linux/sound/soc/soc-dapm.c

4709 lines
118 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0+
//
// soc-dapm.c -- ALSA SoC Dynamic Audio Power Management
//
// Copyright 2005 Wolfson Microelectronics PLC.
// Author: Liam Girdwood <lrg@slimlogic.co.uk>
//
// Features:
// o Changes power status of internal codec blocks depending on the
// dynamic configuration of codec internal audio paths and active
// DACs/ADCs.
// o Platform power domain - can support external components i.e. amps and
// mic/headphone insertion events.
// o Automatic Mic Bias support
// o Jack insertion power event initiation - e.g. hp insertion will enable
// sinks, dacs, etc
// o Delayed power down of audio subsystem to reduce pops between a quick
// device reopen.
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
#include <linux/module.h>
#include <linux/init.h>
#include <linux/async.h>
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
#include <linux/delay.h>
#include <linux/pm.h>
#include <linux/bitops.h>
#include <linux/platform_device.h>
#include <linux/jiffies.h>
#include <linux/debugfs.h>
#include <linux/pm_runtime.h>
#include <linux/regulator/consumer.h>
#include <linux/pinctrl/consumer.h>
#include <linux/clk.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/slab.h>
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
#include <sound/core.h>
#include <sound/pcm.h>
#include <sound/pcm_params.h>
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
#include <sound/soc.h>
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
#include <sound/initval.h>
#include <trace/events/asoc.h>
#define DAPM_UPDATE_STAT(widget, val) widget->dapm->card->dapm_stats.val++;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
#define SND_SOC_DAPM_DIR_REVERSE(x) ((x == SND_SOC_DAPM_DIR_IN) ? \
SND_SOC_DAPM_DIR_OUT : SND_SOC_DAPM_DIR_IN)
#define snd_soc_dapm_for_each_direction(dir) \
for ((dir) = SND_SOC_DAPM_DIR_IN; (dir) <= SND_SOC_DAPM_DIR_OUT; \
(dir)++)
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
static int snd_soc_dapm_add_path(struct snd_soc_dapm_context *dapm,
struct snd_soc_dapm_widget *wsource, struct snd_soc_dapm_widget *wsink,
const char *control,
int (*connected)(struct snd_soc_dapm_widget *source,
struct snd_soc_dapm_widget *sink));
struct snd_soc_dapm_widget *
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
snd_soc_dapm_new_control(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_widget *widget);
struct snd_soc_dapm_widget *
snd_soc_dapm_new_control_unlocked(struct snd_soc_dapm_context *dapm,
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
const struct snd_soc_dapm_widget *widget);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* dapm power sequences - make this per codec in the future */
static int dapm_up_seq[] = {
[snd_soc_dapm_pre] = 1,
[snd_soc_dapm_regulator_supply] = 2,
[snd_soc_dapm_pinctrl] = 2,
[snd_soc_dapm_clock_supply] = 2,
[snd_soc_dapm_supply] = 3,
[snd_soc_dapm_micbias] = 4,
[snd_soc_dapm_vmid] = 4,
[snd_soc_dapm_dai_link] = 3,
[snd_soc_dapm_dai_in] = 5,
[snd_soc_dapm_dai_out] = 5,
[snd_soc_dapm_aif_in] = 5,
[snd_soc_dapm_aif_out] = 5,
[snd_soc_dapm_mic] = 6,
[snd_soc_dapm_siggen] = 6,
[snd_soc_dapm_input] = 6,
[snd_soc_dapm_output] = 6,
[snd_soc_dapm_mux] = 7,
[snd_soc_dapm_demux] = 7,
[snd_soc_dapm_dac] = 8,
[snd_soc_dapm_switch] = 9,
[snd_soc_dapm_mixer] = 9,
[snd_soc_dapm_mixer_named_ctl] = 9,
[snd_soc_dapm_pga] = 10,
[snd_soc_dapm_buffer] = 10,
[snd_soc_dapm_scheduler] = 10,
[snd_soc_dapm_effect] = 10,
[snd_soc_dapm_src] = 10,
[snd_soc_dapm_asrc] = 10,
[snd_soc_dapm_encoder] = 10,
[snd_soc_dapm_decoder] = 10,
[snd_soc_dapm_adc] = 11,
[snd_soc_dapm_out_drv] = 12,
[snd_soc_dapm_hp] = 12,
[snd_soc_dapm_spk] = 12,
[snd_soc_dapm_line] = 12,
[snd_soc_dapm_sink] = 12,
[snd_soc_dapm_kcontrol] = 13,
[snd_soc_dapm_post] = 14,
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
};
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
static int dapm_down_seq[] = {
[snd_soc_dapm_pre] = 1,
[snd_soc_dapm_kcontrol] = 2,
[snd_soc_dapm_adc] = 3,
[snd_soc_dapm_hp] = 4,
[snd_soc_dapm_spk] = 4,
[snd_soc_dapm_line] = 4,
[snd_soc_dapm_out_drv] = 4,
[snd_soc_dapm_sink] = 4,
[snd_soc_dapm_pga] = 5,
[snd_soc_dapm_buffer] = 5,
[snd_soc_dapm_scheduler] = 5,
[snd_soc_dapm_effect] = 5,
[snd_soc_dapm_src] = 5,
[snd_soc_dapm_asrc] = 5,
[snd_soc_dapm_encoder] = 5,
[snd_soc_dapm_decoder] = 5,
[snd_soc_dapm_switch] = 6,
[snd_soc_dapm_mixer_named_ctl] = 6,
[snd_soc_dapm_mixer] = 6,
[snd_soc_dapm_dac] = 7,
[snd_soc_dapm_mic] = 8,
[snd_soc_dapm_siggen] = 8,
[snd_soc_dapm_input] = 8,
[snd_soc_dapm_output] = 8,
[snd_soc_dapm_micbias] = 9,
[snd_soc_dapm_vmid] = 9,
[snd_soc_dapm_mux] = 10,
[snd_soc_dapm_demux] = 10,
[snd_soc_dapm_aif_in] = 11,
[snd_soc_dapm_aif_out] = 11,
[snd_soc_dapm_dai_in] = 11,
[snd_soc_dapm_dai_out] = 11,
[snd_soc_dapm_dai_link] = 12,
[snd_soc_dapm_supply] = 13,
[snd_soc_dapm_clock_supply] = 14,
[snd_soc_dapm_pinctrl] = 14,
[snd_soc_dapm_regulator_supply] = 14,
[snd_soc_dapm_post] = 15,
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
};
static void dapm_assert_locked(struct snd_soc_dapm_context *dapm)
{
if (dapm->card && dapm->card->instantiated)
lockdep_assert_held(&dapm->card->dapm_mutex);
}
static void pop_wait(u32 pop_time)
{
if (pop_time)
schedule_timeout_uninterruptible(msecs_to_jiffies(pop_time));
}
__printf(3, 4)
static void pop_dbg(struct device *dev, u32 pop_time, const char *fmt, ...)
{
va_list args;
char *buf;
if (!pop_time)
return;
buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
if (buf == NULL)
return;
va_start(args, fmt);
vsnprintf(buf, PAGE_SIZE, fmt, args);
dev_info(dev, "%s", buf);
va_end(args);
kfree(buf);
}
static bool dapm_dirty_widget(struct snd_soc_dapm_widget *w)
{
return !list_empty(&w->dirty);
}
static void dapm_mark_dirty(struct snd_soc_dapm_widget *w, const char *reason)
{
dapm_assert_locked(w->dapm);
if (!dapm_dirty_widget(w)) {
dev_vdbg(w->dapm->dev, "Marking %s dirty due to %s\n",
w->name, reason);
list_add_tail(&w->dirty, &w->dapm->card->dapm_dirty);
}
}
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
/*
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
* Common implementation for dapm_widget_invalidate_input_paths() and
* dapm_widget_invalidate_output_paths(). The function is inlined since the
* combined size of the two specialized functions is only marginally larger then
* the size of the generic function and at the same time the fast path of the
* specialized functions is significantly smaller than the generic function.
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
*/
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
static __always_inline void dapm_widget_invalidate_paths(
struct snd_soc_dapm_widget *w, enum snd_soc_dapm_direction dir)
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
{
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
enum snd_soc_dapm_direction rdir = SND_SOC_DAPM_DIR_REVERSE(dir);
struct snd_soc_dapm_widget *node;
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
struct snd_soc_dapm_path *p;
LIST_HEAD(list);
dapm_assert_locked(w->dapm);
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (w->endpoints[dir] == -1)
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
return;
list_add_tail(&w->work_list, &list);
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->endpoints[dir] = -1;
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
list_for_each_entry(w, &list, work_list) {
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
snd_soc_dapm_widget_for_each_path(w, dir, p) {
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
if (p->is_supply || p->weak || !p->connect)
continue;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
node = p->node[rdir];
if (node->endpoints[dir] != -1) {
node->endpoints[dir] = -1;
list_add_tail(&node->work_list, &list);
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
}
}
}
}
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
/*
* dapm_widget_invalidate_input_paths() - Invalidate the cached number of
* input paths
* @w: The widget for which to invalidate the cached number of input paths
*
* Resets the cached number of inputs for the specified widget and all widgets
* that can be reached via outcoming paths from the widget.
*
* This function must be called if the number of output paths for a widget might
* have changed. E.g. if the source state of a widget changes or a path is added
* or activated with the widget as the sink.
*/
static void dapm_widget_invalidate_input_paths(struct snd_soc_dapm_widget *w)
{
dapm_widget_invalidate_paths(w, SND_SOC_DAPM_DIR_IN);
}
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
/*
* dapm_widget_invalidate_output_paths() - Invalidate the cached number of
* output paths
* @w: The widget for which to invalidate the cached number of output paths
*
* Resets the cached number of outputs for the specified widget and all widgets
* that can be reached via incoming paths from the widget.
*
* This function must be called if the number of output paths for a widget might
* have changed. E.g. if the sink state of a widget changes or a path is added
* or activated with the widget as the source.
*/
static void dapm_widget_invalidate_output_paths(struct snd_soc_dapm_widget *w)
{
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
dapm_widget_invalidate_paths(w, SND_SOC_DAPM_DIR_OUT);
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
}
/*
* dapm_path_invalidate() - Invalidates the cached number of inputs and outputs
* for the widgets connected to a path
* @p: The path to invalidate
*
* Resets the cached number of inputs for the sink of the path and the cached
* number of outputs for the source of the path.
*
* This function must be called when a path is added, removed or the connected
* state changes.
*/
static void dapm_path_invalidate(struct snd_soc_dapm_path *p)
{
/*
* Weak paths or supply paths do not influence the number of input or
* output paths of their neighbors.
*/
if (p->weak || p->is_supply)
return;
/*
* The number of connected endpoints is the sum of the number of
* connected endpoints of all neighbors. If a node with 0 connected
* endpoints is either connected or disconnected that sum won't change,
* so there is no need to re-check the path.
*/
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (p->source->endpoints[SND_SOC_DAPM_DIR_IN] != 0)
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
dapm_widget_invalidate_input_paths(p->sink);
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (p->sink->endpoints[SND_SOC_DAPM_DIR_OUT] != 0)
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
dapm_widget_invalidate_output_paths(p->source);
}
void dapm_mark_endpoints_dirty(struct snd_soc_card *card)
{
struct snd_soc_dapm_widget *w;
mutex_lock(&card->dapm_mutex);
list_for_each_entry(w, &card->widgets, list) {
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (w->is_ep) {
dapm_mark_dirty(w, "Rechecking endpoints");
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (w->is_ep & SND_SOC_DAPM_EP_SINK)
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
dapm_widget_invalidate_output_paths(w);
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (w->is_ep & SND_SOC_DAPM_EP_SOURCE)
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
dapm_widget_invalidate_input_paths(w);
}
}
mutex_unlock(&card->dapm_mutex);
}
EXPORT_SYMBOL_GPL(dapm_mark_endpoints_dirty);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* create a new dapm widget */
static inline struct snd_soc_dapm_widget *dapm_cnew_widget(
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
const struct snd_soc_dapm_widget *_widget)
{
struct snd_soc_dapm_widget *w;
w = kmemdup(_widget, sizeof(*_widget), GFP_KERNEL);
if (!w)
return NULL;
/*
* w->name is duplicated in caller, but w->sname isn't.
* Duplicate it here if defined
*/
if (_widget->sname) {
w->sname = kstrdup_const(_widget->sname, GFP_KERNEL);
if (!w->sname) {
kfree(w);
return NULL;
}
}
return w;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
struct dapm_kcontrol_data {
unsigned int value;
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
struct snd_soc_dapm_widget *widget;
struct list_head paths;
struct snd_soc_dapm_widget_list *wlist;
};
static int dapm_kcontrol_data_alloc(struct snd_soc_dapm_widget *widget,
struct snd_kcontrol *kcontrol, const char *ctrl_name)
{
struct dapm_kcontrol_data *data;
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
struct soc_mixer_control *mc;
struct soc_enum *e;
const char *name;
int ret;
data = kzalloc(sizeof(*data), GFP_KERNEL);
if (!data)
return -ENOMEM;
INIT_LIST_HEAD(&data->paths);
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
switch (widget->id) {
case snd_soc_dapm_switch:
case snd_soc_dapm_mixer:
case snd_soc_dapm_mixer_named_ctl:
mc = (struct soc_mixer_control *)kcontrol->private_value;
if (mc->autodisable && snd_soc_volsw_is_stereo(mc))
dev_warn(widget->dapm->dev,
"ASoC: Unsupported stereo autodisable control '%s'\n",
ctrl_name);
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
if (mc->autodisable) {
struct snd_soc_dapm_widget template;
name = kasprintf(GFP_KERNEL, "%s %s", ctrl_name,
"Autodisable");
if (!name) {
ret = -ENOMEM;
goto err_data;
}
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
memset(&template, 0, sizeof(template));
template.reg = mc->reg;
template.mask = (1 << fls(mc->max)) - 1;
template.shift = mc->shift;
if (mc->invert)
template.off_val = mc->max;
else
template.off_val = 0;
template.on_val = template.off_val;
template.id = snd_soc_dapm_kcontrol;
template.name = name;
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
data->value = template.on_val;
data->widget =
snd_soc_dapm_new_control_unlocked(widget->dapm,
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
&template);
kfree(name);
ASoC: dapm: handle probe deferrals This starts to handle probe deferrals on regulators and clocks on the ASoC DAPM. I came to this patch after audio stopped working on Ux500 ages ago and I finally looked into it to see what is wrong. I had messages like this in the console since a while back: ab8500-codec.0: ASoC: Failed to request audioclk: -517 ab8500-codec.0: ASoC: Failed to create DAPM control audioclk ab8500-codec.0: Failed to create new controls -12 snd-soc-mop500.0: ASoC: failed to instantiate card -12 snd-soc-mop500.0: Error: snd_soc_register_card failed (-12)! snd-soc-mop500: probe of snd-soc-mop500.0 failed with error -12 Apparently because the widget table for the codec looks like this (sound/soc/codecs/ab8500-codec.c): static const struct snd_soc_dapm_widget ab8500_dapm_widgets[] = { /* Clocks */ SND_SOC_DAPM_CLOCK_SUPPLY("audioclk"), /* Regulators */ SND_SOC_DAPM_REGULATOR_SUPPLY("V-AUD", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-AMIC1", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-AMIC2", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-DMIC", 0, 0), So when we call snd_soc_register_codec() and any of these widgets get a deferred probe we do not get an -EPROBE_DEFER (-517) back as we should and instead we just fail. Apparently the code assumes that clocks and regulators must be available at this point and not defer. After this patch it rather looks like this: ab8500-codec.0: Failed to create new controls -517 snd-soc-mop500.0: ASoC: failed to instantiate card -517 snd-soc-mop500.0: Error: snd_soc_register_card failed (-517)! (...) abx500-clk.0: registered clocks for ab850x snd-soc-mop500.0: ab8500-codec-dai.0 <-> ux500-msp-i2s.1 mapping ok snd-soc-mop500.0: ab8500-codec-dai.1 <-> ux500-msp-i2s.3 mapping ok I'm pretty happy about the patch as it it, but I'm a bit uncertain on how to proceed: there are a lot of users of the external functions snd_soc_dapm_new_control() (111 sites) and that will now return an occassional error pointer, which is not handled in the calling sites. I want an indication from the maintainers whether I should just go in and augment all these call sites, or if deferred probe is frowned upon when it leads to this much overhead. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Mark Brown <broonie@kernel.org>
2017-01-13 17:23:52 +08:00
if (IS_ERR(data->widget)) {
ret = PTR_ERR(data->widget);
goto err_data;
}
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
}
break;
case snd_soc_dapm_demux:
case snd_soc_dapm_mux:
e = (struct soc_enum *)kcontrol->private_value;
if (e->autodisable) {
struct snd_soc_dapm_widget template;
name = kasprintf(GFP_KERNEL, "%s %s", ctrl_name,
"Autodisable");
if (!name) {
ret = -ENOMEM;
goto err_data;
}
memset(&template, 0, sizeof(template));
template.reg = e->reg;
template.mask = e->mask << e->shift_l;
template.shift = e->shift_l;
template.off_val = snd_soc_enum_item_to_val(e, 0);
template.on_val = template.off_val;
template.id = snd_soc_dapm_kcontrol;
template.name = name;
data->value = template.on_val;
data->widget = snd_soc_dapm_new_control_unlocked(
widget->dapm, &template);
kfree(name);
ASoC: dapm: handle probe deferrals This starts to handle probe deferrals on regulators and clocks on the ASoC DAPM. I came to this patch after audio stopped working on Ux500 ages ago and I finally looked into it to see what is wrong. I had messages like this in the console since a while back: ab8500-codec.0: ASoC: Failed to request audioclk: -517 ab8500-codec.0: ASoC: Failed to create DAPM control audioclk ab8500-codec.0: Failed to create new controls -12 snd-soc-mop500.0: ASoC: failed to instantiate card -12 snd-soc-mop500.0: Error: snd_soc_register_card failed (-12)! snd-soc-mop500: probe of snd-soc-mop500.0 failed with error -12 Apparently because the widget table for the codec looks like this (sound/soc/codecs/ab8500-codec.c): static const struct snd_soc_dapm_widget ab8500_dapm_widgets[] = { /* Clocks */ SND_SOC_DAPM_CLOCK_SUPPLY("audioclk"), /* Regulators */ SND_SOC_DAPM_REGULATOR_SUPPLY("V-AUD", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-AMIC1", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-AMIC2", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-DMIC", 0, 0), So when we call snd_soc_register_codec() and any of these widgets get a deferred probe we do not get an -EPROBE_DEFER (-517) back as we should and instead we just fail. Apparently the code assumes that clocks and regulators must be available at this point and not defer. After this patch it rather looks like this: ab8500-codec.0: Failed to create new controls -517 snd-soc-mop500.0: ASoC: failed to instantiate card -517 snd-soc-mop500.0: Error: snd_soc_register_card failed (-517)! (...) abx500-clk.0: registered clocks for ab850x snd-soc-mop500.0: ab8500-codec-dai.0 <-> ux500-msp-i2s.1 mapping ok snd-soc-mop500.0: ab8500-codec-dai.1 <-> ux500-msp-i2s.3 mapping ok I'm pretty happy about the patch as it it, but I'm a bit uncertain on how to proceed: there are a lot of users of the external functions snd_soc_dapm_new_control() (111 sites) and that will now return an occassional error pointer, which is not handled in the calling sites. I want an indication from the maintainers whether I should just go in and augment all these call sites, or if deferred probe is frowned upon when it leads to this much overhead. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Mark Brown <broonie@kernel.org>
2017-01-13 17:23:52 +08:00
if (IS_ERR(data->widget)) {
ret = PTR_ERR(data->widget);
goto err_data;
}
snd_soc_dapm_add_path(widget->dapm, data->widget,
widget, NULL, NULL);
}
break;
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
default:
break;
}
kcontrol->private_data = data;
return 0;
err_data:
kfree(data);
return ret;
}
static void dapm_kcontrol_free(struct snd_kcontrol *kctl)
{
struct dapm_kcontrol_data *data = snd_kcontrol_chip(kctl);
list_del(&data->paths);
kfree(data->wlist);
kfree(data);
}
static struct snd_soc_dapm_widget_list *dapm_kcontrol_get_wlist(
const struct snd_kcontrol *kcontrol)
{
struct dapm_kcontrol_data *data = snd_kcontrol_chip(kcontrol);
return data->wlist;
}
static int dapm_kcontrol_add_widget(struct snd_kcontrol *kcontrol,
struct snd_soc_dapm_widget *widget)
{
struct dapm_kcontrol_data *data = snd_kcontrol_chip(kcontrol);
struct snd_soc_dapm_widget_list *new_wlist;
unsigned int n;
if (data->wlist)
n = data->wlist->num_widgets + 1;
else
n = 1;
new_wlist = krealloc(data->wlist,
sizeof(*new_wlist) + sizeof(widget) * n, GFP_KERNEL);
if (!new_wlist)
return -ENOMEM;
new_wlist->widgets[n - 1] = widget;
new_wlist->num_widgets = n;
data->wlist = new_wlist;
return 0;
}
static void dapm_kcontrol_add_path(const struct snd_kcontrol *kcontrol,
struct snd_soc_dapm_path *path)
{
struct dapm_kcontrol_data *data = snd_kcontrol_chip(kcontrol);
list_add_tail(&path->list_kcontrol, &data->paths);
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
}
static bool dapm_kcontrol_is_powered(const struct snd_kcontrol *kcontrol)
{
struct dapm_kcontrol_data *data = snd_kcontrol_chip(kcontrol);
if (!data->widget)
return true;
return data->widget->power;
}
static struct list_head *dapm_kcontrol_get_path_list(
const struct snd_kcontrol *kcontrol)
{
struct dapm_kcontrol_data *data = snd_kcontrol_chip(kcontrol);
return &data->paths;
}
#define dapm_kcontrol_for_each_path(path, kcontrol) \
list_for_each_entry(path, dapm_kcontrol_get_path_list(kcontrol), \
list_kcontrol)
unsigned int dapm_kcontrol_get_value(const struct snd_kcontrol *kcontrol)
{
struct dapm_kcontrol_data *data = snd_kcontrol_chip(kcontrol);
return data->value;
}
EXPORT_SYMBOL_GPL(dapm_kcontrol_get_value);
static bool dapm_kcontrol_set_value(const struct snd_kcontrol *kcontrol,
unsigned int value)
{
struct dapm_kcontrol_data *data = snd_kcontrol_chip(kcontrol);
if (data->value == value)
return false;
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
if (data->widget)
data->widget->on_val = value;
data->value = value;
return true;
}
/**
* snd_soc_dapm_kcontrol_widget() - Returns the widget associated to a
* kcontrol
* @kcontrol: The kcontrol
*/
struct snd_soc_dapm_widget *snd_soc_dapm_kcontrol_widget(
struct snd_kcontrol *kcontrol)
{
return dapm_kcontrol_get_wlist(kcontrol)->widgets[0];
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_kcontrol_widget);
/**
* snd_soc_dapm_kcontrol_dapm() - Returns the dapm context associated to a
* kcontrol
* @kcontrol: The kcontrol
*
* Note: This function must only be used on kcontrols that are known to have
* been registered for a CODEC. Otherwise the behaviour is undefined.
*/
struct snd_soc_dapm_context *snd_soc_dapm_kcontrol_dapm(
struct snd_kcontrol *kcontrol)
{
return dapm_kcontrol_get_wlist(kcontrol)->widgets[0]->dapm;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_kcontrol_dapm);
static void dapm_reset(struct snd_soc_card *card)
{
struct snd_soc_dapm_widget *w;
lockdep_assert_held(&card->dapm_mutex);
memset(&card->dapm_stats, 0, sizeof(card->dapm_stats));
list_for_each_entry(w, &card->widgets, list) {
w->new_power = w->power;
w->power_checked = false;
}
}
static const char *soc_dapm_prefix(struct snd_soc_dapm_context *dapm)
{
if (!dapm->component)
return NULL;
return dapm->component->name_prefix;
}
static int soc_dapm_read(struct snd_soc_dapm_context *dapm, int reg,
unsigned int *value)
{
if (!dapm->component)
return -EIO;
return snd_soc_component_read(dapm->component, reg, value);
}
static int soc_dapm_update_bits(struct snd_soc_dapm_context *dapm,
int reg, unsigned int mask, unsigned int value)
{
if (!dapm->component)
return -EIO;
return snd_soc_component_update_bits(dapm->component, reg,
mask, value);
}
static int soc_dapm_test_bits(struct snd_soc_dapm_context *dapm,
int reg, unsigned int mask, unsigned int value)
{
if (!dapm->component)
return -EIO;
return snd_soc_component_test_bits(dapm->component, reg, mask, value);
}
static void soc_dapm_async_complete(struct snd_soc_dapm_context *dapm)
{
if (dapm->component)
snd_soc_component_async_complete(dapm->component);
}
static struct snd_soc_dapm_widget *
dapm_wcache_lookup(struct snd_soc_dapm_wcache *wcache, const char *name)
{
struct snd_soc_dapm_widget *w = wcache->widget;
struct list_head *wlist;
const int depth = 2;
int i = 0;
if (w) {
wlist = &w->dapm->card->widgets;
list_for_each_entry_from(w, wlist, list) {
if (!strcmp(name, w->name))
return w;
if (++i == depth)
break;
}
}
return NULL;
}
static inline void dapm_wcache_update(struct snd_soc_dapm_wcache *wcache,
struct snd_soc_dapm_widget *w)
{
wcache->widget = w;
}
/**
* snd_soc_dapm_force_bias_level() - Sets the DAPM bias level
* @dapm: The DAPM context for which to set the level
* @level: The level to set
*
* Forces the DAPM bias level to a specific state. It will call the bias level
* callback of DAPM context with the specified level. This will even happen if
* the context is already at the same level. Furthermore it will not go through
* the normal bias level sequencing, meaning any intermediate states between the
* current and the target state will not be entered.
*
* Note that the change in bias level is only temporary and the next time
* snd_soc_dapm_sync() is called the state will be set to the level as
* determined by the DAPM core. The function is mainly intended to be used to
* used during probe or resume from suspend to power up the device so
* initialization can be done, before the DAPM core takes over.
*/
int snd_soc_dapm_force_bias_level(struct snd_soc_dapm_context *dapm,
enum snd_soc_bias_level level)
{
int ret = 0;
if (dapm->set_bias_level)
ret = dapm->set_bias_level(dapm, level);
if (ret == 0)
dapm->bias_level = level;
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_force_bias_level);
/**
* snd_soc_dapm_set_bias_level - set the bias level for the system
* @dapm: DAPM context
* @level: level to configure
*
* Configure the bias (power) levels for the SoC audio device.
*
* Returns 0 for success else error.
