We are dropping the only reference here so that the event loop thread
is going to be exited synchronously. In order to avoid deadlocks we
need to unlock the VM so that any handler being called can finish
execution and thus even loop thread be finished too.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
vireventthread: exit thread synchronously on finalize
It it useful to be sure no thread is running after we drop all references to
virEventThread. Otherwise in order to avoid crashes we need to synchronize some
other way or we make extra references in event handler callbacks to all the
object in use. And some of them are not prepared to be refcounted.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
qemu: don't shutdown event thread in monitor EOF callback
This hunk was introduced in [1] in order to avoid loosing
events from monitor on stopping qemu process. But as explained
in [2] on destroy we won't get neither EOF nor any other
events as monitor is just closed. In case of crash/shutdown
we won't get any more events as well and qemuDomainObjStopWorker
will be called by qemuProcessStop eventually. Thus let's
remove qemuDomainObjStopWorker from qemuProcessHandleMonitorEOF
as it is not useful anymore.
[1] e6afacb0f: qemu: start/stop an event loop thread for domains
[2] d2954c072: qemu: ensure domain event thread is always stopped
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Currently we have issues like [1] on libvirtd shutdown as we cleanup while RPC
and other threads are still running. Let's finish all threads other then main
before cleanup.
The approach to finish threads is suggested in [2]. In order to finish RPC
threads serving API calls we let the event loop run but stop accepting new API
calls and block processing any pending API calls. We also inform all drivers of
shutdown so they can prepare for shutdown too. Then we wait for all RPC threads
and driver's background thread to finish. If finishing takes more then 15s we
just exit as we can't safely cleanup in time.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
virNetServerClose and virNetServerShutdownWait are used to start net server
threads shutdown and wait net server threads to actually finish respectively
during net daemon shutdown procedure.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
The function is used to set shutdown prepare and wait callbacks. Prepare
callback is used to inform other threads of the daemon that the daemon will be
closed soon so that they can start to shutdown. Wait callback is used to wait
for other threads to actually finish.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
rpc: don't unref service ref on socket behalf twice
Second unref was added in [1]. We don't need it actually as
we pass free callback to virNetSocketAddIOCallback thus
when we call virNetSocketRemoveIOCallback the extra ref for
callback will be dropped without extra efforts.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Stop just send signal for threads to exit when they finish with
current task. Drain waits when all threads will finish.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Even if we have no priority threads on pool creation we can add them thru
virThreadPoolSetParameters later.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
libvirt: add stateShutdownPrepare/stateShutdownWait to drivers
stateShutdownPrepare is supposed to inform driver that it will be closed soon
so that the driver can prepare and finish all background threads quickly on
stateShutdownWait call.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Give each virtual network bridge its own fixed MAC address
This was a hack to workaround a Linux kernel bug where it would not
honour any attempt to set a MAC address on a bridge. Instead the
bridge would adopt the numerically lowest MAC address of all NICs
attached to the bridge. This lead to the MAC addrss of the bridge
changing over time as NICs were attached/detached.
The Linux bug was actually fixed 3 years before the libvirt
workaround was added in:
Normally, the bridge just chooses the smallest mac address as the
bridge id and mac address of bridge device. But if the administrator
has explictly set the interface address then don't change it.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
but libvirt needed to support RHEL-5 kernels at that time, so
none the less added the workaround.
We have long since dropped support for RHEL-5 vintage distros,
so there's no reason to keep the dummy tap device for the purpose
of setting the bridge MAC address.
Later the dummy TAP device was used for a second purpose related
to IPv6 DAD (Duplicate Address Detection) in:
network: fix dnsmasq/radvd binding to IPv6 on recent kernels
This was again dealing with a regression in the Linux kernel, where
if there were no devices attached to the bridge in the UP state,
IPv6 DAD would not be performed. The virbr0-nic was attached but
in the DOWN state, so the above libvirt fix tenporarily brought
the NIC online. The Linux commit causing the problem was in v2.6.38
IOW, the only reason we need the DAD hack of bringing virbr0-nic
online is because virbr0-nic exists. Once it doesn't exist, then
we hit the "empty bridge" case which works in Linux.
We can rely on distros having Linux kernel >= 3.1, so both things
that the virbr0-nic are doing are redundant.
