Libvirt's backup code has two modes:
1) push - where qemu actively writes the difference since the checkpoint
into the output file
2) pull - where we instruct qemu to expose a frozen disk state along
with a bitmap of blocks which changed since the checkpoint
For push mode qemu needs the temporary bitmap we use where we calculate
the actual changes to be present on the block node backing the disk.
For pull mode where we expose the bitmap via NBD qemu actually wants the
bitmap to be present for the exported block node which is the scratch
file.
Until now we've calculated the bitmap twice and installed it both to the
scratch file and to the disk node, but we don't need to since we know
when it's needed.
Pass in the 'pull' flag and decide where to install the bitmap according
to it and also when to register the bitmap name with the blockjob.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The NBD server used to export pull-mode backups doesn't have any other
form of client authentication on top of the TLS transport, so the only
way to authenticate clients is to verify their certificate.
Enable this option by defauilt when both 'backup_tls_x509_verify' and
'default_tls_x509_verify' were not configured.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1879477
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
The migration stream connection and also the NBD server for non-shared
storage migration don't have any other form of client authentication on
top of the TLS transport, so the only way to authenticate clients is to
verify their certificate.
Enable this option by defauilt when both 'migrate_tls_x509_verify' and
'default_tls_x509_verify' were not configured.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1879477
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Chardevs don't have any other form of client authentication on top of
the TLS transport, so the only way to authenticate clients is to verify
their certificate.
Enable this option by defauilt when both 'chardev_tls_x509_verify' and
'default_tls_x509_verify' were not configured.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1879477
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
If both "vnc_tls_x509_verify" and "default_tls_x509_verify" are missing
from the config file the client certificate validation is disabled. VNC
provides a layer of authentication so client certificate validation is
not strictly required.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Store whether "default_tls_x509_verify" was provided and enhance the
SET_TLS_VERIFY_DEFAULT macro so that indiviual users can provide their
own default if "default_tls_x509_verify" config option was not provided.
For now we keep setting it to 'false'.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Typecast the controller type variable to the appropriate type and add
the missing controller types for future extension.
Note that we currently allow only unplug of
VIR_DOMAIN_CONTROLLER_TYPE_SCSI thus the other controller types which
are not implemented return false now.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Iterate through hostdevs only when the controller type is
VIR_DOMAIN_CONTROLLER_TYPE_SCSI.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The tests which match the disk bus to the controller type were backwards
in this function. This meant that any disk bus type (such as
VIR_DOMAIN_DISK_BUS_SATA) would not skip the controller index comparison
even if the removed controller was of a different type.
Switch the internals to a switch statement with selects the controller
type in the first place and a proper type so that new controller types
are added in the future.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1870072
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
A bad merge while rebasing 74b2834333 caused the @event variable
to be defined twice, inside the 'cleanup' label, causing coverity
errors.
This code was originally moved outside of the label by commit
773c7c4361. Delete the unintended code in the 'cleanup'
label.
Fixes: 74b2834333
Reported-by: John Ferlan <jferlan@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Use g_autoptr() and remove the 'cleanup' label.
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Use VIR_AUTOCLOSE with 'fd' and delete the 'cleanup' label.
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Use g_autoptr() to deprecate the 'cleanup' label.
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Remove obsolete 'cleanup' labels after the changes from the
previous patch.
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Some labels became deprecated after the previous patches.
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Due to failures to unlink on previous rename/undefine we can already have
autolink etc files for the domain to be defined. Remove them.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Let's move objlist restoring to cleanup section so that we can handle failure
of actions between virDomainObjListAdd and virDomainDefSave. We are going
to add such actions in next patch.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
If domain name is changed since snapshot we need to update it to current in
config taken from snapshot.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
This is basically just saves checkpoints metadata on disk after name is changed
in memory as path to domain checkpoints directory depends on name. After that
old checkpoint directory is deleted with checkpoint metadata files.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
This is basically just saves snapshots metadata on disk after name is changed
in memory as path to domain snapshot directory depends on name. After that
old snapshot directory is deleted with snapshot metadata files.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
This patch also changes functionality a bit.
First if unlinking of old config file is failed we rollback and return error
previously and now we return success. I don't think this makes much difference.
I guess in both cases on libvirtd restart we have to deal with both new and old
config existing on disk with different names but same uuid.
