Commit Graph

36388 Commits

Author SHA1 Message Date
Julio Faracco
6cf9577498 lxc: remove domain definition from lxc network struct
Domain definition is useless now inside network structure. This pointer
was required because new network definition was being added each time
that a new network type appeared. So, this should be processed into
old function `lxcNetworkParseDataType()`. Now, as it was moved to an
array, it can be handle together each interface pointer.

Signed-off-by: Julio Faracco <jcfaracco@gmail.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-02-03 16:08:20 +01:00
Julio Faracco
0da46e30a0 lxc: add LXC version 3 network parser
LXC version 3 or higher introduced indexes for network interfaces.
Libvirt should be able to parse entries like `lxc.net.2.KEY`. This
commit adds functions to parse this type of field. That's why array
structures are so important this time.

Signed-off-by: Julio Faracco <jcfaracco@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-02-03 16:07:56 +01:00
Julio Faracco
007925214e lxc: refactor lxcNetworkParseData pointers to use new structures
Struct lxcNetworkParseData is being used as a single pointer which
iterates through LXC config lines. It means that it will be applied as a
network each time that a new type appears. After, the same struct is
used to populate a new network interface. This commit changes this logic
to multiple lxcNetworkParseData to move this strcuture to an array. It
makes more sense if we are using indexes to fill interface settings.
This is better to improve code clarity.

This commit still introduces *Legacy() functions to keep support of
network old style definitions.

Signed-off-by: Julio Faracco <jcfaracco@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-02-03 16:07:41 +01:00
Andrea Bolognani
79ebc31a1b tests: Add capabilities for QEMU 5.0.0 on aarch64
This will be the first QEMU version that will support the
kvm-no-adjvtime CPU feature.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-02-03 09:20:28 +01:00
Michal Privoznik
a37a8c569d Drop virAtomic module
Now, that every use of virAtomic was replaced with its g_atomic
equivalent, let's remove the module.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-02-02 16:36:58 +01:00
Michal Privoznik
7390ff3caa src: Drop virAtomicIntDecAndTest() with g_atomic_int_dec_and_test()
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-02-02 16:36:56 +01:00
Michal Privoznik
574678a27f src: Replace virAtomicIntInc() with g_atomic_int_add()
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-02-02 16:36:54 +01:00
Michal Privoznik
f127acfa34 src: Replace virAtomicIntSet() with g_atomic_int_set()
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-02-02 16:36:50 +01:00
Michal Privoznik
8567ff1f5c src: Replace virAtomicIntGet() with g_atomic_int_get()
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-02-02 16:36:49 +01:00
Michal Privoznik
18a6f18280 tests: Drop viratomictest
In future commits our virAtomic* APIs will be replaced with their
GLib variants. Instead of trying to update the test after each
commit and eventually removing the test anyway, remove it upfront
and save the hassle.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-02-02 16:36:47 +01:00
Michal Privoznik
7c541207bb test_driver: Replace virAtomicIntAdd() with g_atomic_int_add()
Instead of calling virAtomicIntAdd(&var, 1); we can call
g_atomic_int_add() directly.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-02-02 16:35:20 +01:00
Ján Tomko
62d75cdcc6 qemu: drop unused variable
The g_auto conversion made clang realize the variable is unused:
../../src/qemu/qemu_domain.c:10349:36: error: unused variable
    'cfg' [-Werror,-Wunused-variable]
    g_autoptr(virQEMUDriverConfig) cfg = virQEMUDriverGetConfig(driver);

Signed-off-by: Ján Tomko <jtomko@redhat.com>
Fixes: 20fa2bc6e5
2020-01-31 17:02:26 +01:00
Jim Fehlig
72ed254b38 build: Drop execute permissions on virsh bash completion command
rpmlint complains about "script-without-shebang" due to the execute
permissions on /usr/share/bash-completion/completions/vsh. Use
INSTALL_DATA instead of INSTALL_SCRIPT to avoid the unnecessary
execute permissions.

Signed-off-by: Jim Fehlig <jfehlig@suse.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 08:49:04 -07:00
Ján Tomko
e8d5eb0cde qemu: Fix domain ID allocation
The rewrite to use GLib's atomic ops functions changed the behavior
of virAtomicIntInc - before it returned the pre-increment value.

