The parent bridge configuration of the current device
should be read and reset, instead of reading the current
device configuration.
Signed-off-by: He Xin <hexin15@baidu.com>
Signed-off-by: Liu Qi <liuqi16@baidu.com>
Signed-off-by: Zhang Yu <zhangyu31@baidu.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since users can enable/disable drivers at compile time, it may
happen that @drivers array is in fact empty (in both its
occurrences within the function). This means that
ARRAY_CARDINALITY() returns 0UL which makes gcc unhappy because
of loop condition:
i < ARRAY_CARDINALITY(drivers)
GCC complains that @i is unsigned and comparing an unsigned value
against 0 is always false. However, changing the type of @i to
ssize_t is not enough, because compiler still sees the unsigned
zero. The solution is to typecast the ARRAY_CARDINALITY().
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Jim Fehlig <jfehlig@suse.com>
Introduced in commit 4a6ee53581.
Signed-off-by: Andrea Bolognani <abologna@redhat.com>
(cherry picked from commit df1b5cf02e)
Reintroduced-by: fb275b7673
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Add virStorageFileSupportsCreate which allows silent check whether
virStorageFileCreate is implemented.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Modify the return value so that callers don't have to repeat logic.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
This helper extracts common lifecycle action code from both
testDomainShutdownFlags and testDomainReboot.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
All the callers left require virPCIDeviceConfigOpen to be fatal
and only use read-only access to the config file.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
For callers that only need read-only access and don't want
an error reported.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Only a handful of function need write access to the PCI config
space. Create a wrapper function for those so that we can
open it read only by default.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
As a side effect, this also silences the possible:
internal error: Unable to get DBus system bus connection:
Failed to connect to socket /run/dbus/system_bus_socket:
No such file or directory
error, since we check upfront whether dbus is available.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Look up the binary name upfront to avoid the error:
Cannot find 'pm-is-supported' in path: No such file or directory
In that case, we just assume nodesuspend is not available.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Get rid of the ret variable as well as the cleanup label.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
When QEMU supports flushing caches at the end of migration, we can
safely allow migration even if disk/driver/@cache is not none nor
directsync.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Acked-By: Peter Krempa <pkrempa@redhat.com>
QEMU 4.0.0 and newer automatically drops caches at the end of migration.
Let's check for this capability so that we can allow migration when disk
cache is turned on.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Acked-By: Peter Krempa <pkrempa@redhat.com>
The original message was logically incorrect: cache != none or cache !=
directsync is always true. But even replacing "or" with "and" doesn't
make it more readable for humans.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Acked-By: Peter Krempa <pkrempa@redhat.com>
In the first stage of incoming migration (qemuMigrationDstPrepareAny) we
call qemuMigrationEatCookie when there's no vm object created yet and
thus we don't have any private data to pass.
Broken by me in commit v5.6.0-109-gbf15b145ec.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
This reverts commit f38d553e2d.
Gnulib's make coverage (or init-coverage, build-coverage, gen-coverage)
is not a 1-1 replacement for the original configure option. Our old
--enable-test-coverage seems to be close to gnulib's make build-coverage
except gnulib runs lcov in that phase and the build actually fails for
me even before lcov is run. And since we want to be able to just build
libvirt without running lcov, I suggest reverting to our own
implementation.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Acked-By: Eric Blake <eblake@redhat.com>
Back in July 2010, commit 6ea90b84 (meant to resolve
https://bugzilla.redhat.com/571991 ) added code to set the MAC address
of any tap device to the associated guest interface's MAC, but with
the first byte replaced with 0xFE. This was done in order to assure
that
1) the tap MAC and guest interface MAC were different (otherwise L2
forwarding through the tap would not work, and the kernel would
repeatedly issue a warning stating as much).
2) any bridge device that had one of these taps attached would *not*
take on the MAC of the tap (leading to network instability as
guests started and stopped)
A couple years later, https://bugzilla.redhat.com/798467 was filed,
complaining that a user could configure a tap-based guest interface to
have a MAC address that itself had a first byte of 0xFE, silently
(other than the kernel warning messages) resulting in a non-working
configuration. This was fixed by commit 5d571045, which logged an
error and failed the guest start / interface attach if the MAC's first
byte was 0xFE.
Although this restriction only reduces the potential pool of MAC
addresses from 2^46 (last two bits of byte 1 must be set to 10) by
2^32 (still 4 orders of magnitude larger than the entire IPv4 address
space), it also means that management software that autogenerates MAC
addresses must have special code to avoid an 0xFE prefix. Now after 7
years, someone has noticed this restriction and requested that we
remove it.
So instead of failing when 0xFE is found as the first byte, this patch
removes the restriction by just replacing the first byte in the tap
device MAC with 0xFA if the first byte in the guest interface is
0xFE. 0xFA is the next-highest value that still has 10 as the lowest
two bits, and still
2) meets the requirement of "tap MAC must be different from guest
interface MAC", and
3) is high enough that there should never be an issue of the attached
bridge device taking on the MAC of the tap.
The result is that *any* MAC can be chosen by management software
(although it would still not work correctly if a multicast MAC (lowest
bit of first byte set to 1) was chosen), but that's a different
issue).
Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com
QEMU version 2.12.1 introduced a performance feature under commit
be7773268d98 ("target-i386: add KVM_HINTS_DEDICATED performance hint")
This patch adds a new KVM feature 'hint-dedicated' to set this performance
hint for KVM guests. The feature is off by default.
To enable this hint and have libvirt add "-cpu host,kvm-hint-dedicated=on"
to the QEMU command line, the following XML code needs to be added to the
guest's domain description in conjunction with CPU mode='host-passthrough'.
<features>
<kvm>
<hint-dedicated state='on'/>
</kvm>
</features>
...
<cpu mode='host-passthrough ... />
Signed-off-by: Wim ten Have <wim.ten.have@oracle.com>
Signed-off-by: Menno Lageman <menno.lageman@oracle.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
The various distros have the following libxml2 vesions:
CentOS 7: 2.9.1
Debian Stretch: 2.9.4
FreeBSD Ports: 2.9.9
Ubuntu 16.04 LTS: 2.9.3
Based on this sampling, we can reasonably bump libxml2 min
version to 2.9.1
The 'query_raw' struct field was added in version 2.6.28,
so can be assumed to exist.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
QEMU shows a warning message if partial NUMA mapping is set. This patch
adds a warning message in libvirt when editing the XML. It must be an
error in future, when QEMU remove this ability.
Signed-off-by: Maxiwell S. Garcia <maxiwell@linux.ibm.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Historically URIs handled by the remote driver will always connect to
the libvirtd UNIX socket. There will now be one daemon per driver, and
each of these has its own UNIX sockets to connect to.
It will still be possible to run the traditional monolithic libvirtd
though, which will have the original UNIX socket path.
In addition there is a virproxyd daemon that doesn't run any drivers,
but provides proxying for clients accessing libvirt over IP sockets, or
tunnelling to the legacy libvirtd UNIX socket path.
Finally when running inside a daemon, the remote driver must not reject
connections unconditionally. For example, the QEMU driver needs to be
able to connect to the network driver. The remote driver must thus be
willing to handle connections even when inside the daemon, provided no
local driver is registered.
This refactoring enables the remote driver to be able to connect to the
per-driver daemons. The URI parameter "mode" accepts the values "auto",
"direct" and "legacy" to control which daemons are connected to.
The client side libvirt.conf config file also supports a "remote_mode"
setting which is used if the URI parameter is not set.
If neither the config file or URI parameter set a mode, then "auto"
is used, whereby the client looks to see which sockets actually exist
right now.
The remote driver will only ever spawn the per-driver daemons, or
the legacy libvirtd. It won't ever try to spawn virtproxyd, as
that is only there for IP based connectivity, or for access from
legacy remote clients.
If connecting to a remote host over any kind of ssh tunnel, for now we
must assume only the legacy socket exists. A future patch will introduce
a netcat replacement that is tailored for libvirt to make remote
tunnelling easier.
The configure arg '--with-remote-default-mode=legacy|direct' allows
packagers to set a default at build time. If not given, it will default
to legacy mode.
