Michal Privoznik [Tue, 19 Jan 2021 09:16:55 +0000 (10:16 +0100)]
qemu: Build command line for virtio-mem
Nothing special is happening here. All important changes were
done when for 'virtio-pmem' (adjusting the code to put virtio
memory on PCI bus, generating alias using
qemuDomainDeviceAliasIndex(). The only bit that might look
suspicious is no prealloc for virtio-mem. But if you think about
it, the whole purpose of this device is to change amount of
memory exposed to guest on the fly. There is no point in locking
the whole backend in memory.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Michal Privoznik [Mon, 18 Jan 2021 15:13:12 +0000 (16:13 +0100)]
conf: Introduce virtio-mem <memory/> model
The virtio-mem is paravirtualized mechanism of adding/removing
memory to/from a VM. A virtio-mem-pci device is split into blocks
of equal size which are then exposed (all or only a requested
portion of them) to the guest kernel to use as regular memory.
Therefore, the device has two important attributes:
1) block-size, which defines the size of a block
2) requested-size, which defines how much memory (in bytes)
is the device requested to expose to the guest.
The 'block-size' is configured on command line and immutable
throughout device's lifetime. The 'requested-size' can be set on
the command line too, but also is adjustable via monitor. In
fact, that is how management software places its requests to
change the memory allocation. If it wants to give more memory to
the guest it changes 'requested-size' to a bigger value, and if it
wants to shrink guest memory it changes the 'requested-size' to a
smaller value. Note, value of zero means that guest should
release all memory offered by the device. Of course, guest has to
cooperate. Therefore, there is a third attribute 'size' which is
read only and reflects how much memory the guest still has. This
can be different to 'requested-size', obviously. Because of name
clash, I've named it 'current' and it is dealt with in future
commits (it is a runtime information anyway).
In the backend, memory for virtio-mem is backed by usual objects:
memory-backend-{ram,file,memfd} and their size puts the cap on
the amount of memory that a virtio-mem device can offer to a
guest. But we are already able to express this info using <size/>
under <target/>.
Therefore, we need only two more elements to cover 'block-size'
and 'requested-size' attributes. This is the XML I've came up
with:
This capability tracks whether memory-backend-* supports .reserve
attribute which is going to be important for backends associated
with virtio-mem devices.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Michal Privoznik [Wed, 20 Jan 2021 10:30:34 +0000 (11:30 +0100)]
virhostmem: Introduce virHostMemGetTHPSize()
New virHostMemGetTHPSize() is introduced which allows caller to
obtain THP PMD (Page Middle Directory) size, which is equal to
the minimal size that THP can use, taken from kernel doc
(Documentation/admin-guide/mm/transhuge.rst):
Some userspace (such as a test program, or an optimized memory allocation
library) may want to know the size (in bytes) of a transparent hugepage::
Since this size depends on the host architecture and the kernel
it won't change whilst libvirtd is running. Therefore, we can use
virOnce() and cache the value. Of course, we can be running under
kernel that has THP disabled or has no notion of THP at all. In
that case a negative value is returned to signal error.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Michal Privoznik [Tue, 21 Sep 2021 14:26:48 +0000 (16:26 +0200)]
virCPUDefParseXML: Prefer virXMLPropUInt over virXPathUInt
When parsing CPU topology, which is described in <topology/>
attributes we can use virXMLPropUInt() instead of virXPathUInt()
as the former results in shorter code.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Ján Tomko [Fri, 24 Sep 2021 14:04:30 +0000 (16:04 +0200)]
qemu: always assume QEMU_CAPS_SECCOMP_BLACKLIST
elevateprivileges was introduced by QEMU commit: 73a1e64725 "seccomp: add elevateprivileges argument to command line"
released in 2.11.0
and later made conditional on SECCOMP support by: 9d0fdecbad sandbox: disable -sandbox if CONFIG_SECCOMP undefined
Use the existence of the sandbox option as a witness for its support.
Signed-off-by: Ján Tomko <jtomko@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
qemu: ingore the transient domain state in fake reboot
When action for 'on_poweroff' is set to 'restart', 'fake reboot'
is triggered and qemu shutdown state is transient. Domain state
need not to be changed and events not sent in this case.
Fixes: 4ffc807214cb80086d57e1d3e7b60959a41d2874 Signed-off-by: Zhenzhong Duan <zhenzhong.duan@intel.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
virsh: domain: fix mistake in cmdMigrateSetMaxDowntime()
If there was added a new return value indicating success to the
function virDomainMigrateSetMaxDowntime() in the future, because
of the way the function is called it would be treated it as an
error state and would return false (indicating failure). This
patch fixes it, so that the call of the function follows the same
pattern as is currently set in libvirt.
