Not all platforms use the '.so' suffix for shared libraries,
which is how plugins are built. Use the recently introduced
dso_suffix() helper to get the proper host suffix.
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2804
Suggested-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20250220080215.49165-4-philmd@linaro.org>
[AJB: moved plugin_file into testcase.py]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-11-alex.bennee@linaro.org>
Introduce a helper to get the default shared library
suffix used on the host.
Suggested-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20250220080215.49165-3-philmd@linaro.org>
[AJB: dropped whitespace cmd.py damage]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-10-alex.bennee@linaro.org>
./tests/functional/test_aarch64_tcg_plugins.py needs to have plugin
libinsn built. However, it's not listed as a dependency, so meson can't
know it needs to be built.
Thus, we keep track of all plugins, and add them as an explicit
dependency.
Fixes: 4c134d07b9 ("tests: add a new set of tests to exercise plugins")
Signed-off-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-9-alex.bennee@linaro.org>
Update to the most recent aarch64_virt_gpu image. The principle
differences are:
- target a v8.0 baseline CPU
- latest vkmark (2025.1)
- actually uses the rootfs (previously was initrd)
- rootfs includes more testing tools for interactive use
See README.md in https://fileserver.linaro.org/s/ce5jXBFinPxtEdx for
details about the image creation and the buildroot config.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-8-alex.bennee@linaro.org>
Add two more test modes using glmark2-wayland to exercise the OpenGL
pass-through modes with virgl. Virgl can run with or without the
hostmem blob support. To avoid repeating ourselves too much we make
the initial pass a simple --validate pass.
We might want to eventually add more directed tests and individual
features later on but the glmark/vkmark tests are a good general
smoke test for accelerated 3D.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-7-alex.bennee@linaro.org>
While running the new GPU tests it was noted that the proprietary
nVidia driver barfed when run under the sanitiser:
2025-02-20 11:13:08,226: [11:13:07.782] Output 'headless' attempts
EOTF mode SDR and colorimetry mode default.
2025-02-20 11:13:08,227: [11:13:07.784] Output 'headless' using color
profile: stock sRGB color profile
and that's the last thing it outputs.
The sanitizer reports that when the framework sends the SIGTERM
because of the timeout we get a write to a NULL pointer (but
interesting not this time in an atexit callback):
UndefinedBehaviorSanitizer:DEADLYSIGNAL
==471863==ERROR: UndefinedBehaviorSanitizer: SEGV on unknown address
0x000000000000 (pc 0x7a18ceaafe80 bp 0x000000000000 sp 0x7ffe8e3ff6d0
T471863)
==471863==The signal is caused by a WRITE memory access.
==471863==Hint: address points to the zero page.
#0 0x7a18ceaafe80
(/lib/x86_64-linux-gnu/libnvidia-eglcore.so.535.183.01+0x16afe80)
(BuildId: 24b0d0b90369112e3de888a93eb8d7e00304a6db)
#1 0x7a18ce9e72c0
(/lib/x86_64-linux-gnu/libnvidia-eglcore.so.535.183.01+0x15e72c0)
(BuildId: 24b0d0b90369112e3de888a93eb8d7e00304a6db)
#2 0x7a18ce9f11bb
(/lib/x86_64-linux-gnu/libnvidia-eglcore.so.535.183.01+0x15f11bb)
(BuildId: 24b0d0b90369112e3de888a93eb8d7e00304a6db)
#3 0x7a18ce6dc9d1
(/lib/x86_64-linux-gnu/libnvidia-eglcore.so.535.183.01+0x12dc9d1)
(BuildId: 24b0d0b90369112e3de888a93eb8d7e00304a6db)
#4 0x7a18e7d15326 in vrend_renderer_create_fence
/usr/src/virglrenderer-1.0.0-1ubuntu2/obj-x86_64-linux-gnu/../src/vrend_renderer.c:10883:26
#5 0x55bfb6621871 in virtio_gpu_virgl_process_cmd
The #dri-devel channel confirmed:
<digetx> stsquad: nv driver is known to not work with venus, don't use
it for testing
So lets skip running the test to avoid known failures. As we now use
vulkaninfo to probe we also need to handle the case where there is no
Vulkan driver configured for the hardware.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reported-by: Peter Maydell <peter.maydell@linaro.org>
Cc: Dmitry Osipenko <dmitry.osipenko@collabora.com>
[AJB: also skip if vulkaninfo can't find environment]
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-6-alex.bennee@linaro.org>
The message:
MESA-VIRTIO: debug: stuck in fence wait with iter at %d
Seems to occur more often on debug builds. Rather than waiting for our
long timeout to hit we might as well bail as soon as we see the
