platform/kernel/linux-starfive.git
5 years agoMerge branch 'asoc-5.3' into asoc-5.4
Mark Brown [Fri, 26 Jul 2019 12:10:09 +0000 (13:10 +0100)]
Merge branch 'asoc-5.3' into asoc-5.4

5 years agoASoC: ti: davinci-mcasp: Correct slot_width posed constraint
Peter Ujfalusi [Fri, 26 Jul 2019 06:42:43 +0000 (09:42 +0300)]
ASoC: ti: davinci-mcasp: Correct slot_width posed constraint

The slot_width is a property for the bus while the constraint for
SNDRV_PCM_HW_PARAM_SAMPLE_BITS is for the in memory format.

Applying slot_width constraint to sample_bits works most of the time, but
it will blacklist valid formats in some cases.

With slot_width 24 we can support S24_3LE and S24_LE formats as they both
look the same on the bus, but a a 24 constraint on sample_bits would not
allow S24_LE as it is stored in 32bits in memory.

Implement a simple hw_rule function to allow all formats which require less
or equal number of bits on the bus as slot_width (if configured).

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Link: https://lore.kernel.org/r/20190726064244.3762-2-peter.ujfalusi@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: rockchip: Fix mono capture
Cheng-Yi Chiang [Fri, 26 Jul 2019 04:42:02 +0000 (12:42 +0800)]
ASoC: rockchip: Fix mono capture

This reverts commit db51707b9c9aeedd310ebce60f15d5bb006567e0.
Revert "ASoC: rockchip: i2s: Support mono capture"

Previous discussion in

https://patchwork.kernel.org/patch/10147153/

explains the issue of the patch.
While device is configured as 1-ch, hardware is still
generating a 2-ch stream.
When user space reads the data and assumes it is a 1-ch stream,
the rate will be slower by 2x.

Revert the change so 1-ch is not supported.
User space can selectively take one channel data out of two channel
if 1-ch is preferred.
Currently, both channels record identical data.

Signed-off-by: Cheng-Yi Chiang <cychiang@chromium.org>
Link: https://lore.kernel.org/r/20190726044202.26866-1-cychiang@chromium.org
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Intel: Fix some acpi vs apci typo in somme comments
Christophe JAILLET [Thu, 25 Jul 2019 05:35:23 +0000 (07:35 +0200)]
ASoC: Intel: Fix some acpi vs apci typo in somme comments

Fix some typo to have the filaname given in a comment match the real name
of the file.
Some 'acpi' have erroneously been written 'apci'

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/20190725053523.16542-1-christophe.jaillet@wanadoo.fr
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: ti: davinci-mcasp: Fix clk PDIR handling for i2s master mode
Peter Ujfalusi [Thu, 25 Jul 2019 08:34:23 +0000 (11:34 +0300)]
ASoC: ti: davinci-mcasp: Fix clk PDIR handling for i2s master mode

When running McASP as master capture alone will not record any audio unless
a parallel playback stream is running. As soon as the playback stops the
captured data is going to be silent again.

In McASP master mode we need to set the PDIR for the clock pins and fix
the mcasp_set_axr_pdir() to skip the bits in the PDIR registers above
AMUTE.

This went unnoticed as most of the boards uses McASP as slave and neither
of these issues are visible (audible) in those setups.

Fixes: ca3d9433349e ("ASoC: davinci-mcasp: Update PDIR (pin direction) register handling")
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Link: https://lore.kernel.org/r/20190725083423.7321-1-peter.ujfalusi@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: codec2codec: deal with params when necessary
Jerome Brunet [Thu, 25 Jul 2019 16:59:46 +0000 (18:59 +0200)]
ASoC: codec2codec: deal with params when necessary

When there is an event on codec to codec dai_link, we only need to deal
with params if the event is SND_SOC_DAPM_PRE_PMU, when .hw_params() is
called. For the other events, it is useless.