Signed-off-by: Kristina Hanicova <khanicov@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Fri, 24 Sep 2021 09:15:10 +0000 (11:15 +0200)]
vsh: Don't check for OOM in vshGetTypedParamValue()
Both function description and function itself mention check for
OOM which can't happen really. There was a bug in glib where
g_strdup_*() might have not aborted on OOM, but we have our own
implementation when dealing with broken glib (see
vir_g_strdup_printf()). Therefore, checking for OOM is redundant
and can never be true.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Erik Skultety [Tue, 21 Sep 2021 16:48:19 +0000 (18:48 +0200)]
libvirt.spec.in: Replace %global with %define
Commit b5e8db8f14d tuned the SPEC file so that libvirt daemons restart
on package upgrade. In order to do that it added a bunch of
parametrized macros using the %global directive. This caused a problem
when running RPM builds on CentOS Stream 8 resulting in:
error: Too many levels of recursion in macro expansion. It is likely
caused by recursive macro declaration.
error: Macro %libvirt_daemon_perform_restart failed to expand
error: line 1275: %global libvirt_daemon_perform_restart() \
if test %libvirt_daemon_needs_restart %1 \
then \
/bin/systemctl try-restart %1.service >/dev/null 2>&1 || : \
fi \
%libvirt_daemon_finish_restart %1
There are 2 important differences between %global and %define
directives:
1) %define is local-only and does have scope - in reality though, its
scope is apparently not really enforced because it behaves exactly
the same way as %global
2) %define is evaluated at the time of use while %global is evaluated
at the time of definition
The latter and the fact the macro is parametrized is the reason why the
RPM builds fails on CentOS. Strangely enough this only happens on
CentOS Stream, but not Fedora (which is also the main proponent of
replacing %define with %global). Anyhow, replacing %global with %define
makes the rpmbuild to pass on both and along with package upgrades.
Signed-off-by: Erik Skultety <eskultet@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
QEMU added the capability to disable file transfers via spice in commit 5ad24e5f3b ("spice: Add -spice disable-agent-file-transfer cmdline
option (rhbz#961850)") released in qemu-v1.6.0 and the option can't be
disabled.
Remove the unnecessary validation.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Peter Krempa [Thu, 23 Sep 2021 08:10:05 +0000 (10:10 +0200)]
qemu: command: Always assume QEMU_CAPS_VNC_OPTS
The switch to QemuOpts parser which brought the long-form options
happened in qemu commit 4db14629c3 ("vnc: switch to QemuOpts, allow
multiple servers") released in v2.3.0.
We can always assume this capability and remove the old-style
generators.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Peter Krempa [Thu, 23 Sep 2021 08:01:45 +0000 (10:01 +0200)]
qemu: Always assume presence of QEMU_CAPS_OBJECT_TLS_CREDS_X509
The 'tls-creds-x509' object is always registered even when qemu is built
without gnutls for all supported qemu versions. This means we cannot
probe for its support and thus simplify the code using TLS.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
scripts: include function versions in API definition
In order to auto-generate more of the language binding code, it is
desirable to know what libvirt version an API was introduced in.
We can extract this information from the .syms files and expose
it in the API description
Reviewed-by: Michal Privoznik <mprivozn@redhat.com> Tested-by: Victor Toso <victortoso@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Peter Krempa [Wed, 22 Sep 2021 08:41:52 +0000 (10:41 +0200)]
qemuxml2argvtest: Modernize outstanding disk-backend related tests
For backend related tests we need to cover the pre-blockdev and
post-blockdev era, so the fake-capability test is converted to a
combination of DO_TEST_CAPS_VER(..., "4.1.0") and DO_TEST_CAPS_LATEST.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Peter Krempa [Wed, 22 Sep 2021 08:30:56 +0000 (10:30 +0200)]
qemuxml2argvtest: Modernize 'disk-network-rbd-no-colon' case
The test case fails in pre-blockdev scenarios as it would pass RBD
parameters behind our back but succeeds after as we pass it in JSON form
which doesn't have that defect.
Cover both cases instead of the fake-caps version.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Peter Krempa [Wed, 22 Sep 2021 08:08:27 +0000 (10:08 +0200)]
qemuxml2argvtest: Merge 'disk-iscsi' case into 'disk-cache'
Use the iSCSI disk path in one of the disks of the 'disk-cache' test as
it's the only specialty of 'disk-iscsi' case and remove the now
pointless files.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Peter Krempa [Wed, 22 Sep 2021 08:00:00 +0000 (10:00 +0200)]
qemuxml2argvtest: Replace 'disk-network-iscsi' by the more comprehensive 'disk-network-iscsi-modern'
Move the contents of 'disk-network-iscsi-modern' into 'disk-network-iscsi'
to reuse the name and also invocation with real capablities and remove
the leftovers.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Peter Krempa [Tue, 21 Sep 2021 17:14:19 +0000 (19:14 +0200)]
qemuxml2argvtest: Use DO_TEST_CAPS_LATEST_PARSE_ERROR for disk cases
Convert all the disk-related negative cases to use 'latest'
capabilities. The checks are mostly related to validation so using
real capabilities doesn't influence the outcome.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Peter Krempa [Tue, 21 Sep 2021 17:17:25 +0000 (19:17 +0200)]
qemuxml2argvtest: Remove 'disk-usb-nosupport' case
In effor to convert all test cases to real capability testing, this
test doesn't make sense any more as even the oldest QEMU supported
supports USB storage.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Peter Krempa [Tue, 21 Sep 2021 14:35:51 +0000 (16:35 +0200)]
qemuxml2argvtest: Agregate all 'disk'-related test cases
There were a few disk-related test cases spread around in the test
invocation calls. Move them together with disk tests and move one
irrelevant case away from the disk test block.
Note that there are still a few tests having 'disk' in the name but they
belong to different groups mostly.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Implement virConnectGetAllDomainStats in a modular way just like QEMU
driver, though remove some params in GetStatsWorker that we don't need
in test driver currently.
Only add the worker to get state so far, more worker will be added
in the future.
Signed-off-by: Luke Yue <lukedyue@gmail.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
projects / thirdparty / libvirt.git / log