Fixes https://gitlab.com/libvirt/libvirt/-/issues/53 Reviewed-by: Laine Stump <laine@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Scenario 3 works because libvirt checks current affinity first and
skips the sched_setaffinity call, avoiding the SYS_NICE issue
Scenario 4 works only if CAP_SYS_NICE is availalbe
Scenarios 5 & 6 works only if CAP_SYS_NICE is present *AND* the cgroups
cpuset is not set on the container.
If libvirt blindly ignores the sched_setaffinity failure, then scenarios
4, 5 and 6 should all work, but with caveat in case 4 and 6, that
QEMU will only get 2 CPUs instead of the possible 8 and 4 respectively.
This is still better than failing.
Therefore libvirt can blindly ignore the setaffinity failure, but *ONLY*
ignore it when there was no affinity specified in the XML config.
If user specified affinity explicitly, libvirt must report an error if
it can't be honoured.
Suggested-by: Daniel P. Berrangé <berrange@redhat.com> Signed-off-by: Martin Kletzander <mkletzan@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
peer2peer migration: allow connecting to local sockets
Local socket connections were outright disabled because there was no "server"
part in the URI. However, given how requirements and usage scenarios are
evolving, some management apps might need the source libvirt daemon to connect
to the destination daemon over a UNIX socket for peer2peer migration. Since we
cannot know where the socket leads (whether the same daemon or not) let's decide
that based on whether the socket path is non-standard, or rather explicitly
specified in the URI. Checking non-standard path would require to ask the
daemon for configuration and the only misuse that it would prevent would be a
pretty weird one. And that's not worth it. The assumption is that whenever
someone uses explicit UNIX socket paths in the URI for migration they better
know what they are doing.
tests: Add simple test for virDomainMigrateCheckNotLocal
For this we need to make the function accessible (at least privately). The
behaviour will change in following patches and the test helps explaining the
change.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com> Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
Instead of saving some data from a union up front and changing an overlayed
struct before using said data, let's just set the new values after they are
decided. This will increase the readability of future commit(s).
Signed-off-by: Martin Kletzander <mkletzan@redhat.com> Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
qemu_namespace: Be tolerant to non-existent files when populating /dev
In 6.7.0 release I've changed how domain namespace is built and
populated. Previously it used to be done from a pre-exec hook
(ran in the forked off child, just before dropping all privileges
and exec()-ing QEMU), which not only meant we had to have two
different code paths for creating a node in domain's namespace
(one for this pre-exec hook, the other for hotplug ran from the
daemon), it also proved problematic because it was leaking FDs
into QEMU process.
To mitigate this problem, we've not only ditched libdevmapper
from the NS population process, I've also dropped the pre-exec
code and let the NS be populated from the daemon (using the
hotplug code). But, I was not careful when doing so, because the
pre-exec code was tolerant to files that doesn't exist, while
this new code isn't. For instance, the very first thing that is
done when the new NS is created is it's populated with
@defaultDeviceACL which contain files like /dev/null, /dev/zero,
/dev/random and /dev/kvm (and others). While the rest will
probably exist every time, /dev/kvm might not and thus the new
code I wrote has to be tolerant to that.
Of course, users can override the @defaultDeviceACL (by setting
cgroup_device_acl in qemu.conf) and remove /dev/kvm (which is
acceptable workaround), but we definitely want libvirt to work
out of the box even on hosts without KVM.
Fixes: 9048dc4e627ddf33996084167bece7b5fb83b0bc Reported-by: Daniel P. Berrangé <berrange@redhat.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
vsh: Define HAVE_STDARG_H before including readline
As it turned out my previous commits which switched from HAVE_ to
WITH_ and dropped stdarg.h detection were a bit too aggressive.
Because of reasons described in 9ea3424a178 we need to define
HAVE_STDARG_H before including readline otherwise macos build
fails. Honestly, I still don't fully understand the problem so I
am not going to bother you with "explanation".
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
This allows us to drop include of readline header files from
virsh.c and virt-admin.c because they needed it only because of
the add_history() function.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Pavel Hrdina [Tue, 1 Sep 2020 12:54:27 +0000 (14:54 +0200)]
meson: add -Wall and -Wextra explicitly
If someone runs `meson setup --buildtype plain` meson ignores
warning_level=2 that is in our meson.build file. The implication is
that Meson will not automatically add -Wall which enables -Wformat.
This breaks building libvirt from git with the buildtype set to plain.