Also, dealing with the codec to codec params just before calling
.hw_params() callbacks give change to either party on the link to alter
params content in .startup(), which might be useful in some cases

Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20190725165949.29699-4-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: codec2codec: name link using stream direction
Jerome Brunet [Thu, 25 Jul 2019 16:59:45 +0000 (18:59 +0200)]
ASoC: codec2codec: name link using stream direction

At the moment, codec to codec dai link widgets are named after the
cpu dai and the 1st codec valid on the link. This might be confusing
if there is multiple valid codecs on the link for one stream
direction.

Instead, use the dai link name and the stream direction to name the
the dai link widget

Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20190725165949.29699-3-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: codec2codec: run callbacks in order
Jerome Brunet [Thu, 25 Jul 2019 16:59:44 +0000 (18:59 +0200)]
ASoC: codec2codec: run callbacks in order

When handling dai_link events on codec to codec links, run all .startup()
callbacks on sinks and sources before running any .hw_params(). Same goes
for hw_free() and shutdown(). This is closer to the behavior of regular
dai links

Signed-off-by: Jerome Brunet <jbrunet@baylibre.com>
Link: https://lore.kernel.org/r/20190725165949.29699-2-jbrunet@baylibre.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: cs47l92: Add codec driver for Cirrus Logic CS47L92
Stuart Henderson [Thu, 25 Jul 2019 16:39:31 +0000 (17:39 +0100)]
ASoC: cs47l92: Add codec driver for Cirrus Logic CS47L92

Adds the codec driver for the CS47L92 SmartCodec. This is a
multi-functional codec based on the Cirrus Logic Madera platform.

Signed-off-by: Stuart Henderson <stuarth@opensource.wolfsonmicro.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20190725163931.24964-3-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: cs47l15: Add codec driver for Cirrus Logic CS47L15
Richard Fitzgerald [Thu, 25 Jul 2019 16:39:30 +0000 (17:39 +0100)]
ASoC: cs47l15: Add codec driver for Cirrus Logic CS47L15

Adds the codec driver for the CS47L15 SmartCodec. This is a
multi-functional codec based on the Cirrus Logic Madera platform.

Signed-off-by: Richard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: Jaswinder Jassal <jjassal@opensource.wolfsonmicro.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20190725163931.24964-2-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: wm_adsp: Allow bus error handler to be called directly
Charles Keepax [Thu, 25 Jul 2019 16:39:29 +0000 (17:39 +0100)]
ASoC: wm_adsp: Allow bus error handler to be called directly

There is no need for end drivers to add helper functions to allow the
bus error handler to be called, simply update the prototype so it can be
called directly.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20190725163931.24964-1-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Makefile - fix the top-level kernel module names (add snd- prefix)
Jaroslav Kysela [Thu, 25 Jul 2019 13:37:43 +0000 (15:37 +0200)]
ASoC: SOF: Makefile - fix the top-level kernel module names (add snd- prefix)

Use the proper module name. The objs assignments are already there.

Signed-off-by: Jaroslav Kysela <perex@perex.cz>
Cc: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Cc: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Cc: Takashi Iwai <tiwai@suse.de>
Cc: Mark Brown <broonie@kernel.org>0
Link: https://lore.kernel.org/r/20190725133743.22145-1-perex@perex.cz
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: ti: davinci-mcasp: Improve serializer handling in multi AXR setups
Peter Ujfalusi [Thu, 25 Jul 2019 08:34:32 +0000 (11:34 +0300)]
ASoC: ti: davinci-mcasp: Improve serializer handling in multi AXR setups

When multiple serializers are used we need to track the number of
serializers used by the other stream direction to avoid killing data lines
when the first stream used more serializers than the second would need.
We are still protected against the case when the second stream uses more
serializers which had affected the running stream as well.

To take advantage of the improved serializer logic we need to modify the
channel constraints rule as well to allow the use of multiple serializers
for the second stream as additional ones will not affect the FS/BCLK on
the bus.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Link: https://lore.kernel.org/r/20190725083432.7419-1-peter.ujfalusi@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: ti: davinci-mcasp: Add support for RIGHT_J format
Peter Ujfalusi [Thu, 25 Jul 2019 08:34:11 +0000 (11:34 +0300)]
ASoC: ti: davinci-mcasp: Add support for RIGHT_J format

The formater unit's rotation needs to be programmed differently for right
aligned bus format to have the data moved to the correct place.

