Laine Stump [Fri, 21 May 2021 18:32:00 +0000 (14:32 -0400)]
qemu_hotplug.c: add net devices to the domain list earlier
An upcoming patch will be checking if the addition of a new net device
requires adjusting the domain locked memory limit, which must be done
prior to sending the command to qemu to add the new device. But
qemuDomainAdjustMaxMemLock() checks all (and only) the devices that
are currently in the domain definition, and currently we are adding
new net devices to the domain definition only at the very end of the
hotplug operation, after qemu has already executed the device_add
command.
In order for the upcoming patch to work, this patch changes
qemuDomainAttachNetDevice() to add the device to the domain nets list
at an earlier time. It can't be added until after PCI address and
alias name have been determined (because both of those examine
existing devices in the domain to figure out a unique value for the
new device), but must be done before making the qemu monitor call.
Since the device has been added to the list earlier, we need to
potentially remove it on failure. This is done by replacing the
existing call to virDomainNetRemoveHostdev() (which checks if this is
a hostdev net device, and if so removes it from the hostdevs list,
since it could have already been added to that list) with a call to
the new virDomainNetRemoveByObj(), which looks for the device on both
nets and hostdevs lists, and removes it where it finds it.
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Laine Stump [Fri, 21 May 2021 18:28:10 +0000 (14:28 -0400)]
conf: new function virDomainNetRemoveByObj()
virDomainNetRemove() requires the index of the net device you want to
remove from the list, but in some cases you may not have the index
handy, only the object itself (or the object may not have been added
to the domain's list). virDomainNetRemoveByObj() first tries to find
the given object in the nets list, and deletes that if it is found.
As with virDomainNetRemove() it always unconditionally tries to remove
the device from the hostdevs list (in case it is the ridiculous
combined net+hostdev device created for <interface type='hostdev'>).
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Laine Stump [Fri, 21 May 2021 17:03:17 +0000 (13:03 -0400)]
qemu_hotplug.c: don't skip cleanup on failures of qemuDomainAttachNetDevice
We have many places where the earliest error returns from a function
skip any cleanup label at the bottom (the assumption being that it is
so early in the function that there isn't yet anything that needs to
be explicitly undone on failure). But in general it is a bad sign if
there are any direct "return" statements in a function at any time
after there has been a "goto cleanup" - that indicates someone thought
that an earlier point in the code had done something needing cleanup,
so we shouldn't be skipping it.
There were two occurences of a "return -1" after "goto cleanup" in
qemuDomainAttachDeviceNet(). The first of these has been around for a
very long time (since 2013) and my assumption is that the earlier
"goto cleanup" didn't exist at that time (so it was proper), and when
the code further up in the function was added, the this return -1 was
missed. The second was added during a mass change to check the return
from qemuInterfacePrepareSlirp() in several places (commit 99a1cfc43889c6d); in this case it was erroneous from the start.
Change both of these "return -1"s to "goto cleanup". Since we already
have code paths earlier in the function that goto cleanup, this should
not cause any new problem.
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Tue, 25 May 2021 09:32:37 +0000 (11:32 +0200)]
virxml: Avoid double indentation of <metadata/> element
There was a recent change in libxml2 that caused a trouble for
us. To us, <metadata/> in domain or network XMLs are just opaque
value where management application can store whatever data it
finds fit. At XML parser/formatter level, we just make a copy of
the element during parsing and then format it back. For
formatting we use xmlNodeDump() which allows caller to specify
level of indentation. Previously, the indentation was not
applied onto the very first line, but as of v2.9.12-2-g85b1792e
libxml2 is applying indentation also on the first line.
This does not work well with out virBuffer because as soon as we
call virBufferAsprintf() to append <metadata/> element,
virBufferAsprintf() will apply another level of indentation.
Instead of version checking, let's skip any indentation added by
libxml2 before virBufferAsprintf() is called.
