, otherwise at Perform
on the source; with migration v3, this means that validation
will reject things at Begin on the source, or if the source
is old and the XML exposes the problem, then at Prepare on the
destination.

This patch is necessarily over-strict.  Once a later patch
properly handles auto-cleanup of snapshot metadata on the
death of a transient domain, then the only time we actually
need snapshots to prevent migration is when using the
--undefinesource flag on a persistent source domain.

It is possible to recreate snapshot metadata on the destination
with VIR_DOMAIN_SNAPSHOT_CREATE_REDEFINE and
VIR_DOMAIN_SNAPSHOT_CREATE_CURRENT.  But for now, that is limited,
since if we delete the snapshot metadata prior to migration,
then we won't know the name of the current snapshot to pass
along; and if we delete the snapshot metadata after migration
and use the v3 migration cookie to pass along the name of the
current snapshot, then we need a way to bypass the fact that
this patch refuses migration with snapshot metadata present.

So eventually, we may have to introduce migration protocol v4
that allows feature negotiation and an arbitrary number of
handshake exchanges, so as to pass as many rpc calls as needed
to transfer all the snapshot xml hierarchy.

But all of that is thoughts for the future; for now, the best
course of action is to quit early, rather than get into a
funky state of stale metadata; then relax restrictions later.

* src/qemu/qemu_migration.h (qemuMigrationIsAllowed): Make static.
* src/qemu/qemu_migration.c (qemuMigrationIsAllowed): Alter
signature, and allow checks for both outgoing and incoming.
(qemuMigrationBegin, qemuMigrationPrepareAny)
(qemuMigrationPerformJob): Update callers.
2011-09-02 21:57:34 -06:00
Eric Blake
19f8c980ef snapshot: support new undefine flags in qemu
A nice benefit of deleting all snapshots at undefine time is that
you don't have to do any reparenting or subtree identification - since
everything goes, this is an O(n) process, whereas using multiple
virDomainSnapshotDelete calls would be O(n^2) or worse.  But it is
only doable for snapshot metadata, where we are in control of the
data being deleted; for the actual snapshots, there's too much
likelihood of something going wrong, and requiring even more API
calls to figure out what failed in the meantime, so callers are
better off deleting the snapshot data themselves one snapshot at
a time where they can deal with failures as they happen.

* src/qemu/qemu_driver.c (qemuDomainUndefineFlags): Honor new flags.
2011-09-02 21:57:34 -06:00
Eric Blake
3881a47088 snapshot: cache qemu-img location
As more clients start to want to know this information, doing
a PATH stat walk and malloc for every client adds up.

We are only caching the location, not the capabilities, so even
if qemu-img is updated in the meantime, it will still probably
live in the same location.  So there is no need to worry about
clearing this particular cache.

* src/qemu/qemu_conf.h (qemud_driver): Add member.
* src/qemu/qemu_driver.c (qemudShutdown): Cleanup.
(qemuFindQemuImgBinary): Add an argument, and cache result.
(qemuDomainSnapshotForEachQcow2, qemuDomainSnapshotDiscard)
(qemuDomainSnapshotCreateInactive, qemuDomainSnapshotRevertInactive)
(qemuDomainSnapshotCreateXML, qemuDomainRevertToSnapshot): Update
callers.
2011-09-02 21:57:33 -06:00
Eric Blake
8055e5af82 snapshot: refactor some qemu code
Prepare for code sharing.  No semantic change.

* src/qemu/qemu_driver.c (qemuFindQemuImgBinary)
(qemuDomainSnapshotWriteMetadata)
(qemuDomainSnapshotDiscard)
(qemuDomainSnapshotForEachQcow2): Float up.
(qemuDomainSnapshotDiscardDescendant): Likewise, and rename...
(qemuDomainSnapshotDiscardAll): ...for generic use.
(qemuDomainSnapshotDelete): Update caller.
2011-09-02 21:57:33 -06:00
Eric Blake
282fe1f08c snapshot: prevent stranding snapshot data on domain destruction
Just as leaving managed save metadata behind can cause problems
when creating a new domain that happens to collide with the name
of the just-deleted domain, the same is true of leaving any
snapshot metadata behind.  For safety sake, extend the semantic
change of commit b26a9fa9 to also cover snapshot metadata as a
reason to reject undefining an inactive domain.  A future patch
will make sure that shutdown of a transient domain automatically
deletes snapshot metadata (whether by destroy, shutdown, or
guest-initiated action).  Management apps of transient domains
should take care to capture xml of snapshots, if it is necessary
to recreate the snapshot metadata on a later transient domain
with the same name and uuid.

This also documents a new flag that hypervisors can choose to
support as a shortcut for taking care of the metadata as part of
the undefine process; however, nontrivial driver support for these
flags will be deferred to future patches.

Note that ESX and VBox can never be transient; therefore, they
do not have to worry about automatic cleanup after shutdown
(the persistent domain still remains); likewise they never
store snapshot metadata, so the undefine flag is trivial.
The nontrivial work remaining is thus in the qemu driver.

* include/libvirt/libvirt.h.in
(VIR_DOMAIN_UNDEFINE_SNAPSHOTS_METADATA): New flag.
* src/libvirt.c (virDomainUndefine, virDomainUndefineFlags):
Document new limitations and flag.
* src/esx/esx_driver.c (esxDomainUndefineFlags): Trivial
implementation.
* src/vbox/vbox_tmpl.c (vboxDomainUndefineFlags): Likewise.
* src/qemu/qemu_driver.c (qemuDomainUndefineFlags): Enforce
the limitations.
2011-09-02 21:57:33 -06:00
Eric Blake
bd18b9670f snapshot: add qemu snapshot redefine support
Redefining a qemu snapshot requires a bit of a tweak to the common
snapshot parsing code, but the end result is quite nice.

Be careful that redefinitions do not introduce circular parent
chains.  Also, we don't want to allow conversion between online
and offline existing snapshots.  We could probably do some more
validation for snapshots that don't already exist to make sure
they are even feasible, by parsing qemu-img output, but that
can come later.

* src/conf/domain_conf.h (virDomainSnapshotParseFlags): New
internal flags.
* src/conf/domain_conf.c (virDomainSnapshotDefParseString): Alter
signature to take internal flags.
* src/esx/esx_driver.c (esxDomainSnapshotCreateXML): Update caller.
* src/vbox/vbox_tmpl.c (vboxDomainSnapshotCreateXML): Likewise.
* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML): Support
new public flags.
2011-09-02 21:57:33 -06:00
Eric Blake
ece197e9d4 snapshot: add qemu snapshot creation without metadata
Supporting NO_METADATA on snapshot creation is interesting - we must
still return a valid opaque snapshot object, but the user can't get
anything out of it (unless we add a virDomainSnapshotGetName()),
since it is no longer registered with the domain.

Also, virsh now tries to query for secure xml, in anticipation of
when we store <domain> xml inside <domainsnapshot>; for now, we
can trivially support it, since we have nothing secure.

* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML): Support
new flag.
(qemuDomainSnapshotGetXMLDesc): Trivially support VIR_DOMAIN_XML_SECURE.
2011-09-02 21:57:32 -06:00
Eric Blake
af65695af0 snapshot: allow recreation of metadata
The first two flags are essential for being able to replicate
snapshot hierarchies across multiple hosts, which will come in
handy for supervised migrations.  It also allows a management app
to take a snapshot of a transient domain, save the metadata, stop
the domain, recreate a new transient domain by the same name,
redefine the snapshot, then revert to it.