There is an issue reported [1] to not ignore warning_level silently
and the change to ignore it was done by upstream commit [2].
This change makes Meson print warnings about using warning_level which
as described above doesn't work for all build types.
remote: use SocketMode=0600 when polkit is not compiled
The systemd .socket unit files we ship for libvirt daemons use
SocketMode=0666 on the assumption that libvirt is built with
polkit which provides access control.
Some people, however, may have explicitly turned off polkit at
build time and not realize that leaves them insecure unless
they also change the SocketMode. This addresses that problem
by making the SocketMode default to 0600 when polkit is
disabled at compile time.
Note we cannot automatically fix the case where the user
compiles polkit, but then overrides the libvirtd.conf defaults
to disable polkit. This is what lead to CVE-2020-15708 in
Ubuntu 20.10. We can at least improve the inline comments
in the config file to give a clearer warning though, which
may have helped avoid the mistaken config.
Reviewed-by: Jiri Denemark <jdenemar@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Rather than use the names "fial" and "kep", use "fail" and "keep". In
the DO_TEST() macro, to prevent the preprocessor replacing the struct
member names during assignment, use the names "fail_" and "keep_"
instead.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com> Reviewed-by: Laine Stump <laine@redhat.com>
Signed-off-by: Han Han <hhan@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Thomas Huth [Mon, 20 Jul 2020 10:22:33 +0000 (12:22 +0200)]
qemu: Fix domfsinfo for non-PCI device information from guest agent
qemuAgentFSInfoToPublic() currently only sets the devAlias for PCI devices.
However, the QEMU guest agent could also provide the device name in the
"dev" field of the response for other devices instead (well, at least after
fixing another problem in the current QEMU guest agent...). So if creating
the devAlias from the PCI information failed, let's fall back to the name
provided by the guest agent. This helps to fix the empty "Target" fields
that occur when running "virsh domfsinfo" on s390x where CCW devices are
used for the guest instead of PCI devices.
Also add a proper debug message here in case we completely failed to set the
device alias, since this problem here was very hard to debug: The only two
error messages that I've seen were "Unable to get filesystem information"
and "Unable to encode message payload" - which only indicates that something
went wrong in the RPC call. No debug message indicated the real problem, so
I had to learn the hard way why the RPC call failed (it apparently does not
like devAlias left to be NULL) and where the real problem comes from.
Thomas Huth [Wed, 5 Aug 2020 10:22:26 +0000 (12:22 +0200)]
qemu: Do not silently allow non-available timers on non-x86 systems
libvirt currently silently allows <timer name="kvmclock"/> and some
other timer tags in the guest XML definition for timers that do not
exist on non-x86 systems. We should not silently ignore these tags
since the users might not get what they expected otherwise.
Note: The error is only generated if the timer is marked with
present="yes" - otherwise we would suddenly refuse XML definitions
that worked without problems before.
Buglink: https://bugzilla.redhat.com/show_bug.cgi?id=1754887 Signed-off-by: Thomas Huth <thuth@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Boris Fiuczynski <fiuczy@linux.ibm.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Andrea Bolognani [Tue, 25 Aug 2020 15:52:24 +0000 (17:52 +0200)]
AUTHORS: Convert to reStructuredText
Now that we have moved to Meson, we are no longer required to
use a specific name for this file, and since the rest of our
documentation is in reStructuredText format and uses a matching
file extension, we can give the AUTHORS file the same treatment.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Currently, we are mixing: #if HAVE_BLAH with #if WITH_BLAH.
Things got way better with Pavel's work on meson, but apparently,
mixing these two lead to confusing and easy to miss bugs (see 31fb929eca for instance). While we were forced to use HAVE_
prefix with autotools, we are free to chose our own prefix with
meson and since WITH_ prefix appears to be more popular let's use
it everywhere.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
In packet-libvirt.c in wireshark dissector we include rpc/types.h
but guard the include with a condition (that is supposed to be
true if we detected during configure phase that the host system
has the header file). Thing is, it looks like we were never doing
the configure check and thus the file was never included and yet,
the NSS plugin works. Drop the include then.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
meson: Drop checks for some functions and header files
In meson.build, we check for presence of unshare() function
explicitly (even though there is the function usability check a
few hundred lines below), but never have any code depending on
HAVE_UNSHARE. The same applies to stdarg.h and sys/sysctl.h
header files - either we simply include them or guard their
include using different conditionals.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
There are couple of conditional #includes at the beginning of
virfile.c and they try to be nice and document #endifs. But they
are mostly wrong because either they have the condition in the
comment inverted or the comment refers to a different condition
than they belong to. Just remove the comments as these #includes
are single line mostly.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Setting IFF_VNET_HDR for a tap device passes the whole packet to the
host, reducing emulation overhead and improving performance.