*/
static int snd_soc_dapm_set_bias_level(struct snd_soc_dapm_context *dapm,
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
enum snd_soc_bias_level level)
{
struct snd_soc_card *card = dapm->card;
int ret = 0;
trace_snd_soc_bias_level_start(card, level);
ASoC: multi-component - ASoC Multi-Component Support This patch extends the ASoC API to allow sound cards to have more than one CODEC and more than one platform DMA controller. This is achieved by dividing some current ASoC structures that contain both driver data and device data into structures that only either contain device data or driver data. i.e. struct snd_soc_codec ---> struct snd_soc_codec (device data) +-> struct snd_soc_codec_driver (driver data) struct snd_soc_platform ---> struct snd_soc_platform (device data) +-> struct snd_soc_platform_driver (driver data) struct snd_soc_dai ---> struct snd_soc_dai (device data) +-> struct snd_soc_dai_driver (driver data) struct snd_soc_device ---> deleted This now allows ASoC to be more tightly aligned with the Linux driver model and also means that every ASoC codec, platform and (platform) DAI is a kernel device. ASoC component private data is now stored as device private data. The ASoC sound card struct snd_soc_card has also been updated to store lists of it's components rather than a pointer to a codec and platform. The PCM runtime struct soc_pcm_runtime now has pointers to all its components. This patch adds DAPM support for ASoC multi-component and removes struct snd_soc_socdev from DAPM core. All DAPM calls are now made on a card, codec or runtime PCM level basis rather than using snd_soc_socdev. Other notable multi-component changes:- * Stream operations now de-reference less structures. * close_delayed work() now runs on a DAI basis rather than looping all DAIs in a card. * PM suspend()/resume() operations can now handle N CODECs and Platforms per sound card. * Added soc_bind_dai_link() to bind the component devices to the sound card. * Added soc_dai_link_probe() and soc_dai_link_remove() to probe and remove DAI link components. * sysfs entries can now be registered per component per card. * snd_soc_new_pcms() functionailty rolled into dai_link_probe(). * snd_soc_register_codec() now does all the codec list and mutex init. This patch changes the probe() and remove() of the CODEC drivers as follows:- o Make CODEC driver a platform driver o Moved all struct snd_soc_codec list, mutex, etc initialiasation to core. o Removed all static codec pointers (drivers now support > 1 codec dev) o snd_soc_register_pcms() now done by core. o snd_soc_register_dai() folded into snd_soc_register_codec(). CS4270 portions: Acked-by: Timur Tabi <timur@freescale.com> Some TLV320aic23 and Cirrus platform fixes. Signed-off-by: Ryan Mallon <ryan@bluewatersys.com> TI CODEC and OMAP fixes Signed-off-by: Peter Ujfalusi <peter.ujfalusi@nokia.com> Signed-off-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Samsung platform and misc fixes :- Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com> Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Reviewed-by: Jassi Brar <jassi.brar@samsung.com> Signed-off-by: Seungwhan Youn <sw.youn@samsung.com> MPC8610 and PPC fixes. Signed-off-by: Timur Tabi <timur@freescale.com> i.MX fixes and some core fixes. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> J4740 platform fixes:- Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> CC: Tony Lindgren <tony@atomide.com> CC: Nicolas Ferre <nicolas.ferre@atmel.com> CC: Kevin Hilman <khilman@deeprootsystems.com> CC: Sascha Hauer <s.hauer@pengutronix.de> CC: Atsushi Nemoto <anemo@mba.ocn.ne.jp> CC: Kuninori Morimoto <morimoto.kuninori@renesas.com> CC: Daniel Gloeckner <dg@emlix.com> CC: Manuel Lauss <mano@roarinelk.homelinux.net> CC: Mike Frysinger <vapier.adi@gmail.com> CC: Arnaud Patard <apatard@mandriva.com> CC: Wan ZongShun <mcuos.com@gmail.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk>
2010-03-18 04:15:21 +08:00
if (card && card->set_bias_level)
ret = card->set_bias_level(card, dapm, level);
if (ret != 0)
goto out;
if (!card || dapm != &card->dapm)
ret = snd_soc_dapm_force_bias_level(dapm, level);
if (ret != 0)
goto out;
if (card && card->set_bias_level_post)
ret = card->set_bias_level_post(card, dapm, level);
out:
trace_snd_soc_bias_level_done(card, level);
return ret;
}
/* connect mux widget to its interconnecting audio paths */
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
static int dapm_connect_mux(struct snd_soc_dapm_context *dapm,
struct snd_soc_dapm_path *path, const char *control_name,
struct snd_soc_dapm_widget *w)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
const struct snd_kcontrol_new *kcontrol = &w->kcontrol_news[0];
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct soc_enum *e = (struct soc_enum *)kcontrol->private_value;
unsigned int val, item;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
int i;
if (e->reg != SND_SOC_NOPM) {
soc_dapm_read(dapm, e->reg, &val);
val = (val >> e->shift_l) & e->mask;
item = snd_soc_enum_val_to_item(e, val);
} else {
/* since a virtual mux has no backing registers to
* decide which path to connect, it will try to match
* with the first enumeration. This is to ensure
* that the default mux choice (the first) will be
* correctly powered up during initialization.
*/
item = 0;
}
i = match_string(e->texts, e->items, control_name);
if (i < 0)
return -ENODEV;
path->name = e->texts[i];
path->connect = (i == item);
return 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
/* set up initial codec paths */
static void dapm_set_mixer_path_status(struct snd_soc_dapm_path *p, int i,
int nth_path)
{
struct soc_mixer_control *mc = (struct soc_mixer_control *)
p->sink->kcontrol_news[i].private_value;
unsigned int reg = mc->reg;
unsigned int shift = mc->shift;
unsigned int max = mc->max;
unsigned int mask = (1 << fls(max)) - 1;
unsigned int invert = mc->invert;
unsigned int val;
if (reg != SND_SOC_NOPM) {
soc_dapm_read(p->sink->dapm, reg, &val);
/*
* The nth_path argument allows this function to know
* which path of a kcontrol it is setting the initial
* status for. Ideally this would support any number
* of paths and channels. But since kcontrols only come
* in mono and stereo variants, we are limited to 2
* channels.
*
* The following code assumes for stereo controls the
* first path is the left channel, and all remaining
* paths are the right channel.
*/
if (snd_soc_volsw_is_stereo(mc) && nth_path > 0) {
if (reg != mc->rreg)
soc_dapm_read(p->sink->dapm, mc->rreg, &val);
val = (val >> mc->rshift) & mask;
} else {
val = (val >> shift) & mask;
}
if (invert)
val = max - val;
p->connect = !!val;
} else {
p->connect = 0;
}
}
/* connect mixer widget to its interconnecting audio paths */
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
static int dapm_connect_mixer(struct snd_soc_dapm_context *dapm,
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct snd_soc_dapm_path *path, const char *control_name)
{
int i, nth_path = 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* search for mixer kcontrol */
for (i = 0; i < path->sink->num_kcontrols; i++) {
if (!strcmp(control_name, path->sink->kcontrol_news[i].name)) {
path->name = path->sink->kcontrol_news[i].name;
dapm_set_mixer_path_status(path, i, nth_path++);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return 0;
}
}
return -ENODEV;
}
ASoC: Implement mux control sharing Control sharing is enabled when two widgets include pointers to the same kcontrol_new in their definition. Specifically: static const struct snd_kcontrol_new adcinput_mux = SOC_DAPM_ENUM("ADC Input", adcinput_enum); static const struct snd_soc_dapm_widget wm8903_dapm_widgets[] = { SND_SOC_DAPM_MUX("Left ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), SND_SOC_DAPM_MUX("Right ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), }; This is useful when a single register bit or field affects multiple muxes at once. The common case is to have separate control bits or fields for each mux (channel). An alternative way of looking at this is that the mux is a stereo (or even n-channel) mux, rather than independant mono muxes. Without this change, a separate kcontrol will be created for each DAPM_MUX. This has the following disadvantages: * Confuses the user/programmer with redundant controls that don't map to separate hardware. * When one of the controls is changed, ASoC fails to update the DAPM logic for paths solely affected by the other controls impacted by the same register bits. This causes some paths not to be correctly powered up or down. Prior to this change, to work around this, the user or programmer had to manually toggle all duplicate controls away from the intended setting, and then back to it. Control sharing implies that the control is named based on the kcontrol_new itself, not any of the widgets that are affected by it. Control sharing is implemented by: When creating kcontrols, if a kcontrol does not yet exist for a particular kcontrol_new, then a new kcontrol is created with a list of widgets containing just a single entry. This is the normal case. However, if a kcontrol does already exists for the given kcontrol_new, the current widget is simply added to that kcontrol's list of affected widgets. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-04-29 07:38:01 +08:00
static int dapm_is_shared_kcontrol(struct snd_soc_dapm_context *dapm,
ASoC: Fix dapm_is_shared_kcontrol so everything isn't shared Commit af46800 ("ASoC: Implement mux control sharing") introduced function dapm_is_shared_kcontrol. When this function returns true, the naming of DAPM controls is derived from the kcontrol_new. Otherwise, the name comes from the widget (and possibly a widget's naming prefix). A bug in the implementation of dapm_is_shared_kcontrol made it return 1 in all cases. Hence, that commit caused a change in control naming for all controls instead of just shared controls. Specifically, a control is always considered shared because it is always compared against itself. Solve this by never comparing against the widget containing the control being created. Equally, controls should never be shared between DAPM contexts; when the same codec is instantiated multiple times, the same kcontrol_new will be used. However, the control should no be shared between the multiple instances. I tested that with the Tegra WM8903 driver: * Shared is now mostly 0 as expected, and sometimes 1. * The expected controls are still generated after this change. However, I don't have any systems that have a widget/control naming prefix, so I can't test that aspect. Thanks for Jarkko Nikula for pointing out how to fix this. Reported-by: Liam Girdwood <lrg@ti.com> Tested-by: Jarkko Nikula <jhnikula@gmail.com> Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-05-26 23:57:33 +08:00
struct snd_soc_dapm_widget *kcontrolw,
ASoC: Implement mux control sharing Control sharing is enabled when two widgets include pointers to the same kcontrol_new in their definition. Specifically: static const struct snd_kcontrol_new adcinput_mux = SOC_DAPM_ENUM("ADC Input", adcinput_enum); static const struct snd_soc_dapm_widget wm8903_dapm_widgets[] = { SND_SOC_DAPM_MUX("Left ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), SND_SOC_DAPM_MUX("Right ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), }; This is useful when a single register bit or field affects multiple muxes at once. The common case is to have separate control bits or fields for each mux (channel). An alternative way of looking at this is that the mux is a stereo (or even n-channel) mux, rather than independant mono muxes. Without this change, a separate kcontrol will be created for each DAPM_MUX. This has the following disadvantages: * Confuses the user/programmer with redundant controls that don't map to separate hardware. * When one of the controls is changed, ASoC fails to update the DAPM logic for paths solely affected by the other controls impacted by the same register bits. This causes some paths not to be correctly powered up or down. Prior to this change, to work around this, the user or programmer had to manually toggle all duplicate controls away from the intended setting, and then back to it. Control sharing implies that the control is named based on the kcontrol_new itself, not any of the widgets that are affected by it. Control sharing is implemented by: When creating kcontrols, if a kcontrol does not yet exist for a particular kcontrol_new, then a new kcontrol is created with a list of widgets containing just a single entry. This is the normal case. However, if a kcontrol does already exists for the given kcontrol_new, the current widget is simply added to that kcontrol's list of affected widgets. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-04-29 07:38:01 +08:00
const struct snd_kcontrol_new *kcontrol_new,
struct snd_kcontrol **kcontrol)
{
struct snd_soc_dapm_widget *w;
int i;
*kcontrol = NULL;
list_for_each_entry(w, &dapm->card->widgets, list) {
ASoC: Fix dapm_is_shared_kcontrol so everything isn't shared Commit af46800 ("ASoC: Implement mux control sharing") introduced function dapm_is_shared_kcontrol. When this function returns true, the naming of DAPM controls is derived from the kcontrol_new. Otherwise, the name comes from the widget (and possibly a widget's naming prefix). A bug in the implementation of dapm_is_shared_kcontrol made it return 1 in all cases. Hence, that commit caused a change in control naming for all controls instead of just shared controls. Specifically, a control is always considered shared because it is always compared against itself. Solve this by never comparing against the widget containing the control being created. Equally, controls should never be shared between DAPM contexts; when the same codec is instantiated multiple times, the same kcontrol_new will be used. However, the control should no be shared between the multiple instances. I tested that with the Tegra WM8903 driver: * Shared is now mostly 0 as expected, and sometimes 1. * The expected controls are still generated after this change. However, I don't have any systems that have a widget/control naming prefix, so I can't test that aspect. Thanks for Jarkko Nikula for pointing out how to fix this. Reported-by: Liam Girdwood <lrg@ti.com> Tested-by: Jarkko Nikula <jhnikula@gmail.com> Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-05-26 23:57:33 +08:00
if (w == kcontrolw || w->dapm != kcontrolw->dapm)
continue;
ASoC: Implement mux control sharing Control sharing is enabled when two widgets include pointers to the same kcontrol_new in their definition. Specifically: static const struct snd_kcontrol_new adcinput_mux = SOC_DAPM_ENUM("ADC Input", adcinput_enum); static const struct snd_soc_dapm_widget wm8903_dapm_widgets[] = { SND_SOC_DAPM_MUX("Left ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), SND_SOC_DAPM_MUX("Right ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), }; This is useful when a single register bit or field affects multiple muxes at once. The common case is to have separate control bits or fields for each mux (channel). An alternative way of looking at this is that the mux is a stereo (or even n-channel) mux, rather than independant mono muxes. Without this change, a separate kcontrol will be created for each DAPM_MUX. This has the following disadvantages: * Confuses the user/programmer with redundant controls that don't map to separate hardware. * When one of the controls is changed, ASoC fails to update the DAPM logic for paths solely affected by the other controls impacted by the same register bits. This causes some paths not to be correctly powered up or down. Prior to this change, to work around this, the user or programmer had to manually toggle all duplicate controls away from the intended setting, and then back to it. Control sharing implies that the control is named based on the kcontrol_new itself, not any of the widgets that are affected by it. Control sharing is implemented by: When creating kcontrols, if a kcontrol does not yet exist for a particular kcontrol_new, then a new kcontrol is created with a list of widgets containing just a single entry. This is the normal case. However, if a kcontrol does already exists for the given kcontrol_new, the current widget is simply added to that kcontrol's list of affected widgets. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-04-29 07:38:01 +08:00
for (i = 0; i < w->num_kcontrols; i++) {
if (&w->kcontrol_news[i] == kcontrol_new) {
if (w->kcontrols)
*kcontrol = w->kcontrols[i];
return 1;
}
}
}
return 0;
}
/*
* Determine if a kcontrol is shared. If it is, look it up. If it isn't,
* create it. Either way, add the widget into the control's widget list
*/
static int dapm_create_or_share_kcontrol(struct snd_soc_dapm_widget *w,
int kci)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_context *dapm = w->dapm;
struct snd_card *card = dapm->card->snd_card;
const char *prefix;
size_t prefix_len;
int shared;
struct snd_kcontrol *kcontrol;
bool wname_in_long_name, kcname_in_long_name;
char *long_name = NULL;
const char *name;
int ret = 0;
prefix = soc_dapm_prefix(dapm);
if (prefix)
prefix_len = strlen(prefix) + 1;
else
prefix_len = 0;
shared = dapm_is_shared_kcontrol(dapm, w, &w->kcontrol_news[kci],
&kcontrol);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
if (!kcontrol) {
if (shared) {
wname_in_long_name = false;
kcname_in_long_name = true;
} else {
switch (w->id) {
case snd_soc_dapm_switch:
case snd_soc_dapm_mixer:
case snd_soc_dapm_pga:
case snd_soc_dapm_out_drv:
wname_in_long_name = true;
kcname_in_long_name = true;
break;
case snd_soc_dapm_mixer_named_ctl:
wname_in_long_name = false;
kcname_in_long_name = true;
break;
case snd_soc_dapm_demux:
case snd_soc_dapm_mux:
wname_in_long_name = true;
kcname_in_long_name = false;
break;
default:
return -EINVAL;
}
}
if (wname_in_long_name && kcname_in_long_name) {
/*
* The control will get a prefix from the control
* creation process but we're also using the same
* prefix for widgets so cut the prefix off the
* front of the widget name.
*/
long_name = kasprintf(GFP_KERNEL, "%s %s",
w->name + prefix_len,
w->kcontrol_news[kci].name);
if (long_name == NULL)
return -ENOMEM;
name = long_name;
} else if (wname_in_long_name) {
long_name = NULL;
name = w->name + prefix_len;
} else {
long_name = NULL;
name = w->kcontrol_news[kci].name;
}
kcontrol = snd_soc_cnew(&w->kcontrol_news[kci], NULL, name,
prefix);
if (!kcontrol) {
ret = -ENOMEM;
goto exit_free;
}
kcontrol->private_free = dapm_kcontrol_free;
ret = dapm_kcontrol_data_alloc(w, kcontrol, name);
if (ret) {
snd_ctl_free_one(kcontrol);
goto exit_free;
}
ret = snd_ctl_add(card, kcontrol);
if (ret < 0) {
dev_err(dapm->dev,
"ASoC: failed to add widget %s dapm kcontrol %s: %d\n",
w->name, name, ret);
goto exit_free;
}
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
ret = dapm_kcontrol_add_widget(kcontrol, w);
if (ret == 0)
w->kcontrols[kci] = kcontrol;
exit_free:
kfree(long_name);
return ret;
}
/* create new dapm mixer control */
static int dapm_new_mixer(struct snd_soc_dapm_widget *w)
{
int i, ret;
struct snd_soc_dapm_path *path;
struct dapm_kcontrol_data *data;
/* add kcontrol */
for (i = 0; i < w->num_kcontrols; i++) {
/* match name */
snd_soc_dapm_widget_for_each_source_path(w, path) {
/* mixer/mux paths name must match control name */
if (path->name != (char *)w->kcontrol_news[i].name)
continue;
if (!w->kcontrols[i]) {
ret = dapm_create_or_share_kcontrol(w, i);
if (ret < 0)
return ret;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
dapm_kcontrol_add_path(w->kcontrols[i], path);
data = snd_kcontrol_chip(w->kcontrols[i]);
if (data->widget)
snd_soc_dapm_add_path(data->widget->dapm,
data->widget,
path->source,
NULL, NULL);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
}
return 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
/* create new dapm mux control */
static int dapm_new_mux(struct snd_soc_dapm_widget *w)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_context *dapm = w->dapm;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
enum snd_soc_dapm_direction dir;
struct snd_soc_dapm_path *path;
const char *type;
ASoC: Implement mux control sharing Control sharing is enabled when two widgets include pointers to the same kcontrol_new in their definition. Specifically: static const struct snd_kcontrol_new adcinput_mux = SOC_DAPM_ENUM("ADC Input", adcinput_enum); static const struct snd_soc_dapm_widget wm8903_dapm_widgets[] = { SND_SOC_DAPM_MUX("Left ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), SND_SOC_DAPM_MUX("Right ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), }; This is useful when a single register bit or field affects multiple muxes at once. The common case is to have separate control bits or fields for each mux (channel). An alternative way of looking at this is that the mux is a stereo (or even n-channel) mux, rather than independant mono muxes. Without this change, a separate kcontrol will be created for each DAPM_MUX. This has the following disadvantages: * Confuses the user/programmer with redundant controls that don't map to separate hardware. * When one of the controls is changed, ASoC fails to update the DAPM logic for paths solely affected by the other controls impacted by the same register bits. This causes some paths not to be correctly powered up or down. Prior to this change, to work around this, the user or programmer had to manually toggle all duplicate controls away from the intended setting, and then back to it. Control sharing implies that the control is named based on the kcontrol_new itself, not any of the widgets that are affected by it. Control sharing is implemented by: When creating kcontrols, if a kcontrol does not yet exist for a particular kcontrol_new, then a new kcontrol is created with a list of widgets containing just a single entry. This is the normal case. However, if a kcontrol does already exists for the given kcontrol_new, the current widget is simply added to that kcontrol's list of affected widgets. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-04-29 07:38:01 +08:00
int ret;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
switch (w->id) {
case snd_soc_dapm_mux:
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
dir = SND_SOC_DAPM_DIR_OUT;
type = "mux";
break;
case snd_soc_dapm_demux:
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
dir = SND_SOC_DAPM_DIR_IN;
type = "demux";
break;
default:
return -EINVAL;
}
ASoC: Implement mux control sharing Control sharing is enabled when two widgets include pointers to the same kcontrol_new in their definition. Specifically: static const struct snd_kcontrol_new adcinput_mux = SOC_DAPM_ENUM("ADC Input", adcinput_enum); static const struct snd_soc_dapm_widget wm8903_dapm_widgets[] = { SND_SOC_DAPM_MUX("Left ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), SND_SOC_DAPM_MUX("Right ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), }; This is useful when a single register bit or field affects multiple muxes at once. The common case is to have separate control bits or fields for each mux (channel). An alternative way of looking at this is that the mux is a stereo (or even n-channel) mux, rather than independant mono muxes. Without this change, a separate kcontrol will be created for each DAPM_MUX. This has the following disadvantages: * Confuses the user/programmer with redundant controls that don't map to separate hardware. * When one of the controls is changed, ASoC fails to update the DAPM logic for paths solely affected by the other controls impacted by the same register bits. This causes some paths not to be correctly powered up or down. Prior to this change, to work around this, the user or programmer had to manually toggle all duplicate controls away from the intended setting, and then back to it. Control sharing implies that the control is named based on the kcontrol_new itself, not any of the widgets that are affected by it. Control sharing is implemented by: When creating kcontrols, if a kcontrol does not yet exist for a particular kcontrol_new, then a new kcontrol is created with a list of widgets containing just a single entry. This is the normal case. However, if a kcontrol does already exists for the given kcontrol_new, the current widget is simply added to that kcontrol's list of affected widgets. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-04-29 07:38:01 +08:00
if (w->num_kcontrols != 1) {
dev_err(dapm->dev,
"ASoC: %s %s has incorrect number of controls\n", type,
ASoC: Implement mux control sharing Control sharing is enabled when two widgets include pointers to the same kcontrol_new in their definition. Specifically: static const struct snd_kcontrol_new adcinput_mux = SOC_DAPM_ENUM("ADC Input", adcinput_enum); static const struct snd_soc_dapm_widget wm8903_dapm_widgets[] = { SND_SOC_DAPM_MUX("Left ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), SND_SOC_DAPM_MUX("Right ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), }; This is useful when a single register bit or field affects multiple muxes at once. The common case is to have separate control bits or fields for each mux (channel). An alternative way of looking at this is that the mux is a stereo (or even n-channel) mux, rather than independant mono muxes. Without this change, a separate kcontrol will be created for each DAPM_MUX. This has the following disadvantages: * Confuses the user/programmer with redundant controls that don't map to separate hardware. * When one of the controls is changed, ASoC fails to update the DAPM logic for paths solely affected by the other controls impacted by the same register bits. This causes some paths not to be correctly powered up or down. Prior to this change, to work around this, the user or programmer had to manually toggle all duplicate controls away from the intended setting, and then back to it. Control sharing implies that the control is named based on the kcontrol_new itself, not any of the widgets that are affected by it. Control sharing is implemented by: When creating kcontrols, if a kcontrol does not yet exist for a particular kcontrol_new, then a new kcontrol is created with a list of widgets containing just a single entry. This is the normal case. However, if a kcontrol does already exists for the given kcontrol_new, the current widget is simply added to that kcontrol's list of affected widgets. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-04-29 07:38:01 +08:00
w->name);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return -EINVAL;
}
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (list_empty(&w->edges[dir])) {
dev_err(dapm->dev, "ASoC: %s %s has no paths\n", type, w->name);
return -EINVAL;
ASoC: Implement mux control sharing Control sharing is enabled when two widgets include pointers to the same kcontrol_new in their definition. Specifically: static const struct snd_kcontrol_new adcinput_mux = SOC_DAPM_ENUM("ADC Input", adcinput_enum); static const struct snd_soc_dapm_widget wm8903_dapm_widgets[] = { SND_SOC_DAPM_MUX("Left ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), SND_SOC_DAPM_MUX("Right ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), }; This is useful when a single register bit or field affects multiple muxes at once. The common case is to have separate control bits or fields for each mux (channel). An alternative way of looking at this is that the mux is a stereo (or even n-channel) mux, rather than independant mono muxes. Without this change, a separate kcontrol will be created for each DAPM_MUX. This has the following disadvantages: * Confuses the user/programmer with redundant controls that don't map to separate hardware. * When one of the controls is changed, ASoC fails to update the DAPM logic for paths solely affected by the other controls impacted by the same register bits. This causes some paths not to be correctly powered up or down. Prior to this change, to work around this, the user or programmer had to manually toggle all duplicate controls away from the intended setting, and then back to it. Control sharing implies that the control is named based on the kcontrol_new itself, not any of the widgets that are affected by it. Control sharing is implemented by: When creating kcontrols, if a kcontrol does not yet exist for a particular kcontrol_new, then a new kcontrol is created with a list of widgets containing just a single entry. This is the normal case. However, if a kcontrol does already exists for the given kcontrol_new, the current widget is simply added to that kcontrol's list of affected widgets. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-04-29 07:38:01 +08:00
}
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
ret = dapm_create_or_share_kcontrol(w, 0);
if (ret < 0)
return ret;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
snd_soc_dapm_widget_for_each_path(w, dir, path) {
if (path->name)
dapm_kcontrol_add_path(w->kcontrols[0], path);
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
ASoC: Implement mux control sharing Control sharing is enabled when two widgets include pointers to the same kcontrol_new in their definition. Specifically: static const struct snd_kcontrol_new adcinput_mux = SOC_DAPM_ENUM("ADC Input", adcinput_enum); static const struct snd_soc_dapm_widget wm8903_dapm_widgets[] = { SND_SOC_DAPM_MUX("Left ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), SND_SOC_DAPM_MUX("Right ADC Input", SND_SOC_NOPM, 0, 0, &adcinput_mux), }; This is useful when a single register bit or field affects multiple muxes at once. The common case is to have separate control bits or fields for each mux (channel). An alternative way of looking at this is that the mux is a stereo (or even n-channel) mux, rather than independant mono muxes. Without this change, a separate kcontrol will be created for each DAPM_MUX. This has the following disadvantages: * Confuses the user/programmer with redundant controls that don't map to separate hardware. * When one of the controls is changed, ASoC fails to update the DAPM logic for paths solely affected by the other controls impacted by the same register bits. This causes some paths not to be correctly powered up or down. Prior to this change, to work around this, the user or programmer had to manually toggle all duplicate controls away from the intended setting, and then back to it. Control sharing implies that the control is named based on the kcontrol_new itself, not any of the widgets that are affected by it. Control sharing is implemented by: When creating kcontrols, if a kcontrol does not yet exist for a particular kcontrol_new, then a new kcontrol is created with a list of widgets containing just a single entry. This is the normal case. However, if a kcontrol does already exists for the given kcontrol_new, the current widget is simply added to that kcontrol's list of affected widgets. Signed-off-by: Stephen Warren <swarren@nvidia.com> Acked-by: Liam Girdwood <lrg@ti.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2011-04-29 07:38:01 +08:00
return 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
/* create new dapm volume control */
static int dapm_new_pga(struct snd_soc_dapm_widget *w)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
int i, ret;
for (i = 0; i < w->num_kcontrols; i++) {
ret = dapm_create_or_share_kcontrol(w, i);
if (ret < 0)
return ret;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
/* create new dapm dai link control */
static int dapm_new_dai_link(struct snd_soc_dapm_widget *w)
{
int i, ret;
struct snd_kcontrol *kcontrol;
struct snd_soc_dapm_context *dapm = w->dapm;
struct snd_card *card = dapm->card->snd_card;
struct snd_soc_pcm_runtime *rtd = w->priv;
/* create control for links with > 1 config */
if (rtd->dai_link->num_params <= 1)
return 0;
/* add kcontrol */
for (i = 0; i < w->num_kcontrols; i++) {
kcontrol = snd_soc_cnew(&w->kcontrol_news[i], w,
w->name, NULL);
ret = snd_ctl_add(card, kcontrol);
if (ret < 0) {
dev_err(dapm->dev,
"ASoC: failed to add widget %s dapm kcontrol %s: %d\n",
w->name, w->kcontrol_news[i].name, ret);
return ret;
}
kcontrol->private_data = w;
w->kcontrols[i] = kcontrol;
}
return 0;
}
/* We implement power down on suspend by checking the power state of
* the ALSA card - when we are suspending the ALSA state for the card
* is set to D3.
