Commit Graph

16216 Commits

Author SHA1 Message Date
Peter Krempa
9689dfaad3 storage: Refactor storage file initialization to use virStorageSourcePtr
Now that storage source metadata is stored in a single struct we don't
need two initialization functions for different structs.
2014-04-09 14:31:12 +02:00
Peter Krempa
93c1f2cd70 conf: Refactor helpers to retrieve actual storage type
Now that the storage source definition is uniform convert the helpers to
retrieve the actual storage type to a single one.
2014-04-09 14:20:40 +02:00
John Ferlan
0c2305b31c storage: Don't update pool available/allocation if buildVol fails
https://bugzilla.redhat.com/show_bug.cgi?id=1024159

If adding a volume to a storage pool fails during the CreateXML or
CreateXMLFrom API's, we don't want to adjust the available and
allocation values for the storage pool during storageVolDelete
since we haven't adjusted the values for the create.

Refactor storageVolDelete() a bit to create a storageVolDeleteInternal()
which will handle the primary deletion activities.  Add a parameter
updateMeta which will signify whether to update the values or not.

Adjust the calls from CreateXML and CreateXMLFrom to directly call the
DeleteInternal with the pool lock held.  This does bypass the call
to virStorageVolDeleteEnsureACL().
2014-04-09 06:57:37 -04:00
Jean-Baptiste Rouault
a18c713013 build: avoid compiler warning on shadowed name
Introduced in commit d1e55de3.
virstoragetest.c: In function ‘testStorageChain’:
virstoragetest.c:249:10: warning: declaration of ‘abs’ shadows a global
declaration [-Wshadow]
2014-04-09 09:05:42 +02:00
Roman Bogorodskiy
548f21d087 tests: use virBhyveCapsBuild in bhyvexml2argv test
As we can use virBhyveCapsBuild() now, replace
testBhyveBuildCapabilities() with it.
2014-04-09 10:27:16 +04:00
Eric Blake
d1e55de343 conf: another refactor of virstoragetest
Another reduction in the number of structs I have to modify
when I start tracking new fields in virStorageFileMetadata.

* tests/virstoragetest.c (_testFileData): Add fields.
(testStorageChain): Select between fields based on flag.
(mymain): Record both absolute and relative expectations in one
struct.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-08 13:32:55 -06:00
Eric Blake
3486133356 conf: interleave virstoragetest structs
As I add more tests, it's getting harder to follow the split between
a struct in one place and a test using the struct in another.
Interleaving the tests makes changes more localized, and also makes
debugging easier when a test goes wrong during my refactoring work.

* tests/virstoragetest.c (mymain): Modify structs as we go, rather
than up-front.
(testStorageChain): Make failure debugging easier.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-08 13:32:03 -06:00
Eric Blake
fcc7d0ed3a conf: test for more scenarios
Part of the upcoming refactoring will change how broken chains
are detected; it makes sense to test that this works.  In
particular, test the just-fixed infinite loop detection bug.
Also, make sure that detection of directories is sane.

* tests/virstoragetest.c (testStorageChain): Enhance test.
(mymain): Add more tests.
(testCleanupImages, testPrepImages): Populate a directory.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-08 13:29:21 -06:00
Eric Blake
af095bfa0d conf: fix detection of infinite backing loop
While trying to refactor the backing file chain, I noticed that
if you have a self-referential qcow2 file via a relative name:

qemu-img create -f qcow2 loop 10M
qemu-img rebase -u -f qcow2 -F qcow2 -b loop loop

then libvirt was creating a chain 2 deep before realizing it
had hit a loop; furthermore, virStorageFileChainCheckBroken
was not identifying the chain as broken.  With this patch,
the loop is detected when the chain is only 1 deep; still
enough for storage volume XML to display the file, but now
with a proper error report about where the loop was found.

This patch adds a parameter to virStorageFileGetMetadataRecurse,
so that errors at the top of the chain remain unchanged; messages
issued for backing files now use the name provided by the user
instead of the canonical name (for VDSM, which uses relative
symlinks to device mapper block devices, this is actually more
useful).