Take the opportunity and simplify the formater unit setup code.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Link: https://lore.kernel.org/r/20190725083411.7211-1-peter.ujfalusi@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: pcm3168a: Allow all channels in case of parallel DIN/DOUT setup
Peter Ujfalusi [Thu, 25 Jul 2019 08:33:21 +0000 (11:33 +0300)]
ASoC: pcm3168a: Allow all channels in case of parallel DIN/DOUT setup

If multi DIN/DOUT mode is selected (tdm_slots == 2) then configure the
channel constraint to allow all channels.

Signed-off-by: Peter Ujfalusi <peter.ujfalusi@ti.com>
Link: https://lore.kernel.org/r/20190725083321.6776-1-peter.ujfalusi@ti.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Intel: Skylake: Cleanup skl_module_cfg declaration
Cezary Rojewski [Tue, 23 Jul 2019 14:58:54 +0000 (16:58 +0200)]
ASoC: Intel: Skylake: Cleanup skl_module_cfg declaration

With removal of MCPS, CPS and CPC ambiguity, ibs and obs params for
struct skl_module_cfg have been left unused. Update struct declaration
by removing these two.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20190723145854.8527-8-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Intel: Skylake: Make MCPS and CPS params obsolete
Cezary Rojewski [Tue, 23 Jul 2019 14:58:53 +0000 (16:58 +0200)]
ASoC: Intel: Skylake: Make MCPS and CPS params obsolete

As per FW Interface Modules Configuration, init instance IPC request
requires base initial module configuration. This configuration structure
is made of:
- cpc (chunks per cycle)
- ibs (input buffer size)
- obs (output buffer size)
- is_pages (memory pages required)
- audio_fmt (self explanatory)

Skylake topology accepts following tokens: MCPS, CPS and CPC. All of
these are directly connected. Moreover, assigning one of these allows
to calculate the remaining two. In simplest scenario and assuming 1ms
scheduling, following is true:

CPS = CPC times 1000
MCPS = CPS times 1000 000
Note: these calculations vary depending on scenario and scheduling
requirements.

Given the current implementation, userspace is allowed to provide
different values for all three causing informational chaos. On top of
that, struct skl_base_cfg which represents base module configuration,
incorrectly takes CPS param instead of CPC.

This ambiguity may lead to user unintentionally providing improper
values to DSP firmware and thus impacting module scheduling in
unexpected fashion. Fix by making MCPS and CPS topology params obsolete
and relying solely on CPC value.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20190723145854.8527-7-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Intel: Skylake: Do not disable FW notifications
Amadeusz Sławiński [Tue, 23 Jul 2019 14:58:52 +0000 (16:58 +0200)]
ASoC: Intel: Skylake: Do not disable FW notifications

As per FW team recommendation we should not disable notifications.
By default, all notifications are enabled in DSP firmware. These
notifications provide a vital information whenever an error occurs.

Currently, driver disables them during boot sequences. By doing so,
Skylake may silently ignore severe stream errors.
Correct that by removing permissive code.

Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@intel.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20190723145854.8527-6-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Intel: Skylake: Remove memory available check
Cezary Rojewski [Tue, 23 Jul 2019 14:58:51 +0000 (16:58 +0200)]
ASoC: Intel: Skylake: Remove memory available check

Current memory availability check is a stub, while actual memory
management takes place in firmware. Leave this task to firmware entirely
and remove redundant code.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20190723145854.8527-5-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Intel: Skylake: Remove MCPS available check
Cezary Rojewski [Tue, 23 Jul 2019 14:58:50 +0000 (16:58 +0200)]
ASoC: Intel: Skylake: Remove MCPS available check

The entire logic for MCPS calculation and DSP scheduling is found
within DSP firmware. Currently driver implements simplistic, inaccurate
logic itself which may prevent pipeline creation despite firmware being
completely fine its parameters.

Remove that logic and leave the MCPS calculation to DSP alone.

Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20190723145854.8527-4-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Intel: Skylake: Combine snd_soc_skl_ipc and snd_soc_skl
Amadeusz Sławiński [Tue, 23 Jul 2019 14:58:49 +0000 (16:58 +0200)]
ASoC: Intel: Skylake: Combine snd_soc_skl_ipc and snd_soc_skl

As both modules are core part of Skylake driver and none can live
without the other, combine snd_soc_skl_ipc and snd_soc_skl.

It's highly probable IPC module was to be treated as an interface for
platform specific code implementations e.g.: possibility of existence of
BXT specific code without SKL one. However, most funtionalities are
being inherited from one DSP firmware to another, and thus this
assumption fails.

skl-sst, bxt-sst and cnl-sst are not individuals pointing respectively
to SKL (cAVS 1.5), BXT (cAVS 1.5+) & CNL (cAVS 1.8) standalone
implementations. Code found within these is shared among all platforms
whenever necessary to avoid code duplication and reduce development
burden.

Merge also helps in cleaning up internal code in future changes.

Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@intel.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20190723145854.8527-3-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Intel: Skylake: Merge skl_sst and skl into skl_dev struct
Cezary Rojewski [Tue, 23 Jul 2019 14:58:48 +0000 (16:58 +0200)]
ASoC: Intel: Skylake: Merge skl_sst and skl into skl_dev struct

Skylake driver is divided into two modules:
- snd_soc_skl
- snd_soc_skl_ipc

and nothing would be wrong if not for the fact that both cannot exist
without one another. IPC module is not some kind of extension, as it is
the case for snd_hda_ext_core which is separated from snd_hda_core -
legacy hda interface. It's as much core Skylake module as snd_soc_skl
is.

Statement backed up by existence of circular dependency between this
two. To eliminate said problem, struct skl_sst has been created. From
that very momment, Skylake has been plagued by header errors (incomplete
structs, unknown references etc.) whenever something new is to be added
or code is cleaned up.

As this design is being corrected, struct skl_sst is no longer needed,
so combine it with struct skl. To avoid ambiguity when searching for skl
stuff (struct skl *skl) it has also been renamed to skl_dev.

No functional changes.

Signed-off-by: Piotr Maziarz <piotrx.maziarz@intel.com>
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20190723145854.8527-2-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: wm8955: Fix a typo in 'wm8995_pll_factors()' function name
Christophe JAILLET [Wed, 24 Jul 2019 05:26:32 +0000 (07:26 +0200)]
ASoC: wm8955: Fix a typo in 'wm8995_pll_factors()' function name

This should be 'wm8955_pll_factors()' instead.
Fix it and use it.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20190724052632.30476-1-christophe.jaillet@wanadoo.fr
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Fail card instantiation if DAI format setup fails
Ricard Wanderlof [Wed, 24 Jul 2019 09:38:44 +0000 (11:38 +0200)]
ASoC: Fail card instantiation if DAI format setup fails

If the DAI format setup fails, there is no valid communication format
between CPU and CODEC, so fail card instantiation, rather than continue
with a card that will most likely not function properly.

Signed-off-by: Ricard Wanderlof <ricardw@axis.com>
Link: https://lore.kernel.org/r/alpine.DEB.2.20.1907241132350.6338@lnxricardw1.se.axis.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: move snd_soc_dai_stream_valid() to soc-dai.c
Kuninori Morimoto [Mon, 22 Jul 2019 01:36:16 +0000 (10:36 +0900)]
ASoC: soc-dai: move snd_soc_dai_stream_valid() to soc-dai.c

snd_soc_dai_stream_valid() is function to check stream validity.
But, some code is using it, some code are checking stream->channels_min
directly. Doing samethings by different method is confusable.
This patch uses same funcntion for same purpose.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87ftmyhmzz.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-core: use existing snd_soc_dai_digital_mute()
Kuninori Morimoto [Mon, 22 Jul 2019 01:36:27 +0000 (10:36 +0900)]
ASoC: soc-core: use existing snd_soc_dai_digital_mute()

ALSA SoC already has snd_soc_dai_digital_mute() for digital_mute,
let's use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87ef2ihmzo.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_compress_new()
Kuninori Morimoto [Mon, 22 Jul 2019 01:35:29 +0000 (10:35 +0900)]
ASoC: soc-dai: add snd_soc_dai_compress_new()

