snd_pcm_wait() & co checks the current avail value and returns
immediately if it satisfies <= avail_min condition. It's good in
general except for one situation: draining. When the draining is
being performed in the non-blocking mode, apps are supposed to wait
via poll(), typically via snd_pcm_wait(). So this ends up with the
busy loop because of the immediate return from snd_pcm_wait().
A simple workaround is to put the PCM state check and ignore the
avail_min condition if it's DRAINING state. The equivalent check is
found in the kernel xfer code, too.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Without this fix the application pointer would be reseted
whenever an application opens a device with SND_PCM_APPEND.
This would result in an Xrun if the device is already opened and
in running state and the appl_ptr is use.
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Currently, if the config file includes several hw_configs sections,
parse_hw_config_refs() returns after parsing only the first section.
For example, the following config, based on
alsa-lib/src/conf/topology/broadwell/broadwell.conf, is parsed incorrectly:
~~~~
SectionHWConfig."CodecHWConfig" {
id "1"
format "I2S" # physical audio format.
bclk "master" # Platform is master of bit clock
fsync "master" # platform is master of fsync
}
SectionHWConfig."CodecHWConfig2" {
id "2"
format "AC97"
}
SectionLink."Codec" {
# used for binding to the physical link
id "0"
hw_configs [
"CodecHWConfig"
"CodecHWConfig2"
]
default_hw_conf_id "2"
}
~~~~
Signed-off-by: Kirill Marinushkin <k.marinushkin@gmail.com>
Cc: Jaroslav Kysela <perex@perex.cz>
Cc: Takashi Iwai <tiwai@suse.com>
Cc: alsa-devel@alsa-project.org
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
PREPARED should only be set when it is done and it was successfully.
DRAINING should be signalled when starting to drain. There is no need to
check if draining was successfully because it will change to drop (SETUP)
in any case.
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The already existing areas_copy functions do not care about the end of
the source and destination buffer.
Therefore the caller has to take care that the requested offset+size
is not exceeding any buffer limit.
This additional function will take care about the end of an buffer
and will continue at the beginning of the buffer.
For example this is required when copying between buffers with
different sizes (not multiple of).
This is often the case in IO plugins like the JACK plugin.
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This function can be called without calling snd_pcm_avail_update().
The call to snd_pcm_avail_update() can take some time.
Therefore some developers would not like to call it from a real-time
context (e.g. from JACK client context).
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Without this commit compiling fails when THREAD_SAFE_API is not
enabled.
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
if requested by the IO plugin
Without this changes an IO plugin is not able to report
that buffer_size frames were read from the buffer.
When the buffer was full this is a valid action and
has not to be handled as an under run.
For example when the hw_ptr will be updated with
hw_ptr += buffer_size
and it is using the buffer_size as wrap around
hw_ptr %= buffer_size
would result in the same value as before the add operation.
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The 64-bit fast path can be used only in limited conditions:
- destination must be aligned to 64-bit (CPU aligned access)
- step must be equal to width
- physical with must be different than 24 (cannot be multiplied to 64)
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
bits for
24 bit sample cases
A function of 'snd_pcm_area_silence()' has a fast path to copy silent data
efficiently. However, the fast path works well just for a case that target
buffer consists of data samples for which unit of data alignment is
divisors of 64 bits.
At present, the fast path handles sample data aligned to 24 bit. In this
case, the buffer can includes extra 8 bits. This has no issue for 'signed'
case because silent data is zero, however it has an issue for 'unsigned'
case.
This commit fixes the bug by skipping cases of sample data of 24 bit.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
First, after silencing the buffer 64 bits at a time, any remaining samples
need to be silenced by the following width-specific code. However, instead
of silencing the end of the buffer, the code instead re-silences the start
of the buffer, leaving the end unsilenced. To fix this, update the pointer
used by the width-specific code to point to the end of the area just
silenced, instead of leaving it pointing to the start of the buffer.
