Commit Graph

2154 Commits

Author SHA1 Message Date
Michal Privoznik
10594bb311 qemu: Generate cmd line for pipewire audio backend
This is mostly straightforward, except for a teensy-weensy
detail: usually, there's no system wide daemon running, no system
wide available socket that anybody could connect to. PipeWire
uses a per user daemon approach instead. But this in turn means,
that the socket location floats between various locations and is
derived from various environment variables (just like the actual
socket name) and thus we must pass the variables to QEMU.

Resolves: https://gitlab.com/libvirt/libvirt/-/issues/560
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-11-24 17:49:20 +01:00
Michal Privoznik
c472ce024b conf: Introduce pipewire audio backend
QEMU gained support for PipeWire audio backend (see QEMU commit
of v8.0.0-403-gc2d3d1c294). Its configuration knobs are basically
the same as pulseaudio's, except for PA's server name. Therefore,
a lot of code is copied over from pulseadio and fixed by
s/Pulse/Pipewire/ or s/pulseaudio/pipewire/.

There's one ley difference to PA though: pipewire daemon is
usually on per user basis (just like our qemu:///session).
Therefore, introduce this 'runtimeDir' attribute, which allows
specifying path to pipewire daemon socket (useful for
qemu:///system for instance).

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-11-24 17:49:02 +01:00
Tim Wiederhake
784a63c189 Fix cpu-host-model test data
This was broken by the recent addition of vmx-* features.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
2023-11-24 11:05:13 +01:00
Tim Wiederhake
6562669388 cpu_map: Add missing vmx features from MSR 0x491
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:47:05 +01:00
Tim Wiederhake
a155693a40 cpu_map: Add missing vmx features from MSR 0x490
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:47:02 +01:00
Tim Wiederhake
2e49a26158 cpu_map: Add missing vmx features from MSR 0x48F
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:47:00 +01:00
Tim Wiederhake
848e2d0949 cpu_map: Add missing vmx features from MSR 0x48E
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:46:58 +01:00
Tim Wiederhake
23f509c102 cpu_map: Add missing vmx features from MSR 0x48D
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:46:55 +01:00
Tim Wiederhake
102f6335b9 cpu_map: Add missing vmx features from MSR 0x48C
Differences from qemu:

* "vmx-ept-uc" (bit 8) and "vmx-ept-wb" (bit 14) are not added to
qemu's list of named features yet, but used in several qemu cpu
models never the less. Add to libvirt regardless.

* "vmx-invvpid-single-context" (bit 41) is erroneously called
"vmx-invept-single-context" in qemu. This is the name of the
feature associated with bit 25 in both libvirt and qemu.

* "vmx-invvpid-single-context-noglobals" (bit 43) is erroneously
called "vmx-invept-single-context-noglobals". Use the correct name.

Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:46:50 +01:00
Tim Wiederhake
94eacd5a5f cpu_map: Add missing vmx features from MSR 0x48B
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:46:47 +01:00
Tim Wiederhake
a1862e3de2 cpu_map: Add missing vmx features from MSR 0x485
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:46:44 +01:00
Tim Wiederhake
3590bb344e cpu_map: Add missing vmx features from MSR 0x480
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:46:40 +01:00
Tim Wiederhake
20c4a99fc5 cpu_map: Add missing feature "gds-no"
Signed-off-by: Tim Wiederhake <twiederh@redhat.com>
Reviewed-by: Jiri Denemark <jdenemar@redhat.com>
2023-11-24 10:46:21 +01:00
Peter Krempa
658a024127 qemucapabilitiestest: Update capabilities data for 8.2 dev cycle on x86_64
Update to v8.1.0-3111-gad6ef0a42e

Notable changes:
 - migration:
    - multichannel migration added
        - 'channels' array argument of migrate/migrate-incomming added
    - legacy block migration deprecated
        - 'blk', 'inc' arguments of 'migrate' command
        - 'block' migration capability
        - 'block-incremental' migration parameter
    - legacy migration compression deprecated
        - 'compress' migration capability
        - 'compress-level', 'compress-threads', 'compress-wait-thread',
          'decompress-threads' migration paramters
    - migration parameter 'mode' with 'normal' and 'cpr-reboot' options
      added