Most of the callers using its value were adjusted, but the one
in qemuDriverAllocateID was not. If libvirtd would reconnect to
a running domain during startup, the next started domain would get
the same ID:

$ virsh list
 Id   Name       State
--------------------------
 1    f28live    running
 1    f28live1   running

Use the g_atomic_add function directly (as recommended in viratomic.h)
and add 1 to the result.

This also restores the usual numbering from 1 instead of 0.

Signed-off-by: Ján Tomko <jtomko@redhat.com>
Fixes: 7b9645a7d1
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2020-01-31 16:11:50 +01:00
Ján Tomko
5b63cb5abf docs: fix a typo
s/axnd/and/

Signed-off-by: Ján Tomko <jtomko@redhat.com>
Fixes: f0f34056ab
2020-01-31 15:57:00 +01:00
Daniel Henrique Barboza
4427903722 qemu_domain_address.c: turn qemuDomainFillDeviceIsolationGroup to void
Starting on commit 1f43393283, qemuDomainFillDeviceIsolationGroup()
returns 0 in all circunstances. Let's turn it to 'void' make it
clearer that the function will not fail. This also spares a
check for < 0 return in qemu_hotplug.c. The
qemuDomainFillDeviceIsolationGroupIter() callback now returns
0 at all times - which is already happening anyway.

Refer to 1f43393283 commit message for more details on why
the function was changed to never return an error.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 15:25:33 +01:00
Daniel Henrique Barboza
02ad5d2359 qemu_domain.c: turn qemuDomainChrDefDropDefaultPath to void
qemuDomainChrDefDropDefaultPath() returns an int, but it's
always returning 0. Callers are checking for result < 0 to
run their cleanup code needlessly.

Turn the function to 'void' and adjust the callers.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 15:25:33 +01:00
Daniel Henrique Barboza
db780004a9 qemu_domain.c: removing unneeded cleanup labels
Previous patches deprecated some 'cleanup' labels. Let's remove
them.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 15:25:33 +01:00
Daniel Henrique Barboza
20fa2bc6e5 qemu_domain.c: use g_autoptr when possible
Avoid some of the virObjectUnref() calls by using g_autoptr.
Aside from the 'cleanup' label in qemuDomainSetFakeReboot(),
all other now deprecated cleanup labels will be removed in
the next patch.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 15:25:33 +01:00
Daniel Henrique Barboza
ae9f630e50 qemu_domain.c: use g_autofree when possible
Use g_autofree to remove VIR_FREE() calls used for cleanups.
Labels that became deprecated will be removed in a later
patch.

In qemuDomainSetupDisk(), the 'dst' variable is not used at
all and could be removed.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 15:25:33 +01:00
Daniel Henrique Barboza
749ec34a56 qemu_domain.c: remove redundant virObjectUnref()
The 'caps' variable in qemuDomainObjPrivateXMLParseAutomaticPlacement()
is set to auto clean via g_autoptr(), but a 'virObjectUnref(caps)' is
being executed in the 'cleanup' label.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 15:25:33 +01:00
Daniel Henrique Barboza
738bedbf4e conf/domain_addr.c: remove unneeded 'cleanup' labels
Previous patch used 'g_autofree' to eliminate instances of
VIR_FREE(), making some cleanup labels obsolete. This
patch removes them.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 15:25:33 +01:00
Daniel Henrique Barboza
3f9f550061 conf/domain_addr.c: use g_autofree in strings
Use g_autofree in strings when possible to spare a VIR_FREE()
call. Unneeded 'cleanup' labels will be taken care of in the
next patch.

The 'str' string in virDomainVirtioSerialAddrReserve() was
never used by the logic, only being used in cleanup by
VIR_FREE(). Let's remove it.

Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 15:25:33 +01:00
Peter Krempa
b9e87908db qemuMigrationCookieAddNBD: Fix filling of 'capacity' when blockdev is used
With -blockdev we must look up via the nodename rather than the 'drive'
alias which is not present any more.