* src/util/virstoragefile.c (virStorageFileGetMetadataRecurse):
Add parameter, require canonical path up front.  Mark chain
broken on OOM or loop detection.
(virStorageFileGetMetadata): Pass in canonical name.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-08 13:28:19 -06:00
Jean-Baptiste Rouault
2fe658ad90 vmware: set the driver version
Since commit 7457cbe8 the vmware driver version isn't set anymore.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-08 11:16:33 -06:00
Roman Bogorodskiy
6d7a287c5a tests: add bhyve xml2xml test
The only implemented test for now is domain metadata test.
2014-04-08 20:50:36 +04:00
Roman Bogorodskiy
b291cb5e90 bhyve: add domain metadata support
Implement domainSetMetadata and domainGetMetadata driver calls.
2014-04-08 20:40:39 +04:00
Roman Bogorodskiy
986a07c709 bhyve: fix ATTRIBUTE_NONNULL usage
Fix incorrect ATTRIBUTE_NONNULL usage introduced in 17b17565
which caused build failure:

bhyve/bhyve_driver.c:127:48: error: expected ')'
bhyveDriverGetCapabilities(bhyveConnPtr driver ATTRIBUTE_NONNULL)
                                               ^
bhyve/bhyve_driver.c:127:27: note: to match this '('
bhyveDriverGetCapabilities(bhyveConnPtr driver ATTRIBUTE_NONNULL)

Pushed under the build breaker rule.
2014-04-08 20:30:23 +04:00
Richard Weinberger
21ec87f4f1 LXC: Fix return code evaulation in lxcCheckNetNsSupport()
Commit b9dd878f (util: make it easier to grab only regular command exit)
changed the call semantics of virCommandRun() and therefore of virRun()
too. But lxcCheckNetNsSupport() was not updated.
As consequence of this lxcCheckNetNsSupport always failed and broke LXC.

Signed-off-by: Richard Weinberger <richard@nod.at>
2014-04-08 09:19:24 -06:00
Ján Tomko
27fbfc2a17 Fix incorrect values in redirdev ABI check error
My commit c9123fb introduced this copy-and-paste error.
2014-04-08 15:08:20 +02:00
Daniel P. Berrange
3f671e6c52 Use a static initializer for static mutexes
Now that we ditched our custom pthread impl for Win32, we can
use PTHREAD_MUTEX_INITIALIZER for static mutexes. This avoids
the need to use a virOnce one-time global initializer in a
number of places.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-04-08 13:46:29 +01:00
Daniel P. Berrange
25d5d7057d Add syntax check to validate capitalization of abbreviations
Use a syntax-check rule to force PCI, USB and SCSI instead
of Pci, Usb and Scsi.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-04-08 11:15:55 +01:00
Daniel P. Berrange
3cfa19da22 Replace Pci with PCI throughout
Since it is an abbreviation, PCI should always be fully
capitalized or full lower case, never Pci.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-04-08 11:15:55 +01:00
Daniel P. Berrange
edfe82c7f9 Replace Usb with USB throughout
Since it is an abbreviation, USB should always be fully
capitalized or full lower case, never Usb.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-04-08 11:10:59 +01:00
Daniel P. Berrange
21a2446d92 Replace Scsi with SCSI throughout
Since it is an abbreviation, SCSI should always be fully
capitalized or full lower case, never Scsi.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-04-08 11:10:31 +01:00
Daniel P. Berrange
4607168e77 Switch systemd test to use generic dbus mock
Remove the virsystemdmock.la library and instead make use
of the shared virmockdbus.la library

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-04-08 11:04:09 +01:00
Daniel P. Berrange
2c3ce1de94 Create a re-usable DBus LD_PRELOAD mock library
A number of test suites want to mock the DBus APIs. To avoid
re-inventing the wheel create a re-usable virmockdbus.la
library for LD_PRELOAD usage.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-04-08 11:03:59 +01:00
Daniel P. Berrange
442ba49351 Introduce a new set of helper macros for mocking symbols
Introduce virmock.h which provides some macros to assist in
creation of LD_PRELOAD overrides. When these are used, the
LD_PRELOAD code simply has some stubs which forward to a
wrapper function inside the main test case. This means that
logic for the test no longer has to be split between the
virXXXtest.c and virXXXmock.c files. It will also make it
possible to provide some common reusable modules for mocking
code like DBus.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
2014-04-08 11:03:54 +01:00
Wangrui (K)
79bef03a2c virNetDev{Replace,Restore}MacAddress: Fix memory leak
Functions virNetDevRestoreMacAddress() and virNetDevRestoreMacAddress()
allocate memory for variable @path using virAsprintf(), but they
haven't freed that memory before returning out.

Signed-off-by: Zhang bo <oscar.zhangbo@huawei.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-04-08 10:51:53 +02:00
Michal Privoznik
23419a62c2 bhyveConnectGetCapabilities: Fix double caps unref
At the beginning of the function we gain a reference to the driver
capabilities. Then, we call format function (*) which if failed, unref
over caps is called. Then, at the end another unref occurs.

