This is re-applying the fix that went into 5.6 (commit 6c89ffea)
as the changes were wiped out after merging the other code
refactoring.  Basically the same changes, just replacing the
suspicious calls of snprintf() with scnprintf().
Signed-off-by: NTakashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20200310163625.10838-1-tiwai@suse.deSigned-off-by: NMark Brown <broonie@kernel.org>

This patch switch to use plural form macro.

	- for_each_rtd_codec_dai()
	+ for_each_rtd_codec_dais()

	- for_each_rtd_codec_dai_rollback()
	+ for_each_rtd_codec_dais_rollback()

	- for_each_rtd_cpu_dai()
	+ for_each_rtd_cpu_dais()

	- for_each_rtd_cpu_dai_rollback()
	+ for_each_rtd_cpu_dais_rollback()
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/87v9negogr.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

Digital Mute for CPU   is done at soc_pcm_close(), and
Digital Mute for Codec is done at soc_pcm_hw_free().
It is just confusable.
This patch do Digital Mute for both CPU/Codec in same timing.
Then, it cares DAI activity
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87imjip9ty.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

soc_pcm_hw_free() want to call snd_soc_dai_digital_mute()
if it was last user of Playback or Capture.

	bool playback = substream->stream == SNDRV_PCM_STREAM_PLAYBACK;
	int playback_active = dai->stream_active[SNDRV_PCM_STREAM_PLAYBACK];
	int capture_active  = dai->stream_active[SNDRV_PCM_STREAM_CAPTURE];

	if ((playback  && playback_active == 1) ||
	    (!playback && capture_active  == 1))
		snd_soc_dai_digital_mute(...)

But it is same as

	int active = dai->stream_active[substream->stream];

	if (active == 1)
		snd_soc_dai_digital_mute(...)

This patch simplify the code.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87k13yp9ub.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

error handling at dpcm_fe_dai_startup() has duplicate code.
This patch tidyup it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/87lfoep9ug.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

We already have snd_soc_dai_get_pcm_stream(),
let's use it
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/87mu8up9ul.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

dpcm_fe_dai_close() and error case of dpcm_fe_dai_open()
need to do same cleanup operation.
To avoid duplicate code, this patch adds dpcm_fe_dai_cleanup()
and use it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/87o8tap9uq.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

move dpcm_fe_dai_close() next to dpcm_fe_dai_open().
This is prepare for dpcm_fe_dai_open() cleanup
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/87pndqp9uv.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

The logic to calculate the subset of stream parameters supported by all
DAIs associated with a PCM stream is nontrivial. Export a helper
function so it can be used to set up simple codec2codec DAI links.
Signed-off-by: NSamuel Holland <samuel@sholland.org>
Link: https://lore.kernel.org/r/20200305051143.60691-3-samuel@sholland.orgSigned-off-by: NMark Brown <broonie@kernel.org>

dpcm_add_paths() checks returned be from dpcm_get_be()

	static int dpcm_add_paths(...)
	{
		...
		for_each_dapm_widgets(list, i, widget) {
			...
			be = dpcm_get_be(...);
			...
			/* make sure BE is a real BE */
=>			if (!be->dai_link->no_pcm)
				continue;
			...
		}
		...
	}

But, dpcm_get_be() itself is checking it already.

	dpcm_get_be(...)
	{
		...
		for_each_card_rtds(card, be) {
=>			if (!be->dai_link->no_pcm)
				continue;
			...
			if (...)
=>				return be;
		}

		return NULL
	}

This patch removes duplicate check
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/87lfoo7q1j.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

Many functions defines "stream = substream->stream", but
some of them is using "substream->stream" instead of "stream".
It is pointless. This patch uses defined stream.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/87mu947q1t.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

Now multi-cpu-dais are supported, we can skip cpi-dais which don't
support the current stream, following the example of multi-codec-dais.
Signed-off-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Link: https://lore.kernel.org/r/20200225133917.21314-7-yung-chuan.liao@linux.intel.comSigned-off-by: NMark Brown <broonie@kernel.org>

Multi cpu is not supported by all functions yet. Add an error message
and return.
Suggested-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Link: https://lore.kernel.org/r/20200225133917.21314-6-yung-chuan.liao@linux.intel.comSigned-off-by: NMark Brown <broonie@kernel.org>

