1
0
mirror of https://gitlab.com/libvirt/libvirt.git synced 2024-12-22 17:34:18 +03:00
Commit Graph

4762 Commits

Author SHA1 Message Date
Chris Lalancette
54971d9170 Domain snapshot RNG and tests.
Signed-off-by: Chris Lalancette <clalance@redhat.com>
2010-05-20 13:50:03 -04:00
Chris Lalancette
8749256d0b Fix up the python bindings for snapshotting.
This involved a few fixes.  To start with,
an virDomainSnapshot object is really tied to a
domain, not a connection, so we have to generate
a slightly different object so that we can get
at self._dom for the object.

Next, we had to "dummy" up an override piece of
XML with a bogus argument that the function doesn't
actually take.  That's so that the generator places
virDomainRevertToSnapshot underneath the correct
class (namely, the virDomain class).

Finally, we had to hand-implement the
virDomainRevertToSnapshot implementation, ignoring the
bogus pointer we are being passed.

With all of this in place, I was able to successfully
take a snapshot and revert to it using only the
Python bindings.

Signed-off-by: Chris Lalancette <clalance@redhat.com>
2010-05-20 13:49:38 -04:00
Jim Meyering
20beb0c0db qemu_conf.c: also recognize new first line of qemu -help output
* src/qemu/qemu_conf.c (QEMU_VERSION_STR_1, QEMU_VERSION_STR_2):
Define these instead of...
(QEMU_VERSION_STR): ... this.  Remove definition.
(qemudParseHelpStr): Check first for the new, shorter prefix,
"QEMU emulator version", and then for the old one,
"QEMU PC emulator version" when trying to parse the version number.
Based on a patch by Chris Wright.
2010-05-20 19:13:26 +02:00
Jim Meyering
1f6fc519cd lxc_controller.c: don't ignore failed "accept"
* src/lxc/lxc_controller.c (ignorable_epoll_accept_errno): New function.
(lxcControllerMain): Handle a failed accept carefully:
most errno values indicate legitimate failure and must be fatal.
However, ignore a special case: that in which an incoming client quits
between the poll() indicating its presence, and our accept() which
is trying to process it.
2010-05-20 19:12:39 +02:00
Cole Robinson
348c6fc0ab qemu: Don't deny ShutdownVMDaemon for non-running VMs
Clients that require this already seem to do so. Calling this function
with pid < 1 also should not cause problems.
2010-05-20 11:45:20 -04:00
Cole Robinson
4da188ea3d qemu: Remove explicit VNC XML cleanup
This only exists for a certain cleanup path in StartVMDaemon, but is
unneeded since domain_conf.c handles this for us automatically.
2010-05-20 11:45:19 -04:00
Cole Robinson
b6992818c6 qemu: Properly cleanup in security startup error path
Everything after hostdev setup needs to jump to cleanup on error.
2010-05-20 11:45:19 -04:00
Jim Meyering
6752d11997 libvirtd: don't ignore virInitialize failure
* daemon/libvirtd.c (main): Diagnose virInitialize failure
and exit nonzero.
2010-05-20 16:19:38 +02:00
Alex Williamson
7e1249f259 Rename qemuBuildCommandLine tapfds -> vmfds.
There doesn't seem to be anything specific to tap devices for this
array of file descriptors which need to stay open of the guest to use.
Rename then for others to make use of.

Signed-off-by: Alex Williamson <alex.williamson@redhat.com>
Signed-off-by: Chris Lalancette <clalance@redhat.com>
2010-05-20 10:05:20 -04:00
Jim Meyering
97d5c95d8e maint: prohibit newline at end of diagnostic
* cfg.mk (sc_prohibit_newline_at_end_of_diagnostic): New rule.
Idea proposed by Jiri Denemark.
2010-05-20 15:44:44 +02:00
Jim Meyering
e1b419ea25 maint: remove unwanted newline at end of diagnostic
* src/xen/xend_internal.c (xenDaemonDomainDefineXML): Remove \n.
* src/network/bridge_driver.c (networkAddMasqueradingIptablesRules):
Likewise.
2010-05-20 15:44:40 +02:00
Jim Meyering
6aacf84a3b qemudDomainMigrateFinish2: handle a case of virDomainSaveStatus failure
* src/qemu/qemu_driver.c (qemudDomainMigrateFinish2): Don't ignore
virDomainSaveStatus failure.
* src/conf/domain_conf.h (virDomainSaveStatus): Use
ATTRIBUTE_RETURN_CHECK, so this doesn't happen again.
2010-05-20 07:51:21 +02:00
Eric Blake
33c721d323 docs: distribute more coding convention documentation
These files may be useful for anyone making modifications to
source files in a tarball distribution.