 - new formats for 'dump-guest-memory'
    - 'dump-raw-lzo', 'kdump-raw-snappy', 'kdump-raw-zlib' added

 - 'block-job-change' command added

 - 'HV_BALLOON_STATUS_REPORT' event and
   'query-hv-balloon-status-report' command added

 - deprecated old i440fx machine types:
    - 'pc-i440fx-2.0', 'pc-i440fx-2.1', 'pc-i440fx-2.2', 'pc-i440fx-2.3'

 - 'sha-ni' and few other ring 0 instruction are now exposed to tcg
   guests

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-11-10 10:04:04 +01:00
Peter Krempa
bdb516901c qemuxml2argvtest: Add '8.1.0' variants of 'cpu-host-model*' tests
Upcoming test bump will cause some changes thus preserve the existing
state.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-11-10 10:04:04 +01:00
Peter Krempa
f90675087f qemuxml2argvtest: Add complementary 'kvm/tcg' test cases for cpu-host-model*
CPU host model expansion depends on the CPU data from the capabilities
and can change based on emulation type. Add complementary tests to the
ones we already have to ensure full coverage.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-11-10 10:04:04 +01:00
Peter Krempa
3b1e855505 qemuxml2argvtest: Rename 'cpu-host-model' tests according to emulation type
The cpu-host-model.xml test case uses 'kvm' whereas the
fallback/nofallback cases use tcg in the definition. Rename them
accordingly so that the complement cases can be added later.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-11-10 10:04:04 +01:00
Peter Krempa
80494dfb74 qemu: block: Use qemuBlockStorageSourceAddBlockdevCommonProps for storage slice
Use the new helper in qemuBlockStorageSourceGetBlockdevStorageSliceProps
to format the common bits.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-11-02 15:31:40 +01:00
Peter Krempa
84df50dc87 qemuBlockStorageSourceGetBackendProps: Unify ordering of fields
Use the same ordering of the relevant fields as we do for the format
layer -blockdev so that later they can be refactored without test
fallout.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-11-02 15:31:10 +01:00
Andrea Bolognani
3653eb5dcf tests: Fix some test cases on macOS
Test cases that depend on duplicating fds are using fairly big
values as targets.

This works fine on Linux, where RLIMIT_NOFILE is 1024 by
default, but fails on macOS which uses 256 as the default.

Decrease the values so that they're valid across all platforms.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
2023-10-26 11:31:06 +02:00
Peter Krempa
829d9ed829 qemuxml2(argv|xml)test: Add network backed disk type='sd'
Add a few examples of SD cards backed with network storage to capture
the current state as the formatter code is about to be refactored.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-10-26 09:02:23 +02:00
Michal Privoznik
3fd64fb0e2 virDomainMemoryDefValidate: Check for overlapping memory devices
As of v9.4.0-rc2~5 it is possible to specify guest address where
a virtio-mem/virtio-pmem memory device is mapped to. What that
commit forgot to introduce was a check for overlaps.

And yes, this is technically an O(n^2) algorithm, as
virDomainMemoryDefValidate() is called over each memory device
and after this, virDomainMemoryDefValidate() also iterates over
each memory device. But given there's usually only a handful of
such devices, and this runs only when parsing domain XML I guess
code readability wins over some less obvious solution.

Resolves: https://issues.redhat.com/browse/RHEL-4452
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-22 16:37:10 +02:00
Andrea Bolognani
1839c5c46e qemu: Improve error message for failed firmware autoselection
The current message can be misleading, because it seems to suggest
that no firmware of the requested type is available on the system.

What actually happens most of the time, however, is that despite
having multiple firmwares of the right type to choose from, none
of them is suitable because of lacking some specific feature or
being incompatible with some setting that the user has explicitly
enabled.

Providing an error message that describes exactly the problem is
not feasible, since we would have to list each candidate along
with the reason why we rejected it, which would get out of hand
quickly.

As a small but hopefully helpful improvement over the current
situation, reword the error message to make it clearer that the
culprit is not necessarily the firmware type, but rather the
overall domain configuration.

Suggested-by: Michael Kjörling <7d1340278307@ewoof.net>
Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
2023-09-22 15:50:35 +02:00
Jonathon Jongsma
3310b12d52 qemu: implement ssh-agent auth for ssh disks with nbdkit
It's not possible to use password-protected ssh keys directly with
libvirt because libvirt doesn't have any way to prompt a user for the
password. To accomodate password-protected key files, an administrator
can add these keys to an ssh agent and then configure the domain with
the path to the ssh-agent socket.

