For IRQ based scheduling we might otherwise be woken up after we only
added one of the fds to the poll loop and then we get an error when we
try to update it afterwards. Instead, add the fds from the data thread
to get things nicely in sync.
In IRQ mode, disable the ALSA fds while we wait for the graph to produce
data. Otherwise we will wake up very quickly over and over.
When we get more data, we activate the sources again to start the next
cycle.
Provides configuration to disable timer-based scheduling. This can be
useful at low latencies, for example, where period-based interrupts
might be more reliable than timers.
Drivers should only read the target_ values in the timeout, update the
timeout with the new duration and then update the position.
For the position we simply need to add the previous duration to the
position and then set the new duration + rate.
Otherwise, everything else should read the duration/rate and not use
the target_ values.
Place the target rate and duration in the io clock area.
The driver is meant to read these new values at the start of the cycle
and update the position rate and duration.
This used to be done by the pipewire server when it received the ready
callback from the driver but this is in fact too late. Most driver would
start processing and set the next timeout based on the old rate/duration
instead of the new pending ones.
There is still a fallback for the old behaviour (with a warning) when
the driver doesn't yet update the position.
The code that removes the mixer path if probing fails can be called in
the path that sets a non-off device profile on hotplug *before*
card->active_profile is updated, which results in spuriously removing
the mixer path. By this point, context->ucm->active_verb would be set
to the same as the profile name, so we can use that instead to avoid
the issue.
On Apple Silicon machines with the UCM profiles in the Asahi Linux repo,
this manifests as the headphones jack having hardware volume controls
*only* if PA is started with headphones connected and until they are
disconnected. Hotplugs end up triggering the bad codepath, and it falls
back to software volume (which is particularly a problem when the
hardware volume happens to be very low or 0 at that point).
Fixes: a9cc1373e2a7 ("alsa: ucm - update the mixer path also after volume probe")
Signed-off-by: Hector Martin <marcan@marcan.st>
The ucm_get_device_property() function adds to each UCM device's
playback_volumes (or capture_volumes) hash map an associated volume
mixer keyed with the UCM verb. These key-value pairs are then iterated
over in various places which assume the key is a profile name. This
assumption is no longer true since we can generate multiple profiles to
use conflicting devices.
A previous commit 45278904167f ("alsa-ucm: Stop conflating profile name
with UCM verb name") fixes some instances of this assumption, but misses
the relation explained above. Fix more instances of misleading
"profile"s where the UCM verb name is actually meant.
Fixes: 45278904167f ("alsa-ucm: Stop conflating profile name with UCM verb name")
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Although it's a valid UCM configuration to have multiple devices using
the same PlaybackPCM or CapturePCM, it's unclear how PulseAudio should
handle the cases where multiple of these devices are enabled. Some
options I can think of are:
- Merge all devices sharing the same PCM into the same mapping, open
only one PCM substream for this mapping, and add 'combination ports'
that enable combinations of the devices. This has been the case until
recently, although the combination port logic was broken. A problem
with this is that we can't independently control device volumes. We
most likely cannot use hardware volumes either.
- Have one mapping for each device in the same profile, and open one PCM
substream for each mapping. This is the current state, and it fails
when there are fewer substreams than devices. Otherwise it works, but
it's still confusing, as sound directed to a device-specific mapping
might end up playing at multiple devices.
- Make multiple profiles each with combinations of upto-substream-count
devices, and have one mapping/substream per device. This still causes
the confusion mentioned above. And it's likely that the substream
count will almost always be one, where this case degenerates into the
last one.
- Have one mapping for each device in the same profile, but open only
one PCM substream. I assume this is possible with software mixing, but
it is still confusing like the above, and probably less performant.
- Generate multiple profiles each with one of the shared-PCM devices,
again with one mapping/substream for that one device. The trade-off
with this is that we can't use multiple of these devices at the same
time. However, this doesn't have the output device confusion,
combination port's volume problems, or the substream count limitation.
This patch takes a short-cut to achieve the last option, by considering
shared-PCM devices implicitly conflicting with each other.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
While switching profiles of the same UCM verb, existing code first
disables devices that are only on the first profile to avoid conflicts.
However, it only disables devices, not modifiers. Even worse, modifiers
which have PlaybackPCM/CapturePCM are incorrectly treated as devices and
result in a segmentation fault.