Second if unlinking of old autolink is failed we rollback previously which
was not right as at this point we already unlink old config file. So this
is fixed now.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Going to cleanup label is mere return -1 thus let's just return
instead of goto to this label.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
We can simplify cleanup section by moving sending events to success path only
because only on sucess path events are not NULL.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
For example if saving config file with new name fails we send false undefine
event currently.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Some CPUs provide a way to read exact TSC frequency, while measuring it
is required on other CPUs. However, measuring is never exact and the
result may slightly differ across reboots. For this reason both Linux
kernel and QEMU recently started allowing for guests TSC frequency to
fall into +/- 250 ppm tolerance interval around the host TSC frequency.
Let's do the same to avoid unnecessary failures (esp. during migration)
in case the host frequency does not exactly match the frequency
configured in a domain XML.
https://bugzilla.redhat.com/show_bug.cgi?id=1839095
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
There are two types of vhostuser ports:
dpdkvhostuser - OVS creates the socket and QEMU connects to it
dpdkvhostuserclient - QEMU creates the socket and OVS connects to it
But of course ovs-vsctl syntax for fetching ifname is different.
So far, we've implemented the former. The lack of implementation
for the latter means that we are not detecting the interface name
and thus not reporting it in domain XML, or failing to get
interface statistics.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1767013
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Although the code in qemuProcessStartValidateTSC works as if the
timer frequency was already unsigned long long (by using an appropriate
temporary variable), the virDomainTimerDef structure actually defines
frequency as unsigned long, which is not guaranteed to be 64b.
Fixes support for frequencies higher than 2^32 - 1 on 32b systems.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
A qemu guest which has virtiofs config fails to start if the previous
starting failed because of invalid option or something.
That's because the virtiofsd isn't killed by virPidFileForceCleanupPath()
on the former failure because the pidfile was already removed by
virFileDeleteTree(priv->libDir) in qemuProcessStop(), so
virPidFileForceCleanupPath() just returned.
Move qemuExtDevicesStop() before virFileDeleteTree(priv->libDir) so that
virPidFileForceCleanupPath() can kill virtiofsd correctly.
For example of the reproduction:
# virsh start guest
error: Failed to start domain guest
error: internal error: process exited while connecting to monitor: qemu-system-x86_64: -foo: invalid option
... fix the option ...
# virsh start guest
error: Failed to start domain guest
error: Cannot open log file: '/var/log/libvirt/qemu/guest-fs0-virtiofsd.log': Device or resource busy
#
Signed-off-by: Masayoshi Mizuma <m.mizuma@jp.fujitsu.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
This function always returns zero, so it might as well be void.
Signed-off-by: Matt Coleman <matt@datto.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
This function always returns zero, so it might as well be void.
Signed-off-by: Matt Coleman <matt@datto.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
This function always returns zero, so it might as well be void.
Signed-off-by: Matt Coleman <matt@datto.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
This function always returns zero, so it might as well be void.
Signed-off-by: Matt Coleman <matt@datto.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
This function always returns zero, so it might as well be void.
Signed-off-by: Matt Coleman <matt@datto.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
The function only returns zero or aborts, so it might as well be void.
This has the added benefit of simplifying the code that calls it.
Signed-off-by: Matt Coleman <matt@datto.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
When using .path() for an argument to a python script meson will not
setup dependancies on the file. This means that changes to the generator
script will not trigger a rebiuld
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The enum constant names should all have a prefix that matches the enum
name. VIR_DOMAIN_CHECKPOINT_REDEFINE_VALIDATE was missing the "CREATE_"
part of the name prefix.
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Verify that the checkpoint requested by an incremental backup exists.
Unfortunately validating whether the checkpoint configuration actually
matches the disk may not be reasonably feasible as the disk may have
been renamed/snapshotted/etc. We still rely on bitmap presence.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Validate that the bitmaps are present when redefining a checkpoint.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
First one prepares and validates the definition, the second one actually
either updates an existing checkpoint or assigns definition for the new
one.