Second, the code for 24 bit samples can only silence a single sample, as
there's no loop for multiple samples as with other formats. To fix this,
add a loop similar to the ones used for every other width.
The symptoms of these bugs are random data at the end of every supposedly
silenced buffer with certain format/buffer size combinations, resulting in
pops and noise.
Signed-off-by: furrywolf <alsa2@bushytails.net>
Reviewed-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Some PCM functions have the sanity check of the expected PCM states,
and most of them return -EBADFD if the current state doesn't match.
This is bad for some programs like aplay that expect the function
returning a proper code corresponding to the state, e.g. -ESTRPIPE for
the suspend.
This patch is an attempt to address such inconsistencies. The sanity
checker bad_pcm_state() now returns the error code instead of bool, so
that the caller can pass the returned code as is. And it calls a new
helper, pcm_state_to_error(), for obtaining the error code to certain
known PCM error state.
While we're at it, use the new pcm_state_to_error() for simplifying
the existing code to retrieve the error code, too.
Tested-by: Mirza Krak <mirza.krak@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Add a configuration for Cherry Trail boards which use a rt5645 codec
connected to a mono speaker and with an analog mic on IN2N + IN2P.
The chtrt5645-mono-speaker-analog-mic/HiFi.conf for this is based on the
latest version from https://github.com/plbossart/UCM/tree/master/chtrt5645
with all the unused input options removed and some changes made to make
the analog mic work.
This commit also adds 2 ucm dirs with the longname of 2 boards known to use
this setup, which simply contain a symlink to the generic
chtrt5645-mono-speaker-analog-mic entry.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
snd_dlopen() was recently rewritten to be versioned symbols, and we
have to call it with INTERNAL() wrapper from the library itself.
Add the proper declaration in the local header and fix the callers
appropriately.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The multiply defined versioned symbols have to be called with
INTERNAL() wrapper.
Add the missing declarations of versioned timer API functions in the
local header, and use them in the callers in PCM.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The multiply defined versioned symbols have to be called with
INTERNAL() wrapper.
Add the missing declarations of the internal forms of
snd_ctl_elem_info_get_dimension*() in the local header, and use them
in the (still remaining) callers in alisp.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
While calculating the size of data to be written into the topology
binary file, the size of the compound elements is added as well. This
results in wrong file offset calculation and topology build failure.
The compound elements shouldn't be written to the topology as these are
already embedded as part of other elements. So, skip adding the size of
compound elements to the file offset size calculation.
Signed-off-by: GuruprasadX Pawse <guruprasadx.pawse@intel.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Guneshwor Singh <guneshwor.o.singh@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A commit 16b3bf447c ('Enhanced bitmasks in PCM - added support for more
formats by Takashi and me') adds support for some cases of linear
interpolation of PCM samples, however some of added comments are not
proper. This commit fixes them.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A commit fcd164e622 ("Permit to PCM plug configuration to specify unchanged
parameters. Added support for RT signals to async interface. Added ops for
PCM mix.") added a pair of NORMS_LABELS/NORMS_END, however they have been
no longer used.
This commit removes them in a point to reduce the amount of code to
maintain.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A commit 07c07da44f ("Fixed signess for route conversion") obsoletes
usage of a pair of GETU_LABEL/GETU_END, but it did not remove some
actual codes in 'src/pcm_plugin_ops.h'.
This commit removes them in a point to reduce the amount of code to
maintain.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
A commit 7b054f4dce obsoleted usage of a pair of COPY_LABELS/COPY_END,
however it did not remove some codes in 'src/pcm/plugin_ops.h'.
This commit removes them in a point to reduce the amount of code to
maintain.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit adds the recently added formats SND_PCM_FORMAT_{S,U}20 to
the linear_preferred_formats array in pcm_plug.
Let's give them lower priority than more standard S24 formats but a higher
priority than less typical 3-byte versions.
Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
Reviewed-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The previous patch has added 20-bit PCM formats SND_PCM_FORMAT_{S,U}20 to
alsa-lib.