Eventually the default will switch to direct mode. Distros can choose
to do the switch earlier if desired. The main blocker is testing and
suitable SELinux/AppArmor policies.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The ssh, libssh, libssh2 & unix transports all need to use a UNIX socket
path, and duplicate some of the same logic for error checking. Pull this
out into a separate method to increase code sharing.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Instead of open-coding a string -> enum conversion, use the enum helpers
for the remote driver transport. The old code uses STRCASEEQ, so we must
force the URI transport to lowercase for sake of back-compatibility.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtproxyd daemon is merely responsible for forwarding RPC calls to
one of the other per-driver daemons. As such, it does not have any
drivers loaded and so regular auto-probing logic will not work. We need
it to be able to handle NULL URIs though, so must implement some kind of
alternative probing logic.
When running as root this is quite crude. If a per-driver daemon is
running, its UNIX socket will exist and we can assume it will accept
connections. If the per-driver daemon is not running, but socket
autostart is enabled, we again just assume it will accept connections.
The is not great, however, because a default install may well have
all sockets available for activation. IOW, the virtxend socket may
exist, despite the fact that the libxl driver will not actually work.
When running as non-root this is slightly easier as we only have two
drivers, QEMU and VirtualBox. These daemons will likely not be running
and socket activation won't be used either, as libvirt spawns the
daemon on demand. So we just check whether the daemon actually is
installed.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
When the client has a connection to one of the hypervisor specific
daemons (eg virtqemud), the app may still expect to use the secondary
driver APIs (storage, network, etc). None of these will be registered in
the hypervisor daemon, so we must explicitly open a connection to each
of the daemons for the secondary drivers we need.
We don't want to open these secondary driver connections at the same
time as the primary connection is opened though. That would mean that
establishing a connection to virtqemud would immediately trigger
activation of virtnetworkd, virnwfilterd, etc despite that that these
drivers may never be used by the app.
Thus we only open the secondary driver connections at time of first use
by an API call.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The driver dispatch methods access the priv->conn variables directly.
In future we want to dynamically open the connections for the secondary
driver. Thus we want the methods to call a method to get the connection
handle instead of assuming the private variable is non-NULL.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
If the event (un)registration methods are invoked while no connection is
open, they jump to a cleanup block which unlocks a mutex which is not
currently locked.
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The driver dispatch methods access the priv->conn variables directly.
In future we want to dynamically open the connections for the secondary
driver. Thus we want the methods to call a method to get the connection
handle instead of assuming the private variable is non-NULL.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The client parameter is always used to get access to the private data
struct.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The admin client now supports addressing the per-driver daemons using
the obvious URI schemes for each daemon. eg virtqemud:///system
virtqemud:///session, etc.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtvzd daemon will be responsible for providing the vz API
driver functionality. The vz driver is still loaded by the main
libvirtd daemon at this stage, so virtvzd must not be running at
the same time.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtbhyved daemon will be responsible for providing the bhyve API
driver functionality. The bhyve driver is still loaded by the main
libvirtd daemon at this stage, so virtbhyved must not be running at
the same time.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtvboxd daemon will be responsible for providing the vbox API
driver functionality. The vbox driver is still loaded by the main
libvirtd daemon at this stage, so virtvboxd must not be running at
the same time.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtlxcd daemon will be responsible for providing the lxc API
driver functionality. The lxc driver is still loaded by the main
libvirtd daemon at this stage, so virtlxcd must not be running at
the same time.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtqemud daemon will be responsible for providing the qemu API
driver functionality. The qemu driver is still loaded by the main
libvirtd daemon at this stage, so virtqemud must not be running at
the same time.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtxend daemon will be responsible for providing the libxl API
driver functionality. The libxl driver is still loaded by the main
libvirtd daemon at this stage, so virtxend must not be running at
the same time.
This naming is slightly different than other drivers. With the libxl
driver, the user still has a 'xen:///system' URI, and we provide it
in a libvirt-daemon-xen RPM, which pulls in a
libvirt-daemon-driver-libxl RPM.
Arguably we could rename the libxl driver to "xen" since it is the
only xen driver we have these days, and that matches how we expose it
to users in the URI naming.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtnwfilterd daemon will be responsible for providing the nwfilter API
driver functionality. The nwfilter driver is still loaded by the main
libvirtd daemon at this stage, so virtnwfilterd must not be running at
the same time.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtnodedevd daemon will be responsible for providing the nodedev API
driver functionality. The nodedev driver is still loaded by the main
libvirtd daemon at this stage, so virtnodedevd must not be running at
the same time.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtstoraged daemon will be responsible for providing the storage API
driver functionality. The storage driver is still loaded by the main
libvirtd daemon at this stage, so virtstoraged must not be running at
the same time.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtinterfaced daemon will be responsible for providing the interface API
driver functionality. The interface driver is still loaded by the main
libvirtd daemon at this stage, so virtinterfaced must not be running at
the same time.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtnetworkd daemon will be responsible for providing the network API
driver functionality. The network driver is still loaded by the main
libvirtd daemon at this stage, so virtnetworkd must not be running at
the same time.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virtsecretd daemon will be responsible for providing the secret API
driver functionality. The secret driver is still loaded by the main
libvirtd daemon at this stage, so virtsecretd must not be running at
the same time.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The libvirtd daemon provides the traditional libvirt experience where
all the drivers are in a single daemon, and is accessible over both
local UNIX sockets and remote IP sockets.
In the new world we're having a set of per-driver daemons which will
primarily be accessed locally via their own UNIX sockets.
We still, however, need to allow for case of applications which will
connect to libvirt remotely. These remote connections can be done as
TCP/TLS sockets, or by SSH tunnelling to the UNIX socket.
In the later case, the old libvirt.so clients will only know about
the path to the old libvirtd socket /var/run/libvirt/libvirt-sock,
and not the new driver sockets /var/run/libvirt/virtqemud-sock.
It is also not desirable to expose the main driver specific daemons
over IP directly to minimize their attack service.
Thus the virtproxyd daemon steps into place, to provide TCP/TLS sockets,
and back compat for the old libvirtd UNIX socket path(s). It will then
forward all RPC calls made to the appropriate driver specific daemon.
Essentially it is equivalent to the old libvirtd with absolutely no
drivers registered except for the remote driver (and other stateless
drivers in libvirt.so).
We could have modified libvirtd so none of the drivers are registed
to get the same end result. We could even add a libvirtd.conf parameter
to control whether the drivers are loaded to enable users to switch back
to the old world if we discover bugs in the split-daemon model. Using a
new daemon though has some advantages
- We can make virtproxyd and the virtXXXd per-driver daemons all
have "Conflicts: libvirtd.service" in their systemd unit files.
This will guarantee that libvirtd is never started at the same
time, as this would result in two daemons running the same driver.
Fortunately drivers use locking to protect themselves, but it is
better to avoid starting a daemon we know will conflict.
- It allows us to break CLI compat to remove the --listen parameter.
Both listen_tcp and listen_tls parameters in /etc/libvirtd/virtd.conf
will default to zero. Either TLS or TCP can be enabled exclusively
though virtd.conf without requiring the extra step of adding --listen.
- It allows us to set a strict SELinux policy over virtproxyd. For
back compat the libvirtd policy must continue to allow all drivers
to run. We can't easily give a second policy to libvirtd which
locks it down. By introducing a new virtproxyd we can set a strict
policy for that daemon only.
- It gets rid of the weird naming of having a daemon with "lib" in
its name. Now all normal daemons libvirt ships will have "virt"
as their prefix not "libvirt".
- Distros can more easily choose their upgrade path. They can
ship both sets of daemons in their packages, and choose to
either enable libvirtd, or enable the per-driver daemons and
virtproxyd out of the box. Users can easily override this if
desired by just tweaking which systemd units are active.
After some time we can deprecate use of libvirtd and after some more
time delete it entirely, leaving us in a pretty world filled with
prancing unicorns.
The main downside with introducing a new daemon, and with the
per-driver daemons in general, is figuring out the correct upgrade
path.
The conservative option is to leave libvirtd running if it was
an existing installation. Only use the new daemons & virtproxyd
on completely new installs.
The aggressive option is to disable libvirtd if already running
and activate all the new daemons.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
When running in libvirtd, we are happy for any of the drivers to simply
skip their initialization in virStateInitialize, as other drivers are
still potentially useful.