This is not quite as convenient as leaving the metadata behind
after a domain is no longer around, but doing that has a few
problems: 1. the libvirt API can only delete snapshot metadata
if there is a valid domain handle to use to get to that snapshot
object - if stale data is left behind without a domain, there is
no way to request that the data be cleaned up. 2. creating a new
domain with the same name but different uuid than the older
domain where a snapshot existed cannot use the older snapshot
data; this risks confusing libvirt, and forbidding the stale
data is similar to the recent patch to forbid stale managed save.

The first two flags might be useful on hypervisors with no metadata,
but only for modifying the notion of the current snapshot;
however, I don't know how to do that for ESX or VBox.

The third flag is a convenience option, to combine a creation with
a delete metadata into one step.  It is trivial for hypervisors
with no metadata.

The qemu changes will be involved enough to warrant a separate patch.

* include/libvirt/libvirt.h.in
(VIR_DOMAIN_SNAPSHOT_CREATE_REDEFINE)
(VIR_DOMAIN_SNAPSHOT_CREATE_CURRENT)
(VIR_DOMAIN_SNAPSHOT_CREATE_NO_METADATA): New flags.
* src/libvirt.c (virDomainSnapshotCreateXML): Document them, and
enforce mutual exclusion.
* src/esx/esx_driver.c (esxDomainSnapshotCreateXML): Trivial
implementation.
* src/vbox/vbox_tmpl.c (vboxDomainSnapshotCreateXML): Likewise.
* docs/formatsnapshot.html.in: Document re-creation.
2011-09-02 17:44:42 -06:00
Eric Blake
6b3801b042 snapshot: identify qemu snapshot roots
Filtering for roots is pretty easy to do.

* src/conf/domain_conf.h (virDomainSnapshotObjListGetNames)
(virDomainSnapshotObjListNum): Update prototype.
* src/conf/domain_conf.c (virDomainSnapshotObjListCopyNames)
(virDomainSnapshotObjListGetNames, virDomainSnapshotObjListCount)
(virDomainSnapshotObjListNum): Support filtering.
* src/qemu/qemu_driver.c (qemuDomainSnapshotNum)
(qemuDomainSnapshotListNames): Update callers.
2011-09-02 17:35:02 -06:00
Eric Blake
9f5e53e211 snapshot: identify which snapshots have metadata
To make it easier to know when undefine will fail because of existing
snapshot metadata, we need to know how many snapshots have metadata.

Also, it is handy to filter the list of snapshots to just those that
have no parents; document that flag now, but implement it in later patches.

* include/libvirt/libvirt.h.in (VIR_DOMAIN_SNAPSHOT_LIST_ROOTS)
(VIR_DOMAIN_SNAPSHOT_LIST_METADATA): New flags.
* src/libvirt.c (virDomainSnapshotNum)
(virDomainSnapshotListNames): Document them.
* src/esx/esx_driver.c (esxDomainSnapshotNum)
(esxDomainSnapshotListNames): Implement trivial flag.
* src/vbox/vbox_tmpl.c (vboxDomainSnapshotNum)
(vboxDomainSnapshotListNames): Likewise.
* src/qemu/qemu_driver.c (qemuDomainSnapshotNum)
(qemuDomainSnapshotListNames): Likewise.
2011-09-02 16:58:36 -06:00
Eric Blake
7afa6b4129 snapshot: let qemu discard only snapshot metadata
Adding this was trivial compared to the previous patch for fixing
qemu snapshot deletion in the first place.

* src/qemu/qemu_driver.c (qemuDomainSnapshotDiscard): Add
parameter.
(qemuDomainSnapshotDiscardDescendant, qemuDomainSnapshotDelete):
Update callers.
2011-09-02 16:56:48 -06:00
Eric Blake
795fe9b2fa snapshot: allow deletion of just snapshot metadata
A future patch will make it impossible to remove a domain if it
would leave behind any libvirt-tracked metadata about snapshots,
since stale metadata interferes with a new domain by the same name.
But requiring snaphot contents to be deleted before removing a
domain is harsh; with qemu, qemu-img can still make use of the
contents after the libvirt domain is gone.  Therefore, we need
an option to get rid of libvirt tracking information, but not
the actual contents.  For hypervisors that do not track any
metadata in libvirt, the implementation is trivial; all remaining
hypervisors (really, just qemu) will be dealt with separately.

* include/libvirt/libvirt.h.in
(VIR_DOMAIN_SNAPSHOT_DELETE_METADATA_ONLY): New flag.
* src/libvirt.c (virDomainSnapshotDelete): Document it.
* src/esx/esx_driver.c (esxDomainSnapshotDelete): Trivially
supported when there is no libvirt metadata.
* src/vbox/vbox_tmpl.c (vboxDomainSnapshotDelete): Likewise.
2011-09-02 16:51:33 -06:00
Eric Blake
67555b2434 snapshot: simplify acting on just children
Similar to the last patch in isolating the filtering from the
client actions, so that clients don't have to reinvent the
filtering.

* src/conf/domain_conf.h (virDomainSnapshotForEachChild): New
prototype.
* src/libvirt_private.syms (domain_conf.h): Export it.
* src/conf/domain_conf.c (virDomainSnapshotActOnChild)
(virDomainSnapshotForEachChild): New functions.
(virDomainSnapshotCountChildren): Delete.
(virDomainSnapshotHasChildren): Simplify.
* src/qemu/qemu_driver.c (qemuDomainSnapshotReparentChildren)
(qemuDomainSnapshotDelete): Likewise.
2011-09-02 16:09:18 -06:00
Eric Blake
90ea06b88a snapshot: track current domain across deletion of children
Deleting a snapshot and all its descendants had problems with
tracking the current snapshot.  The deletion does not necessarily
proceed in depth-first order, so a parent could be deleted
before a child, wreaking havoc on passing the notion of the
current snapshot to the parent.  Furthermore, even if traversal
were depth-first, doing multiple file writes to pass current up
the chain one snapshot at a time is wasteful, comparing to a
single update to the current snapshot at the end of the algorithm.

* src/qemu/qemu_driver.c (snap_remove): Add field.
(qemuDomainSnapshotDiscard): Add parameter.
(qemuDomainSnapshotDiscardDescendant): Adjust accordingly.
(qemuDomainSnapshotDelete): Properly reset current.
2011-09-02 16:07:41 -06:00
Eric Blake
cb231b4bee snapshot: avoid crash when deleting qemu snapshots
This one's nasty.  Ever since we fixed virHashForEach to prevent
nested hash iterations for safety reasons (commit fba550f6),
virDomainSnapshotDelete with VIR_DOMAIN_SNAPSHOT_DELETE_CHILDREN
has been broken for qemu: it deletes children, while leaving
grandchildren intact but pointing to a no-longer-present parent.
But even before then, the code would often appear to succeed to
clean up grandchildren, but risked memory corruption if you have
a large and deep hierarchy of snapshots.

For acting on just children, a single virHashForEach is sufficient.
But for acting on an entire subtree, it requires iteration; and
since we declared recursion as invalid, we have to switch to a
while loop.  Doing this correctly requires quite a bit of overhaul,
so I added a new helper function to isolate the algorithm from the
actions, so that callers do not have to reinvent the iteration.