Note that this requires an administrator or management app to
configure the ssh-agent with an appropriate socket path and add the
necessary keys to it. In addition, it does not currently work with
selinux enabled. The ssh-agent socket would need a label that libvirt
would be allowed to access rather than unconfined_t.

Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2023-09-19 14:28:50 -05:00
Jonathon Jongsma
68599168ea qemu: implement keyfile auth for ssh disks with nbdkit
For ssh disks that are served by nbdkit, we can support logging in with
an ssh key file. Pass the path to the configured key file and the
username to the nbdkit process.

Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2023-09-19 14:28:50 -05:00
Jonathon Jongsma
21b377a31b qemu: implement knownHosts for ssh disks with nbdkit
For ssh disks that are served by nbdkit, use the configured value for
knownHosts and pass it to the nbdkit process.

Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2023-09-19 14:28:50 -05:00
Jonathon Jongsma
36f591402f qemu: implement password auth for ssh disks with nbdkit
For ssh disks that are served by nbdkit, lookup the password from the
configured secret and securely pass it to the nbdkit process using fd
passing.

Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2023-09-19 14:28:50 -05:00
Jonathon Jongsma
0993f9f905 qemu: add test for authenticating a https network disk
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2023-09-19 14:28:50 -05:00
Jonathon Jongsma
3361cfd79e qemu: use nbdkit to serve network disks if available
For virStorageSource objects that contain an nbdkitProcess, start that
nbdkit process to serve that network drive and then pass the nbdkit
socket to qemu rather than sending the network url to qemu directly.

Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2023-09-19 14:28:50 -05:00
Jonathon Jongsma
7c5a5366ff schema: allow 'ssh' as a protocol for network disks
There was support in the code for parsing protocol='ssh' on network disk
sources, but it was not present in the xml schema. Add this to the
schema.

Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2023-09-19 14:28:49 -05:00
Jonathon Jongsma
4ef2bcfd3f qemu: Implement support for vDPA block devices
Requires recent qemu with support for the virtio-blk-vhost-vdpa device
and the ability to pass a /dev/fdset/N path for the vdpa path (8.1.0)

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1900770
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com>
Reviewed-by: Peter Krempa <pkrempa@redhat.com>
2023-09-12 11:06:41 -05:00
Peter Krempa
54257ed51b qemuxml2argvtest: Modernize rest of the test cases
The rest of the test cases has no change in the output now that we've
assumed some flags.

Remove the fake-caps test macros after conversion.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
fd1a18f645 qemuxml2argvtest: Modernize 'cpu-exact' cases
Use real capabilities, but select the fake 'Haswell' host CPU for test
stability.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
f1e288f6d4 qemuxml2argvtest: Modernize 'cpu-(no)fallback' test
Previously without modern capabilities the test was relying on a CPU
model which was not entered into a fake list of supported cpus.

With real capabilities we have to pick a CPU model which is supported by
libvirt but in some version is not supported by qemu. I've picked
EPYC-Milan, which was introduced into qemu-6.0.

This test configures a CPU which is equivalent to EPYC-Rome by disabling
features from EPYC-Milan and uses a versioned real caps test to check it
against a qemu which doesn't support EPYC-Milan.

With real capabilities though, we can also do a positive test case by
using a version whic doesh support it. I've specifically not used the
LATEST caps so that it doesn't change once capabilities are bumped.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
36072d873e qemuxml2argvtest: Modernize 'cpu-strict' case
Use the fake Haswell processor definition and augment the list of
features to make the test pass.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
457b6ee0c8 qemuxml2argvtest: Modernize 'cpu-minimum' cases
Use real capabilities, but select the fake 'Haswell' host CPU for test
stability.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
dcad2c05f5 qemuxml2argvtest: Modernize 'cpu-Haswell' test cases
Modernize all test cases which set 'Haswell' as the host cpu model.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
dbd99066a4 qemuxml2argvtest: Modernize test cases using 'host-model' CPU
Modernize test cases using 'host-mode' cpu type, where the actual CPU
doesn't isn't important.