When running in per-driver daemons though, we want the daemon to abort
startup if the driver cannot initialize itself, as the daemon will be
useless without it.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The make logic assumes that the SYSTEMD_UNIT_FILES var can be built from
SYSTEMD_UNIT_FILES_IN by simply dropping the directory prefix and the
.in suffix.
This won't work in future when a single .in unit file can be used to
generate multiple different units.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Prepare for reusing libvirtd socket unit files with other daemons by
making various parts of their config conditionally defined by the make
rules.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The make rules for the systemd socket unit files are all essentially
identical and can be collapsed into a single generic rule. The service
unit file rule can be simplified too.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Substitute in the @sysconfigdir@ value instead of /etc.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The same make variables will be useful for building both libvirtd and
the split daemons, so refactor & rename variables to facilitate reuse.
Automake gets annoyed if you define a variable ending LDFLAGS:
src/remote/Makefile.inc.am:53: warning: variable 'REMOTE_DAEMON_LDFLAGS' is defined but no program or
src/remote/Makefile.inc.am:53: library has 'REMOTE_DAEMON' as canonical name (possible typo)
So we trick it by using an LD_FLAGS or LD_ADD suffix instead.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
GNULIB sets $(LIBSOCKET) on mingw to pull in the windows socket
APIs. This is trivially not required, since we don't build libvirtd
on mingw.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Prepare for reusing libvirtd augeas defintions with other daemons by
making the config parameters for IP sockets conditionally defined by
the make rules.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Prepare for reusing libvirtd config to create other daemons by making
the config parameters for IP sockets conditionally defined by the make
rules.
The main libvirtd daemon will retain IP listen ability, but all the
driver specific daemons will be local UNIX sockets only. Apps needing
IP connectivity will connect via the libvirtd daemon which will proxy
to the driver specfic daemon.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Using @VARNAME@ is a normal style of automake, so lets match that.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Prepare for reusing libvirtd source to create other daemons by making
the use of IP sockets conditionally defined by the make rules.
The main libvirtd daemon will retain IP listen ability, but all the
driver specific daemons will be local UNIX sockets only. Apps needing
IP connectivity will connect via the libvirtd daemon which will proxy
to the driver specfic daemon.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Prepare for reusing libvirtd source to create other daemons by making
the driver(s) to load conditionally defined by the make rules.
If nothing is set, all drivers will be loaded, ignoring any missing ones
as historically done.
If MODULE_NAME is set only one driver will be loaded and that one must
succeed.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Prepare for reusing libvirtd source to create other daemons by making
the daemon name conditionally defined by the make rules.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The remote daemon tries to print out its help text in a couple of giant
blocks of text. This has already lead to duplication of the text for the
privileged vs unprivileged execution mode. With the introduction of more
daemons, this text is going to be duplicated many more times with small
variations. This is very unfriendly to translators as they have to
translate approximately the same text many times with small tweaks.
Splitting the text up into individual strings to print means that each
piece will only need translating once. It also gets rid of all the
layout information from the translated strings, so avoids the problem of
translators breaking formatting by mistake.
Reviewed-by: Christophe de Dinechin <dinechin@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Instead of adding generated config files to CLEANFILES and BUILT_SOURCES
in each makefile, add them all at once.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Instead of each subdir containing its own custom rule for checking the
augeas tests, use common rule for all.
The new rule searches both src + build dirs for include files, since
some augeas files will be auto-generated very shortly.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The current make rules are inconsistent about which directory the
augeas test files are created in. Put them all in the same dir as
their source.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
We already have a variable that lists all augeas test files, so we can
add everything to CLEANFILES at once.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The augeas-gentest.pl program merges a config file into a augeas
file, saving the output to a new file. It is going to be useful
to further process the output file, and it would be easier if this can
be done with a pipeline, so change augeas-gentest.pl to write to stdout
instead of a file.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Each public API is required to log all arguments it was called
with. Except, there are some missing. Fix them.
Signed-off-by: Han Han <hhan@redhat.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Redefining a domain via virDomainDefineXML should not give different results
based on an already existing definition.
Also, there's a crasher somewhere in the code:
https://bugzilla.redhat.com/show_bug.cgi?id=1739338
This reverts commit 94b3aa55f8
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
Since its introduction in commit
8737578d11, the TPM version format is
"2.0" and not "2".
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDeviceDefPostParse callback requires qemuCaps, we need
to make sure it gets the capabilities stored in the domain's private
data if the domain is running. Passing NULL may cause QEMU capabilities
probing to be triggered in case QEMU binary changed in the meantime.
When this happens while a running domain object is locked, QMP event
delivered to the domain before QEMU capabilities probing finishes will
deadlock the event loop.
QEMU capabilities lookup (via domainPostParseDataAlloc callback) is
hidden inside virDomainDeviceDefPostParseOne with no way to pass
qemuCaps to virDomainDeviceDef* functions. This patch fixes all
remaining paths leading to virDomainDeviceDefPostParse.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
Several general snapshot and checkpoint APIs were lazily passing NULL as
the parseOpaque pointer instead of letting their callers pass the right
data. This patch fixes all paths leading to virDomainDefParseNode.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
This patch fixes all paths leading to virDomainDefPostParse.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
Several general functions from domain_conf.c were lazily passing NULL as
the parseOpaque pointer instead of letting their callers pass the right
data. This patch fixes all paths leading to virDomainDefCopy to do the
right thing.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
This patch fixes all paths leading to qemuMigrationCookieXMLParse.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
This patch fixes all paths leading to virDomainDefParseString.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
This patch fixes all paths leading to qemuMigrationAnyPrepareDef.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
This patch fixes all paths leading to qemuDomainSaveImageOpen.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
This patch fixes all paths leading to qemuDomainDefFormatBufInternal.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Since qemuDomainDefPostParse callback requires qemuCaps, we need to make
sure it gets the capabilities stored in the domain's private data if the
domain is running. Passing NULL may cause QEMU capabilities probing to
be triggered in case QEMU binary changed in the meantime. When this
happens while a running domain object is locked, QMP event delivered to
the domain before QEMU capabilities probing finishes will deadlock the
event loop.
This patch fixes all paths leading to qemuDomainDefCopy.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Use the correct enum constant when validating vlan usage.
This fixes a merge error in
commit 6cb0ec48bd
Author: Daniel P. Berrangé <berrange@redhat.com>
Date: Mon Sep 3 17:34:22 2018 +0100
network: convert networkAllocateActualDevice to virNetworkPortDef
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Requires adjustments to use verify_expr() which replaces
verify_true(), and to disable the new syntax check
'sc_prohibit_gnu_make_extensions' since we require GNU make.
Signed-off-by: Eric Blake <eblake@redhat.com>
Commit fed58d83 was a hack to fix a mingw build failure due to header
inclusion order resulting in a clash over the use of DATADIR,
repeating a trick made several other times in the past. Better is to
revert that, and instead use pragmas to avoid the clash in the first
place, regardless of header ordering, solving it for everyone.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Now that the code does not refer to any libvirt headers,
except internal.h macros, it does not need to link to
any libvirt code, nor gnulib either. The only thing it
needs is yajl.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Now that 100% of libvirt code is forbidden in a SUID environment,
we no longer need to worry about whether env variables are
trustworthy or not. The virt-login-shell setuid program, which
does not link to any libvirt code, will purge all environment
variables, except $TERM, before invoking the virt-login-shell-helper
program which uses libvirt.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Now that 100% of libvirt code is forbidden in a SUID environment,
we no longer need to worry about whether env variables are
trustworthy or not. The virt-login-shell setuid program, which
does not link to any libvirt code, will purge all environment
variables, except $TERM, before invoking the virt-login-shell-helper
program which uses libvirt.
Thus we only need one API for env passthrough in virCommand.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Now that none of the libvirt.so code will ever run in a setuid
context, we can remove the virIsSUID() method. The global
initializer function can just inline the check itself. The new
inlined check is slightly stronger as it also looks for a
setgid situation.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The virt-login-shell setuid program is now a tiny piece of code
that only uses standard libc functions, and santizes the execution
environment before invoking the real virt-login-shell-helper.