message.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-5-alex.bennee@linaro.org>
It's possible to build QEMU without support for the GL enabled GPU
devices and we can catch that earlier with an explicit check.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-4-alex.bennee@linaro.org>
In preparation for handling more tests split out the common machine
setup details from the test specific stuff and add a helper for
launching the weston test. Instead of searching for "vkmark score" we
set a custom PS1 and wait for a successful completion. This ensures we
capture the score in the console log which otherwise wouldn't log
anything.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-3-alex.bennee@linaro.org>
I want to expand the number of tests to cover a wide range of
configurations. That starts with splitting off from the normal virt
test from which it doesn't really share much code. We can also reduce
the timeout of the original virt test now it is now longer burdened
with testing the GPU.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-Id: <20250304222439.2035603-2-alex.bennee@linaro.org>
In ASPEED SDK v09.05, the naming convention for pre-built images has been
updated. The pre-built image for AST2700 A0 has been renamed to
ast2700-a0-default, while ast2700-default is now used for AST2700 A1.
Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250307035945.3698802-28-jamin_lin@aspeedtech.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Modified the temperature hwmon path to use a wildcard to handle different SDK
versions: "cat /sys/bus/i2c/devices/1-004d/hwmon/hwmon*/temp1_input".
Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250307035945.3698802-27-jamin_lin@aspeedtech.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Added a new method "start_ast2700_test" to the "AST2x00MachineSDK" class and
this method centralizes the logic for starting the AST2700 test, making it
reusable for different test cases.
Signed-off-by: Jamin Lin <jamin_lin@aspeedtech.com>
Reviewed-by: Cédric Le Goater <clg@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250307035945.3698802-26-jamin_lin@aspeedtech.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Use do_test_arm_aspeed_openbmc() to run the latest OpenBMC firmware
build of the bletchley BMC.
Reviewed-by: Patrick Williams <patrick@stwcx.xyz>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250129071820.1258133-6-clg@redhat.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Use the new do_test_arm_aspeed_openbmc() routine to run the latest
OpenBMC firmware build of the romulus BMC. Remove the older routine
which is now unused.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250129071820.1258133-4-clg@redhat.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
The OpenBMC images currently used by QEMU to test the Aspeed machines
are rather old. To prepare an update to the latest builds, we need to
adjust the console patterns. Introduce a new routine to preserve the
current tests.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250129071820.1258133-2-clg@redhat.com
Signed-off-by: Cédric Le Goater <clg@redhat.com>
The kernel that is used in the sx1 test prints the usual Linux log
onto the serial console, but this test currently ignores it. To
make sure that the serial device is working properly, let's check
for some strings in the output here.
While we're at it, also add the test to the corresponding section
in the MAINTAINERS file.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-id: 20250226104833.1176253-1-thuth@redhat.com
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
The virtio_balloon test is currently hanging for unknown reasons
when being run on the shared gitlab CI runners (which don't provide
KVM, thus it's running in TCG mode there). All other functional tests
that use the same asset (the Fedora 31 kernel) have already been
marked to work only with KVM in the past, so those other tests are
skipped on the shared gitlab CI runners. As long as the problem isn't
fully understood and fixed, let's do the same with the virtio_balloon
test to avoid that the CI is failing here.
Message-ID: <20250307063904.1081961-1-thuth@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
There are two race conditions in the recently added virtio balloon
test
* The /dev/vda device node is not ready
* The virtio-balloon driver has not issued the first stats refresh
To fix the former, monitor dmesg for a line about 'vda'.
To fix the latter, retry the stats query until seeing fresh data.