*/
static int snd_soc_dapm_suspend_check(struct snd_soc_dapm_widget *widget)
{
int level = snd_power_get_state(widget->dapm->card->snd_card);
ASoC: multi-component - ASoC Multi-Component Support This patch extends the ASoC API to allow sound cards to have more than one CODEC and more than one platform DMA controller. This is achieved by dividing some current ASoC structures that contain both driver data and device data into structures that only either contain device data or driver data. i.e. struct snd_soc_codec ---> struct snd_soc_codec (device data) +-> struct snd_soc_codec_driver (driver data) struct snd_soc_platform ---> struct snd_soc_platform (device data) +-> struct snd_soc_platform_driver (driver data) struct snd_soc_dai ---> struct snd_soc_dai (device data) +-> struct snd_soc_dai_driver (driver data) struct snd_soc_device ---> deleted This now allows ASoC to be more tightly aligned with the Linux driver model and also means that every ASoC codec, platform and (platform) DAI is a kernel device. ASoC component private data is now stored as device private data. The ASoC sound card struct snd_soc_card has also been updated to store lists of it's components rather than a pointer to a codec and platform. The PCM runtime struct soc_pcm_runtime now has pointers to all its components. This patch adds DAPM support for ASoC multi-component and removes struct snd_soc_socdev from DAPM core. All DAPM calls are now made on a card, codec or runtime PCM level basis rather than using snd_soc_socdev. Other notable multi-component changes:- * Stream operations now de-reference less structures. * close_delayed work() now runs on a DAI basis rather than looping all DAIs in a card. * PM suspend()/resume() operations can now handle N CODECs and Platforms per sound card. * Added soc_bind_dai_link() to bind the component devices to the sound card. * Added soc_dai_link_probe() and soc_dai_link_remove() to probe and remove DAI link components. * sysfs entries can now be registered per component per card. * snd_soc_new_pcms() functionailty rolled into dai_link_probe(). * snd_soc_register_codec() now does all the codec list and mutex init. This patch changes the probe() and remove() of the CODEC drivers as follows:- o Make CODEC driver a platform driver o Moved all struct snd_soc_codec list, mutex, etc initialiasation to core. o Removed all static codec pointers (drivers now support > 1 codec dev) o snd_soc_register_pcms() now done by core. o snd_soc_register_dai() folded into snd_soc_register_codec(). CS4270 portions: Acked-by: Timur Tabi <timur@freescale.com> Some TLV320aic23 and Cirrus platform fixes. Signed-off-by: Ryan Mallon <ryan@bluewatersys.com> TI CODEC and OMAP fixes Signed-off-by: Peter Ujfalusi <peter.ujfalusi@nokia.com> Signed-off-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Samsung platform and misc fixes :- Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com> Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Reviewed-by: Jassi Brar <jassi.brar@samsung.com> Signed-off-by: Seungwhan Youn <sw.youn@samsung.com> MPC8610 and PPC fixes. Signed-off-by: Timur Tabi <timur@freescale.com> i.MX fixes and some core fixes. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> J4740 platform fixes:- Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> CC: Tony Lindgren <tony@atomide.com> CC: Nicolas Ferre <nicolas.ferre@atmel.com> CC: Kevin Hilman <khilman@deeprootsystems.com> CC: Sascha Hauer <s.hauer@pengutronix.de> CC: Atsushi Nemoto <anemo@mba.ocn.ne.jp> CC: Kuninori Morimoto <morimoto.kuninori@renesas.com> CC: Daniel Gloeckner <dg@emlix.com> CC: Manuel Lauss <mano@roarinelk.homelinux.net> CC: Mike Frysinger <vapier.adi@gmail.com> CC: Arnaud Patard <apatard@mandriva.com> CC: Wan ZongShun <mcuos.com@gmail.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk>
2010-03-18 04:15:21 +08:00
switch (level) {
case SNDRV_CTL_POWER_D3hot:
case SNDRV_CTL_POWER_D3cold:
if (widget->ignore_suspend)
dev_dbg(widget->dapm->dev, "ASoC: %s ignoring suspend\n",
widget->name);
return widget->ignore_suspend;
default:
return 1;
}
}
static int dapm_widget_list_create(struct snd_soc_dapm_widget_list **list,
struct list_head *widgets)
{
struct snd_soc_dapm_widget *w;
struct list_head *it;
unsigned int size = 0;
unsigned int i = 0;
list_for_each(it, widgets)
size++;
treewide: Use struct_size() for kmalloc()-family One of the more common cases of allocation size calculations is finding the size of a structure that has a zero-sized array at the end, along with memory for some number of elements for that array. For example: struct foo { int stuff; void *entry[]; }; instance = kmalloc(sizeof(struct foo) + sizeof(void *) * count, GFP_KERNEL); Instead of leaving these open-coded and prone to type mistakes, we can now use the new struct_size() helper: instance = kmalloc(struct_size(instance, entry, count), GFP_KERNEL); This patch makes the changes for kmalloc()-family (and kvmalloc()-family) uses. It was done via automatic conversion with manual review for the "CHECKME" non-standard cases noted below, using the following Coccinelle script: // pkey_cache = kmalloc(sizeof *pkey_cache + tprops->pkey_tbl_len * // sizeof *pkey_cache->table, GFP_KERNEL); @@ identifier alloc =~ "kmalloc|kzalloc|kvmalloc|kvzalloc"; expression GFP; identifier VAR, ELEMENT; expression COUNT; @@ - alloc(sizeof(*VAR) + COUNT * sizeof(*VAR->ELEMENT), GFP) + alloc(struct_size(VAR, ELEMENT, COUNT), GFP) // mr = kzalloc(sizeof(*mr) + m * sizeof(mr->map[0]), GFP_KERNEL); @@ identifier alloc =~ "kmalloc|kzalloc|kvmalloc|kvzalloc"; expression GFP; identifier VAR, ELEMENT; expression COUNT; @@ - alloc(sizeof(*VAR) + COUNT * sizeof(VAR->ELEMENT[0]), GFP) + alloc(struct_size(VAR, ELEMENT, COUNT), GFP) // Same pattern, but can't trivially locate the trailing element name, // or variable name. @@ identifier alloc =~ "kmalloc|kzalloc|kvmalloc|kvzalloc"; expression GFP; expression SOMETHING, COUNT, ELEMENT; @@ - alloc(sizeof(SOMETHING) + COUNT * sizeof(ELEMENT), GFP) + alloc(CHECKME_struct_size(&SOMETHING, ELEMENT, COUNT), GFP) Signed-off-by: Kees Cook <keescook@chromium.org>
2018-05-09 04:45:50 +08:00
*list = kzalloc(struct_size(*list, widgets, size), GFP_KERNEL);
if (*list == NULL)
return -ENOMEM;
list_for_each_entry(w, widgets, work_list)
(*list)->widgets[i++] = w;
(*list)->num_widgets = i;
return 0;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/*
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
* Common implementation for is_connected_output_ep() and
* is_connected_input_ep(). The function is inlined since the combined size of
* the two specialized functions is only marginally larger then the size of the
* generic function and at the same time the fast path of the specialized
* functions is significantly smaller than the generic function.
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*/
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
static __always_inline int is_connected_ep(struct snd_soc_dapm_widget *widget,
struct list_head *list, enum snd_soc_dapm_direction dir,
int (*fn)(struct snd_soc_dapm_widget *, struct list_head *,
bool (*custom_stop_condition)(struct snd_soc_dapm_widget *,
enum snd_soc_dapm_direction)),
bool (*custom_stop_condition)(struct snd_soc_dapm_widget *,
enum snd_soc_dapm_direction))
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
enum snd_soc_dapm_direction rdir = SND_SOC_DAPM_DIR_REVERSE(dir);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct snd_soc_dapm_path *path;
int con = 0;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (widget->endpoints[dir] >= 0)
return widget->endpoints[dir];
DAPM_UPDATE_STAT(widget, path_checks);
/* do we need to add this widget to the list ? */
if (list)
list_add_tail(&widget->work_list, list);
if (custom_stop_condition && custom_stop_condition(widget, dir)) {
widget->endpoints[dir] = 1;
return widget->endpoints[dir];
}
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if ((widget->is_ep & SND_SOC_DAPM_DIR_TO_EP(dir)) && widget->connected) {
widget->endpoints[dir] = snd_soc_dapm_suspend_check(widget);
return widget->endpoints[dir];
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
snd_soc_dapm_widget_for_each_path(widget, rdir, path) {
DAPM_UPDATE_STAT(widget, neighbour_checks);
2014-10-25 23:42:00 +08:00
if (path->weak || path->is_supply)
continue;
if (path->walking)
return 1;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
trace_snd_soc_dapm_path(widget, dir, path);
if (path->connect) {
path->walking = 1;
con += fn(path->node[dir], list, custom_stop_condition);
path->walking = 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
}
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
widget->endpoints[dir] = con;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return con;
}
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
/*
* Recursively check for a completed path to an active or physically connected
* output widget. Returns number of complete paths.
*
* Optionally, can be supplied with a function acting as a stopping condition.
* This function takes the dapm widget currently being examined and the walk
* direction as an arguments, it should return true if the walk should be
* stopped and false otherwise.
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
*/
static int is_connected_output_ep(struct snd_soc_dapm_widget *widget,
struct list_head *list,
bool (*custom_stop_condition)(struct snd_soc_dapm_widget *i,
enum snd_soc_dapm_direction))
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
{
return is_connected_ep(widget, list, SND_SOC_DAPM_DIR_OUT,
is_connected_output_ep, custom_stop_condition);
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/*
* Recursively check for a completed path to an active or physically connected
* input widget. Returns number of complete paths.
*
* Optionally, can be supplied with a function acting as a stopping condition.
* This function takes the dapm widget currently being examined and the walk
* direction as an arguments, it should return true if the walk should be
* stopped and false otherwise.
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*/
static int is_connected_input_ep(struct snd_soc_dapm_widget *widget,
struct list_head *list,
bool (*custom_stop_condition)(struct snd_soc_dapm_widget *i,
enum snd_soc_dapm_direction))
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
return is_connected_ep(widget, list, SND_SOC_DAPM_DIR_IN,
is_connected_input_ep, custom_stop_condition);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
/**
* snd_soc_dapm_get_connected_widgets - query audio path and it's widgets.
* @dai: the soc DAI.
* @stream: stream direction.
* @list: list of active widgets for this stream.
* @custom_stop_condition: (optional) a function meant to stop the widget graph
* walk based on custom logic.
*
* Queries DAPM graph as to whether a valid audio stream path exists for
* the initial stream specified by name. This takes into account
* current mixer and mux kcontrol settings. Creates list of valid widgets.
*
* Optionally, can be supplied with a function acting as a stopping condition.
* This function takes the dapm widget currently being examined and the walk
* direction as an arguments, it should return true if the walk should be
* stopped and false otherwise.
*
* Returns the number of valid paths or negative error.
*/
int snd_soc_dapm_dai_get_connected_widgets(struct snd_soc_dai *dai, int stream,
struct snd_soc_dapm_widget_list **list,
bool (*custom_stop_condition)(struct snd_soc_dapm_widget *,
enum snd_soc_dapm_direction))
{
struct snd_soc_card *card = dai->component->card;
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
struct snd_soc_dapm_widget *w;
LIST_HEAD(widgets);
int paths;
int ret;
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
/*
* For is_connected_{output,input}_ep fully discover the graph we need
* to reset the cached number of inputs and outputs.
*/
list_for_each_entry(w, &card->widgets, list) {
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->endpoints[SND_SOC_DAPM_DIR_IN] = -1;
w->endpoints[SND_SOC_DAPM_DIR_OUT] = -1;
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
}
ASoC: dapm: Remove path 'walked' flag The 'walked' flag was used to avoid walking paths that have already been walked. But since we started caching the number of inputs and outputs of a path we never actually get into a situation where we try to walk a path that has the 'walked' flag set. There are two cases in which we can end up walking a path multiple times within a single run of is_connected_output_ep() or is_connected_input_ep(). 1) If a path splits up and rejoins later: .--> C ---v A -> B E --> F '--> D ---^ When walking from A to F we'll end up at E twice, once via C and once via D. But since we do a depth first search we'll fully discover the path and initialize the number of outputs/inputs of the widget the first time we get there. The second time we get there we'll use the cached value and not bother to check any of the paths again. So we'll never see a path where 'walked' is set in this case. 2) If there is a circle: A --> B <-- C <-.--> F '--> D ---' When walking from A to F we'll end up twice at B. But since there is a circle the 'walking' flag will still be set on B once we get there the second time. This means we won't look at any of it's outgoing paths. So in this case we won't ever see a path where 'walked' is set either. So it is safe to remove the flag. This on one hand means we remove some always true checks from one of the hottest paths of the DAPM algorithm and on the other hand means we do not have to do the tedious clearing of the flag after checking the number inputs or outputs of a widget. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-21 01:36:39 +08:00
if (stream == SNDRV_PCM_STREAM_PLAYBACK)
paths = is_connected_output_ep(dai->playback_widget, &widgets,
custom_stop_condition);
ASoC: dapm: Remove path 'walked' flag The 'walked' flag was used to avoid walking paths that have already been walked. But since we started caching the number of inputs and outputs of a path we never actually get into a situation where we try to walk a path that has the 'walked' flag set. There are two cases in which we can end up walking a path multiple times within a single run of is_connected_output_ep() or is_connected_input_ep(). 1) If a path splits up and rejoins later: .--> C ---v A -> B E --> F '--> D ---^ When walking from A to F we'll end up at E twice, once via C and once via D. But since we do a depth first search we'll fully discover the path and initialize the number of outputs/inputs of the widget the first time we get there. The second time we get there we'll use the cached value and not bother to check any of the paths again. So we'll never see a path where 'walked' is set in this case. 2) If there is a circle: A --> B <-- C <-.--> F '--> D ---' When walking from A to F we'll end up twice at B. But since there is a circle the 'walking' flag will still be set on B once we get there the second time. This means we won't look at any of it's outgoing paths. So in this case we won't ever see a path where 'walked' is set either. So it is safe to remove the flag. This on one hand means we remove some always true checks from one of the hottest paths of the DAPM algorithm and on the other hand means we do not have to do the tedious clearing of the flag after checking the number inputs or outputs of a widget. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-21 01:36:39 +08:00
else
paths = is_connected_input_ep(dai->capture_widget, &widgets,
custom_stop_condition);
/* Drop starting point */
list_del(widgets.next);
ret = dapm_widget_list_create(list, &widgets);
if (ret)
paths = ret;
trace_snd_soc_dapm_connected(paths, stream);
mutex_unlock(&card->dapm_mutex);
return paths;
}
/*
* Handler for regulator supply widget.
*/
int dapm_regulator_event(struct snd_soc_dapm_widget *w,
struct snd_kcontrol *kcontrol, int event)
{
int ret;
soc_dapm_async_complete(w->dapm);
if (SND_SOC_DAPM_EVENT_ON(event)) {
if (w->on_val & SND_SOC_DAPM_REGULATOR_BYPASS) {
ret = regulator_allow_bypass(w->regulator, false);
if (ret != 0)
dev_warn(w->dapm->dev,
"ASoC: Failed to unbypass %s: %d\n",
w->name, ret);
}
return regulator_enable(w->regulator);
} else {
if (w->on_val & SND_SOC_DAPM_REGULATOR_BYPASS) {
ret = regulator_allow_bypass(w->regulator, true);
if (ret != 0)
dev_warn(w->dapm->dev,
"ASoC: Failed to bypass %s: %d\n",
w->name, ret);
}
return regulator_disable_deferred(w->regulator, w->shift);
}
}
EXPORT_SYMBOL_GPL(dapm_regulator_event);
/*
* Handler for pinctrl widget.
*/
int dapm_pinctrl_event(struct snd_soc_dapm_widget *w,
struct snd_kcontrol *kcontrol, int event)
{
struct snd_soc_dapm_pinctrl_priv *priv = w->priv;
struct pinctrl *p = w->pinctrl;
struct pinctrl_state *s;
if (!p || !priv)
return -EIO;
if (SND_SOC_DAPM_EVENT_ON(event))
s = pinctrl_lookup_state(p, priv->active_state);
else
s = pinctrl_lookup_state(p, priv->sleep_state);
if (IS_ERR(s))
return PTR_ERR(s);
return pinctrl_select_state(p, s);
}
EXPORT_SYMBOL_GPL(dapm_pinctrl_event);
/*
* Handler for clock supply widget.
*/
int dapm_clock_event(struct snd_soc_dapm_widget *w,
struct snd_kcontrol *kcontrol, int event)
{
if (!w->clk)
return -EIO;
soc_dapm_async_complete(w->dapm);
if (SND_SOC_DAPM_EVENT_ON(event)) {
return clk_prepare_enable(w->clk);
} else {
clk_disable_unprepare(w->clk);
return 0;
}
return 0;
}
EXPORT_SYMBOL_GPL(dapm_clock_event);
static int dapm_widget_power_check(struct snd_soc_dapm_widget *w)
{
if (w->power_checked)
return w->new_power;
if (w->force)
w->new_power = 1;
else
w->new_power = w->power_check(w);
w->power_checked = true;
return w->new_power;
}
/* Generic check to see if a widget should be powered. */
static int dapm_generic_check_power(struct snd_soc_dapm_widget *w)
{
int in, out;
DAPM_UPDATE_STAT(w, power_checks);
in = is_connected_input_ep(w, NULL, NULL);
out = is_connected_output_ep(w, NULL, NULL);
return out != 0 && in != 0;
}
/* Check to see if a power supply is needed */
static int dapm_supply_check_power(struct snd_soc_dapm_widget *w)
{
struct snd_soc_dapm_path *path;
DAPM_UPDATE_STAT(w, power_checks);
/* Check if one of our outputs is connected */
snd_soc_dapm_widget_for_each_sink_path(w, path) {
DAPM_UPDATE_STAT(w, neighbour_checks);
if (path->weak)
continue;
if (path->connected &&
!path->connected(path->source, path->sink))
continue;
if (dapm_widget_power_check(path->sink))
return 1;
}
return 0;
}
static int dapm_always_on_check_power(struct snd_soc_dapm_widget *w)
{
return w->connected;
}
static int dapm_seq_compare(struct snd_soc_dapm_widget *a,
struct snd_soc_dapm_widget *b,
bool power_up)
{
int *sort;
BUILD_BUG_ON(ARRAY_SIZE(dapm_up_seq) != SND_SOC_DAPM_TYPE_COUNT);
BUILD_BUG_ON(ARRAY_SIZE(dapm_down_seq) != SND_SOC_DAPM_TYPE_COUNT);
if (power_up)
sort = dapm_up_seq;
else
sort = dapm_down_seq;
WARN_ONCE(sort[a->id] == 0, "offset a->id %d not initialized\n", a->id);
WARN_ONCE(sort[b->id] == 0, "offset b->id %d not initialized\n", b->id);
if (sort[a->id] != sort[b->id])
return sort[a->id] - sort[b->id];
if (a->subseq != b->subseq) {
if (power_up)
return a->subseq - b->subseq;
else
return b->subseq - a->subseq;
}
if (a->reg != b->reg)
return a->reg - b->reg;
if (a->dapm != b->dapm)
return (unsigned long)a->dapm - (unsigned long)b->dapm;
return 0;
}
/* Insert a widget in order into a DAPM power sequence. */
static void dapm_seq_insert(struct snd_soc_dapm_widget *new_widget,
struct list_head *list,
bool power_up)
{
struct snd_soc_dapm_widget *w;
list_for_each_entry(w, list, power_list)
if (dapm_seq_compare(new_widget, w, power_up) < 0) {
list_add_tail(&new_widget->power_list, &w->power_list);
return;
}
list_add_tail(&new_widget->power_list, list);
}
static void dapm_seq_check_event(struct snd_soc_card *card,
struct snd_soc_dapm_widget *w, int event)
{
const char *ev_name;
int power, ret;
switch (event) {
case SND_SOC_DAPM_PRE_PMU:
ev_name = "PRE_PMU";
power = 1;
break;
case SND_SOC_DAPM_POST_PMU:
ev_name = "POST_PMU";
power = 1;
break;
case SND_SOC_DAPM_PRE_PMD:
ev_name = "PRE_PMD";
power = 0;
break;
case SND_SOC_DAPM_POST_PMD:
ev_name = "POST_PMD";
power = 0;
break;
case SND_SOC_DAPM_WILL_PMU:
ev_name = "WILL_PMU";
power = 1;
break;
case SND_SOC_DAPM_WILL_PMD:
ev_name = "WILL_PMD";
power = 0;
break;
default:
WARN(1, "Unknown event %d\n", event);
return;
}
if (w->new_power != power)
return;
if (w->event && (w->event_flags & event)) {
pop_dbg(w->dapm->dev, card->pop_time, "pop test : %s %s\n",
w->name, ev_name);
soc_dapm_async_complete(w->dapm);
trace_snd_soc_dapm_widget_event_start(w, event);
ret = w->event(w, NULL, event);
trace_snd_soc_dapm_widget_event_done(w, event);
if (ret < 0)
dev_err(w->dapm->dev, "ASoC: %s: %s event failed: %d\n",
ev_name, w->name, ret);
}
}
/* Apply the coalesced changes from a DAPM sequence */
static void dapm_seq_run_coalesced(struct snd_soc_card *card,
struct list_head *pending)
{
struct snd_soc_dapm_context *dapm;
struct snd_soc_dapm_widget *w;
int reg;
unsigned int value = 0;
unsigned int mask = 0;
w = list_first_entry(pending, struct snd_soc_dapm_widget, power_list);
reg = w->reg;
dapm = w->dapm;
list_for_each_entry(w, pending, power_list) {
WARN_ON(reg != w->reg || dapm != w->dapm);
w->power = w->new_power;
mask |= w->mask << w->shift;
if (w->power)
value |= w->on_val << w->shift;
else
value |= w->off_val << w->shift;
pop_dbg(dapm->dev, card->pop_time,
"pop test : Queue %s: reg=0x%x, 0x%x/0x%x\n",
w->name, reg, value, mask);
/* Check for events */
dapm_seq_check_event(card, w, SND_SOC_DAPM_PRE_PMU);
dapm_seq_check_event(card, w, SND_SOC_DAPM_PRE_PMD);
}
if (reg >= 0) {
/* Any widget will do, they should all be updating the
* same register.
*/
pop_dbg(dapm->dev, card->pop_time,
"pop test : Applying 0x%x/0x%x to %x in %dms\n",
value, mask, reg, card->pop_time);
pop_wait(card->pop_time);
soc_dapm_update_bits(dapm, reg, mask, value);
}
list_for_each_entry(w, pending, power_list) {
dapm_seq_check_event(card, w, SND_SOC_DAPM_POST_PMU);
dapm_seq_check_event(card, w, SND_SOC_DAPM_POST_PMD);
}
}
/* Apply a DAPM power sequence.
*
* We walk over a pre-sorted list of widgets to apply power to. In
* order to minimise the number of writes to the device required
* multiple widgets will be updated in a single write where possible.
* Currently anything that requires more than a single write is not
* handled.
*/
static void dapm_seq_run(struct snd_soc_card *card,
struct list_head *list, int event, bool power_up)
{
struct snd_soc_dapm_widget *w, *n;
struct snd_soc_dapm_context *d;
LIST_HEAD(pending);
int cur_sort = -1;
int cur_subseq = -1;
int cur_reg = SND_SOC_NOPM;
struct snd_soc_dapm_context *cur_dapm = NULL;
int ret, i;
int *sort;
if (power_up)
sort = dapm_up_seq;
else
sort = dapm_down_seq;
list_for_each_entry_safe(w, n, list, power_list) {
ret = 0;
/* Do we need to apply any queued changes? */
if (sort[w->id] != cur_sort || w->reg != cur_reg ||
w->dapm != cur_dapm || w->subseq != cur_subseq) {
if (!list_empty(&pending))
dapm_seq_run_coalesced(card, &pending);
if (cur_dapm && cur_dapm->seq_notifier) {
for (i = 0; i < ARRAY_SIZE(dapm_up_seq); i++)
if (sort[i] == cur_sort)
cur_dapm->seq_notifier(cur_dapm,
i,
cur_subseq);
}
if (cur_dapm && w->dapm != cur_dapm)
soc_dapm_async_complete(cur_dapm);
INIT_LIST_HEAD(&pending);
cur_sort = -1;
cur_subseq = INT_MIN;
cur_reg = SND_SOC_NOPM;
cur_dapm = NULL;
}
switch (w->id) {
case snd_soc_dapm_pre:
if (!w->event)
list_for_each_entry_safe_continue(w, n, list,
power_list);
if (event == SND_SOC_DAPM_STREAM_START)
ret = w->event(w,
NULL, SND_SOC_DAPM_PRE_PMU);
else if (event == SND_SOC_DAPM_STREAM_STOP)
ret = w->event(w,
NULL, SND_SOC_DAPM_PRE_PMD);
break;
case snd_soc_dapm_post:
if (!w->event)
list_for_each_entry_safe_continue(w, n, list,
power_list);
if (event == SND_SOC_DAPM_STREAM_START)
ret = w->event(w,
NULL, SND_SOC_DAPM_POST_PMU);
else if (event == SND_SOC_DAPM_STREAM_STOP)
ret = w->event(w,
NULL, SND_SOC_DAPM_POST_PMD);
break;
default:
/* Queue it up for application */
cur_sort = sort[w->id];
cur_subseq = w->subseq;
cur_reg = w->reg;
cur_dapm = w->dapm;
list_move(&w->power_list, &pending);
break;
}
if (ret < 0)
dev_err(w->dapm->dev,
"ASoC: Failed to apply widget power: %d\n", ret);
}
if (!list_empty(&pending))
dapm_seq_run_coalesced(card, &pending);
if (cur_dapm && cur_dapm->seq_notifier) {
for (i = 0; i < ARRAY_SIZE(dapm_up_seq); i++)
if (sort[i] == cur_sort)
cur_dapm->seq_notifier(cur_dapm,
i, cur_subseq);
}
list_for_each_entry(d, &card->dapm_list, list) {
soc_dapm_async_complete(d);
}
}
static void dapm_widget_update(struct snd_soc_card *card)
{
struct snd_soc_dapm_update *update = card->update;
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
struct snd_soc_dapm_widget_list *wlist;
struct snd_soc_dapm_widget *w = NULL;
unsigned int wi;
int ret;
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
if (!update || !dapm_kcontrol_is_powered(update->kcontrol))
return;
wlist = dapm_kcontrol_get_wlist(update->kcontrol);
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
for (wi = 0; wi < wlist->num_widgets; wi++) {
w = wlist->widgets[wi];
if (w->event && (w->event_flags & SND_SOC_DAPM_PRE_REG)) {
ret = w->event(w, update->kcontrol, SND_SOC_DAPM_PRE_REG);
if (ret != 0)
dev_err(w->dapm->dev, "ASoC: %s DAPM pre-event failed: %d\n",
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
w->name, ret);
}
}
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
if (!w)
return;
ret = soc_dapm_update_bits(w->dapm, update->reg, update->mask,
update->val);
if (ret < 0)
dev_err(w->dapm->dev, "ASoC: %s DAPM update failed: %d\n",
w->name, ret);
if (update->has_second_set) {
ret = soc_dapm_update_bits(w->dapm, update->reg2,
update->mask2, update->val2);
if (ret < 0)
dev_err(w->dapm->dev,
"ASoC: %s DAPM update failed: %d\n",
w->name, ret);
}
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
for (wi = 0; wi < wlist->num_widgets; wi++) {
w = wlist->widgets[wi];
if (w->event && (w->event_flags & SND_SOC_DAPM_POST_REG)) {
ret = w->event(w, update->kcontrol, SND_SOC_DAPM_POST_REG);
if (ret != 0)
dev_err(w->dapm->dev, "ASoC: %s DAPM post-event failed: %d\n",
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
w->name, ret);
}
}
}
/* Async callback run prior to DAPM sequences - brings to _PREPARE if
* they're changing state.
*/
static void dapm_pre_sequence_async(void *data, async_cookie_t cookie)
{
struct snd_soc_dapm_context *d = data;
int ret;
/* If we're off and we're not supposed to go into STANDBY */
if (d->bias_level == SND_SOC_BIAS_OFF &&
d->target_bias_level != SND_SOC_BIAS_OFF) {
if (d->dev)
pm_runtime_get_sync(d->dev);
ret = snd_soc_dapm_set_bias_level(d, SND_SOC_BIAS_STANDBY);
if (ret != 0)
dev_err(d->dev,
"ASoC: Failed to turn on bias: %d\n", ret);
}
/* Prepare for a transition to ON or away from ON */
if ((d->target_bias_level == SND_SOC_BIAS_ON &&
d->bias_level != SND_SOC_BIAS_ON) ||
(d->target_bias_level != SND_SOC_BIAS_ON &&
d->bias_level == SND_SOC_BIAS_ON)) {
ret = snd_soc_dapm_set_bias_level(d, SND_SOC_BIAS_PREPARE);
if (ret != 0)
dev_err(d->dev,
"ASoC: Failed to prepare bias: %d\n", ret);
}
}
/* Async callback run prior to DAPM sequences - brings to their final
* state.
*/
static void dapm_post_sequence_async(void *data, async_cookie_t cookie)
{
struct snd_soc_dapm_context *d = data;
int ret;
/* If we just powered the last thing off drop to standby bias */
if (d->bias_level == SND_SOC_BIAS_PREPARE &&
(d->target_bias_level == SND_SOC_BIAS_STANDBY ||
d->target_bias_level == SND_SOC_BIAS_OFF)) {
ret = snd_soc_dapm_set_bias_level(d, SND_SOC_BIAS_STANDBY);
if (ret != 0)
dev_err(d->dev, "ASoC: Failed to apply standby bias: %d\n",
ret);
}
/* If we're in standby and can support bias off then do that */
if (d->bias_level == SND_SOC_BIAS_STANDBY &&
d->target_bias_level == SND_SOC_BIAS_OFF) {
ret = snd_soc_dapm_set_bias_level(d, SND_SOC_BIAS_OFF);
if (ret != 0)
dev_err(d->dev, "ASoC: Failed to turn off bias: %d\n",
ret);
if (d->dev)
pm_runtime_put(d->dev);
}
/* If we just powered up then move to active bias */
if (d->bias_level == SND_SOC_BIAS_PREPARE &&
d->target_bias_level == SND_SOC_BIAS_ON) {
ret = snd_soc_dapm_set_bias_level(d, SND_SOC_BIAS_ON);
if (ret != 0)
dev_err(d->dev, "ASoC: Failed to apply active bias: %d\n",
ret);
}
}
static void dapm_widget_set_peer_power(struct snd_soc_dapm_widget *peer,
bool power, bool connect)
{
/* If a connection is being made or broken then that update
* will have marked the peer dirty, otherwise the widgets are
* not connected and this update has no impact. */
if (!connect)
return;
/* If the peer is already in the state we're moving to then we
* won't have an impact on it. */
if (power != peer->power)
dapm_mark_dirty(peer, "peer state change");
}
static void dapm_widget_set_power(struct snd_soc_dapm_widget *w, bool power,
struct list_head *up_list,
struct list_head *down_list)
{
struct snd_soc_dapm_path *path;
if (w->power == power)
return;
trace_snd_soc_dapm_widget_power(w, power);
/* If we changed our power state perhaps our neigbours changed
* also.