This will allow driver code to add extra validation between those
steps.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
If we don't have a consistent chain of bitmaps for the backup to proceed
we'd report VIR_ERR_INVALID_ARG error code, which makes it hard to
decide whether an incremental backup makes even sense.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
In testing code we don't properly populate the job sometimes. If it
isn't populated we should not touch it though in the migration cookie
code.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Include qemu_domain.h and qemu_domainjob.h as the types from those
headers are used by this header.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
qemuMigrationSrcCleanup uses qemuDomainObjDiscardAsyncJob currently. But
discard does not reduce jobs_queued counter so it leaks. Also discard does not
notify other threads that job condition is available. Discard does reset nested
job but nested job is not possible in this conditions.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
Don't hide our use of GHashTable behind our typedef. This will also
promote the use of glibs hash function directly.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Matt Coleman <matt@datto.com>
The simplest way to write tests is to check the output against expected
output, but we must ensure that the output is stable. We can use
virHashForEachSorted as a hash iterator to ensure stable ordering.
This patch fixes 3 instances of hash iteration which is tested in
various parts, including test output changes in appropriate places.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Matt Coleman <matt@datto.com>
All but one of the callers either use the list in arbitrary order or
sorted by key. Rewrite the function so that it supports sorting by key
natively and make it return the element count. This in turn allows to
rewrite the only caller to sort by value internally.
This allows to remove multiple sorting functions which were sorting by
key and the function will be also later reused for some hash operations
internally.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Matt Coleman <matt@datto.com>
The remember owner feature uses XATTRs to store original
seclabels. But that means we don't want a regular user to be able
to change what we stored and thus trick us into setting different
seclabel. Therefore, we use namespaces that are reserved to
CAP_SYS_ADMIN only. Such namespaces exist on Linux and FreeBSD.
That also means, that the whole feature is enabled only for
qemu:///system. Now, while the secdriver code is capable of
dealing with XATTRs being unsupported (it has to, not all
filesystems support them) if the feature is enabled users will
get an harmless error message in the logs and the feature
disables itself.
Since we have virSecurityXATTRNamespaceDefined() we can use it to
make a wiser decision on the default state of the feature.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Roman Bolshakov <r.bolshakov@yadro.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Add logic to validate and then pass through 'fmode' and 'dmode' to the
QEMU call.
Signed-off-by: Brian Turek <brian.turek@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
The QEMU 9pfs 'fmode' and 'dmode' options have existed since QEMU 2.10.
Probe QEMU's command line set to check whether these options are
available, and if yes, enable this new QEMU_CAPS_FSDEV_CREATEMODE
capability on libvirt side.
Signed-off-by: Brian Turek <brian.turek@gmail.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
All other helper processes are moved to cgroup with QEMU emulator
thread as we keep the root VM cgroup without any processes. This
assumption is validated in qemuRestoreCgroupState() which is called
when libvirtd is restarted and reconnected to all running VMs.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
In both cases priv->cgroup will always be NULL because it is called
before the QEMU process is started and cgroups are configured.
In qemuProcessLaunch() the call order is following:
qemuExtDevicesStart()
...
virCommandRun()
...
qemuSetupCgroup()
where qemuDBusStart() is called from qemuExtDevicesStart() but we
cgroups are created in qemuSetupCgroup().
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
After converting all DIR* to g_autoptr(DIR), many cleanup: labels
ended up just having "return ret", and every place that set ret would
just immediately goto cleanup. Remove the cleanup label and its
return, and just return the set value immediately, thus eliminating
the need for the return variable itself.
Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
All of these conversions are trivial - VIR_DIR_CLOSE() (aka
virDirClose()) is called only once on the DIR*, and it happens just
before going out of scope.
Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
This will make it easier to review upcoming patches that use g_autoptr
to auto-close all DIRs.
Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Now that qemu stabilized it's interface and we've switched to the new
design we can re-enable use of 'block-export-add'
This reverts commit b87cfc957f
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
qemu decided to modify the arguments of 'block-export-add' to include an
array of bitmaps rather than a single bitmap.
Since we've added the code prior to qemu setting the interface in stone
and thus it will be changed incompatibly and we already have tests for
the new interface we need to update the code and qemu capabilities data
at the same time.
Use a array of bitmaps as the 'bitmaps' argument instead of 'bitmap' and
bump qemu capabilities for the upcoming 5.2.0 release to
v5.1.0-2827-g2c6605389c
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
We use the capability to switch to using 'block-export-add' in the
upcoming qemu release instead of the at the same time deprecated
'nbd-server-add'.
Unfortunately qemu wants to change the interface of 'block-export-add'
before the release. Since we've tried to stay up to date and added the
code before it was written in stone, we need to disable the use of the
new interface for the upcoming libvirt release so that we don't have a
version of libvirt which would not work with the upcoming qemu version.