Add support in PCM operations to invoke multiple cpu dais as we do
for multiple codec dais. Also the symmetry calculations are updated to
reflect multiple cpu dais.
Signed-off-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: NBard Liao <yung-chuan.liao@linux.intel.com>
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NShreyas NC <shreyas.nc@intel.com>
Signed-off-by: NVinod Koul <vkoul@kernel.org>
Link: https://lore.kernel.org/r/20200225133917.21314-3-yung-chuan.liao@linux.intel.comSigned-off-by: NMark Brown <broonie@kernel.org>

ASoC component open/close and snd_soc_component_module_get/put are called
independently for each component-substream pair, so the logic added in
commit dd03907b ("ASoC: soc-pcm: call snd_soc_component_open/close()
once") was not sufficient and led to PCM playback and module unload errors.

Implement handling of failures directly in soc_pcm_components_open(),
so that any successfully opened components are closed upon error with
other components. This allows to clean up error handling in
soc_pcm_open() without adding more state tracking.

Fixes: dd03907b ("ASoC: soc-pcm: call snd_soc_component_open/close() once")
Signed-off-by: NKai Vehmanen <kai.vehmanen@linux.intel.com>
Tested-by: NDmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/20200220094955.16968-1-kai.vehmanen@linux.intel.comSigned-off-by: NMark Brown <broonie@kernel.org>

soc-pcm.c has implementation which depends on CONFIG_DEBUG_FS.
But, we don't want to have random #ifdef.
This patch adds dpcm_create/remove_debugfs_state() and care it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/877e0jjc9b.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

This is prepare for CONFIG_DEBUG_FS cleanup
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/878skzjc9k.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

dpcm_path_put() (A) is calling kfree(*list).
The freed list is created by dapm_widget_list_create() (B) which is called
from snd_soc_dapm_dai_get_connected_widgets() (C) which is called from
dpcm_path_get() (D).

(B)	dapm_widget_list_create(**list, ...)
	{
		...
=>		*list = kzalloc();
		...
	}

(C)	snd_soc_dapm_dai_get_connected_widgets(..., **list, ...)
	{
		...
		dapm_widget_list_create(list, ...);
		...
	}

(D)	dpcm_path_get(..., **list)
	{
		...
		snd_soc_dapm_dai_get_connected_widgets(..., list, ...);
		...
	}

(A)	dpcm_path_put(**list)
	{
=>		kfree(*list);
	}

This kind of unbalance code is very difficult to read/understand.
To avoid this issue, this patch adds each missing paired function
dapm_widget_list_free()         for dapm_widget_list_create() (B), and
snd_soc_dapm_dai_free_widgets() for snd_soc_dapm_dai_get_connected_widgets() (C).

This patch uses these, and moves dpcm_path_put() next to dpcm_path_get().
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87a75fjc9q.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

soc-pcm has dpcm_run_new/old_update(), but these are used from
dpcm_fe_runtime_update() only, and are very verbose functions.
This patch disassembles these.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87blpvjc9v.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

soc-pcm.c has dai_get_widget(), but it can be more generic.
This patch renames it to snd_soc_dai_get_widget(), and use it.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87d0abjca1.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

soc_pcm_apply_msb() apply msb for CPU/Codec,
but, it has duplicate code. The difference is only
SNDRV_PCM_STREAM_PLAYBACK and SNDRV_PCM_STEAM_CAPTURE.

It is very verbose and duplicate code.
This patch simplify code by using snd_soc_dai_get_pcm_stream().
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87eeurjca6.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

DAI driver has playback/capture stream.
OTOH, we have SNDRV_PCM_STREAM_PLAYBACK/CAPTURE.
Because of this kind of implementation,
ALSA SoC needs to have many verbose code.

