WirePlumber 0.5.8
~~~~~~~~~~~~~~~~~
Additions & Enhancements:
- Added support for handling UCM SplitPCM nodes in the ALSA monitor, which
allows native PipeWire channel remapping using loopbacks for devices that
use this feature (!685)
- Introduced new functions to mark WpSpaDevice child objects as pending.
This allows properly associating asynchronously created loopback nodes with
their parent WpSpaDevice without losing ObjectConfig events (!687, !689)
- Improved the node name deduplication logic in the ALSA monitor to prevent
node names with .2, .3, etc appended to them in some more cases (!688)
- Added a new script to populate ``session.services``. This is a step towards
implementing detection of features that PipeWire can service (!686)
Fixes:
- Fixed an issue that was causing duplicate Bluetooth SCO (HSP/HFP) source
nodes to be shown in UIs (#701, !683)
- In the BlueZ monitor, marked the source loopback node as non-virtual,
addressing how it appears on UIs (#729)
- Disabled stream-restore for device loopback nodes to prevent unwanted
property changes (!691)
- Fixed ``wp_lua_log_topic_copy()`` to correctly copy topic names (#757)
- Updated script tests to handle differences in object identifiers
(``object.serial`` vs ``node.id``), ensuring proper test behavior (#761)
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
- remove backport patch
- install libspa.so to libpipewire package
PipeWire 1.4.0 (2025-03-06)
This is the 1.4 release that is API and ABI compatible with previous
1.2.x and 1.0.x releases.
This release contains some of the bigger changes that happened since
the 1.2 release last year, including:
* client-rt.conf was removed, all clients now use client.conf and
are given RT priority in the data threads.
* UMP (aka MIDI2) support was added and is now the default format
to carry MIDI1 and MIDI2 around in PipeWire. There are helper
functions to convert between legacy MIDI and UMP.
* The resampler can now precompute (at compile time) some common
conversion filters. Delay reporting in the resampler was fixed and
improved.
* Bluetooth support for BAP broadcast links and support for hearing aids
using ASHA was added. A new G722 codec was also added.
Delay reporting and configuration in Bluetooth was improved.
* The ALSA plugin now supports DSD playback when explicitly allowed
with the alsa.formats property.
* A PipeWire JACK control API was added.
* A system service was added for pipewire-pulse.
* Many documentation and translation updates.
* Many of the SPA macros are converted to inline functions. All SPA
inline functions are now also compiled into a libspa.so library to
make it easier to access them from bindings.
* The module-filter-chain graph code was moved to a separate
filter-graph SPA plugin so that it becomes usable in more places.
EBUR128, param_eq and dcblock plugins were added to filter-graph.
The filter graph can now also use fftw for doing convolutions.
The audioconvert plugin was optimized and support was added to
audioconvert to insert extra filter-graphs in the processing pipeline.
* New helper functions were added to parse JSON format descriptions.
* The profiler now also includes the clock of the followers.
* RISCV CPU support and assembler optimisations were added.
* The clock used for logging timestamps can be configured now.
* The JSON parser was split into core functions and helper.
* Support for UCM split PCMs was added. Instead of alsa-lib splitting
up PCMs, PipeWire can mark the PCMs with the correct metadata so that
the session manager can use native PipeWire features to do this.
* Support for webrtc2 was added to echo-cancel.
* IEC958 codecs are now detected from the HDMI ELD data.
* Conversion between floating point and 32 bits now preserve 25 bits of
precision instead of 24 bits.
* A new Telephony D-BUS API compatible with ofono was added.
* The invoke queues are now more efficient and can be called from multiple
threads concurrently.
* Clock information in v4l2 was improved.
* An ffmpeg based videoconvert plugin was added that can be used with the
videoadapter.
* The GStreamer elements have improved buffer pool handling and rate
matching.
* The combine-stream module can now also mix streams.
* link-factory now checks that the port and node belong together.
* The netjack-manager module has support for autoconnecting streams.
* The native-protocol has support for abstract sockets.
* The pulse server has support for blocking playback and capture in
pulse.rules.
* The corked state of stream is now reported correctly in pulse-server.
* Fix backwards jumps in pulse-server.
* Latency configuration support was added in loopback and raop-sink.