The latter is thus able to use the normal libvirt.so build,
allowing us to delete the special cut down setuid library build.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
If virQEMUDriverGetCapabilities returns NULL, then a subsequent
deref of @caps would cause an error, so we just return failure.
Found by Coverity
Signed-off-by: John Ferlan <jferlan@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
The previous bump to 4.4 was done in:
commit 24241c236e
Author: Daniel P. Berrange <berrange@redhat.com>
Date: Wed Jul 5 10:35:32 2017 +0100
Require use of GCC 4.4 or CLang compilers
with 4.4 picked due to RHEL-6. Since we dropped RHEL-6, the
next oldest distro is RHEL-7 (4.8.5), and thus we pick 4.8
as the new min.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Until now, testDomainGetTime would always return the same fixed values
everytime it was called. By using domain-private data we can make this
API return the values previously set with testDomainSetTime, or use the
same old fixed values in case testDomainSetTime hasn't been called at all.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
The qemu and vz implementations don't emit any signals when this API is
called, so we can do the same here for now and succeed by doing nothing.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
The qemu driver already does some <rng> model validation, based on
qemuCaps. However, the logic for exposing <rng> model values in domcaps
is basically identical. This drops the qemuCaps checking and compares
against the domCaps data directly.
This approach makes it basically impossible to add a new <rng> model to
the qemu driver without extending domcaps. The validation can also
be shared with other drivers eventually.
Reviewed-by: Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
Fill in virDomainCaps at Validate time and use it to call
virDomainCapsDeviceDefValidate
Reviewed-by: Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
This is an entrypoint to validate a virDomainDeviceDef against
values filled into virDomainCaps.
Currently it's just a stub
Reviewed-by: Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
qemuCaps is tied to a binary on disk. domCaps is tied to a combo
of binary+machine+arch+virttype values. For the qemu driver this almost
entirely translates to a permutation of qemuCaps though
Upcoming patches want to use the domCaps data store at XML validate
time, but we need to cache the data so we aren't repeatedly
regenerating it.
Add a domCapsCache hash table to qemuCaps. This ensures that the domCaps
cache is blown away whenever qemuCaps needs to be regenerated. Similarly
when qemuCaps is invalidated, the next call to virQEMUCapsCacheLookup
will unref qemuCaps and free our cache as well.
Adjust virQEMUDriverGetDomainCapabilities to search the cache and add
to it if we don't find a hit.
Signed-off-by: Cole Robinson <crobinso@redhat.com>
Reviewed-by: Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
For now it's just a helper for building a qemu virDomainCapsPtr,
used in qemuConnectGetDomainCapabilities
Reviewed-by: Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
The model logic is taken from qemuDomainRNGDefValidate
Reviewed-by: Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
This way it is obvious when adding a new resource control type
that stats helper func needs to be updated too.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
If the host doesn't have resctrl then the monitor is going to be
NULL and we must avoid dereferencing it.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The capabilities object must be unrefed when no longer needed.
Use VIR_AUTOUNREF() for that.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Current code doesn't allow us to add sub-features as we always print the
closing '/>'. As a preparatory change to implementing 'direct' sub-feature
for 'stimer' feature switch to printing closing tag individually.
No functional change.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Some Hyper-V features (like the upcoming Direct Synthetic timers) are
announced by feature bits in Edx but KVM_FEATURE_DEF() supports only Eax.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
I find this function more readable if checks for passed storage
source are done first and backing chain is done last. Mixing them
together does not hurt, but is less readable.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
ACKed-by: Peter Krempa <pkrempa@redhat.com>
This function does not change any of the passed addresses. It
just reads them.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
ACKed-by: Peter Krempa <pkrempa@redhat.com>
This function does not change any of the passed addresses. It
just reads them.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
ACKed-by: Peter Krempa <pkrempa@redhat.com>
On success update the domain-private data. Consider / and /boot to be
the only mountpoints avaiable in order to be consistent with the other
FS-related calls.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
vm-specific data can be used by APIs that need to preserve some state
between calls
Some of them are:
- FS-related APIs for remembering which mountpoints are frozen
- virDomainSetTime / virDomainGetTime for maintaining time information
- virDomainSetIOThreadParams for storing the I/O thread parameters
- virDomainManagedSaveDefineXML for internally storing the VM definition
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
The main value here is the current balloon value which taken from the
config. All the other values (except for period) are derived by 2^n
division so that compiler prefers bitwise operations. Period value was
kept fixed in order to produce predictable results in a test environment.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
This should just forward the call to testDomainCreateXML since we
can't do anything with the provided file descriptors in the test driver.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
This should just forward the call to testDomainCreateWithFlags since we
can't do anything with the provided file descriptors in the test driver.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
The xencommons service provides all the essential services such as
xenstored, xenconsoled, etc. needed by the libvirt Xen driver, so
libvirtd should be started after xencommons.
The xendomains service uses Xen's xl tool to operate on any domains it
finds running, even those managed by libvirt. Add a conflicts on the
xendomains service to ensure it is not enabled when libvirtd is enabled.
Signed-off-by: Jim Fehlig <jfehlig@suse.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
This code that executes virPCIDeviceReattach in all
virPCIDevicePtr objects of a given virPCIDeviceListPtr
list is replicated twice in the code. Putting it in a helper
function helps with readability.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
virHostdevReattachPCIDevice() is a static that simply does
a wait loop with virPCIDeviceWaitForCleanup() before
calling virPCIDeviceReattach().
This loop traces back to commit d1e5676c0d, aiming to
solve a race condition between Libvirt returning the
device back to the host and QEMU trying to access it in
the meantime, which resulted in QEMU exiting on error
and killing the guest. This happens because device_del
is asynchronous, returning OK even if the guest didn't
release the device. Commit 01abc8a1b8 moved this code
to qemu_hostdev.c, 82e8dd4cf8 added the pci-stub conditional
for the loop, 899b261127 moved the code to virhostdev.c
where it stood until now.
The intent of this wait loop is still valid: device_del
is still not bullet proof into preventing the conditions
that commit d1e5676c0d aimed to fix, especially when considering
all the architectures we must support. However, this loop
is executed only in virHostdevReattachPCIDevice(), leaving
every other virPCIDeviceReattach() call prone to that error.
Let's move the wait loop code to virPCIDeviceReattach(). This
will:
- make every reattach call safe from this race condition
with the pci-stub;
- allow for a bit of code cleanup (virHostdevReattachPCIDevice()
can be erased, and virHostdevReAttachPCIDevices() can use
virPCIDeviceReattach() directly);
- make it easier to understand the overall reattach mechanisms in
Libvirt, without the risk of a newcomer wondering why reattach
is done slightly different in some instances.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
This code that executes virPCIDeviceReset in all virPCIDevicePtr
objects of a given virPCIDeviceListPtr list is replicated twice
in the code. Putting it in a helper function helps with
readability.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
There is no restriction on maximum value of PCI domain. In fact,
Linux kernel uses plain atomic inc when assigning PCI domains:
drivers/pci/pci.c:static int pci_get_new_domain_nr(void)
drivers/pci/pci.c-{
drivers/pci/pci.c- return atomic_inc_return(&__domain_nr);
drivers/pci/pci.c-}
Of course, this function is called only if kernel was compiled
without PCI domain support or ACPI did not provide PCI domain.
However, QEMU still has the same restriction as us: in
set_pci_host_devaddr() QEMU checks if domain isn't greater than
0xffff. But one can argue that that's a QEMU limitation. We still
want to be able to cope with other hypervisors that don't have
this limitation (possibly).
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Currently, the way we format PCI address is using printf-s
precision, e.g. "%.4x". This works if we don't want to print any
value outside of bounds (which is usually the case). However,
turns out, PCI domain can be 0x10000 which doesn't work well with
our format strings. However, if we change the format string to
"%04x" then we still pad small values with zeroes but also we are
able to print values that are larger than four digits. In fact,
this format string used by kernel to print a PCI address:
"%04x:%02x:%02x.%d"
The other three format strings (for bus, device and function) are
changed too, so that we use the same format string as kernel.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The format string for a PCI address is copied over and over
again, often with slight adjustments. Introduce global
VIR_PCI_DEVICE_ADDRESS_FMT macro that holds the formatting string
and use it wherever possible.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
In near future, the length restriction of PCI domain is going to
be lifted. This means that our assumption that PCI address is 13
bytes long is no longer true. We can avoid this problem by making
@name dynamically allocated and thus not bother with actual
length of stringified PCI address.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
This function declares @ret variable and then uses
VIR_STEAL_PTR() to avoid freeing temporary variable @dev which is
constructed. Well, as of 267f1e6da5 we have VIR_RETURN_PTR()
macro so that we can avoid this pattern.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
While it's true that older QEMUs were not able to deal with PCI
domains, we don't support those versions anymore (see
4a42ece13a). Therefore it is safe to always format fully
expanded PCI address. Format PCI domain always as it will
simplify next commits.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
A new algorithm for detecting the vcpus and monitor type conflicts
between new monitor an existing allocation and monitor groups.