Adding 'quiet' to the kernel command line reduces serial output
which otherwise slows boot, making it less likely to hit the former
race too.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20250304183340.3749797-1-berrange@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: David Hildenbrand <david@redhat.com>
[thuth: Break long line to avoid checkpatch error]
Signed-off-by: Thomas Huth <thuth@redhat.com>
We run the gitlab-CI with the untrusted tests enabled, and
the test_replay_mips64el_malta_5KEc_cpio subtest is rather slow,
so this already hit the standard 90 seconds timeout in the CI.
Increase the timeout for more headroom.
Reported-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20250305074353.52552-1-thuth@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
The kernel has a driver for the pcnet NIC included, and the initrd has
a "tftp" command, so we can test a simple network transfer here, too.
Message-ID: <20250227103915.19795-3-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
We are going to use this code in other tests, too, so let's move it
to the qemu_test module to be able to re-use it more easily.
Message-ID: <20250227103915.19795-2-thuth@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
The zstd command will print incremental decompression progress to stderr
when running. Fortunately it is not on stdout as that would confuse the
TAP parsing, but we should still not have this printed. By switching
from 'check_call' to 'run' with the check=True and capture_output=True
we'll get the desired silence on success, and on failure the raised
exception will automatically include stdout/stderr data for diagnosis
purposes.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250228102738.3064045-8-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250228102738.3064045-7-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
If the QEMU binary was built for a 32-bit ELF target we cannot run the
memory address space tests as they all require ability to address more
RAM that can be represented on 32-bit.
We can't use a decorator to skip the tests as we need setUp() to run to
pick the QEMU binary, thus we must call a method at the start of each
test to check and skip it. The functional result is effectively the
same as using a decorator, just less pretty. This code will go away when
32-bit hosts are full dropped from QEMU.
The code allows any non-ELF target since all macOS versions supported
at 64-bit only and we already dropped support for 32-bit Windows.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20250228102738.3064045-6-berrange@redhat.com>
[thuth: Add missing byteorder='little' to from_bytes()]
Signed-off-by: Thomas Huth <thuth@redhat.com>
maxmem=4G is too large to address on 32-bit hosts, so reduce it
to 2G since the tuxrun tests don't actually need such an elevated
memory limit.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250228102738.3064045-5-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
The 'qemu_bin' field is currently set on the class, despite being
accessed as if it were an object instance field with 'self.qemu_bin'.
This is no obvious need to have it as a class field, so move it into
the object instance.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250228102738.3064045-3-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
This was copied over from avocado but has not been used in the new
functional tests.
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20250228102738.3064045-2-berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Replace the advent calendar test with a buildroot image built with
qemu_ppc64_e5500_defconfig. Unlike the advent calendar image, this
newer buildroot image supports networking, too. Thus boot a ppce500
machine from kernel and disk, test network and poweroff.
Add '-no-shutdown' to the command line to avoid exiting from QEMU
as it seems to bother the functional framework.
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Message-ID: <20250226065013.196052-1-clg@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
[thuth: Add some wording about network support to the commit message]
Signed-off-by: Thomas Huth <thuth@redhat.com>
On my local machine, for a debug build, sbsaref_alpine takes
nearly 900s:
$ (cd build/x86 && ./pyvenv/bin/meson test --setup thorough --suite func-thorough func-aarch64-aarch64_sbsaref_alpine
)
1/1 qemu:func-thorough+func-aarch64-thorough+thorough / func-aarch64-aarch64_sbsaref_alpine
OK 896.90s
arm_aspeed_rainier can also run close to its current timeout:
6/44 qemu:func-thorough+func-arm-thorough+thorough / func-arm-arm_aspeed_rainier
OK 215.75s
and arm_aspeed_ast2500 and arm_aspeed_ast2600 can go over:
13/44 qemu:func-thorough+func-arm-thorough+thorough / func-arm-arm_aspeed_ast2600
OK 792.94s
27/44 qemu:func-thorough+func-arm-thorough+thorough / func-arm-arm_aspeed_ast2500
TIMEOUT 480.01s
The sx1 test fails not on the overall meson timeout but on the
60 second timeout in some of the subtests.
