When a link enters the "ERROR" state, it is scheduled for destruction in
`module-link-factory.c:link_state_changed()`, which queues `destroy_link()`
to be executed on the context's work queue.
However, if the link is destroyed by means of `pw_impl_link_destroy()`
directly after that, then `link_destroy()` unregisters the associated
`pw_global`'s event hook, resulting in `global_destroy()` not being called
when `pw_impl_link_destroy()` proceeds to call `pw_global_destroy()` some
time later. This causes the scheduled async work to not be cancelled. When
it runs later, it will trigger a use-after-free since the `link_data` object
is directly tied to the `pw_impl_link` object.
For example, if the link is destroyed when the client disconnects:
==259313==ERROR: AddressSanitizer: heap-use-after-free on address 0x7ce753028af0 at pc 0x7f475354a565 bp 0x7ffd71501930 sp 0x7ffd71501920
READ of size 8 at 0x7ce753028af0 thread T0
#0 0x7f475354a564 in destroy_link ../src/modules/module-link-factory.c:253
#1 0x7f475575a234 in process_work_queue ../src/pipewire/work-queue.c:67
#2 0x7b47504e7f24 in source_event_func ../spa/plugins/support/loop.c:1011
[...]
0x7ce753028af0 is located 1136 bytes inside of 1208-byte region [0x7ce753028680,0x7ce753028b38)
freed by thread T0 here:
#0 0x7f475631f79d in free /usr/src/debug/gcc/gcc/libsanitizer/asan/asan_malloc_linux.cpp:51
#1 0x7f4755594a44 in pw_impl_link_destroy ../src/pipewire/impl-link.c:1742
#2 0x7f475569dc11 in do_destroy_link ../src/pipewire/impl-port.c:1386
#3 0x7f47556a428b in pw_impl_port_for_each_link ../src/pipewire/impl-port.c:1673
#4 0x7f475569dc3e in pw_impl_port_unlink ../src/pipewire/impl-port.c:1392
#5 0x7f47556a02d8 in pw_impl_port_destroy ../src/pipewire/impl-port.c:1453
#6 0x7f4755634f79 in pw_impl_node_destroy ../src/pipewire/impl-node.c:2447
#7 0x7b474f722ba8 in client_node_resource_destroy ../src/modules/module-client-node/client-node.c:1253
#8 0x7f47556d7c6c in pw_resource_destroy ../src/pipewire/resource.c:325
#9 0x7f475545f07d in destroy_resource ../src/pipewire/impl-client.c:627
#10 0x7f47554550cd in pw_map_for_each ../src/pipewire/map.h:222
#11 0x7f4755460aa4 in pw_impl_client_destroy ../src/pipewire/impl-client.c:681
#12 0x7b474fb0658b in handle_client_error ../src/modules/module-protocol-native.c:471
[...]
Fix this by cancelling the work queue item in `link_destroy()`, which should
always run, regardless of the ordering of events.
Fixes #4691
|
||
|---|---|---|
| .gitlab | ||
| doc | ||
| include/valgrind | ||
| pipewire-alsa | ||
| pipewire-jack | ||
| pipewire-v4l2 | ||
| po | ||
| spa | ||
| src | ||
| subprojects | ||
| test | ||
| .codespell-ignore | ||
| .editorconfig | ||
| .gitattributes | ||
| .gitignore | ||
| .gitlab-ci.yml | ||
| autogen.sh | ||
| CODE_OF_CONDUCT.md | ||
| COPYING | ||
| INSTALL.md | ||
| LICENSE | ||
| Makefile.in | ||
| meson.build | ||
| meson_options.txt | ||
| NEWS | ||
| pw-uninstalled.sh | ||
| README.md | ||
| template.test.in | ||
PipeWire
PipeWire is a server and user space API to deal with multimedia pipelines. This includes:
- Making available sources of video (such as from a capture devices or application provided streams) and multiplexing this with clients.
- Accessing sources of video for consumption.
- Generating graphs for audio and video processing.
Nodes in the graph can be implemented as separate processes, communicating with sockets and exchanging multimedia content using fd passing.
Building and installation
The preferred way to install PipeWire is to install it with your distribution package system. This ensures PipeWire is integrated into the rest of your system for the best experience.
If you want to build and install PipeWire yourself, refer to install for instructions.
Usage
The most important purpose of PipeWire is to run your favorite apps.
Some applications use the native PipeWire API, such as most compositors (gnome-shell, wayland, ...) to implement screen sharing. These apps will just work automatically.
Most audio applications can use either ALSA, JACK or PulseAudio as a backend. PipeWire provides support for all 3 backends. Depending on how your distribution has configured things this should just work automatically or with the provided scripts shown below.
