To allow future change of virStorageFileMetadata to virStorageSource we
need to store a full path in the "path" variable as rest of the code
expects it to be a full path. Rename the "path" field to "relPath" to
keep tracking the info but allowing a real "path" field.

Avoid passing lot of arguments into guts of metadata retrieval to fill
the actual structure. Temporarily fill the structure before passing it
down to the actual metadata extractor.

This will later help the inversion of the steps taken to extract the
metadata so that this function can be fully converted to
virStorageSource as the data struct.

This patch also fixes regression when starting a gluster storage pool
where the volumes don't have local representation so that the
canonicalization of the volume's file name failed. Broken by commit
79f11b35

Move the code checking the presence of the backing file to the recursive
worker function instead of the metadata parser. The recursive worker
will later be changed to parse more than just local files and this
change will help the separation.

As we already pass the whole structure down the call path there's no
need to return some stuff in a separate argument. Remove the argument
and tweak callers to avoid breaking semantics.

virStorageFileGetMetadataFromBuf will be refactored later along with the
storage driver.

Don't use the backingStoreRaw as a indication of broken chains. Fill it
always and tweak the broken image chain detector to avoid changing the
semantics.

The new semantics to detect a broken chain is the presence of string in
backingStoreRaw but the lack of the backing chain metadata structure in
the chain.

Now that the raw backing store name is always filled there's no need to
pass the raw name variable separately to fill in case the backing is not
a file. Tweak the function so that it can handle a NULL in that case.

The function isn't used in any other source file. Move it so that it
doesn't need a declaration.

Fixes https://bugzilla.redhat.com/show_bug.cgi?id=1088748

The hotplugged vcpus could fail to show up in the guest if
they were added during boot, see:
https://bugzilla.redhat.com/show_bug.cgi?id=807508

Also change arbitrary to arbitrarily.

The live change of vcpus was not reflected in the domain status
xml and it got lost during libvirtd restart.

https://bugzilla.redhat.com/show_bug.cgi?id=1088703

- do not lose new definition for an active domain
- do not leak oldDef
- do not set dom->id if virDomainSaveConfig() fails
- do not call virObjectUnlock(vm) if vm is NULL

Currently, libvirt is using legacy USB controller as default. There
are problems with VGA which can't work correctly with USB Keyboard and
USB Mouse.

While providing -nodefaults, ppc64 should be specifying the usb
controller explicitly in place of using the legacy
controller(-usb). Qemu spapr initialization code when sees "-usb" adds a
USB Keyboard and USB Mouse by default. And libvirt has added a USB
keyboard and USB mouse.

A recent fix in the in qemu VGA code uncoverd this problem, which
resulted in addition of extra keyboard and mouse to the qemu machine.

This patch is to set pci-ohci as USB default controller.
Signed-off-by: NLi Zhang <zhlcindy@linux.vnet.ibm.com>

Changes:
error: invalid argument: vcpu number out of range 2 > 2
to slightly less confusing:
error: invalid argument: vcpu number out of range 2 > 1

Before, it only returned -1 on failure to shrink the array.
Since the switch to VIR_DELETE_ELEMENT in commit 2133441a,
it returns either 0 or 0.

Remove the pointer from def->cputune.vcpupin after unplugging
the CPU and also free the bitmap contained in the structure
by calling virDomainVcpuPinDel instead of VIR_FREE.

Introduced by commit 0df1a790.

This makes virDomainLookupVcpuPin redundant.

https://bugzilla.redhat.com/show_bug.cgi?id=1088165

Fix minor typos in source comments
Signed-off-by: NEric Blake <eblake@redhat.com>

Among other patches, Natanael Copa contributed a fix for
pyshmem_available when using musl libc that directly affects
libvirt.

* .gnulib: Update to latest, in part for musl libc compat.
Signed-off-by: NEric Blake <eblake@redhat.com>

Implement bhyveDomainCreateXML function.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
Signed-off-by: NRoman Bogorodskiy <bogorodskiy@gmail.com>

When the default was changed from kvm to vfio, the documentation for
hostdev and interface was changed, but the documentation in <network>
was forgotten.

Also document when the default was changed from "always kvm" to "vfio
if available, else kvm" (1.0.5).

The version check is for >=1.3.0, but ComputeCmdFlags
is only called for QEMU <1.2.0.

