In 2010, ALSA PCM interface got an flag of hardware parameters to suppress
periodical interrupts, according to a request from PulseAudio developer.
In typical PCM operation for usual hardware, PCM drivers configure the
hardware to generate the periodical interrupts to notify that the same
amount of data frames as a period of PCM buffer is actually transferred
via serial sound interface. The flag can suppress this if the driver
support it.
There's some merits of this configuration:
- No interrupt context run for PCM substream. The PCM substream is
handled in any process context only. No need to care of race
conditions between interrupt/process contexts. This is good for
developers of drivers and applications.
- CPU time is not used for handlers on the interrupt context. The CPU
time can be dedicated for the other tasks. This is good in a point
of Time Sharing System.
- Hardware is not configured to generate interrupts. This is good in a
point of reduction of overall power consumption.
Disabling period interrupt is used for 'Timer-based scheduling' to
consume data frames on PCM buffer independently of interrupt context. As
noted, no interrupt context runs for PCM substream, thus any blocking
operation is not released. Furthermore, system calls for multiplexed I/O
is not also released without timeout.
In this scheduling model, applications need to care of available space on
PCM buffer by lapse of time, typically by yielding CPU and wait for
rescheduling. For the yielding, timeout is calculated for preferable
amount of PCM frames to process. This is an additional merit for
applications, like sound servers. when an I/O thread of the server wait
for the timeout, the other threads can process data frames for server
clients. Furthermore, with usage of rewinding/forwarding, applications
can achieve low latency between transmission position and handling
position even if they uses large size of PCM buffers.
But the timeout should be calculated with enough care of hardware
capabilities. To disable period interrupt, used hardware should satisfy
some requirements for data transmission:
1. Even if drivers don't handle interrupts to queue next data transmission,
hardware voluntarily perform the data transmission when needed
(typically by requesting DMA automatically).
2. hardware has a capability to report current position of data
transmission with enough accuracy against the data transmission.
developers refer this as 'granularity'. If hardware can always
reports updated position after the data transmission finishes, the
granularity equals to the size of period of PCM buffer.
3. a fine size of data transmission in one time. This size is decided
depending on configuration of hardware or DMA controller, but for
efficiency it may not be one byte. Thus some amount of data frame is
transferred by one data transmission. Developers refer this as
'burst-ness'.
The timeout should be calculated according to the item 2 and 3, however
in current ALSA PCM interface supplemental information is not delivered
from drivers to applications. Although at present userspace applications
should be written by a speculative way for this point, there's few
problems because there're a few hardware which satisfy the above items.
However, when more drivers supports this feature, the problem may largely
be exposed and bothers application developers.
This commit adds an option to use 'timer-based scheduling' for data
transmission. This commit adds '--sched-model' option, and the scheduling
mode is enabled when 'timer' is assigned to the option by equal sign.
Although there's some TODOs, you can see the scheduling mode in this
simple program, like:
$ axfer transfer --sched-model=timer -P -d 2 -D hw:0,3 /dev/urandom -f dat -vvv
$ axfer transfer --sched-model=timer -C -d 2 -D hw:1,0 /dev/null -r 48000 -vvv
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit is an integration to add an option for users to choose waiter
type. Users give the type to value to '--waiter-type' ('-w') option to
choose it. Currently, 'snd_pcm_wait()' is just supported as a default.
This alsa-lib API is implemented with a call of poll(2).
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
There're several types of system calls for multiplexed I/O. They're used to
receive notifications of I/O events. Typically, userspace applications call
them against file descriptor to yield CPU. When I/O is enabled on any of
the descriptors, a task of the application is rescheduled, then the
application execute I/O calls.
This commit adds a common interface for this type of system calls, named as
'waiter'. This is expected to be used with non-blocking file operation and
operations on mapped page frame.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
As of 2017, two userspace library implementations are known; alsa-lib and
tinyalsa. The latter is simple I/O library to use ALSA PCM interface. On
the other hand, alsa-lib is more complicated than it. This is because it's
designed to add features to transmission of data frames; e.g. sample
resampling. To achieve this, alsa-lib has its configuration space and
plugin system.