*/
snd_soc_dapm_widget_for_each_source_path(w, path)
dapm_widget_set_peer_power(path->source, power, path->connect);
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
/* Supplies can't affect their outputs, only their inputs */
if (!w->is_supply) {
snd_soc_dapm_widget_for_each_sink_path(w, path)
dapm_widget_set_peer_power(path->sink, power,
path->connect);
}
if (power)
dapm_seq_insert(w, up_list, true);
else
dapm_seq_insert(w, down_list, false);
}
static void dapm_power_one_widget(struct snd_soc_dapm_widget *w,
struct list_head *up_list,
struct list_head *down_list)
{
int power;
switch (w->id) {
case snd_soc_dapm_pre:
dapm_seq_insert(w, down_list, false);
break;
case snd_soc_dapm_post:
dapm_seq_insert(w, up_list, true);
break;
default:
power = dapm_widget_power_check(w);
dapm_widget_set_power(w, power, up_list, down_list);
break;
}
}
static bool dapm_idle_bias_off(struct snd_soc_dapm_context *dapm)
{
if (dapm->idle_bias_off)
return true;
switch (snd_power_get_state(dapm->card->snd_card)) {
case SNDRV_CTL_POWER_D3hot:
case SNDRV_CTL_POWER_D3cold:
return dapm->suspend_bias_off;
default:
break;
}
return false;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/*
* Scan each dapm widget for complete audio path.
* A complete path is a route that has valid endpoints i.e.:-
*
* o DAC to output pin.
* o Input pin to ADC.
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
* o Input pin to Output pin (bypass, sidetone)
* o DAC to ADC (loopback).
*/
static int dapm_power_widgets(struct snd_soc_card *card, int event)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_widget *w;
struct snd_soc_dapm_context *d;
LIST_HEAD(up_list);
LIST_HEAD(down_list);
ASYNC_DOMAIN_EXCLUSIVE(async_domain);
enum snd_soc_bias_level bias;
lockdep_assert_held(&card->dapm_mutex);
trace_snd_soc_dapm_start(card);
list_for_each_entry(d, &card->dapm_list, list) {
if (dapm_idle_bias_off(d))
d->target_bias_level = SND_SOC_BIAS_OFF;
else
d->target_bias_level = SND_SOC_BIAS_STANDBY;
}
dapm_reset(card);
/* Check which widgets we need to power and store them in
* lists indicating if they should be powered up or down. We
* only check widgets that have been flagged as dirty but note
* that new widgets may be added to the dirty list while we
* iterate.
*/
list_for_each_entry(w, &card->dapm_dirty, dirty) {
dapm_power_one_widget(w, &up_list, &down_list);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
list_for_each_entry(w, &card->widgets, list) {
switch (w->id) {
case snd_soc_dapm_pre:
case snd_soc_dapm_post:
/* These widgets always need to be powered */
break;
default:
list_del_init(&w->dirty);
break;
}
if (w->new_power) {
d = w->dapm;
/* Supplies and micbiases only bring the
* context up to STANDBY as unless something
* else is active and passing audio they
* generally don't require full power. Signal
* generators are virtual pins and have no
* power impact themselves.
*/
switch (w->id) {
case snd_soc_dapm_siggen:
case snd_soc_dapm_vmid:
break;
case snd_soc_dapm_supply:
case snd_soc_dapm_regulator_supply:
case snd_soc_dapm_pinctrl:
case snd_soc_dapm_clock_supply:
case snd_soc_dapm_micbias:
if (d->target_bias_level < SND_SOC_BIAS_STANDBY)
d->target_bias_level = SND_SOC_BIAS_STANDBY;
break;
default:
d->target_bias_level = SND_SOC_BIAS_ON;
break;
}
}
}
/* Force all contexts in the card to the same bias state if
* they're not ground referenced.
*/
bias = SND_SOC_BIAS_OFF;
list_for_each_entry(d, &card->dapm_list, list)
if (d->target_bias_level > bias)
bias = d->target_bias_level;
list_for_each_entry(d, &card->dapm_list, list)
if (!dapm_idle_bias_off(d))
d->target_bias_level = bias;
trace_snd_soc_dapm_walk_done(card);
/* Run card bias changes at first */
dapm_pre_sequence_async(&card->dapm, 0);
/* Run other bias changes in parallel */
list_for_each_entry(d, &card->dapm_list, list) {
ASoC: core: Don't schedule DAPM work if already in target state When dapm_power_widgets() is called, the dapm_pre_sequence_async() and dapm_post_sequence_async() functions are scheduled for all DAPM contexts (apart from the card DAPM context) regardless of whether the DAPM context is already in the desired state. The overhead of this is not insignificant and the more DAPM contexts there are the more overhead there is. For example, on the Tegra124 Jetson TK1, when profiling the time taken to execute the dapm_power_widgets() the following times were observed. Times for function dapm_power_widgets() are (us): Min 23, Ave 190, Max 434, Count 39 Here 'Count' is the number of times that dapm_power_widgets() has been called. Please note that the above time were measured using ktime_get() to log the time on entry and exit from dapm_power_widgets(). So it should be noted that these times may not be purely the time take to execute this function if it is preempted. However, after applying this patch and measuring the time taken to execute dapm_power_widgets() again a significant improvement is seen as shown below. Times for function dapm_power_widgets() are (us): Min 4, Ave 16, Max 82, Count 39 Therefore, optimise the dapm_power_widgets() function by only scheduling the dapm_pre/post_sequence_async() work if the DAPM context is not in the desired state. Signed-off-by: Jon Hunter <jonathanh@nvidia.com> Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com> Signed-off-by: Mark Brown <broonie@kernel.org>
2018-08-17 23:35:43 +08:00
if (d != &card->dapm && d->bias_level != d->target_bias_level)
async_schedule_domain(dapm_pre_sequence_async, d,
&async_domain);
}
async_synchronize_full_domain(&async_domain);
list_for_each_entry(w, &down_list, power_list) {
dapm_seq_check_event(card, w, SND_SOC_DAPM_WILL_PMD);
}
list_for_each_entry(w, &up_list, power_list) {
dapm_seq_check_event(card, w, SND_SOC_DAPM_WILL_PMU);
}
/* Power down widgets first; try to avoid amplifying pops. */
dapm_seq_run(card, &down_list, event, false);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
dapm_widget_update(card);
/* Now power up. */
dapm_seq_run(card, &up_list, event, true);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* Run all the bias changes in parallel */
list_for_each_entry(d, &card->dapm_list, list) {
ASoC: core: Don't schedule DAPM work if already in target state When dapm_power_widgets() is called, the dapm_pre_sequence_async() and dapm_post_sequence_async() functions are scheduled for all DAPM contexts (apart from the card DAPM context) regardless of whether the DAPM context is already in the desired state. The overhead of this is not insignificant and the more DAPM contexts there are the more overhead there is. For example, on the Tegra124 Jetson TK1, when profiling the time taken to execute the dapm_power_widgets() the following times were observed. Times for function dapm_power_widgets() are (us): Min 23, Ave 190, Max 434, Count 39 Here 'Count' is the number of times that dapm_power_widgets() has been called. Please note that the above time were measured using ktime_get() to log the time on entry and exit from dapm_power_widgets(). So it should be noted that these times may not be purely the time take to execute this function if it is preempted. However, after applying this patch and measuring the time taken to execute dapm_power_widgets() again a significant improvement is seen as shown below. Times for function dapm_power_widgets() are (us): Min 4, Ave 16, Max 82, Count 39 Therefore, optimise the dapm_power_widgets() function by only scheduling the dapm_pre/post_sequence_async() work if the DAPM context is not in the desired state. Signed-off-by: Jon Hunter <jonathanh@nvidia.com> Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com> Signed-off-by: Mark Brown <broonie@kernel.org>
2018-08-17 23:35:43 +08:00
if (d != &card->dapm && d->bias_level != d->target_bias_level)
async_schedule_domain(dapm_post_sequence_async, d,
&async_domain);
}
async_synchronize_full_domain(&async_domain);
/* Run card bias changes at last */
dapm_post_sequence_async(&card->dapm, 0);
/* do we need to notify any clients that DAPM event is complete */
list_for_each_entry(d, &card->dapm_list, list) {
if (d->stream_event)
d->stream_event(d, event);
}
pop_dbg(card->dev, card->pop_time,
"DAPM sequencing finished, waiting %dms\n", card->pop_time);
pop_wait(card->pop_time);
trace_snd_soc_dapm_done(card);
return 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
#ifdef CONFIG_DEBUG_FS
static ssize_t dapm_widget_power_read_file(struct file *file,
char __user *user_buf,
size_t count, loff_t *ppos)
{
struct snd_soc_dapm_widget *w = file->private_data;
struct snd_soc_card *card = w->dapm->card;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
enum snd_soc_dapm_direction dir, rdir;
char *buf;
int in, out;
ssize_t ret;
struct snd_soc_dapm_path *p = NULL;
buf = kmalloc(PAGE_SIZE, GFP_KERNEL);
if (!buf)
return -ENOMEM;
mutex_lock(&card->dapm_mutex);
2014-10-25 23:42:00 +08:00
/* Supply widgets are not handled by is_connected_{input,output}_ep() */
if (w->is_supply) {
in = 0;
out = 0;
} else {
in = is_connected_input_ep(w, NULL, NULL);
out = is_connected_output_ep(w, NULL, NULL);
2014-10-25 23:42:00 +08:00
}
ASoC: dapm: change snprintf to scnprintf for possible overflow Change snprintf to scnprintf. There are generally two cases where using snprintf causes problems. 1) Uses of size += snprintf(buf, SIZE - size, fmt, ...) In this case, if snprintf would have written more characters than what the buffer size (SIZE) is, then size will end up larger than SIZE. In later uses of snprintf, SIZE - size will result in a negative number, leading to problems. Note that size might already be too large by using size = snprintf before the code reaches a case of size += snprintf. 2) If size is ultimately used as a length parameter for a copy back to user space, then it will potentially allow for a buffer overflow and information disclosure when size is greater than SIZE. When the size is used to index the buffer directly, we can have memory corruption. This also means when size = snprintf... is used, it may also cause problems since size may become large. Copying to userspace is mitigated by the HARDENED_USERCOPY kernel configuration. The solution to these issues is to use scnprintf which returns the number of characters actually written to the buffer, so the size variable will never exceed SIZE. Signed-off-by: Silvio Cesare <silvio.cesare@gmail.com> Cc: Liam Girdwood <lgirdwood@gmail.com> Cc: Mark Brown <broonie@kernel.org> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Kees Cook <keescook@chromium.org> Cc: Will Deacon <will.deacon@arm.com> Cc: Greg KH <greg@kroah.com> Signed-off-by: Willy Tarreau <w@1wt.eu> Signed-off-by: Mark Brown <broonie@kernel.org>
2019-01-12 23:28:43 +08:00
ret = scnprintf(buf, PAGE_SIZE, "%s: %s%s in %d out %d",
w->name, w->power ? "On" : "Off",
w->force ? " (forced)" : "", in, out);
if (w->reg >= 0)
ASoC: dapm: change snprintf to scnprintf for possible overflow Change snprintf to scnprintf. There are generally two cases where using snprintf causes problems. 1) Uses of size += snprintf(buf, SIZE - size, fmt, ...) In this case, if snprintf would have written more characters than what the buffer size (SIZE) is, then size will end up larger than SIZE. In later uses of snprintf, SIZE - size will result in a negative number, leading to problems. Note that size might already be too large by using size = snprintf before the code reaches a case of size += snprintf. 2) If size is ultimately used as a length parameter for a copy back to user space, then it will potentially allow for a buffer overflow and information disclosure when size is greater than SIZE. When the size is used to index the buffer directly, we can have memory corruption. This also means when size = snprintf... is used, it may also cause problems since size may become large. Copying to userspace is mitigated by the HARDENED_USERCOPY kernel configuration. The solution to these issues is to use scnprintf which returns the number of characters actually written to the buffer, so the size variable will never exceed SIZE. Signed-off-by: Silvio Cesare <silvio.cesare@gmail.com> Cc: Liam Girdwood <lgirdwood@gmail.com> Cc: Mark Brown <broonie@kernel.org> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Kees Cook <keescook@chromium.org> Cc: Will Deacon <will.deacon@arm.com> Cc: Greg KH <greg@kroah.com> Signed-off-by: Willy Tarreau <w@1wt.eu> Signed-off-by: Mark Brown <broonie@kernel.org>
2019-01-12 23:28:43 +08:00
ret += scnprintf(buf + ret, PAGE_SIZE - ret,
" - R%d(0x%x) mask 0x%x",
w->reg, w->reg, w->mask << w->shift);
ASoC: dapm: change snprintf to scnprintf for possible overflow Change snprintf to scnprintf. There are generally two cases where using snprintf causes problems. 1) Uses of size += snprintf(buf, SIZE - size, fmt, ...) In this case, if snprintf would have written more characters than what the buffer size (SIZE) is, then size will end up larger than SIZE. In later uses of snprintf, SIZE - size will result in a negative number, leading to problems. Note that size might already be too large by using size = snprintf before the code reaches a case of size += snprintf. 2) If size is ultimately used as a length parameter for a copy back to user space, then it will potentially allow for a buffer overflow and information disclosure when size is greater than SIZE. When the size is used to index the buffer directly, we can have memory corruption. This also means when size = snprintf... is used, it may also cause problems since size may become large. Copying to userspace is mitigated by the HARDENED_USERCOPY kernel configuration. The solution to these issues is to use scnprintf which returns the number of characters actually written to the buffer, so the size variable will never exceed SIZE. Signed-off-by: Silvio Cesare <silvio.cesare@gmail.com> Cc: Liam Girdwood <lgirdwood@gmail.com> Cc: Mark Brown <broonie@kernel.org> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Kees Cook <keescook@chromium.org> Cc: Will Deacon <will.deacon@arm.com> Cc: Greg KH <greg@kroah.com> Signed-off-by: Willy Tarreau <w@1wt.eu> Signed-off-by: Mark Brown <broonie@kernel.org>
2019-01-12 23:28:43 +08:00
ret += scnprintf(buf + ret, PAGE_SIZE - ret, "\n");
if (w->sname)
ASoC: dapm: change snprintf to scnprintf for possible overflow Change snprintf to scnprintf. There are generally two cases where using snprintf causes problems. 1) Uses of size += snprintf(buf, SIZE - size, fmt, ...) In this case, if snprintf would have written more characters than what the buffer size (SIZE) is, then size will end up larger than SIZE. In later uses of snprintf, SIZE - size will result in a negative number, leading to problems. Note that size might already be too large by using size = snprintf before the code reaches a case of size += snprintf. 2) If size is ultimately used as a length parameter for a copy back to user space, then it will potentially allow for a buffer overflow and information disclosure when size is greater than SIZE. When the size is used to index the buffer directly, we can have memory corruption. This also means when size = snprintf... is used, it may also cause problems since size may become large. Copying to userspace is mitigated by the HARDENED_USERCOPY kernel configuration. The solution to these issues is to use scnprintf which returns the number of characters actually written to the buffer, so the size variable will never exceed SIZE. Signed-off-by: Silvio Cesare <silvio.cesare@gmail.com> Cc: Liam Girdwood <lgirdwood@gmail.com> Cc: Mark Brown <broonie@kernel.org> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Kees Cook <keescook@chromium.org> Cc: Will Deacon <will.deacon@arm.com> Cc: Greg KH <greg@kroah.com> Signed-off-by: Willy Tarreau <w@1wt.eu> Signed-off-by: Mark Brown <broonie@kernel.org>
2019-01-12 23:28:43 +08:00
ret += scnprintf(buf + ret, PAGE_SIZE - ret, " stream %s %s\n",
w->sname,
w->active ? "active" : "inactive");
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
snd_soc_dapm_for_each_direction(dir) {
rdir = SND_SOC_DAPM_DIR_REVERSE(dir);
snd_soc_dapm_widget_for_each_path(w, dir, p) {
if (p->connected && !p->connected(p->source, p->sink))
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
continue;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (!p->connect)
continue;
ASoC: dapm: change snprintf to scnprintf for possible overflow Change snprintf to scnprintf. There are generally two cases where using snprintf causes problems. 1) Uses of size += snprintf(buf, SIZE - size, fmt, ...) In this case, if snprintf would have written more characters than what the buffer size (SIZE) is, then size will end up larger than SIZE. In later uses of snprintf, SIZE - size will result in a negative number, leading to problems. Note that size might already be too large by using size = snprintf before the code reaches a case of size += snprintf. 2) If size is ultimately used as a length parameter for a copy back to user space, then it will potentially allow for a buffer overflow and information disclosure when size is greater than SIZE. When the size is used to index the buffer directly, we can have memory corruption. This also means when size = snprintf... is used, it may also cause problems since size may become large. Copying to userspace is mitigated by the HARDENED_USERCOPY kernel configuration. The solution to these issues is to use scnprintf which returns the number of characters actually written to the buffer, so the size variable will never exceed SIZE. Signed-off-by: Silvio Cesare <silvio.cesare@gmail.com> Cc: Liam Girdwood <lgirdwood@gmail.com> Cc: Mark Brown <broonie@kernel.org> Cc: Dan Carpenter <dan.carpenter@oracle.com> Cc: Kees Cook <keescook@chromium.org> Cc: Will Deacon <will.deacon@arm.com> Cc: Greg KH <greg@kroah.com> Signed-off-by: Willy Tarreau <w@1wt.eu> Signed-off-by: Mark Brown <broonie@kernel.org>
2019-01-12 23:28:43 +08:00
ret += scnprintf(buf + ret, PAGE_SIZE - ret,
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
" %s \"%s\" \"%s\"\n",
(rdir == SND_SOC_DAPM_DIR_IN) ? "in" : "out",
p->name ? p->name : "static",
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
p->node[rdir]->name);
}
}
mutex_unlock(&card->dapm_mutex);
ret = simple_read_from_buffer(user_buf, count, ppos, buf, ret);
kfree(buf);
return ret;
}
static const struct file_operations dapm_widget_power_fops = {
.open = simple_open,
.read = dapm_widget_power_read_file,
llseek: automatically add .llseek fop All file_operations should get a .llseek operation so we can make nonseekable_open the default for future file operations without a .llseek pointer. The three cases that we can automatically detect are no_llseek, seq_lseek and default_llseek. For cases where we can we can automatically prove that the file offset is always ignored, we use noop_llseek, which maintains the current behavior of not returning an error from a seek. New drivers should normally not use noop_llseek but instead use no_llseek and call nonseekable_open at open time. Existing drivers can be converted to do the same when the maintainer knows for certain that no user code relies on calling seek on the device file. The generated code is often incorrectly indented and right now contains comments that clarify for each added line why a specific variant was chosen. In the version that gets submitted upstream, the comments will be gone and I will manually fix the indentation, because there does not seem to be a way to do that using coccinelle. Some amount of new code is currently sitting in linux-next that should get the same modifications, which I will do at the end of the merge window. Many thanks to Julia Lawall for helping me learn to write a semantic patch that does all this. ===== begin semantic patch ===== // This adds an llseek= method to all file operations, // as a preparation for making no_llseek the default. // // The rules are // - use no_llseek explicitly if we do nonseekable_open // - use seq_lseek for sequential files // - use default_llseek if we know we access f_pos // - use noop_llseek if we know we don't access f_pos, // but we still want to allow users to call lseek // @ open1 exists @ identifier nested_open; @@ nested_open(...) { <+... nonseekable_open(...) ...+> } @ open exists@ identifier open_f; identifier i, f; identifier open1.nested_open; @@ int open_f(struct inode *i, struct file *f) { <+... ( nonseekable_open(...) | nested_open(...) ) ...+> } @ read disable optional_qualifier exists @ identifier read_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; expression E; identifier func; @@ ssize_t read_f(struct file *f, char *p, size_t s, loff_t *off) { <+... ( *off = E | *off += E | func(..., off, ...) | E = *off ) ...+> } @ read_no_fpos disable optional_qualifier exists @ identifier read_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; @@ ssize_t read_f(struct file *f, char *p, size_t s, loff_t *off) { ... when != off } @ write @ identifier write_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; expression E; identifier func; @@ ssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off) { <+... ( *off = E | *off += E | func(..., off, ...) | E = *off ) ...+> } @ write_no_fpos @ identifier write_f; identifier f, p, s, off; type ssize_t, size_t, loff_t; @@ ssize_t write_f(struct file *f, const char *p, size_t s, loff_t *off) { ... when != off } @ fops0 @ identifier fops; @@ struct file_operations fops = { ... }; @ has_llseek depends on fops0 @ identifier fops0.fops; identifier llseek_f; @@ struct file_operations fops = { ... .llseek = llseek_f, ... }; @ has_read depends on fops0 @ identifier fops0.fops; identifier read_f; @@ struct file_operations fops = { ... .read = read_f, ... }; @ has_write depends on fops0 @ identifier fops0.fops; identifier write_f; @@ struct file_operations fops = { ... .write = write_f, ... }; @ has_open depends on fops0 @ identifier fops0.fops; identifier open_f; @@ struct file_operations fops = { ... .open = open_f, ... }; // use no_llseek if we call nonseekable_open //////////////////////////////////////////// @ nonseekable1 depends on !has_llseek && has_open @ identifier fops0.fops; identifier nso ~= "nonseekable_open"; @@ struct file_operations fops = { ... .open = nso, ... +.llseek = no_llseek, /* nonseekable */ }; @ nonseekable2 depends on !has_llseek @ identifier fops0.fops; identifier open.open_f; @@ struct file_operations fops = { ... .open = open_f, ... +.llseek = no_llseek, /* open uses nonseekable */ }; // use seq_lseek for sequential files ///////////////////////////////////// @ seq depends on !has_llseek @ identifier fops0.fops; identifier sr ~= "seq_read"; @@ struct file_operations fops = { ... .read = sr, ... +.llseek = seq_lseek, /* we have seq_read */ }; // use default_llseek if there is a readdir /////////////////////////////////////////// @ fops1 depends on !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier readdir_e; @@ // any other fop is used that changes pos struct file_operations fops = { ... .readdir = readdir_e, ... +.llseek = default_llseek, /* readdir is present */ }; // use default_llseek if at least one of read/write touches f_pos ///////////////////////////////////////////////////////////////// @ fops2 depends on !fops1 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read.read_f; @@ // read fops use offset struct file_operations fops = { ... .read = read_f, ... +.llseek = default_llseek, /* read accesses f_pos */ }; @ fops3 depends on !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier write.write_f; @@ // write fops use offset struct file_operations fops = { ... .write = write_f, ... + .llseek = default_llseek, /* write accesses f_pos */ }; // Use noop_llseek if neither read nor write accesses f_pos /////////////////////////////////////////////////////////// @ fops4 depends on !fops1 && !fops2 && !fops3 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read_no_fpos.read_f; identifier write_no_fpos.write_f; @@ // write fops use offset struct file_operations fops = { ... .write = write_f, .read = read_f, ... +.llseek = noop_llseek, /* read and write both use no f_pos */ }; @ depends on has_write && !has_read && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier write_no_fpos.write_f; @@ struct file_operations fops = { ... .write = write_f, ... +.llseek = noop_llseek, /* write uses no f_pos */ }; @ depends on has_read && !has_write && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; identifier read_no_fpos.read_f; @@ struct file_operations fops = { ... .read = read_f, ... +.llseek = noop_llseek, /* read uses no f_pos */ }; @ depends on !has_read && !has_write && !fops1 && !fops2 && !has_llseek && !nonseekable1 && !nonseekable2 && !seq @ identifier fops0.fops; @@ struct file_operations fops = { ... +.llseek = noop_llseek, /* no read or write fn */ }; ===== End semantic patch ===== Signed-off-by: Arnd Bergmann <arnd@arndb.de> Cc: Julia Lawall <julia@diku.dk> Cc: Christoph Hellwig <hch@infradead.org>
2010-08-16 00:52:59 +08:00
.llseek = default_llseek,
};
static ssize_t dapm_bias_read_file(struct file *file, char __user *user_buf,
size_t count, loff_t *ppos)
{
struct snd_soc_dapm_context *dapm = file->private_data;
char *level;
switch (dapm->bias_level) {
case SND_SOC_BIAS_ON:
level = "On\n";
break;
case SND_SOC_BIAS_PREPARE:
level = "Prepare\n";
break;
case SND_SOC_BIAS_STANDBY:
level = "Standby\n";
break;
case SND_SOC_BIAS_OFF:
level = "Off\n";
break;
default:
WARN(1, "Unknown bias_level %d\n", dapm->bias_level);
level = "Unknown\n";
break;
}
return simple_read_from_buffer(user_buf, count, ppos, level,
strlen(level));
}
static const struct file_operations dapm_bias_fops = {
.open = simple_open,
.read = dapm_bias_read_file,
.llseek = default_llseek,
};
void snd_soc_dapm_debugfs_init(struct snd_soc_dapm_context *dapm,
struct dentry *parent)
{
struct dentry *d;
if (!parent)
return;
dapm->debugfs_dapm = debugfs_create_dir("dapm", parent);
if (!dapm->debugfs_dapm) {
dev_warn(dapm->dev,
"ASoC: Failed to create DAPM debugfs directory\n");
return;
}
d = debugfs_create_file("bias_level", 0444,
dapm->debugfs_dapm, dapm,
&dapm_bias_fops);
if (!d)
dev_warn(dapm->dev,
"ASoC: Failed to create bias level debugfs file\n");
}
static void dapm_debugfs_add_widget(struct snd_soc_dapm_widget *w)
{
struct snd_soc_dapm_context *dapm = w->dapm;
struct dentry *d;
if (!dapm->debugfs_dapm || !w->name)
return;
d = debugfs_create_file(w->name, 0444,
dapm->debugfs_dapm, w,
&dapm_widget_power_fops);
if (!d)
dev_warn(w->dapm->dev,
"ASoC: Failed to create %s debugfs file\n",
w->name);
}
static void dapm_debugfs_cleanup(struct snd_soc_dapm_context *dapm)
{
debugfs_remove_recursive(dapm->debugfs_dapm);
}
#else
void snd_soc_dapm_debugfs_init(struct snd_soc_dapm_context *dapm,
struct dentry *parent)
{
}
static inline void dapm_debugfs_add_widget(struct snd_soc_dapm_widget *w)
{
}
static inline void dapm_debugfs_cleanup(struct snd_soc_dapm_context *dapm)
{
}
#endif
/*
* soc_dapm_connect_path() - Connects or disconnects a path
* @path: The path to update
* @connect: The new connect state of the path. True if the path is connected,
* false if it is disconnected.
* @reason: The reason why the path changed (for debugging only)
*/
static void soc_dapm_connect_path(struct snd_soc_dapm_path *path,
bool connect, const char *reason)
{
if (path->connect == connect)
return;
path->connect = connect;
dapm_mark_dirty(path->source, reason);
dapm_mark_dirty(path->sink, reason);
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
dapm_path_invalidate(path);
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* test and update the power status of a mux widget */
static int soc_dapm_mux_update_power(struct snd_soc_card *card,
struct snd_kcontrol *kcontrol, int mux, struct soc_enum *e)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_path *path;
int found = 0;
bool connect;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
lockdep_assert_held(&card->dapm_mutex);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* find dapm widget path assoc with kcontrol */
dapm_kcontrol_for_each_path(path, kcontrol) {
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
found = 1;
/* we now need to match the string in the enum to the path */
if (!(strcmp(path->name, e->texts[mux])))
connect = true;
else
connect = false;
soc_dapm_connect_path(path, connect, "mux update");
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
if (found)
dapm_power_widgets(card, SND_SOC_DAPM_STREAM_NOP);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return found;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
int snd_soc_dapm_mux_update_power(struct snd_soc_dapm_context *dapm,
struct snd_kcontrol *kcontrol, int mux, struct soc_enum *e,
struct snd_soc_dapm_update *update)
{
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
struct snd_soc_card *card = dapm->card;
int ret;
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
card->update = update;
ret = soc_dapm_mux_update_power(card, kcontrol, mux, e);
card->update = NULL;
mutex_unlock(&card->dapm_mutex);
if (ret > 0)
soc_dpcm_runtime_update(card);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_mux_update_power);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* test and update the power status of a mixer or switch widget */
static int soc_dapm_mixer_update_power(struct snd_soc_card *card,
struct snd_kcontrol *kcontrol,
int connect, int rconnect)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_path *path;
int found = 0;
lockdep_assert_held(&card->dapm_mutex);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* find dapm widget path assoc with kcontrol */
dapm_kcontrol_for_each_path(path, kcontrol) {
/*
* Ideally this function should support any number of
* paths and channels. But since kcontrols only come
* in mono and stereo variants, we are limited to 2
* channels.