Note, the problem is only when telling xmlNodeDump() to use
indentation, i.e. level argument is not zero. Therefore,
virXMLNodeToString() which also calls xmlNodeDump() is safe as it
passes zero.
Tested-by: Bjoern Walk <bwalk@linux.ibm.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Michal Privoznik [Tue, 25 May 2021 09:30:33 +0000 (11:30 +0200)]
virxml: Report error if virXMLFormatMetadata() fails
I guess this is more of an academic problem, because if
<metadata/> content was problematic we would have caught the
error during parsing. Anyway, as is this function returns -1
without any error reported. Fix it by reporting one.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Michal Privoznik [Tue, 25 May 2021 09:21:02 +0000 (11:21 +0200)]
virxml: Introduce and use virXMLFormatMetadata()
So far, we have to places where we format <metadata/> into XMLs:
domain and network. Bot places share the same code. Move it into
a helper function and just call it from those places.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Support for glusterfs with KVM is being dropped in RHEL-9 in the
virtualization stack.
Reviewed-by: Neal Gompa <ngompa13@gmail.com> Reviewed-by: Andrea Bolognani <abologna@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
it also removed the logic that applied to RHEL-8 wrt arch list
and lost PowerPC 64 support on 8. This reverts that part of the
change but with the condition reversed to prioritize the future
state.
Reviewed-by: Neal Gompa <ngompa13@gmail.com> Reviewed-by: Andrea Bolognani <abologna@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Peter Krempa [Tue, 18 May 2021 09:15:50 +0000 (11:15 +0200)]
qemu: Allow <transient> disks with images shared accross VMs
Implement this behaviour by skipping the disks on traditional
commandline and hotplug them before resuming CPUs. That allows to use
the support for hotplugging of transient disks which inherently allows
sharing of the backing image as we open it read-only.
This commit implements the validation code to allow it only with buses
supporting hotplug and the hotplug code while starting up the VM.
When we have such disk we need to issue a system-reset so that firmware
tables are regenerated to allow booting from such device.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Tue, 18 May 2021 09:11:02 +0000 (11:11 +0200)]
conf: Introduce 'shareBacking' for <transient> disks
In case the user wants to share the disk image between multiple VMs the
qemu driver needs to hotplug such disks to instantiate the backends.
Since that doesn't work for all disk configs add a switch to force this
behaviour.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Tue, 18 May 2021 14:18:53 +0000 (16:18 +0200)]
qemuDomainAttachDiskGeneric: Pass the qemu async job type
The qemuDomainAttachDiskGeneric will also be used on startup for
transient disks which share the overlay. The VM startup code passes the
asyncJob around so we need to pass it into qemuDomainAttachDiskGeneric.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Tue, 18 May 2021 15:47:48 +0000 (17:47 +0200)]
qemu: Track creation of <transient> disk overlay individually
In preparation for hotplug of <transient> disks we'll need to track
whether the overlay file was created individually per-disk.
