Add the capability for QEMU's packed virtqueues for virtio that supposedly have
better cache utilization and performance compared to the default split queues.
Reviewed-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NBoris Fiuczynski <fiuczy@linux.ibm.com>
Signed-off-by: NBjoern Walk <bwalk@linux.ibm.com>
Signed-off-by: NJán Tomko <jtomko@redhat.com>

Update s390x capabilities for QEMU 4.2 with the actual GA version for
QEMU and on the latest z15 machine.

This picks up the new blockdev capability, so we need to refresh a bunch
of test cases as well.
Reviewed-by: NBoris Fiuczynski <fiuczy@linux.ibm.com>
Acked-by: NChristian Borntraeger <borntraeger@de.ibm.com>
Reviewed-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>
Signed-off-by: NJán Tomko <jtomko@redhat.com>

The virDomainGenerateMachineName() function doesn't belong in
src/conf/ really, because it has nothing to do with domain XML
parsing. It landed there because of lack of better place in the
past. But now that we have src/hypervisor/ the function should
live there. At the same time, the function name is changed to
match new location.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NAndrea Bolognani <abologna@redhat.com>
Reviewed-by: NDaniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: NDaniel P. Berrangé <berrange@redhat.com>

Initially introduced in v3.10.0-rc1~172.

When generating a path for memory-backend-file or -mem-path, qemu
driver will use the following pattern:

  $memoryBackingDir/libvirt/qemu/$id-$shortName

where $memoryBackingDir defaults to /var/lib/libvirt/qemu/ram but
can be overridden in qemu.conf. Anyway, the "/libvirt/qemu/" part
looks redundant, because it's already contained in the default,
or creates unnecessary nesting if overridden in qemu.conf.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NAndrea Bolognani <abologna@redhat.com>

The virQEMUDriverConfigNew() accepts path to root directory for
embed mode as an argument. If the argument is not NULL it uses
the passed value as prefix for some internal paths (e.g.
cfg->libDir). If it is NULL though, it looks if the other
argument, @privileged is true or false and generates internal
paths accordingly. But when calling the function from the test
suite, instead of passing NULL for @root, an empty string is
passed. Fortunately, this doesn't create a problem because in
both problematic cases the generated paths are "fixed" to point
somewhere into build dir or the code which is tested doesn't
access them.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NAndrea Bolognani <abologna@redhat.com>
Reviewed-by: NDaniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: NDaniel P. Berrangé <berrange@redhat.com>

This option prevents misbehaviours on guest if a qemu 9pfs export
contains multiple devices, due to the potential file ID collisions
this otherwise may cause.
Signed-off-by: NChristian Schoenebeck <qemu_oss@crudebyte.com>
Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Introduce new 'multidevs' option for filesystem.

  <filesystem type='mount' accessmode='mapped' multidevs='remap'>
    <source dir='/path'/>
    <target dir='mount_tag'>
  </filesystem>

This option prevents misbehaviours on guest if a qemu 9pfs export
contains multiple devices, due to the potential file ID collisions
this otherwise may cause.
Signed-off-by: NChristian Schoenebeck <qemu_oss@crudebyte.com>
Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The QEMU 9pfs 'multidevs' option exists since QEMU 4.2. Probe QEMU's
command line set though to check whether this option is really
available, and if yes enable this new QEMU_CAPS_FSDEV_MULTIDEVS
capability on libvirt side.
Signed-off-by: NChristian Schoenebeck <qemu_oss@crudebyte.com>
Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

This function will remain public due to its usage in qemublocktest.c
even after moving qemuDomainDeviceDefValidate(). The position of its
header in qemu_validate.h is no accident.
Signed-off-by: NDaniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>
Signed-off-by: NJán Tomko <jtomko@redhat.com>

Now that v5.0.0-rc1 was tagged, update the capabilities to make sure
that everything works as expected.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

QEMU now formalized the arguments of netdev-add in the schema, so we
must use a real type to pass the schema validation once the schema is
updated.