We need to extend the linear format conversion code with handling of these
sample formats so they can also be utilized by applications that only
recognize the more typical ones like SND_PCM_FORMAT_S16.
Since the conversion arrays are indexed by a format bit width divided by 8
the easiest way to handle these formats is to treat them like they were
40-bit wide (the next free integer multiple of 8 bits).
This doesn't create a collision risk with a future format since there can't
really be a 40-bit sample format that occupies 4 bytes.
Make sure we use the getput conversion method for these formats since a
direct conversion from / to them is not supported.
Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
Reviewed-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This patch adds and describes in various functions that query format
properties SND_PCM_FORMAT_{S,U}20 formats that were recently added to the
kernel as SNDRV_PCM_FORMAT_{S,U}20.
These formats are similar to existing 20-bit PCM formats
SND_PCM_FORMAT_{S,U}20_3, however they occupy 4 bytes instead of 3.
Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
Reviewed-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This patch adds definitions of 20-bit, 4-byte PCM formats
SNDRV_PCM_FORMAT_{S,U}20, that were recently added to the kernel, to
alsa-lib's own copy of asound.h.
Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
Reviewed-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Like the previous Lenovo laptops, some Gigabyte mobos have dual
HD-audio codecs and need to switch dynamically via UCM profile.
Reuse the same profile as Lenovo dual codecs, so far.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Some recent Lenovo laptops have dual codecs and we need to switch them
accordingly. The kernel side already contains a fix and gives the
unique longname string for identifying the board, and here we hook up
the corresponding UCM profile.
The profile was corrected and tested by Hui Wang on Lenovo p520.
Tested-by: Kailang <kailang@realtek.com>
Tested-by: Hui Wang <hui.wang@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The dlopen() function might fail also for another reason than
a missing file, thus return the error string from dlerror().
Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Current implementation of channel-map TLV on test program is not valid.
Furthermore, it brings buffer-over-run due to byte counting.
This commit fixes it.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Commit de63b942ac ("pcm: route: Use get/put labels for all 3 byte formats")
wanted to make the route plugin use get / put labels not only for 24-bit
physical width formats but also for 18 and 20-bit ones.
There was, however, a typo in that commit so a check for these widths
didn't really work.
Let's fix it now.
Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
APIs of dimensional information are deprecated for future removal. This
commit removes test codes for user-defined element set in an aspect of
the feature.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In ALSA control interface in asound.h, 'struct snd_ctl_elem_info' has
'dimen' member to deliver information for multi-dimensional array, however
there's no common way to handle the member. As a result, drivers can
force userspace applications to handle the information by inconsistent
ways.
This issue was reported by a commit 51db452df07b ('Revert "ALSA: echoaudio:
purge contradictions between dimension matrix members and total number of
members"') to Linux kernel. As a result of discussion at Linux
miniconference 2017, usage of 'dimen' member of 'struct snd_ctl_elem_info'
is going to be deprecated for future removal.
A removal of kernel interface can cause regression issues. However no ALSA
driver in kernel land except for 'echoaudio' series utilizes this feature.
Actually it's reasonable to assume that 'echomixer' program is an unique
consumer of the interface in user land and the removal rarely brings any
impact to user land.
This commit deprecates some APIs corresponding to the kernel interface. The
kernel interface is kept till Linux kernel v4.20 at least, but actual
timing of removal is not fixed yet. After that, these APIs may also be
removed at a reasonable timing.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The UCM configuration is enabling the speakers in the SectionDefaults.
This is a problem when booting with an headset already connected since
the audio output is routed at the same time both on speakers and
heaphones until the jack is disconnected and reconnected again.
Fix this disabling all the outputs in the default mixer configuration.
Signed-off-by: Carlo Caione <carlo@endlessm.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The conf HiFi file name is HiFi.conf, fix the name in the main
configuration file.
Signed-off-by: Carlo Caione <carlo@endlessm.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Add a UCM configuration for the rt5651 codec on Intel's Cherry-Trail
platform. Adapted from [0].
