qemu: Remove qemu_hotplugpriv.h and qemuDomainRemoveDeviceWaitTime

qemu_hotplugpriv.h is a header file created to share a global variable
called 'qemuDomainRemoveDeviceWaitTime', declared in qemu_hotplug.c,
to other files that would want to change the timeout value
(currently, only tests/qemuhotplugtest.c).

Previous patch deprecated the variable, using qemu_driver->unplugTimeout
to set the timeout instead. This means that the header file is now
unused, and can be safely discarded.

Reviewed-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>

qemu_hotplug.c: adding qemuDomainGetUnplugTimeout

For some architectures and setups, device removal can take
longer than the default 5 seconds. This results in commands
such as 'virsh setvcpus' to fire timeout messages even if
the operation were successful in the guest, confusing the
user.

This patch sets a new 10 seconds unplug timeout for PPC64
guests. All other archs will keep the default 5 seconds
timeout.

Instead of putting 'if PPC64' conditionals inside qemu_hotplug.c
to set the new timeout value, a new function called
qemuDomainGetUnplugTimeout was added. The timeout value is then
retrieved when needed, by passing the correspondent DomainDef
object. This approach allows for different guest architectures
to have distint unplug timeout intervals, regardless of the
host architecture. This design also makes it easier to
modify/enhance the unplug timeout logic in the future
(allow for special timeouts for TCG domains, for example).

A new mock file was created to work with qemuhotplugtest.c,
given that the test timeout is significantly shorter than
the actual timeout value in qemu_hotplug.c.

The now unused 'qemuDomainRemoveDeviceWaitTime' global can't
be simply erased from qemu_hotplug.c though. Next patch will
remove it properly.

Reviewed-by: Cole Robinson <crobinso@redhat.com>
Suggested-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>

conf: use glib allocation when parsing video props

In preparation for some other improvements, switch to using glib
allocation and g_autofree when parsing the 'acceleration' and
'resolution' properties of the video device.

Reviewed-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>

conf: remove unnecessary NULL checks

Just above in the function, we return from the function if either x or y
are NULL, so there's no need to re-check whether x or y are NULL.

Reviewed-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>

qemu: fix documentation for video resolution

The video resolution support that was introduced in
7286279797a34b3083d85bc4556432b5e7ad9fff is specified as a <resolution>
sub-element of <model>, not optional attributes of model.

Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>

conf: Fix memory leak caused by missing VIR_FREE for video resolution.

Commit 72862797 introduced resolution settings for QEMU video drivers.
It includes a new structure inside video definition. So, the code needs
to clear pointer allocation for that structure into clear function
virDomainVideoDefClear(). This commit adds this missing VIR_FREE().

Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Signed-off-by: Julio Faracco <jcfaracco@gmail.com>

Remove VIR_STRNDUP usage that subtracts from a non-NULL pointer

Use g_strndup in all the cases where we check upfront whether a pointer
is non-NULL and then use it to calculate the copied length.

Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

Remove VIR_STRNDUP usage that passes -1

Replace all the usage of
  VIR_STRNDUP(dest, b, p ? p - b : -1)
with separate calls to g_strndup/g_strdup.

Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

Remove VIR_STRDUP usage that snuck in

Fixes: 224d269f19f0a6c496dd2218f934a54742d51708
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

locking: fix build with older sanlock

../../src/locking/lock_driver_sanlock.c:106:17: error: incompatible pointer types
assigning to 'char **' from 'char *' [-Werror,-Wincompatible-pointer-types]
        message = g_strdup_printf(_("sanlock error %d"), err);
                ^ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fixes: b1d58418aa5051c7aafa4519b332007e7c73c261
Signed-off-by: Ján Tomko <jtomko@redhat.com>

util: buffer: remove virBufferSetChildIndent

Promote usage of separate buffers for separate formatting passes by
removing the now unused virBufferSetChildIndent.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: domain: Convert child buffers to use VIR_BUFFER_INIT_CHILD

Use the new helper to initialize child XML element buffers.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: cpu: Convert child buffers to use VIR_BUFFER_INIT_CHILD

Use the new helper to initialize child XML element buffers.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: caps: sysinfo: Convert child buffers to use VIR_BUFFER_INIT_CHILD