After refactoring, since we are verifying both @vcpus and monitor
type @tag at the same time, the validating function name has been
renamed from virDomainResctrlMonValidateVcpus to
virDomainResctrlValidateMonitor.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Export virResctrlMonitorGetStats and make
virResctrlMonitorGetCacheOccupancy obsoleted.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Refactor 'virResctrlMonitorStats' to track multiple statistical
records.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Refactor and rename 'virResctrlMonitorFreeStats' to
'virResctrlMonitorStatsFree' to free one
'virResctrlMonitorStatsPtr' object.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
'virResctrlAllocIsEmpty' checks if cache allocation or memory
bandwidth allocation settings are specified in configuration
file. It is not proper to be used in checking memory bandwidth
allocation is specified in XML settings because this function
could not distinguish memory bandwidth allocations from cache
allocations.
Here using the local variable @n, which indicates the cache
allocation groups or memory bandwidth groups depending on the
context it is in, to decide if append a new @resctrl object.
If @n is zero and no monitors groups specified in XML, then
we should not append a new @resctrl object to @def->resctrls.
This kind of replacement is also more efficient and avoiding
a long function calling path.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Let 'virDomainResctrlVcpuMatch' to retrieve a pointer of
virDomainResctrlDefPtr in its third parameter instead
of virResctrlAllocPtr, if @vcpus is matched with the vcpus
of some resctrl allocation in list of @def->resctrls.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Creating object and judging if it is successfully created in fewer
lines.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
code cleanup for 'virDomainCachetuneDefParse' and
'virDomainMemorytuneDefParse'.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
'default monitor of an allocation' is defined as the resctrl
monitor group that created along with an resctrl allocation,
which is created by resctrl file system. If the monitor group
specified in domain configuration file is happened to be a
default monitor group of an allocation, then it is not necessary
to create monitor group since it is already created. But if
an monitor group is not an allocation default group, you
should create the group under folder
'/sys/fs/resctrl/mon_groups' and fill the vcpu PIDs to 'tasks'
file.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Fortunately, the code that handles metadata getting or setting is
driver agnostic, so all that is needed from individual hypervisor
drivers is to call the right functions.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1732306
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
I messed up formatting during conflict resolution across rebasing
while preparing my checkpoint patches :)
Signed-off-by: Eric Blake <eblake@redhat.com>
hv-spinlocks is not a CPUID feature and should not be checked as such.
While starting a domain with hv-spinlocks enabled, we would report a
warning about unsupported hyperv spinlocks feature even though it was
set properly.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
CI flagged a failing mingw build, due to:
In file included from ../../src/conf/checkpoint_conf.c:24:
../gnulib/lib/configmake.h:8:17: error: expected identifier or '(' before string constant
8 | #define DATADIR "/usr/i686-w64-mingw32/sys-root/mingw/share"
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
As previously learned in commits bd205a90 and 976abdf6, gnulib's
configmake.h header does #define DATADIR "string...", while mingw's
<winsock2.h> expects to declare a type named DATADIR. As long as the
mingw system header is included first before configmake.h, the two
uses do not conflict, but until gnulib is patched to make configmake.h
automatically work around the issue, our immediate fix is the
workaround of rearranging our include order to insure no conflict.
Copy the paradigm used in domain_conf.c of using <unistd.h> to trigger
the indirect inclusion of <winsock2.h> on mingw.
Fixes: 1a4df34a
Signed-off-by: Eric Blake <eblake@redhat.com>
Turning a NULL URI instead the empty string is very misleading when
reading the debug logs as the distinction between the two is
functionally important.
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Originally the names of the KVM CPU features were only used internally
for looking up their CPUID bits. So we used "__kvm_" prefix for them to
make sure the names do not collide with normal CPU features stored in
our CPU map.
But with QEMU 4.1 we check which features were enabled or disabled by a
freshly started QEMU process using their names rather than their CPUID
bits (mostly because of MSR features). Thus we need to change our made
up internal names into the actual names used by QEMU.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Tested-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Most of the internally defined KVM CPUID features are not actually used
by libvirt. The QEMU driver may enable or disable them on the command
line, but we don't check for the associated CPU properties or CPUID
bits. They would be useless with QEMU 4.1 anyway since their names were
only remotely similar to the actual feature names.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Tested-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
All the features are hyperv features even though they are provided by
KVM with QEMU. The "KVM" part in the macro names does not make a lot of
sense.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Tested-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Starting with QEMU 4.1, we're using the canonical feature names on the
command line and avoid aliases to prepare for possible deprecation of
all aliases in QEMU. But we do so only for features from our CPU map,
hyperv features defined in the code were unchanged and this patch fixes
it. Some features use "hv-" prefix unconditionally because they were
introduced recently enough to always support spelling with a dash.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Tested-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Originally the names of the hyperv CPU features were only used
internally for looking up their CPUID bits. So we used "__kvm_hv_"
prefix for them to make sure the names do not collide with normal CPU
features stored in our CPU map.
But with QEMU 4.1 we check which features were enabled or disabled by a
freshly started QEMU process using their names rather than their CPUID
bits (mostly because of MSR features). Thus we need to change our made
up internal names into the actual names used by QEMU. Most of the names
are only used with QEMU 4.1 and newer and the reset was introduced with
QEMU recently enough to already support spelling with "-". Thus we don't
need to define them as "hv_*" with a translation to "hv-*" for new QEMU.
Without this patch libvirt would mistakenly report all hyperv features
as unavailable and refuse to start any domain using them with QEMU 4.1.
Reported-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Tested-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Earlier patches mentioned that the initial implementation will prevent
snapshots and checkpoints from being used on the same domain at once.
However, the actual restriction is done in this separate patch to make
it easier to lift that restriction via a revert, when we are finally
ready to tackle that integration in the future.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Time to actually issue the QMP transactions that create and delete
persistent checkpoints, resolving TODOs intentionally left earlier in
the series. For create, we only need one transaction: inside, we
visit all disks affected by the checkpoint, and create a new enabled
bitmap, as well as disabling the bitmap of the first ancestor
checkpoint (if any) that also had a bitmap. For deletion, we need
multiple QMP calls: for each disk, if there is an ancestor checkpoint
with a bitmap, then the bitmap must be merged (including activating
the ancestor bitmap if the leaf node changes), all before deleting the
bitmap from the checkpoint being removed.
Signed-off-by: Eric Blake <eblake@redhat.com>
Qemu bitmap operations require knowing the node name associated with
the format layer (the qcow2 file); as upcoming patches will be
grabbing that information frequently, make a helper function to access
it.
Another potential benefit of this function is that we have a single
place where we could insert a QMP node-name scraping call if we don't
currently know the node name, when -blockdev is not supported;
however, the goal is that we hopefully don't ever have to do that
because we instead scrape node names only at the point where they
change.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
A lot of this work heavily copies from the existing snapshot APIs.
What's more, this patch is (intentionally) very similar to the
checkpoint code just added in the test driver, to the point that qemu
checkpoints are not fully usable in this patch, but it at least
bisects and builds cleanly. The separation between patches is done
because the grunt work of saving and restoring XML and tracking
relations between checkpoints is common to the test driver, while the
later patch adding integration with QMP is specific to qemu.
Also note that the interlocking to prevent checkpoints and snapshots
from existing at the same time will be a separate patch, to make it
easier to revert that restriction when we finally round out the design
for supporting interaction between the two concepts.
