Commit 500131154d ("exec.c: Add new address_space_ld*/st*
functions") added a new API to fix a shortcoming of the
ld/st*_phys() API, which does blind bus access, not reporting
failure (and it also allow to provide transaction attributes).
Later commit 42874d3a8c ("Switch non-CPU callers from ld/st*_phys
to address_space_ld/st*") automatically converted the legacy uses
to the new API, not precising transaction attributes
(MEMTXATTRS_UNSPECIFIED) and ignoring the transation result (passing
NULL pointer as MemTxResult).
While this is a faithful replacement, without any logical change,
we later realized better is to not use MEMTXATTRS_UNSPECIFIED or
NULL MemTxResult, and adapt each call site on a pair basis, looking
at the device model datasheet to do the correct behavior (which is
unlikely to ignore transaction failures).
Since this is quite some work, we defer that to device model
maintainers. Meanwhile we introduce a definition, to allow a
target which removed all legacy API call to prohibit further
legacy API uses, named "TARGET_NOT_USING_LEGACY_LDST_PHYS_API".
Since all targets should be able to check this definition, we
take care to not poison it.
Suggested-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20251224151351.86733-6-philmd@linaro.org>
AioContexts are used as a generic event loop even outside the block
layer; move the header file out of block/ just like the implementation
is in util/.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
With the bump to Meson 1.10.0, C objects can be passed to rust targets.
This way, the Rust libstd will be added by rustc itself in its final
linker invocation. Use that to eliminate the staticlib and allow
dynamic linking with libstd (also introduced by Meson 1.9.0, but not
for staticlib crates due to lack of support in rustc).
The main() function is still provided by C, which is possible by
declaring the main source file of the Rust executable (which is
still created by scripts/rust/rust_root_crate.sh) as #![no_main].
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Implement DTrace/SystemTap SDT by emitting the following:
- The probe crate's probe!() macro is used to emit a DTrace/SystemTap
SDT probe.
- Every trace event gets a corresponding trace_<name>_enabled() -> bool
generated function that Rust code can use to avoid expensive
computation when a trace event is disabled. This API works for other
trace backends too.
`#[allow(dead_code)]` additions are necessary for QEMU's dstate in
generated trace-<dir>.rs files since they are unused by the dtrace
backend. `./configure --enable-trace-backends=` can enable multiple
backends, so keep it simple and just silence the warning instead of
trying to detect the condition when generating the dstate code can be
skipped.
The tracetool tests are updated. Take a look at
tests/tracetool/dtrace.rs to see what the new generated code looks like.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Link: https://lore.kernel.org/r/20251119205200.173170-5-stefanha@redhat.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The probe crate (https://crates.io/crates/probe) provides a probe!()
macro that defines SystemTap SDT probes on Linux hosts or does nothing
on other host OSes.
This crate will be used to implement DTrace support for Rust.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Link: https://lore.kernel.org/r/20251119205200.173170-4-stefanha@redhat.com
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Simplify rustc_args.py, and align its code with what Meson's own Cargo.toml
translator does in v1.10.
Bump unknown_lints to "forbid", so that it will certainly override Cargo.toml's
"allow" level.
Reviewed-by: Zhao Liu <zhao1.liu@intel.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The script has not been updated to read mapped-ram snapshots and is currently
crashing when trying to read such a file.
With this commit, it can now read a snapshot created with:
(qemu) migrate_set_capability x-ignore-shared on
(qemu) migrate_set_capability mapped-ram on
(qemu) migrate -d file:vm.state
Signed-off-by: Pawel Zmarzly <pzmarzly0@gmail.com>
Link: https://lore.kernel.org/r/20251126155015.941129-1-pzmarzly0@gmail.com
[peterx: space fixes, introduce parseMappedRamBlob(), add comments, etc.]
Signed-off-by: Peter Xu <peterx@redhat.com>
It has been renamed on the C side a few years ago. In modern QEMU versions,
fill_byte must be zero. Updating the Python script to make grepping and
understanding the code easier.
Signed-off-by: Pawel Zmarzly <pzmarzly0@gmail.com>
Link: https://lore.kernel.org/r/20251125173007.245607-1-pzmarzly0@gmail.com
[peterx: fix over-long line]
Signed-off-by: Peter Xu <peterx@redhat.com>
This fixes some settings like the default installation path
for the QEMU installation on Windows on ARM (WoA).