In aplay, some options are implemented as a flag for the plugins in
alsa-lib. The flag is given to snd_pcm_open(). This commit adds support
for the flags.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In ALSA PCM interface, some parameters are used to configure runtime of
PCM substream independently of actual hardware. These parameters are
mainly used to decide the detailed timing to start/stop PCM substream and
release I/O blocking state of application. These parameters are
represented and delivered by a structure.
In alsa-lib PCM API, the structure is hidden from userspace applications.
The applications can set/get actual parameters by helper functions.
In aplay, three of the parameters are configurable. This commit adds
support for them. When no options are given, default values are used.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In ALSA PCM interface, two parameters are used for size of intermediate
buffer for data frames; period size and buffer size. Actual effects of
these sizes differs depending on hardware, but basically the size of
period is used for intervals of hardware interrupts and the size of buffer
is used to maintain the intermediate buffer as ring buffer. These
parameters can be configured as a part of hardware parameters by
data frame unit or micro second. PCM API in alsa-lib also includes
helper functions to configure them by the two units.
This commit adds support for options to the parameters by both units. When
no options are given, default values are applied according to current
aplay; available maximum size of buffer up to 500msec, a quarter of the
size of buffer for period size. However, these calculation should be
reconsidered somehow.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In aplay, '--test-nowait' is used to suppress calls of snd_pcm_wait()
when I/O operations return -EAGAIN or process truncated number of data
frames. This seems to be for debugging purpose. In this program, this
option is equivalent to suppress event waiting.
This commit adds support for this option.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In alsa-lib PCM API, data frames can be handled in mapped page frame,
instead of calling any system calls.
This commit support for this type of operation. To reduce CPU usage,
this commit uses 'snd_pcm_wait()' to wait for event notification.
Below lines are examples to execute:
$ axfer transfer -M -P -d 2 -D hw:0,3 /dev/urandom -f dat -vvv
$ axfer transfer -M -C -d 2 -D hw:1,0 /dev/null -r 48000 -vvv
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In alsa-lib PCM API, snd_pcm_read[i|n]() and snd_pcm_write[i|n] can be
used with non-blocking mode. This is available when SND_PCM_NONBLOCK is
used as 'mode' argument for a call of snd_pcm_open().
This commit adds support this type of operation. To reduce CPU usage, this
commit uses 'snd_pcm_wait()' to wait for event notification.
Below lines are examples to execute:
$ axfer transfer -N -P -d 2 -D hw:0,3 /dev/urandom -f dat -vvv
$ axfer transfer -N -C -d 2 -D hw:1,0 /dev/null -r 48000 -vvv
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In aplay, '--fatal-errors' option has an effect to give up recovery of PCM
substream from XRUN state. This commit adds support for this option.
In original implementation, this option brings program abort. This seems
to generate core dump of process VMA. However, typically, XRUN comes from
timing mismatch between hardware and application, therefore core dump has
less helpful. This commit finishes this program in usual way with this
option at XRUN.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
In alsa-lib PCM API, snd_pcm_read[i|n]() and snd_pcm_write[i|n]() are used
to transfer data frames from/to hardware. When a handler is not opened with
specific flags, these functions perform blocking operation; i.e. the
function call doesn't return till all of request number of data frames are
actually handled, or call is interrupted by Unix signals, or PCM substeam
corrupts due to hardware reasons.
This commit adds support for this type of data transmission. For cases that
requested data frames are not processed by container interface, this commit
adds internal cache mechanism to handle rest of data frames in next timing.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
This commit adds support fo alsa-lib PCM API as a backend of 'xfer'
module. In a set of alsa-lib PCM API, there're two ways to handle data
frames; by calling ioctl(2) with some specific commands with buffer in
user space, or copying data frames on mapped page frames. To support
both ways, this commit adds an operation structure as abstraction.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