Signed-off-by: Eric Blake <eblake@redhat.com>
This is similar to the existing directory for snapshots; the domain
will save one xml file per checkpoint, for reloading on the next
libvirtd restart. Fortunately, since checkpoints mandate RNG
validation, we are assured that the checkpoint name will be usable as
a file name (no abuse of '../escape' as a checkpoint name, for
example).
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Introduce the handler for finalizing a block commit and active bloc
commit job which will allow to use it with blockdev.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Introduce the handler for finalizing a block pull job which will allow
to use it with blockdev.
This patch also contains some additional machinery which is required to
store all the relevant job data in the status XML which will also be
reused with other block job types.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
In case of an incoming migration we do not need to run swtpm_setup
with all the parameters but only want to get the benefit of it
creating a TPM state file for us that we can then label with an
SELinux label. The actual state will be overwritten by the in-
coming state. So we have to pass an indicator for incomingMigration
all the way to the command line parameter generation for swtpm_setup.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
A lot of this work heavily copies from the existing snapshot APIs.
The test driver doesn't really have to do anything more than just
expose an interface into libvirt metadata, making it possible to test
saving and restoring XML, and tracking relations between multiple
checkpoints.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
The remote code generator had to be taught about the new
virDomainCheckpointPtr type, at which point the remote driver code for
checkpoints can be generated.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Creating a checkpoint does not modify guest-visible state,
but does modify host resources. Rather than reuse existing
domain:write, domain:block_write, or domain:snapshot access
controls, it seems better to introduce a new access control
specific to tasks related to checkpoints and incremental
backups of guest disk state.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Wire up the use of a checkpoint list into each domain, similar to the
existing snapshot list. This includes adding a function for checking
that a redefine operation fits in with the existing list, as well as
various filtering capabilities over the list contents.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Create a new file for managing a list of checkpoint objects, borrowing
heavily from existing virDomainSnapshotObjList paradigms.
Note that while snapshots definitely have a use case for multiple
children to a single parent (create a base snapshot, create a child
snapshot, revert to the base, then create another child snapshot),
it's harder to predict how checkpoints will play out with reverting to
prior points in time. Thus, in initial use, given a list of
checkpoints, you never have more than one child, and we can treat the
most-recent leaf node as the parent of the next node creation, without
having to expose a notion of a current node in XML or public API.
However, as the snapshot machinery is already generic, it is easier to
reuse the generic machinery that tracks relations between domain
moments than it is to open-code a new list-management scheme just for
checkpoints (hence, we still have internal functions related to a
current checkpoint, even though that has no observable effect
externally, as well as the addition of a function to easily find the
lone leaf in the list to use as the current checkpoint).
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Add a new file checkpoint_conf.c that performs the translation to and
from new XML describing a checkpoint. The code shares a common base
class with snapshots, since a checkpoint similarly represents the
domain state at a moment in time. Add some basic testing of round trip
XML handling through the new code.
Of note - this code intentionally differs from snapshots in that XML
schema validation is unconditional, rather than based on a public API
flag. We have many existing interfaces that still need to add a flag
for opt-in schema validation, but those interfaces have existing
clients that may not have been producing strictly-compliant XML, or we
may still uncover bugs where our RNG grammar is inconsistent with our
code (where omitting the opt-in flag allows existing apps to keep
working while waiting for an RNG patch). But since checkpoints are
brand-new, it's easier to ensure the code matches the schema by always
using the schema. If needed, a later patch could extend the API and
add a flag to turn on to request schema validation, rather than having
it forced (possibly just the validation of the <domain> sub-element
during REDEFINE) - but if a user encounters XML that looks like it
should be good but fails to validate with our RNG schema, they would
either have to upgrade to a new libvirt that adds the new flag, or
upgrade to a new libvirt that fixes the RNG schema, which implies
adding such a flag won't help much.
Also, the redefine flag requires the <domain> sub-element to be
present, rather than catering to historical back-compat to older
versions.
Signed-off-by: Eric Blake <eblake@redhat.com>
Introduce a bunch of new public APIs related to backup checkpoints.
Checkpoints are modeled heavily after virDomainSnapshotPtr (both
represent a point in time of the guest), although a snapshot exists
with the intent of rolling back to that state, while a checkpoint
exists to make it possible to create an incremental backup at a later
time. We may have a future hypervisor that can completely manage
checkpoints without libvirt metadata, but the first two planned
hypervisors (qemu and test) both always use libvirt for tracking
metadata relations between checkpoints, so for now, I've deferred
the counterpart of virDomainSnapshotHasMetadata for a separate
API addition at a later date if there is ever a need for it.
Note that until we allow snapshots and checkpoints to exist
simultaneously on the same domain (although the actual prevention of
this will be in a separate patch for the sake of an easier revert down
the road), that it is not possible to branch out to create more than
one checkpoint child to a given parent, although it may become
possible later when we revert to a snapshot that coincides with a
checkpoint. This also means that for now, the decision of which
checkpoint becomes the parent of a newly created one is the only
checkpoint with no child (so while there are APIs for dealing with a
current snapshot, we do not need those for checkpoints). We may end
up exposing a notion of a current checkpoint later, but it's easier to
add stuff when proven needed than to blindly support it now and wish
we hadn't exposed it.
The following map shows the API relations to snapshots, with new APIs
on the right:
Operate on a domain object to create/redefine a child:
virDomainSnapshotCreateXML virDomainCheckpointCreateXML
Operate on a child object for lifetime management:
virDomainSnapshotDelete virDomainCheckpointDelete
virDomainSnapshotFree virDomainCheckpointFree
virDomainSnapshotRef virDomainCheckpointRef
Operate on a child object to learn more about it:
virDomainSnapshotGetXMLDesc virDomainCheckpointGetXMLDesc
virDomainSnapshotGetConnect virDomainCheckpointGetConnect
virDomainSnapshotGetDomain virDomainCheckpointGetDomain
virDomainSnapshotGetName virDomainCheckpiontGetName
virDomainSnapshotGetParent virDomainCheckpiontGetParent
virDomainSnapshotHasMetadata (deferred for later)
virDomainSnapshotIsCurrent (no counterpart, see note above)
Operate on a domain object to list all children:
virDomainSnapshotNum (no counterparts, these are the old
virDomainSnapshotListNames racy interfaces)
virDomainSnapshotListAllSnapshots virDomainListAllCheckpoints
Operate on a child object to list descendents:
virDomainSnapshotNumChildren (no counterparts, these are the old
virDomainSnapshotListChildrenNames racy interfaces)
virDomainSnapshotListAllChildren virDomainCheckpointListAllChildren
Operate on a domain to locate a particular child:
virDomainSnapshotLookupByName virDomainCheckpointLookupByName
virDomainSnapshotCurrent (no counterpart, see note above)
virDomainHasCurrentSnapshot (no counterpart, old racy interface)
Operate on a snapshot to roll back to earlier state:
virDomainSnapshotRevert (no counterpart, instead checkpoints
are used in incremental backups via
XML to virDomainBackupBegin)
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
If we are using -blockdev, then node names are always available
(because we set them). But when not using it, we have to scrape node
names from QMP, and want to do so as infrequently as possible. We
were scraping node names after reconnecting a new libvirtd to an
existing guest (see qemuProcessReconnect), and after any block job
that may have changed the set of node names we care about (legacy
block jobs), but forgot to scrape the names when first starting a
guest. Do so now in order to allow the checkpoint code to always have
access to a node name without having to repeat a node name scrape
itself.
Future patches may need to clean up qemuDomainSetBlockThreshold (if
node names are always available, then it doesn't need to repeat a
scrape) and/or hotplug and media changes (if the addition of new nodes
can result in a null node name, then scraping at that point in time
would be appropriate). But for now, this patch addresses only the
most common instance of a missing node name.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Since we are checking the 2nd parameter in the function for NULL,
we need to remove ATTRIBUTE_NONNULL(2) from the prototype.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Message-Id: <20190726205633.2041912-5-stefanb@linux.vnet.ibm.com>
Remove the ATTRIBUTE_NONNULL(1) from virCommandSetSendBuffer()
prototype since we are checking for '!cmd' and move the initialization
if 'i' after the test for '!cmd'.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Message-Id: <20190726205633.2041912-4-stefanb@linux.vnet.ibm.com>
Use the existing variables rather then calling virTPMSwtpmXYZ().