Check what we are disabling, and call the appropriate disable function
for both devices and modifiers. Modifiers are disabled before devices,
because _dismod calls fail when the modifier's supported devices are
disabled.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
These are mostly the same as the device helpers added in c83b34516929
("alsa-ucm: Add enable, disable, status helpers for devices"), but for
modifiers instead.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Looks like original intention was to scan over sample formats supported by PA,
but code does the scan by list of alsa formats. Reverse the map and adjust
fallback case which now can use the same map.
Perform detection of supported sample format and rates just after device is
opened, before `snd_pcm_hw_params()` is called for the first time. This fixes a
problem where device restricts available sample rates after HW params are set
preventing sample rate detection (seen with UAC2 devices and kernel 6.1.9)
The full identifier check must be executed for the new melem
creation, otherwise the duplicate control element code check
is reached.
Example (using the snd-aloop driver):
numid=56,iface=PCM,name='PCM Notify',device=1,subdevice=1
numid=62,iface=PCM,name='PCM Notify',device=1,subdevice=2
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
The new helem must be tracked and old helem must be cleared
to make the code work properly. Introduce the pointer to helem
as the private value for melem and add the necessary code.
Also, add a check for the duplicate mixer elements. The duplicate
mixer element invokes the abort check in alsa-lib. Print a warning
instead and handle the exit gracefully.
Fixes: def8eb074 ("alsa-mixer: allow to re-attach the mixer control element")
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Right now we try to add all UCM devices of a verb to a single profile.
But if some devices using different PCMs are configured as conflicting
with one another, we will only be able to utilize one of them, chosen
seemingly based on the order in the UCM config file.
This is not a problem with conflicting devices sharing a PCM, as they
are assigned to the same mapping and the ports mechanism only enables
one of them to be active at a time.
To utilize all devices in a UCM verb even when there are conflicting
devices using different PCMs, calculate subsets of devices which
can be simultaneously used and create a profile for each such set.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
While switching profiles, it was enough to switch UCM verbs since that
disables all enabled UCM devices and every profile had a distinct verb.
However, switching to the current verb does not disable any devices.
To support multiple profiles for a verb we need to explicitly disable
the old profile's devices, since they might be conflicting with the new
profile's devices and will prevent them from being enabled. Compare both
profiles' mappings, and disable the devices not in the new mappings.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
After previous patches, we should be generating no combination ports, so
we don't need to store multiple modifiers per mapping. Simplify the code
based on this.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
After previous patches, we should be generating no combination ports, so
we don't need to store multiple devices per mapping. Simplify the code
based on this.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
After previous patches, we should be generating no combination ports, so
we don't need to store multiple devices per port. Simplify the code
based on this.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
A previous commit makes mapping names depend on the UCM device name.
Since UCM device names are unique, this means a mapping will at most
have one port and thus no combination ports can be generated.
This removes the dead code in the pa_alsa_ucm_add_ports_combination()
function, unrolls the remaining code in its helper functions that it
used, and renames it to pa_alsa_ucm_add_port() to signal that it no
longer generates combinations.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
PulseAudio combines UCM devices that have the same PlaybackPCM or
CapturePCM value into a single mapping with multiple ports. It also
creates ports in the same mapping for each valid combination of those
UCM devices.
Since mappings are the things we put in profiles, we can put in a
profile either all devices of a joint mapping or none of them. This
causes some complications with device conflicts. For example, a
different UCM device might be marked as conflicting with some (but not
all) of the devices in a joint mapping. In this case we can do one of
three things:
- Include all devices in one profile, and hope the conflicting device
isn't chosen as the mapping's active port. We shouldn't do this as it
puts conflicting devices in the same profile.
- Make one profile with the joint group, and one with the other device.
This is somewhat acceptable as we have no conflicts, but we sacrifice
some compatible combinations of devices.
- Do not group the devices into the same mapping, and make one profile
for each compatible combination of devices. This appears to be the
best option, one where we can always have the maximum number of
working devices.
This patch chooses the third option and makes one input and/or output
mapping per UCM device, by using UCM device names instead of PCM device
strings in the mapping names.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Combination port logic calculates some useful properties for device
groups that we could reuse while generating multiple profiles to support
conflicting devices. Split them into their own functions.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Right now this check is rejecting devices whose UCM config specifies
neither a conflicting device nor a supported device list, and accepting
devices which specify both. However, a device without neither list is
actually unrestricted, and a device with both lists is a configuration
error. Fix the check to accept the former.