* The ROC module has more configuration options.
* The SAP module now only send updated SDP when something changed.
* RTP source now has a standby mode where it idles when there is no
data received.
* Support for PTP clocking was added the RTP streams.
* The VBAN receiver can now dynamically create streams when they are
detected.
* Error reporting when making links was improved.
* Support for returning (canceling) a dequeued buffer in pw-stream.
* Support for emiting events in pw-stream was added.
* pw-cat now support stdin and stdout.
Highlights (since the previous 1.3.83 release)
- Small fixes and improvements.
PipeWire
- Fix some missing includes in metadata.h
- Pass the current error in errno when a stream is in error (#4574)
modules
- Evaluate node rules before loading adapter follower to ensure
properties are set correctly. (#4562)
SPA
- Avoid a use after free when building PODs. (#4445)
- Take headroom into account when calculating resync.
Bluetooth
- Fix +CLCC parsing.
GStreamer
- Notify about default device changes in deviceprovider.
- Copy frames between pools and avoid splitting video buffers.
JACK
- Add an option to disable the MIDI2 port flags. (#4584)
Older versions:
PipeWire 1.3.83 (2025-02-20)
This is the third and hopefully last 1.4 release candidate that
is (almost) API and (entirely) ABI compatible with previous 1.2.x
and 1.0.x releases.
We note that in the 1.3.x series, the API is slighty not backwards
compatible because some methods previously used to accept void* as
a parameter while they now require the correct type. We think this
is however a good kind of API breakage and expect projects to patch
their code to get things compiled with newer version (which will also
compile for older versions). Note also that this is not an ABI break.
- Handle JACK transport updates in a better way.
- Fix a SAP regression when starting.
- Fix regression in rate scaling.
- Improve bluetooth source rate handling.
- More small bugfixes and improvements.
- Handle JACK transport updates in a better way. (#4543)
- Check that the link factory port and nodes match. Deprecate the
port.id when making links.
- Improve profiler output by scaling the quantum with the node
rate so that we don't end up with confusing information. (#4555)
- Fix sending of the SAP SDP. Handle some SDP parsing errors.
- Add some more options to the ROC source module. (#4516)
- Fix firewire quirks in udev rules. (#4528)
- Fix a bug in the rate scaling in some cases that would make things
run with the wrong samplerate.
- Improve introspection of control types.
- Use the G722 codec from Android instead of FFmpeg for ASHA.
- Use the A2DP source rate as the graph rate. (#4555)
- Specify the bluetooth source latency property in the rate of the
stream to avoid conversions and rounding errors.
This is the second 1.4 release candidate that is API and ABI
compatible with previous 1.2.x and 1.0.x releases.
- Various pw-stream improvements: timing information fixes,
avoid locking buffers in some cases and an improved drain
event.
- A new Telephony D-BUS API compatible with ofono.
- Documentation fixes and updates.
- More small fixes and improvements.
- Improve timing information when rate is unknown.
- Avoid locked buffers in pw_stream in some cases.
- Improve pw_stream drain event emission.
- Improve manager socket handling. Applications can avoid hardcoding
the sockets so that they will respect the config settings.
- Fix header size calculation when using ipv6. (#4524)
- Optimize byteswapped s16 conversions.
- Improve event handling for internal events.
- Optimize negiotiation when in convert mode, prefer the format
of the follower in adapter.
- Fix EnumPortConfig for videoadapter without converter.
- Fix libcamera property buffer size.
- Add systemwide systemd files.
- Add a UMP example.
- Use the new JackPortMIDI2 flag to mark UMP ports to JACK.
- Support BAP hardware volume.
- Add a Telephony DBUS API.
- Disable buffer pools for audio by default.
- Improve the module documentation.
This is the first 1.4 release candidate that is API and ABI
compatible with previous 1.2.x and 1.0.x releases.
In addition to all the changes backported to 1.2.x, this release
also contains some new features:
- UMP support was added with MIDI 1.0 and MIDI 2.0 support in the ALSA
sequencer plugin. By default PipeWire will now use MIDI 2.0 in UMP
messages to transport MIDI in the graph, with conversions to/from legacy
MIDI where required. This requires UMP support in the kernel.