*
* The following code assumes for stereo controls the
* first path (when 'found == 0') is the left channel,
* and all remaining paths (when 'found == 1') are the
* right channel.
*
* A stereo control is signified by a valid 'rconnect'
* value, either 0 for unconnected, or >= 0 for connected.
* This is chosen instead of using snd_soc_volsw_is_stereo,
* so that the behavior of snd_soc_dapm_mixer_update_power
* doesn't change even when the kcontrol passed in is
* stereo.
*
* It passes 'connect' as the path connect status for
* the left channel, and 'rconnect' for the right
* channel.
*/
if (found && rconnect >= 0)
soc_dapm_connect_path(path, rconnect, "mixer update");
else
soc_dapm_connect_path(path, connect, "mixer update");
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
found = 1;
}
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
if (found)
dapm_power_widgets(card, SND_SOC_DAPM_STREAM_NOP);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return found;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
int snd_soc_dapm_mixer_update_power(struct snd_soc_dapm_context *dapm,
struct snd_kcontrol *kcontrol, int connect,
struct snd_soc_dapm_update *update)
{
ASoC: dapm: Run widget updates for shared controls at the same time Currently when updating a control that is shared between multiple widgets the whole power-up/power-down sequence is being run once for each widget. The control register is updated during the first run, which means the CODEC internal routing is also updated for all widgets during this first run. The input and output paths for each widgets are only updated though during the respective run for that widget. This leads to a slight inconsistency between the CODEC's internal state and ASoC's state, which causes non optimal behavior in regard to click and pop avoidance. E.g. consider the following setup where two MUXs share the same control. +------+ A1 ------| | | MUX1 |----- C1 B1 ------| | +------+ | control ---+ | +------+ A2 ------| | | MUX2 |----- C2 B2 ------| | +------+ If the control is updated to switch the MUXs from input A to input B with the current code the power-up/power-down sequence will look like this: Run soc_dapm_mux_update_power for MUX1 Power-down A1 Update MUXing Power-up B1 Run soc_dapm_mux_update_power for MUX2 Power-down A2 (Update MUXing) Power-up B2 Note that the second 'Update Muxing' is a no-op, since the register was already updated. While the preferred order for avoiding pops and clicks should be: Run soc_dapm_mux_update_power for control Power-down A1 Power-down A2 Update MUXing Power-up B1 Power-up B2 This patch changes the behavior to the later by running the updates for all widgets that the control is attached to at the same time. The new code is also a bit simpler since callers of soc_dapm_{mux,muxer}_update_power don't have to loop over each widget anymore and neither do we need to keep track for which of the kcontrol's widgets the current update is. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-07-24 21:27:37 +08:00
struct snd_soc_card *card = dapm->card;
int ret;
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
card->update = update;
ret = soc_dapm_mixer_update_power(card, kcontrol, connect, -1);
card->update = NULL;
mutex_unlock(&card->dapm_mutex);
if (ret > 0)
soc_dpcm_runtime_update(card);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_mixer_update_power);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
static ssize_t dapm_widget_show_component(struct snd_soc_component *cmpnt,
char *buf)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_context *dapm = snd_soc_component_get_dapm(cmpnt);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct snd_soc_dapm_widget *w;
int count = 0;
char *state = "not set";
/* card won't be set for the dummy component, as a spot fix
* we're checking for that case specifically here but in future
* we will ensure that the dummy component looks like others.
*/
if (!cmpnt->card)
return 0;
list_for_each_entry(w, &cmpnt->card->widgets, list) {
if (w->dapm != dapm)
continue;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* only display widgets that burn power */
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
switch (w->id) {
case snd_soc_dapm_hp:
case snd_soc_dapm_mic:
case snd_soc_dapm_spk:
case snd_soc_dapm_line:
case snd_soc_dapm_micbias:
case snd_soc_dapm_dac:
case snd_soc_dapm_adc:
case snd_soc_dapm_pga:
case snd_soc_dapm_out_drv:
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
case snd_soc_dapm_mixer:
case snd_soc_dapm_mixer_named_ctl:
case snd_soc_dapm_supply:
case snd_soc_dapm_regulator_supply:
case snd_soc_dapm_pinctrl:
case snd_soc_dapm_clock_supply:
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
if (w->name)
count += sprintf(buf + count, "%s: %s\n",
w->name, w->power ? "On":"Off");
break;
default:
break;
}
}
switch (snd_soc_dapm_get_bias_level(dapm)) {
case SND_SOC_BIAS_ON:
state = "On";
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
break;
case SND_SOC_BIAS_PREPARE:
state = "Prepare";
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
break;
case SND_SOC_BIAS_STANDBY:
state = "Standby";
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
break;
case SND_SOC_BIAS_OFF:
state = "Off";
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
break;
}
count += sprintf(buf + count, "PM State: %s\n", state);
return count;
}
/* show dapm widget status in sys fs */
static ssize_t dapm_widget_show(struct device *dev,
struct device_attribute *attr, char *buf)
{
struct snd_soc_pcm_runtime *rtd = dev_get_drvdata(dev);
struct snd_soc_dai *codec_dai;
int i, count = 0;
mutex_lock(&rtd->card->dapm_mutex);
for_each_rtd_codec_dai(rtd, i, codec_dai) {
struct snd_soc_component *cmpnt = codec_dai->component;
count += dapm_widget_show_component(cmpnt, buf + count);
}
mutex_unlock(&rtd->card->dapm_mutex);
return count;
}
static DEVICE_ATTR_RO(dapm_widget);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct attribute *soc_dapm_dev_attrs[] = {
&dev_attr_dapm_widget.attr,
NULL
};
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
static void dapm_free_path(struct snd_soc_dapm_path *path)
{
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
list_del(&path->list_node[SND_SOC_DAPM_DIR_IN]);
list_del(&path->list_node[SND_SOC_DAPM_DIR_OUT]);
list_del(&path->list_kcontrol);
list_del(&path->list);
kfree(path);
}
void snd_soc_dapm_free_widget(struct snd_soc_dapm_widget *w)
{
struct snd_soc_dapm_path *p, *next_p;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
enum snd_soc_dapm_direction dir;
list_del(&w->list);
/*
* remove source and sink paths associated to this widget.
* While removing the path, remove reference to it from both
* source and sink widgets so that path is removed only once.
*/
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
snd_soc_dapm_for_each_direction(dir) {
snd_soc_dapm_widget_for_each_path_safe(w, dir, p, next_p)
dapm_free_path(p);
}
kfree(w->kcontrols);
kfree_const(w->name);
kfree_const(w->sname);
kfree(w);
}
void snd_soc_dapm_reset_cache(struct snd_soc_dapm_context *dapm)
{
dapm->path_sink_cache.widget = NULL;
dapm->path_source_cache.widget = NULL;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* free all dapm widgets and resources */
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
static void dapm_free_widgets(struct snd_soc_dapm_context *dapm)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_widget *w, *next_w;
list_for_each_entry_safe(w, next_w, &dapm->card->widgets, list) {
if (w->dapm != dapm)
continue;
snd_soc_dapm_free_widget(w);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
snd_soc_dapm_reset_cache(dapm);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
static struct snd_soc_dapm_widget *dapm_find_widget(
struct snd_soc_dapm_context *dapm, const char *pin,
bool search_other_contexts)
{
struct snd_soc_dapm_widget *w;
struct snd_soc_dapm_widget *fallback = NULL;
list_for_each_entry(w, &dapm->card->widgets, list) {
if (!strcmp(w->name, pin)) {
if (w->dapm == dapm)
return w;
else
fallback = w;
}
}
if (search_other_contexts)
return fallback;
return NULL;
}
static int snd_soc_dapm_set_pin(struct snd_soc_dapm_context *dapm,
const char *pin, int status)
{
struct snd_soc_dapm_widget *w = dapm_find_widget(dapm, pin, true);
dapm_assert_locked(dapm);
if (!w) {
dev_err(dapm->dev, "ASoC: DAPM unknown pin %s\n", pin);
return -EINVAL;
}
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
if (w->connected != status) {
dapm_mark_dirty(w, "pin configuration");
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
dapm_widget_invalidate_input_paths(w);
dapm_widget_invalidate_output_paths(w);
}
w->connected = status;
if (status == 0)
w->force = 0;
return 0;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/**
* snd_soc_dapm_sync_unlocked - scan and power dapm paths
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*
* Walks all dapm audio paths and powers widgets according to their
* stream or path usage.
*
* Requires external locking.
*
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
* Returns 0 for success.
*/
int snd_soc_dapm_sync_unlocked(struct snd_soc_dapm_context *dapm)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
/*
* Suppress early reports (eg, jacks syncing their state) to avoid
* silly DAPM runs during card startup.
*/
if (!dapm->card || !dapm->card->instantiated)
return 0;
return dapm_power_widgets(dapm->card, SND_SOC_DAPM_STREAM_NOP);
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_sync_unlocked);
/**
* snd_soc_dapm_sync - scan and power dapm paths
* @dapm: DAPM context
*
* Walks all dapm audio paths and powers widgets according to their
* stream or path usage.
*
* Returns 0 for success.
*/
int snd_soc_dapm_sync(struct snd_soc_dapm_context *dapm)
{
int ret;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
ret = snd_soc_dapm_sync_unlocked(dapm);
mutex_unlock(&dapm->card->dapm_mutex);
return ret;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_sync);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
static int dapm_update_dai_chan(struct snd_soc_dapm_path *p,
struct snd_soc_dapm_widget *w,
int channels)
{
switch (w->id) {
case snd_soc_dapm_aif_out:
case snd_soc_dapm_aif_in:
break;
default:
return 0;
}
dev_dbg(w->dapm->dev, "%s DAI route %s -> %s\n",
w->channel < channels ? "Connecting" : "Disconnecting",
p->source->name, p->sink->name);
if (w->channel < channels)
soc_dapm_connect_path(p, true, "dai update");
else
soc_dapm_connect_path(p, false, "dai update");
return 0;
}
static int dapm_update_dai_unlocked(struct snd_pcm_substream *substream,
struct snd_pcm_hw_params *params,
struct snd_soc_dai *dai)
{
int dir = substream->stream;
int channels = params_channels(params);
struct snd_soc_dapm_path *p;
struct snd_soc_dapm_widget *w;
int ret;
if (dir == SNDRV_PCM_STREAM_PLAYBACK)
w = dai->playback_widget;
else
w = dai->capture_widget;
if (!w)
return 0;
dev_dbg(dai->dev, "Update DAI routes for %s %s\n", dai->name,
dir == SNDRV_PCM_STREAM_PLAYBACK ? "playback" : "capture");
snd_soc_dapm_widget_for_each_sink_path(w, p) {
ret = dapm_update_dai_chan(p, p->sink, channels);
if (ret < 0)
return ret;
}
snd_soc_dapm_widget_for_each_source_path(w, p) {
ret = dapm_update_dai_chan(p, p->source, channels);
if (ret < 0)
return ret;
}
return 0;
}
int snd_soc_dapm_update_dai(struct snd_pcm_substream *substream,
struct snd_pcm_hw_params *params,
struct snd_soc_dai *dai)
{
struct snd_soc_pcm_runtime *rtd = substream->private_data;
int ret;
mutex_lock_nested(&rtd->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
ret = dapm_update_dai_unlocked(substream, params, dai);
mutex_unlock(&rtd->card->dapm_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_update_dai);
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
/*
* dapm_update_widget_flags() - Re-compute widget sink and source flags
* @w: The widget for which to update the flags
*
* Some widgets have a dynamic category which depends on which neighbors they
* are connected to. This function update the category for these widgets.
*
* This function must be called whenever a path is added or removed to a widget.
*/
static void dapm_update_widget_flags(struct snd_soc_dapm_widget *w)
{
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
enum snd_soc_dapm_direction dir;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
struct snd_soc_dapm_path *p;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
unsigned int ep;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
switch (w->id) {
case snd_soc_dapm_input:
/* On a fully routed card an input is never a source */
if (w->dapm->card->fully_routed)
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
return;
ep = SND_SOC_DAPM_EP_SOURCE;
snd_soc_dapm_widget_for_each_source_path(w, p) {
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
if (p->source->id == snd_soc_dapm_micbias ||
p->source->id == snd_soc_dapm_mic ||
p->source->id == snd_soc_dapm_line ||
p->source->id == snd_soc_dapm_output) {
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
ep = 0;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
break;
}
}
break;
case snd_soc_dapm_output:
/* On a fully routed card a output is never a sink */
if (w->dapm->card->fully_routed)
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
return;
ep = SND_SOC_DAPM_EP_SINK;
snd_soc_dapm_widget_for_each_sink_path(w, p) {
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
if (p->sink->id == snd_soc_dapm_spk ||
p->sink->id == snd_soc_dapm_hp ||
p->sink->id == snd_soc_dapm_line ||
p->sink->id == snd_soc_dapm_input) {
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
ep = 0;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
break;
}
}
break;
case snd_soc_dapm_line:
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
ep = 0;
snd_soc_dapm_for_each_direction(dir) {
if (!list_empty(&w->edges[dir]))
ep |= SND_SOC_DAPM_DIR_TO_EP(dir);
}
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
break;
default:
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
return;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
}
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->is_ep = ep;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
}
static int snd_soc_dapm_check_dynamic_path(struct snd_soc_dapm_context *dapm,
struct snd_soc_dapm_widget *source, struct snd_soc_dapm_widget *sink,
const char *control)
{
bool dynamic_source = false;
bool dynamic_sink = false;
if (!control)
return 0;
switch (source->id) {
case snd_soc_dapm_demux:
dynamic_source = true;
break;
default:
break;
}
switch (sink->id) {
case snd_soc_dapm_mux:
case snd_soc_dapm_switch:
case snd_soc_dapm_mixer:
case snd_soc_dapm_mixer_named_ctl:
dynamic_sink = true;
break;
default:
break;
}
if (dynamic_source && dynamic_sink) {
dev_err(dapm->dev,
"Direct connection between demux and mixer/mux not supported for path %s -> [%s] -> %s\n",
source->name, control, sink->name);
return -EINVAL;
} else if (!dynamic_source && !dynamic_sink) {
dev_err(dapm->dev,
"Control not supported for path %s -> [%s] -> %s\n",
source->name, control, sink->name);
return -EINVAL;
}
return 0;
}
static int snd_soc_dapm_add_path(struct snd_soc_dapm_context *dapm,
struct snd_soc_dapm_widget *wsource, struct snd_soc_dapm_widget *wsink,
const char *control,
int (*connected)(struct snd_soc_dapm_widget *source,
struct snd_soc_dapm_widget *sink))
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
struct snd_soc_dapm_widget *widgets[2];
enum snd_soc_dapm_direction dir;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct snd_soc_dapm_path *path;
int ret;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
if (wsink->is_supply && !wsource->is_supply) {
dev_err(dapm->dev,
"Connecting non-supply widget to supply widget is not supported (%s -> %s)\n",
wsource->name, wsink->name);
return -EINVAL;
}
if (connected && !wsource->is_supply) {
dev_err(dapm->dev,
"connected() callback only supported for supply widgets (%s -> %s)\n",
wsource->name, wsink->name);
return -EINVAL;
}
if (wsource->is_supply && control) {
dev_err(dapm->dev,
"Conditional paths are not supported for supply widgets (%s -> [%s] -> %s)\n",
wsource->name, control, wsink->name);
return -EINVAL;
}
ret = snd_soc_dapm_check_dynamic_path(dapm, wsource, wsink, control);
if (ret)
return ret;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
path = kzalloc(sizeof(struct snd_soc_dapm_path), GFP_KERNEL);
if (!path)
return -ENOMEM;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
path->node[SND_SOC_DAPM_DIR_IN] = wsource;
path->node[SND_SOC_DAPM_DIR_OUT] = wsink;
widgets[SND_SOC_DAPM_DIR_IN] = wsource;
widgets[SND_SOC_DAPM_DIR_OUT] = wsink;
path->connected = connected;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
INIT_LIST_HEAD(&path->list);
INIT_LIST_HEAD(&path->list_kcontrol);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
2014-10-25 23:42:00 +08:00
if (wsource->is_supply || wsink->is_supply)
path->is_supply = 1;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* connect static paths */
if (control == NULL) {
path->connect = 1;
} else {
switch (wsource->id) {
case snd_soc_dapm_demux:
ret = dapm_connect_mux(dapm, path, control, wsource);
if (ret)
goto err;
break;
default:
break;
}
switch (wsink->id) {
case snd_soc_dapm_mux:
ret = dapm_connect_mux(dapm, path, control, wsink);
if (ret != 0)
goto err;
break;
case snd_soc_dapm_switch:
case snd_soc_dapm_mixer:
case snd_soc_dapm_mixer_named_ctl:
ret = dapm_connect_mixer(dapm, path, control);
if (ret != 0)
goto err;
break;
default:
break;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
list_add(&path->list, &dapm->card->paths);
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
snd_soc_dapm_for_each_direction(dir)
list_add(&path->list_node[dir], &widgets[dir]->edges[dir]);
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
snd_soc_dapm_for_each_direction(dir) {
dapm_update_widget_flags(widgets[dir]);
dapm_mark_dirty(widgets[dir], "Route added");
}
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
if (dapm->card->instantiated && path->connect)
dapm_path_invalidate(path);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return 0;
err:
kfree(path);
return ret;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
static int snd_soc_dapm_add_route(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_route *route)
{
struct snd_soc_dapm_widget *wsource = NULL, *wsink = NULL, *w;
struct snd_soc_dapm_widget *wtsource = NULL, *wtsink = NULL;
const char *sink;
const char *source;
char prefixed_sink[80];
char prefixed_source[80];
const char *prefix;
unsigned int sink_ref = 0;
unsigned int source_ref = 0;
int ret;
prefix = soc_dapm_prefix(dapm);
if (prefix) {
snprintf(prefixed_sink, sizeof(prefixed_sink), "%s %s",
prefix, route->sink);
sink = prefixed_sink;
snprintf(prefixed_source, sizeof(prefixed_source), "%s %s",
prefix, route->source);
source = prefixed_source;
} else {
sink = route->sink;
source = route->source;
}
wsource = dapm_wcache_lookup(&dapm->path_source_cache, source);
wsink = dapm_wcache_lookup(&dapm->path_sink_cache, sink);
if (wsink && wsource)
goto skip_search;
/*
* find src and dest widgets over all widgets but favor a widget from
* current DAPM context
*/
list_for_each_entry(w, &dapm->card->widgets, list) {
if (!wsink && !(strcmp(w->name, sink))) {
wtsink = w;
if (w->dapm == dapm) {
wsink = w;
if (wsource)
break;
}
sink_ref++;
if (sink_ref > 1)
dev_warn(dapm->dev,
"ASoC: sink widget %s overwritten\n",
w->name);
continue;
}
if (!wsource && !(strcmp(w->name, source))) {
wtsource = w;
if (w->dapm == dapm) {
wsource = w;
if (wsink)
break;
}
source_ref++;
if (source_ref > 1)
dev_warn(dapm->dev,
"ASoC: source widget %s overwritten\n",
w->name);
}
}
/* use widget from another DAPM context if not found from this */
if (!wsink)
wsink = wtsink;
if (!wsource)
wsource = wtsource;
if (wsource == NULL) {
dev_err(dapm->dev, "ASoC: no source widget found for %s\n",
route->source);
return -ENODEV;
}
if (wsink == NULL) {
dev_err(dapm->dev, "ASoC: no sink widget found for %s\n",
route->sink);
return -ENODEV;
}
skip_search:
dapm_wcache_update(&dapm->path_sink_cache, wsink);
dapm_wcache_update(&dapm->path_source_cache, wsource);
ret = snd_soc_dapm_add_path(dapm, wsource, wsink, route->control,
route->connected);
if (ret)
goto err;
return 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
err:
dev_warn(dapm->dev, "ASoC: no dapm match for %s --> %s --> %s\n",
source, route->control, sink);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return ret;
}
static int snd_soc_dapm_del_route(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_route *route)
{
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
struct snd_soc_dapm_widget *wsource, *wsink;
struct snd_soc_dapm_path *path, *p;
const char *sink;
const char *source;
char prefixed_sink[80];
char prefixed_source[80];
const char *prefix;
if (route->control) {
dev_err(dapm->dev,
"ASoC: Removal of routes with controls not supported\n");
return -EINVAL;
}
prefix = soc_dapm_prefix(dapm);
if (prefix) {
snprintf(prefixed_sink, sizeof(prefixed_sink), "%s %s",
prefix, route->sink);
sink = prefixed_sink;
snprintf(prefixed_source, sizeof(prefixed_source), "%s %s",
prefix, route->source);
source = prefixed_source;
} else {
sink = route->sink;
source = route->source;
}
path = NULL;
list_for_each_entry(p, &dapm->card->paths, list) {
if (strcmp(p->source->name, source) != 0)
continue;
if (strcmp(p->sink->name, sink) != 0)
continue;
path = p;
break;
}
if (path) {
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
wsource = path->source;
wsink = path->sink;
dapm_mark_dirty(wsource, "Route removed");
dapm_mark_dirty(wsink, "Route removed");
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
if (path->connect)
dapm_path_invalidate(path);
dapm_free_path(path);
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
/* Update any path related flags */
dapm_update_widget_flags(wsource);
dapm_update_widget_flags(wsink);
} else {
dev_warn(dapm->dev, "ASoC: Route %s->%s does not exist\n",
source, sink);
}
return 0;
}
/**
* snd_soc_dapm_add_routes - Add routes between DAPM widgets
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
* @route: audio routes
* @num: number of routes
*
* Connects 2 dapm widgets together via a named audio path. The sink is
* the widget receiving the audio signal, whilst the source is the sender
* of the audio signal.
*
* Returns 0 for success else error. On error all resources can be freed
* with a call to snd_soc_card_free().
*/
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
int snd_soc_dapm_add_routes(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_route *route, int num)
{
int i, r, ret = 0;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
for (i = 0; i < num; i++) {
r = snd_soc_dapm_add_route(dapm, route);
if (r < 0) {
dev_err(dapm->dev, "ASoC: Failed to add route %s -> %s -> %s\n",
route->source,
route->control ? route->control : "direct",
route->sink);
ret = r;
}
route++;
}
mutex_unlock(&dapm->card->dapm_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_add_routes);
/**
* snd_soc_dapm_del_routes - Remove routes between DAPM widgets
* @dapm: DAPM context
* @route: audio routes
* @num: number of routes
*
* Removes routes from the DAPM context.
*/
int snd_soc_dapm_del_routes(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_route *route, int num)
{
int i;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
for (i = 0; i < num; i++) {
snd_soc_dapm_del_route(dapm, route);
route++;
}
mutex_unlock(&dapm->card->dapm_mutex);
return 0;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_del_routes);
static int snd_soc_dapm_weak_route(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_route *route)
{
struct snd_soc_dapm_widget *source = dapm_find_widget(dapm,
route->source,
true);
struct snd_soc_dapm_widget *sink = dapm_find_widget(dapm,
route->sink,
true);
struct snd_soc_dapm_path *path;
int count = 0;
if (!source) {
dev_err(dapm->dev, "ASoC: Unable to find source %s for weak route\n",
route->source);
return -ENODEV;
}
if (!sink) {
dev_err(dapm->dev, "ASoC: Unable to find sink %s for weak route\n",
route->sink);
return -ENODEV;
}
if (route->control || route->connected)
dev_warn(dapm->dev, "ASoC: Ignoring control for weak route %s->%s\n",
route->source, route->sink);
snd_soc_dapm_widget_for_each_sink_path(source, path) {
if (path->sink == sink) {
path->weak = 1;
count++;
}
}
if (count == 0)
dev_err(dapm->dev, "ASoC: No path found for weak route %s->%s\n",
route->source, route->sink);
if (count > 1)
dev_warn(dapm->dev, "ASoC: %d paths found for weak route %s->%s\n",
count, route->source, route->sink);
return 0;
}
/**
* snd_soc_dapm_weak_routes - Mark routes between DAPM widgets as weak
* @dapm: DAPM context
* @route: audio routes
* @num: number of routes
*
* Mark existing routes matching those specified in the passed array
* as being weak, meaning that they are ignored for the purpose of
* power decisions. The main intended use case is for sidetone paths
* which couple audio between other independent paths if they are both
* active in order to make the combination work better at the user
* level but which aren't intended to be "used".
*
* Note that CODEC drivers should not use this as sidetone type paths
* can frequently also be used as bypass paths.
*/
int snd_soc_dapm_weak_routes(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_route *route, int num)
{
int i, err;
int ret = 0;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_INIT);
for (i = 0; i < num; i++) {
err = snd_soc_dapm_weak_route(dapm, route);
if (err)
ret = err;
route++;
}
mutex_unlock(&dapm->card->dapm_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_weak_routes);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/**
* snd_soc_dapm_new_widgets - add new dapm widgets
* @card: card to be checked for new dapm widgets
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*
* Checks the codec for any new dapm widgets and creates them if found.
*
* Returns 0 for success.