Furthermore, this is missing another case where an already selected
device might have a supported devices list that doesn't have the
candidate device. Make this function also check against that, and also
make it accept devices already in the set.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
The existing code meant to generate device groups for combination ports
is tightly coupled to port creation. Similar functionality would be
useful to generate nonconflicting device groups for multiple profiles as
well, so this tries to rewrite it into a more reusable state.
Several things (e.g devices, mapping contexts) use idxsets to store a
device selection. This also switches this conformance check and device
group generation to using idxsets to make it easier to work with those,
with the eventual aim to unify device group representations.
Also try to adjust users of these functions to use idxsets these will
need/return, without causing too much interference.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
To support having multiple profiles per UCM verb, split the profile
creation into two parts based on whether they should run once for each
verb or for each profile (maybe multiple times per verb).
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
The ucm_create_mapping() function is not idempotent. It looks like it
was meant to be called once per device for the devices of a UCM verb
and takes a profile argument simply because a verb has generated a
single profile so far.
Make sure creating mappings per device and adding those mappings to the
profiles happens as separate steps to make it easier to split UCM verbs
and profiles as concepts.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
So far each profile had the exact name as their associated UCM verb,
which caused the one to be used where the other should have been.
Explicitly get and use the verb name where that was intended, and make
sure things about profiles aren't named after verbs.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Currently each UCM verb generates one profile named the same as the
verb, meaning it's trivial to know which verb the profile belongs to.
This will be slightly harder to do when we generate multiple profiles
per UCM verb (e.g. to make use of conflicting devices).
It would still be possible to parse the profile name to get the UCM
verb, but instead let's keep track of the struct instance representing
the profile's associated verb. This also lets us remove a block of code
searching for the verb by its name.
Co-authored-by: Jaroslav Kysela <perex@perex.cz>
[Alper: Reused Jaroslav's UCM profile context changes for UCM verb
instead of combined devices.]
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
While switching profiles, it's possible that we will want to do more
work besides switching UCM verbs. The alsa-card module already has our
profiles as structs, but passes in only the names instead of the entire
struct. Make things work with the struct instead, so we can add other
things (like a UCM context) to it and use those here.
Co-authored-by: Tanu Kaskinen <tanuk@iki.fi>
[Alper: Split into its own commit and integrated Tanu's snippet.]
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Right now manipulating device status is done inline once while setting a
port. However, we will need to reuse this code to disable conflicting
devices of a device we want to enable. Split it into enable and disable
helper functions.
There is another issue with the device enable logic, where trying to
disabling an already disabled device sometimes fails. To avoid that,
implement a status helper and check if the device we want to enable is
already enabled/disabled before trying to do so.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
Modifiers currently keep their conflicting and supported devices's
names, and these names are resolved to devices every time we need to use
them. Instead, resolve these device names while creating the modifier
struct and keep track of the resulting device structs in idxsets, same
as how device structs keep track of their support relations.
Signed-off-by: Alper Nebi Yasak <alpernebiyasak@gmail.com>
This is intended to make the current and upcoming code a bit clearer, as
we won't need to constantly check for the existence of these idxsets
before using or operating on them.
It may be possible that the ALSA control element appears
again. Allow this combination by checking, if the pulseaudio
mixer element already exists. Do not create the duplicate
mixer element in this case.
Don't just limit the max delay of samples we keep in the ALSA ringbuffer
to the buffer_size but to half of it. Make this into a max_delay
variable.
If we have a buffer size of 8192 samples and a headroom of 8192 samples,
when capturing, we would wait for the ringbuffer to contain at least
8192 samples, which would always xrun. When we limit the size to
half, we can still read the data without xruns.
Fixes#2972
* Add support for running the sink as a driver
* Detect which compressed formats are actually supported
* Correctly open/close/start/stop device according to the node commands
* Shift away from tinycompress and use Compress-Offload ioctls directly
to be able to access various caps information (including fragment sizes)
which are unavailable in the tinycompress API
* Implement SPA_PARAM_PropInfo and SPA_PARAM_Props support