Libvirt bases its decision about applying IFF_VNET_HDR to the tap
interface on whether or not the model of the emulated network device
is virtio. Originally, virtio was the only model to support
IFF_VNET_HDR in QEMU; however, the e1000e & vmxnet3 adapters have also
supported it since their introductions - QEMU commit 786fd2b0f87 for vmxnet3, and QEMU commit 6f3fbe4ed0 for e1000e, so it
should be set for those models too.
Signed-off-by: Patrick Magauran <patmagauran.j@gmail.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Laine Stump <laine@redhat.com>
Scott Shambarger [Fri, 28 Aug 2020 19:35:01 +0000 (12:35 -0700)]
tests/meson: fix mock library link on MacOS
MacOS can not pre-load modules, so mock libraries must be built
as shared libraries (without asneeded striping, and undefined
symbols allowed).
Signed-off-by: Scott Shambarger <scott-libvirt@shambarger.net> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Signed-off-by: Ján Tomko <jtomko@redhat.com>
Ján Tomko [Tue, 1 Sep 2020 21:30:35 +0000 (23:30 +0200)]
conf: fix enum conversion
../src/conf/domain_conf.c:8144:78: error: result of comparison of unsigned enum expression < 0 is always false [-Werror,-Wtautological-unsigned-enum-zero-compare]
if ((def->writeFiltering = virTristateBoolTypeFromString(filtering)) < 0) {
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~
Signed-off-by: Ján Tomko <jtomko@redhat.com> Fixes: 9d15647dcb96831c93ac8c1d67c47265b5ed9072
Jim Fehlig [Fri, 14 Aug 2020 20:47:09 +0000 (14:47 -0600)]
Xen: Add support for writeFiltering in config converter
Add support for the writeFiltering attribute in the domXML to native
config converter. Also include a test.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Jim Fehlig [Fri, 14 Aug 2020 16:28:38 +0000 (10:28 -0600)]
Xen: Add writeFiltering option for PCI devices
By default Xen only allows guests to write "known safe" values into PCI
configuration space, yet many devices require writes to other areas of
the configuration space in order to operate properly. To allow writing
any values Xen supports the 'permissive' setting, see xl.cfg(5) man page.
This change models Xen's permissive setting by adding a writeFiltering
attribute on the <source> element of a PCI hostdev. When writeFiltering
is set to 'no', the Xen permissive setting will be enabled and guests
will be able to write any values into the device's configuration space.
The permissive setting remains disabled in the absense of the
writeFiltering attribute, of if it is explicitly set to 'yes'.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Signed-off-by: Simon Gaiser <simon@invisiblethingslab.com> Signed-off-by: Marek Marczykowski-Górecki <marmarek@invisiblethingslab.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Jim Fehlig [Thu, 20 Aug 2020 21:52:17 +0000 (15:52 -0600)]
qemu: Check for changes in qemu modules directory
Add a configuration option for specifying location of the qemu modules
directory, defaulting to /usr/lib64/qemu. Then use this location to
check for changes in the directory, indicating that a qemu module has
changed and capabilities need to be reprobed.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Laine Stump [Mon, 24 Aug 2020 01:20:13 +0000 (21:20 -0400)]
util: assign tap device names using a monotonically increasing integer
When creating a standard tap device, if provided with an ifname that
contains "%d", rather than taking that literally as the name to use
for the new device, the kernel will instead use that string as a
template, and search for the lowest number that could be put in place
of %d and produce an otherwise unused and unique name for the new
device. For example, if there is no tap device name given in the XML,
libvirt will always send "vnet%d" as the device name, and the kernel
will create new devices named "vnet0", "vnet1", etc. If one of those
devices is deleted, creating a "hole" in the name list, the kernel
will always attempt to reuse the name in the hole first before using a
name with a higher number (i.e. it finds the lowest possible unused
number).