*/
int snd_soc_dapm_new_widgets(struct snd_soc_card *card)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_widget *w;
unsigned int val;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_INIT);
list_for_each_entry(w, &card->widgets, list)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
if (w->new)
continue;
if (w->num_kcontrols) {
treewide: kzalloc() -> kcalloc() The kzalloc() function has a 2-factor argument form, kcalloc(). This patch replaces cases of: kzalloc(a * b, gfp) with: kcalloc(a * b, gfp) as well as handling cases of: kzalloc(a * b * c, gfp) with: kzalloc(array3_size(a, b, c), gfp) as it's slightly less ugly than: kzalloc_array(array_size(a, b), c, gfp) This does, however, attempt to ignore constant size factors like: kzalloc(4 * 1024, gfp) though any constants defined via macros get caught up in the conversion. Any factors with a sizeof() of "unsigned char", "char", and "u8" were dropped, since they're redundant. The Coccinelle script used for this was: // Fix redundant parens around sizeof(). @@ type TYPE; expression THING, E; @@ ( kzalloc( - (sizeof(TYPE)) * E + sizeof(TYPE) * E , ...) | kzalloc( - (sizeof(THING)) * E + sizeof(THING) * E , ...) ) // Drop single-byte sizes and redundant parens. @@ expression COUNT; typedef u8; typedef __u8; @@ ( kzalloc( - sizeof(u8) * (COUNT) + COUNT , ...) | kzalloc( - sizeof(__u8) * (COUNT) + COUNT , ...) | kzalloc( - sizeof(char) * (COUNT) + COUNT , ...) | kzalloc( - sizeof(unsigned char) * (COUNT) + COUNT , ...) | kzalloc( - sizeof(u8) * COUNT + COUNT , ...) | kzalloc( - sizeof(__u8) * COUNT + COUNT , ...) | kzalloc( - sizeof(char) * COUNT + COUNT , ...) | kzalloc( - sizeof(unsigned char) * COUNT + COUNT , ...) ) // 2-factor product with sizeof(type/expression) and identifier or constant. @@ type TYPE; expression THING; identifier COUNT_ID; constant COUNT_CONST; @@ ( - kzalloc + kcalloc ( - sizeof(TYPE) * (COUNT_ID) + COUNT_ID, sizeof(TYPE) , ...) | - kzalloc + kcalloc ( - sizeof(TYPE) * COUNT_ID + COUNT_ID, sizeof(TYPE) , ...) | - kzalloc + kcalloc ( - sizeof(TYPE) * (COUNT_CONST) + COUNT_CONST, sizeof(TYPE) , ...) | - kzalloc + kcalloc ( - sizeof(TYPE) * COUNT_CONST + COUNT_CONST, sizeof(TYPE) , ...) | - kzalloc + kcalloc ( - sizeof(THING) * (COUNT_ID) + COUNT_ID, sizeof(THING) , ...) | - kzalloc + kcalloc ( - sizeof(THING) * COUNT_ID + COUNT_ID, sizeof(THING) , ...) | - kzalloc + kcalloc ( - sizeof(THING) * (COUNT_CONST) + COUNT_CONST, sizeof(THING) , ...) | - kzalloc + kcalloc ( - sizeof(THING) * COUNT_CONST + COUNT_CONST, sizeof(THING) , ...) ) // 2-factor product, only identifiers. @@ identifier SIZE, COUNT; @@ - kzalloc + kcalloc ( - SIZE * COUNT + COUNT, SIZE , ...) // 3-factor product with 1 sizeof(type) or sizeof(expression), with // redundant parens removed. @@ expression THING; identifier STRIDE, COUNT; type TYPE; @@ ( kzalloc( - sizeof(TYPE) * (COUNT) * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kzalloc( - sizeof(TYPE) * (COUNT) * STRIDE + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kzalloc( - sizeof(TYPE) * COUNT * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kzalloc( - sizeof(TYPE) * COUNT * STRIDE + array3_size(COUNT, STRIDE, sizeof(TYPE)) , ...) | kzalloc( - sizeof(THING) * (COUNT) * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kzalloc( - sizeof(THING) * (COUNT) * STRIDE + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kzalloc( - sizeof(THING) * COUNT * (STRIDE) + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) | kzalloc( - sizeof(THING) * COUNT * STRIDE + array3_size(COUNT, STRIDE, sizeof(THING)) , ...) ) // 3-factor product with 2 sizeof(variable), with redundant parens removed. @@ expression THING1, THING2; identifier COUNT; type TYPE1, TYPE2; @@ ( kzalloc( - sizeof(TYPE1) * sizeof(TYPE2) * COUNT + array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2)) , ...) | kzalloc( - sizeof(TYPE1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(TYPE1), sizeof(TYPE2)) , ...) | kzalloc( - sizeof(THING1) * sizeof(THING2) * COUNT + array3_size(COUNT, sizeof(THING1), sizeof(THING2)) , ...) | kzalloc( - sizeof(THING1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(THING1), sizeof(THING2)) , ...) | kzalloc( - sizeof(TYPE1) * sizeof(THING2) * COUNT + array3_size(COUNT, sizeof(TYPE1), sizeof(THING2)) , ...) | kzalloc( - sizeof(TYPE1) * sizeof(THING2) * (COUNT) + array3_size(COUNT, sizeof(TYPE1), sizeof(THING2)) , ...) ) // 3-factor product, only identifiers, with redundant parens removed. @@ identifier STRIDE, SIZE, COUNT; @@ ( kzalloc( - (COUNT) * STRIDE * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kzalloc( - COUNT * (STRIDE) * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kzalloc( - COUNT * STRIDE * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kzalloc( - (COUNT) * (STRIDE) * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) | kzalloc( - COUNT * (STRIDE) * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kzalloc( - (COUNT) * STRIDE * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kzalloc( - (COUNT) * (STRIDE) * (SIZE) + array3_size(COUNT, STRIDE, SIZE) , ...) | kzalloc( - COUNT * STRIDE * SIZE + array3_size(COUNT, STRIDE, SIZE) , ...) ) // Any remaining multi-factor products, first at least 3-factor products, // when they're not all constants... @@ expression E1, E2, E3; constant C1, C2, C3; @@ ( kzalloc(C1 * C2 * C3, ...) | kzalloc( - (E1) * E2 * E3 + array3_size(E1, E2, E3) , ...) | kzalloc( - (E1) * (E2) * E3 + array3_size(E1, E2, E3) , ...) | kzalloc( - (E1) * (E2) * (E3) + array3_size(E1, E2, E3) , ...) | kzalloc( - E1 * E2 * E3 + array3_size(E1, E2, E3) , ...) ) // And then all remaining 2 factors products when they're not all constants, // keeping sizeof() as the second factor argument. @@ expression THING, E1, E2; type TYPE; constant C1, C2, C3; @@ ( kzalloc(sizeof(THING) * C2, ...) | kzalloc(sizeof(TYPE) * C2, ...) | kzalloc(C1 * C2 * C3, ...) | kzalloc(C1 * C2, ...) | - kzalloc + kcalloc ( - sizeof(TYPE) * (E2) + E2, sizeof(TYPE) , ...) | - kzalloc + kcalloc ( - sizeof(TYPE) * E2 + E2, sizeof(TYPE) , ...) | - kzalloc + kcalloc ( - sizeof(THING) * (E2) + E2, sizeof(THING) , ...) | - kzalloc + kcalloc ( - sizeof(THING) * E2 + E2, sizeof(THING) , ...) | - kzalloc + kcalloc ( - (E1) * E2 + E1, E2 , ...) | - kzalloc + kcalloc ( - (E1) * (E2) + E1, E2 , ...) | - kzalloc + kcalloc ( - E1 * E2 + E1, E2 , ...) ) Signed-off-by: Kees Cook <keescook@chromium.org>
2018-06-13 05:03:40 +08:00
w->kcontrols = kcalloc(w->num_kcontrols,
sizeof(struct snd_kcontrol *),
GFP_KERNEL);
if (!w->kcontrols) {
mutex_unlock(&card->dapm_mutex);
return -ENOMEM;
}
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
switch(w->id) {
case snd_soc_dapm_switch:
case snd_soc_dapm_mixer:
case snd_soc_dapm_mixer_named_ctl:
dapm_new_mixer(w);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
break;
case snd_soc_dapm_mux:
case snd_soc_dapm_demux:
dapm_new_mux(w);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
break;
case snd_soc_dapm_pga:
case snd_soc_dapm_out_drv:
dapm_new_pga(w);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
break;
case snd_soc_dapm_dai_link:
dapm_new_dai_link(w);
break;
default:
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
break;
}
/* Read the initial power state from the device */
if (w->reg >= 0) {
soc_dapm_read(w->dapm, w->reg, &val);
val = val >> w->shift;
val &= w->mask;
if (val == w->on_val)
w->power = 1;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
w->new = 1;
dapm_mark_dirty(w, "new widget");
dapm_debugfs_add_widget(w);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
dapm_power_widgets(card, SND_SOC_DAPM_STREAM_NOP);
mutex_unlock(&card->dapm_mutex);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
return 0;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_new_widgets);
/**
* snd_soc_dapm_get_volsw - dapm mixer get callback
* @kcontrol: mixer control
* @ucontrol: control element information
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*
* Callback to get the value of a dapm mixer control.
*
* Returns 0 for success.
*/
int snd_soc_dapm_get_volsw(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_value *ucontrol)
{
struct snd_soc_dapm_context *dapm = snd_soc_dapm_kcontrol_dapm(kcontrol);
struct snd_soc_card *card = dapm->card;
struct soc_mixer_control *mc =
(struct soc_mixer_control *)kcontrol->private_value;
int reg = mc->reg;
unsigned int shift = mc->shift;
int max = mc->max;
unsigned int width = fls(max);
unsigned int mask = (1 << fls(max)) - 1;
unsigned int invert = mc->invert;
unsigned int reg_val, val, rval = 0;
int ret = 0;
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
if (dapm_kcontrol_is_powered(kcontrol) && reg != SND_SOC_NOPM) {
ret = soc_dapm_read(dapm, reg, &reg_val);
val = (reg_val >> shift) & mask;
if (ret == 0 && reg != mc->rreg)
ret = soc_dapm_read(dapm, mc->rreg, &reg_val);
if (snd_soc_volsw_is_stereo(mc))
rval = (reg_val >> mc->rshift) & mask;
} else {
reg_val = dapm_kcontrol_get_value(kcontrol);
val = reg_val & mask;
if (snd_soc_volsw_is_stereo(mc))
rval = (reg_val >> width) & mask;
}
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
mutex_unlock(&card->dapm_mutex);
if (ret)
return ret;
if (invert)
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
ucontrol->value.integer.value[0] = max - val;
else
ucontrol->value.integer.value[0] = val;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
if (snd_soc_volsw_is_stereo(mc)) {
if (invert)
ucontrol->value.integer.value[1] = max - rval;
else
ucontrol->value.integer.value[1] = rval;
}
return ret;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_get_volsw);
/**
* snd_soc_dapm_put_volsw - dapm mixer set callback
* @kcontrol: mixer control
* @ucontrol: control element information
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*
* Callback to set the value of a dapm mixer control.
*
* Returns 0 for success.
*/
int snd_soc_dapm_put_volsw(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_value *ucontrol)
{
struct snd_soc_dapm_context *dapm = snd_soc_dapm_kcontrol_dapm(kcontrol);
struct snd_soc_card *card = dapm->card;
struct soc_mixer_control *mc =
(struct soc_mixer_control *)kcontrol->private_value;
int reg = mc->reg;
unsigned int shift = mc->shift;
int max = mc->max;
unsigned int width = fls(max);
unsigned int mask = (1 << width) - 1;
unsigned int invert = mc->invert;
unsigned int val, rval = 0;
int connect, rconnect = -1, change, reg_change = 0;
struct snd_soc_dapm_update update = {};
int ret = 0;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
val = (ucontrol->value.integer.value[0] & mask);
connect = !!val;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
if (invert)
val = max - val;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
if (snd_soc_volsw_is_stereo(mc)) {
rval = (ucontrol->value.integer.value[1] & mask);
rconnect = !!rval;
if (invert)
rval = max - rval;
}
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* This assumes field width < (bits in unsigned int / 2) */
if (width > sizeof(unsigned int) * 8 / 2)
dev_warn(dapm->dev,
"ASoC: control %s field width limit exceeded\n",
kcontrol->id.name);
change = dapm_kcontrol_set_value(kcontrol, val | (rval << width));
if (reg != SND_SOC_NOPM) {
val = val << shift;
rval = rval << mc->rshift;
reg_change = soc_dapm_test_bits(dapm, reg, mask << shift, val);
if (snd_soc_volsw_is_stereo(mc))
reg_change |= soc_dapm_test_bits(dapm, mc->rreg,
mask << mc->rshift,
rval);
}
if (change || reg_change) {
if (reg_change) {
if (snd_soc_volsw_is_stereo(mc)) {
update.has_second_set = true;
update.reg2 = mc->rreg;
update.mask2 = mask << mc->rshift;
update.val2 = rval;
}
update.kcontrol = kcontrol;
update.reg = reg;
update.mask = mask << shift;
update.val = val;
card->update = &update;
}
change |= reg_change;
ret = soc_dapm_mixer_update_power(card, kcontrol, connect,
rconnect);
card->update = NULL;
}
mutex_unlock(&card->dapm_mutex);
if (ret > 0)
soc_dpcm_runtime_update(card);
return change;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_put_volsw);
/**
* snd_soc_dapm_get_enum_double - dapm enumerated double mixer get callback
* @kcontrol: mixer control
* @ucontrol: control element information
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*
* Callback to get the value of a dapm enumerated double mixer control.
*
* Returns 0 for success.
*/
int snd_soc_dapm_get_enum_double(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_value *ucontrol)
{
struct snd_soc_dapm_context *dapm = snd_soc_dapm_kcontrol_dapm(kcontrol);
struct snd_soc_card *card = dapm->card;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct soc_enum *e = (struct soc_enum *)kcontrol->private_value;
unsigned int reg_val, val;
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
if (e->reg != SND_SOC_NOPM && dapm_kcontrol_is_powered(kcontrol)) {
int ret = soc_dapm_read(dapm, e->reg, &reg_val);
if (ret) {
mutex_unlock(&card->dapm_mutex);
return ret;
}
} else {
reg_val = dapm_kcontrol_get_value(kcontrol);
}
mutex_unlock(&card->dapm_mutex);
val = (reg_val >> e->shift_l) & e->mask;
ucontrol->value.enumerated.item[0] = snd_soc_enum_val_to_item(e, val);
if (e->shift_l != e->shift_r) {
val = (reg_val >> e->shift_r) & e->mask;
val = snd_soc_enum_val_to_item(e, val);
ucontrol->value.enumerated.item[1] = val;
}
return 0;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
EXPORT_SYMBOL_GPL(snd_soc_dapm_get_enum_double);
/**
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
* snd_soc_dapm_put_enum_double - dapm enumerated double mixer set callback
* @kcontrol: mixer control
* @ucontrol: control element information
*
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
* Callback to set the value of a dapm enumerated double mixer control.
*
* Returns 0 for success.
*/
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
int snd_soc_dapm_put_enum_double(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_value *ucontrol)
{
struct snd_soc_dapm_context *dapm = snd_soc_dapm_kcontrol_dapm(kcontrol);
struct snd_soc_card *card = dapm->card;
struct soc_enum *e = (struct soc_enum *)kcontrol->private_value;
unsigned int *item = ucontrol->value.enumerated.item;
unsigned int val, change, reg_change = 0;
unsigned int mask;
struct snd_soc_dapm_update update = {};
int ret = 0;
if (item[0] >= e->items)
return -EINVAL;
val = snd_soc_enum_item_to_val(e, item[0]) << e->shift_l;
mask = e->mask << e->shift_l;
if (e->shift_l != e->shift_r) {
if (item[1] > e->items)
return -EINVAL;
val |= snd_soc_enum_item_to_val(e, item[1]) << e->shift_r;
mask |= e->mask << e->shift_r;
}
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
change = dapm_kcontrol_set_value(kcontrol, val);
if (e->reg != SND_SOC_NOPM)
reg_change = soc_dapm_test_bits(dapm, e->reg, mask, val);
if (change || reg_change) {
if (reg_change) {
update.kcontrol = kcontrol;
update.reg = e->reg;
update.mask = mask;
update.val = val;
card->update = &update;
}
change |= reg_change;
ret = soc_dapm_mux_update_power(card, kcontrol, item[0], e);
card->update = NULL;
}
mutex_unlock(&card->dapm_mutex);
if (ret > 0)
soc_dpcm_runtime_update(card);
return change;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
EXPORT_SYMBOL_GPL(snd_soc_dapm_put_enum_double);
/**
* snd_soc_dapm_info_pin_switch - Info for a pin switch
*
* @kcontrol: mixer control
* @uinfo: control element information
*
* Callback to provide information about a pin switch control.
*/
int snd_soc_dapm_info_pin_switch(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_info *uinfo)
{
uinfo->type = SNDRV_CTL_ELEM_TYPE_BOOLEAN;
uinfo->count = 1;
uinfo->value.integer.min = 0;
uinfo->value.integer.max = 1;
return 0;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_info_pin_switch);
/**
* snd_soc_dapm_get_pin_switch - Get information for a pin switch
*
* @kcontrol: mixer control
* @ucontrol: Value
*/
int snd_soc_dapm_get_pin_switch(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_value *ucontrol)
{
struct snd_soc_card *card = snd_kcontrol_chip(kcontrol);
const char *pin = (const char *)kcontrol->private_value;
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
ucontrol->value.integer.value[0] =
snd_soc_dapm_get_pin_status(&card->dapm, pin);
mutex_unlock(&card->dapm_mutex);
return 0;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_get_pin_switch);
/**
* snd_soc_dapm_put_pin_switch - Set information for a pin switch
*
* @kcontrol: mixer control
* @ucontrol: Value
*/
int snd_soc_dapm_put_pin_switch(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_value *ucontrol)
{
struct snd_soc_card *card = snd_kcontrol_chip(kcontrol);
const char *pin = (const char *)kcontrol->private_value;
if (ucontrol->value.integer.value[0])
snd_soc_dapm_enable_pin(&card->dapm, pin);
else
snd_soc_dapm_disable_pin(&card->dapm, pin);
snd_soc_dapm_sync(&card->dapm);
return 0;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_put_pin_switch);
struct snd_soc_dapm_widget *
snd_soc_dapm_new_control_unlocked(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_widget *widget)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
enum snd_soc_dapm_direction dir;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct snd_soc_dapm_widget *w;
const char *prefix;
int ret;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
if ((w = dapm_cnew_widget(widget)) == NULL)
return ERR_PTR(-ENOMEM);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
switch (w->id) {
case snd_soc_dapm_regulator_supply:
w->regulator = devm_regulator_get(dapm->dev, w->name);
if (IS_ERR(w->regulator)) {
ret = PTR_ERR(w->regulator);
goto request_failed;
}
if (w->on_val & SND_SOC_DAPM_REGULATOR_BYPASS) {
ret = regulator_allow_bypass(w->regulator, true);
if (ret != 0)
dev_warn(dapm->dev,
"ASoC: Failed to bypass %s: %d\n",
w->name, ret);
}
break;
case snd_soc_dapm_pinctrl:
w->pinctrl = devm_pinctrl_get(dapm->dev);
if (IS_ERR(w->pinctrl)) {
ret = PTR_ERR(w->pinctrl);
goto request_failed;
}
break;
case snd_soc_dapm_clock_supply:
w->clk = devm_clk_get(dapm->dev, w->name);
if (IS_ERR(w->clk)) {
ret = PTR_ERR(w->clk);
goto request_failed;
}
break;
default:
break;
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
prefix = soc_dapm_prefix(dapm);
ASoC: dapm: Don't add prefix to widget stream name Commit fdb6eb0a1287 ("ASoC: dapm: Modify widget stream name according to prefix") fixed the case where a DAPM route between a DAI widget and a DAC/ADC/AIF widget with a matching stream name was not created when the DAPM context was using a prefix. Unfortunately the patch introduced a few issues on its own like leaking the dynamically allocated stream name memory and also not checking whether the allocation succeeded in the first place. It is also incomplete in that it still does not handle the case where stream name of the widget is a substring of the stream name of the DAI, which is explicitly allowed and works fine if no DAPM prefix is used. Revert the commit and take a slightly different approach to solving the issue. Instead of comparing the widget's stream name to the name of the DAI widget compare it to the stream name of the DAI widget. The stream name of the DAI widget is identical to the name of the DAI widget except that it wont have the DAPM prefix added. So this approach behaves identical regardless to whether the DAPM context uses a prefix or not. We don't have to worry about potentially matching with a widget with the same stream name, but from a different DAPM context with a different prefix, since the code already makes sure that both the DAI widget and the matched widget are from the same DAPM context. Fixes: fdb6eb0a1287 ("ASoC: dapm: Modify widget stream name according to prefix") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org> Cc: stable@vger.kernel.org
2015-07-21 17:51:35 +08:00
if (prefix)
w->name = kasprintf(GFP_KERNEL, "%s %s", prefix, widget->name);
ASoC: dapm: Don't add prefix to widget stream name Commit fdb6eb0a1287 ("ASoC: dapm: Modify widget stream name according to prefix") fixed the case where a DAPM route between a DAI widget and a DAC/ADC/AIF widget with a matching stream name was not created when the DAPM context was using a prefix. Unfortunately the patch introduced a few issues on its own like leaking the dynamically allocated stream name memory and also not checking whether the allocation succeeded in the first place. It is also incomplete in that it still does not handle the case where stream name of the widget is a substring of the stream name of the DAI, which is explicitly allowed and works fine if no DAPM prefix is used. Revert the commit and take a slightly different approach to solving the issue. Instead of comparing the widget's stream name to the name of the DAI widget compare it to the stream name of the DAI widget. The stream name of the DAI widget is identical to the name of the DAI widget except that it wont have the DAPM prefix added. So this approach behaves identical regardless to whether the DAPM context uses a prefix or not. We don't have to worry about potentially matching with a widget with the same stream name, but from a different DAPM context with a different prefix, since the code already makes sure that both the DAI widget and the matched widget are from the same DAPM context. Fixes: fdb6eb0a1287 ("ASoC: dapm: Modify widget stream name according to prefix") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org> Cc: stable@vger.kernel.org
2015-07-21 17:51:35 +08:00
else
w->name = kstrdup_const(widget->name, GFP_KERNEL);
if (w->name == NULL) {
kfree_const(w->sname);
kfree(w);
return ERR_PTR(-ENOMEM);
}
switch (w->id) {
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
case snd_soc_dapm_mic:
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->is_ep = SND_SOC_DAPM_EP_SOURCE;
w->power_check = dapm_generic_check_power;
break;
case snd_soc_dapm_input:
if (!dapm->card->fully_routed)
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->is_ep = SND_SOC_DAPM_EP_SOURCE;
w->power_check = dapm_generic_check_power;
break;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
case snd_soc_dapm_spk:
case snd_soc_dapm_hp:
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->is_ep = SND_SOC_DAPM_EP_SINK;
w->power_check = dapm_generic_check_power;
break;
case snd_soc_dapm_output:
if (!dapm->card->fully_routed)
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->is_ep = SND_SOC_DAPM_EP_SINK;
w->power_check = dapm_generic_check_power;
break;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
case snd_soc_dapm_vmid:
case snd_soc_dapm_siggen:
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->is_ep = SND_SOC_DAPM_EP_SOURCE;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
w->power_check = dapm_always_on_check_power;
break;
case snd_soc_dapm_sink:
w->is_ep = SND_SOC_DAPM_EP_SINK;
w->power_check = dapm_always_on_check_power;
break;
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
case snd_soc_dapm_mux:
case snd_soc_dapm_demux:
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
case snd_soc_dapm_switch:
case snd_soc_dapm_mixer:
case snd_soc_dapm_mixer_named_ctl:
case snd_soc_dapm_adc:
case snd_soc_dapm_aif_out:
case snd_soc_dapm_dac:
case snd_soc_dapm_aif_in:
case snd_soc_dapm_pga:
case snd_soc_dapm_out_drv:
case snd_soc_dapm_micbias:
case snd_soc_dapm_line:
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
case snd_soc_dapm_dai_link:
case snd_soc_dapm_dai_out:
case snd_soc_dapm_dai_in:
w->power_check = dapm_generic_check_power;
break;
case snd_soc_dapm_supply:
case snd_soc_dapm_regulator_supply:
case snd_soc_dapm_pinctrl:
case snd_soc_dapm_clock_supply:
ASoC: dapm: Implement mixer input auto-disable Some devices have the problem that if a internal audio signal source is disabled the output of the source becomes undefined or goes to a undesired state (E.g. DAC output goes to ground instead of VMID). In this case it is necessary, in order to avoid unwanted clicks and pops, to disable any mixer input the signal feeds into or to active a mute control along the path to the output. Often it is still desirable to expose the same mixer input control to userspace, so cerain paths can sill be disabled manually. This means we can not use conventional DAPM to manage the mixer input control. This patch implements a method for letting DAPM overwrite the state of a userspace visible control. I.e. DAPM will disable the control if the path on which the control sits becomes inactive. Userspace will then only see a cached copy of the controls state. Once DAPM powers the path up again it will sync the userspace setting with the hardware and give control back to userspace. To implement this a new widget type is introduced. One widget of this type will be created for each DAPM kcontrol which has the auto-disable feature enabled. For each path that is controlled by the kcontrol the widget will be connected to the source of that path. The new widget type behaves like a supply widget, which means it will power up if one of its sinks are powered up and will only power down if all of its sinks are powered down. In order to only have the mixer input enabled when the source signal is valid the new widget type will be disabled before all other widget types and only be enabled after all other widget types. E.g. consider the following simplified example. A DAC is connected to a mixer and the mixer has a control to enable or disable the signal from the DAC. +-------+ +-----+ | | | DAC |-----[Ctrl]-| Mixer | +-----+ : | | | : +-------+ | : +-------------+ | Ctrl widget | +-------------+ If the control has the auto-disable feature enabled we'll create a widget for the control. This widget is connected to the DAC as it is the source for the mixer input. If the DAC powers up the control widget powers up and if the DAC powers down the control widget is powered down. As long as the control widget is powered down the hardware input control is kept disabled and if it is enabled userspace can freely change the control's state. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@linaro.org>
2013-08-05 17:27:31 +08:00
case snd_soc_dapm_kcontrol:
ASoC: dapm: Introduce toplevel widget categories DAPM widgets can be classified into four categories: * supply: Supply widgets do not affect the power state of their non-supply widget neighbors and unlike other widgets a supply widget is not powered up when it is on an active path, but when at least on of its neighbors is powered up. * source: A source is a widget that receives data from outside the DAPM graph or generates data. This can for example be a microphone, the playback DMA or a signal generator. A source widget will be considered powered up if there is an active path to a sink widget. * sink: A sink is a widget that transmits data to somewhere outside of the DAPM graph. This can e.g. be a speaker or the capture DMA. A sink widget will be considered powered up if there is an active path from a source widget. * normal: Normal widgets are widgets not covered by the categories above. A normal widget will be considered powered up if it is on an active path between a source widget and a sink widget. The way the number of input and output paths for a widget is calculated depends on its category. There are a bunch of factors which decide which category a widget is. Currently there is no formal classification of these categories and we calculate the category of the widget based on these factors whenever we want to know it. This is at least once for every widget during each power update sequence. The factors which determine the category of the widgets are mostly static though and if at all change rather seldom. This patch introduces three new per widget flags, one for each of non-normal widgets categories. Instead of re-computing the category each time we want to know them the flags will be checked. For the majority of widgets the category is solely determined by the widget id, which means it never changes and only has to be set once when the widget is created. The only widgets with dynamic categories are: snd_soc_dapm_dai_out: Is considered a sink iff the capture stream is active, otherwise normal. snd_soc_dapm_dai_in: Is considered a source iff the playback stream is active, otherwise normal. snd_soc_dapm_input: Is considered a sink iff it has no outgoing paths, otherwise normal. snd_soc_dapm_output: Is considered a source iff it has no incoming paths, otherwise normal. snd_soc_dapm_line: Is considered a sink iff it has no outgoing paths and is considered a source iff it has no incoming paths, otherwise normal. For snd_soc_dapm_dai_out/snd_soc_dapm_dai_in widgets the category will be updated when a stream is started or stopped. For the other dynamic widgets the category will be updated when a path connecting to it is added or removed. Introducing those new widget categories allows to make is_connected_{output,input}_ep, which are among the hottest paths of the DAPM algorithm, more generic and significantly shorter. The before and after sizes for is_connected_{output,input}_ep are: On ARM (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 480 340 -140 is_connected_input_ep 456 352 -104 On amd64 (defconfig + CONFIG_SND_SOC): function old new delta is_connected_output_ep 579 427 -152 is_connected_input_ep 563 427 -136 Which is about a 25%-30% decrease, other architectures are expected to have similar numbers. At the same time the size of the snd_soc_dapm_widget struct does not change since the new flags are stored in the same word as the existing flags. Note: that since the per widget 'ext' flag was only used to decide whether a snd_soc_dapm_input or snd_soc_dapm_output widget was a source or a sink it is now unused and can be removed. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:41:59 +08:00
w->is_supply = 1;
w->power_check = dapm_supply_check_power;
break;
default:
w->power_check = dapm_always_on_check_power;
break;
}
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
w->dapm = dapm;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
INIT_LIST_HEAD(&w->list);
INIT_LIST_HEAD(&w->dirty);
list_add_tail(&w->list, &dapm->card->widgets);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
snd_soc_dapm_for_each_direction(dir) {
INIT_LIST_HEAD(&w->edges[dir]);
w->endpoints[dir] = -1;
}
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
/* machine layer sets up unconnected pins and insertions */
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
w->connected = 1;
return w;
request_failed:
if (ret != -EPROBE_DEFER)
dev_err(dapm->dev, "ASoC: Failed to request %s: %d\n",
w->name, ret);
return ERR_PTR(ret);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
/**
* snd_soc_dapm_new_control - create new dapm control
* @dapm: DAPM context
* @widget: widget template
*
* Creates new DAPM control based upon a template.
*
* Returns a widget pointer on success or an error pointer on failure
*/
struct snd_soc_dapm_widget *
snd_soc_dapm_new_control(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_widget *widget)
{
struct snd_soc_dapm_widget *w;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
w = snd_soc_dapm_new_control_unlocked(dapm, widget);
mutex_unlock(&dapm->card->dapm_mutex);
return w;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_new_control);
/**
* snd_soc_dapm_new_controls - create new dapm controls
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
* @widget: widget array
* @num: number of widgets
*
* Creates new DAPM controls based upon the templates.
*
* Returns 0 for success else error.