The 'user' variant doesn't have any other mandatory fields.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Trying to squeeze the validator output into the monitor reply message
doesn't make sense and doesn't work well as it's not well formed JSON:

54) qemuMonitorJSONAddNetdev                                          ... libvirt:  error : internal error: cannot parse json { "error":  { "desc": "failed to validate arguments of 'netdev_add' against QAPI schema: {
   ERROR: variant 'test' for discriminator 'type' not found
",    "class": "UnexpectedCommand" } }: lexical error: invalid character inside string.
          ev_add' against QAPI schema: {    ERROR: variant 'test' for
                     (right here) ------^
FAILED

Output it to stderr if requested and just note that schema validation
failed in the error message:

54) qemuMonitorJSONAddNetdev                                          ...
failed to validate arguments of 'netdev_add' against QAPI schema
args:
{
  "id": "net0",
  "type": "test"
}

validator output:
 {
   ERROR: variant 'test' for discriminator 'type' not found

libvirt: QEMU Driver error : internal error: unable to execute QEMU command 'netdev_add': failed to validate arguments of 'netdev_add' against QAPI schema (to see debug output use VIR_TEST_DEBUG=2)
FAILED
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The test name is concatenated from a prefix with the test name, but no
space was added so the output looked wrong.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

When moving the formatting of this attributes from -drive
to -device, the QEMU_CAPS_USB_STORAGE_WERROR capability
was used, because usb-storage was the last device to gain
this capability.

However this lead to the assumption that QEMU binaries
without the usb-storage device do not support this,
leading to breakage on s390x with blockdev.

Fixes: bb4f3543
https://bugzilla.redhat.com/show_bug.cgi?id=1819250Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NPeter Krempa <pkrempa@redhat.com>

Detect the werror property on SCSI and virtio disks.
But clear it if the QEMU supports usb-storage device without it
also supporting this option for usb-storage.
Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NPeter Krempa <pkrempa@redhat.com>

To demonstrate the move of these attributes from -drive to -device.
Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NPeter Krempa <pkrempa@redhat.com>

For http/https URIs we need to preserve the query part as it may be
important to refer to the image.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

If the storage source has the query part set, format it in the output.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Add a new attribute for holding the query part for http(s) disks.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The quotes are forbidden only inside the value, but the value itself may
be enclosed in quotes. Fix the RNG schema and validator and add a test
case.

https://bugzilla.redhat.com/show_bug.cgi?id=1804750Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

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

Host-model CPU definitions (and domain capabilities) will use the
original CPU models (without noTSX in their name) and explicitly disable
hle and rtm features. This way domains with host-model CPUs will be
migratable even to older versions of libvirt which do not support the
noTSX model variants.

The new models will be advertised in host capabilities and they may
be used explicitly with custom CPUs.
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>
Reviewed-by: NChristian Ehrhardt <christian.ehrhardt@canonical.com>
Tested-by: NChristian Ehrhardt <christian.ehrhardt@canonical.com>

One of the mitigation methods for TAA[1] is to disable TSX
support on the host system.  Linux added a mechanism to disable
TSX globally through the kernel command line, and many Linux
distributions now default to tsx=off.  This makes existing CPU
models that have HLE and RTM enabled not usable anymore.

Add new versions of all CPU models that have the HLE and RTM
features enabled, that can be used when TSX is disabled in the
host system.

On systems disabling the features without those types defined
in cpu-maps users end up without modern CPU types in the list
of usable CPUs to use in the likes of virsh domcapabilities
or tools higher in the stack like virt-manager.

This adds:
-Cascadelake-Server-noTSX
-Icelake-Client-noTSX
-Icelake-Server-noTSX
-Skylake-Server-noTSX-IBRS
-Skylake-Client-noTSX-IBRS

Introduced in QEMU by commit v4.2.0-rc2-3-g9ab2237f19 (function)
                  and commit v4.2.0-rc2-4-g02fa60d101 (names)

References:

    [1] TAA, TSX asynchronous Abort:
        https://software.intel.com/security-software-guidance/insights/deep-dive-intel-transactional-synchronization-extensions-intel-tsx-asynchronous-abort
        https://www.kernel.org/doc/html/latest/admin-guide/hw-vuln/tsx_async_abort.html

Fixes: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1853200Signed-off-by: NChristian Ehrhardt <christian.ehrhardt@canonical.com>
Message-Id: <20200310104806.2723-2-christian.ehrhardt@canonical.com>
Reviewed-by: NJiri Denemark <jdenemar@redhat.com>

Use qemuAgentCommandFull so that callers of qemuAgentGetFSInfo can
suppress error reports if the function is not supported by the guest
agent.

Since this patch removes the last use of
qemuAgentErrorCommandUnsupported the whole function is deleted as well.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Use qemuAgentCommandFull so that callers of qemuAgentGetTimezone can
suppress error reports if the function is not supported by the guest
agent.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Use qemuAgentCommandFull so that callers of qemuAgentGetOSInfo can
suppress error reports if the function is not supported by the guest
agent.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Use qemuAgentCommandFull so that callers of qemuAgentGetUsers can
suppress error reports if the function is not supported by the guest
agent.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Unfortunately, advisory record locking lose the lock if any fd refering
to the file is closed. There doesn't seem to be a way to preserve the
lock atomically. We could eventually retake the lock if low pidfilefd
is required.