* - Moreover, the format was not called over gained caps, but over
privconn->caps directly which is not allowed anymore.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-04-08 09:17:01 +02:00
Michal Privoznik
17b17565e6 Simplify bhyveDriverGetCapabilities()
The driver passed as the only argument to the function should never be
NULL so there's no need to check it. After removing it, the whole
function collapses to a single line doing ref over driver
capabilities.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-04-08 09:17:01 +02:00
Michal Privoznik
87f24f3434 bhyve_capabilities: Add Semihalf to Copyright
Since b15a2bbd we have the new bhyve_capabilities.[ch] files.
However, the copyright is held by both Roman and Semihalf.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-04-08 09:17:00 +02:00
Guido Günther
ff32ac06f5 tests: Don't crash when creating the config object fails
As observed when building in a chroot and QEMU_USER doesn't exist
2014-04-08 08:42:16 +02:00
Guido Günther
ed1b5d8e8d tests: link against libxml2
to avoid

CCLD     storagevolxml2argvtest
/usr/bin/ld: ../src/.libs/libvirt_driver_storage_impl.a(libvirt_driver_storage_impl_la-storage_backend.o): undefined reference to symbol 'xmlFreeDoc@@LIBXML2_2.4.30'
//usr/lib/x86_64-linux-gnu/libxml2.so.2: error adding symbols: DSO missing from command line
2014-04-08 08:42:16 +02:00
Eric Blake
8733303933 conf: avoid memleak on NULL path
I noticed that the apparmor code could request metadata even
for a cdrom with no media, which would cause a memory leak of
the hash table used to look for loops in the backing chain.
But even before that, we blindly dereferenced the path for
printing a debug statement, so it is just better to enforce
that this is only used on non-NULL names.

* src/util/virstoragefile.c (virStorageFileGetMetadata): Assume
non-NULL path.
* src/util/virstoragefile.h: Annotate this.
* src/security/virt-aa-helper.c (get_files): Fix caller.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-07 16:31:45 -06:00
Cédric Bosdonnat
8f4ef7e6b6 lxc conf2xml: don't let current vcpus at 0: define won't like it 2014-04-07 10:52:28 -06:00
Wojciech Macek
627d1439c2 bhyve: connectCompareCPU support
Implement support for connectCompareCPU.
2014-04-07 15:34:53 +02:00
Wojciech Macek
b15a2bbd64 bhyve: create capabilities submodule
- Move all capabilities functions to separate file
- Add initCPU
2014-04-07 15:34:53 +02:00
Wojciech Macek
36cf8174b6 bhyve: support for connectBaselineCPU
Implement bhyveConnectBaselineCPU to support OpenStack/Nova
2014-04-07 15:34:53 +02:00
Laine Stump
7284c499e5 interface: dump inactive xml when interface isn't active
Other drivers in libvirt (e.g. network, qemu) will automatically
return the "inactive" (persistent configuration) XML of an object when
that object is inactive. The netcf backend of the interface driver
would always try to return the live status XML of the interface, even
when it was down. Although netcf does return valid XML in that case,
for bond interfaces it is missing almost all of its content, including
the <bond> subelement itself, leading to this error message from
"virsh iface-dumpxml" of a bond interface that is inactive:

  error: XML error: bond interface misses the bond element

(this is because libvirt's validation of the XML returned by netcf
always requires a <bond> element be present).

This patch modifies the interface driver netcf backend to check if the
interface is inactive, and in that case always return the inactive XML
(which will always have a <bond> element, thus eliminating the error
message, as well as making operation more in line with other drivers.

This fixes the following bug:

  https://bugzilla.redhat.com/show_bug.cgi?id=878394
2014-04-07 16:25:40 +03:00
Eric Blake
09567144d6 hash: add common utility functions
I almost wrote a hash value free function that just called
VIR_FREE, then realized I couldn't be the first person to
do that.  Sure enough, it was worth factoring into a common
helper routine.