Note that this _still_ does not handle CHILDREN correctly if one
of the children is the current snapshot; that will be next.

* src/conf/domain_conf.h (_virDomainSnapshotDef): Add mark.
(virDomainSnapshotForEachDescendant): New prototype.
* src/libvirt_private.syms (domain_conf.h): Export it.
* src/conf/domain_conf.c (virDomainSnapshotMarkDescendant)
(virDomainSnapshotActOnDescendant)
(virDomainSnapshotForEachDescendant): New functions.
* src/qemu/qemu_driver.c (qemuDomainSnapshotDiscardChildren):
Replace...
(qemuDomainSnapshotDiscardDescenent): ...with callback that
doesn't nest hash traversal.
(qemuDomainSnapshotDelete): Use new function.
2011-09-02 16:04:32 -06:00
Eric Blake
a31d65695d snapshot: speed up snapshot location
Each snapshot lookup was iterating over the entire hash table, O(n),
instead of honing in directly on the hash key, amortized O(1).

Besides, fixing this means that virDomainSnapshotFindByName can now
be used inside another virHashForeach iteration (without this patch,
attempts to lookup a snapshot by name during a hash iteration will
fail due to nested iteration).

* src/conf/domain_conf.c (virDomainSnapshotFindByName): Simplify.
(virDomainSnapshotObjListSearchName): Delete unused function.
2011-09-02 16:03:50 -06:00
Eric Blake
7dc44eb059 snapshot: fine-tune qemu snapshot revert states
For a system checkpoint of a running or paused domain, it's fairly
easy to honor new flags for altering which state to use after the
revert.  For an inactive snapshot, the revert has to be done while
there is no qemu process, so do back-to-back transitions; this also
lets us revert to inactive snapshots even for transient domains.

* src/qemu/qemu_driver.c (qemuDomainRevertToSnapshot): Support new
flags.
2011-09-02 12:33:37 -06:00
Eric Blake
25fb3ef1e1 snapshot: properly revert qemu to offline snapshots
Commit 5e47785 broke reverts to offline system checkpoint snapshots
with older qemu, since there is no longer any code path to use
qemu -loadvm on next boot.  Meanwhile, reverts to offline system
checkpoints have been broken for newer qemu, both before and
after that commit, since -loadvm no longer works to revert to
disk state without accompanying vm state.  Fix both of these by
using qemu-img to revert disk state.

Meanwhile, consolidate the (now 3) clients of a qemu-img iteration
over all disks of a VM into one function, so that any future
algorithmic fixes to the FIXMEs in that function after partial
loop iterations are dealt with at once.  That does mean that this
patch doesn't handle partial reverts very well, but we're not
making the situation any worse in this patch.

* src/qemu/qemu_driver.c (qemuDomainRevertToSnapshot): Use
qemu-img rather than 'qemu -loadvm' to revert to offline snapshot.
(qemuDomainSnapshotRevertInactive): New helper.
(qemuDomainSnapshotCreateInactive): Factor guts...
(qemuDomainSnapshotForEachQcow2): ...into new helper.
(qemuDomainSnapshotDiscard): Use it.
2011-09-02 12:30:11 -06:00
Eric Blake
88fe7a4ba5 snapshot: improve reverting to qemu paused snapshots
If you take a checkpoint snapshot of a running domain, then pause
qemu, then restore the snapshot, the result should be a running
domain, but the code was leaving things paused.  Furthermore, if
you take a checkpoint of a paused domain, then run, then restore,
there was a brief but non-deterministic window of time where the
domain was running rather than paused.  Fix both of these
discrepancies by always pausing before restoring.

Also, check that the VM is active every time lock is dropped
between two monitor calls.

Finally, straighten out the events that get emitted on each
transition.

* src/qemu/qemu_driver.c (qemuDomainRevertToSnapshot): Always
pause before reversion, and improve events.
2011-09-02 12:05:08 -06:00
Eric Blake
7381aaff33 snapshot: fine-tune qemu saved images starting paused
Implement the new running/paused overrides for saved state management.

Unfortunately, for virDomainSaveImageDefineXML, the saved state
updates are write-only - I don't know of any way to expose a way
to query the current run/pause setting of an existing save image
file to the user without adding a new API or modifying the domain
xml of virDomainSaveImageGetXMLDesc to include a new element to
reflect the state bit encoded into the save image.  However, I
don't think this is a show-stopper, since the API is designed to
leave the state bit alone unless an explicit flag is used to
change it.

* src/qemu/qemu_driver.c (qemuDomainSaveInternal)
(qemuDomainSaveImageOpen): Adjust signature.
(qemuDomainSaveFlags, qemuDomainManagedSave)
(qemuDomainRestoreFlags, qemuDomainSaveImageGetXMLDesc)
(qemuDomainSaveImageDefineXML, qemuDomainObjRestore): Adjust
callers.
2011-09-02 10:00:06 -06:00
Eric Blake
3cff66f487 snapshot: fine-tune ability to start paused
While it is nice that snapshots and saved images remember whether
the domain was running or paused, sometimes the restoration phase
wants to guarantee a particular state (paused to allow hot-plugging,
or running without needing to call resume).  This introduces new
flags to allow the control, and a later patch will implement the
flags for qemu.

* include/libvirt/libvirt.h.in (VIR_DOMAIN_SAVE_RUNNING)
(VIR_DOMAIN_SAVE_PAUSED, VIR_DOMAIN_SNAPSHOT_REVERT_RUNNING)
(VIR_DOMAIN_SNAPSHOT_REVERT_PAUSED): New flags.
* src/libvirt.c (virDomainSaveFlags, virDomainRestoreFlags)
(virDomainManagedSave, virDomainSaveImageDefineXML)
(virDomainRevertToSnapshot): Document their use, and enforce
mutual exclusion.
2011-09-02 10:00:06 -06:00
Eric Blake
c1ff5dc63d snapshot: better events when starting paused
There are two classes of management apps that track events - one
that only cares about on/off (and only needs to track EVENT_STARTED
and EVENT_STOPPED), and one that cares about paused/running (also
tracks EVENT_SUSPENDED/EVENT_RESUMED).  To keep both classes happy,
any transition that can go from inactive to paused must emit two
back-to-back events - one for started and one for suspended (since
later resuming of the domain will only send RESUMED, but the first
class isn't tracking that).

This also fixes a bug where virDomainCreateWithFlags with the
VIR_DOMAIN_START_PAUSED flag failed to start paused when restoring
from a managed save image.

* include/libvirt/libvirt.h.in (VIR_DOMAIN_EVENT_SUSPENDED_RESTORED)
(VIR_DOMAIN_EVENT_SUSPENDED_FROM_SNAPSHOT)
(VIR_DOMAIN_EVENT_RESUMED_FROM_SNAPSHOT): New sub-events.
* src/qemu/qemu_driver.c (qemuDomainRevertToSnapshot): Use them.
(qemuDomainSaveImageStartVM): Likewise, and add parameter.
(qemudDomainCreate, qemuDomainObjStart): Send suspended event when
starting paused.
(qemuDomainObjRestore): Add parameter.
(qemuDomainObjStart, qemuDomainRestoreFlags): Update callers.
* examples/domain-events/events-c/event-test.c
(eventDetailToString): Map new detail strings.
2011-09-02 10:00:06 -06:00
Marc-André Lureau
4813b3f094 Learn to use spicevmc as a redirection type for usb-redir 2011-09-02 23:39:03 +08:00
Marc-André Lureau
162efa1a7c Add "redirdev" redirection device
- create a new "redirdev" element for this purpose
2011-09-02 23:39:03 +08:00
Marc-André Lureau
fdd14a9d05 qemu: Don't append 0 at usb id, so that it is compatible with legacy -usb
QEMU uses USB bus name "usb.0" when using the legacy -usb argument.
If we want to allow USB devices to specify their addresses with legacy
-usb, we should either in case of legacy bus name drop the 0 from the
address bus, or just drop the 0 from device id. This patch does the
later.