- client-rt.conf is no longer supported. Custom changes made to this
config should be moved to client.conf. Clients that try to load the
client-rt.conf will emit a warning and be directed to client.conf
automatically for backwards compatibility.
- The module-filter-chain code was moved to a new filter-graph plugin. This
made it possible to add filter-graph support directly in audioconvert. It
is now possible to run up to 8 run-time swappable filter-graphs inside
streams and nodes. This should make it easier to add effects to streams
and device nodes.
- Bluetooth support for BAP broadcast links and support for hearing aids
using ASHA was added.
- Many more bugfixes and improvements.
- Nodes are now only scheduled when ready to signal the driver.
- Add slovenian translation. (#4156)
- Link errors are handled better.
- The videoadapter is now enabled by default but no videoconverter
is loaded yet by default.
- Streams now have support for ProcessLatency.
- Streams now have a method to emit events.
- The RequestProcess event and command can now pass around extra
properties.
- Local timestamps are now used for logging.
- client-rt.conf is no longer supported. Custom changes made to this
config should be moved to client.conf. Clients that try to load the
client-rt.conf will emit a warning and be directed to cliert.conf
automatically to preserve backwards compatibility.
- pw_stream now has an API to return unused buffers.
- module-combine-stream can now mix streams.
- Links in error are now destroyed by link-factory.
- The netjack2 driver can now also create streams that autoconnect when
specified. (#4125)
- Many updates and bugfixes to the RTP modules.
- The netjack2 driver can now bind to a custom IP and port pair. (#4144)
- The loopback module and module-raop have support for ProcessLatency, which
can be used to query and update the latency.
- The profiler module can now reduce the sampling rate.
- The filter-chain was optimized some more.
- The filter-chain gained some more plugins: param_eq, ebur128, dcblock.
- Support for fftw based convolver was added.
- Some module arguments can now be overridden.
- The VBAN receiver now creates new streams per stream name. (#4400)
- The RTP SAP module is now smarter with generating new SAP messages.
- The RTP source can now be paused when no data is received. (#4456)
- pw-cat can now stream most formats from stdin/stdout.
- pw-profiler has a JSON dump option to dump the raw profiler data.
- pw-cli now supports unload-module. (#4276)
- The resampler can precompute some common coeficients now at compile
time.
- UMP support was added with MIDI 1.0 and MIDI 2.0 support in the ALSA
sequencer plugin. By default PipeWire will now use MIDI 2.0 in UMP
messages to transport MIDI in the graph, with conversions to/from legacy
MIDI where required.
- Control types can now be negotiated.
- Support for writing ALSA bind controls was added.
- The ALSA sequencer now has better names for the ports.
- The F32 to S32 conversion now uses 25 bits for an extra bit of
precision.
- libcamera controls can now be set in all cases.
- The videoadapter has been improved and a dummy and ffmpeg based
videoconverter plugin was added.
- Negotiation was improved in audioadapter. First a passthrough format
is tried.
- Some JSON helper functions were added and some duplicate code removed
or simplified.
- Add support for RISC V CPU detection and add many optimizations in
the audio converters.
- Add an option to disable ALSA mixer path select. (#4311)
- Fix a potential bug with the cleanup of the loop queues.
- ALSA nodes now dynamically adjust the DLL bandwidth based on average
measured variance.
- The loop invoke queue was made more efficient and make it possible to
invoke from multiple threads.
- The filter-chain code was moved to a new filter-graph plugin.
- Most function macros are now static inlined and can also be built into a
libspa.so file. This should improve language bindings.
- V4l2 clock information was improved.
- Supported IEC958 codecs are now autodetected via ELD info.
- Audioconvert was optimized some more.
- Audioconvert can now include filter-graphs in its processing.
- webrtc-audio-processing-2 is now supported in AEC.
- The resampler now reports the delay and subsample delay. Also the
delay is reported in the samplerate of the input.
- The ALSA sequencer now handle kernels without UMP support. (#4507)
- Add quirk to block clients from making record and playback streams.
- The corked state is now set on stream to always report this state
correctly to other clients.
- Readiness notification was added to the pulse server with the
PIPEWIRE_PULSE_NOTIFICATION_FD environment variable. (#4347)
- The pulse.cmd config now supports conditions.