Current ALSA SoC is directly using dai->driver->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_compress_new() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87h87ehn1a.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_remove()
Kuninori Morimoto [Mon, 22 Jul 2019 01:35:05 +0000 (10:35 +0900)]
ASoC: soc-dai: add snd_soc_dai_remove()

Current ALSA SoC is directly using dai->driver->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_remvoe() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87imruhn1x.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_probe()
Kuninori Morimoto [Mon, 22 Jul 2019 01:34:56 +0000 (10:34 +0900)]
ASoC: soc-dai: add snd_soc_dai_probe()

Current ALSA SoC is directly using dai->driver->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_probe() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87k1cahn26.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_resume()
Kuninori Morimoto [Mon, 22 Jul 2019 01:34:43 +0000 (10:34 +0900)]
ASoC: soc-dai: add snd_soc_dai_resume()

Current ALSA SoC is directly using dai->driver->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_resume() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87lfwqhn2j.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_suspend()
Kuninori Morimoto [Mon, 22 Jul 2019 01:34:29 +0000 (10:34 +0900)]
ASoC: soc-dai: add snd_soc_dai_suspend()

Current ALSA SoC is directly using dai->driver->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_suspend() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87muh6hn2x.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_delay()
Kuninori Morimoto [Mon, 22 Jul 2019 01:34:09 +0000 (10:34 +0900)]
ASoC: soc-dai: add snd_soc_dai_delay()

Current ALSA SoC is directly using dai->driver->ops->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_delay() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87o91mhn3i.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_bespoke_trigger()
Kuninori Morimoto [Mon, 22 Jul 2019 01:33:56 +0000 (10:33 +0900)]
ASoC: soc-dai: add snd_soc_dai_bespoke_trigger()

Current ALSA SoC is directly using dai->driver->ops->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_bespoke_trigger() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87r26ihn3u.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_trigger()
Kuninori Morimoto [Mon, 22 Jul 2019 01:33:51 +0000 (10:33 +0900)]
ASoC: soc-dai: add snd_soc_dai_trigger()

Current ALSA SoC is directly using dai->driver->ops->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_trigger() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87sgqyhn40.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_prepare()
Kuninori Morimoto [Mon, 22 Jul 2019 01:33:45 +0000 (10:33 +0900)]
ASoC: soc-dai: add snd_soc_dai_prepare()

Current ALSA SoC is directly using dai->driver->ops->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_prepare() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87tvbehn46.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_shutdown()
Kuninori Morimoto [Mon, 22 Jul 2019 01:33:39 +0000 (10:33 +0900)]
ASoC: soc-dai: add snd_soc_dai_shutdown()

Current ALSA SoC is directly using dai->driver->ops->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_shutdown() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87v9vuhn4b.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_startup()
Kuninori Morimoto [Mon, 22 Jul 2019 01:33:32 +0000 (10:33 +0900)]
ASoC: soc-dai: add snd_soc_dai_startup()

Current ALSA SoC is directly using dai->driver->ops->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_startup() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87wogahn4i.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: add snd_soc_dai_hw_free()
Kuninori Morimoto [Mon, 22 Jul 2019 01:33:19 +0000 (10:33 +0900)]
ASoC: soc-dai: add snd_soc_dai_hw_free()

Current ALSA SoC is directly using dai->driver->ops->xxx,
thus, it has deep nested bracket, and it makes code unreadable.
This patch adds new snd_soc_dai_hw_free() and use it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87y30qhn4w.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: soc-dai: mv soc_dai_hw_params() to soc-dai
Kuninori Morimoto [Mon, 22 Jul 2019 01:33:04 +0000 (10:33 +0900)]
ASoC: soc-dai: mv soc_dai_hw_params() to soc-dai

Sometimes ALSA SoC naming is very random.
Current soc_dai_hw_params() should use snd_soc_dai_xxx() style.
And then, 1st parameter should be dai. Otherwise it is confusable.
 - soc_dai_hw_params(..., dai);
 + snd_soc_dai_hw_params(dai, ...);

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87zhl6hn5b.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: add soc-dai.c
Kuninori Morimoto [Mon, 22 Jul 2019 01:32:12 +0000 (10:32 +0900)]
ASoC: add soc-dai.c

Current ALSA SoC has many snd_soc_dai_xxx() function which is
using dai->driver->ops->xxx.
But, some of them are implemented as snd_soc_dai_xxx(),
but others are directly using dai->driver->ops->xxx.
Because of it, the code is not easy to read.