Remove the detection of 'block-export-add' until we are more sure how
the qemu interface will look.
This patch partially reverts commit adb9f7123a
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Currently all errors from qemuInterfacePrepareSlirp() are completely
ignored by the callers. The intention is that missing qemu-slirp binary
should cause the caller to fallback to the built-in slirp impl.
Many of the possible errors though should indeed be considered fatal.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Eliminate cleanup code by using g_autofree.
Signed-off-by: Yi Li <yili@winhong.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
In recent commit v6.8.0-135-g518be41aaa the formatting of NBD
into migration cookie was moved into a separate function and with
it it was switched from direct printing into the output buffer to
virXMLFormatElement(). But there was a typo. The
virXMLFormatElement() accepts two buffers on input, one for
element attributes and another for child elements. Well, the line
that was supposed to add NBD port into the attributes buffer
printed the attribute directly into the output buffer which
produced this mangled XML:
<qemu-migration>
port='49153'<nbd>
<disk target='vda' capacity='8589934592'/>
<disk target='vdb' capacity='12746752000'/>
</nbd>
</qemu-migration>
Changing the incriminated line to print into the attributes
buffer fixes the problem.
Fixes: 518be41aaa
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
In one of recent patches the way that we start NBD server for
incoming migration was reworked (v6.8.0-rc1~298). A new boolean
was introduced that tracks whether the NBD server was started so
that we don't start it twice nor record in the port in the port
allocator twice. Well, this idea is good, but in the
implementation the boolean is never set, so we are reserving the
port twice and would be starting the NBD server twice too if it
wasn't for port reservation fail.
Fixes: e74d627bb3
Reported-by: Vjaceslavs Klimovs <vklimovs@gmail.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
In qemuDomainLogContextNew() the domain log file is opened.
Twice, the first time for writing, and the second time for
reading (if required by caller). When opening the log file for
reading a mode is provided. This doesn't do much harm, but is
unnecessary. Drop the mode.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Trivial fix to improve readability by combining these into a compound
conditional.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
Coverity reported a potential resource leak. While it's probably not
a real-world scenario, the code could technically jump to cleanup
between the time that vdpafd is opened and when it is used. Ensure that
it gets cleaned up in that case.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
Since QEMU 5.2 (commit-77b285f7f6), QEMU supports 'memory failure'
event, posts event to monitor if hitting a hardware memory error.
Fully support this feature for QEMU.
Test with commit 'libvirt: support memory failure event', build a
little complex environment(nested KVM):
1, install newly built libvirt in L1, and start a L2 vm. run command
in L1:
~# virsh event l2 --event memory-failure
2, run command in L0 to inject MCE to L1:
~# virsh qemu-monitor-command l1 --hmp mce 0 9 0xbd000000000000c0 0xd 0x62000000 0x8c
Test result in l1(recipient hypervisor case):
event 'memory-failure' for domain l2:
recipient: hypervisor
action: ignore
flags:
action required: 0
recursive: 0
Test result in l1(recipient guest case):
event 'memory-failure' for domain l2:
recipient: guest
action: inject
flags:
action required: 0
recursive: 0
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
All users of virHashTable pass strings as the name/key of the entry.
Make this an official requirement by turning the variables to 'const
char *'.
For any other case it's better to use glib's GHashTable.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
It doesn't make much sense to configure the bucket count in the hash
table for each case specifically. Replace all calls of virHashCreate
with virHashNew which has a pre-set size and remove virHashCreate
completely.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
virHashCreate will be removed in upcoming patches. This change has an
impact on ordering of the blockjob entries in one of the status XML->XML
tests.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Use of the -enable-fips option is being deprecated in QEMU >= 5.2.0. If
FIPS compliance is required, QEMU must be built with libcrypt which will
unconditionally enforce it.
Thus there is no need for libvirt to pass -enable-fips to modern QEMU.
Unfortunately there was never any way to probe for -enable-fips in the
first instance, it was enabled by libvirt based on version number
originally, and then later unconditionally enabled when libvirt dropped
support for older QEMU. Similarly we now use a version number check to
decide when to stop passing -enable-fips.