*/
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
int snd_soc_dapm_new_controls(struct snd_soc_dapm_context *dapm,
const struct snd_soc_dapm_widget *widget,
int num)
{
struct snd_soc_dapm_widget *w;
int i;
int ret = 0;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_INIT);
for (i = 0; i < num; i++) {
w = snd_soc_dapm_new_control_unlocked(dapm, widget);
ASoC: dapm: handle probe deferrals This starts to handle probe deferrals on regulators and clocks on the ASoC DAPM. I came to this patch after audio stopped working on Ux500 ages ago and I finally looked into it to see what is wrong. I had messages like this in the console since a while back: ab8500-codec.0: ASoC: Failed to request audioclk: -517 ab8500-codec.0: ASoC: Failed to create DAPM control audioclk ab8500-codec.0: Failed to create new controls -12 snd-soc-mop500.0: ASoC: failed to instantiate card -12 snd-soc-mop500.0: Error: snd_soc_register_card failed (-12)! snd-soc-mop500: probe of snd-soc-mop500.0 failed with error -12 Apparently because the widget table for the codec looks like this (sound/soc/codecs/ab8500-codec.c): static const struct snd_soc_dapm_widget ab8500_dapm_widgets[] = { /* Clocks */ SND_SOC_DAPM_CLOCK_SUPPLY("audioclk"), /* Regulators */ SND_SOC_DAPM_REGULATOR_SUPPLY("V-AUD", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-AMIC1", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-AMIC2", 0, 0), SND_SOC_DAPM_REGULATOR_SUPPLY("V-DMIC", 0, 0), So when we call snd_soc_register_codec() and any of these widgets get a deferred probe we do not get an -EPROBE_DEFER (-517) back as we should and instead we just fail. Apparently the code assumes that clocks and regulators must be available at this point and not defer. After this patch it rather looks like this: ab8500-codec.0: Failed to create new controls -517 snd-soc-mop500.0: ASoC: failed to instantiate card -517 snd-soc-mop500.0: Error: snd_soc_register_card failed (-517)! (...) abx500-clk.0: registered clocks for ab850x snd-soc-mop500.0: ab8500-codec-dai.0 <-> ux500-msp-i2s.1 mapping ok snd-soc-mop500.0: ab8500-codec-dai.1 <-> ux500-msp-i2s.3 mapping ok I'm pretty happy about the patch as it it, but I'm a bit uncertain on how to proceed: there are a lot of users of the external functions snd_soc_dapm_new_control() (111 sites) and that will now return an occassional error pointer, which is not handled in the calling sites. I want an indication from the maintainers whether I should just go in and augment all these call sites, or if deferred probe is frowned upon when it leads to this much overhead. Signed-off-by: Linus Walleij <linus.walleij@linaro.org> Signed-off-by: Mark Brown <broonie@kernel.org>
2017-01-13 17:23:52 +08:00
if (IS_ERR(w)) {
ret = PTR_ERR(w);
break;
}
widget++;
}
mutex_unlock(&dapm->card->dapm_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_new_controls);
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
static int snd_soc_dai_link_event(struct snd_soc_dapm_widget *w,
struct snd_kcontrol *kcontrol, int event)
{
struct snd_soc_dapm_path *path;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
struct snd_soc_dai *source, *sink;
struct snd_soc_pcm_runtime *rtd = w->priv;
const struct snd_soc_pcm_stream *config;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
struct snd_pcm_substream substream;
struct snd_pcm_hw_params *params = NULL;
ASoC: dapm: Add a dummy snd_pcm_runtime to avoid NULL pointer access The SND_SOC_DAPM_PRE_PMU case would call startup()/hw_params() that might access substream->runtime through other functions. For example: Unable to handle kernel NULL pointer dereference at virtual address [....] PC is at snd_pcm_hw_rule_add+0x24/0x1b0 LR is at snd_pcm_hw_constraint_list+0x20/0x28 [....] Process arecord (pid: 424, stack limit = 0xffffffc1ecaf0020) Call trace: [<ffffffc00086be68>] snd_pcm_hw_rule_add+0x24/0x1b0 [<ffffffc00086c014>] snd_pcm_hw_constraint_list+0x20/0x28 [<ffffffc0008b47a4>] cs53l30_pcm_startup+0x24/0x30 [<ffffffc0008a6260>] snd_soc_dai_link_event+0x290/0x354 [<ffffffc0008a7528>] dapm_seq_check_event.isra.31+0x134/0x2c8 [<ffffffc0008a7768>] dapm_seq_run_coalesced+0x94/0x1c8 [<ffffffc0008a7940>] dapm_seq_run+0xa4/0x404 [<ffffffc0008a8bac>] dapm_power_widgets+0x524/0x984 [<ffffffc0008ab1c4>] snd_soc_dapm_stream_event+0x8c/0xa8 [<ffffffc0008ac7f4>] soc_pcm_prepare+0x10c/0x1ec [<ffffffc000865b9c>] snd_pcm_do_prepare+0x1c/0x38 [<ffffffc000865600>] snd_pcm_action_single+0x40/0x88 [<ffffffc0008656b8>] snd_pcm_action_nonatomic+0x70/0x90 [<ffffffc000868d28>] snd_pcm_common_ioctl1+0xb6c/0xdd8 [<ffffffc000869508>] snd_pcm_capture_ioctl1+0x200/0x334 [<ffffffc00086a084>] snd_pcm_ioctl_compat+0x648/0x95c [<ffffffc0001ff4b4>] compat_SyS_ioctl+0xac/0xfc4 [<ffffffc000084cf0>] el0_svc_naked+0x24/0x28 ---[ end trace 0dc4f99c2759c35c ]--- So this patch adds a dummy runtime for the original dummy substream to merely avoid the NULL pointer access. Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com> Signed-off-by: Mark Brown <broonie@kernel.org>
2016-07-27 05:55:51 +08:00
struct snd_pcm_runtime *runtime = NULL;
unsigned int fmt;
int ret = 0;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
config = rtd->dai_link->params + rtd->params_select;
if (WARN_ON(!config) ||
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
WARN_ON(list_empty(&w->edges[SND_SOC_DAPM_DIR_OUT]) ||
list_empty(&w->edges[SND_SOC_DAPM_DIR_IN])))
return -EINVAL;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
/* Be a little careful as we don't want to overflow the mask array */
if (config->formats) {
fmt = ffs(config->formats) - 1;
} else {
dev_warn(w->dapm->dev, "ASoC: Invalid format %llx specified\n",
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
config->formats);
fmt = 0;
}
/* Currently very limited parameter selection */
params = kzalloc(sizeof(*params), GFP_KERNEL);
if (!params) {
ret = -ENOMEM;
goto out;
}
snd_mask_set(hw_param_mask(params, SNDRV_PCM_HW_PARAM_FORMAT), fmt);
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
hw_param_interval(params, SNDRV_PCM_HW_PARAM_RATE)->min =
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
config->rate_min;
hw_param_interval(params, SNDRV_PCM_HW_PARAM_RATE)->max =
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
config->rate_max;
hw_param_interval(params, SNDRV_PCM_HW_PARAM_CHANNELS)->min
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
= config->channels_min;
hw_param_interval(params, SNDRV_PCM_HW_PARAM_CHANNELS)->max
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
= config->channels_max;
memset(&substream, 0, sizeof(substream));
ASoC: dapm: Add a dummy snd_pcm_runtime to avoid NULL pointer access The SND_SOC_DAPM_PRE_PMU case would call startup()/hw_params() that might access substream->runtime through other functions. For example: Unable to handle kernel NULL pointer dereference at virtual address [....] PC is at snd_pcm_hw_rule_add+0x24/0x1b0 LR is at snd_pcm_hw_constraint_list+0x20/0x28 [....] Process arecord (pid: 424, stack limit = 0xffffffc1ecaf0020) Call trace: [<ffffffc00086be68>] snd_pcm_hw_rule_add+0x24/0x1b0 [<ffffffc00086c014>] snd_pcm_hw_constraint_list+0x20/0x28 [<ffffffc0008b47a4>] cs53l30_pcm_startup+0x24/0x30 [<ffffffc0008a6260>] snd_soc_dai_link_event+0x290/0x354 [<ffffffc0008a7528>] dapm_seq_check_event.isra.31+0x134/0x2c8 [<ffffffc0008a7768>] dapm_seq_run_coalesced+0x94/0x1c8 [<ffffffc0008a7940>] dapm_seq_run+0xa4/0x404 [<ffffffc0008a8bac>] dapm_power_widgets+0x524/0x984 [<ffffffc0008ab1c4>] snd_soc_dapm_stream_event+0x8c/0xa8 [<ffffffc0008ac7f4>] soc_pcm_prepare+0x10c/0x1ec [<ffffffc000865b9c>] snd_pcm_do_prepare+0x1c/0x38 [<ffffffc000865600>] snd_pcm_action_single+0x40/0x88 [<ffffffc0008656b8>] snd_pcm_action_nonatomic+0x70/0x90 [<ffffffc000868d28>] snd_pcm_common_ioctl1+0xb6c/0xdd8 [<ffffffc000869508>] snd_pcm_capture_ioctl1+0x200/0x334 [<ffffffc00086a084>] snd_pcm_ioctl_compat+0x648/0x95c [<ffffffc0001ff4b4>] compat_SyS_ioctl+0xac/0xfc4 [<ffffffc000084cf0>] el0_svc_naked+0x24/0x28 ---[ end trace 0dc4f99c2759c35c ]--- So this patch adds a dummy runtime for the original dummy substream to merely avoid the NULL pointer access. Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com> Signed-off-by: Mark Brown <broonie@kernel.org>
2016-07-27 05:55:51 +08:00
/* Allocate a dummy snd_pcm_runtime for startup() and other ops() */
runtime = kzalloc(sizeof(*runtime), GFP_KERNEL);
if (!runtime) {
ret = -ENOMEM;
goto out;
}
substream.runtime = runtime;
substream.private_data = rtd;
ASoC: dapm: Add a dummy snd_pcm_runtime to avoid NULL pointer access The SND_SOC_DAPM_PRE_PMU case would call startup()/hw_params() that might access substream->runtime through other functions. For example: Unable to handle kernel NULL pointer dereference at virtual address [....] PC is at snd_pcm_hw_rule_add+0x24/0x1b0 LR is at snd_pcm_hw_constraint_list+0x20/0x28 [....] Process arecord (pid: 424, stack limit = 0xffffffc1ecaf0020) Call trace: [<ffffffc00086be68>] snd_pcm_hw_rule_add+0x24/0x1b0 [<ffffffc00086c014>] snd_pcm_hw_constraint_list+0x20/0x28 [<ffffffc0008b47a4>] cs53l30_pcm_startup+0x24/0x30 [<ffffffc0008a6260>] snd_soc_dai_link_event+0x290/0x354 [<ffffffc0008a7528>] dapm_seq_check_event.isra.31+0x134/0x2c8 [<ffffffc0008a7768>] dapm_seq_run_coalesced+0x94/0x1c8 [<ffffffc0008a7940>] dapm_seq_run+0xa4/0x404 [<ffffffc0008a8bac>] dapm_power_widgets+0x524/0x984 [<ffffffc0008ab1c4>] snd_soc_dapm_stream_event+0x8c/0xa8 [<ffffffc0008ac7f4>] soc_pcm_prepare+0x10c/0x1ec [<ffffffc000865b9c>] snd_pcm_do_prepare+0x1c/0x38 [<ffffffc000865600>] snd_pcm_action_single+0x40/0x88 [<ffffffc0008656b8>] snd_pcm_action_nonatomic+0x70/0x90 [<ffffffc000868d28>] snd_pcm_common_ioctl1+0xb6c/0xdd8 [<ffffffc000869508>] snd_pcm_capture_ioctl1+0x200/0x334 [<ffffffc00086a084>] snd_pcm_ioctl_compat+0x648/0x95c [<ffffffc0001ff4b4>] compat_SyS_ioctl+0xac/0xfc4 [<ffffffc000084cf0>] el0_svc_naked+0x24/0x28 ---[ end trace 0dc4f99c2759c35c ]--- So this patch adds a dummy runtime for the original dummy substream to merely avoid the NULL pointer access. Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com> Signed-off-by: Mark Brown <broonie@kernel.org>
2016-07-27 05:55:51 +08:00
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
switch (event) {
case SND_SOC_DAPM_PRE_PMU:
substream.stream = SNDRV_PCM_STREAM_CAPTURE;
snd_soc_dapm_widget_for_each_source_path(w, path) {
source = path->source->priv;
if (source->driver->ops->startup) {
ret = source->driver->ops->startup(&substream,
source);
if (ret < 0) {
dev_err(source->dev,
"ASoC: startup() failed: %d\n",
ret);
goto out;
}
source->active++;
}
ret = soc_dai_hw_params(&substream, params, source);
if (ret < 0)
goto out;
dapm_update_dai_unlocked(&substream, params, source);
}
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
substream.stream = SNDRV_PCM_STREAM_PLAYBACK;
snd_soc_dapm_widget_for_each_sink_path(w, path) {
sink = path->sink->priv;
if (sink->driver->ops->startup) {
ret = sink->driver->ops->startup(&substream,
sink);
if (ret < 0) {
dev_err(sink->dev,
"ASoC: startup() failed: %d\n",
ret);
goto out;
}
sink->active++;
}
ret = soc_dai_hw_params(&substream, params, sink);
if (ret < 0)
goto out;
dapm_update_dai_unlocked(&substream, params, sink);
}
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
break;
case SND_SOC_DAPM_POST_PMU:
snd_soc_dapm_widget_for_each_sink_path(w, path) {
sink = path->sink->priv;
ret = snd_soc_dai_digital_mute(sink, 0,
SNDRV_PCM_STREAM_PLAYBACK);
if (ret != 0 && ret != -ENOTSUPP)
dev_warn(sink->dev,
"ASoC: Failed to unmute: %d\n", ret);
ret = 0;
}
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
break;
case SND_SOC_DAPM_PRE_PMD:
snd_soc_dapm_widget_for_each_sink_path(w, path) {
sink = path->sink->priv;
ret = snd_soc_dai_digital_mute(sink, 1,
SNDRV_PCM_STREAM_PLAYBACK);
if (ret != 0 && ret != -ENOTSUPP)
dev_warn(sink->dev,
"ASoC: Failed to mute: %d\n", ret);
ret = 0;
}
substream.stream = SNDRV_PCM_STREAM_CAPTURE;
snd_soc_dapm_widget_for_each_source_path(w, path) {
source = path->source->priv;
if (source->driver->ops->hw_free)
source->driver->ops->hw_free(&substream,
source);
source->active--;
if (source->driver->ops->shutdown)
source->driver->ops->shutdown(&substream,
source);
}
substream.stream = SNDRV_PCM_STREAM_PLAYBACK;
snd_soc_dapm_widget_for_each_sink_path(w, path) {
sink = path->sink->priv;
if (sink->driver->ops->hw_free)
sink->driver->ops->hw_free(&substream, sink);
sink->active--;
if (sink->driver->ops->shutdown)
sink->driver->ops->shutdown(&substream, sink);
}
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
break;
default:
WARN(1, "Unknown event %d\n", event);
ret = -EINVAL;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
}
out:
ASoC: dapm: Add a dummy snd_pcm_runtime to avoid NULL pointer access The SND_SOC_DAPM_PRE_PMU case would call startup()/hw_params() that might access substream->runtime through other functions. For example: Unable to handle kernel NULL pointer dereference at virtual address [....] PC is at snd_pcm_hw_rule_add+0x24/0x1b0 LR is at snd_pcm_hw_constraint_list+0x20/0x28 [....] Process arecord (pid: 424, stack limit = 0xffffffc1ecaf0020) Call trace: [<ffffffc00086be68>] snd_pcm_hw_rule_add+0x24/0x1b0 [<ffffffc00086c014>] snd_pcm_hw_constraint_list+0x20/0x28 [<ffffffc0008b47a4>] cs53l30_pcm_startup+0x24/0x30 [<ffffffc0008a6260>] snd_soc_dai_link_event+0x290/0x354 [<ffffffc0008a7528>] dapm_seq_check_event.isra.31+0x134/0x2c8 [<ffffffc0008a7768>] dapm_seq_run_coalesced+0x94/0x1c8 [<ffffffc0008a7940>] dapm_seq_run+0xa4/0x404 [<ffffffc0008a8bac>] dapm_power_widgets+0x524/0x984 [<ffffffc0008ab1c4>] snd_soc_dapm_stream_event+0x8c/0xa8 [<ffffffc0008ac7f4>] soc_pcm_prepare+0x10c/0x1ec [<ffffffc000865b9c>] snd_pcm_do_prepare+0x1c/0x38 [<ffffffc000865600>] snd_pcm_action_single+0x40/0x88 [<ffffffc0008656b8>] snd_pcm_action_nonatomic+0x70/0x90 [<ffffffc000868d28>] snd_pcm_common_ioctl1+0xb6c/0xdd8 [<ffffffc000869508>] snd_pcm_capture_ioctl1+0x200/0x334 [<ffffffc00086a084>] snd_pcm_ioctl_compat+0x648/0x95c [<ffffffc0001ff4b4>] compat_SyS_ioctl+0xac/0xfc4 [<ffffffc000084cf0>] el0_svc_naked+0x24/0x28 ---[ end trace 0dc4f99c2759c35c ]--- So this patch adds a dummy runtime for the original dummy substream to merely avoid the NULL pointer access. Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com> Signed-off-by: Mark Brown <broonie@kernel.org>
2016-07-27 05:55:51 +08:00
kfree(runtime);
kfree(params);
return ret;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
}
static int snd_soc_dapm_dai_link_get(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_value *ucontrol)
{
struct snd_soc_dapm_widget *w = snd_kcontrol_chip(kcontrol);
struct snd_soc_pcm_runtime *rtd = w->priv;
ucontrol->value.enumerated.item[0] = rtd->params_select;
return 0;
}
static int snd_soc_dapm_dai_link_put(struct snd_kcontrol *kcontrol,
struct snd_ctl_elem_value *ucontrol)
{
struct snd_soc_dapm_widget *w = snd_kcontrol_chip(kcontrol);
struct snd_soc_pcm_runtime *rtd = w->priv;
/* Can't change the config when widget is already powered */
if (w->power)
return -EBUSY;
if (ucontrol->value.enumerated.item[0] == rtd->params_select)
return 0;
if (ucontrol->value.enumerated.item[0] >= rtd->dai_link->num_params)
return -EINVAL;
rtd->params_select = ucontrol->value.enumerated.item[0];
return 0;
}
static void
snd_soc_dapm_free_kcontrol(struct snd_soc_card *card,
unsigned long *private_value,
int num_params,
const char **w_param_text)
{
int count;
devm_kfree(card->dev, (void *)*private_value);
for (count = 0 ; count < num_params; count++)
devm_kfree(card->dev, (void *)w_param_text[count]);
devm_kfree(card->dev, w_param_text);
}
static struct snd_kcontrol_new *
snd_soc_dapm_alloc_kcontrol(struct snd_soc_card *card,
char *link_name,
const struct snd_soc_pcm_stream *params,
int num_params, const char **w_param_text,
unsigned long *private_value)
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
{
struct soc_enum w_param_enum[] = {
SOC_ENUM_SINGLE(0, 0, 0, NULL),
};
struct snd_kcontrol_new kcontrol_dai_link[] = {
SOC_ENUM_EXT(NULL, w_param_enum[0],
snd_soc_dapm_dai_link_get,
snd_soc_dapm_dai_link_put),
};
struct snd_kcontrol_new *kcontrol_news;
const struct snd_soc_pcm_stream *config = params;
int count;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
for (count = 0 ; count < num_params; count++) {
if (!config->stream_name) {
dev_warn(card->dapm.dev,
"ASoC: anonymous config %d for dai link %s\n",
count, link_name);
w_param_text[count] =
devm_kasprintf(card->dev, GFP_KERNEL,
"Anonymous Configuration %d",
count);
} else {
w_param_text[count] = devm_kmemdup(card->dev,
config->stream_name,
strlen(config->stream_name) + 1,
GFP_KERNEL);
}
if (!w_param_text[count])
goto outfree_w_param;
config++;
}
w_param_enum[0].items = num_params;
w_param_enum[0].texts = w_param_text;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
*private_value =
(unsigned long) devm_kmemdup(card->dev,
(void *)(kcontrol_dai_link[0].private_value),
sizeof(struct soc_enum), GFP_KERNEL);
if (!*private_value) {
dev_err(card->dev, "ASoC: Failed to create control for %s widget\n",
link_name);
goto outfree_w_param;
}
kcontrol_dai_link[0].private_value = *private_value;
/* duplicate kcontrol_dai_link on heap so that memory persists */
kcontrol_news = devm_kmemdup(card->dev, &kcontrol_dai_link[0],
sizeof(struct snd_kcontrol_new),
GFP_KERNEL);
if (!kcontrol_news) {
dev_err(card->dev, "ASoC: Failed to create control for %s widget\n",
link_name);
goto outfree_w_param;
}
return kcontrol_news;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
outfree_w_param:
snd_soc_dapm_free_kcontrol(card, private_value, num_params, w_param_text);
return NULL;
}
static struct snd_soc_dapm_widget *
snd_soc_dapm_new_dai(struct snd_soc_card *card, struct snd_soc_pcm_runtime *rtd,
struct snd_soc_dapm_widget *source,
struct snd_soc_dapm_widget *sink)
{
struct snd_soc_dapm_widget template;
struct snd_soc_dapm_widget *w;
const char **w_param_text;
unsigned long private_value;
char *link_name;
int ret;
link_name = devm_kasprintf(card->dev, GFP_KERNEL, "%s-%s",
source->name, sink->name);
if (!link_name)
return ERR_PTR(-ENOMEM);
memset(&template, 0, sizeof(template));
template.reg = SND_SOC_NOPM;
template.id = snd_soc_dapm_dai_link;
template.name = link_name;
template.event = snd_soc_dai_link_event;
template.event_flags = SND_SOC_DAPM_PRE_PMU | SND_SOC_DAPM_POST_PMU |
SND_SOC_DAPM_PRE_PMD;
template.kcontrol_news = NULL;
/* allocate memory for control, only in case of multiple configs */
if (rtd->dai_link->num_params > 1) {
w_param_text = devm_kcalloc(card->dev,
rtd->dai_link->num_params,
sizeof(char *), GFP_KERNEL);
if (!w_param_text) {
ret = -ENOMEM;
goto param_fail;
}
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
template.num_kcontrols = 1;
template.kcontrol_news =
snd_soc_dapm_alloc_kcontrol(card,
link_name,
rtd->dai_link->params,
rtd->dai_link->num_params,
w_param_text, &private_value);
if (!template.kcontrol_news) {
ret = -ENOMEM;
goto param_fail;
}
} else {
w_param_text = NULL;
}
dev_dbg(card->dev, "ASoC: adding %s widget\n", link_name);
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
w = snd_soc_dapm_new_control_unlocked(&card->dapm, &template);
if (IS_ERR(w)) {
ret = PTR_ERR(w);
goto outfree_kcontrol_news;
}
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
w->priv = rtd;
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
return w;
outfree_kcontrol_news:
devm_kfree(card->dev, (void *)template.kcontrol_news);
snd_soc_dapm_free_kcontrol(card, &private_value,
rtd->dai_link->num_params, w_param_text);
param_fail:
devm_kfree(card->dev, link_name);
return ERR_PTR(ret);
ASoC: core: Support transparent CODEC<->CODEC DAI links Rather than having the user half start a stream but avoid any DMA to trigger data flow on links which don't pass through the CPU create a DAPM route between the two DAI widgets using a hw_params configuration provided by the machine driver with the new 'params' member of the dai_link struct. If no configuration is provided in the dai_link then use the old style even for CODEC<->CODEC links to avoid breaking systems. This greatly simplifies the userspace usage of such links, making them as simple as analogue connections with the stream configuration being completely transparent to them. This is achieved by defining a new dai_link widget type which is created when CODECs are linked and triggering the configuration of the link via the normal PCM operations from there. It is expected that the bias level callbacks will be used for clock configuration. Currently only the DAI format, rate and channel count can be configured and currently the only DAI operations which can be called are hw_params and digital_mute(). This corresponds well to the majority of CODEC drivers which only use other callbacks for constraint setting but there is obviously much room for extension here. We can't simply call hw_params() on startup as things like the system clocking configuration may change at runtime and in future it will be desirable to offer some configurability of the link parameters. At present we are also restricted to a single DAPM link for the entire DAI. Once we have better support for channel mapping it would also be desirable to extend this feature so that we can propagate per-channel power state over the link. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-04-05 05:12:09 +08:00
}
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
int snd_soc_dapm_new_dai_widgets(struct snd_soc_dapm_context *dapm,
struct snd_soc_dai *dai)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
struct snd_soc_dapm_widget template;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
struct snd_soc_dapm_widget *w;
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
WARN_ON(dapm->dev != dai->dev);
memset(&template, 0, sizeof(template));
template.reg = SND_SOC_NOPM;
if (dai->driver->playback.stream_name) {
template.id = snd_soc_dapm_dai_in;
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
template.name = dai->driver->playback.stream_name;
template.sname = dai->driver->playback.stream_name;
dev_dbg(dai->dev, "ASoC: adding %s widget\n",
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
template.name);
w = snd_soc_dapm_new_control_unlocked(dapm, &template);
if (IS_ERR(w))
return PTR_ERR(w);
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
w->priv = dai;
dai->playback_widget = w;
}
if (dai->driver->capture.stream_name) {
template.id = snd_soc_dapm_dai_out;
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
template.name = dai->driver->capture.stream_name;
template.sname = dai->driver->capture.stream_name;
dev_dbg(dai->dev, "ASoC: adding %s widget\n",
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
template.name);
w = snd_soc_dapm_new_control_unlocked(dapm, &template);
if (IS_ERR(w))
return PTR_ERR(w);
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
w->priv = dai;
dai->capture_widget = w;
}
return 0;
}
int snd_soc_dapm_link_dai_widgets(struct snd_soc_card *card)
{
struct snd_soc_dapm_widget *dai_w, *w;
struct snd_soc_dapm_widget *src, *sink;
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
struct snd_soc_dai *dai;
/* For each DAI widget... */
list_for_each_entry(dai_w, &card->widgets, list) {
switch (dai_w->id) {
case snd_soc_dapm_dai_in:
case snd_soc_dapm_dai_out:
break;
default:
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
continue;
}
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
/* let users know there is no DAI to link */
if (!dai_w->priv) {
dev_dbg(card->dev, "dai widget %s has no DAI\n",
dai_w->name);
continue;
}
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
dai = dai_w->priv;
/* ...find all widgets with the same stream and link them */
list_for_each_entry(w, &card->widgets, list) {
if (w->dapm != dai_w->dapm)
continue;
switch (w->id) {
case snd_soc_dapm_dai_in:
case snd_soc_dapm_dai_out:
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
continue;
default:
break;
}
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
ASoC: dapm: Don't add prefix to widget stream name Commit fdb6eb0a1287 ("ASoC: dapm: Modify widget stream name according to prefix") fixed the case where a DAPM route between a DAI widget and a DAC/ADC/AIF widget with a matching stream name was not created when the DAPM context was using a prefix. Unfortunately the patch introduced a few issues on its own like leaking the dynamically allocated stream name memory and also not checking whether the allocation succeeded in the first place. It is also incomplete in that it still does not handle the case where stream name of the widget is a substring of the stream name of the DAI, which is explicitly allowed and works fine if no DAPM prefix is used. Revert the commit and take a slightly different approach to solving the issue. Instead of comparing the widget's stream name to the name of the DAI widget compare it to the stream name of the DAI widget. The stream name of the DAI widget is identical to the name of the DAI widget except that it wont have the DAPM prefix added. So this approach behaves identical regardless to whether the DAPM context uses a prefix or not. We don't have to worry about potentially matching with a widget with the same stream name, but from a different DAPM context with a different prefix, since the code already makes sure that both the DAI widget and the matched widget are from the same DAPM context. Fixes: fdb6eb0a1287 ("ASoC: dapm: Modify widget stream name according to prefix") Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org> Cc: stable@vger.kernel.org
2015-07-21 17:51:35 +08:00
if (!w->sname || !strstr(w->sname, dai_w->sname))
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
continue;
if (dai_w->id == snd_soc_dapm_dai_in) {
src = dai_w;
sink = w;
} else {
src = w;
sink = dai_w;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
dev_dbg(dai->dev, "%s -> %s\n", src->name, sink->name);
snd_soc_dapm_add_path(w->dapm, src, sink, NULL, NULL);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
}
ASoC: dapm: Implement and instantiate DAI widgets In order to allow us to do smarter things with DAI links create DAPM widgets which directly represent the DAIs in the DAPM graph. These are automatically created from the DAIs as we probe the card with references held in both directions between the widget and the DAI. The widgets are not made available for direct instantiation by drivers, they are created automatically from the DAIs. Drivers should be updated to create stream routes using DAPM maps rather than by annotating AIF and DAC widgets with streams. In order to ease transition to this model from existing drivers we automatically create DAPM routes between the DAI widgets and the existing stream widgets which are started and stopped by the DAI widgets, though the old stream handling mechanism is still in place. This also has the nice effect of removing non-DAPM devices as any device with a DAI acquires a widget automatically which will allow future simplifications to the core DAPM logic. The intention is that in future the AIF and DAI widgets will gain the ability to interact such that we are able to manage activity on individual channels independantly rather than powering up and down the entire AIF as we do currently. Currently we only generate these for CODECs, mostly as I have no systems with non-CODEC DAPM to integrate with. It should be a simple matter of programming to add the additional hookup for these. Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Acked-by: Liam Girdwood <lrg@ti.com>
2012-02-17 11:37:51 +08:00
return 0;
}
static void dapm_connect_dai_link_widgets(struct snd_soc_card *card,
struct snd_soc_pcm_runtime *rtd)
{
struct snd_soc_dai *cpu_dai = rtd->cpu_dai;
struct snd_soc_dai *codec_dai;
struct snd_soc_dapm_widget *playback = NULL, *capture = NULL;
struct snd_soc_dapm_widget *codec, *playback_cpu, *capture_cpu;
int i;
if (rtd->dai_link->params) {
playback_cpu = cpu_dai->capture_widget;
capture_cpu = cpu_dai->playback_widget;
} else {
playback = cpu_dai->playback_widget;
capture = cpu_dai->capture_widget;
playback_cpu = playback;
capture_cpu = capture;
}
for_each_rtd_codec_dai(rtd, i, codec_dai) {
/* connect BE DAI playback if widgets are valid */
codec = codec_dai->playback_widget;
if (playback_cpu && codec) {
if (!playback) {
playback = snd_soc_dapm_new_dai(card, rtd,
playback_cpu,
codec);
if (IS_ERR(playback)) {
dev_err(rtd->dev,
"ASoC: Failed to create DAI %s: %ld\n",
codec_dai->name,
PTR_ERR(playback));
continue;
}
snd_soc_dapm_add_path(&card->dapm, playback_cpu,
playback, NULL, NULL);
}
dev_dbg(rtd->dev, "connected DAI link %s:%s -> %s:%s\n",
cpu_dai->component->name, playback_cpu->name,
codec_dai->component->name, codec->name);
snd_soc_dapm_add_path(&card->dapm, playback, codec,
NULL, NULL);
}
}
for_each_rtd_codec_dai(rtd, i, codec_dai) {
/* connect BE DAI capture if widgets are valid */
codec = codec_dai->capture_widget;
if (codec && capture_cpu) {
if (!capture) {
capture = snd_soc_dapm_new_dai(card, rtd,
codec,
capture_cpu);
if (IS_ERR(capture)) {
dev_err(rtd->dev,
"ASoC: Failed to create DAI %s: %ld\n",
codec_dai->name,
PTR_ERR(capture));
continue;
}
snd_soc_dapm_add_path(&card->dapm, capture,
capture_cpu, NULL, NULL);
}
dev_dbg(rtd->dev, "connected DAI link %s:%s -> %s:%s\n",
codec_dai->component->name, codec->name,
cpu_dai->component->name, capture_cpu->name);
snd_soc_dapm_add_path(&card->dapm, codec, capture,
NULL, NULL);
}
}
}
static void soc_dapm_dai_stream_event(struct snd_soc_dai *dai, int stream,
int event)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
struct snd_soc_dapm_widget *w;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
unsigned int ep;
if (stream == SNDRV_PCM_STREAM_PLAYBACK)
w = dai->playback_widget;
else
w = dai->capture_widget;
if (w) {
dapm_mark_dirty(w, "stream event");
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
if (w->id == snd_soc_dapm_dai_in) {
ep = SND_SOC_DAPM_EP_SOURCE;
dapm_widget_invalidate_input_paths(w);
} else {
ep = SND_SOC_DAPM_EP_SINK;
dapm_widget_invalidate_output_paths(w);
}
switch (event) {
case SND_SOC_DAPM_STREAM_START:
w->active = 1;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->is_ep = ep;
break;
case SND_SOC_DAPM_STREAM_STOP:
w->active = 0;
ASoC: dapm: Consolidate input and output path handling After the recent cleanups and generalizations of the DAPM algorithm the handling of input and output paths is now fully symmetric. This means by making some slight changes to the data structure and using arrays with one entry for each direction, rather than separate fields, it is possible to create a generic implementation that is capable of handling both input and output paths. Unfortunately this generalization significantly increases the code size on the hot path of is_connected_{input,output}_ep() and dapm_widget_invalidate_{input,output}_paths(), which has a negative impact on the overall performance. The inner loops of those functions are quite small and the generic implementation adds extra pointer arithmetic in a few places. Testing on ARM shows that the combined code size of the specialized functions is about 50% larger than the generalized function in relative numbers. But in absolute numbers its less than 200 bytes, which is still quite small. On the other hand the generalized function increases the execution time of dapm_power_one_widget() by 30%. Given that this function is one of the most often called functions of the DAPM framework the trade-off of getting better performance at expense of generating slightly larger code at seems to be worth it. To avoid this still keep two versions of these functions around, one for input and one for output. But have a generic implementation of the algorithm which gets inlined by those two versions. And then let the compiler take care of optimizing it and removing he extra instructions. This still reduces the source code size as well as the makes making changes to the implementation more straight forward since the same change does no longer need to be done in two separate places. Also on the slow paths we can use a generic implementations that handle both input and output paths. Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2015-08-12 03:38:00 +08:00
w->is_ep = 0;
break;
case SND_SOC_DAPM_STREAM_SUSPEND:
case SND_SOC_DAPM_STREAM_RESUME:
case SND_SOC_DAPM_STREAM_PAUSE_PUSH:
case SND_SOC_DAPM_STREAM_PAUSE_RELEASE:
break;
}
}
}
void snd_soc_dapm_connect_dai_link_widgets(struct snd_soc_card *card)
{
2015-11-18 15:34:11 +08:00
struct snd_soc_pcm_runtime *rtd;
/* for each BE DAI link... */
for_each_card_rtds(card, rtd) {
/*
* dynamic FE links have no fixed DAI mapping.