* src/util/virhash.h (virHashValueFree): New function.
* src/util/virhash.c (virHashValueFree): Implement it.
* src/util/virobject.h (virObjectFreeHashData): New function.
* src/libvirt_private.syms (virhash.h, virobject.h): Export them.
* src/nwfilter/nwfilter_learnipaddr.c (virNWFilterLearnInit): Use
common function.
* src/qemu/qemu_capabilities.c (virQEMUCapsCacheNew): Likewise.
* src/qemu/qemu_command.c (qemuDomainCCWAddressSetCreate):
Likewise.
* src/qemu/qemu_monitor.c (qemuMonitorGetBlockInfo): Likewise.
* src/qemu/qemu_process.c (qemuProcessWaitForMonitor): Likewise.
* src/util/virclosecallbacks.c (virCloseCallbacksNew): Likewise.
* src/util/virkeyfile.c (virKeyFileParseGroup): Likewise.
* tests/qemumonitorjsontest.c
(testQemuMonitorJSONqemuMonitorJSONGetBlockInfo): Likewise.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-07 06:14:45 -06:00
Jincheng Miao
cdc5f3f1a3 tests: build viridentitytest only WITH_ATTR.
Commit d7c4e0036 assumed all SELinux tests depended upon
securityselinuxhelper need xattr support, but forgot to
move viridentitytest under WITH_ATTR.

Reported-by: Nehal J Wani <nehaljw.kkd1@gmail.com>
Signed-off-by: Jincheng Miao <jmiao@redhat.com>
2014-04-07 13:49:24 +02:00
Peter Krempa
56746d8ad6 maint: Correctly detect wether "gluster" cli tool is accessible
The configure definition previously always defined the GLUSTER_CLI macro
and thus the code needing it wasn't compiled out if the tool wasn't
accessible.
2014-04-07 13:20:28 +02:00
Guido Günther
5a05865773 libvirt-guests: avoid bashism
At least Debian uses dash to run the init scripts
2014-04-07 12:51:05 +02:00
Antoni S. Puimedon
d9de144385 QoS: make tc filters match all traffic
Up until now the traffic control filters for the vNIC QoS were
matching only ip traffic. For egress traffic that was unnoticed
because the unmatched traffic would just go to the default htb class
and be shaped anyway. For ingress, though, since the policing of the
rate is done by the filter itself.

The problem is solved by changing protocol to all and making anything
match the filter.

Bug-Url: https://bugzilla.redhat.com/1084444
Signed-off-by: Antoni S. Puimedon <asegurap@redhat.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
2014-04-07 11:35:29 +02:00
Roman Bogorodskiy
4463e438e2 bhyve: add xml2argv tests for console 2014-04-05 19:12:18 +04:00
Roman Bogorodskiy
6c91134de4 bhyve: add console support through nmdm device
nmdm is a FreeBSD driver which allows to create a pair of tty
devices one of which is passed to the guest and second is used
by the client.

This patch adds new 'nmdm' character device type. Its definition
looks this way:

<serial type='nmdm'>
  <source master='/dev/nmdm0A' slave='/dev/nmdm0B'/>
</serial>

Master is passed to the hypervisior and slave is used for client
connection.

Also implement domainOpenConsole() for bhyve driver based on that.
2014-04-05 19:12:18 +04:00
David Shane Holden
6318a80168 bhyve: domain autostart support
Signed-off-by: Roman Bogorodskiy <bogorodskiy@gmail.com>
2014-04-05 14:44:49 +04:00
Eric Blake
9673418ce5 conf: track when storage type is still undetermined
Right now, virStorageFileMetadata tracks bool backingStoreIsFile
for whether the backing string specified in metadata can be
resolved as a file (covering both block and regular file
resources) or is treated as a network protocol.  But when
merging this struct with virStorageSource, it will be easier
to just actually track which type of resource it is, as well
as have a reserved value for the case where the resource type
is unknown (or had an error during probing).

* src/util/virstoragefile.h (virStorageType): Add a placeholder
value, swap order to match similar public enum.
* src/util/virstoragefile.c (virStorage): Update string mapping.
* src/conf/domain_conf.c (virDomainDiskSourceParse)
(virDomainDiskDefParseXML, virDomainDiskDefFormat)
(virDomainDiskSourceFormat): Adjust clients.
* src/conf/snapshot_conf.c (virDomainSnapshotDiskDefParseXML):
Likewise.
* src/qemu/qemu_driver.c
(qemuDomainSnapshotPrepareDiskExternalBackingInactive)
(qemuDomainSnapshotPrepareDiskExternalOverlayActive)
(qemuDomainSnapshotPrepareDiskExternalOverlayInactive)
(qemuDomainSnapshotPrepareDiskInternal)
(qemuDomainSnapshotCreateSingleDiskActive): Likewise.
* src/qemu/qemu_command.c (qemuGetDriveSourceString): Likewise.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-04 10:58:23 -06:00
Eric Blake
d1a1d841db tests: refactor virstoragetest for less stack space
I'm about to add fields to virStorageFileMetadata, which means
also adding fields to the testFileData struct in virstoragetest.
Alas, adding even one pointer on an x86_64 machine gave me a
dreaded compiler error:

virstoragetest.c:712:1: error: the frame size of 4208 bytes is larger than 4096 bytes [-Werror=frame-larger-than=]