Signed-off-by: Stefan Weil <sw@weilnetz.de>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20251209130212.764443-1-sw@weilnetz.de>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
sys.stderr.print is dropped long ago and should not be used. Official
replacement is sys.stderr.write
The problem has been found debugging building on some fancy platform
derived from Debian.
Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: John Snow <jsnow@redhat.com>
CC: Cleber Rosa <crosa@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20251203220138.159656-1-den@openvz.org>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Missed a spot with isort, which now causes the python-minreqs test on
GitLab to fail. Fix it.
(Hint: the commands in python/tests/qapi-isort.sh can be run without the
"-c" parameter to automatically adjust import statements according to
our style rules. Maybe I should make a pre-submit hook that makes this
adjustment automatically. What do you think?)
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/3200
Fixes: 5bd89761a4 ("qapi/command: Avoid generating unused qmp_marshal_output_T")
Signed-off-by: John Snow <jsnow@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20251118200657.1043688-4-jsnow@redhat.com>
Pylint 4.x wants to use a different regex for _Stub for some reason;
just silence this.
Signed-off-by: John Snow <jsnow@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20251118200657.1043688-3-jsnow@redhat.com>
There are just two of them (slow and thorough; quick is simply the
default). Avoid repeating them for as many times as there are tests.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Thorough tests may have more dependencies than faster ones.
Dependencies are now looked up based on the suites being
executed, not on the suites passed as goals to the makefile.
Therefore, it is possible to limit dependencies to the
speeds that need them.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Remove the "--suite" argument from the .*.mtest-suites variables, and
add it only when actually computing the arguments to "meson test".
This makes it possible to set ninja-cmd-goals from the set of suites,
instead of doing it via many different .ninja-goals.* variables.
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
We need coreutils to run the IO tests so we need to include it in the
package list. Now we have the latest libvirt we can do that.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20251117115523.3993105-5-alex.bennee@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
re.match(r'^ *', ...) can't fail, but mypy doesn't know that and
complains:
scripts/qapi/parser.py:444: error: Item "None" of "Match[str] | None" has no attribute "end" [union-attr]
Work around by using must_match() instead.
Fixes: 8107ba47fd (qapi: Add documentation format validation)
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20251105152219.311154-1-armbru@redhat.com>
As an eMMC block device image may consist of more than just the user
data partition, provide a helper script that can compose the image from
boot partitions, an RPMB partition and the user data image. The script
also does the required size validation and/or rounding.
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Tested-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <eecefa8e-44ae-45ff-85d0-3f9b786948e0@siemens.com>
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Ubuntu is now including updated versions of Rust (up to 1.85) for
its LTS releases. Adjust the CI containers and re-add --enable-rust
to the Ubuntu jobs.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add explicit validation for QAPI documentation formatting rules:
1. Lines must not exceed 70 columns in width (including '# ' prefix)
2. Sentences must be separated by two spaces
Example sections and literal :: blocks (seldom case) are excluded, we
don't require them to be <= 70, that would be too restrictive. Anyway,
they share common 80-columns recommendations (not requirements).
Add two simple tests, illustrating the change.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Message-ID: <20251031183129.246814-1-vsementsov@yandex-team.ru>
The detection of example and literal blocks isn't quite correct, but
it works well enough, and we can improve on top.
Reviewed-by: Markus Armbruster <armbru@redhat.com>
[Comments, error messages, and test file names tweaked]
Signed-off-by: Markus Armbruster <armbru@redhat.com>
qmp_marshal_output_T() is only ever called by qmp_marshal_C() for a
command C that returns type T.
We've always generated it as a static function on demand, i.e. when we
generate a call.
Since we split up monolithic generated code into modules (commit
252dc3105f "qapi: Generate separate .h, .c for each module"), we do
this per module. As noted in the commit message, this can result in
identical (static) qmp_marshal_output_T() in several modules. Was
deemed not worth avoiding.
A bit later, we added 'if' conditionals to the schema language (merge
commit 5dafaf4fbc).
When a conditional definition uses a type, then its condition must
imply the type's condition. We made this the user's responsibility.
Hasn't been an issue in practice.
However, the sharing of qmp_marshal_output_T() among commands
complicates matters. To avoid both undefined function errors and
unused function warnings, qmp_marshal_output_T() must be defined
exactly when it's used. It is used when any of the qmp_marshal_C()
calling it is defined, i.e. when any C's condition holds.
The generator uses T's condition instead. To avoid both error and
warning, T's condition must be the conjunction of all C's conditions.
Unfortunately, this can be impossible:
* Conditional command returning a builtin type
A builtin type cannot be conditional. This is noted in a FIXME
comment.