This fixes the pre-creation of storage volumes on migration with
non-shared storage.

https://bugzilla.redhat.com/show_bug.cgi?id=1793263

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:29 +01:00
Peter Krempa
d409411213 qemuMigrationCookieAddNBD: Remove 'ret' variable and 'cleanup' label
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:29 +01:00
Peter Krempa
45eefb2c78 qemuMigrationCookieAddNBD: Use virHashNew and automatic freeing of virHashTablePtr
Swithc to the helper which doesn't require checking of the return value.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:29 +01:00
Peter Krempa
464345e153 qemuMigrationCookieAddNBD: Move monitor call out of the loop
The data is gathered only once so we can move the whole block which
fetches the data out of the loop and get rid of the logic which
prevents multiple calls.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:29 +01:00
Peter Krempa
8efeeb59a6 qemuMigrationCookieAddNBD: Use glib memory allocators
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:29 +01:00
Peter Krempa
3093822d1d qemuMigrationCookieNBD: Extract embedded struct
Extract the struct so that it's type has a name.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:28 +01:00
Peter Krempa
bdff9d4513 qemuMigrationCookieAddNBD: Exit early if there are no disks
Refactor the logic to skip the body of the function if there's nothing
to do.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:28 +01:00
Peter Krempa
6eab924daa Remove checking of return value of virHashNew
There are two calls to virHashNew which check the return value. It's not
necessary any more as virHashNew always returns a valid pointer.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:28 +01:00
Peter Krempa
2a5ea0a0c1 conf: domain: Remove checking of return value of virHashCreateFull
This module has last two direct checks whether the value returned by
virHashCreateFull is NULL. Remove them so that static analyzers don't
get the false idea that checking the value is necessary.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:28 +01:00
Peter Krempa
50f7483a0d util: hash: Use g_new0 for allocating hash internals
Use the glib helpers and remove the mention of returning NULL on failure
of virHashNew, virHashCreate and virHashCreateFull.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2020-01-31 14:28:28 +01:00
Sahid Orentino Ferdjaoui
b918bfba6f docs: update Rust releases and resources links
This is updating the releases and resources links so they point now to
crates.io for the releases and docs.rs for the api ref.

Signed-off-by: Sahid Orentino Ferdjaoui <sahid.ferdjaoui@canonical.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
2020-01-31 12:01:36 +01:00
Andrea Bolognani
193ad36406 docs: Update godoc URLs for Go packages
Since

  commit 201156ec6e
  Author: Daniel P. Berrangé <berrange@redhat.com>
  Date:   Mon Dec 16 10:06:50 2019 +0000

    docs: add pages to support Go module package resolution

it's possible to use

  import "libvirt.org/libvirt-go"
  import "libvirt.org/libvirt-go-xml"

in Go programs, which will result in the Go packages being
fetched directly from libvirt.org instead of GitHub.

The godoc website offers the API documentation for the updated
imports, and that's where we should send users.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-01-31 11:59:43 +01:00
Andrea Bolognani
87b811279a docs: Update download location for Go packages
We haven't ever made releases of the Go components, which makes
sense because that's just not how that specific ecosystem works.

For people who are looking to use libvirt from Go, the brief
introductory pages located at the same URLs used to import the
packages are a better entry point than an empty directory
listing, so let's send them there instead.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2020-01-31 11:59:04 +01:00
Christian Ehrhardt
5a21fd513a
apparmor: fix qemu_bridge_helper for named profile
Since a3ab6d42 "apparmor: convert libvirtd profile to a named profile"
the detection of the subelement for qemu_bridge_helper is wrong.

In combination with the older 123cc3e1 "apparmor: allow
/usr/lib/qemu/qemu-bridge-helper" it now detects qemu-bridge-helper no
more with its path, but instead as a proper subelement of the named profile
like: label=libvirtd//qemu_bridge_helper

In the same fashion the reverse rule in the qemu_bridge_helper
sub-profile still uses the path and not the named profile label.

Triggering denies like:
apparmor="DENIED" operation="file_inherit"
  profile="libvirtd//qemu_bridge_helper" pid=5629 comm="qemu-bridge-hel"
  family="unix" sock_type="stream" protocol=0 requested_mask="send receive"
  denied_mask="send receive" addr=none peer_addr=none peer="libvirtd"

This patch fixes the unix socket rules for the communication between
libvirtd and qemu-bridge-helper to match that.