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Message-Id: <20190726205633.2041912-2-stefanb@linux.vnet.ibm.com>
Create an empty log file if the log file was removed, otherwise the
transaction to set the security labels on the file will fail.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20190726210706.24440-3-stefanb@linux.ibm.com>
Set the transactionStarted to false if the commit failed. If this is not
done, then the failure path will report 'no transaction is set' and hide
more useful error reports.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20190726210706.24440-2-stefanb@linux.ibm.com>
Starting with QEMU 4.1 qemuMonitorCPUModelInfo structure in virQEMUCaps
stores only canonical feature names which may differ from the name used
by libvirt. We need translate these canonical names into libvirt names
for further consumption.
This fixes a bug in qemuConnectBaselineHypervisorCPU which would remove
all features for which libvirt's spelling differs from the QEMU's
preferred name. For example, the following result of
qemuConnectBaselineHypervisorCPU on my host with QEMU 4.1 is wrong:
<cpu mode='custom' match='exact'>
<model fallback='forbid'>Skylake-Client</model>
<vendor>Intel</vendor>
<feature policy='require' name='ss'/>
<feature policy='require' name='vmx'/>
<feature policy='require' name='hypervisor'/>
<feature policy='require' name='clflushopt'/>
<feature policy='require' name='umip'/>
<feature policy='require' name='arch-capabilities'/>
<feature policy='require' name='xsaves'/>
<feature policy='require' name='pdpe1gb'/>
<feature policy='require' name='invtsc'/>
<feature policy='disable' name='pclmuldq'/>
<feature policy='disable' name='lahf_lm'/>
</cpu>
The 'pclmuldq' and 'lahf_lm' should not be disabled in the baseline CPU
as they are supported by QEMU on this host.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Since swtpm does not support getting started without password
once it was created with encryption enabled, we don't allow
encryption to be removed. Similarly, we do not allow encryption
to be added once swtpm has run. We also prevent chaning the type
of the TPM backend since the encrypted state is still around and
the next time one was to switch back to the emulator backend
and forgot the encryption the TPM would not work.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
This patch now passes the passphrase as a migration key to swtpm.
This now encrypts the state of the TPM while a VM is migrated between
hosts or when suspended into a file. Since the migration key secret
is the same as the state encryption secret, this now requires that
the migration destination host has the same secret value.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Allow vTPM state encryption when swtpm_setup and swtpm support
passing a passphrase using a file descriptor.
This patch enables the encryption of the vTPM state only. It does
not encrypt the state during migration, so the destination secret
does not need to have the same password at this point.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Extend virCommandProcessIO to include the send buffers in the poll
loop.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Mark a virCommand's inpipe (write-end of pipe) as non-blocking so that it
will never block when we were to try to write too many bytes to it while
it doesn't have the capacity to hold them.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Convert the struct pollfd *fds to be allocated rather than residing
on the stack. This prepares it for the next patch where the size of
the array of fds becomes dynamic.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Implement virCommandSetSendBuffer() that allows the caller to pass a
file descriptor and buffer to virCommand. virCommand will write the
buffer into the file descriptor. That file descriptor could be the
write end of a pipe or one of the file descriptors of a socketpair.
The other file descriptor should be passed to the launched process to
read the data from.
Only implement the function to allocate memory for send buffers
and to free them later on.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Run 'swtpm socket --print-capabilities' and
'swtpm_setup --print-capabilities' to get the JSON object of the
features the programs are supporting and parse them into a bitmap.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Check whether previously found executables were updated and if
so look for them again. This helps to use updated features of
swtpm and its tools upon updating them.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Refactor virTPMEmulatorInit to use a loop with parameters. This allows
for easier extension later on.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Move qemuTPMEmulatorInit to virTPMEmulatorInit in virtpm.c and introduce
a few functions to query the executables needed for virCommands.
Add locking to protect the tool paths and return a copy of the tool paths
to callers wanting to access them so that we can run the initialization
function multiples time later on and detect when the executable gets updated.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Extend the TPM device XML parser and XML generator with emulator
state encryption support.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Add support for usage type vTPM to secret.
Extend the schema for the Secret to support the vTPM usage type
and add a test case for parsing the Secret with usage type vTPM.
Signed-off-by: Stefan Berger <stefanb@linux.ibm.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
When using the ENUM macros, the compiler guards that the declaration
and implementation are in sync.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Acked-by: Peter Krempa <pkrempa@redhat.com>
qemuBlockJobRewriteConfigDiskSource rewrites the disk source only
according to the 'target'. This means that if someone would change the
inactive config of the VM to refer to a different disk a block job would
rewrite it when finishing a job which modifies the disk source.
Make sure that this does not happen by verifying that the source of the
config disk is the same.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
ACKed-by: Eric Blake <eblake@redhat.com>
Since we copy everything from the original storage source including some
runtime data which are not relevant for the config we should clear them.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
ACKed-by: Eric Blake <eblake@redhat.com>
Both active block commit and block copy modify the disk source of the
active definition and thus also must modify the corresponding inactive
definition source so that the VM starts up later. This is currently
implemented in the legacy block job handler but the logic will be useful
also for the new handlers. Split it out which also simplifies it.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
ACKed-by: Eric Blake <eblake@redhat.com>
The <mirror> subelement is used in two ways: in a commit job to point to
existing storage, and in a block-copy job to point to additional
storage. We need a way to track only the distinct storage.
This patch introduces qemuBlockJobDiskRegisterMirror which registers the
mirror chain separately only for jobs which require it. This also comes
with remembering that in the status XML.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
ACKed-by: Eric Blake <eblake@redhat.com>
Commit c412383796 used a value from wrong enum when setting the disk's
mirrorState variable. This meant that a 'READY' job would show up as
'PIVOTING'.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
ACKed-by: Eric Blake <eblake@redhat.com>
When returning to asynchronous block job handling the flag which
determines the handling method should be reset prior to flushing
outstanding events. If there's an event to process the handler may
invoke the monitor and another event may be received. We'd not process
that one. Reset the flag earlier.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
qemuDomainSnapshotDiskDataCollect copies the source of the disk from the
live config into the inactive config. Move this operation earlier so
that if we initialize it for use for the particular instance the
run-time-only data is not copied.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
In case when the backing store can be represented with something
simpler such as a URI we can use it rather than falling back to the
json: pseudo-protocol.
In cases when it's not worth it (e.g. with the old ugly NBD or RBD
strings) let's switch to json.
The function is exported as we'll need it when overwriting the ugly
strings qemu would come up with during blockjobs.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The block commit API checked 'disk->src->path' to see whether there
is a reasonable disk source to be committed. As the top image can be
e.g. backed by NBD the check is not good enough. Replace it by
virStorageSourceIsEmpty.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
For the modern use cases we are going to use 'blockdev-snapshot' instead
of 'blockdev-snapshot-sync'.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
qemuBuildStorageSourceChainAttachPrepareBlockdev prepares the full
backing chain for attachment via blockdev. For snapshots we'll need to
prepare one image only as it needs to be plugged on top of the existing
chain.
This patch introduces qemuBuildStorageSourceChainAttachPrepareBlockdevTop
which prepares only @top similarly to the original function by splitting
out the functionality into an internal function so that the API does not
change.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
In commit 042c95bd19 qemuBuildStorageSourceChainAttachPrepareBlockdev
was added but the comment for the function mentions
qemuBuildStorageSourceChainAttachPrepareDrive. Fix the mistake.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Now that we track the job separately we watch only for the abort of the
one single block job so the comment is no longer accurate. Also
describing asynchronous operation is not really necessary.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
With -blockdev:
- we track the job and check it after restart
- have the ability to ask qemu to persist it to collect result
- have the ability to report errors.
This solves all points the comment outlined so remove it. Also all jobs
handle the disk state modification along with the event so there's
nothing special the comment says.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Use job-complete/job-abort instead of the blockjob-* variants for
blockdev.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
As the error message is now available and we know whether the job failed
we can report an error straight away rather than having the user check
the event.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Set the correct job states after the operation is requested in qemu.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
When initiating a pivot or abort of a block job we need to track which
one was initiated. Currently it was done via data stashed in
virDomainDiskDef. Add possibility to track this also together with the
job itself.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Do decisions based on the configuration of the job rather than the data
stored with the disk.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Use the stored job name rather than passing in the disk alias when
referring to the job which allows the same code to work also when
-blockdev will be used.
