Peter Krempa [Fri, 30 Mar 2018 10:35:52 +0000 (12:35 +0200)]
qemu: monitor: Don't resist stealing 'actions' in qemuMonitorJSONTransaction
Rather than trying to prevent stealing of the 'actions' virJSONValue
into the monitor command replace the code so that it does the same
thing, since 'actions' was actually not really used after calling the
monitor.
Peter Krempa [Thu, 29 Mar 2018 18:38:35 +0000 (20:38 +0200)]
util: json: Add accessor for geting a VIR_JSON_TYPE_NUMBER as string
Sometimes it's desired to get a JSON number as string. Add a helper.
This will help in cases where we'd want to convert the internal type from
string to something else.
Peter Krempa [Fri, 30 Mar 2018 09:12:57 +0000 (11:12 +0200)]
util: json: Fix freeing of objects appended to virJSONValue
It was not possible to determine whether virJSONValueObjectAddVArgs and
the functions using it would consume a virJSONValue or not when used
with the 'a' or 'A' modifier depending on when the loop failed.
Fix this by passing in a pointer to the pointer so that it can be
cleared once it's successfully consumed and the callers don't have to
second-guess leaving a chance of leaking or double freeing the value
depending on the ordering.
docs: formatdomain: Clarify CPU feature policy option "require"
Using the 'require' value for the 'policy' attribute indicates
that the guest will have the feature so the host CPU does not
need to support it if the hypervisor can emulate it.
E.g. 'x2apic' is emulated by QEMU even if the host does not support it:
Pino Toscano [Wed, 28 Mar 2018 11:40:21 +0000 (13:40 +0200)]
vmx: convert any amount of NICs
Scan the parsed VMX file, and gather the biggest index of the network
interfaces there: this way, it is possible to parse all the available
network interfaces, instead of just 4 maximum.
Add the VMX file attached to RHBZ#1560917 as testcase esx-in-the-wild-8.
Pino Toscano [Wed, 28 Mar 2018 11:40:19 +0000 (13:40 +0200)]
vmx: allocate space for network interfaces if needed
Dynamically grow the array of network interfaces for each interface
read, instead of using a single array of size 4. This way, in the
future it will be easier to not limit the number of network interfaces
(which this patch still does not change).
Pino Toscano [Wed, 28 Mar 2018 11:40:18 +0000 (13:40 +0200)]
vmx: check for present/enabled devices earlier
When parsing filesystems, network interfaces, serial ports, and
parallel ports, check earlier whether they are present/enabled, delaying
the allocation of the objects.
This is mostly a small optimization, with no behaviour change.
For historical reasons we've used 32 bytes long static buffer for
storing PTY aliases. This breaks users scenario where they try to
start a machine with user alias consisting of "ua-$uuid".
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Rainer Müller [Fri, 23 Mar 2018 20:36:34 +0000 (21:36 +0100)]
build: Remove --with-xml-catalog-file option
After validation against XHTML 1.0 was dropped in f802c9de0,
the XML_CATALOG_FILE is not in use anymore. Therefore the checks in
configure can be removed.
Andrea Bolognani [Fri, 23 Mar 2018 12:31:14 +0000 (13:31 +0100)]
tests: qemucapabilities: Drop mostly duplicated data
When GIC support was introduced (QEMU 2.6 timeframe) we needed
to make sure both GICv2 hardware and GICv3 hardware were handled
correctly, and that was achieved by having separate capabilities
data for each.
Now that we have capabilities data for several QEMU versions we
can stop storing data for GICv2 and GICv3 hardware separately,
and instead have GICv2 data for QEMU <= 2.10 and GICv3 data for
QEMU >= 2.12, without losing any coverage.
Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Farhan Ali [Fri, 23 Mar 2018 17:22:40 +0000 (13:22 -0400)]
qemu: Change default video model type to virtio for S390
S390 guests can only support a virtio-gpu-ccw device as a video
device. So set default video model type to VIR_DOMAIN_VIDEO_TYPE_VIRTIO
for S390 guests.