* Commands in multiple modules where the conjunction differs between
modules
An instance of this came up recently. we have unconditional
commands returning HumanReadableText. If we add a conditional one
to a module that does not have unconditional ones, compilation fails
with "defined but not used". If we make HumanReadableText
conditional to fix this module, we break the others.
Instead of complicating the code to compute the conjunction, simplify
it: generate the output marshaling code right into qmp_marshal_C().
This duplicates it when multiple commands return the same type. The
impact on code size is negligible: qemu-system-x86_64's text segment
grows by 1448 bytes.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-ID: <20250804130602.903904-1-armbru@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
[Commit message typos fixed]
Remove the 'patch prefix exists, appears to be a -p0 patch' warning
entirely as it is fundamentally flawed and can only produce false
positives.
Sometimes I create test files with names 'a' and 'b', and then get
surprised seeing this warning. It was not easy to understand where it
comes from.
How it works:
1. It extracts prefixes (a/, b/) from standard diff output
2. Checks if files/directories with these names exist in the project
root
3. Warns if they exist, claiming it's a '-p0 patch' issue
This logic is wrong because:
- Standard diff/patch tools always use a/ and b/ prefixes by default
- The existence of files named 'a' or 'b' in the working directory is
completely unrelated to patch format
- The working directory state may not correspond to the patch content
(different commits, branches, etc.)
- In QEMU project, there are no single-letter files/directories in root,
so this check can only generate false positives
The correct way to detect -p0 patches would be to analyze the path
format within the patch itself (e.g., absolute paths or paths without
prefixes), not check filesystem state.
So, let's finally drop it.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Link: https://lore.kernel.org/r/20251030201319.858480-1-vsementsov@yandex-team.ru
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The argparse.FileType() type has been deprecated in the latest argparse
version (e.g. the one from Fedora 43), now causing the test_bad_vmstate
functional test to fail since there are unexpected strings in the output.
Change the script to use pathlib.Path instead to fix the test_bad_vmstate
test and to be prepared for the future when the deprecated FileType gets
removed completely.
Reported-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Link: https://lore.kernel.org/r/20251030092638.39505-1-thuth@redhat.com
Signed-off-by: Peter Xu <peterx@redhat.com>
When the QMP library was updated to match the standalone repository in
094ded52, I neglected to update the logging filter(s) in
device-crash-test, which allowed the spurious messages to leak through.
Update the log filter to re-suppress these messages.
Fixes: 094ded52
Reported-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: John Snow <jsnow@redhat.com>
Message-ID: <20251022213109.395149-1-jsnow@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
With the current wording, users might think that the -F option is not
required as long as the script is placed in the default path. Be clear
that the option is always required. Also includes some minor language
improvements in the rest of the comment.
Signed-off-by: Fiona Ebner <f.ebner@proxmox.com>
Reviewed-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
Link: https://lore.kernel.org/qemu-devel/20250916122111.36019-1-f.ebner@proxmox.com
Signed-off-by: Kostiantyn Kostiuk <kkostiuk@redhat.com>
All the functions are about "-audio model=" handling, a simpler
way to setup audio. Rename functions/variables to reflect this better.
audio_register_model_with_cb() dropped "pci" from the name, since it
will be generalized next.
deprecated_register_soundhw() was actually not a function to be
removed since it's used for "-audio model=" aliasing.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Both Debian and Ubuntu are setup the same way.
Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Message-ID: <20251027110344.2289945-7-alex.bennee@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Unlike the Aarch64 runners this comes with pure Debian out of the box.
We need a minor tweak to build-environment to deal with the
differences in naming convention.
Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Message-ID: <20251027110344.2289945-6-alex.bennee@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
There is no practical difference between the systems when it comes to
updating the installed system.
Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Message-ID: <20251027110344.2289945-5-alex.bennee@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
We can share the setup of the build environment with multiple
operating systems as we just need to check the YAML for each env is
present in the directory structure.
Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Message-ID: <20251027110344.2289945-4-alex.bennee@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
We inadvertently updated the base libvirt-ci project which has
resulted in changes. Make sure the output matches what we generate.
Fixes: 0d4fb8f746 (configure: set the bindgen cross target)
Reviewed-by: Manos Pitsidianakis <manos.pitsidianakis@linaro.org>
Message-ID: <20251027110344.2289945-3-alex.bennee@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
We no longer need to support 32 bit builds and we missed this while
cleaning up.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-ID: <20251027110344.2289945-2-alex.bennee@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
While working out what hoops to jump through to get a full set of
aarch32 packages installed on the aarch64 runner it was pointed out 32
bit host support is deprecated. As the extra packages where needed for
system emulation (marked deprecated since 8.0!) there didn't seem much
point keeping this in.