Use the new helper to initialize child XML element buffers.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

util: sysinfo: Convert child buffers to use VIR_BUFFER_INIT_CHILD

Use the new helper to initialize child XML element buffers.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

qemu: domain: Convert child buffers to use VIR_BUFFER_INIT_CHILD

Use the new helper to initialize child XML element buffers.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

util: buffer: Add init macro for automatically setting child XML indent

Add a new macro which initializes a virBuffer on the stack and also sets
the indent level to be used for child XML element formatting.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: capabilities: Refactor API for setting guest capability features

Remove the need to pass around strings and switch to the enum values
instead.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: Refactor storage of guest capabilities

The capabilities are declared in the XML schema so passing feature names
as strings from hypervisor drivers makes no sense.

Additionally some of the features expose so called 'toggles' while
others not. This knowledge was encoded by a bunch of 'STREQ's in the
formatter.

Change all of this by declaring the features as an enum and use it
instead of a dynamically allocated array.

Presence of 'toggles' is encoded together with the conversion strings
rather than in the formatter directly.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

schema: capabilities: Add 'hap' feature flag

The libxl driver exposes a 'hap' feature in the capability XML but our
schema didn't cover it.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

qemu: driver: Remove unused 'driver' from qemuDomainSnapshotFSFreeze/Thaw

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>

cpu_map: Drop pconfig from Icelake-Server CPU model

The pconfig feature was enabled in QEMU by accident in 3.1.0. All other
newer versions do not support it and it was removed from the
Icelake-Server CPU model in QEMU.

We don't normally change our CPU models even when QEMU does so to avoid
breaking migrations between different versions of libvirt. But we can
safely do so in this specific case. QEMU never supported enabling
pconfig so any domain which was able to start has pconfig disabled.

With a small compatibility hack which explicitly disables pconfig when
CPU model equals Icelake-Server in migratable domain definition, only
one migration scenario stays broken (and there's nothing we can do about
it): from any host to a host with libvirt < 5.10.0 and QEMU > 3.1.0.

https://bugzilla.redhat.com/show_bug.cgi?id=1749672

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

cputest: Add data for Ice Lake Server CPU

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

cpu_map: Drop comments about ospke

QEMU does not support setting this feature on the command line anymore.
We don't need to explain why it is not included in CPU models then.

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qemu: Drop disabled CPU features unknown to QEMU

When a CPU definition wants to explicitly disable some features that are
unknown to QEMU, we can safely drop them from the definition before
starting QEMU. Naturally QEMU won't enable such features implicitly.

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qemuxml2*test: Add tests for Icelake-Server,-pconfig

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

cpu_conf: Pass policy to CPU feature filtering callbacks

Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

bhyve_device.c: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

tools: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

suspend.c: remove unneeded cleanup label

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

bridge_driver.c: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

libxl_driver.c: remove unneeded cleanup label

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

libvirt.c: remove unneeded cleanup label

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

util: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

vz_sdk.c: remove unneeded cleanup label

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

rpc: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

qemu_monitor_json.c: remove unneeded cleanup label

qemuMonitorJSONBlockIoThrottleInfo uses a macro called
GET_THROTTLE_STATS that's defined outside of the function,
which references a 'cleanup' label. GET_THROTTLE_STATS is
only used inside qemuMonitorJSONBlockIoThrottleInfo (in fact,
the macro is undef right after it) thus it is safe to erase
the 'cleanup' reference inside the macro, then proceed
with the usual cleanup label removal inside
qemuMonitorJSONBlockIoThrottleInfo.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

secret_driver.c: remove unneeded cleanup label

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

node_device: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

lxc: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

remote: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

nwfilter: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

storage: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: remove unneeded cleanup labels

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

docs: hacking: add missing code element

Wrap the 'g_renew()' call for VIR_SHRINK_N in <code>.

Signed-off-by: Ján Tomko <jtomko@redhat.com>

qemu: hotplug: ensure address generation for vfio-ccw

When attaching a mediated host device of model vfio-ccw without
specifying a guest-address, none is generated by libvirt. Let's fix this
and make sure to generate a device address during live-hotplug.