Signed-off-by: Farhan Ali <alifm@linux.vnet.ibm.com>
conf: avoid reporting errors when network driver is disabled
In previous releases all these methods were a no-op if the network
driver is disabled. These helper methods are called unconditionally for
all types of network interface, so must be no-ops if missing. Other code
will already generate an error if the network driver is disabled and a
NIC with type=network is used.
Reviewed-by: Laine Stump <laine@laine.org> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The libvirtd daemon has some arbitrary logic to drop privileges, but
only on Solaris platforms. This was added during Xen days, when Xen was
the only driver running in libvirtd. There's no expectation or testing
that this works with the new libxl stack, nor whether dropping
privileges breaks any of the secondary drivers. Finally, we'll be
splitting drivers out into their own independant daemons, so this won't
be applicable to libvirtd in future anyway.
The remote driver client meanwhile arbitrarily disables daemon
auto-spawn when connecting as non-root, breaking a key feature of
libvirt unprivileged connections.
Since we've not had any contributions for Solaris since circa 2012
and we don't do any CI testing we should consider this platform
unmaintained and thus reasonable to remove this cruft. If someone steps
forward to maintain Solaris again, this code would need re-evaluating to
come up with something more targetted.
There's various __sun conditionals in the Xen driver code, but those are
not touched. This is all for the legacy Xen driver, which will be
entirely removed at some point in future, so not benefit to hacking out
just the Solaris parts.
Reviewed-by: Andrea Bolognani <abologna@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Similarly to b133fac356324c3 we need to look up alias of CCID
controller when constructing smartcard command line instead of
relying on broken assumption it will always be 'ccid0'. After
user aliases it can be anything.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Peter Krempa [Thu, 22 Mar 2018 18:05:26 +0000 (19:05 +0100)]
tests: qemumonitor: Allow testing schema for fake monitor interactions
Add infrastructure that will allow testing schema of the commands we
pass to the fake monitor object, so that we can make sure that it
actually does something.
Peter Krempa [Thu, 22 Mar 2018 16:07:43 +0000 (17:07 +0100)]
tests: qemu: Add infrastructure for QAPI schema testing
Add a function which will allow to test whether a JSON object conforms
to the QAPI schema. This greatly helps when developing formatters for
new JSON objects and will help make sure that the code will not break in
cases which have unit tests but were actually not function-tested
(mostly various disk access protocols).
Peter Krempa [Wed, 21 Mar 2018 15:51:29 +0000 (16:51 +0100)]
qemu: qapi: Return correct entry in virQEMUQAPISchemaTraverse
virQEMUQAPISchemaTraverse would return previous-to-last queried item on
a query. It would not be a problem if checking if the given path exists
since error reporting works properly but if the caller is interested in
the result, it would be wrong.
Prafull [Thu, 22 Mar 2018 14:34:35 +0000 (20:04 +0530)]
lxc: report error message raised by the failing function
The code that calls VIR_WARN after a function fails, doesn't
report the error message raised by the failing function.
Such error messages are now reported in lxc/lxc_driver.c
Signed-off-by: Prafullkumar Tale <talep158@gmail.com>
rpc: avoid duplicating config in virtlockd/virtlogd augeas tests
Most of the augeas test files use ::CONFIG:: to pull in the master
config file for testing. This ensures that entries added to the config
file are actually tested by augeas.
This identified the missing admin_max_clients example in the virtlogd
config file, which in turn prompted a change in description of the
max_clients parameter, since these daemons don't have separate
readonly & readwrite sockets.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
A earlier commit changed the global log buffer so that it only
records messages that are explicitly requested via the log
filters setting. This removes the performance burden, and
improves the signal/noise ratio for messages in the global
buffer. At the same time though, it is somewhat pointless, since
all the recorded log messages are already going to be sent to an
explicit log output like syslog, stderr or the journal. The
global log buffer is thus just duplicating this data on stderr
upon crash.