Another solution would be to permit addressing on non-legacy USB
controllers only.
2011-09-02 23:39:03 +08:00
Marc-André Lureau
f35bbf7be7 qemu: don't reserve slot 1 if a PIIX3 USB controller is defined there
Applies only to piix3 and check if piix3 controller is on correct
address, or report error
2011-09-02 23:39:03 +08:00
Marc-André Lureau
31710a5389 Modify USB port to be defined as a port path
So that devices can be attached to hubs. Example, to attach to first
port of a usb-hub on port 1.

      <hub type='usb'>
         <address type='usb' bus='0' port='1'/>
      </hub>

      <input type='mouse' type='usb'>
         <address type='usb' bus='0' port='1.1'/>
      </hub>

also add a test entry
2011-09-02 23:39:03 +08:00
Marc-André Lureau
fdabeb3c5f Add USB hub device
domain parsing and serialization code, qemu driver backend and
a couple of test
2011-09-02 23:38:52 +08:00
Marc-André Lureau
f3ce59621f Add USB companion controllers support
Companion controllers take an extra 'master' attribute to associate
them.

Also add tests for this
2011-09-02 23:22:56 +08:00
Marc-André Lureau
22c0d433ab USB devices gain a new USB address child element
Expand the domain and the QEmu driver code
Adds a couple of tests
2011-09-02 23:22:56 +08:00
Marc-André Lureau
d6d54cd19e Add a new controller type 'usb' with optionnal 'model'
The model by default is piix3-uchi.

Example:
<controller type='usb' index='0' model='ich9-ehci'/>
2011-09-02 23:22:56 +08:00
Marc-André Lureau
2e4b5243b2 Add USB controller models
List is: piix3-uhci piix4-uhci ehci ich9-ehci1 ich9-uhci1 ich9-uhci2
ich9-uhci3 vt82c686b-uhci pci-ohci
2011-09-02 23:22:56 +08:00
Marc-André Lureau
8631bdc0c8 Rename virDomainControllerModel to virDomainControllerModelSCSI
Since we are about to add USB controller support let's remove the
ambiguity
2011-09-02 23:22:56 +08:00
Marc-André Lureau
329f907b99 Add various USB devices QEMU_CAPS 2011-09-02 23:22:56 +08:00
Eric Blake
c554f6e18b snapshot: fix corner case on OOM during creation
Commit 6766ff10 introduced a corner case bug with snapshot creation:
if a snapshot is created, but then we hit OOM while trying to
create the return value of the function, then we have polluted the
internal directory with the snapshot metadata with no way to clean
it up from the running libvirtd.

* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML): Don't
write metadata file on OOM condition.
2011-09-02 08:50:01 -06:00
Osier Yang
6ee52c1b76 Add directsync cache mode support for disk driver
Newer QEMU introduced cache=directsync for -drive, this patchset
is to expose it in libvirt layer.

  * Introduced a new QEMU capability flag ($prefix_CACHE_DIRECTSYNC),
    As even $prefix_CACHE_V2 is set, we can't known if directsync
    is supported.
2011-09-02 21:36:58 +08:00
Osier Yang
27758859c7 storage: Add fs pool formatting
This patch adds the ability to make the filesystem for a filesystem
pool during a pool build.

The patch adds two new flags, no overwrite and overwrite, to control
when mkfs gets executed.  By default, the patch preserves the
current behavior, i.e., if no flags are specified, pool build on a
filesystem pool only makes the directory on which the filesystem
will be mounted.

If the no overwrite flag is specified, the target device is checked
to determine if a filesystem of the type specified in the pool is
present.  If a filesystem of that type is already present, mkfs is
not executed and the build call returns an error.  Otherwise, mkfs
is executed and any data present on the device is overwritten.

If the overwrite flag is specified, mkfs is always executed, and any
existing data on the target device is overwritten unconditionally.
2011-09-02 21:16:58 +08:00
Osier Yang
50c82157e1 API: Init conn in case of it might be used uninitialized
There is a goto before "conn" is initialized.
2011-09-02 15:41:29 +08:00
Eric Blake
55d88def95 qemu: detect incomplete save files
Several users have reported problems with 'virsh start' failing because
it was encountering a managed save situation where the managed save file
was incomplete.  Be more robust to this by using two different magic
numbers, so that newer libvirt can gracefully handle an incomplete file
differently than a complete one, while older libvirt will at least fail
up front rather than trying to load only to have qemu fail at the end.

Managed save is a convenience - it exists to preserve as much state
as possible; if the state was not preserved, it is reasonable to just
log that fact, then proceed with a fresh boot.  On the other hand,
user saves are under user control, so we must fail, but by making
the failure message distinct, the user can better decide how to handle
the situation of an incomplete save file.

* src/qemu/qemu_driver.c (QEMUD_SAVE_PARTIAL): New define.
(qemuDomainSaveInternal): Use it to mark incomplete images.
(qemuDomainSaveImageOpen, qemuDomainObjRestore): Add parameter
that controls what to do with partial images.
(qemuDomainRestoreFlags, qemuDomainSaveImageGetXMLDesc)
(qemuDomainSaveImageDefineXML, qemuDomainObjStart): Update callers.
Based on an initial idea by Osier Yang.
2011-09-01 22:08:13 -06:00
Eric Blake
449ae9c2f1 qemu: refactor file opening
In a SELinux or root-squashing NFS environment, libvirt has to go
through some hoops to create a new file that qemu can then open()
by name.  Snapshots are a case where we want to guarantee an empty
file that qemu can open; also, reopening a save file to convert it
from being marked partial to complete requires a reopen to avoid
O_DIRECT headaches.  Refactor some existing code to make it easier
to reuse in later patches.

* src/qemu/qemu_migration.h (qemuMigrationToFile): Drop parameter.
* src/qemu/qemu_migration.c (qemuMigrationToFile): Let cgroup do
the stat, rather than asking caller to do it and pass info down.
* src/qemu/qemu_driver.c (qemuOpenFile): New function, pulled from...
(qemuDomainSaveInternal): ...here.
(doCoreDump, qemuDomainSaveImageOpen): Use it here as well.
2011-09-01 22:08:13 -06:00
Wen Congyang
deff02a365 reserve slot 1 on pci bus0
After supporting multi function pci device, we only reserve function 1 on slot 1.
The user can use the other function on slot 1 in the xml config file. We should
detect this wrong usage.
2011-09-02 11:33:04 +08:00
Scott Moser
f0fe28cb8d lxc: do not require 'ifconfig' or 'ipconfig' in container
Currently, the lxc implementation invokes 'ip' and 'ifconfig' commands
inside a container using 'virRun'.  That has the side effect of requiring
those commands to be present and to function in a manner consistent with
the usage.  Some small roots (such as ttylinux) may not have 'ip' or
'ifconfig'.