Bump all these timeouts up a bit.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250221140640.786341-1-peter.maydell@linaro.org>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Put the tests into a separate file now (in the functional framework,
each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com
Message-ID: <20250218152744.228335-14-thuth@redhat.com>
Put the tests into a separate file now (in the functional framework,
each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-13-thuth@redhat.com>
Put the tests into a separate file now (in the functional framework,
each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-12-thuth@redhat.com>
Put the tests into a separate file now (since in the functional
framework, each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-11-thuth@redhat.com>
Put the tests into a separate file now (since in the functional
framework, each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-10-thuth@redhat.com>
Put the tests into a separate file now (since in the functional
framework, each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-9-thuth@redhat.com>
Put the tests into a separate file now (since in the functional
framework, each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-8-thuth@redhat.com>
Put the tests into a separate file now (since in the functional
framework, each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-7-thuth@redhat.com>
Put the tests into a separate file now (since in the functional
framework, each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-6-thuth@redhat.com>
Put the tests into a separate file now (since in the functional
framework, each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-5-thuth@redhat.com>
While we're at it, change the machine from SS-20 to SS-10 to
increase the test coverage a little bit (SS-20 is already
tested in the test_sparc_sun4m.py file).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-4-thuth@redhat.com>
Put the tests into a separate file now (since in the functional
framework, each file is run with one specific qemu-system-* binary).
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-3-thuth@redhat.com>
With a proper name the log files get a more meaningful name.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250218152744.228335-2-thuth@redhat.com>
Have the MicroblazeMachine class being common to both
MicroblazeBigEndianMachine and MicroblazeLittleEndianMachine
classes. Move the xmaton and ballerina tests to the parent class.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20250212123659.52764-11-philmd@linaro.org>
[thuth: Add missing ASSET statements to the leaf classes]
Signed-off-by: Thomas Huth <thuth@redhat.com>
Features:
SR-IOV emulation for pci
virtio-mem-pci support for s390
interleave support for cxl
big endian support for vdpa svq
new QAPI events for vhost-user
Also vIOMMU reset order fixups are in.
Fixes, cleanups all over the place.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
-----BEGIN PGP SIGNATURE-----
iQFDBAABCAAtFiEEXQn9CHHI+FuUyooNKB8NuNKNVGkFAme4b8sPHG1zdEByZWRo
YXQuY29tAAoJECgfDbjSjVRpHKcIAKPJsVqPdda2dJ7b7FdyRT0Q+uwezXqaGHd4
7Lzih1wsxYNkwIAyPtEb76/21qiS7BluqlUCfCB66R9xWjP5/KfvAFj4/r4AEduE
fxAgYzotNpv55zcRbcflMyvQ42WGiZZHC+o5Lp7vDXUP3pIyHrl0Ydh5WmcD+hwS
BjXvda58TirQpPJ7rUL+sSfLih17zQkkDcfv5/AgorDy1wK09RBKwMx/gq7wG8yJ
twy8eBY2CmfmFD7eTM+EKqBD2T0kwLEeLfS/F/tl5Fyg6lAiYgYtCbGLpAmWErsg
XZvfZmwqL7CNzWexGvPFnnLyqwC33WUP0k0kT88Y5wh3/h98blw=
=tej8
-----END PGP SIGNATURE-----
Merge tag 'for_upstream' of https://git.kernel.org/pub/scm/virt/kvm/mst/qemu into staging
virtio,pc,pci: features, fixes, cleanups
Features:
SR-IOV emulation for pci
virtio-mem-pci support for s390
interleave support for cxl
big endian support for vdpa svq
new QAPI events for vhost-user
Also vIOMMU reset order fixups are in.