PipeWire can use environment variables to control the behaviour of applications:
PIPEWIRE_DEBUG=<level>to increase the debug level (or use one ofXEWIDTfor none, error, warnings, info, debug, or trace, respectively).PIPEWIRE_LOG=<filename>to redirect log to filenamePIPEWIRE_LOG_SYSTEMD=falseto disable logging to systemd journalPIPEWIRE_LATENCY=<num/denom>to configure latency as a fraction. 10/1000 configures a 10ms latency. Usually this is expressed as a fraction of the samplerate, like 256/48000, which uses 256 samples at a samplerate of 48KHz for a latency of 5.33ms. This function does not attempt to configure the samplerate.PIPEWIRE_RATE=<num/denom>to configure a rate for the graph.PIPEWIRE_QUANTUM=<num/denom>to configure latency as a fraction and a samplerate. This function will force the graph samplerate todenomand force the specifiednumas the buffer size.PIPEWIRE_NODE=<id>to request a link to the specified node. The id can be a node.name or object.serial of the target node.
Using tools
pw-cat can be used to play and record audio and midi. Use pw-cat -h to get
some more help. There are some aliases like pw-play and pw-record to make
things easier:
$ pw-play /home/wim/data/01.\ Firepower.wav
Running JACK applications
Depending on how the system was configured, you can either run PipeWire and JACK side-by-side or have PipeWire take over the functionality of JACK completely.
In dual mode, JACK apps will by default use the JACK server. To direct a JACK
app to PipeWire, you can use the pw-jack script like this:
$ pw-jack <appname>
If you replaced JACK with PipeWire completely, pw-jack does not have any
effect and can be omitted.
JACK applications will automatically use the buffer-size chosen by the
server. You can force a maximum buffer size (latency) by setting the
PIPEWIRE_LATENCY environment variable like so:
PIPEWIRE_LATENCY=128/48000 jack_simple_client
Requests the jack_simple_client to run with a buffer of 128 or
less samples.
Running PulseAudio applications
PipeWire can run a PulseAudio compatible replacement server. You can't use both servers at the same time. Usually your package manager will make the server conflict so that you can only install one or the other.
PulseAudio applications still use the regular PulseAudio client libraries and you don't need to do anything else than change the server implementation.
A successful swap of the server can be verified by checking the output of
pactl info
It should include the string:
...
Server Name: PulseAudio (on PipeWire 0.3.x)
...
You can use pavucontrol to change profiles and ports, change volumes or redirect streams, just like with PulseAudio.
Running ALSA applications
If the PipeWire alsa module is installed, it can be seen with
$ aplay -L
ALSA applications can then use the pipewire: device to use PipeWire
as the audio system.
Running GStreamer applications
PipeWire includes 2 GStreamer elements called pipewiresrc and
pipewiresink. They can be used in pipelines such as this:
$ gst-launch-1.0 pipewiresrc ! videoconvert ! autovideosink
Or to play a beeping sound:
$ gst-launch-1.0 audiotestsrc ! pipewiresink
PipeWire provides a device monitor as well so that
$ gst-device-monitor-1.0
shows the PipeWire devices and applications like cheese will automatically use the PipeWire video source when possible.
Inspecting the PipeWire state
To inspect and manipulate the PipeWire graph via GUI, you can use Helvum.
Alternatively, you can use use one of the excellent JACK tools, such as Carla,
catia, qjackctl, ...
However, you will not be able to see all features like the video
ports.
pw-mon dumps and monitors the state of the PipeWire daemon.
pw-dot can dump a graph of the pipeline, check out the help for
how to do this.
pw-top monitors the real-time status of the graph. This is handy to
find out what clients are running and how much DSP resources they
use.
pw-dump dumps the state of the PipeWire daemon in JSON format. This
can be used to find out the properties and parameters of the objects
in the PipeWire daemon.
There is a more complicated tool to inspect the state of the server
with pw-cli. This tool can be used interactively or it can execute
single commands like this to get the server information:
$ pw-cli info 0
Documentation
Find tutorials and design documentation here.
The (incomplete) autogenerated API docs are here.
The Wiki can be found here
Contributing
PipeWire is Free Software and is developed in the open. It is mostly licensed under the MIT license. Check LICENSE for more details about the exceptions.
Contributors are encouraged to submit merge requests or file bugs on gitlab.
Join us on IRC at #pipewire on OFTC.
We adhere to the Contributor Covenant for our code of conduct.
Getting help
You can ask for help on the IRC channel (see above). You can also ask questions by raising a gitlab issue.