- A bug in clearing the ringbuffer was fixed. (#4464)
- Support for the default devices was added to the deviceprovider. (#4268)
- The graph clock is now used as the source for the GStreamer clock.
- The sink now does some rate control.
- The ALSA plugin now supports DSD when explicitly enabled.
- JACK now supports 2 new extension formats for OSC and UMP.
- JACK clients can receive UMP MIDI1 or MIDI2 messages when using
the new UMP port format extension.
- JACK now reports the PipeWire version in the minor/micro/proto.
- Implement more jackserver functions.
- Support BAP broadcast links.
- Support for ASHA was added.
- Delay reporting in A2DP sources was improved.
- 2 new examples of pw-stream using spa_ringbuffer were added.
- Many updates to the man pages.
- More documentation about thread safety of functions in stream
and filters. (#4521)
Signed-off-by: Markus Volk <f_l_k@t-online.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 0.39.1:
- [fix] Make LightQueue and derivatives subscriptable
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 7.45.6:
- Update the wheels to incorporate libcurl 8.12.1 for security
fixes, as well as enable some additional libraries in wheel
builds.
- The previous release was accidentally built without CA bundle
autodetection in Linux wheels - this restores that behavior
(no changes to macOS or Linux).
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 2.22.4:
- Ensure dev-dependencies are added to the correct group when the
tool.pdm.dev-dependencies table has groups.
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
There is a bug in uutils-coreutils causing stdbuf to include the HOST version of libstdbuf.so, even when cross-compiled.
As a side-effect, the uutils-coreutils binary was containing the buildpath, because RPATH was set on libstdbuf.so.
Thus the buildpath error can be re-enabled.
This is tracked upstream by https://github.com/uutils/coreutils/issues/6591
Signed-off-by: Etienne Cordonnier <ecordonnier@snap.com>
Co-authored-by: Gyorgy Sarvari <skandigraun@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The library version to link against can be automatically detected.
This fixes the following runtime warning:
>>> import h5py
/usr/lib/python3.13/site-packages/h5py/__init__.py:36: UserWarning:
h5py is running against HDF5 1.14.4 when it was built against
1.14.0, this may cause problems
_warn(("h5py is running against HDF5 {0} when it was built against
{1}, "
Signed-off-by: Oliver Kästner <okaestner@rosen-nxt.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upstream moved to PEP-517 builds, so use the appropriate bbclass.
Signed-off-by: Oliver Kästner <okaestner@rosen-nxt.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* hardcoding:
${WORKDIR}/recipe-sysroot-native${libdir}
is just wrong when there is a better variable
* This path won't be correct for multilib builds where target ${libdir}
might be e.g. /usr/lib32 while native libdir is /usr/lib.
* use STAGING_LIBDIR_NATIVE variable which uses correct ${libdir_native}
* fixes:
error: failed to run custom build command for `devicemapper-sys v0.3.0`
Caused by:
process didn't exit successfully: `lib32-thin-provisioning-tools/1.1.0/build/target/release/build/devicemapper-sys-88de590bc55acb43/build-script-build` (exit status: 101)
--- stderr
thread 'main' panicked at lib32-thin-provisioning-tools/1.1.0/sources-unpack/cargo_home/bitbake/bindgen-0.69.4/lib.rs:622:31:
Unable to find libclang: "couldn't find any valid shared libraries matching: ['libclang.so', 'libclang-*.so', 'libclang.so.*', 'libclang-*.so.*'], set the `LIBCLANG_PATH` environment variable to a path where one of these files can be found (invalid: [])"
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 4.2.0:
- Support for Python 3.11 has been added.
- Support for Python 3.12 has been added.
- Support for Python 3.13 has been added.
- Add an ability to send outbound cookies separately to improve
headers compression.
- Updated packaging and testing infrastructure.
- Fix repr() checks for Python 3.11
- Fix asyncio / wsgi examples.
- Clarify docs on using curl with http2.
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 1.15.1:
- Fixed an issue in the new PEP 621 pyproject.toml layout that
prevented Alembic's template files from being included in the
.whl file in the distribution.