* src/Makefile.am (EXTRA_DIST): Add THREADS.txt.
* daemon/Makefile.am (EXTRA_DIST): Add THREADING.txt.
2010-05-19 16:28:50 -06:00
Eric Blake
92036c2edf build: distribute missing file
Commit f30ccb2458 was incomplete.

* .gitignore: Ignore intermediate directory during failed 'make
distcheck'.
* tests/Makefile.am (qemuhelpdata): Distribute new file.
2010-05-19 16:28:49 -06:00
Matthias Bolte
894baa0af8 esx: Make esxVI_*_CastFromAnyType dynamically dispatched
This will be used in the ESX storage driver in order to handle
the DatastoreInfo type and inheriting types properly.
2010-05-20 00:00:34 +02:00
Matthias Bolte
90fce23f24 esx: Allow esxVI_X_DynamicCast to be called successfully on X objects
This semantic will be used in the ESX storage driver.
2010-05-20 00:00:31 +02:00
Jim Meyering
10c681622a tests: the remote_protocol check also accommodates older pdwtags
This test was failing on systems using pdwtags from dwarves-1.3.
Reported by Matthias Bolte.
Two-pronged fix:
  - use --verbose to work also with dwarves-1.3; adapt regular
    expressions to handle now-varying separators
  - require a minimum number of post-split clauses, in order to
    skip upon any future format change.
    Currently there are 318; if there are 300 or fewer,
    give a warning similar to when pdwtags is missing.
* src/Makefile.am (remote_protocol-structs): Use pdwtags' --verbose
option to make 1.3 emit member sizes and offsets.
Consistently output WARNING messages to stderr.
2010-05-19 19:12:44 +02:00
Jim Meyering
dcf30d9c25 initialize "meta" in virStorageFileGetMetadata, not in each caller
Do not require each caller of virStorageFileGetMetadata and
virStorageFileGetMetadataFromFD to first clear the storage of the
"meta" buffer.  Instead, initialize that storage in
virStorageFileGetMetadataFromFD.
* src/util/storage_file.c (virStorageFileGetMetadataFromFD): Clear
"meta" here, not before each of the following callers.
* src/qemu/qemu_driver.c (qemuSetupDiskCgroup): Don't clear "meta" here.
(qemuTeardownDiskCgroup): Likewise.
* src/qemu/qemu_security_dac.c (qemuSecurityDACSetSecurityImageLabel):
Likewise.
* src/security/security_selinux.c (SELinuxSetSecurityImageLabel):
Likewise.
* src/security/virt-aa-helper.c (get_files): Likewise.
2010-05-19 17:07:08 +02:00
Jim Meyering
e638a2461d (qemu*DiskCgroup): avoid dead code
* src/qemu/qemu_driver.c (qemuTeardownDiskCgroup): Remove
bogus empty-body while-loop.
(qemuSetupDiskCgroup): Likewise.
2010-05-19 17:07:08 +02:00
Jim Meyering
acd981e37f maint: more VIR_WARN corrections: now manually
* po/POTFILES.in: Remove src/util/logging.c and src/util/uuid.c.
* src/phyp/phyp_driver.c (phypUUIDTable_ReadFile): Correct more
VIR_WARN uses, now manually.
(phypUUIDTable_Init, phypUUIDTable_Pull): Likewise.
2010-05-19 13:03:35 +02:00
Jim Meyering
3f76a993c2 maint: use VIR_WARN0("...") rather than VIR_WARN("%s", "...")
Run this command:
  git grep -l 'VIR_WARN("%s", "'|xargs perl -pi -e \
    's/VIR_WARN\("%s", "/VIR_WARN0("/'
* src/phyp/phyp_driver.c (phypDomainGetInfo): Perform the above.
(phypDomainCreateAndStart, phypUUIDTable_ReadFile): Likewise.
2010-05-19 12:59:15 +02:00
Jim Meyering
052893cf66 maint: enforce no-markup policy wrt VIR_WARN-like macros
* cfg.mk (sc_prohibit_gettext_markup): New rule, to enforce
this policy.  Contrary to most diagnostic-emitting functions,
where we require _(...) markup, here, we require that _() *not*
be used for certain functions (or function-like macros).
2010-05-19 12:52:47 +02:00
Jim Meyering
feb93e9830 maint: remove _(...) from VIR_WARN arg manually
* src/util/uuid.c (virUUIDGenerate): Remove _(...) manually.
2010-05-19 12:52:47 +02:00
Jim Meyering
c5a2fe243b maint: don't mark VIR_WARN or VIR_WARN0 diagnostics for translation
Approximately 60 messages were marked.  Since these diagnostics are
intended solely for developers and maintainers, encouraging translation
is deemed to be counterproductive:
http://thread.gmane.org/gmane.comp.emulators.libvirt/25050/focus=25052