QEMU only supports it on x86, but we've been assuming it for
all QEMUs when doing QMP capability detection.

https://bugzilla.redhat.com/show_bug.cgi?id=1066145

While running virstoragetest, valgrind pointed out the following
memory leak:

==8142== 2 bytes in 1 blocks are definitely lost in loss record 1 of 92
==8142==    at 0x4A069EE: malloc (vg_replace_malloc.c:270)
==8142==    by 0x4E7B53E: mdir_name (dirname-lgpl.c:78)
==8142==    by 0x4CBE2B0: virStorageFileGetMetadataInternal (virstoragefile.c:595)
==8142==    by 0x4CBE651: virStorageFileGetMetadataFromFDInternal (virstoragefile.c:1086)
==8142==    by 0x4CBEEB4: virStorageFileGetMetadataRecurse (virstoragefile.c:1175)
==8142==    by 0x4CBF1DE: virStorageFileGetMetadata (virstoragefile.c:1270)
==8142==    by 0x4028AD: testStorageChain (virstoragetest.c:275)
==8142==    by 0x407B91: virtTestRun (testutils.c:201)
==8142==    by 0x4039D7: mymain (virstoragetest.c:534)
==8142==    by 0x40830D: virtTestMain (testutils.c:789)
==8142==    by 0x3E6CE1ED1C: (below main) (libc-start.c:226)

...62 times

Fix a locking problem by leaving the function via the cleanup label.
Signed-off-by: NStefan Berger <stefanb@linux.vnet.ibm.com>

Commit id '18642d10' caused a virt-test regression for NFS backend
storage error path checks when running the command:

    'virsh find-storage-pool-sources-as netfs Unknown  '

when the host did not have Gluster installed. Prior to the commit,
the test would fail with the error:

    error: internal error: Child process (/usr/sbin/showmount --no-headers
    --exports Unknown) unexpected exit status 1: clnt_create: RPC: Unknown host

After the commit, the error would be ignored, the call would succeed,
and an empty list of pool sources returned. This was tucked into the
commit message as an expected outcome.

When the target host does not have a GLUSTER_CLI this is a regression
over the previous release. Furthermore, even if Gluster CLI was present,
but had a failure to get devices, the API would return a failure even if
the NFS backend had found devices.

Modify the logic to return failure when the NFS backend check fails and
there's no GLUSTER_CLI or when both backend checks fail.

If either returns success and GLUSTER_CLI is defined, then fetch and return
a list of source devices even if it's empty

I noticed that depending on the <driver> attributes the user passed
in, the output may omit the <driver> element altogether.  For example,
the rerror_policy has had this problem since commit 4bb4109f in Oct
2011.  But in adding testsuite coverage to expose it, I found another
problem: the C code is just fine without a driver name, but the
XML validator required either a name or a cache mode.

* src/conf/domain_conf.c (virDomainDiskDefFormat): Update
conditional.
* docs/schemas/domaincommon.rng (diskDriver): Simplify.
* tests/qemuxml2argvdata/qemuxml2argv-disk-drive-copy-on-read.xml:
* tests/qemuxml2argvdata/qemuxml2argv-disk-drive-copy-on-read.args:
New files.
* tests/qemuxml2argvdata/qemuxml2argv-disk-drive-discard.xml:
Enhance test.
* tests/qemuxml2xmloutdata/qemuxml2xmlout-disk-drive-discard.xml:
Likewise.
* tests/qemuxml2argvtest.c (mymain): New test.
* tests/qemuxml2xmltest.c (mymain): Likewise.
Signed-off-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

To make <disk> schema more maintainable and to allow for moving the
pieces to a common file in the future. It relies on the ability to
override definitions as part of an include, set up in the previous
patch.

The diff is a bit hard to read, because it mixes reindentation
with refactoring; 'git diff -b --patience' may help.

* docs/schemas/domaincommon.rng (disk): Refactor into pieces.
(diskSource, diskSourceFile, diskSourceBlock, diskSourceDir)
(diskSourceVolume: New defines.
(diskSourceNetwork): Revise scope.
* docs/schemas/domainsnapshot.rng (disksnapshot): Adjust.
* tests/domainsnapshotxml2xmlin/disk-seclabel-invalid.xml,
tests/domainsnapshotxml2xmlin/disk-network-seclabel-invalid.xml: New
tests to check seclabel is forbidden in domain snapshot by schema.
Signed-off-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

This patch is my first experience playing with nested grammars,
as documented in http://relaxng.org/tutorial-20011203.html#IDA3PZR.
I plan on doing more overrides in order to make the RelaxNG
grammar mirror the C code refactoring into a common
virStorageSource, but where different clients of that source do
not support the same subset of functionality.  By starting with
something fairly easy to validate, I can make sure my later
patches will be possible.