The log_buffer_size config parameter is left in the augeas
lens to prevent breakage for users on upgrade. It is however
completely ignored hereafter.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
This was in the 1.2.3 release, and 4 years is sufficient time for a
graceful upgrade path for augeas, so all remaining traces are now
removed.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
virt-aa-helper: test: check for expected profile content
So far the virt-aa-helper tests only checked the return code and thereby
catched aborts like issues failing to parse the XML. But there is one
category of virt-aa-helper issues so far untested - not generating the
expected rule.
This adds a basic grep based checks after each test to match against the
rule that is expected to be added by the test.
Acked-by: Jamie Strandboge <jamie@canonical.com> Signed-off-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
nvdimm memory is backed by a path on the host. This currently works only via
hotplug where the AppArmor label is created via the domain label callbacks.
This adds the virt-aa-helper support for nvdimm memory devices to generate
rules for the needed paths from the initial guest definition as well.
Example in domain xml:
<memory model='nvdimm'>
<source>
<path>/tmp/nvdimm-base</path>
</source>
<target>
<size unit='KiB'>524288</size>
<node>0</node>
</target>
</memory>
Works to start now and creates:
"/tmp/nvdimm-base" rw,
Fixes: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1757085 Acked-by: Jamie Strandboge <jamie@canonical.com> Signed-off-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
virt-aa-helper: generate rules for passthrough input devices
Input devices can passthrough an event device. This currently works only via
hotplug where the AppArmor label is created via the domain label callbacks.
This adds the virt-aa-helper support for passthrough input devices to generate
rules for the needed paths from the initial guest definition as well.
Example in domain xml:
<input type='passthrough' bus='virtio'>
<source evdev='/dev/input/event0' />
</input>
Works to start now and creates:
"/dev/input/event0" rw,
Fixes: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1757085 Acked-by: Jamie Strandboge <jamie@canonical.com> Signed-off-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
d8116b5a "security: Introduce functions for input device hot(un)plug"
implemented the code (Set|Restore)InputLabel for several security modules,
this patch adds an AppArmor implementation for it as well.
That fixes hot-plugging event input devices by generating a rule for the
path that needs to be accessed.
Example hot adding:
<input type='passthrough' bus='virtio'>
<source evdev='/dev/input/event0' />
</input>
Creates now:
"/dev/input/event0" rwk,
Fixes: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1755153 Acked-by: Jamie Strandboge <jamie@canonical.com> Signed-off-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
Recent changes have made implementing this mandatory to hot add any
memory.
Implementing this in apparmor fixes this as well as allows hot-add of nvdimm
tpye memory with an nvdimmPath set generating a AppArmor rule for that
path.
Example hot adding:
<memory model='nvdimm'>
<source>
<path>/tmp/nvdimm-test</path>
</source>
<target>
<size unit='KiB'>524288</size>
<node>0</node>
</target>
</memory>
Creates now:
"/tmp/nvdimm-test" rwk,
Fixes: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1755153 Acked-by: Jamie Strandboge <jamie@canonical.com> Signed-off-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
Michal Privoznik [Wed, 21 Mar 2018 16:17:54 +0000 (17:17 +0100)]
virCommandPassFD: Give name to flags
The flags passed to virCommandPassFD() are unnamed and
documentation to this function doesn't list them either.
Give them name and mention it in documentation to functions
using them.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Laine Stump <laine@laine.org>
Jim Fehlig [Thu, 22 Feb 2018 18:52:56 +0000 (11:52 -0700)]
libxl: don't hardcode scheduler weight
Long ago in commit dfa1e1dd53 the scheduler weight was accidentally
hardcoded to 1000. Weight is a setting with no unit since it is
relative to the weight of other domains. If no weight is specified,
libxl defaults to 256.
Instead of hardcoding the weight to 1000, honor any <shares> specified
in <cputune>. libvirt's notion of shares is synonomous to libxl's
scheduler weight setting. If shares is unspecified, defer default
weight setting to libxl.