Run this command:
  git grep -l VIR_WARN|xargs perl -pi -e \
    's/(VIR_WARN0?)\s*\(_\((".*?")\)/$1($2/'
2010-05-19 12:00:18 +02:00
Jim Meyering
d54bad5d0e note a typo: VIR_MIGRATE_TUNNELLED should be VIR_MIGRATE_TUNNELED,
so that searches for properly spelled "TUNNELED" turn up
the surprising spelling of this public symbol.
* include/libvirt/libvirt.h.in: Add a comment.
2010-05-19 08:08:52 +02:00
Jiri Denemark
b02db3518e Add support for SSE4.1 and SSE4.2 CPU features 2010-05-18 20:53:31 +02:00
Jiri Denemark
272d5a9653 Fix potential NULL dereference in remoteDomainMigratePrepare2 2010-05-18 20:53:31 +02:00
Jim Meyering
49ed0a2e1b do not ignore qemuMonitorAddDrive failure; make uses identical
There were three very similar uses of qemuMonitorAddDrive.
This change makes the three 17-line sequences identical.
* src/qemu/qemu_driver.c (qemudDomainAttachPciDiskDevice): Detect
failure.  Add VIR_WARN and braces.
(qemudDomainAttachSCSIDisk): Add VIR_WARN and braces.
(qemudDomainAttachUsbMassstorageDevice): Likewise.
2010-05-18 19:23:33 +02:00
Jim Meyering
d564fcb3e1 ebtablesAddRemoveRule, iptablesAddRemoveRule: don't skip va_end
* src/util/ebtables.c (ebtablesAddRemoveRule): Don't skip
va_end(args) on an error path.
* src/util/iptables.c (iptablesAddRemoveRule): Identical change.
2010-05-18 19:23:33 +02:00
Jim Meyering
11eeabd71b qemudDomainRestore: handle a case of virDomainSaveStatus failure
* src/qemu/qemu_driver.c (qemudDomainRestore): Don't ignore
virDomainSaveStatus failure.
2010-05-18 19:23:33 +02:00
Jim Meyering
a23675d36d python: don't ignore virInitialize failure in module initialization
* python/libvirt-override.c (initlibvirtmod): Upon virInitialize
failure, skip the Py_InitModule call.
2010-05-18 19:23:33 +02:00
Jim Meyering
6ab131bae2 tests: do not ignore virInitialize failure
* tests/nodeinfotest.c (mymain): Do not ignore virInitialize failure.
Most other callers of virInitialize test for failure.
2010-05-18 19:23:32 +02:00
Jim Meyering
cff218f353 qemuMonitorTextMigrate: avoid leak on OOM-error path
* src/qemu/qemu_monitor_text.c (qemuMonitorTextMigrate): Also
free "safedest" buffer when failing.
2010-05-18 19:23:32 +02:00
Jim Meyering
bc5443f4c6 virNWFilterDefParseXML: avoid leak on error paths
* src/conf/nwfilter_conf.c (virNWFilterDefParseXML): Also free "ret"
via cleanup.
2010-05-18 19:23:32 +02:00
Jim Meyering
43beb82eca linuxNodeInfoCPUPopulate: avoid used-uninitialized via a test
* tests/nodeinfotest.c (linuxTestCompareFiles): Don't use
nodeinfo->member uninitialized.  linuxNodeInfoCPUPopulate requires
that some of its nodeinfo members (including threads) be initialized
upon input.  The nodeinfotest.c program lacked the initialization,
while the only other use (nodeGetInfo) did perform it.
It's not trivial to move the initialization into the function,
since nodeGetInfo sets at least one member after clearing the
buffer but before calling linuxNodeInfoCPUPopulate.
2010-05-18 19:23:32 +02:00
Jim Meyering
4d2adaa1d5 virDomainNetDefParseXML: avoid leak upon multiple "filterref"
* src/conf/domain_conf.c (virDomainNetDefParseXML): Don't leak
memory when parsing two or more "filterref" elements.
2010-05-18 19:23:27 +02:00
Jim Meyering
0058184c4e ebiptablesWriteToTempFile: don't close a negative file descriptor
* src/nwfilter/nwfilter_ebiptables_driver.c (ebiptablesWriteToTempFile):
Skip the close if "fd" is negative.
2010-05-18 16:55:26 +02:00
Daniel P. Berrange
31e29fe524 Protect against NULL pointer flaws in monitor usage
History has shown that there are frequent bugs in the QEMU driver
code leading to the monitor being invoked with a NULL pointer.
Although the QEMU driver code should always report an error in
this case before invoking the monitor, as a safety net put in a
generic check in the monitor code entry points.