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 5.4.1:
- Skip installation to speed up pep8
- reno: Update master for unmaintained/2023.1
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Since module crypt has been removed from python 3.13, add a renamed copy
of it that other packages may require it, such as python3-anaconda.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
python 3.13 removed some modules such as cgi which is required by python3-requests-ftp:
https://docs.python.org/3/whatsnew/3.13.html
Important removals:
PEP 594: The remaining 19 “dead batteries” (legacy stdlib modules) have
been removed from the standard library: aifc, audioop, cgi, cgitb,
chunk, crypt, imghdr, mailcap, msilib, nis, nntplib, ossaudiodev, pipes,
sndhdr, spwd, sunau, telnetlib, uu and xdrlib.
Backport and rebase a patch from Fedora to remove use of module cgi.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
mkdir -p /build-dir/202503/build-poky/tmp/work/core2-64-poky-linux/dialog/1.3-20250116/image/usr/share/man/man1
/build-dir/202503/build-poky/tmp/hosttools/install -c dialog /build-dir/202503/build-poky/tmp/work/core2-64-poky-linux/dialog/1.3-20250116/image/usr/bin/`echo dialog |sed s,^,x86_64-poky-linux-,`
This change goes against established user habits. So, create a symbolic link named /usr/bin/dialog.
Signed-off-by: Lei Maohui <leimaohui@fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Add ptest for Wolfssl package.
Set IMAGE_ROOTFS_EXTRA_SPACE:virtclass-mcextend-wolfssl to 700M
enough to avoid a "No space left on device".
BEGIN: /usr/lib/wolfssl/ptest
Wolfssl ptest logs are stored in /tmp/wolfss_temp.qvuQ9h/ptest.log
Test script returned: 0
unit_test: Success for all configured tests.
PASS: Wolfssl
DURATION: 7
END: /usr/lib/wolfssl/ptest
Signed-off-by: Sofiane HAMAM <sofiane.hamam@smile.fr>
Reviewed-by: Yoann Congal <yoann.congal@smile.fr>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* don't seem to be needed, I don't see any failures or changes after removing them
* they were needed before in older version as explained by Etienne in:
https://lists.openembedded.org/g/openembedded-devel/message/115756
regarding SELINUX_INCLUDE_DIR, it was definitely needed as I implemented the recipe. During implementation I submitted this patch to selinux-sys ( https://github.com/koutheir/selinux-sys/commit/a6667c404ecc33266f66e9516bb8e30150c41ec2 ) to correctly use this variable, and was getting a build error without this fix because the rust crate selinux-sys was not finding the selinux header-file (specifically, the bindgen tool).
However, this was in 2022, and since then there has been this patch in selinux-sys: https://github.com/koutheir/selinux-sys/commit/d3ea4674e18c5240918348bea4ec6248e50516e5 , and it seems that the variable is not needed any more.
At least I'm not able to reproduce the build error which I was getting, even with commenting out SELINUX_LIB_DIR and SELINUX_INCLUDE_DIR.
The header-file is used for generating rust code from the C header-file using bindgen, so I don't think it makes a difference whether the native version or the target version is used.
This is the old build error I was getting:
process didn't exit successfully: `/workdir/build/tmp-glibc/work/armv8a-linux/uutils-coreutils/0.0.14.AUTOINC+43a5b8c7d7-r0/build/target/release/build/selinux-sys-0849198bc1fef4a4/build-script-build` (exit status: 101)
--- stdout
cargo:root=/workdir/build/tmp-glibc/work/armv8a-linux/uutils-coreutils/0.0.14.AUTOINC+43a5b8c7d7-r0/build/target/aarch64-snap-linux/release/build/selinux-sys-d96748d49908959f/out
cargo:rerun-if-env-changed=SELINUX_STATIC_aarch64-snap-linux
cargo:rerun-if-env-changed=SELINUX_STATIC_aarch64_snap_linux
cargo:rerun-if-env-changed=TARGET_SELINUX_STATIC
| --- stderr
| src/selinux-sys.h:1:10: fatal error: 'selinux/selinux.h' file not found
| src/selinux-sys.h:1:10: fatal error: 'selinux/selinux.h' file not found, err: true
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 2.9.0:
- Add warnings about too few or too many samples
- .github: Add ARM64 CI for pyperf
- Add support for profiling benchmarks using perf-record
- Prepare 2.9.0 release
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 5.2.0:
- :func:`unittest.mock.Mock.attach_mock` no longer triggers a call
to a PropertyMock being attached.