Removing the hardcoded weight required some test fixup. While at it,
add an explicit test for <shares> conversion to scheduler weight.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Jim Fehlig [Fri, 16 Mar 2018 21:00:17 +0000 (15:00 -0600)]
libxl: dont dereference NULL libxlDomainObjPrivatePtr
In libxlDomainMigrationPrepare it is possible to dereference a NULL
libxlDomainObjPrivatePtr in early error paths. Check for a valid
'priv' before using it.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Jim Fehlig [Fri, 16 Mar 2018 21:22:45 +0000 (15:22 -0600)]
libxl: call EndJob in error case
If starting the domain fails in libxlDomainCreateXML, we mistakenly
jumped to cleanup without calling libxlDomainObjEndJob. Remove the
jump to 'cleanup'.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Jim Fehlig [Fri, 16 Mar 2018 21:15:07 +0000 (15:15 -0600)]
libxl: lock virDomainObj after ListRemove
Most libxl driver API use the pattern of lock and add a ref to
virDomainObj, perform API, then decrement ref and unlock in
virDomainEndAPI. In some cases the API may call
virDomainObjListRemove, which unlocks the virDomainObj. Relock
the object in such cases so EndAPI is called with a locked object.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
docs: update all GIT repo examples to use https:// protocol
The https:// protocol is much more reliably usable than git:// when
faced with unreasonably strict firewalls. The libvirt.org web server is
now setup to support the smart https:// protocol, which is just as fast
as git://, so change all the docs to use https://
Reviewed-by: Pavel Hrdina <phrdina@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
docs: remove obsolete docs about gitorious and cvs access
The gitorious.org service went away a long time ago now, and our main
download.html page tells people where all the official mirrors are
for every component.
Meanwhile telling people about CVS is a bad joke in 2018, and the CVS
server no longer exists on libvirt.org
Reviewed-by: Pavel Hrdina <phrdina@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
When defining a domain that has <interface type='hostdev'/> our
parser creates two entries in virDomainDef: one for <interface/>
and one for <hostdev/>. However, some info is shared between the
two which makes user alias validation fail because alias belongs
to the set of shared info.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
Andrea Bolognani [Mon, 19 Mar 2018 17:06:09 +0000 (18:06 +0100)]
mingw: Use Python 3 for building
While RHEL / CentOS are still using Python 2 for the time being,
Fedora has already switched to Python 3 as the default Python
interpreter a while ago, so on that OS it doesn't make sense to
drag in Python 2 anymore.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Andrea Bolognani [Mon, 19 Mar 2018 16:32:47 +0000 (17:32 +0100)]
spec: Use Python 3 for building when possible
While RHEL / CentOS are still using Python 2 for the time being,
Fedora has already switched to Python 3 as the default Python
interpreter a while ago, so on that OS it doesn't make sense to
drag in Python 2 anymore; the same applies to future RHEL versions.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Andrea Bolognani [Mon, 19 Mar 2018 16:12:39 +0000 (17:12 +0100)]
configure: Allow (and prefer) Python 3
Our build process no longer depends on Python 2, so we can
finally allow Python 3 to satisfy our requirement for a Python
interpreter.
Since several distributions have now switched to installing
Python 3 by default and Python 2 is on its way out, prefer the
former when both are available.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Andrea Bolognani [Mon, 19 Mar 2018 17:04:34 +0000 (18:04 +0100)]
mingw: Require Fedora
Although it was never formally specified, it was always expected
that the mingw RPM build would happen on Fedora, if anything
because RHEL / CentOS don't ship the necessary mingw dependencies.
Make this fact explicit by erroring out if that's not the case,
the same way we already do in the main spec file.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Improve readability and reduce the complexity of the code that is
searching for string tokens (i.e. characters surrounded by a single
or double quote).
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com> Signed-off-by: Radostin Stoyanov <rstoyanov1@gmail.com>
The method strip_lead_star() removes a single leading asterisk
character from a string by ignoring leading whitespace, otherwise it
returns the original string.
This could be achieved with a single if-statement followed by replace.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com> Signed-off-by: Radostin Stoyanov <rstoyanov1@gmail.com>
projects / libvirt.git / log