After some experimentation, I realized that each test was creating
yet another testChainData (which contains testFileData) on the stack;
forcing the reuse of one of these structures instead of creating a
fresh one each time drastically reduces the size requirements.  While
at it, I also got rid of a lot of intermediate structs, with some
macro magic that lets me directly build up the destination chains
inline.

For a bit more insight into what this patch does:
The old code uses an intermediate variable as a fixed-size array
of structs:
testFileData chain[] = { a, b };
data.files = chain;

In the new code, the use of VIR_FLATTEN_* allows the TEST_CHAIN()
macro to still take a single argument for each chain, but now of
the form '(a, b)', where it is turned into the var-args 'a, b'
multiple arguments understood by TEST_ONE_CHAIN().  Thus, the
new code avoids an intermediate variable, and directly provides
the list of pointers to be assigned into array elements:
data.files = { &a, &b };

* tests/virstoragetest.c (mymain): Rewrite TEST_ONE_CHAIN to
reuse the same struct for each test, and to take the data
inline rather than via intermediate variables.
(testChainData): Use bounded array of pointers instead of
unlimited array of struct.
(testStorageChain): Reflect struct change.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-04 10:48:58 -06:00
Eric Blake
7da61b30a0 tests: use C99 initialization for storage test
Writing this test with C99 initializers will make it easier to test
additions and deletions to struct members as I refactor the code.

* tests/virstoragetest.c (mymain): Rewrite initializers.

Signed-off-by: Eric Blake <eblake@redhat.com>
2014-04-04 10:48:53 -06:00
John Ferlan
30283b8640 NFS storage pool: Fix libvirtd crash due to refactor edit
Commit id '18642d10' refactored the call to virCommandRunRegex()
inside a new function virStorageBackendFileSystemNetFindNFSPoolSources(),
but the cut-n-paste didn't remove the "&state".  Now that state is passed
by reference, it results in a libvirtd core with a messages entry:

"...internal error: unknown storage pool type Unknow"
2014-04-04 09:55:00 -04:00
Olivia Yin
82d858ce42 Define CPUINFO_FILE_LEN and fix maxlen of cpuinfo file for all uses
For example, the file /proc/cpuinfo for 24 cores PowerPC platform is larger than
the previous maximum size 2KB.
It will fail to start libvirtd with the error message as below:
virFileReadAll: Failed to read file '/proc/cpuinfo': Value too large for defined
data type
virSysinfoRead: internal error Failed to open /proc/cpuinfo

This patch defines CPUINFO_FILE_LEN as 10KB which is enough for most architectures.

Signed-off-by: Olivia Yin <Hong-Hua.Yin@freescale.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
2014-04-04 08:08:52 +02:00
Nehal J Wani
34542473e5 Fix Memory Leak in daemon/libvirtd.c
Fixes leak introduced by e562e82f

==4937== 64 bytes in 1 blocks are definitely lost in loss record 270 of 405
==4937==    at 0x4A06BE0: realloc (vg_replace_malloc.c:662)
==4937==    by 0x6FA41C4: __vasprintf_chk (vasprintf_chk.c:90)
==4937==    by 0x50C8D29: virVasprintfInternal (stdio2.h:199)
==4937==    by 0x50C8E3A: virAsprintfInternal (virstring.c:362)
==4937==    by 0x11D01A: main (libvirtd.c:1170)

Signed-off-by: Ján Tomko <jtomko@redhat.com>
2014-04-04 07:45:41 +02:00
Stefan Bader
7a1452f533 libxl: Set disk format for empty cdrom device
The XML config for a CDROM device can be without a source path,
indicating that there is no media present. Without this change
the libxl driver fails to start a guest in that case because
the libxl library checks for the LIBXL_DISK_FORMAT_EMPTY format
type and tries to stat the NULL pointer that gets passed on.

> libxl: error: libxl_device.c:265:libxl__device_disk_set_backend:
> Disk vdev=hdc failed to stat: (null): Bad address

Signed-off-by: Stefan Bader <stefan.bader@canonical.com>
2014-04-03 17:27:46 +02:00