- Improved documentation for :func:`unittest.mock.patch.dict`
- Add support for :func:`dataclasses.dataclass` in
:func:`unittest.mock.create_autospec`. Now create_autospec will
check for potential dataclasses and use :func:`dataclasses.fields`
function to retrieve the spec information.
- Fix :class:`unittest.mock.MagicMock` reseting magic methods
return values after .reset_mock(return_value=True) was called.
- Fixed :func:`unittest.mock.create_autospec` to configure parent
mock with keyword arguments.
- Indicate if there were no actual calls in unittest
:meth:`~unittest.mock.Mock.assert_has_calls` failure.
- Deprecate :func:`!asyncio.iscoroutinefunction` in favor of
:func:`inspect.iscoroutinefunction`.
- Limit starting a patcher (from :func:`unittest.mock.patch` or
:func:`unittest.mock.patch.object`) more than once without
stopping it
- Fixed :func:`unittest.mock.create_autospec` to pass the call
through to the wrapped object to return the real result.
- Fix :func:`unittest.mock.patch` to not read attributes of the
target when new_callable is set. Patch by Robert Collins.
- Fix import of :mod:`unittest.mock` when CPython is built without
docstrings.
- Fix name passing to :class:`unittest.mock.Mock` object when using
:func:`unittest.mock.create_autospec`.
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 2.2.2:
- Bugfix where it was impossible to use a signal as an ok_code
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 3.11.13:
- Removed a break statement inside the finally block in
:py:class:~aiohttp.web.RequestHandler
- Changed connection creation to explicitly close sockets if an
exception is raised in the event loop's create_connection method
- Fixed test test_write_large_payload_deflate_compression_data_in_eof_writelines
failing with Python 3.12.9+ or 3.13.2+
- Added human-readable error messages to the exceptions for
WebSocket disconnects due to PONG not being received
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* use STAGING_LIBDIR_NATIVE variable for LIBCLANG_PATH and SELINUX_LIB_DIR
and STAGING_INCDIR_NATIVE variable for SELINUX_INCLUDE_DIR
not sure why it uses native version and not target, but at least use
correct variable
* hardcoding:
${WORKDIR}/recipe-sysroot-native${libdir}
is just wrong when there is a better variable
* This path won't be correct for multilib builds where target ${libdir}
might be e.g. /usr/lib32 while native libdir is /usr/lib.
* even with meta-selinux added to my builds I don't see any failure
maybe these variables aren't really used?
* use STAGING_LIBDIR_NATIVE variable which uses correct ${libdir_native}
OE qemuarm64-multilib@ /OE/build/oe-core $ bitbake-getvar LIBCLANG_PATH -r uutils-coreutils
#
# $LIBCLANG_PATH [2 operations]
# exported /OE/build/oe-core/meta-openembedded/meta-oe/recipes-core/uutils-coreutils/uutils-coreutils_0.0.29.bb:33
# [export] "1"
# set /OE/build/oe-core/meta-openembedded/meta-oe/recipes-core/uutils-coreutils/uutils-coreutils_0.0.29.bb:33
# "${WORKDIR}/recipe-sysroot-native${libdir}"
# pre-expansion value:
# "${WORKDIR}/recipe-sysroot-native${libdir}"
export LIBCLANG_PATH="/OE/build/oe-core/tmp/work/cortexa57-oe-linux/uutils-coreutils/0.0.29/recipe-sysroot-native/usr/lib64"
OE qemuarm64-multilib@ /OE/build/oe-core $ bitbake-getvar STAGING_LIBDIR_NATIVE -r uutils-coreutils
#
# $STAGING_LIBDIR_NATIVE
# set /OE/build/oe-core/openembedded-core/meta/conf/bitbake.conf:425
# "${STAGING_DIR_NATIVE}${libdir_native}"
STAGING_LIBDIR_NATIVE="/OE/build/oe-core/tmp/work/cortexa57-oe-linux/uutils-coreutils/0.0.29/recipe-sysroot-native/usr/lib"
OE qemuarm64-multilib@ /OE/build/oe-core $ bitbake-getvar LIBCLANG_PATH -r lib32-uutils-coreutils