The problem with this, as described in the previous patch dealing with
macvtap device naming, is that it makes "immediate reuse" of a newly
freed tap device name *much* more common, and in the aftermath of
deleting a tap device, there is some other necessary cleanup of things
which are named based on the device name (nwfilter rules, bandwidth
rules, OVS switch ports, to name a few) that could end up stomping
over the top of the setup of a new device of the same name for a
different guest.
Since the kernel "create a name based on a template" functionality for
tap devices doesn't exist for macvtap, this patch for standard tap
devices is a bit different from the previous patch for macvtap - in
particular there was no previous "bitmap ID reservation system" or
overly-complex retry loop that needed to be removed. We simply find
and unused name, and pass that name on to the kernel instead of
"vnet%d".
This counter is also wrapped when either it gets to INT_MAX or if the
full name would overflow IFNAMSIZ-1 characters. In the case of
"vnet%d" and a 32 bit int, we would reach INT_MAX first, but possibly
someday someone will change the name from vnet to something else.
(NB: It is still possible for a user to provide their own
parameterized template name (e.g. "mytap%d") in the XML, and libvirt
will just pass that through to the kernel as it always has.)
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Laine Stump [Sun, 23 Aug 2020 18:57:19 +0000 (14:57 -0400)]
util: replace macvtap name reservation bitmap with a simple counter
There have been some reports that, due to libvirt always trying to
assign the lowest numbered macvtap / tap device name possible, a new
guest would sometimes be started using the same tap device name as
previously used by another guest that is in the process of being
destroyed *as the new guest is starting.
In some cases this has led to, for example, the old guest's
qemuProcessStop() code deleting a port from an OVS switch that had
just been re-added by the new guest (because the port name is based on
only the device name using the port). Similar problems can happen (and
I believe have) with nwfilter rules and bandwidth rules (which are
both instantiated based on the name of the tap device).
A couple patches have been previously proposed to change the ordering
of startup and shutdown processing, or to put a mutex around
everything related to the tap/macvtap device name usage, but in the
end no matter what you do there will still be possible holes, because
the device could be deleted outside libvirt's control (for example,
regular tap devices are automatically deleted when the qemu process
terminates, and that isn't always initiated by libvirt but could
instead happen completely asynchronously - libvirt then has no control
over the ordering of shutdown operations, and no opportunity to
protect it with a mutex.)
But this only happens if a new device is created at the same time as
one is being deleted. We can effectively eliminate the chance of this
happening if we end the practice of always looking for the lowest
numbered available device name, and instead just keep an integer that
is incremented each time we need a new device name. At some point it
will need to wrap back around to 0 (in order to avoid the IFNAMSIZ 15
character limit if nothing else), and we can't guarantee that the new
name really will be the *least* recently used name, but "math"
suggests that it will be *much* less common that we'll try to re-use
the *most* recently used name.
This patch implements such a counter for macvtap/macvlan, replacing
the existing, and much more complicated, "ID reservation" system. The
counter is set according to whatever macvtap/macvlan devices are
already in use by guests when libvirtd is started, incremented each
time a new device name is needed, and wraps back to 0 when either
INT_MAX is reached, or when the resulting device name would be longer
than IFNAMSIZ-1 characters (which actually is what happens when the
template for the device name is "maccvtap%d"). The result is that no
macvtap name will be re-used until the host has created (and possibly
destroyed) 99,999,999 devices.
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Laine Stump [Wed, 26 Aug 2020 18:22:07 +0000 (14:22 -0400)]
meson: link libm
On some platforms libm (needed for the pow() function) isn't being
linked in somehow. This patch adds the necessary bits to assure that
it's linked in when necessary.
Suggested-by: Michal Privoznik <mprivozn@redhat.com> Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
(cherry picked from commit 20a62b42ec001310a6329d7ee2021f0737d534ef)
Andrea Bolognani [Tue, 25 Aug 2020 16:30:57 +0000 (18:30 +0200)]
meson: Use @BASENAME@ more
This patch takes care of just the obvious cases: there are
many more situations where the data we pass to configure_file()
could likely be obtained in a more effective way, but we can
address the low-hanging fruits as a first approximation.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
When switching to meson, some of HAVE_* macros were renamed to
WITH_ because they did not reflect whether the build platform has
or doesn't have something, but whether we are building with some
functionality turned on or off. This is the case with
HAVE_BSD_NSS macro too. As a result, the NSS plugin built on BSD
did not expose nss_module_register() function which made the
plugin unusable:
Neal Gompa [Sun, 30 Aug 2020 00:09:18 +0000 (20:09 -0400)]
rpm: Fix conditional for defining %_vpath_builddir for RHEL <= 7
The conditional was incorrectly overriding %_vpath_builddir when
%rhel is not defined, which led to surprising behavior when the
global %_vpath_builddir path is set on Fedora already.