Add 'transientOverlayCreated' to 'struct _qemuDomainDiskPrivate' and
remove 'inhibitDiskTransientDelete' from 'qemuDomainObjPrivate' and
adjust the code for the change.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Fri, 21 May 2021 08:34:55 +0000 (10:34 +0200)]
qemuDomainAttachDiskGeneric: Split up frontend and backend attachment
Split up the monitor contexts to attach the backend of the disk and the
frontend device in preparation for hotplugging transient disks where
we'll need to add the code for adding the transient overlay between
these two steps.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Modify the rollback section to use its own monitor context so that we
can later split up the hotplug into multiple steps and move the
detachment of the extension device into the rollback section rather than
doing it inline.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Fri, 21 May 2021 08:37:36 +0000 (10:37 +0200)]
qemuDomainAttachDiskGeneric: Move PR helper attach into qemuDomainAttachDeviceDiskLiveInternal
Similarly to previous refactors we want to move all hotplug related
setup which isn't strictly relevant to attaching the disk into
qemuDomainAttachDeviceDiskLiveInternal.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Tue, 18 May 2021 14:11:52 +0000 (16:11 +0200)]
qemu: hotplug: Move post-insertion steps of disk hotplug to qemuDomainAttachDeviceDiskLiveInternal
Move the auditing entry and insertion into the disk definition from the
function which deals with qemu to 'qemuDomainAttachDeviceDiskLiveInternal'
which deals with the hotplug related specifics.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Tue, 18 May 2021 13:50:06 +0000 (15:50 +0200)]
qemuDomainAttachDiskGeneric: Move setup of disk into qemuDomainAttachDeviceDiskLiveInternal
qemuDomainAttachDeviceDiskLiveInternal already sets up certain pieces of
the disk definition so it's better suited to move the setup of the
virStorageSource structs, granting access to the storage and allocation
of the alias from qemuDomainAttachDiskGeneric which will be just
handling the qemu interaction.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Move the validation of the SCSI device address and the attachment of the
controller into qemuDomainAttachDeviceDiskLiveInternal as there's no
specific need for a special helper.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Fri, 7 May 2021 12:48:02 +0000 (14:48 +0200)]
qemuBlockStorageSourceChainData: Add handling of 'copy-on-read' filter layer
qemuBlockStorageSourceChainData encapsulates the backend of the disk for
startup and hotplug operations. Add the handling for the copy-on-read
filter so that the hotplug code doesn't need to have separate cleanup.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Fri, 30 Apr 2021 14:29:45 +0000 (16:29 +0200)]
qemuxml2argvtest: Remove pointless tests for keywrapping on s390
There were two negative tests for the keywrapping feature on s390 when
the feature flag was missing. For now both shared the error message thus
worked fine, but with the upcoming patch to move some disk validation
code from the command line formatter to validation code will change the
error message in case the disk capabilities are missing.
Drop the test cases which don't provide any capability and keep those
that have the disk capabilities present as they are sufficient to prove
the feature.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Fri, 30 Apr 2021 06:58:45 +0000 (08:58 +0200)]
qemu: Move 'bootindex' handling for disks out of command line formatter
The logic assigning the bootindices from the legacy boot order
configuration was spread through the command line formatters for the
disk device and for the floppy controller.
This patch adds 'effectiveBootindex' property to the disk private data
which holds the calculated boot index and moves the logic of determining
the boot index into 'qemuProcessPrepareDomainDiskBootorder' called from
'qemuProcessPrepareDomainStorage'.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Fri, 21 May 2021 12:53:43 +0000 (14:53 +0200)]
schema: Allow '0' offset for a <slice> of <disk>
Using slice to cut off the end of the image is a perfectly vaid
configuration. Use 'unsignedInt' instead of 'positiveInteger' for the
'offset' attribute in the XML schema and modify one test case to cover
this use case.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1960993 Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
After previous patches we have two structures:
virCapsHostNUMACellDistance and virNumaDistance which express the
same thing. And have the exact same members (modulo their names).
Drop the former in favor of the latter.
This change means that distances with value of 0 are no longer
printed out into capabilities XML, because domain XML code allows
partial distance specification and thus threats value of 0 as
unspecified by user (see virDomainNumaGetNodeDistance() which
returns the default LOCAL/REMOTE distance for value of 0).
Also, from ACPI 6.1 specification, section 5.2.17 System Locality
Distance Information Table (SLIT):
Distance values of 0-9 are reserved and have no meaning.
Thus we shouldn't be ever reporting 0 in neither domain nor
capabilities XML.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
The virCapsHostNUMACellSiblingInfo structure really represents
distance to other NUMA node. Rename the structure and variables
of that type to make it more obvious.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
First of all, the reverted commit is incomplete. It only sets
cpuset.mems in the VM root cgroup when the API is used but there is no
code that would do the same when the VM is started.