#
# $LIBCLANG_PATH [2 operations]
# exported /OE/build/oe-core/meta-openembedded/meta-oe/recipes-core/uutils-coreutils/uutils-coreutils_0.0.29.bb:33
# [export] "1"
# set /OE/build/oe-core/meta-openembedded/meta-oe/recipes-core/uutils-coreutils/uutils-coreutils_0.0.29.bb:33
# "${WORKDIR}/recipe-sysroot-native${libdir}"
# pre-expansion value:
# "${WORKDIR}/recipe-sysroot-native${libdir}"
export LIBCLANG_PATH="/OE/build/oe-core/tmp/work/armv7at2-neon-oemllib32-linux-gnueabi/lib32-uutils-coreutils/0.0.29/recipe-sysroot-native/usr/lib32"
OE qemuarm64-multilib@ /OE/build/oe-core $ bitbake-getvar STAGING_LIBDIR_NATIVE -r lib32-uutils-coreutils
#
# $STAGING_LIBDIR_NATIVE
# set /OE/build/oe-core/openembedded-core/meta/conf/bitbake.conf:425
# "${STAGING_DIR_NATIVE}${libdir_native}"
STAGING_LIBDIR_NATIVE="/OE/build/oe-core/tmp/work/armv7at2-neon-oemllib32-linux-gnueabi/lib32-uutils-coreutils/0.0.29/recipe-sysroot-native/usr/lib"
Signed-off-by: Martin Jansa <martin.jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 8.3:
- Replace setup.py with hatchling. To install from source at least
version 21.3 of pip is needed.
- Drop support for Python 3.8.
- Add condition coverage to text summary report.
- Add --include to search files in search paths which should be
added to report.
- Add option to generate LCOV format produced by version 1.x of
LCOV tool.
- Extend logging for data merge errors with info about the data
sources.
- Add condition coverage merge mode option --merge-mode-conditions
- Add --gcov-suspicious-hits-threshold to configure the value for
detecting suspicious hits in GCOV files.
- Renamed JSON element destination_blockno to destination_block_id.
- Add --html-block-ids to show the block ids of the lines and
branches in HTML report.
- Fixed an error handling bug throwing a TypeError exception on a
gcov merge assertion failure instead of reporting the error and
(if requested by the user) continuing execution
- Check format version of external generated gcov JSON files.
- Fix crash on Windows when trying to fix the case of the files.
- Fix LCOV report. Excluded lines where added with a count of 0.
- Fix line exclusion not clearing all child coverage data.
- Fix summary stats in JaCoCo report.
- Fix path issue when reading/writing Coveralls report.
- Fix issue with negative counters in GCOV JSON export.
License-Update: Update year
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 3.15.1:
- Fix IndexError for add_rows() with an empty list
- Fix typing: replace Iterable with indexable Sequence
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Upgrade to release 0.15.2:
- Allow custom styles for commands in help output.
- Avoid the unnecessary import of typing_extensions in newer Python
versions.
- Fix shell completions for the fish shell.
- Rename test to corner-cases to make it more explicit.
- Fix small typos in the tutorial documentation.
- Update optional CLI argument section in tutorial with Annotated.
- Clarify the need for mix_stderr when accessing the output of
stderr in tests.
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The OPC UA schema definitions and nodeset descriptions make up 90MB of
the "normal" package and contain mostly text files not needed during
runtime.
Split them into a -tools package. The "normal" package size is reduced
drastically.
Signed-off-by: Johannes Kauffmann <johanneskauffmann@hotmail.com>
The default is still the FULL namespace, but it is now possible to
switch to REDUCED.
With the REDUCED namespace, the binary shrinks about 80% in size, from
4.5MB to ~800kB.
Signed-off-by: Johannes Kauffmann <johanneskauffmann@hotmail.com>
open62541@9e0b766fbba73be8d703b6ccbfdeec5bfd200b48 made it an error to
install the package while the amalgamation option is enabled.
On the 1.3 branch, this is not yet an issue. However, it is clear that
this option is not deemed suitable for building open62541 in the context
of a distribution.