Note that the qemu-5.2 capabilities are currently from the pre-release
version and will be updated once qemu-5.2 is released.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
libvirt can retrieve traffic stats for emulated interfaces that are
backed by tap or macvtap devices, but this information wasn't
available for hostdev interfaces (those that are implemented by
assigning an SR-IOV VF device to a guest using vfio):
#virsh domifstat instance --interface=52:54:00:2d:b2:35
error: Failed to get interface stats instance 52:54:00:2d:b2:35
error: internal error: Interface name not provided
For some SR-IOV VF devices this information is available via the
netlink VFINFO_LIST request/response, and that is what this patch uses
to implement stats retrieval for VF. Not that this is dependent on
support in the PF driver - for example, the Mellanox ConnectX-4 Lx
(mlx5) driver reports usable stats, while Intel 82599 (ixgbe) and
82576 (igb) just report all stats as 0. (this is the same result as
"ip -s link show").
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
Reviewed-by: Laine Stump <laine@redhat.com>
By using the new qemu monitor functions to handle passing and removing
file descriptors, we can support hotplug of vdpa devices.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
add-fd, remove-fd, and query-fdsets provide functionality that can be
used for passing fds to qemu and closing fdsets that are no longer
necessary.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
Enable <interface type='vdpa'> for qemu domains. This provides basic
support and does not support hotplug or migration.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
Recent versions of qemu added the -netdev vhost-vdpa device. This
capability allows libvirt to know whether this is supported.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
This patch adds new schema and adds support for parsing and formatting
domain configurations that include vdpa devices.
vDPA network devices allow high-performance networking in a virtual
machine by providing a wire-speed data path. These devices require a
vendor-specific host driver but the data path follows the virtio
specification.
When a device on the host is bound to an appropriate vendor-specific
driver, it will create a chardev on the host at e.g. /dev/vhost-vdpa-0.
That chardev path can then be used to define a new interface with
type='vdpa'.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
My code movement changed the type of ifaces_ret from
virDomainInterfacePtr * to virDomainInterfacePtr **,
but failed to adjust the condition or dereference the
array correctly.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Fixes: 6ddb1f803e
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
SCSI hostdev setup requires querying the host os for the actual path of
the configured hostdev. This was historically done in the command line
formatter. Our new approach is to split out this part into
'qemuProcessPrepareHost' which is designed to be skipped in tests.
Refactor the hostdev code to use this new semantics, and add appropriate
handlers filling in the data for tests and the qemuConnectDomainXMLToNative
users.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
qemuBuildHostdevSCSIAttachPrepare is supposed to prepare the data
structure used for attaching the hostdev not preparing the hostdev
definition itself. Move the corresponding bits to qemuDomainPrepareHostdev
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Host preparation steps which are deliberately skipped when
pretend-creating a commandline are normally executed after VM object
preparation. In the test code we are faking some of the host
preparation steps, but we were doing that prior to the call to
qemuProcessPrepareDomain embedded in qemuProcessCreatePretendCmd.
By splitting up qemuProcessCreatePretendCmd into two functions we can
ensure that the ordering of the prepare steps stays consistent.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Switch to the new QMP command once it becomes available. Since the code
was refactored to have just one central location to do this we can
contain the ugly bits to just this one function.
Since we now use the replacement for 'nbd-server-add' mark the test case
as being OK with removal of the command.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Add the monitor code, corresponding generator of properties for NBD and
tests validating it against the schema.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
The 'block-export-add' QMP command is a replacement for 'nbd-server-add'
and will allow greater flexibility. Add a capability so that we can
switch to it.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Centralize the logic deciding which arguments to use when exporting a
block backend via NBD to a single place so that it can be centrally
fixed in upcoming commits to support the new export method via
'block-export-add'.
Additionally this allows simplification of the caller from migration as
the logic deciding which arguments to use is extracted too.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
These XML attributes have been mandatory since the introduction of SEV
support to libvirt. This design decision was based on QEMU's
requirement for these to be mandatory for migration purposes, as
differences in these values across platforms must result in the
pre-migration checks failing (not that migration with SEV works at the
time of this patch).
This patch enables autofill of these attributes right before launching
QEMU and thus updating the live XML.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Checks such as this one should be done at domain def validation time,
not before starting the QEMU process.
As for this change, existing domains will see some QEMU error when
starting as opposed to a libvirt error that this QEMU binary doesn't
support SEV, but that's okay, we never guaranteed error messages to
remain the same.