To solve this issue, this patch adds snd_soc_dai_get_pcm_stream() macro
to get playback/capture stream pointer from stream.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87ftf7jcab.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

commit b0edff42 ("ASoC: soc-pcm/soc-compress: use
snd_soc_dapm_stream_stop() for SND_SOC_DAPM_STREAM_STOP")
uses snd_soc_dapm_stream_stop() for soc_compr_free_fe()
and dpcm_fe_dai_shutdown() because it didn't care about pmdown_time.
But, it didn't need to care.
This patch rollback to original code.
Some system will wait unneeded timed-out without this patch.
Special Thanks for reporting to Chris Gorman.
	...
	intel_sst_acpi 808622A8:00: Wait timed-out condition:0x0, msg_id:0x1 fw_state 0x3
	intel_sst_acpi 808622A8:00: fw returned err -16
	sst-mfld-platform sst-mfld-platform: ASoC: PRE_PMD: pcm0_in event failed: -16
	...

Fixes: commit b0edff42 ("ASoC: soc-pcm/soc-compress: use snd_soc_dapm_stream_stop() for SND_SOC_DAPM_STREAM_STOP")
Reported-by: NChris Gorman <chrisjohgorman@gmail.com>
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87lfowspeb.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

Now that dpcm_get_be() is used in dpcm_end_walk_at_be(), it is not a error
if this function does not find a BE for the provided widget. Remove the
related dev_err() trace which is confusing since things might be working
as expected.

When called from dpcm_add_paths(), it is an error if dpcm_get_be() fails to
find a BE for the provided widget. The necessary error trace is already
done in this case.

Fixes: 027a4838 ("ASoC: soc-pcm: use dpcm_get_be() at dpcm_end_walk_at_be()")
Signed-off-by: NJerome Brunet <jbrunet@baylibre.com>
Tested-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Acked-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Cc: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/20200219115048.934678-1-jbrunet@baylibre.comSigned-off-by: NMark Brown <broonie@kernel.org>

dpcm_show_state() invokes multiple snprintf() calls to concatenate
formatted strings on the fixed size buffer.  The usage of snprintf()
is supposed for avoiding the buffer overflow, but it doesn't work as
expected because snprintf() doesn't return the actual output size but
the size to be written.

Fix this bug by replacing all snprintf() calls with scnprintf()
calls.

Fixes: f86dcef8 ("ASoC: dpcm: Add debugFS support for DPCM")
Signed-off-by: NTakashi Iwai <tiwai@suse.de>
Acked-by: NCezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20200218111737.14193-4-tiwai@suse.deSigned-off-by: NMark Brown <broonie@kernel.org>

Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/871rqtboth.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

DAI has playback_active and capture_active to care usage count.
OTOH, we have SNDRV_PCM_STREAM_PLAYBACK/CAPTURE.
But because of this kind of implementation mismatch,
ALSA SoC has many verbose code.

To solve this issue, this patch merge playback_active/capture_active
into stream_active[2];
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/875zg5botu.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

When we use some kind of lock, we need to do unlock.
In that time, multi unlock/return is not good implementation.
This patch add label and use goto at dpcm_fe_dai_open()
to reduce such code.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/877e0lboty.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

Below functions are doing very similar things, the difference is
used state only.

	snd_soc_dpcm_can_be_free_stop()
	snd_soc_dpcm_can_be_params()

This patch adds common snd_soc_dpcm_check_state(), and use it from
snd_soc_dpcm_can_be_free_stop() / snd_soc_dpcm_can_be_params().
It can reduce duplicate code.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/878sl1bou2.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

No one is using snd_soc_dpcm_be_get/set_state().
If it exists only by assumption that "it may be necessary someday",
let's remove it now. Otherwise code maintenance will be difficult.
We can revive it when we really needed it.
Let's remove it, so far.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/87a75hbou7.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

No one is using soc_dpcm_be_digital_mute().
If it exists only by assumption that "it may be necessary someday",
let's remove it now. Otherwise code maintenance will be difficult.
We can revive it when we really needed it.
Let's remove it, so far.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/87blpxbouc.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

dpcm_end_walk_at_be() and dpcm_get_be() are almost same code.
This patch uses dpcm_get_be() from dpcm_end_walk_at_be().
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/87d0adbouh.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

dpcm_end_walk_at_be() has very duplicate code.

	dpcm_end_walk_at_be() {
		...
		if (stream == SNDRV_PCM_STREAM_PLAYBACK) {
(1)			/* code for Playback */
		} else  {
(2)			/* code for Capture */
		}
	}