Fixes: a3ab6d42d8
Fixes: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1655111

Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
2020-01-31 08:32:00 +01:00
Erik Skultety
2b082d875d nwfilter: Use immediate packet delivery mode rather than buffering
Our nwfilter code doesn't set any timeout on the pcap packet buffer which
means that when DHCP snooping is enabled on a guest interface and
libvirt is trying to learn the IP address from guest's DHCP traffic, it
takes up to 4x longer to ping a guest successfully compared to a case
where nwfilter isn't enabled at all or libvirt uses the cached nwfilter
leases to populate the corresponding rules to ebtables.
With the pcap filter and rate limiting already in place, we should be
able to afford enabling the immediate packet delivery, FWIW immediate
mode was actually the default prior libpcap-1.5.0 (CentOS 6) regardless
of whether a buffer was requested.

The lack of any kind of timeout on the pcap buffer messed with the
libvirt TCK test suite which, even with a generous timeout in place,
timeouts every single time simply because it takes a while until
guest actually starts producing any kind of traffic to fill up
the buffer in place (apart from the DHCP traffic which happens fairly
early on).

Signed-off-by: Erik Skultety <eskultet@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2020-01-30 16:30:18 +01:00
Erik Skultety
77c534033c libpcap: Bump the minimum required version to >= 1.5.0
libpcap-1.5.0 introduced a function to enforce immediate mode (on all
platforms) which the follow-up patches will rely on.

Signed-off-by: Erik Skultety <eskultet@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2020-01-30 16:30:12 +01:00
Ján Tomko
49882b3337 Add a space before ending a comment
Also add a space after the start in some of the cases.

Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2020-01-30 12:32:03 +01:00
Michal Privoznik
2f74105d2c apparmor: Drop 'Last modified' comment from profiles
At the beginning of each profile we have a comment that says when
the profile was last updated. In theory, it makes sense because
one can see immediately if they are using an outdated profile.
However, we don't do a good job in keeping the comments in sync
with reality and also sysadmins should rather use their package
manager to find out libvirt version which installed the profiles.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Acked-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
2020-01-30 08:57:59 +01:00
Michal Privoznik
8f204fb4da apparmor: Allow some more BIOS/UEFI paths
There are two more paths that we are missing in the default
domain profile: /usr/share/edk2-ovmf/ and /usr/share/sgabios/.
These exist on my Gentoo box and contain UEFI and BIOS images
respectively.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Acked-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
2020-01-30 08:57:59 +01:00
Michal Privoznik
07af71ad99 apparmor: Sort paths in blocks in libvirt-qemu profile
Even though we construct a domain specific profile for each
domain we start (which should cover domain specific paths), there
is also another file that is included from the profile and which
contains domain agnostic paths (e.g. to cover libraries that qemu
links with). The paths in the file are split into blocks divided
by comments. Sort the paths in each block individually (ignoring
case sensitivity).

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Acked-by: Christian Ehrhardt <christian.ehrhardt@canonical.com>
2020-01-30 08:57:59 +01:00
Dario Faggioli
849052ec61 libxl: support getting and setting parameters for the Credit2
With Credit2 being Xen default scheduler, it's definitely the case to
allow Credit2's scheduling parameters to be get and set via libvirt.

This is easy, as Credit and Credit2 have (at least as of now) the very
same parameters ('weight' and 'cap'). So we can just let credit2 pass
the scheduler-type check and the same code will work for both.

Signed-off-by: Dario Faggioli <dfaggioli@suse.com>
Reviewed-by: Jim Fehlig <jfehlig@suse.com>
2020-01-29 17:27:08 -07:00
Laine Stump
f0f34056ab docs: document <interface> subelement <teaming>
Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2020-01-29 15:33:29 -05:00
Laine Stump
8a226ddb36 qemu: add wait-unplug to qemu migration status enum
Aside from itinerant error (actually warning) messages due to an
unrecognized response from qemu, this isn't even necessary - the
migration proceeds successfully to completion anyway.

(I'm not sure where to see this status reported in the API though - do
we need to add an extra state, or recognition of a new event somewhere?)

Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2020-01-29 15:33:29 -05:00
Laine Stump
2758f680b7 qemu: allow migration with assigned PCI hostdev if <teaming> is set
Normally a PCI hostdev can't be migrated, so
qemuMigrationSrcIsAllowedHostdev() won't permit it. In the case of a a
hostdev network interface that has <teaming type='transient'/> set,
QEMU will automatically unplug the device prior to migration, and
re-plug a corresponding device on the destination. This patch modifies
qemuMigrationSrcIsAllowedHostdev() to allow domains with those devices
to be migrated.

Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2020-01-29 15:33:29 -05:00
Laine Stump
eb9f6cc4b3 qemu: support interface <teaming> functionality
The QEMU driver uses the <teaming type='persistent|transient'
persistent='blah'/> element to setup a "failover" pair of devices -
the persistent device must be a virtio emulated NIC, with the only
extra configuration being the addition of ",failover=on" to the device
commandline, and the transient device must be a hostdev NIC
(<interface type='hostdev'> or <interface type='network'> with a
network that is a pool of SRIOV VFs) where the extra configuration is
the addition of ",failover_pair_id=$aliasOfVirtio" to the device
commandline. These new options are supported in QEMU 4.2.0 and later.

Extra qemu-specific validation is added to ensure that the device
type/model is appropriate and that the qemu binary supports these
commandline options.

The result of this will be:

1) The virtio device presented to the guest will have an extra bit set
in its PCI capabilities indicating that it can be used as a failover
backup device. The virtio guest driver will need to be equipped to do
something with this information - this is included in the Linux
virtio-net driver in kernel 4.18 and above (and also backported to
some older distro kernels). Unfortunately there is no way for libvirt
to learn whether or not the guest driver supports failover - if it
doesn't then the extra PCI capability will be ignored and the guest OS
will just see two independent devices. (NB: the current virtio guest
driver also requires that the MAC addresses of the two NICs match in
order to pair them into a bond).

2) When a migration is requested, QEMu will automatically unplug the
transient/hostdev NIC from the guest on the source host before
starting migration, and automatically re-plug a similar device after
restarting the guest CPUs on the destination host. While the transient
NIC is unplugged, all network traffic will go through the
persistent/virtio device, but when the hostdev NIC is plugged in, it
will get all the traffic. This means that in normal circumstances the
guest gets the performance advantage of vfio-assigned "real hardware"
networking, but it can still be migrated with the only downside being
a performance penalty (due to using an emulated NIC) during the
migration.

Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2020-01-29 15:33:29 -05:00
Laine Stump
fb0509d06a conf: parse/format <teaming> subelement of <interface>
The subelement <teaming> of <interface> devices is used to configure a
simple teaming association between two interfaces in a domain. Example:

  <interface type='bridge'>
    <source bridge='br0'/>
    <model type='virtio'/>
    <mac address='00:11:22:33:44:55'/>
    <alias name='ua-backup0'/>
    <teaming type='persistent'/>
  </interface>
  <interface type='hostdev'>
    <source>
      <address type='pci' bus='0x02' slot='0x10' function='0x4'/>
    </source>
    <mac address='00:11:22:33:44:55'/>
    <teaming type='transient' persistent='ua-backup0'/>
  </interface>

The interface with <teaming type='persistent'/> is assumed to always
be present, while the interface with type='transient' may be be
unplugged and later re-plugged; the persistent='blah' attribute (and
in the one currently available implementation, also the matching MAC
addresses) is what associates the two devices with each other. It is
up to the hypervisor and the guest network drivers to determine what
to do with this information.

Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2020-01-29 15:33:29 -05:00
Laine Stump
cad65f222f qemu: add capabilities flag for failover feature
Presence of the virtio-net-pci option called "failover" indicates
support in a qemu binary of a simplistic bonding of a virtio-net
device with another PCI device. This feature allows migration of
guests that have a network device assigned to a guest with VFIO, by
creating a network bond device in the guest consisting of the
VFIO-assigned device and a virtio-net-pci device, then temporarily
(and automatically) unplugging the VFIO net device prior to migration
(and hotplugging an equivalent device on the migration
destination). (The feature is called "failover" because the bond
device uses the vfio-pci netdev for normal guest networking, but
"fails over" to the virtio-net-pci netdev once the vfio-pci device is
unplugged for migration.)

Full functioning of the feature also requires support in the
virtio-net driver in the guest OS (since that is where the bond device
resides), but if the "failover" commandline option is present for the
virtio-net-pci device in qemu, at least the qemu part of the feature
is available, and libvirt can add the proper options to both the
virtio-net-pci and vfio-pci device commandlines to indicate qemu
should attempt doing the failover during migration.

This patch just adds the qemu capabilities flag "virtio-net.failover".

Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2020-01-29 15:33:29 -05:00