As using the host cpu from the 'latest' capabilities data would cause
test churn in case the host cpu changes in the future, convert them
using the overriden Haswell cpu.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
e485f8bd4e qemuxml2argvtest: Consolidate and modernize CPU topology testing
Move the 'smp-dies' test case into 'cpu-topology4' and remove
unnecessary cruft.

Remove cpu definition from 'cpu-topology2' as it's not relevant to the
test case.

Remove 'smp' case as it's covered by the rest.

Use real capabilities for all of them.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
07e6396c6a qemuxml2argvtest: Modernize migration/restore tests
As these were using DO_TEST_FULL the churn-reducing patches didn't
influence these.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
035b9bde6a qemuxml2argvtest: Properly modernize 'clock-hpet-off' test case
Since qemu-8.0 a new way to disable 'hpet' via -machine was added.
Properly test both branches with real capabilities.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
da68054702 qemuxml2argvtest: Assume QEMU_CAPS_QXL_VRAM64, QEMU_CAPS_(QXL|VGA)_VGAMEM for fake-caps tests
Assume the features modern qemus have to bring the test data closer
to the 'latest' real-caps versions.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
84a14d1b76 qemuxml2argvtest: Assume QEMU_CAPS_USB_HOST_HOSTDEVICE and QEMU_CAPS_DEVICE_QEMU_XHCI for fake-caps tests
Assume the two features modern qemus have to bring the test data closer
to the 'latest' real-caps versions.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
c159cecd4c qemuxml2argvtest: Assume QEMU_CAPS_CPU_MIGRATABLE and QEMU_CAPS_SMP_DIES for fake-caps tests
Assume the two features modern qemus have to bring the test data closer
to the 'latest' real-caps versions.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
9e7907e78b qemuxml2argvtest: Assume QEMU_CAPS_DEVICE_PCIE_ROOT_PORT for fake-caps tests
Assume the support for the 'pcie-root-port' all modern qemus have to
bring the test data closer to the 'latest' real-caps versions.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:45 +02:00
Peter Krempa
2516394892 qemuxml2argvtest: Assume QEMU_CAPS_PIIX3_USB_UHCI for all fake-caps tests
To minimize further churn when coverting to real capabilities, assume
that all fake-caps machines support the piix3 USB controller.

Since we already have solid testing of USB controllers, this will have
effect only in cases when it's not relevant to the test itself.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:44 +02:00
Peter Krempa
28820bf5c0 qemuxml2argvtest: Consolidate testing of 'xhci' controllers
Rename the 'usb-controller-explicit-(unavailable-)q35' test case to
'usb-controller-nec-xhci'. Since this also covers what
'usb-controller-xhci' was testing the latter is removed.

Other 'usb-xhci' test cases which were using the NEC controller are also
renamed to contain the name.

In case of 'usb-controller-qemu-xhci' the negative test case is deleted
as we don't need two cases for missing explicitly specified controller
and the positive case is modernized to use real capabilities.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:44 +02:00
Peter Krempa
136abdc992 qemuxml2argvtest: Improve default usb controller testing
Improve testing of an explicitly requested USB controller without a
model being provided.

For this purpose the 'usb-controller' case is renamed to
'usb-controller-default-i440fx'; 'usb-controller-default-q35' is moved
up to form a group. In both cases tests are covnerted to use
DO_TEST_CAPS_LATEST.

A new 'usb-controller-default-isapc' negative test case is added for
symmetry.

The negative test case 'usb-controller-default-unavailable-q35' is
converted to use latest caps, but stripping the default controller
instead of using fake caps. Additionally for symmetry
'usb-controller-default-unavailable-i440fx' is added although that
doesn't cause failure, but rather a graceful downgrade to use '-usb'.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:44 +02:00
Peter Krempa
123f57c947 qemuxml2argvtest: Add tests for auto-added USB controllers
There are per-machine type variations on which usb controller will be
picked on an x86_64 machine. Add test cases where a USB controller is
completely missing to cover all 3 cases ('isapc', 'pc', 'q35') when an
USB controller is not explicitly requested by the user.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:44 +02:00
Peter Krempa
2b8ccecbc3 qemuxml2argvtest: Consolidate tests for usb controllers
Move all controller related tests together and consolidate naming of the
test cases.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2023-09-06 14:09:44 +02:00