This patch adds a use of the no-op <ref
name='sourceStartupPolicy'/> to the disksnapshot definition, so
that the snapshot version of a type='file' <source> more closely
resembles the version in domaincommon.  A future patch will merge
the two files into using a common define, but this patch is
sufficient for testing that adding <source
startupPolicy='optional'/> in any of the
tests/domainsnapshotxml2xmlin/*.xml files still gets rejected
unless it occurs within the <domain> subelement, because the
definition of startupPolicy is empty outside of domain.rng.

* docs/schemas/storagecommon.rng (storageStartupPolicy)
(storageSourceExtra): Create no-op defaults.
* docs/schemas/domainsnapshot.rng (domain): Use nested grammar
to avoid restricting <domain>.
(storageSourceExtra): Create new override.
(disksnapshot): Access overrides through common names.
* docs/schemas/domaincommon.rng (disk): Access overrides through
common names.
* docs/schemas/domain.rng (storageStartupPolicy)
(storageSourceExtra): Create new overrides.
Signed-off-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

Domain snapshots should only permit an external snapshot into
a storage format that permits a backing chain, since the new
snapshot file necessarily must be backed by the existing file.
The C code for the qemu driver is a little bit stricter in
currently enforcing only qcow2 or qed, but at the XML parser
level, including virt-xml-validate, it is fairly easy to
enforce that a user can't request a 'raw' external snapshot.

* docs/schemas/storagecommon.rng (storageFormat): Split out...
(storageFormatBacking): ...new sublist.
* docs/schemas/domainsnapshot.rng (disksnapshotdriver): Use new
type.
* src/util/virstoragefile.h (virStorageFileFormat): Rearrange for
easier code management.
* src/util/virstoragefile.c (virStorageFileFormat, fileTypeInfo):
Likewise.
* src/conf/snapshot_conf.c (virDomainSnapshotDiskDefParseXML): Use
new marker to limit selection of formats.
Signed-off-by: NEric Blake <eblake@redhat.com>
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

We had incomplete RelaxNG support for storage formats listed
in virstoragefile.h: commit 027bf2ea added 'vdi' but forgot
to update the <volume> and <domain> xml lists; the <volume>
list was also missing 'fat' and 'vhd'.  Maintaining two lists
is a recipe for them getting out of sync, so make the list
common so that both contexts benefit the next time we add a
format in a single location.

* docs/schemas/domaincommon.rng (storageFormat): Move...
* docs/schemas/storagecommon.rng: ...here, and add vdi.
* docs/schemas/storagevol.rng (formatfile): Use common list.
Signed-off-by: NEric Blake <eblake@redhat.com>

In general, we try to make virt-xml-validate tolerant of input
elements in any order when possible.  However, as written, the
RNG grammar did not permit <source> unless there was an explicit
type= attribute (even though the C code manages just fine by
defaulting to type='file').  After making the attribute optional
on the 'file' branch, I noticed that the use of diskspec was now
redundant with the branch when no <source> was supplied.

View this patch with 'git diff -b' for a better picture of the
schema change.

* docs/schemas/domaincommon.rng (disk): Hoist 'diskspec' out of
choice, make type='file' default, and still preserve interleave.
* tests/qemuxml2xmloutdata/qemuxml2xmlout-disk-source-pool.xml:
* tests/qemuxml2xmloutdata/qemuxml2xmlout-disk-drive-discard.xml:
New files.
* tests/qemuxml2argvdata/qemuxml2argv-disk-source-pool.xml:
* tests/qemuxml2argvdata/qemuxml2argv-disk-drive-discard.xml:
Reorder XML.
* tests/qemuxml2xmltest.c (mymain): Cover new files.
Signed-off-by: NEric Blake <eblake@redhat.com>

Commit 48976981 fixed the build without dbus by only building
the virSystemdPMSupportTarget with SYSTEMD_DAEMON.

Introduce a virDBusMessageUnref wrapper for dbus_message_unref
to let virsystemd.c build without dbus, while still allowing
virsystemdtest to run without SYSTEM_DAEMON.

In order to do that, virNodeSuspendSupportsTargetPMUtils() and
virSystemdPMSupportTarget() are created even when pm-utils and dbus
are compiled out, respectively, but in that case returning -2 meaning
"unavailable" (this return code was already used for unavailability
before).  Error is reported in virNodeSuspendSupportsTarget() only if
both functions returned -2, otherwise the error (or success) is properly
propagated up the stack.
Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

Having two tiny files with a couple definitions didn't make
as much sense as one common file, especially since I plan to
add more definitions and use it in more places.