This patch replaces the use of these commands with usage of
netdevice.  The result is that lxc containers do not have to implement
those commands, and lxc in libvirt is only dependent on the netdevice
interface.

I've tested this patch locally against the ubuntu libvirt version enough
to verify its generally sane.  I attempted to build upstream today, but
failed with:
  /usr/bin/ld:
    ../src/.libs/libvirt_driver_qemu.a(libvirt_driver_qemu_la-qemu_domain.o):
   undefined reference to symbol 'xmlXPathRegisterNs@@LIBXML2_2.4.30

Thats probably a local issue only, but I wanted to get this patch up and
see what others thought of it.  This is ubuntu bug
https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/828211 .
2011-09-01 20:11:50 -06:00
Serge Hallyn
c1665ba872 Create ptmx as a device
Hi,

I'm seeing an issue with udev and libvirt-lxc.  Libvirt-lxc creates
/dev/ptmx as a symlink to /dev/pts/ptmx.  When udev starts up, it
checks the device type, sees ptmx is 'not right', and replaces it
with a 'proper' ptmx.

In lxc, /dev/ptmx is bind-mounted from /dev/pts/ptmx instead of being
symlinked, so udev sees the right device type and leaves it alone.

A patch like the following seems to work for me.  Would there be
any objections to this?

>From 4c5035de52de7e06a0de9c5d0bab8c87a806cba7 Mon Sep 17 00:00:00 2001
From: Ubuntu <ubuntu@domU-12-31-39-14-F0-B3.compute-1.internal>
Date: Wed, 31 Aug 2011 18:15:54 +0000
Subject: [PATCH 1/1] make ptmx a bind mount rather than symlink

udev on some systems checks the device type of /dev/ptmx, and replaces it if
not as expected.  The symlink created by libvirt-lxc therefore gets replaced.
By creating it as a bind mount, the device type is correct and udev leaves it
alone.

Signed-off-by: Serge Hallyn <serge.hallyn@canonical.com>
2011-09-01 20:11:50 -06:00
Adam Litke
d4b9e06256 BlockPull: Set initial bandwidth limit if specified
The libvirt BlockPull API supports the use of an initial bandwidth limit but the
qemu block_stream API does not.  To get the desired behavior we use the two APIs
strung together: first BlockPull, then BlockJobSetSpeed.  We can do this at the
driver level to avoid duplicated code in each monitor path.

Signed-off-by: Adam Litke <agl@us.ibm.com>
2011-09-01 20:11:50 -06:00
Adam Litke
78d9325d1e BlockJob: Bandwidth parameter is in MB when using text monitor
Due to an unfortunate precedent in qemu, the units for the bandwidth parameter
to block_job_set_speed are different between the text monitor and the qmp
monitor.  While the qmp monitor uses bytes/s, the text monitor expects MB/s.

Correct the units for the text interface.

Signed-off-by: Adam Litke <agl@us.ibm.com>
2011-09-01 20:11:50 -06:00
Jim Fehlig
57c95175e2 Increase size of buffer for xend response
On systems with many pcpus, the sexpr returned by xend can be quite
large for dom0 when it is configured to have #vcpus = #pcpus (default).
E.g. on a 80 pcpu system, where dom0 had 80 vcpus, the sexpr details
for dom0 was 73817 bytes!  Increase maximum buffer size to 256k.
2011-09-01 19:57:30 -06:00
Jim Fehlig
32620dabb1 Don't overwrite errors from xend_{get,req}
xenDaemonDomainFetch() was overwriting errors reported by
xend_get() and xend_req().  E.g. without patch

error: failed Xen syscall xenDaemonDomainFetch failed to find this domain

with patch

error: internal error Xend returned HTTP Content-Length of 73817, which exceeds
maximum of 65536
2011-09-01 18:19:33 -06:00
Eric Blake
7bc1c5cefe build: fix 'make check' with pdwtags
Problem introduced by commit b12354b.

* src/remote_protocol-structs: Remove spurious blank line.
2011-09-01 12:33:46 -06:00
Jim Fehlig
b12354befe Add public API for getting migration speed
Includes impl of python binding since the generator was not
able to cope.

Note: Requires gendispatch.pl patch from Matthias Bolte

https://www.redhat.com/archives/libvir-list/2011-August/msg01367.html
2011-09-01 11:26:21 -06:00
Daniel P. Berrange
b3fb288e52 Fix tracking of RPC messages wrt streams
Commit 2c85644b0b attempted to
fix a problem with tracking RPC messages from streams by doing

-            if (msg->header.type == VIR_NET_REPLY) {
+            if (msg->header.type == VIR_NET_REPLY ||
+                (msg->header.type == VIR_NET_STREAM &&
+                 msg->header.status != VIR_NET_CONTINUE)) {
                 client->nrequests--;

In other words any stream packet, with status NET_OK or NET_ERROR
would cause nrequests to be decremented. This is great if the
packet from from a synchronous virStreamFinish or virStreamAbort
API call, but wildly wrong if from a server initiated abort.
The latter resulted in 'nrequests' being decremented below zero.
This then causes all I/O for that client to be stopped.

Instead of trying to infer whether we need to decrement the
nrequests field, from the message type/status, introduce an
explicit 'bool tracked' field to mark whether the virNetMessagePtr
object is subject to tracking.

Also add a virNetMessageClear function to allow a message
contents to be cleared out, without adversely impacting the
'tracked' field as a naive memset() would do

* src/rpc/virnetmessage.c, src/rpc/virnetmessage.h: Add
  a 'bool tracked' field and virNetMessageClear() API
* daemon/remote.c, daemon/stream.c, src/rpc/virnetclientprogram.c,
  src/rpc/virnetclientstream.c, src/rpc/virnetserverclient.c,
  src/rpc/virnetserverprogram.c: Switch over to use
  virNetMessageClear() and pass in the 'bool tracked' value
  when creating messages.
2011-09-01 10:52:35 +01:00
Daniel P. Berrange
b6263c1801 Fix parted sector size assumption
Parted does not report disk size in 512 byte units, but
rather the disks' logical sector size, which with modern
drives might be 4k.

* src/storage/parthelper.c: Remove hardcoded 512 byte sector
  size
2011-09-01 10:46:31 +01:00
Osier Yang
6f2581edd7 qemu: Fix a regression of domain save
* src/qemu/qemu_driver.c - qemuDomainSaveInternal: Return directly
will keep the domain object locked, introduced by 173015bec6.
2011-09-01 17:38:20 +08:00
Osier Yang
9f3e724339 Revert "test: Cleanup improper VIR_ERR_NO_SUPPORT use"
This reverts commit 172214bd30.
2011-09-01 17:37:11 +08:00
Osier Yang
ffafede112 storage: Fix incorrect error codes
Commit 0376f4a69b intended to fix incorrect use of VIR_ERR_NO_SUPPORT,
but replacing it with VIR_ERR_OPERATION_INVALID is not proper either.
2011-09-01 17:36:38 +08:00
Osier Yang
fd038a337b remote: Fix incorrect error codes
Introduced by d4b53ef6c. For "no internalFlags support", the
error code is changed into INTERNAL_ERROR.
2011-09-01 17:35:56 +08:00
Osier Yang
03388b6424 nodeinfo: Fix incorrect error codes
Introduced by 5e495c8b, except the ones for checking if numa
is supported by host, all the NO_SUPPORT are changed back. For
the ones about numa checking, change them into INTERNAL_ERROR.
2011-09-01 17:35:23 +08:00
Osier Yang
6af0c3e82b lxc: Fix incorrect changes on error codes.
Fix incorrect changes introduced by commit 6ac47762bb.
2011-09-01 17:34:31 +08:00
Osier Yang
c2c713dd00 conf: Substitute OPERATION_INVALID with INTERNAL_ERROR 2011-09-01 17:31:24 +08:00
Daniel P. Berrange
6ff9fc26d3 Stop libxl driver polluting logs on non-Xen hosts
If the libxl driver is compiled in, then everytime libvirtd
starts up on a non-Xen Dom0 host, it logs a error message.
Since this is an expected condition, we should not log at
'error' level, only 'info'.