While the full expunging of 32 bit host support will probably be done
for 11.0 we can at least reduce the CI burden a bit now.
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Message-ID: <20251016150357.876415-3-alex.bennee@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
The download cache of the functional tests is currently only growing.
But sometimes tests get removed or changed to use different assets,
thus we should clean up the stale old assets after a while when they
are not in use anymore. So add a script that looks at the time stamps
of the assets and removes them if they haven't been touched for more
than half of a year. Since there might also be some assets around that
have been added to the cache before we added the time stamp files,
assume a default time stamp that is close to the creation date of this
patch, so that we don't delete these files too early (so we still have
all assets around in case we have to bisect an issue in the recent past
of QEMU).
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Thomas Huth <thuth@redhat.com>
Message-ID: <20251014083424.103202-3-thuth@redhat.com>
Introduce headers for the Microsoft Hypervisor (MSHV) userspace ABI,
including IOCTLs and structures used to interface with the hypervisor.
These definitions are based on the upstream Linux MSHV interface and
will be used by the MSHV accelerator backend in later patches.
Signed-off-by: Magnus Kulke <magnuskulke@linux.microsoft.com>
Link: https://lore.kernel.org/r/20250916164847.77883-6-magnuskulke@linux.microsoft.com
[Do not use __uN types. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Introduce a Meson feature option and default-config entry to allow
building QEMU with MSHV (Microsoft Hypervisor) acceleration support.
This is the first step toward implementing an MSHV backend in QEMU.
Signed-off-by: Magnus Kulke <magnuskulke@linux.microsoft.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Link: https://lore.kernel.org/r/20250916164847.77883-2-magnuskulke@linux.microsoft.com
[Add error for unavailable accelerator. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The probe of GDB is done in 'configure' and the full path is passed to
meson.build via the -Dgdb=option.
Meson then can pass the location of GDB to the functional tests via an
environment variable: QEMU_TEST_GDB.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Gustavo Romero <gustavo.romero@linaro.org>
Message-ID: <20251003141820.85278-4-gustavo.romero@linaro.org>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
The registration-token method is being deprecated:
https://docs.gitlab.com/ci/runners/new_creation_workflow/
As a result we can drop a bunch of the descriptive stuff as that is
entered on the web UI. We don't need a secondary runner if we just
create one with both aarch64 and aarch32 tags.
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20251001170947.2769296-8-alex.bennee@linaro.org>
In anticipation of new runners lets move to a newer Ubuntu LTS.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
Message-ID: <20251001170947.2769296-6-alex.bennee@linaro.org>
The legacy cpu_physical_memory_rw() method is no more used,
remove it.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20251002084203.63899-16-philmd@linaro.org>
Following the mechanical changes of commit adeefe0167 ("Avoid
cpu_physical_memory_rw() with a constant is_write argument"),
replace:
- cpu_physical_memory_rw(, is_write=false) -> address_space_read()
- cpu_physical_memory_rw(, is_write=true) -> address_space_write()
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20251002084203.63899-15-philmd@linaro.org>
qemu_bh_new_guarded() is considered legacy since commit 9c86c97f12
("async: Add an optional reentrancy guard to the BH API"); recommend
the new API: aio_bh_new_guarded().
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-ID: <20250924163911.51479-1-philmd@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Such annotation is present into the kernel uAPI headers since
v6.7, and will be used soon by the vhost_type.h. Deal with it
just stripping it.
Reviewed-by: Akihiko Odaki <odaki@rsg.ci.i.u-tokyo.ac.jp>
Acked-by: Jason Wang <jasowang@redhat.com>
Acked-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Tested-by: Lei Yang <leiyang@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Message-ID: <a1430f43cc954d2a931fa60581bda6d6af4bc771.1758549625.git.pabeni@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Philippe Mathieu-Daudé
Paolo Bonzini
Stefan Hajnoczi
Pawel Zmarzly
Stefan Weil
Denis V. Lunev
John Snow
Djordje Todorovic
Alex Bennée
Markus Armbruster
Jan Kiszka
Vladimir Sementsov-Ogievskiy
Thomas Huth
Marc-André Lureau
Fiona Ebner
Magnus Kulke
Gustavo Romero
Mauro Carvalho Chehab
Paolo Abeni