Signed-off-by: Erik Skultety <eskultet@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Rename the function to qemuValidateDomainVCpuTopology() to reflect
what it is currently doing as well.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
All but VIR_CPU_MODE_HOST_MODEL were moved. 'host_model' mode
has nuances that forbid the verification to be moved to parse
time.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
We have a lot of "if (usingVirtio)" checks being done while
constructing the NIC command line. Let's put all of them in
a single "if".
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
A few tweaks were made during the move:
- the error messages were changed to mention 'sata controller'
instead of 'ide controller';
- a check for address type 'drive' was added like it is done
with other bus types. The error message of qemuxml2argdata was
updated to reflect that now, instead of erroring it out from the
common code in virDomainDiskDefValidate(), we're failing earlier
with a different error message.
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
In fee8a61d29 a new attribute to <memballoon/> was introduced:
free-page-reporting. We don't really like hyphens in attribute
names. Use camelCase instead.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
These variables seem to be left over from a previous refactoring and
they don't add anything to the code.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Laine Stump <laine@redhat.com>
...if a machine memory-backend using shared memory is configured for
the guest. This is especially important for QEMU machine types that
don't have NUMA but virtiofs support.
An example snippet:
<domain type='kvm'>
<name>test</name>
<memory unit='KiB'>2097152</memory>
<memoryBacking>
<access mode='shared'/>
</memoryBacking>
<devices>
<filesystem type='mount' accessmode='passthrough'>
<driver type='virtiofs'/>
<source dir='/tmp/test'/>
<target dir='coffee'/>
</filesystem>
...
</devices>
...
</domain>
and the corresponding QEMU command line:
/usr/bin/qemu-system-s390x \
-machine s390-ccw-virtio-5.2,memory-backend=s390.ram \
-m 2048 \
-object
memory-backend-file,id=s390.ram,mem-path=/var/lib/libvirt/qemu/ram/46-test/s390.ram,share=yes,size=2147483648 \
...
Signed-off-by: Marc Hartmayer <mhartmay@linux.ibm.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
This patch enables the free-page-reporting in qemu.
Signed-off-by: Nico Pache <npache@redhat.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
This patch will introduce the free-page-reporting feature capabilities
that are in qemu 5.1
Signed-off-by: Nico Pache <npache@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
By default, pfifo_fast queueing discipline (qdisc) is set on
newly created interfaces (including TAPs). This qdisc has three
queues and packets that want to be sent through given NIC are
placed into one of the queues based on TOS field. Queues are then
emptied based on their priority allowing interactive sessions
stay interactive whilst something else is downloading a large
file.
Obviously, this means that kernel has to be involved and some
locking has to happen (when placing packets into queues). If
virtualization is taken into account then the above algorithm
happens twice - once in the guest and the second time in the
host.
This is arguably not optimal as it burns host CPU cycles
needlessly. Guest already made it choice and sent packets in the
order it wants.
To resolve this, Linux kernel offers 'noqueue' qdisc which can be
applied on virtual interfaces and in fact for 'lo' it is by
default:
lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue
Set it for other TAP devices we create for domains too. With this
change I was able to squeeze 1Mbps more from a macvtap attached
to a guest and to my 1Gbps LAN (as measured by iperf3).
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1329644
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
If storage migration is requested, and the destination storage does
not exist on the remote host, qemu's migration support will call
into the libvirt storage driver to precreate the destination storage.
The storage driver virConnectPtr is opened too early though, adding
an unnecessary dependency on the storage driver for several cases
that don't require it. This currently requires kubevirt to install
the storage driver even though they aren't actually using it.
Push the virGetConnectStorage calls to right before the cases they are
actually needed.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
As preparation for g_autoptr() we need to change the function to take
only virCgroupPtr.
Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
"cpu_map.xml" was moved to a directory "cpu_map" and split up into
several files.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
The original descirption for *_tls_x509_verify is a little misleading
by saying that "Enabling this option will reject any client who does
not have a ca-cert.pem certificate".
Signed-off-by: Fangge Jin <fjin@redhat.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
In 88957116c9 I've switched to -machine memory-backend=ID and
-object memory-backend-* because QEMU is obsoleting -mem-path
and -mem-prealloc. However, what I did not foresee was that using
-machine memory-backend in combination with -numa is not allowed
in QEMU. This was reported upstream and fortunately not released
yet.