* src/libxl/libxl_driver.c: Lower log level for certain
  expected errors during driver init
2011-08-31 17:53:01 +01:00
Daniel P. Berrange
d07aa6a96f Fix memory leak parsing 'relabel' attribute in domain security XML
* src/conf/domain_conf.c: Free the 'relabel' attribute
2011-08-31 17:51:09 +01:00
Daniel P. Berrange
c32536e7da Don't leak memory if a cgroup is mounted multiple times
It is possible (expected/likely in Fedora 15) for a cgroup controller
to be mounted in multiple locations at the same time, due to bind
mounts. Currently we leak memory if this happens, because we overwrite
the previous 'mountPoint' string. Instead just accept the first match
we find.

* src/util/cgroup.c: Only accept first match for a cgroup
  controller mount
2011-08-31 17:51:09 +01:00
Eric Blake
cab55fa0a8 security: fix build
Regression introduced in commit 183383889.

* src/libvirt_private.syms (security_manager.h): Drop deleted
symbol. Detected by build-bot.
2011-08-31 08:33:17 -06:00
Daniel P. Berrange
183383889a Remove bogus virSecurityManagerSetProcessFDLabel method
The virSecurityManagerSetProcessFDLabel method was introduced
after a mis-understanding from a conversation about SELinux
socket labelling. The virSecurityManagerSetSocketLabel method
should have been used for all such scenarios.

* src/security/security_apparmor.c, src/security/security_apparmor.c,
  src/security/security_driver.h, src/security/security_manager.c,
  src/security/security_manager.h, src/security/security_selinux.c,
  src/security/security_stack.c: Remove SetProcessFDLabel driver
2011-08-31 11:07:31 +01:00
Daniel P. Berrange
64bdec3841 Fix sanlock socket security labelling
It is not possible to change the label of a TCP socket once it
has been opened. When creating a TCP socket care must be taken
to ensure the socket creation label is set & then cleared.
Remove the bogus call to virSecurityManagerSetProcessFDLabel
from the lock driver guest setup code and instead make use of
virSecurityManagerSetSocketLabel
2011-08-31 11:07:31 +01:00
Daniel P. Berrange
2223b1f71f Fix incorrect path length check in sanlock lockspace setup
The code for creating a sanlock lockspace accidentally used
SANLK_NAME_LEN instead of SANLK_PATH_LEN for a size check.
This meant disk paths were limited to 48 bytes !

* src/locking/lock_driver_sanlock.c: Fix disk path length
  check
2011-08-31 11:07:31 +01:00
Eric Blake
173015bec6 snapshot: forbid snapshot on autodestroy domain
There is no reason to forbid pausing an autodestroy domain
(not to mention that 'virsh start --paused --autodestroy'
succeeds in creating a paused autodestroy domain).

Meanwhile, qemu was failing to enforce the API documentation that
autodestroy domains cannot be saved.  And while the original
documentation only mentioned save/restore, snapshots are another
form of saving that are close enough in semantics as to make no
sense on one-shot domains.

* src/qemu/qemu_driver.c (qemudDomainSuspend): Drop bogus check.
(qemuDomainSaveInternal, qemuDomainSnapshotCreateXML): Forbid
saves of autodestroy domains.
* src/libvirt.c (virDomainCreateWithFlags, virDomainCreateXML):
Document snapshot interaction.
2011-08-30 11:03:54 -06:00
Philipp Hahn
4521ffabeb Fix error detection in device change
According to qemu-kvm/qerror.c all messages start with a capital
"Device ", but the current code only scans for the lower case "device ".
This results in "virDomainUpdateDeviceFlags()" to not detect locked
CD-ROMs and reporting success even in the case of a failure:
	# virsh qemu-monitor-command "$VM" change\ drive-ide0-0-0\ \"/var/lib/libvirt/images/ucs_2.4-0-sec4-20110714145916-dvd-amd64.iso\"
	Device 'drive-ide0-0-0' is locked
	# virsh update-device "$VM" /dev/stdin <<<"<disk type='file' device='cdrom'><driver name='qemu' type='raw'/><source file='/var/lib/libvirt/images/ucs_2.4-0-sec4-20110714145916-dvd-amd64.iso'/><target dev='hda' bus='ide'/><readonly/><alias name='ide0-0-0'/><address type='drive' controller='0' bus='0' unit='0'/></disk>"
	Device updated successfully

Signed-off-by: Philipp Hahn <hahn@univention.de>
2011-08-30 10:32:13 -06:00
Eric Blake
27c8526053 start: allow discarding managed save
There have been several instances of people having problems with
a broken managed save file, and not aware that they could use
'virsh managedsave-remove dom' to fix things.  Making it possible
to do this as part of starting a domain makes the same functionality
easier to find, and one less API call.

* include/libvirt/libvirt.h.in (VIR_DOMAIN_START_FORCE_BOOT): New
flag.
* src/libvirt.c (virDomainCreateWithFlags): Document it.
* src/qemu/qemu_driver.c (qemuDomainObjStart): Alter signature.
(qemuAutostartDomain, qemuDomainStartWithFlags): Update callers.
* tools/virsh.c (cmdStart): Expose it in virsh.
* tools/virsh.pod (start): Document it.
2011-08-30 09:26:47 -06:00
Eric Blake
71a0beaf3a build: simplify use of verify
Back in 2008 when this line of util.h was written, gnulib's verify
module didn't allow the use of multiple verify() in one file
in combination with our choice of gcc -W options.  But that has
since been fixed in gnulib, and newer gnulib even maps verify()
to the C1x feature of _Static_assert, which gives even nicer
diagnostics with a new enough compiler, so we might as well go
with the simpler verify().

* src/util/util.h (VIR_ENUM_IMPL): Use simpler verify, now that
gnulib module is smarter.
2011-08-30 09:23:20 -06:00
Eric Blake
e6b8bc812a qemu: properly label outgoing pipe for tunneled migration
Commit 3261761 made it possible to use pipes instead of sockets
for outgoing tunneled migration; however, it caused a regression
because the pipe was never given a SELinux label.