Fixes, cleanups all over the place.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
# -----BEGIN PGP SIGNATURE-----
#
# iQFDBAABCAAtFiEEXQn9CHHI+FuUyooNKB8NuNKNVGkFAme4b8sPHG1zdEByZWRo
# YXQuY29tAAoJECgfDbjSjVRpHKcIAKPJsVqPdda2dJ7b7FdyRT0Q+uwezXqaGHd4
# 7Lzih1wsxYNkwIAyPtEb76/21qiS7BluqlUCfCB66R9xWjP5/KfvAFj4/r4AEduE
# fxAgYzotNpv55zcRbcflMyvQ42WGiZZHC+o5Lp7vDXUP3pIyHrl0Ydh5WmcD+hwS
# BjXvda58TirQpPJ7rUL+sSfLih17zQkkDcfv5/AgorDy1wK09RBKwMx/gq7wG8yJ
# twy8eBY2CmfmFD7eTM+EKqBD2T0kwLEeLfS/F/tl5Fyg6lAiYgYtCbGLpAmWErsg
# XZvfZmwqL7CNzWexGvPFnnLyqwC33WUP0k0kT88Y5wh3/h98blw=
# =tej8
# -----END PGP SIGNATURE-----
# gpg: Signature made Fri 21 Feb 2025 20:21:31 HKT
# gpg: using RSA key 5D09FD0871C8F85B94CA8A0D281F0DB8D28D5469
# gpg: issuer "mst@redhat.com"
# gpg: Good signature from "Michael S. Tsirkin <mst@kernel.org>" [full]
# gpg: aka "Michael S. Tsirkin <mst@redhat.com>" [full]
# Primary key fingerprint: 0270 606B 6F3C DF3D 0B17 0970 C350 3912 AFBE 8E67
# Subkey fingerprint: 5D09 FD08 71C8 F85B 94CA 8A0D 281F 0DB8 D28D 5469
* tag 'for_upstream' of https://git.kernel.org/pub/scm/virt/kvm/mst/qemu: (41 commits)
docs/devel/reset: Document reset expectations for DMA and IOMMU
hw/vfio/common: Add a trace point in vfio_reset_handler
hw/arm/smmuv3: Move reset to exit phase
hw/i386/intel-iommu: Migrate to 3-phase reset
hw/virtio/virtio-iommu: Migrate to 3-phase reset
vhost-user-snd: correct the calculation of config_size
net: vhost-user: add QAPI events to report connection state
hw/virtio/virtio-nsm: Respond with correct length
vdpa: Fix endian bugs in shadow virtqueue
MAINTAINERS: add more files to `vhost`
cryptodev/vhost: allocate CryptoDevBackendVhost using g_mem0()
vhost-iova-tree: Update documentation
vhost-iova-tree, svq: Implement GPA->IOVA & partial IOVA->HVA trees
vhost-iova-tree: Implement an IOVA-only tree
amd_iommu: Use correct bitmask to set capability BAR
amd_iommu: Use correct DTE field for interrupt passthrough
hw/virtio: reset virtio balloon stats on machine reset
mem/cxl_type3: support 3, 6, 12 and 16 interleave ways
hw/mem/cxl_type3: Ensure errp is set on realization failure
hw/mem/cxl_type3: Fix special_ops memory leak on msix_init_exclusive_bar() failure
...
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
When a machine is first booted, all virtio balloon stats are initialized
to their default value -1 (18446744073709551615 when represented as
unsigned).
They remain that way while the firmware is loading, and early phase of
guest OS boot, until the virtio-balloon driver is activated. Thereafter
the reported stats reflect the guest OS activity.
When a machine reset is performed, however, the virtio-balloon stats are
left unchanged by QEMU, despite the guest OS no longer updating them,
nor indeed even still existing.
IOW, the mgmt app keeps getting stale stats until the guest OS starts
once more and loads the virtio-balloon driver (if ever). At that point
the app will see a discontinuity in the reported values as they sudden
jump from the stale value to the new value. This jump is indigituishable
from a valid data update.
While there is an "last-updated" field to report on the freshness of
the stats, that does not unambiguously tell the mgmt app whether the
stats are still conceptually relevant to the current running workload.
It is more conceptually useful to reset the stats to their default
values on machine reset, given that the previous guest workload the
stats reflect no longer exists. The mgmt app can now clearly identify
that there are is no stats information available from the current
executing workload.
The 'last-updated' time is also reset back to 0.
IOW, on every machine reset, the virtio stats are in the same clean
state they were when the macine first powered on.
A functional test is added to validate this behaviour with a real
world guest OS.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20250204094202.2183262-1-berrange@redhat.com>
Acked-by: David Hildenbrand <david@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