* src/qemu/qemu_monitor.c: Safety net to check for NULL monitor
  object
2010-05-18 06:03:06 -04:00
Daniel P. Berrange
c4b2a93907 Fix multiple potential NULL pointer references in monitor usage
Any method which intends to invoke a monitor command must have
a check for virDomainObjIsActive() before using the monitor to
ensure that priv->mon != NULL.

There is one subtle edge case in this though. If a method invokes
multiple monitor commands, and calls qemuDomainObjExitMonitor()
in between two of these commands then there is no guarentee that
priv->mon != NULL anymore. This is because the QEMU process may
exit or die at any time, and because qemuDomainObjEnterMonitor()
releases the lock on virDomainObj, it is possible for the background
thread to close the monitor handle and thus qemuDomainObjExitMonitor
will release the last reference allowing priv->mon to become NULL.

This affects several methods, most notably migration but also some
hotplug methods. This patch takes a variety of approaches to solve
the problem, depending on the particular usage scenario. Generally
though it suffices to add an extra virDomainObjIsActive() check
if qemuDomainObjExitMonitor() was called during the method.

* src/qemu/qemu_driver.c: Fix multiple potential NULL pointer flaws
  in usage of the monitor
2010-05-18 06:03:06 -04:00
Jim Meyering
a986892e61 maint: add more free-like functions to the list and deal with fallout
* cfg.mk (useless_free_options): Add many vir*Free* function names,
and then remove the useless if-before-free tests exposed by running
make syntax-check.
* src/conf/interface_conf.c (virInterfaceDefFree): Remove useless "if".
(virInterfaceAssignDef): Likewise.
* src/conf/network_conf.c (virNetworkAssignDef): Likewise.
* src/conf/storage_conf.c (virStoragePoolObjAssignDef): Likewise.
* src/node_device/node_device_hal.c (dev_create): Likewise.
* src/security/virt-aa-helper.c (vahDeinit): Likewise.
* src/test/test_driver.c (testNodeDeviceCreateXML): Likewise.
* src/util/conf.c (virConfSetValue): Likewise.
2010-05-18 07:53:42 +02:00
Jim Meyering
933522a341 maint: add virCgroupFree to the list of free-like functions
This makes the useless-if-before-free test in maint.mk spot
uses of virCgroupFree just like it does for free and the other
listed functions.
* cfg.mk (useless_free_options): Add virCgroupFree.
Prompted by suggestion from Eric Blake.
2010-05-18 07:53:42 +02:00
Jim Meyering
20701b17e2 qemudDomainSetVcpus: avoid NULL-deref on failed uuid look-up
* src/qemu/qemu_driver.c (qemudDomainSetVcpus): Upon look-up failure,
i.e., vm==NULL, goto cleanup, rather than to "endjob", superficially
since the latter would dereference vm, but more fundamentally because
we certainly don't want to call qemuDomainObjEndJob before we've
even attempted qemuDomainObjBeginJob.
2010-05-18 07:53:42 +02:00
Jim Meyering
93fedcf20f lxcFreezeContainer: avoid test-after-deref of never-NULL pointer