The difference between Playback (1) and Capture (2) code is pointer only
(= "playback_widget" or "caputre_widget").
OTOH, now we already has dai_get_widget() for it.
This means we can merge (1) and (2).
This patch do it and remove duplicated code.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/87eeutboul.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

dpcm_get_be() has very duplicate code.

	dpcm_get_be() {
		...
		if (stream == SNDRV_PCM_STREAM_PLAYBACK) {
(1)			/* code for Playback */
		} else  {
(2)			/* code for Capture */
		}
	}

The difference between Playback (1) and Capture (2) code is pointer only
(= "playback_widget" or "caputre_widget").
OTOH, now we already has dai_get_widget() for it.
This means we can merge (1) and (2).
This patch do it and remove duplicated code.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/87ftf9bouq.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

This patch moves dai_get_widget() to top side.
This is prepare for cleanup soc-pcm.c
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/87h7zpbouu.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

Commit af4bac11 ("ASoC: soc-pcm: crash in snd_soc_dapm_new_dai")
swapped the SNDRV_PCM_STREAM_* parameter in the
snd_soc_dai_stream_valid(cpu_dai, ...) checks. But that works only
for codec2codec links. For normal links it breaks registration of
playback/capture-only PCM devices.

E.g. on qcom/apq8016_sbc there is usually one playback-only and one
capture-only PCM device, but they disappeared after the commit.

The codec2codec case was added in commit a342031c
("ASoC: create pcm for codec2codec links as well") as an extra check
(e.g. `playback = playback && cpu_playback->channels_min`).

We should be able to simplify the code by checking directly for
the correct stream type in the loop.
This also fixes the regression because we check for PLAYBACK for
both codec and cpu dai again when codec2codec is not used.

Fixes: af4bac11 ("ASoC: soc-pcm: crash in snd_soc_dapm_new_dai")
Signed-off-by: NStephan Gerhold <stephan@gerhold.net>
Tested-by: NJerome Brunet <jbrunet@baylibre.com>
Reviewed-by: NJerome Brunet <jbrunet@baylibre.com>
Cc: Jerome Brunet <jbrunet@baylibre.com>
Cc: Sameer Pujar <spujar@nvidia.com>
Link: https://lore.kernel.org/r/20200218103824.26708-1-stephan@gerhold.netSigned-off-by: NMark Brown <broonie@kernel.org>

soc_pcm_open() operation order is not good.
At first, soc_pcm_open() operation order is

	1) CPU DAI startup
	2) Component open
	3) Codec DAI startup
	4) rtd startup

But here, 2) will call try_module_get() if component has
module_get_upon_open flags. This means 1) CPU DAI startup
will be operated *before* its module was loaded.
DAI should be called *after* Component.

Second, soc_pcm_close() operation order is
	1) CPU DAI shutdown
	2) Codec DAI shutdown
	3) rtd shutdown
	4) Component close

soc_pcm_open() and soc_pcm_close() are paired function,
but, its operation order is unbalance.
This patch tidyup soc_pcm_open() order to Component -> rtd -> DAI.
This is one of prepare for cleanup soc-pcm-open()
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/871rr3cext.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

This patch moves soc_pcm_close() next to soc_pcm_open().
This is prepare for soc_pcm_open() cleanup.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NRanjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/8736bjcexx.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

Current soc_pcm_open() calls snd_soc_component_open() under loop.
Thus, it needs to care about opened/not-yet-opened Component.

But, if soc-component.c is handling it, soc-pcm.c don't need to care
about it.
This patch adds opened flag to soc-component.h, and simplify soc-pcm.c.
This is one of prepare for cleanup soc-pcm-open()
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Reviewed-by: NPierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/874kvzcey1.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>

Current soc_pcm_open() calls snd_soc_dai_startup() under loop.
Thus, it needs to care about started/not-yet-started codec DAI.

But, if soc-dai.c is handling it, soc-pcm.c don't need to care
about it.
This patch adds started flag to soc-dai.h, and simplify soc-pcm.c.
This is one of prepare for cleanup soc-pcm-open()
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/875zgfcey5.wl-kuninori.morimoto.gx@renesas.comSigned-off-by: NMark Brown <broonie@kernel.org>