This fixes processes being leaked, as they are not killed in
virPidFileForceCleanupPath() if the lock can be taken. Here also, we may
consider this is not good enough, as a process may leak by simply
closing the pidfilefd.

Fixes commit d146105f ("virCommand:
Actually acquire pidfile instead of just writing it")
Signed-off-by: NMarc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

Our virCommand module allows us to set a pidfile for commands we
want to spawn. The caller constructs the string of pidfile path
and then uses virCommandSetPidFile() to tell the module to write
the pidfile once the command is ran. This usually works, but has
two flaws:

1) the child process does not hold the pidfile open & locked.
Therefore, the caller (or anybody else) can't use our fancy
virPidFileForceCleanupPath() function to kill the command
afterwards. Also, for everybody else on the system it's
needlessly harder to check if the pid from the pidfile is still
alive or not.

2) if the caller ever makes a mistake and passes the same pidfile
path for two different commands, the start of the second command
will overwrite the pidfile even though the first command might
still be running.

NOTE that this temporarily renders some command spawning
unusable, specifically those code patterns where both
virCommandSetPidFile() is used together with instructing spawned
command to acquire pidfile itself. Fortunately, there is only one
occurrence of such pattern and it is in
qemuProcessStartManagedPRDaemon(). This is fixed in next commit.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NMarc-André Lureau <marcandre.lureau@redhat.com>

As a part of c799d150 I've introduced a test case that
tests whether passing error object between processes works. The
test spawns a child which reports a system error, parent process
then reads the error and compares with expected output. Problem
with this approach is that error message contains stringified
errno which is not portable. FreeBSD has generally different
messages than Linux. Therefore, use g_strerror() to do the errno
to string translation for us.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NDaniel P. Berrangé <berrange@redhat.com>

Our code allows snapshots of NVMe based disks which means we create
overlay file with a 'json:{}' pseudo-uri refering to the NVME device.
Our parser code doesn't handle them though. Add the parser and test it
via the XML->json->XML round-trip and reference data.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

Format cookies into the backing store string without encryption as they
will not be visible on the command line when formatting a 'target' only
string. In cases when cookies or other options are used we must use the
JSON format rather than pure URI.

Add tests to validate the scenario.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

QEMU requires an extra wrapper object where only the "file" member is
populated. This is basically a placeholder for establishing the format
layer. We did the same in qemuDiskSourceGetProps for the old-school
JSON usage with -drive but forgot to adopt this for -blockdev.

https://bugzilla.redhat.com/show_bug.cgi?id=1804617Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

qemublocktest showed that we don't add the "fat:" prefix for directory
storage when formatting the backing store string. While it's unlikely to
be used it's simple enough to actually implement the support rather than
trying to forbid it.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

With -blockdev libvirt provides the string which is recorded  as
'backing store' property of an image to qemu. Add testing for
qemuBlockGetBackingStoreString which generates these strings as there's
logic which determines which format to use.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

We had two non-syncrhonized arrays holding the individual data. This was
a lazy way to do it when I was adding new tests recently. Since it's
hard to extend with new data to test refactor the storage of test data
to use a new struct where all per-image data are kept and can be
extended easily.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

The function was misplaced. Group it together with other helper
functions for testing disk XML to qemu JSON props conversion.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

Use automatic variable clearing and remove the cleanup sections of
testQemuDiskXMLToProps, testQemuDiskXMLToPropsValidateSchema and
testQemuDiskXMLToPropsValidateFile.

testQemuDiskXMLToPropsValidateFileSrcOnly already uses new helpers.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

Based on the configuration from the only qemuxml2argv test.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>

virHostCPUGetStatsLinux walks through every cpu in /proc/stat until it
finds cpu%cpuNum that matches with the requested cpu.
If none is found it logs the error but it should return -1, instead of 0.
Otherwise virsh nodecpustats --cpu <invalid cpu number> and API bindings
don't fail properly, printing a blank line instead of an error message.

This patch also includes an additional test for virhostcputest to avoid
this regression to happen again in the future.

Fixes: 93af79fbReported-by: NSatheesh Rajendran <satheera@in.ibm.com>
Signed-off-by: NMauro S. M. Rodrigues <maurosr@linux.vnet.ibm.com>
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NChristian Ehrhardt <christian.ehrhardt@canonical.com>
Tested-by: NChristian Ehrhardt <christian.ehrhardt@canonical.com>