The problem is that if domain has NUMA nodes then we will
generate memory-backend-* objects for NUMA nodes (because if QEMU
is new enough to expose default RAM ID it also supports -numa
memdev=) and adding non-NUMA memory backend is wrong.
Reported-by: Masayoshi Mizuma <msys.mizuma@gmail.com>
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
For readability, and to ensure we do allocation when
returning 0.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
We only care about the first part of the 'ifname' string,
splitting it is overkill.
Instead, just replace the ':' with a '\0' in a copy of the string.
This reduces the count of the varaibles containing some form
of the interface name to two.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
The error check for ValueObjectGet("return") is redundant,
qemuAgentCommand already checked for us that the reply contains
a "return" object.
It does not guarantee, that it is an array.
Use virJSONValueObjectGetArray that combines getting the object
with checking for its type and return the more helpful of
the two error messages.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
Convert one interface from the "return" array returned by
"guest-network-get-interfaces" to virDomainInterface.
Due to the functionality of squashing interface aliases together,
this is not a pure function - it either:
* Adds the interface to ifaces_ret, incrementing ifaces_count
and adds a pointer to it into the ifaces_store hash table.
* Adds the additional IP addresses from the interface alias
to the existing interface entry, found through the hash table.
This does not increment ifaces_count or extend the array.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
We have a local 'iface' variable that contains the same value
eventually. Initialize it early instead of indexing two more
times.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
We're passing 'ifaces_count' to virHashCreate as the initial
hash table size just after we've initialized it to zero.
This translates to a default of 256 inside virHashCreateFull.
Instead of this obfuscation, use virHashNew (default of 32),
to make it obvious we don't care about the initial hash size.
Also remove the error handling, since neither of the functions
return any errors after switching to g_new0.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
This lets us conveniently reduce its scope to the outer loop.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
qemuAgentGetInterfaceOneAddress returns exactly one address
for every iteration of the loop (and we error out if not).
Instead of expanding the addrs by one on every iteration,
do it upfront since we know how many times the loop will
execute.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
For both 'ip_addr_arr' and 'ret_array', we:
1) already checked that they are arrays
2) only iterate up to the array size
Remove the duplicate checks.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
A function that takes one entry from the "ip-addresses" array
returned by "guest-network-get-interfaces" and converts it
into virDomainIPAddress.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
virJSONValueObjectGetArray returns NULL if the object with
the supplied key is not an array.
Calling virJSONValueIsArray right after is redundant.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Neal Gompa <ngompa13@gmail.com>
This adds a new value to virConnectCompareCPUFlags,
"VIR_CONNECT_CPU_VALIDATE_XML", that governs XML document validation in
virCPUDefParseXML.
In src/conf/cpu_conf.c, include configmake.h for PKGDATADIR and
virfile.h for virFileFindResource.
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
I left in a 'return' or 'goto cleanup' in a few places
where I did the conversion manually.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reported-by: John Ferlan <jferlan@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
Move them to separate conditions to reduce churn
in following patches.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
Previous refactors allow us to plainly replace all VIR_FREE by g_free to
finish the modernization of the file.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Restructure the control-flow a bit using an temporary variable to avoid
the need to use VIR_FREE.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Use modern allocators, automatic memory feeing, and decrease the scope
of some variables to remove the 'cleanup' label.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Use modern allocators, automatic memory feeing, and decrease the scope
of some variables to remove the 'error' and 'cleanup' labels.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Use modern allocators, automatic memory feeing, and decrease the scope
of some variables to remove the 'error' label.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Use automatic memory freeing to get rid of the 'error' label. Since the
'tmp' variable was used only in one instance, rename it appropriately.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Use modern memory handling approach to simplify the code.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Use modern memory handling approach to simplify the code.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Register the the cleanup functions for 'qemuMigrationCookieGraphics',
'qemuMigrationCookieNetwork', 'qemuMigrationCookieNBD', and
'qemuMigrationCookieCaps'.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Switch to automatic memory cleaning, use g_new0 for allocation and get
rid of the 'error' label.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Move the code into 'qemuMigrationCookieNBDXMLFormat' and use modern XML
formatting code patterns.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Use 'virXMLFormatElement' both for formating the whole <network> element
but also for formatting the <interface> subelements. This alows to
remove the crazy logic which was determining which element was already
formatted.