[0] https://github.com/plbossart/UCM/tree/master/bytcr-rt5651
Signed-off-by: Carlo Caione <carlo@endlessm.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
PulseAudio assumes that the "front" pcm device always refers to an
analog device, not HDMI. While that assumption is not really valid, the
reality is that without that assumption PulseAudio can't know whether
"front" and "hdmi" refer to a different or the same device.
The HDMI LPE driver doesn't allow audio streaming while the HDMI cable
is unplugged, so PulseAudio has to know when it's plugged in and when
it's not. If both "front" and "hdmi" devices exist, PulseAudio will
notice that HDMI is unplugged, but it doesn't know that "front" refers
to the same device, and PulseAudio will try to use the "front" device
with bad consequences. The kernel driver's refusal to stream any audio
makes PulseAudio enter an infinite loop and then the kernel kills
PulseAudio, because it consumes too much CPU time in a realtime thread.
While the looping in PulseAudio could probably be fixed, that wouldn't
change the fact that PulseAudio thinks that there is an analog device. I
believe it's best to avoid having the same device as both "front" and
"hdmi" in alsa-lib.
I removed also the surround configuration includes. I don't think they
had any effect anyway, so I wonder why they were there in the first
place.
BugLink: https://bugs.freedesktop.org/show_bug.cgi?id=100488
Signed-off-by: Takashi Iwai <tiwai@suse.de>
The snd_ctl_ext_callback::read_event() callback is only optional
if no poll descriptor was given via
snd_ctl_ext_t::poll_fd
or
snd_ctl_ext_callback::snd_ctl_ext_poll_descriptors().
If a poll descriptor is given the
snd_ctl_ext_callback::read_event()
callback has also to be defined
because there is no minigful default behavior.
This callback will be called when ever the poll() on
the file descriptor indicates that there is an event pending.
Therefore returning a 0 which indicates that there is no event makes no
sense.
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Events sent by external clients subscribed to the input port are not
timestamped.
This inconsistent behavior may surprise newbies who look at seq-decoder as
a reference example.
See the example below using "vkeybd --addr 128:0" to connect to seq-decoder,
the events sent by vkeybd are on a different queue with no timestamps:
...
EVENT>>> Type = 66, flags = 0x0, time = 0 ticks
Source = 0.1, dest = 128.0, queue = 253
Event = Port Subscribed; 129:0 -> 128:0
EVENT>>> Type = 66, flags = 0x1, time = 4.829712627
Source = 0.1, dest = 128.0, queue = 0
Event = Port Subscribed; 129:0 -> 128:0
EVENT>>> Type = 10, flags = 0x0, time = 0 ticks
Source = 129.0, dest = 128.0, queue = 253
Event = Controller; ch=0, param=0, value=0
EVENT>>> Type = 11, flags = 0x0, time = 0 ticks
Source = 129.0, dest = 128.0, queue = 253
Event = Program Change; ch=0, program=0
...
After the change events are on the main queue and are timestamped:
...
EVENT>>> Type = 66, flags = 0x1, time = 4.280907223
Source = 0.1, dest = 128.0, queue = 0
Event = Port Subscribed; 129:0 -> 128:0
EVENT>>> Type = 66, flags = 0x1, time = 4.280912063
Source = 0.1, dest = 128.0, queue = 0
Event = Port Subscribed; 129:0 -> 128:0
EVENT>>> Type = 10, flags = 0x1, time = 4.280990702
Source = 129.0, dest = 128.0, queue = 0
Event = Controller; ch=0, param=0, value=0
EVENT>>> Type = 11, flags = 0x1, time = 4.280994862
Source = 129.0, dest = 128.0, queue = 0
Event = Program Change; ch=0, program=0
...
Signed-off-by: Antonio Ospite <ao2@ao2.it>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
snd_seq_set_queue_tempo() requires a queue id as the second argument,
fix the example in documentation to reflect that.
Also add the queue id as an argument of the set_tempo() function, just
to keep the whole example compilable.
Signed-off-by: Antonio Ospite <ao2@ao2.it>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