Michal Privoznik [Tue, 28 Jul 2020 08:51:32 +0000 (10:51 +0200)]
qemu: Validate memory hotplug in domainValidateCallback instead of cmd line generator
When editing a domain with hotplug enabled, I removed the only
NUMA node it had and got no error. I got the error later though,
when starting the domain. This is not as user friendly as it can
be. Move the validation call out from command line generator and
into domain validator (which is called prior to starting cmd line
generation anyway).
When doing this, I had to remove memory-hotplug-nonuma xml2xml
test case because there is no way the test case can succeed,
obviously.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Michal Privoznik [Mon, 31 Aug 2020 09:35:47 +0000 (11:35 +0200)]
RNG: Allow interleaving of /domain/cpu/numa/cell children
So far, the <cell/> element can have two types of children
elements: <distances/> and <cache/> (which can be repeated more
times). However, there is no reason to require specific order in
input XML. Allow elements to be interleaved.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Fangge Jin [Thu, 20 Aug 2020 10:09:24 +0000 (18:09 +0800)]
docs: add kbase entry for migrationinternals
Commit c051e56d27 added migrationinternals.rst in kbase, but the
entry was missing.
Signed-off-by: Fangge Jin <fjin@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Signed-off-by: Ján Tomko <jtomko@redhat.com>
Scott Shambarger [Tue, 25 Aug 2020 23:47:07 +0000 (16:47 -0700)]
util: use host module suffix when loading drivers
Driver module loaders current hardcode ".so" as the file
extension. On MacOS, meson uses ".dylib" as a module file extension.
This patch adds VIR_FILE_MODULE_EXT to virfile.h defined as the
hosts module extension, and updates driver module loaders to make
use of it.
Signed-off-by: Scott Shambarger <scott-libvirt@shambarger.net> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Previous patch handled the runtime case where a non-x86 host is
fetching /proc/cpuinfo data for a microcode info that we know
it doesn't exist. This change alone speeded everything by a
bit for non-x86, but there is at least one major culprit left.
qemuxml2argvtest does several arch-specific tests, and a good
chunk of them are x86 exclusive. This means that 'hostArch'
will be seen as x86 for these tests, even when running in
non-x86 hosts. In a Power 9 server with 128 CPUs, qemuxml2argvtest
takes 298 seconds to complete in average, and 'perf record'
indicates that 95% of the time is spent in
virHostCPUGetMicrocodeVersion().
This patch mocks virHostCPUGetMicrocodeVersion() to always return
0 in the tests, avoiding /proc/cpuinfo reads. This will make all
tests behave arch-agnostic, and the microcode value being 0 has no
impact on any existing test.
This is a CI speed across the board for all archs, including x86,
given that we're not reading /proc/cpuinfo in the tests. For
a Thinkpad T480 laptop with 8 Intel i7 CPUs, qemuxml2argvtest
went from 15.50 sec to 12.50 seconds. The performance gain is even
more noticeable for huge servers with lots of CPUs. For the
Power 9 server mentioned above, this patch speeds qemuxml2argvtest
to 9 seconds, down from 298 sec.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Ján Tomko <jtomko@redhat.com> Signed-off-by: Ján Tomko <jtomko@redhat.com>
virhostcpu.c: skip non x86 hosts in virHostCPUGetMicrocodeVersion()
Non-x86 archs does not have a 'microcode' version like x86. This is
covered already inside the function - just return 0 if no microcode
is found. Regardless of that, a read of /proc/cpuinfo is always made.
Each read will invoke the kernel to fill in the CPU details every time.
Now let's consider a non-x86 host, like a Power 9 server with 128 CPUs.
Each /proc/cpuinfo read will need to fetch data for each CPU and it
won't even matter because we know beforehand that PowerPC chips don't
have microcode information.
We can do better for non-x86 hosts by skipping this process entirely.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com> Signed-off-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
projects / thirdparty / libvirt.git / log