Libvirt never places any process into the VM root cgroup directly. All
the supporting processes like slirp-helper or dbus-daemon are placed
into the emulator sub-cgroup and all the QEMU threads are distributed
between emulator, vcpu* and iothread* sub-cgroups. The scenario
described in the reverted commit can happen only if someone manually
adds any process there which we should not care about.
If we would like to set the limit in the VM root cgroup we need to
introduce better logic:
- set both (old and new) numa group in the VM root cgroup
- change the numa group in all sub-cgroups to new value
- finally set only the new value in the VM root cgroup
The simplest fix now is to revert the commit.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Olaf Hering [Thu, 20 May 2021 10:17:20 +0000 (12:17 +0200)]
libxl: adjust handling of libxl_device_disk objects
libxl objects are supposed to be initialized and disposed.
Correct the usage of libxl_device_disk objects which are allocated on
the stack. Initialize each one prior usage, and dispose them once done.
Adjust libxlMakeDisk to use an already initialized object, it is owned
by the caller.
Adjust libxlMakeDiskList to initialize the list of objects, before they
are filled by libxlMakeDisk. In case of error, the objects are disposed
by libxl_domain_config_dispose.
Signed-off-by: Olaf Hering <olaf@aepfle.de> Reviewed-by: Jim Fehlig <jfehlig@suse.com>
conf: Report alias name of the detached device in error
This is v2 from:
https://listman.redhat.com/archives/libvir-list/2021-May/msg00481.html
I have reworked the code a bit to have only one error report
instead of multiple ones with different combinations of possible
matching items. Suggested by Laine.
qemu: Return -EINVAL to keep qemuDomainOpenFile() consistent
The description of the function says that the return value is a
file descriptor on success and negative errno on failure which is
not true. If the 'if' case with check on security labels fails,
the return value is -1 not -errno. The solution is to return
'-EINVAL' instead.
Signed-off-by: Kristina Hanicova <khanicov@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
qemu: Use qemuDomainOpenFile() in qemuPrepareNVRAM()
Previously, nvram file was created with user/group owner as
'root', rather than specifications defined in libvirtd.conf. The
solution is to call qemuDomainOpenFile(), which creates file with
defined permissions and qemuSecurityDomainSetPathLabel() to set
security label for created nvram file.
Pavel Hrdina [Wed, 12 May 2021 15:43:36 +0000 (17:43 +0200)]
storage: add support for QCOW2 cluster_size option
The default value hard-coded in QEMU (64KiB) is not always the ideal.
Having a possibility to set the cluster_size by user may in specific
use-cases improve performance for QCOW2 images.
QEMU internally has some limits, the value has to be between 512B and
2048KiB and must by power of two, except when the image has Extended L2
Entries the minimal value has to be 16KiB.
Since qemu-img ensures the value is correct and the limit is not always
the same libvirt will not duplicate any of these checks as the error
message from qemu-img is good enough:
Cluster size must be a power of two between 512 and 2048k
qemu: Add check for needed paths for memory devices
When building a commandline for a DIMM memory device with
non-default access mode, the qemuBuildMemoryBackendProps() will
tell QEMU to allocate memory from per-domain memory backing dir.
But later, when preparing the host, the
qemuProcessNeedMemoryBackingPath() does not check for memory
devices at all resulting in per-domain memory backing dir not
being created which upsets QEMU.
Michal Privoznik [Thu, 20 May 2021 12:32:11 +0000 (14:32 +0200)]
lib: Add win-dmp crashdump format
QEMU gained support for 'win-dmp' format in it's release of 3.0,
but libvirt doesn't implement it yet. Fortunately, there not much
needed: new value to virDomainCoreDumpFormat public enum, which
unfortunately means that QEMU driver has to be updated in the
same commit, because of VIR_ENUM_IMPL().