* src/qemu/qemu_migration.c (doTunnelMigrate): Label outgoing pipe.
2011-08-30 09:15:26 -06:00
Guannan Ren
bae460fc56 rpc: fix a typo in debugging log in virNetServerProgramSendStreamData
The bufferOffset has been initialized to zero in virNetMessageEncodePayloadRaw(),
so, we use bufferLength to represent the length of message which is going to be
sent to client side.
2011-08-29 11:35:49 -06:00
Eric Blake
a71f8fc70f maint: fix spelling errors on lose
* docs/drvqemu.html.in: Fix typo.
* src/libvirt.c (virDomainCreateXML, virDomainCreateWithFlags):
Likewise.
2011-08-26 16:48:24 -06:00
Jim Fehlig
7e5f6a516c Fix generator to cope with call-by-ref long types
From: Matthias Bolte <matthias.bolte@googlemail.com>
Tested-by: Jim Fehlig <jfehlig@novell.com>

Matthias provided this patch to fix an issue I encountered in the
generator with APIs containing call-by-ref long type, e.g.

int virDomainMigrateGetMaxSpeed(virDomainPtr domain,
                                unsigned long *bandwidth,
                                unsigned int flags);
2011-08-26 11:46:41 -06:00
Matthias Bolte
2137cb1911 hyperv: Add basic documentation 2011-08-26 17:52:55 +02:00
Matthias Bolte
5e3b0f8b57 hyperv: Add basic driver for Microsoft Hyper-V
Domain listing, basic information retrieval and domain life cycle
management is implemented. But currently the domain XML output
lacks the complete devices section.

The driver uses OpenWSMAN to directly communicate with a Hyper-V
server over its WS-Management interface exposed via Microsoft WinRM.

The driver is based on the work of Michael Sievers. This started in
the same master program project group at the University of Paderborn
as the ESX driver.

See Michael's blog for details: http://hyperv4libvirt.wordpress.com/
2011-08-26 17:52:55 +02:00
Matthias Bolte
e224b6f8fb hyperv: Add OpenWSMAN based client for the Hyper-V WMI API
Add a generator script to generate the structs and serialization
information for OpenWSMAN.

openwsman.h collects workarounds for problems in OpenWSMAN <= 2.2.6.
There are also disabled sections that would use ws_serializer_free_mem
but can't because it's broken in OpenWSMAN <= 2.2.6. Patches to fix
this have been posted upstream.
2011-08-26 17:52:55 +02:00
Matthias Bolte
4d6e6f4aa9 hyperv: Add driver skeleton 2011-08-26 17:52:55 +02:00
KAMEZAWA Hiroyuki
709b4c500d Fix persistent migration config save
When a user migrates a domain by command as

libvirt saves vm's domain XML config in destination host after migration.
But it saves vm->def. Then, the saved XML contains some garbage.

  <domain type='kvm' id='50'>
                     ^^^^^^^^
  ...
   <console type='pty' tty='/dev/pts/5'>
                       ^^^^^^^^^^^^^^^^^

Avoid saving unnecessary things by saving persistent vm definition.
2011-08-26 09:45:57 -06:00
Jiri Denemark
22d744d0c7 rpc: Don't close connection if program is unknown
In case we add a new program in the future (we did that in the past and
we are going to do it again soon) current daemon will behave badly with
new client that wants to use the new program. Before the RPC rewrite we
used to just send an error reply to any request with unknown program.
With the RPC rewrite in 0.9.3 the daemon just closes the connection
through which such request was sent. This patch fixes this regression.
2011-08-26 17:29:44 +02:00
Michal Privoznik
c4f91b144c remote: Refuse connecting to remote socket
If users wants to connect to remote unix socket, e.g.
'qemu+unix://<remote>/system' currently the <remote> part is ignored,
ending up connecting to localhost. Connecting to remote socket is not
supported and user should have used TLS/TCP/SSH instead.
2011-08-26 16:40:58 +02:00
Michal Privoznik
6c7299d47d virterror: Fix error message for VIR_ERR_INVALID_ARG
When a detail message is presented, nobody expects prefix 'invalid
argument in' but something more general, like 'invalid argument:'.
2011-08-26 16:40:42 +02:00
Daniel P. Berrange
6b434da6bf Detect errors from the 'sendkey' command
On success, the 'sendkey' command does not return any data, so
any data in the reply should be considered to be an error
message

* src/qemu/qemu_monitor_text.c: Treat non-"" reply data as an
  error message for 'sendkey' command
2011-08-26 14:18:57 +01:00
Daniel P. Berrange
ce93f64b1e Fix keymap used to talk with QEMU
The QEMU 'sendkey' command expects keys to be encoded in the same
way as the RFB extended keycode set. Specifically it wants extended
keys to have the high bit of the first byte set, while the Linux
XT KBD driver codeset uses the low bit of the second byte. To deal
with this we introduce a new keymap 'RFB' and use that in the QEMU
driver

* include/libvirt/libvirt.h.in: Add VIR_KEYCODE_SET_RFB
* src/qemu/qemu_driver.c: Use RFB keycode set instead of XT KBD
* src/util/virkeycode-mapgen.py: Auto-generate the RFB keycode
  set from the XT KBD set
* src/util/virkeycode.c: Add RFB keycode entry to table. Add a
  verify check on cardinality of the codeOffset table
2011-08-26 14:18:57 +01:00
Jiri Denemark
855f768996 qemu: Correctly label migration TCP socket 2011-08-26 11:52:54 +02:00
Jiri Denemark
520d91f8bd security: Introduce SetSocketLabel
This API labels all sockets created until ClearSocketLabel is called in
a way that a vm can access them (i.e., they are labeled with svirt_t
based label in SELinux).
2011-08-26 11:52:48 +02:00
Jiri Denemark
4c85d96f27 security: Rename SetSocketLabel APIs to SetDaemonSocketLabel
The APIs are designed to label a socket in a way that the libvirt daemon
itself is able to access it (i.e., in SELinux the label is virtd_t based
as opposed to svirt_* we use for labeling resources that need to be
accessed by a vm). The new name reflects this.
2011-08-26 11:51:09 +02:00
Jiri Denemark
b136266d57 Ignore unused streams in virStreamAbort
When virStreamAbort is called on a stream that has not been used yet,
quite confusing error is returned: "this function is not supported by
the connection driver". Let's just ignore such streams as there's
nothing to abort anyway.
2011-08-26 11:25:01 +02:00
Jiri Denemark
3398eeda75 Do not try to cancel non-existent migration on source
If migration failed on source daemon, the migration is automatically
canceled by the daemon itself. Thus we don't need to call
virDomainMigrateConfirm3(cancelled=1). Calling it doesn't cause any harm
but the resulting error message printed in logs may confuse people.
2011-08-26 11:18:27 +02:00
Eric Blake
6766ff10dd snapshot: track current snapshot across restarts
Audit all changes to the qemu vm->current_snapshot, and make them
update the saved xml file for both the previous and the new
snapshot, so that there is always at most one snapshot with
<active>1</active> in the xml, and that snapshot is used as the
current snapshot even across libvirtd restarts.

This patch does not fix the case of virDomainSnapshotDelete(,CHILDREN)
where one of the children is the current snapshot; that will be later.