Reviewed-by: Boris Fiuczynski <fiuczy@linux.ibm.com>
Signed-off-by: Bjoern Walk <bwalk@linux.ibm.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

virpcimock: Make @fakerootdir static

Since we are not passing the @fakerootdir variable to any inline function
anymore, we can make the variable static.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

Drop virVasprintf()

Now that function is no longer used, it can be dropped.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

virstring: Drop virVasprintfQuiet()

This macro is no longer used and therefore can be dropped.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>

Drop virAsprintf()

Now that function is no longer used, it can be dropped.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

virstring: Drop virAsprintfQuiet()

This macro is no longer used and therefore can be dropped.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>

Use g_strdup_vprintf() instead of virVasprintf() everywhere

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

Use g_strdup_printf instead of virAsprintf everywhere

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

tools: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

tests: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

vz: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

vmx: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

vmware: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

vbox: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

secret: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

util: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

test: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

storage: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

security: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

rpc: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

remote: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

qemu: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

phyp: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

openvz: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

nwfilter: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

node_device: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

network: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

lxc: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

logging: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

locking: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

libxl: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

interface: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

hyperv: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

esx: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

cpu: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>

bhyve: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

access: Use g_strdup_printf() instead of virAsprintf()

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

src: Wrap long lines in Makefiles

In my previous commit of v5.9.0-83-g4ae7181376 I've fixed
check-aclrules but whilst doing so, I forgot to wrap long
lines that I've added.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>

spec: fix vpath build on RHEL 7

Macro _vpath_builddir is not defined so we have to define it ourselves.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

make check-driverimpls work again

Previously we generated all source files into $srcdir which is no
longer true. This means that we can't just blindly prepend each
source file with $srcdir.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

src: Make check-aclrules work again

Previously we generated all source files into $srcdir which is no
longer true. This means that we can't just blindly prepend each
source file with $srcdir.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

conf: capabilities: Modernize virCapabilitiesFormatMemoryBandwidth

Use virXMLFormatElement and the automatic memory handlers to simplfy the
code.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

conf: caps: Modernize virCapabilitiesFormatCaches

Use automatic memory freeing and use virXMLFormatElement instead of open
coding it.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: turn virDomainMemtuneFormat void

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: domain: Split up formatting of <memtune> and <memoryBacking>

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: caps: Automaticaly free 'cpus_str'

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: Rename virDomainCapsFeature to virDomainProcessCapsFeature

The enum name sounds too generic. It in fact describes the capabilities
of the process, thus add 'Process' to the name.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

conf: storagecaps: Fix broken attempt at being const-correct

The code formatting storage capabilities faithfully copied the wrong use
of 'const' from domain capabilities.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

conf: domaincaps: Fix broken attempt at being const-correct

'virBlahPtr const blah' results into modification to the value of 'blah'
triggering compilation error rather than the modification of the virBlah
struct the pointer points to.

All of the domain capability formatting code was broken in this regard.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

qemu: caps: Make capability filler functions void

Most of them don't have anything to report so we can simplify the logic.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

qemu: caps: Rework memory allocation in virQEMUCapsFillDomainFeatureSEVCaps

Use g_new0 instead of VIR_ALLOC to avoid error cases.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>

storage: improve the while loop virStorageBackendFileSystemIsMounted

Move virStorageBackendFileSystemGetPoolSource outside of the while loop

Signed-off-by: Yi Li <yili@winhong.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

qemu: remove duplicate header files

"#include vircgroup.h" appears in both qemu_cgroup.h and
qemu_cgroup.c, and qemu_cgroup.c contains qemu_cgroup.h,
so remove the duplicate declarations.

Signed-off-by: Mao Zhongyi <maozhongyi@cmss.chinamobile.com>
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>

lxc: remove duplicate header files

"#include vircgroup.h" appears in both lxc_cgroup.h and
lxc_cgroup.c, and lxc_cgroup.c contains lxc_cgroup.h,
so remove the duplicate declarations.

Signed-off-by: Mao Zhongyi <maozhongyi@cmss.chinamobile.com>
Signed-off-by: Zhang Shengju <zhangshengju@cmss.chinamobile.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>