Luckily, we don't need any extra QEMU capability - the code
already checks supported formats via
'query-dump-guest-memory-capability' just before issuing
'dump-guest-memory'.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Michal Privoznik [Thu, 20 May 2021 12:20:15 +0000 (14:20 +0200)]
include: Fix copy-paste error in comment to virDomainCoreDumpFormat enum
The comment to virDomainCoreDumpFormat enum says that new values
can be introduced in the future "as new events are added". Well,
it should have been "formats" instead of "events", obviously.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Peter Krempa [Thu, 20 May 2021 10:17:09 +0000 (12:17 +0200)]
conf: node_device: Fix build with clang
Clang complains:
../libvirt/src/conf/node_device_conf.c:1945:74: error: result of comparison of unsigned enum expression < 0 is always false [-Werror,-Wtautological-unsigned-enum-zero-compare]
if ((mdev->start = virNodeDevMdevStartTypeFromString(starttype)) < 0) {
Fixes: 42a55854993 Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Tim Wiederhake [Wed, 19 May 2021 14:10:12 +0000 (16:10 +0200)]
virDomainDeviceSpaprVioAddressParseXML: Use virXMLProp*
This strictens the parser to disallow negative values (interpreted as
`ULLONG_MAX + value + 1`) for attribute `reg`. Allowing negative
numbers to be interpreted this way makes no sense for this attribute, as it
refers to a 32 bit address space.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Tim Wiederhake [Wed, 19 May 2021 14:10:11 +0000 (16:10 +0200)]
virStorageAdapterParseXML: Use virXMLProp*
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Tim Wiederhake [Wed, 19 May 2021 14:10:08 +0000 (16:10 +0200)]
virDomainIOMMUDefParseXML: Use virXMLProp*
This strictens the parser to disallow negative values (interpreted as
`UINT_MAX + value + 1`) for attribute `aw_bits`. Allowing negative
numbers to be interpreted this way makes no sense for this attribute.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Tim Wiederhake [Wed, 19 May 2021 14:10:06 +0000 (16:10 +0200)]
virDomainAudioDefParseXML: Use virXMLProp*
This strictens the parser to disallow negative values (interpreted as
`UINT_MAX + value + 1`) for attribute `id`. Allowing negative
numbers to be interpreted this way makes no sense for this attribute.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Tim Wiederhake [Wed, 19 May 2021 14:10:05 +0000 (16:10 +0200)]
virDomainAudioDef: Change type of "sdl.driver" to virDomainAudioSDLDriver
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Tim Wiederhake [Wed, 19 May 2021 14:10:03 +0000 (16:10 +0200)]
virDomainAudioPulseAudioParse: Use virXMLProp*
This strictens the parser to disallow negative values (interpreted as
`UINT_MAX + value + 1`) for attribute `latency`. Allowing negative
numbers to be interpreted this way makes no sense for this attribute.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Jonathon Jongsma [Fri, 14 May 2021 21:29:00 +0000 (16:29 -0500)]
tests: nodedevxml2xmltest: test more mdev files
Add the rest of the mdev xml files to the xml2xml test, and include 2
new test cases: one that explicitly specifies 'manual' start, and one
that explicitly specifies 'auto' start.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Boris Fiuczynski [Fri, 14 May 2021 21:28:59 +0000 (16:28 -0500)]
nodedev: support auto-start property for mdevs
This adds a new element to the mdev capabilities xml schema that
represents the start policy for a defined mediated device. The actual
auto-start functionality is handled behind the scenes by mdevctl, but it
wasn't yet hooked up in libvirt.
Signed-off-by: Boris Fiuczynski <fiuczy@linux.ibm.com> Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Jonathon Jongsma [Fri, 14 May 2021 21:28:58 +0000 (16:28 -0500)]
test: move nodedev xml2xml output to a separate dir
Currently, we're loading and parsing the xml from the input file, and
then formatting it and then comparing it directly back to the input
file. This works for now, but is severely limiting as it relies on the
input file being fully-specified and in the exact order as the output
xml format.