Additional simplification is achieved by switching to skipping the loop
using 'continue' rather than putting everything in a giant block.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Switch to the two buffer approach to simplify the logic for terminating
the element.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Now it only returns -1 so we can do that directly.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Most of our functions report errors so there's no need to mention it
here again.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Make sure that 'virXPathNodeSet' returns '1' as the only expected value
rather than relying on the fact that the previous check for the number
of elements ensures success of the subsequent call.
The error message no longer mentions the number of <domain> elements in
the cookie, but this is a very unlikely internal error anyways.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Don't reuse 'tmp' over and over, but switch to single use automaticaly
freed variables instead.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Move the code into 'qemuMigrationCookieXMLParseMandatoryFeatures' to
simplify 'qemuMigrationCookieXMLParse'.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
After recent refactors the function can be refactored to remove the
'cleanup' label by using autoptr for the 'map' variable.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
There are only 3 places using the function. Two can use virBitmapNewCopy
directly. In case of the qemu capabilities code we need to free the old
bitmap first.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Instead of saving the interesting pieces of each existing NetDef,
clearing it, and then copying back the saved pieces after setting the
type to ethernet, just create a new NetDef, copy in the interesting
bits, and replace the old one. (The end game is to eliminate
virDomainNetDefClear() completely, since this is the only real use)
Signed-off-by: Laine Stump <laine@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
So far, Libvirt configures memory-backend-* for memory hotplug,
possibly NUMA nodes and in a few other cases. This patch
switches to constructing the memory-backend-* command line for
all cases. To keep ability to migrate guests a little hack is
used: the ID of the object is set to the one that QEMU uses
internally anyways. These IDs are stable (first started to appear
somewhere around v0.13.0-rc0~96) and can't change.
In fact, this patch does exactly what QEMU does internally. The
reason for moving the logic into Libvirt is that QEMU wants to
deprecate the old style of specifying memory.
So far, only x84_64 test cases are changed, because tests for
other architectures use older capabilities, which still lack the
QEMU_CAPS_MACHINE_MEMORY_BACKEND capability and they don't report
the RAM ID.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1836043
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
The machine structure has another (optional) attribute:
default-ram-id, which specifies the alias of the default RAM
object. While the alias is private, it can never change in order
to not break migration. QEMU uses the alias when allocating
regular, not NUMA memory. In order to switch to new command line
and maintain migration, save this ID.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
The objects at @def and @mem pointers are only read and not
written. Make the arguments const to make that explicit.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
If a domain was using hugepages through memory-backend-file or
via -mem-path, we would turn prealloc on. But we are not doing
that for memory-backend-memfd. Fix this, because we need QEMU to
fully allocate hugepages.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
If user specifies immediate memory allocation in the domain XML,
they want QEMU to fully allocate its memory. And if the memory
was allocated using plain '-m' then we would honour it. But, if a
memory backend is used, then we don't set the prealloc attribute
of the backend.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
All three memory backends (-file, -ram and -memfd) have .prealloc
attribute. Since we are setting it only for -file, the
corresponding code lives only under if() that handles that
specific backend. But in near future we will want to set the
attribute for other backends too. Therefore, move the
corresponding code outside of the if().
This causes some .argv files to be changed, but the only change
happening there is move of the attribute (best viewed with:
'git show --color-words=.').
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
This originally started as bug 1595525 in which namespaces and
libusb used in QEMU were not playing nicely with each other. The
problem was that libusb built a cache of USB devices it saw
(which was a very limited set because of the namespace) and then
expected to receive udev events to keep the cache in sync. But
those udev events didn't come because on hotplug when we mknod()
devices in the namespace no udev event is generated. And what is
worse - libusb failed to open a device that wasn't in the cache.
Without going further into what the problem was, libusb added a
new API for opening USB devices that avoids using cache which
QEMU incorporated and exposes under "hostdevice" attribute.
What is even nicer is that QEMU uses qemu_open() for path
provided in the attribute and thus FD passing could be used.
Except qemu_open() expects so called FD sets instead of `getfd'
and these are not implemented in libvirt, yet.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1877218
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
This capability tracks whether "usb-host" device has "hostdevice"
attribute. This attribute allows us to specify full path to the
USB device ("/dev/bus/usb/$bus/$dev") but more importantly, since
QEMU uses qemu_open() for this attribute it allows us to pass
pre-opened FD and have QEMU not bother with opening the file at
all.
The attribute was added in v5.1.0-rc0~71^2~1 QEMU commit.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