* CODEC<->CODEC links have no direct connection.
*/
if (rtd->dai_link->dynamic)
continue;
dapm_connect_dai_link_widgets(card, rtd);
}
}
static void soc_dapm_stream_event(struct snd_soc_pcm_runtime *rtd, int stream,
int event)
{
struct snd_soc_dai *codec_dai;
int i;
soc_dapm_dai_stream_event(rtd->cpu_dai, stream, event);
for_each_rtd_codec_dai(rtd, i, codec_dai)
soc_dapm_dai_stream_event(codec_dai, stream, event);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
dapm_power_widgets(rtd->card, event);
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
}
/**
* snd_soc_dapm_stream_event - send a stream event to the dapm core
* @rtd: PCM runtime data
* @stream: stream name
* @event: stream event
*
* Sends a stream event to the dapm core. The core then makes any
* necessary widget power changes.
*
* Returns 0 for success else error.
*/
void snd_soc_dapm_stream_event(struct snd_soc_pcm_runtime *rtd, int stream,
int event)
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
{
struct snd_soc_card *card = rtd->card;
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
mutex_lock_nested(&card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
soc_dapm_stream_event(rtd, stream, event);
mutex_unlock(&card->dapm_mutex);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
/**
* snd_soc_dapm_enable_pin_unlocked - enable pin.
* @dapm: DAPM context
* @pin: pin name
*
* Enables input/output pin and its parents or children widgets iff there is
* a valid audio route and active audio stream.
*
* Requires external locking.
*
* NOTE: snd_soc_dapm_sync() needs to be called after this for DAPM to
* do any widget power switching.
*/
int snd_soc_dapm_enable_pin_unlocked(struct snd_soc_dapm_context *dapm,
const char *pin)
{
return snd_soc_dapm_set_pin(dapm, pin, 1);
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_enable_pin_unlocked);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/**
* snd_soc_dapm_enable_pin - enable pin.
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
* @pin: pin name
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*
* Enables input/output pin and its parents or children widgets iff there is
* a valid audio route and active audio stream.
*
* NOTE: snd_soc_dapm_sync() needs to be called after this for DAPM to
* do any widget power switching.
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*/
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
int snd_soc_dapm_enable_pin(struct snd_soc_dapm_context *dapm, const char *pin)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
int ret;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
ret = snd_soc_dapm_set_pin(dapm, pin, 1);
mutex_unlock(&dapm->card->dapm_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_enable_pin);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/**
* snd_soc_dapm_force_enable_pin_unlocked - force a pin to be enabled
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
* @pin: pin name
*
* Enables input/output pin regardless of any other state. This is
* intended for use with microphone bias supplies used in microphone
* jack detection.
*
* Requires external locking.
*
* NOTE: snd_soc_dapm_sync() needs to be called after this for DAPM to
* do any widget power switching.
*/
int snd_soc_dapm_force_enable_pin_unlocked(struct snd_soc_dapm_context *dapm,
const char *pin)
{
struct snd_soc_dapm_widget *w = dapm_find_widget(dapm, pin, true);
if (!w) {
dev_err(dapm->dev, "ASoC: unknown pin %s\n", pin);
return -EINVAL;
}
dev_dbg(w->dapm->dev, "ASoC: force enable pin %s\n", pin);
ASoC: dapm: Use more aggressive caching Currently we cache the number of input and output paths going to/from a widget only within a power update sequence. But not in between power update sequences. But we know how changes to the DAPM graph affect the number of input (form a source) and output (to a sink) paths of a widget and only need to recalculate them if a operation has been performed that might have changed them. * Adding/removing or connecting/disconnecting a path means that the for the source of the path the number of output paths can change and for the sink the number of input paths can change. * Connecting/disconnecting a widget has the same effect has connecting/ disconnecting all paths of the widget. So for the widget itself the number of inputs and outputs can change, for all sinks of the widget the number of inputs can change and for all sources of the widget the number of outputs can change. * Activating/Deactivating a stream can either change the number of outputs on the sources of the widget associated with the stream or the number of inputs on the sinks. Instead of always invalidating all cached numbers of input and output paths for each power up or down sequence this patch restructures the code to only invalidate the cached numbers when a operation that might change them has been performed. This can greatly reduce the number of DAPM power checks for some very common operations. Since per DAPM operation typically only either change the number of inputs or outputs the number of path checks is reduced by at least 50%. The number of neighbor checks is also reduced about the same percentage, but since the number of neighbors encountered when walking from sink to source is not the same as when walking from source to sink the actual numbers will slightly vary from card to card (e.g. for a mixer we see 1 neighbor when walking from source to sink, but the number of inputs neighbors when walking from source to sink). Bigger improvements can be observed for widgets with multiple connected inputs and output (e.g. mixers probably being the most widespread form of this). Previously we had to re-calculate the number of inputs and outputs on all input and output paths. With this change we only have to re-calculate the number of outputs on the input path that got changed and the number of inputs on the output paths. E.g. imagine the following example: A --> B ----. v M --> N --> Z <-- S <-- R | v X Widget Z has multiple input paths, if any change was made that cause Z to be marked as dirty the power state of Z has to be re-computed. This requires to know the number of inputs and outputs of Z, which requires to know the number of inputs and outputs of all widgets on all paths from or to Z. Previously this meant re-computing all inputs and outputs of all the path going into or out of Z. With this patch in place only paths that actually have changed need to be re-computed. If the system is idle (or the part of the system affected by the changed path) the number of path checks drops to either 0 or 1, regardless of how large or complex the DAPM context is. 0 if there is no connected sink and no connected source. 1 if there is either a connected source or sink, but not both. The number of neighbor checks again will scale accordingly and will be a constant number that is the number of inputs or outputs of the widget for which we did the path check. When loading a state file or switching between different profiles typically multiple mixer and mux settings are changed, so we see the benefit of this patch multiplied for these kinds of operations. Testing with the ADAU1761 shows the following changes in DAPM stats for changing a single Mixer switch for a Mixer with 5 inputs while the DAPM context is idle. Power Path Neighbour Before: 2 12 30 After: 2 1 2 For the same switch, but with a active playback stream the stat changed are as follows. Power Path Neighbour Before: 10 20 54 After: 10 7 21 Cumulative numbers for switching the audio profile which changes 7 controls while the system is idle: Power Path Neighbour Before: 16 80 170 After: 16 7 23 Cumulative numbers for switching the audio profile which changes 7 controls while playback is active: Power Path Neighbour Before: 51 123 273 After: 51 29 109 Starting (or stopping) the playback stream: Power Path Neighbour Before: 34 34 117 After: 34 17 69 Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Mark Brown <broonie@kernel.org>
2014-10-25 23:42:03 +08:00
if (!w->connected) {
/*
* w->force does not affect the number of input or output paths,
* so we only have to recheck if w->connected is changed
*/
dapm_widget_invalidate_input_paths(w);
dapm_widget_invalidate_output_paths(w);
w->connected = 1;
}
w->force = 1;
dapm_mark_dirty(w, "force enable");
return 0;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_force_enable_pin_unlocked);
/**
* snd_soc_dapm_force_enable_pin - force a pin to be enabled
* @dapm: DAPM context
* @pin: pin name
*
* Enables input/output pin regardless of any other state. This is
* intended for use with microphone bias supplies used in microphone
* jack detection.
*
* NOTE: snd_soc_dapm_sync() needs to be called after this for DAPM to
* do any widget power switching.
*/
int snd_soc_dapm_force_enable_pin(struct snd_soc_dapm_context *dapm,
const char *pin)
{
int ret;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
ret = snd_soc_dapm_force_enable_pin_unlocked(dapm, pin);
mutex_unlock(&dapm->card->dapm_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_force_enable_pin);
/**
* snd_soc_dapm_disable_pin_unlocked - disable pin.
* @dapm: DAPM context
* @pin: pin name
*
* Disables input/output pin and its parents or children widgets.
*
* Requires external locking.
*
* NOTE: snd_soc_dapm_sync() needs to be called after this for DAPM to
* do any widget power switching.
*/
int snd_soc_dapm_disable_pin_unlocked(struct snd_soc_dapm_context *dapm,
const char *pin)
{
return snd_soc_dapm_set_pin(dapm, pin, 0);
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_disable_pin_unlocked);
/**
* snd_soc_dapm_disable_pin - disable pin.
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
* @pin: pin name
*
* Disables input/output pin and its parents or children widgets.
*
* NOTE: snd_soc_dapm_sync() needs to be called after this for DAPM to
* do any widget power switching.
*/
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
int snd_soc_dapm_disable_pin(struct snd_soc_dapm_context *dapm,
const char *pin)
{
int ret;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
ret = snd_soc_dapm_set_pin(dapm, pin, 0);
mutex_unlock(&dapm->card->dapm_mutex);
return ret;
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_disable_pin);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/**
* snd_soc_dapm_nc_pin_unlocked - permanently disable pin.
* @dapm: DAPM context
* @pin: pin name
*
* Marks the specified pin as being not connected, disabling it along
* any parent or child widgets. At present this is identical to
* snd_soc_dapm_disable_pin() but in future it will be extended to do
* additional things such as disabling controls which only affect
* paths through the pin.
*
* Requires external locking.
*
* NOTE: snd_soc_dapm_sync() needs to be called after this for DAPM to
* do any widget power switching.
*/
int snd_soc_dapm_nc_pin_unlocked(struct snd_soc_dapm_context *dapm,
const char *pin)
{
return snd_soc_dapm_set_pin(dapm, pin, 0);
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_nc_pin_unlocked);
/**
* snd_soc_dapm_nc_pin - permanently disable pin.
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
* @pin: pin name
*
* Marks the specified pin as being not connected, disabling it along
* any parent or child widgets. At present this is identical to
* snd_soc_dapm_disable_pin() but in future it will be extended to do
* additional things such as disabling controls which only affect
* paths through the pin.
*
* NOTE: snd_soc_dapm_sync() needs to be called after this for DAPM to
* do any widget power switching.
*/
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
int snd_soc_dapm_nc_pin(struct snd_soc_dapm_context *dapm, const char *pin)
{
int ret;
mutex_lock_nested(&dapm->card->dapm_mutex, SND_SOC_DAPM_CLASS_RUNTIME);
ret = snd_soc_dapm_set_pin(dapm, pin, 0);
mutex_unlock(&dapm->card->dapm_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_nc_pin);
/**
* snd_soc_dapm_get_pin_status - get audio pin status
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
* @pin: audio signal pin endpoint (or start point)
*
* Get audio pin status - connected or disconnected.
*
* Returns 1 for connected otherwise 0.
*/
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
int snd_soc_dapm_get_pin_status(struct snd_soc_dapm_context *dapm,
const char *pin)
{
struct snd_soc_dapm_widget *w = dapm_find_widget(dapm, pin, true);
if (w)
return w->connected;
return 0;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_get_pin_status);
/**
* snd_soc_dapm_ignore_suspend - ignore suspend status for DAPM endpoint
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
* @dapm: DAPM context
* @pin: audio signal pin endpoint (or start point)
*
* Mark the given endpoint or pin as ignoring suspend. When the
* system is disabled a path between two endpoints flagged as ignoring
* suspend will not be disabled. The path must already be enabled via
* normal means at suspend time, it will not be turned on if it was not
* already enabled.
*/
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
int snd_soc_dapm_ignore_suspend(struct snd_soc_dapm_context *dapm,
const char *pin)
{
struct snd_soc_dapm_widget *w = dapm_find_widget(dapm, pin, false);
if (!w) {
dev_err(dapm->dev, "ASoC: unknown pin %s\n", pin);
return -EINVAL;
}
w->ignore_suspend = 1;
return 0;
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_ignore_suspend);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/**
* snd_soc_dapm_free - free dapm resources
* @dapm: DAPM context
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
*
* Free all dapm widgets and resources.
*/
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
void snd_soc_dapm_free(struct snd_soc_dapm_context *dapm)
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
{
dapm_debugfs_cleanup(dapm);
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
dapm_free_widgets(dapm);
list_del(&dapm->list);
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
}
EXPORT_SYMBOL_GPL(snd_soc_dapm_free);
static void soc_dapm_shutdown_dapm(struct snd_soc_dapm_context *dapm)
{
struct snd_soc_card *card = dapm->card;
struct snd_soc_dapm_widget *w;
LIST_HEAD(down_list);
int powerdown = 0;
mutex_lock(&card->dapm_mutex);
list_for_each_entry(w, &dapm->card->widgets, list) {
if (w->dapm != dapm)
continue;
if (w->power) {
dapm_seq_insert(w, &down_list, false);
w->power = 0;
powerdown = 1;
}
}
/* If there were no widgets to power down we're already in
* standby.
*/
if (powerdown) {
if (dapm->bias_level == SND_SOC_BIAS_ON)
snd_soc_dapm_set_bias_level(dapm,
SND_SOC_BIAS_PREPARE);
dapm_seq_run(card, &down_list, 0, false);
if (dapm->bias_level == SND_SOC_BIAS_PREPARE)
snd_soc_dapm_set_bias_level(dapm,
SND_SOC_BIAS_STANDBY);
}
mutex_unlock(&card->dapm_mutex);
ASoC: multi-component - ASoC Multi-Component Support This patch extends the ASoC API to allow sound cards to have more than one CODEC and more than one platform DMA controller. This is achieved by dividing some current ASoC structures that contain both driver data and device data into structures that only either contain device data or driver data. i.e. struct snd_soc_codec ---> struct snd_soc_codec (device data) +-> struct snd_soc_codec_driver (driver data) struct snd_soc_platform ---> struct snd_soc_platform (device data) +-> struct snd_soc_platform_driver (driver data) struct snd_soc_dai ---> struct snd_soc_dai (device data) +-> struct snd_soc_dai_driver (driver data) struct snd_soc_device ---> deleted This now allows ASoC to be more tightly aligned with the Linux driver model and also means that every ASoC codec, platform and (platform) DAI is a kernel device. ASoC component private data is now stored as device private data. The ASoC sound card struct snd_soc_card has also been updated to store lists of it's components rather than a pointer to a codec and platform. The PCM runtime struct soc_pcm_runtime now has pointers to all its components. This patch adds DAPM support for ASoC multi-component and removes struct snd_soc_socdev from DAPM core. All DAPM calls are now made on a card, codec or runtime PCM level basis rather than using snd_soc_socdev. Other notable multi-component changes:- * Stream operations now de-reference less structures. * close_delayed work() now runs on a DAI basis rather than looping all DAIs in a card. * PM suspend()/resume() operations can now handle N CODECs and Platforms per sound card. * Added soc_bind_dai_link() to bind the component devices to the sound card. * Added soc_dai_link_probe() and soc_dai_link_remove() to probe and remove DAI link components. * sysfs entries can now be registered per component per card. * snd_soc_new_pcms() functionailty rolled into dai_link_probe(). * snd_soc_register_codec() now does all the codec list and mutex init. This patch changes the probe() and remove() of the CODEC drivers as follows:- o Make CODEC driver a platform driver o Moved all struct snd_soc_codec list, mutex, etc initialiasation to core. o Removed all static codec pointers (drivers now support > 1 codec dev) o snd_soc_register_pcms() now done by core. o snd_soc_register_dai() folded into snd_soc_register_codec(). CS4270 portions: Acked-by: Timur Tabi <timur@freescale.com> Some TLV320aic23 and Cirrus platform fixes. Signed-off-by: Ryan Mallon <ryan@bluewatersys.com> TI CODEC and OMAP fixes Signed-off-by: Peter Ujfalusi <peter.ujfalusi@nokia.com> Signed-off-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Samsung platform and misc fixes :- Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com> Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Reviewed-by: Jassi Brar <jassi.brar@samsung.com> Signed-off-by: Seungwhan Youn <sw.youn@samsung.com> MPC8610 and PPC fixes. Signed-off-by: Timur Tabi <timur@freescale.com> i.MX fixes and some core fixes. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> J4740 platform fixes:- Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> CC: Tony Lindgren <tony@atomide.com> CC: Nicolas Ferre <nicolas.ferre@atmel.com> CC: Kevin Hilman <khilman@deeprootsystems.com> CC: Sascha Hauer <s.hauer@pengutronix.de> CC: Atsushi Nemoto <anemo@mba.ocn.ne.jp> CC: Kuninori Morimoto <morimoto.kuninori@renesas.com> CC: Daniel Gloeckner <dg@emlix.com> CC: Manuel Lauss <mano@roarinelk.homelinux.net> CC: Mike Frysinger <vapier.adi@gmail.com> CC: Arnaud Patard <apatard@mandriva.com> CC: Wan ZongShun <mcuos.com@gmail.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk>
2010-03-18 04:15:21 +08:00
}
/*
* snd_soc_dapm_shutdown - callback for system shutdown
*/
void snd_soc_dapm_shutdown(struct snd_soc_card *card)
{
struct snd_soc_dapm_context *dapm;
ASoC: multi-component - ASoC Multi-Component Support This patch extends the ASoC API to allow sound cards to have more than one CODEC and more than one platform DMA controller. This is achieved by dividing some current ASoC structures that contain both driver data and device data into structures that only either contain device data or driver data. i.e. struct snd_soc_codec ---> struct snd_soc_codec (device data) +-> struct snd_soc_codec_driver (driver data) struct snd_soc_platform ---> struct snd_soc_platform (device data) +-> struct snd_soc_platform_driver (driver data) struct snd_soc_dai ---> struct snd_soc_dai (device data) +-> struct snd_soc_dai_driver (driver data) struct snd_soc_device ---> deleted This now allows ASoC to be more tightly aligned with the Linux driver model and also means that every ASoC codec, platform and (platform) DAI is a kernel device. ASoC component private data is now stored as device private data. The ASoC sound card struct snd_soc_card has also been updated to store lists of it's components rather than a pointer to a codec and platform. The PCM runtime struct soc_pcm_runtime now has pointers to all its components. This patch adds DAPM support for ASoC multi-component and removes struct snd_soc_socdev from DAPM core. All DAPM calls are now made on a card, codec or runtime PCM level basis rather than using snd_soc_socdev. Other notable multi-component changes:- * Stream operations now de-reference less structures. * close_delayed work() now runs on a DAI basis rather than looping all DAIs in a card. * PM suspend()/resume() operations can now handle N CODECs and Platforms per sound card. * Added soc_bind_dai_link() to bind the component devices to the sound card. * Added soc_dai_link_probe() and soc_dai_link_remove() to probe and remove DAI link components. * sysfs entries can now be registered per component per card. * snd_soc_new_pcms() functionailty rolled into dai_link_probe(). * snd_soc_register_codec() now does all the codec list and mutex init. This patch changes the probe() and remove() of the CODEC drivers as follows:- o Make CODEC driver a platform driver o Moved all struct snd_soc_codec list, mutex, etc initialiasation to core. o Removed all static codec pointers (drivers now support > 1 codec dev) o snd_soc_register_pcms() now done by core. o snd_soc_register_dai() folded into snd_soc_register_codec(). CS4270 portions: Acked-by: Timur Tabi <timur@freescale.com> Some TLV320aic23 and Cirrus platform fixes. Signed-off-by: Ryan Mallon <ryan@bluewatersys.com> TI CODEC and OMAP fixes Signed-off-by: Peter Ujfalusi <peter.ujfalusi@nokia.com> Signed-off-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Samsung platform and misc fixes :- Signed-off-by: Chanwoo Choi <cw00.choi@samsung.com> Signed-off-by: Joonyoung Shim <jy0922.shim@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Reviewed-by: Jassi Brar <jassi.brar@samsung.com> Signed-off-by: Seungwhan Youn <sw.youn@samsung.com> MPC8610 and PPC fixes. Signed-off-by: Timur Tabi <timur@freescale.com> i.MX fixes and some core fixes. Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de> J4740 platform fixes:- Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> CC: Tony Lindgren <tony@atomide.com> CC: Nicolas Ferre <nicolas.ferre@atmel.com> CC: Kevin Hilman <khilman@deeprootsystems.com> CC: Sascha Hauer <s.hauer@pengutronix.de> CC: Atsushi Nemoto <anemo@mba.ocn.ne.jp> CC: Kuninori Morimoto <morimoto.kuninori@renesas.com> CC: Daniel Gloeckner <dg@emlix.com> CC: Manuel Lauss <mano@roarinelk.homelinux.net> CC: Mike Frysinger <vapier.adi@gmail.com> CC: Arnaud Patard <apatard@mandriva.com> CC: Wan ZongShun <mcuos.com@gmail.com> Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk>
2010-03-18 04:15:21 +08:00
list_for_each_entry(dapm, &card->dapm_list, list) {
if (dapm != &card->dapm) {
soc_dapm_shutdown_dapm(dapm);
if (dapm->bias_level == SND_SOC_BIAS_STANDBY)
snd_soc_dapm_set_bias_level(dapm,
SND_SOC_BIAS_OFF);
}
ASoC: Decouple DAPM from CODECs Decoupling Dynamic Audio Power Management (DAPM) from codec devices is required when developing ASoC further. Such as for other ASoC components to have DAPM widgets or when extending DAPM to handle cross-device paths. This patch decouples DAPM related variables from struct snd_soc_codec and moves them to new struct snd_soc_dapm_context that is used to encapsulate DAPM context of a device. ASoC core and API of DAPM functions are modified to use DAPM context instead of codec. This patch does not change current functionality and a large part of changes come because of structure and internal API changes. Core implementation is from Liam Girdwood <lrg@slimlogic.co.uk> with some minor core changes, codecs and machine driver conversions from Jarkko Nikula <jhnikula@gmail.com>. Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk> Signed-off-by: Jarkko Nikula <jhnikula@gmail.com> Cc: Nicolas Ferre <nicolas.ferre@atmel.com> Cc: Manuel Lauss <manuel.lauss@googlemail.com> Cc: Mike Frysinger <vapier.adi@gmail.com> Cc: Cliff Cai <cliff.cai@analog.com> Cc: Kevin Hilman <khilman@deeprootsystems.com> Cc: Ryan Mallon <ryan@bluewatersys.com> Cc: Timur Tabi <timur@freescale.com> Cc: Sascha Hauer <s.hauer@pengutronix.de> Cc: Lars-Peter Clausen <lars@metafoo.de> Cc: Arnaud Patard (Rtp) <arnaud.patard@rtp-net.org> Cc: Wan ZongShun <mcuos.com@gmail.com> Cc: Eric Miao <eric.y.miao@gmail.com> Cc: Jassi Brar <jassi.brar@samsung.com> Cc: Daniel Gloeckner <dg@emlix.com> Cc: Kuninori Morimoto <morimoto.kuninori@renesas.com> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2010-11-05 21:53:46 +08:00
}
soc_dapm_shutdown_dapm(&card->dapm);
if (card->dapm.bias_level == SND_SOC_BIAS_STANDBY)
snd_soc_dapm_set_bias_level(&card->dapm,
SND_SOC_BIAS_OFF);
}
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
/* Module information */
MODULE_AUTHOR("Liam Girdwood, lrg@slimlogic.co.uk");
[ALSA] ASoC: dynamic audio power management (DAPM) This patch adds Dynamic Audio Power Management (DAPM) to ASoC. Dynamic Audio Power Management (DAPM) is designed to allow portable and handheld Linux devices to use the minimum amount of power within the audio subsystem at all times. It is independent of other kernel PM and as such, can easily co-exist with the other PM systems. DAPM is also completely transparent to all user space applications as all power switching is done within the ASoC core. No code changes or recompiling are required for user space applications. DAPM makes power switching decisions based upon any audio stream (capture/playback) activity and audio mixer settings within the device. DAPM spans the whole machine. It covers power control within the entire audio subsystem, this includes internal codec power blocks and machine level power systems. There are 4 power domains within DAPM:- 1. Codec domain - VREF, VMID (core codec and audio power) Usually controlled at codec probe/remove and suspend/resume, although can be set at stream time if power is not needed for sidetone, etc. 2. Platform/Machine domain - physically connected inputs and outputs Is platform/machine and user action specific, is configured by the machine driver and responds to asynchronous events e.g when HP are inserted 3. Path domain - audio subsystem signal paths Automatically set when mixer and mux settings are changed by the user. e.g. alsamixer, amixer. 4. Stream domain - DAC's and ADC's. Enabled and disabled when stream playback/capture is started and stopped respectively. e.g. aplay, arecord. All DAPM power switching decisions are made automatically by consulting an audio routing map of the whole machine. This map is specific to each machine and consists of the interconnections between every audio component (including internal codec components). Signed-off-by: Richard Purdie <rpurdie@rpsys.net> Signed-off-by: Liam Girdwood <liam.girdwood@wolfsonmicro.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Signed-off-by: Jaroslav Kysela <perex@suse.cz>
2006-10-07 00:32:18 +08:00
MODULE_DESCRIPTION("Dynamic Audio Power Management core for ALSA SoC");
MODULE_LICENSE("GPL");