If optional elements are ommitted in the input XML, the output xml
may include default values for the ommitted elements and thus the output
will not match the input.
In order to allow more flexibility in testing, save the expected output
to a seprate 'out' directory similar to what most of the other xml2xml
tests are already doing.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Wed, 19 May 2021 09:58:21 +0000 (11:58 +0200)]
virsh-domain: Fix @ret handling in cmdSetmem and cmdSetmaxmem
These functions initialize @ret to true and only after something
fails either they call cleanup code (which consists only from
virshDomainFree()) and return false, or they set ret = false and
carry on (when the failure occurred close to cleanup code).
Switch them to the usual pattern in which ret is initialized to
failure, goto cleanup is used and ret is set to true only after
everything succeeded.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Michal Privoznik [Tue, 18 May 2021 12:24:01 +0000 (14:24 +0200)]
virsh: Fix logic wrt to --current flag in cmdSetmem
In my commit of v7.1.0-rc1~376 I've simplified the logic of
handling @flags. My assumption back then was that calling
virDomainSetMemory() is equivalent to
virDomainSetMemoryFlags(flags = 0). But that is not the case,
because it is equivalent to virDomainSetMemoryFlags(flags =
VIR_DOMAIN_AFFECT_LIVE). Fix the condition that calls the old
API.
Fixes: b5e267e8c59a257652f88d034cb1e0ce1ed4b58a
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1961118 Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Peter Krempa [Mon, 17 May 2021 08:38:25 +0000 (10:38 +0200)]
qemuxml2argvtest: Limit 'disk-network-sheepdog' testcase to qemu-6.0.0
QEMU is dropping sheepdog support in 6.1 so we need to limit the test
case to the latest version supporting sheepdog as it won't be described
by the QMP schema any more.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Mon, 17 May 2021 08:51:15 +0000 (10:51 +0200)]
testQemuInfoSetArgs: Strip default machine alias only for 'latest' test cases
For the real-capabilities test cases testing 'latest' capabilities we
strip off the alias from 'pc' to the appropriate versioned machine type
to prevent update to all tests when bumping qemu capabilities.
Recenly we also started caching the capabilities to prevent re-parsing
the XML all the time. The commit adding the caching kept the alias
stripping prior to cache insertion, thus the cache contains the stripped
alias.
This leads to problem when a test case is added where the 'latest'
equals to the selected version.
Move the machine alias stripping after we create a local copy thus
stripping it only for 'latest' tests.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Tim Wiederhake [Tue, 18 May 2021 15:04:51 +0000 (17:04 +0200)]
virNodeDevCapPCIDevIommuGroupParseXML: Use virXMLProp*
This strictens the parser to disallow negative values (interpreted as
`UINT_MAX + value + 1`) for attribute `number`. Allowing negative
numbers to be interpreted this way makes no sense for this attribute.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Reviewed-by: Laine Stump <laine@redhat.com>
Tim Wiederhake [Tue, 18 May 2021 15:04:50 +0000 (17:04 +0200)]
virDomainAudioOSSParse: Use virXMLProp*
This strictens the parser to disallow negative values (interpreted as
`UINT_MAX + value + 1`) for attribute `bufferCount`. Allowing negative
numbers to be interpreted this way makes no sense for this attribute.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Reviewed-by: Laine Stump <laine@redhat.com>
Tim Wiederhake [Tue, 18 May 2021 15:04:49 +0000 (17:04 +0200)]
virDomainAudioCoreAudioParse: Use virXMLProp*
This strictens the parser to disallow negative values (interpreted as
`UINT_MAX + value + 1`) for attribute `bufferCount`. Allowing negative
numbers to be interpreted this way makes no sense for this attribute.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com> Reviewed-by: Laine Stump <laine@redhat.com>
projects / thirdparty / libvirt.git / log