Note that this API does not require the change to use 'query-job' as it
will ever only work with blockjobs bound to disks due to the arguments
which allow only referring to a disk. For the disk-less jobs we'll need
to add a separate API later.
The change to qemuMonitorGetBlockJobInfo is required as the API was
stripping the 'drive-' prefix when returning the data which is not
desired any more.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
In cgroups v2 when a new group is created by default no controller is
enabled so the detection code will not detect any controllers.
When enabling the controllers we should also store them for the group.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Acked-by: Peter Krempa <pkrempa@redhat.com>
When creating new group for cgroups v2 the we cannot check
cgroups.controllers for that cgroup because the directory is created
later. In that case we should check cgroups.subtree_control of parent
group to get list of controllers enabled for child cgroups.
In order to achieve that we will prefer the parent group if it exists,
the current group will be used only for root group.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Acked-by: Peter Krempa <pkrempa@redhat.com>
When commit 6ac402c456 added the API whenever VIR_DOMAIN_MEM_MAXIMUM
was passed the code always checked whether the domain was active and
therefore failed with an error even though only a config change was
requested. Fix the issue by replacing virDomainObjGetOneDef with
virDomainObjGetOneDefState which tells us what definition we're
performing the change on.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Acked-by: Peter Krempa <pkrempa@redhat.com>
Commit d5572f62e3 forgot to add maxthreads to the non-Linux definition
of the function, thus breaking the MinGW build.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Acked-by: Peter Krempa <pkrempa@redhat.com>
Any message that is easy to trigger (as evidenced by the testsuite
update) should not use 'internal error' as its category.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
virDomainSnapshotFindByName(list, NULL) should return NULL, rather
than the internal-use-only metaroot. Most existing callers pass in a
non-NULL name; the few external callers that don't are immediately
calling virDomainMomentSetParent (which indeed needs the metaroot
rather than NULL if the parent name is NULL); but as the leaky
abstraction is ugly, it is worth instead making
virDomainMomentSetParent static and adding a new function for
resolving the parent link of a brand new moment within its list. The
existing external uses of virDomainMomentSetParent always succeed
(either the new moment has parent_name of NULL to become a new root,
or has parent_name set to a strdup of the previous current moment);
hence, our new function does not need a return value (but it still has
a VIR_WARN in case future uses break our assumptions about failure
being impossible).
Missed when commit 02c4e24d refactored things to attempt to remove
direct metaroot manipulations out of the qemu and test drivers into
internal-only details, and made more obvious when commit dc8d3dc6
factored it out into a separate file.
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Some VM configurations may result in a large number of threads created by
the associated qemu process which can exceed the system default limit. The
maximum number of threads allowed per process is controlled by the pids
cgroup controller and is set to 16k when creating VMs with systemd's
machined service. The maximum number of threads per process is recorded
in the pids.max file under the machine's pids controller cgroup hierarchy,
e.g.
$cgrp-mnt/pids/machine.slice/machine-qemu\\x2d1\\x2dtest.scope/pids.max
Maximum threads per process is controlled with the TasksMax property of
the systemd scope for the machine. This patch adds an option to qemu.conf
which can be used to override the maximum number of threads allowed per
qemu process. If the value of option is greater than zero, it will be set
in the TasksMax property of the machine's scope after creating the machine.
Signed-off-by: Jim Fehlig <jfehlig@suse.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Update the current or max memory, on the persistent or live definition
depending on the flags which are currently ignored.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
Make the test driver only support the VIR_TYPED_PARAM_STRING flag for
now.
Signed-off-by: Ilias Stamatis <stamatis.iliass@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
Avoid the chance that sysconf(_SC_OPEN_MAX) returns -1 and thus
would cause virBitmapNew would attempt to allocate a very large
bitmap.
Found by Coverity
Signed-off-by: John Ferlan <jferlan@redhat.com>
ACKed-by: Peter Krempa <pkrempa@redhat.com>
It's already dereffed in the initialization and shouldn't be NULL
unless virJSONValueArraySize after a virJSONValueObjectGetArray
could return a NULL data entry.
Found by Coverity
Signed-off-by: John Ferlan <jferlan@redhat.com>
ACKed-by: Peter Krempa <pkrempa@redhat.com>
Report in logs when we don't find existing block job data and create it
just to handle the job.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
While this function does start a block job in case when we'd not be able
to get our internal data for it, the handler sets the job state to
QEMU_BLOCKJOB_STATE_RUNNING anyways, thus qemuBlockJobStartupFinalize
would just unref the job.
Since the other usage of qemuBlockJobStartupFinalize in the other part
of the event handler was a bug replace this one anyways even if it would
not cause problems.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The block job event handler qemuProcessHandleBlockJob looks at the block
job data to see whether the job requires synchronous handling. Since the
block job event may arrive before we continue the job handling (if the
job has no data to copy) we could hit the state when the job is still
set as QEMU_BLOCKJOB_STATE_NEW (as we move it to the
QEMU_BLOCKJOB_STATE_RUNNING state only after returning from monitor).
If the event handler uses qemuBlockJobStartupFinalize it would
unregister and free the job. Thankfully this is not a big problem for
legacy blockjobs as we don't need much data for them but since we'd
re-instantiate the job data structure we'd report wrong job type for
active commit as qemu reports it as a regular commit job.
Fix it by not using qemuBlockJobStartupFinalize function in
qemuProcessHandleBlockJob as it is not starting the job anyways.
https://bugzilla.redhat.com/show_bug.cgi?id=1721375
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
virCgroupRemove return -1 when removing cgroup failed.
But there are retry code to remove cgroup in QemuProcessStop:
retry:
if ((ret = qemuRemoveCgroup(vm)) < 0) {
if (ret == -EBUSY && (retries++ < 5)) {
usleep(200*1000);
goto retry;
}
VIR_WARN("Failed to remove cgroup for %s",
vm->def->name);
}
The return value of qemuRemoveCgroup will never be equal to "-EBUSY",
so change the return value of virCgroupRemove if failed.
Signed-off-by: Wang Yechao <wang.yechao255@zte.com.cn>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Shutting down the daemon after 30 seconds of being idle is a little bit
too aggressive. Especially when using 'virsh' in single-shot mode, as
opposed to interactive shell mode, it would not be unusual to have
more than 30 seconds between commands. This will lead to the daemon
shutting down and starting up between a series of commands.
Increasing the shutdown timer to 2 minutes will make it less likely that
the daemon will shutdown while the user is in the middle of a series of
commands.
Reviewed-by: Jim Fehlig <jfehlig@suse.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Instead of having each caller pass in the desired logfile name, pass in
the binary name instead. The logging code can then just derive a logfile
name by appending ".log".
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
This adds detection of a Quobyte as a shared file system for live
migration.
Signed-off-by: Silvan Kaiser <silvan@quobyte.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
We have two functions: virPCIDeviceAddressIsEqual() defined only
on Linux and virPCIDeviceAddressEqual() defined everywhere. And
both of them do the same. Drop the former in favour of the
latter.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
Commit 5ff46aaa7f added a new parameter but neglected to fix the NONNULL
declarations.
Reported-by: Julio Faracco <jcfaracco@gmail.com>
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
When removing the disk fronted while any block job is still active we
need to transfer the ownership of the backing chain to the job itself as
the job still holds the reference to the chain members and thus attempts
to remove them would fail.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
In cases when the disk frontend was unplugged while a blockjob was
running the blockjob inherits the backing chain. When the blockjob is
then terminated we need to unplug the chain as it will not be used any
more.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The PR manager is a property of the format layer in qemu so we need to
be able to track it also in the chains of orphaned block jobs.
Add a helper for qemu to look also into the blockjob state.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
When the guest unplugs the disk frontend libvirt is responsible for
deleting the backend. Since a blockjob may still have a reference to the
backing chain when it is running we'll have to store the metadata for
the unplugged disk for future reference.
This patch adds 'chain' and 'mirrorChain' fields to 'qemuBlockJobData'
to keep them around with the job along with status XML machinery and
tests. Later patches will then add code to change the ownership of the
chain when unplugging the disk backend.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