* src/lxc/lxc_driver.c (lxcFreezeContainer): Remove test-after-deref.
Correct indentation in expression.
2010-05-18 07:53:42 +02:00
Matthias Bolte
61fb697977 Add CIFS to the list of network file systems
ESX supports NFS and CIFS. The ESX storage driver will reflect this.
2010-05-18 01:34:34 +02:00
Matthias Bolte
32d9e0707c Add VIR_STORAGE_POOL_INACCESSIBLE to denote inaccessible storage pools
This status will be used by the ESX storage driver.

For example a running NFS pool is inaccessible when the NFS server is
currently unreachable.
2010-05-18 01:34:34 +02:00
Eric Blake
f30ccb2458 qemu_conf: fix flag value
(gdb) p/x QEMUD_CMD_FLAG_VNET_HOST
$7 = 0xffffffff80000000

Oops - that meant we were incorrectly setting QEMU_CMD_FLAG_RTC_TD_HACK
for qemu-kvm-0.12.3 (and probably botching a few other settings as well).

Fixes Red Hat BZ#592070

* src/qemu/qemu_conf.h (QEMUD_CMD_FLAG_VNET_HOST): Avoid sign
extension.
* tests/qemuhelpdata/qemu-kvm-0.12.3: New file.
* tests/qemuhelptest.c (mymain): Add another case.
2010-05-17 16:28:02 -06:00
Cole Robinson
07c621d09c qemu: Clarify a couple error messages
A fedora translator filed:

https://bugzilla.redhat.com/show_bug.cgi?id=580816

Pointing out these two error messages as unclear: "write save" sounds
like a typo without context, and lack of a colon made the second message
difficult to parse.
2010-05-17 17:22:08 -04:00
Eric Blake
d533a98ed6 virFileResolveLink: fix return value
virFileResolveLink was returning a positive value on error,
thus confusing callers that assumed failure was < 0.  The
confusion is further evidenced by callers that would have
ended up calling virReportSystemError with a negative value
instead of a valid errno.

Fixes Red Hat BZ #591363.

* src/util/util.c (virFileResolveLink): Live up to documentation.
* src/qemu/qemu_security_dac.c
(qemuSecurityDACRestoreSecurityFileLabel): Adjust callers.
* src/security/security_selinux.c
(SELinuxRestoreSecurityFileLabel): Likewise.
* src/storage/storage_backend_disk.c
(virStorageBackendDiskDeleteVol): Likewise.
2010-05-17 14:48:27 -06:00
Cole Robinson
df5944ff02 tests: Skip daemon-conf test if dir exceeds UNIX_PATH_MAX
The max path length for unix sockets is pretty small (108, see man 7 unix).
If 'make check' is run from a directory that exceeds this, one of the tests
will fail, and in such a way that requires manually editting the test to
determine why.

There are certainly other ways to handle this, but I've chosen just to skip
the offending test if we will exceed the length limitation.

v2: Drop bashism, use test infrastructure to warn and skip
2010-05-17 15:01:59 -04:00
Cole Robinson
5679c844de pci: Give an explicit error if device not found
v2: Use intended F_OK. Drop devdir param, just check dev->path for device
existence.

v3: Use virReportSystemError, include dev->path in error message.
2010-05-17 15:01:59 -04:00