* docs/schemas/storageencryption.rng: Merge this...
* docs/schemas/storagefilefeatures.rng: ...and this, into...
* docs/schemas/storagecommon.rng: ...this new file.
* docs/schemas/Makefile.am (schema_DATA): Reflect renames.
* docs/schemas/storagevol.rng: Likewise.
* docs/schemas/domaincommon.rng: Likewise.
* libvirt.spec.in: Likewise.
* mingw-libvirt.spec.in: Likewise.
Signed-off-by: NEric Blake <eblake@redhat.com>

When passing migration bandwidth to QEMU, we multiply it by 1024 * 1024
to convert the speed to B/s and the result still needs to fit in
int64_t.

https://bugzilla.redhat.com/show_bug.cgi?id=1083483Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

Another field no longer needed, getting us one step closer to
merging virStorageFileMetadata and virStorageSource.

* src/util/virstoragefile.h (_virStorageFileMetadata): Drop
field.
* src/util/virstoragefile.c (virStorageFileGetMetadataInternal)
(virStorageFileGetMetadataFromFDInternal): Alter signature.
(virStorageFileFreeMetadata, virStorageFileGetMetadataFromBuf)
(virStorageFileGetMetadataFromFD): Adjust clients.
* tests/virstoragetest.c (_testFileData, testStorageChain)
(mymain): Simplify test.
Signed-off-by: NEric Blake <eblake@redhat.com>

Thanks to the testsuite, I feel quite confident that this rewrite
is correct; it gives the same results for all cases except for one.
I can make the argument that _that_ case was a pre-existing bug:
when looking up relative names, the lookup is supposed to be
pegged to the directory that contains the parent qcow2 file.  Thus,
this resolves the fixme first mentioned in commit 367cd69d (even
though I accidentally removed the fixme comment early in 74430fe3).

* src/util/virstoragefile.c (virStorageFileChainLookup): Depend on
new rather than old fields.
* tests/virstoragetest.c (mymain): Adjust test to match fix.
Signed-off-by: NEric Blake <eblake@redhat.com>

The original chain lookup code had to pass in the starting name,
because it was not available in the chain.  But now that we have
added fields to the struct, this parameter is redundant.

* src/util/virstoragefile.h (virStorageFileChainLookup): Alter
signature.
* src/util/virstoragefile.c (virStorageFileChainLookup): Adjust
handling of top of chain.
* src/qemu/qemu_driver.c (qemuDomainBlockCommit): Adjust caller.
* tests/virstoragetest.c (testStorageLookup, mymain): Likewise.
Signed-off-by: NEric Blake <eblake@redhat.com>

The chain lookup function was inconsistent on whether it left
a message in the log when looking up a name that is not found
on the chain (leaving a message for OOM or if name was
relative but not part of the chain), and could litter the log
even when successful (when name was relative but deep in the
chain, use of virFindBackingFile early in the chain would complain
about a file not found).  It's easier to make the function
consistently emit a message exactly once on failure, and to let
all callers rely on the clean semantics.

* src/util/virstoragefile.c (virStorageFileChainLookup): Always
report error on failure.  Simplify relative lookups.
* src/qemu/qemu_driver.c (qemuDomainBlockCommit): Avoid
overwriting error.
Signed-off-by: NEric Blake <eblake@redhat.com>

When checking if two filenames point to the same inode (whether
by hardlink or symlink), sometimes one of the names might be
relative.  This convenience function makes it easier to check.

* src/util/virfile.h (virFileRelLinkPointsTo): New prototype.
* src/util/virfile.c (virFileRelLinkPointsTo): New function.
* src/libvirt_private.syms (virfile.h): Export it.
* src/xen/xm_internal.c (xenXMDomainGetAutostart): Use it.
Signed-off-by: NEric Blake <eblake@redhat.com>

I realized that we had no good test coverage of looking up a
name from within a backing chain, even though code like
block-commit is relying on it.

* tests/virstoragetest.c (testStorageLookup): New function.
(mymain): New tests.
Signed-off-by: NEric Blake <eblake@redhat.com>

This uses the dbus api of systemd to check the power management
capabilities of the node.

Noticed while tweaking the RelaxNG grammar for <disk> elements.

* tests/qemuxml2xmloutdata/qemuxml2xmlout-numad-static-vcpu-no-numatune.xml:
* tests/qemuxml2xmloutdata/qemuxml2xmlout-disk-cdrom-empty.xml:
Drop unused files.
* tests/qemuargv2xmltest.c (testInfo, DO_TEST_FULL): Drop unused
field.
Signed-off-by: NEric Blake <eblake@redhat.com>