* src/conf/domain_conf.h (_virDomainSnapshotDef): Alter member
type and name.
* src/conf/domain_conf.c (virDomainSnapshotDefParseString)
(virDomainSnapshotDefFormat): Update clients.
* docs/schemas/domainsnapshot.rng: Tighten rng.
* src/qemu/qemu_driver.c (qemuDomainSnapshotLoad): Reload current
snapshot.
(qemuDomainSnapshotCreateXML, qemuDomainRevertToSnapshot)
(qemuDomainSnapshotDiscard): Track current snapshot.
2011-08-25 15:11:03 -06:00
Eric Blake
5e47785b85 snapshot: only pass snapshot to qemu command line when reverting
Changing the current vm, and writing that change to the file
system, all before a new qemu starts, is risky; it's hard to
roll back if starting the new qemu fails for some reason.
Instead of abusing vm->current_snapshot and making the command
line generator decide whether the current snapshot warrants
using -loadvm, it is better to just directly pass a snapshot all
the way through the call chain if it is to be loaded.

This frees up the last use of snapshot->def->active for qemu's
use, so the next patch can repurpose that field for tracking
which snapshot is current.

* src/qemu/qemu_command.c (qemuBuildCommandLine): Don't use active
field of snapshot.
* src/qemu/qemu_process.c (qemuProcessStart): Add a parameter.
* src/qemu/qemu_process.h (qemuProcessStart): Update prototype.
* src/qemu/qemu_migration.c (qemuMigrationPrepareAny): Update
callers.
* src/qemu/qemu_driver.c (qemudDomainCreate)
(qemuDomainSaveImageStartVM, qemuDomainObjStart)
(qemuDomainRevertToSnapshot): Likewise.
(qemuDomainSnapshotSetCurrentActive)
(qemuDomainSnapshotSetCurrentInactive): Delete unused functions.
2011-08-25 15:07:57 -06:00
Eric Blake
861dc84bb5 snapshot: don't leak resources on qemu snapshot failure
https://bugzilla.redhat.com/show_bug.cgi?id=727709
mentions that if qemu fails to create the snapshot (such as what
happens on Fedora 15 qemu, which has qmp but where savevm is only
in hmp, and where libvirt is old enough to not try the hmp fallback),
then 'virsh snapshot-list dom' will show a garbage snapshot entry,
and the libvirt internal directory for storing snapshot metadata
will have a bogus file.

This fixes the fallout bug of polluting the snapshot-list with
garbage on failure (the root cause of the F15 bug of not having
fallback to hmp has already been fixed in newer libvirt releases).

* src/qemu/qemu_driver.c (qemuDomainSnapshotCreateXML): Allocate
memory before making snapshot, and cleanup on failure.  Don't
dereference NULL if transient domain exited during snapshot creation.
2011-08-25 14:01:36 -06:00
Alex Jia
5495e45e70 libvirt: avoid dead store in virDomainMigrateVersion3
* src/qemu/qemu_migration.c: avoid dead 'ret' assignment and silence
  clang warning.

Detected by ccc-analyzer:

libvirt.c:4277:5: warning: Value stored to 'ret' is never read
    ret = domain->conn->driver->domainMigrateConfirm3
    ^     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2011-08-25 13:28:10 -06:00
Alex Jia
d69d321086 qemu: avoid dead store in doPeer2PeerMigrate3
* src/qemu/qemu_migration.c: avoid dead 'ret' assignment and silence
  clang warning.

Detected by ccc-analyzer:

  CC     libvirt_driver_qemu_la-qemu_migration.lo
qemu/qemu_migration.c:2046:5: warning: Value stored to 'ret' is never read
    ret = qemuMigrationConfirm(driver, sconn, vm,
    ^     ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
2011-08-25 13:28:08 -06:00
Michal Privoznik
eaddec976e daemon: Move TLS initialization to virInitialize
My previous patch 74c7567133
introduced a regression by removing TLS initialization from client.
2011-08-25 10:22:03 +02:00
Laine Stump
b1643dc15c util: only fchown newly created files in virFileOpenAs
virFileOpenAs takes desired uid:gid as arguments, and not only uses
them for a fork/setuid/setgid when retrying failed open operations,
but additionally always forces the opened file to be owned by the
given uid:gid.

One example of the problems this causes is that, when restoring a
domain from a file that is owned by the qemu user, opening the file
chowns it to root. if dynamic_ownership=1 this is coincidentally
expected, but if dynamic_ownership=0, no existing file should ever
have its ownership changed.

This patch adds an extra check before calling fchown() - it only does
it if O_CREAT was passed to virFileOpenAs() in the openflags.
2011-08-24 15:32:00 -04:00
Shradha Shah
7ae740fcb1 qemu: fix off-by-one in pci slot reservation
Signed-off-by: Steve Hodgson <shodgson@solarflare.com>
2011-08-24 10:49:02 -06:00
Shradha Shah
c0f025b8ba pci: fix pciDeviceListSteal on multiple devices
pciDeviceListSteal(pcidevs, dev) removes dev from pcidevs reducing
the length of pcidevs, so moving onto what was the next dev is wrong.

Instead callers should pop entry 0 repeatedly until pcidevs is empty.

Signed-off-by: Steve Hodgson <shodgson@solarflare.com>
Signed-off-by: Shradha Shah <sshah@solarflare.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
2011-08-24 10:31:29 -06:00
Eric Blake
ba0c219902 libvirt: log all flags parameters
I was testing a virsh patch, and wanted to see if I had passed the
flags I thought.  But with LIBVIRT_DEBUG in the environment, I just
saw:

14:24:52.359: 15022: debug : virDomainSnapshotNum:15586 : dom=0xc9c180, (VM: name=rhel_6-64, uuid=48f8e8e7-e14f-0e14-02f0-ce71997bdcab),

including a trailing space.  This fixes the issues.

* src/libvirt.c: Log flag parameters, even if currently unused.
(VIR_DOMAIN_DEBUG_0): Drop trailing comma in log.
(VIR_DOMAIN_DEBUG_1): Split guts into...
(VIR_DOMAIN_DEBUG_2): ...new macro.
2011-08-24 08:34:49 -06:00
Matthias Bolte
6aa57af3e4 esx: Use $(PYTHON) instead of the shebang to run the generator 2011-08-24 10:45:03 +02:00
Osier Yang
10b100240f qemu: Report error if qemu monitor command not found for BlockJob
* src/qemu/qemu_monitor_json.c: Handle error "CommandNotFound" and
  report the error.

* src/qemu/qemu_monitor_text.c: If a sub info command is not found,
  it prints the output of "help info", for other commands,
  "unknown command" is printed.

Without this patch, libvirt always report:

  An error occurred, but the cause is unknown

This patch was adapted from a patch by Osier Yang <jyang@redhat.com> to
break out detection of unrecognized text monitor commands into a separate
function.

Signed-off-by: Adam Litke <agl@us.ibm.com>
2011-08-24 14:39:42 +08:00
Matthias Bolte
678cd0f04b esx: Refactor a repeated string in the generator 2011-08-23 23:15:21 +02:00
Eric Blake
3a52b864dd maint: fix comment typos
* src/qemu/qemu_driver.c (qemuDomainSaveInternal): Fix typo.
* src/conf/domain_event.c (virDomainEventDispatchMatchCallback):
Likewise.
* daemon/libvirtd.c (daemonRunStateInit): Likewise.
* src/lxc/lxc_container.c (lxcContainerChildMountSort): Likewise.
* src/util/virterror.c (virCopyError, virRaiseErrorFull): Likewise.
* src/xenxs/xen_sxpr.c (xenParseSxprSound): Likewise.
2011-08-23 11:31:28 -06:00
Osier Yang
0756e5ad92 xen: Cleanup improper VIR_ERR_NO_SUPPORT use 2011-08-23 16:53:15 +08:00