Thus, remove the PACKAGECONFIG outright, before someone depends on the
different (incompatible) header includes required by the amalgamation
option.
Signed-off-by: Johannes Kauffmann <johanneskauffmann@hotmail.com>
ChangeLog:
Version 2.03.31 - 27th February 2025
====================================
Reduce 'mandoc -T lint' reported issues for man pages.
Restore support for LVM_SUPPRESS_FD_WARNINGS (2.03.24).
Fix uncache and split cache restoring original state of volume.
Extend use of lockopt skip to more scenarios.
Enhance error path resolving in polling code.
Disallow shared activation of LV with CoW snapshot.
Fix lvmlockd use in lvremove of CoW snapshot, VDO pool, and uncache.
Improve mirror split with opened temporary volumes.
Improve pvmove finish with opened temporary volumes.
Fix backup limit for devices file, handle over 10,000 files.
Ignore reported optimal_io_size not divisible by 4096.
Fix busy-loop in config reading when read returned 0.
Fix DM cache preserving logic (2.03.28).
Improve use of lvmlockd for usecases involving thin volumes and pools.
Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Native build on aarch64 breaks with:
| {standard input}: Assembler messages:
| {standard input}:169: Error: selected processor does not support `xpaclri'
| {standard input}:411: Error: selected processor does not support `xpaclri'
| {standard input}:859: Error: selected processor does not support `xpaclri'
| {standard input}:1140: Error: selected processor does not support `xpaclri'
Signed-off-by: Alex Kiernan <alex.kiernan@gmail.com>
Signed-off-by: Alex Kiernan <alexk@a-squared-projects.uk>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Currently, gnuplot recipe requires the x11 distro feature to be built.
However, X11 is not needed for running gnuplot as long as we don't use
a gnuplot X11 terminal.
This commit adds the x11 package config used to make the need of x11
distro feature optional to build gnuplot. `features_check` inherit is removed as
any of the functionalities of this class is now used.
Signed-off-by: Paul Le Guen de Kerneizon <paul.leguendekerneizon@savoirfairelinux.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Set GRPC_PYTHON_BUILD_EXT_COMPILER_JOBS to limit spawned compiler
processes. Without this it uses all available CPUs (via
multiprocessing.cpu_count()) and can exhaust build host since there are
lot of files to compile (e.g. with 128 cores it manages to spawn 128 gcc
processes)
Note that this is a general problem for all setuptools based builds with
build_ext compilation which can either compile with 1 thread or
cpu_count threads. grpcio hot-patches setuptools and allows to set
specific build concurrency value.
Signed-off-by: Peter Marko <peter.marko@siemens.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
disable_libunwind_aarch64.patch is dropped because from my runtime
testing, this patch is no longer needed.
0001-disbale-heap-checkers-and-debug-allocator-on-musl.patch is
dropped because I don't see it necessary now. Things still build
without it.
0001-src-mmap_hook.cc-Fix-build-for-32bit-machine.patch is dropped
because this oe specific patch is no longer needed because of the
following two commits:
02adc8c also disable _TIME_BITS in mmap_hook.cc
198b3dd disable _FILE_OFFSET_BITS in mmap_hook.cc
ppc-musl.patch is rebased.
0002-src-base-elf_mem_image.cc-fix-build-for-musl.patch is added
to fix build failure for musl + ppc64.
0003-Makefile.am-disable-building-noinst-tests-for-musl.patch is
added as a workaround to fix build failure on musl + ppc64.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
As of right now there are no applications available for benchmarking
OpenCL. Clpeak is a simple synthetic benchmark for OpenCL that is
designed to determine the peak capabilities of OpenCL devices.
It's currently one of the more popular CL benchmarks supported by
OpenBenchmarking.org and licensed Apache-2.0.
Signed-off-by: Randolph Sapp <rs@ti.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Khem Raj
Markus Volk
Etienne Cordonnier
Leon Anavi
Oliver Kästner
Martin Jansa
Kai Kang
Lei Maohui
Peter Marko
Sofiane HAMAM
Johannes Kauffmann
Yi Zhao
Alex Kiernan
Paul Le Guen de Kerneizon
Chen Qi
Randolph Sapp