This patch creats new soc-dai.c and moves snd_soc_dai_xxx()
functions into it.
One exception is snd_soc_dai_is_dummy() which is based on
soc-utils local variable. We need to keep it as-is there.

Others which is directly using dai->driver->ops->xxx will be
implemented at soc-dai.c by incremental patches.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/871ryij1r6.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: rsnd: Support hw_free() callback at DAI level
Timo Wischer [Mon, 22 Jul 2019 07:24:01 +0000 (16:24 +0900)]
ASoC: rsnd: Support hw_free() callback at DAI level

This patch provides the needed infrastructure to support calling hw_free()
at the DAI level. This is for example required to free resources allocated
in hw_params() callback.

The modification of __rsnd_mod_add_hw_params does not have any side
effects because rsnd_mod_ops::hw_params callback is not used by anyone
until now.

Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Link: https://lore.kernel.org/r/20190722072403.11008-2-jiada_wang@mentor.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: rt1308: Convert headers to SPDX
Shuming Fan [Tue, 23 Jul 2019 10:12:47 +0000 (18:12 +0800)]
ASoC: rt1308: Convert headers to SPDX

Convert include and source files to SPDX.

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Link: https://lore.kernel.org/r/20190723101247.19741-1-shumingf@realtek.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: rt5665: Fix a typo in the name of a function
Christophe JAILLET [Mon, 22 Jul 2019 21:26:39 +0000 (23:26 +0200)]
ASoC: rt5665: Fix a typo in the name of a function

All function names start with rt5665_, except 'rt5655_set_verf()'.
It is likely a typo.

Fix it to be consistent.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/20190722212639.26954-1-christophe.jaillet@wanadoo.fr
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: fix stream id setting
Rander Wang [Mon, 22 Jul 2019 14:14:00 +0000 (09:14 -0500)]
ASoC: SOF: Intel: hda: fix stream id setting

snd_hdac_ext_link_clear_stream_id maps stream id to
link output, which is for playback, not capture.

Tested on Whiskey Lake platform.

Signed-off-by: Rander Wang <rander.wang@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-20-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: fix link DMA config
Rander Wang [Mon, 22 Jul 2019 14:13:59 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: fix link DMA config

For this bug, there are two capture pcm streams active, with one
stream and its related stream tag released before suspend. Later
when system suspend is done, the stream tag for the remaining
active stream is released by SOF driver. After system resume, hda
codec driver restores the stream tag for the active pcm stream,
but SOF goes to assign a new one, which now doesn't match with the
stream tag used by codec driver, and this causes DMA to fail
receiving data, leading to unrecoverable XRUN condition in FW.

For stream tag is stored in both hda codec and SOF driver, it
shouldn't be released only in SOF driver. This patch just keeps the
stream information in dma data and checks whether there is a stored
DMA data for stream resuming from S3 and restores it. And it also
removes DMA data when the stream is released.

Tested on Whiskey Lake platform.

GitHub issue: https://github.com/thesofproject/sof/issues/1594
Signed-off-by: Rander Wang <rander.wang@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-19-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: fsl_esai: recover the channel swap after xrun
Shengjiu Wang [Thu, 11 Jul 2019 10:49:46 +0000 (18:49 +0800)]
ASoC: fsl_esai: recover the channel swap after xrun

There is chip errata ERR008000, the reference doc is
(https://www.nxp.com/docs/en/errata/IMX6DQCE.pdf),

The issue is "While using ESAI transmit or receive and
an underrun/overrun happens, channel swap may occur.
The only recovery mechanism is to reset the ESAI."

This issue exist in imx3/imx5/imx6(partial) series.

In this commit add a tasklet to handle reset of ESAI
after xrun happens to recover the channel swap.

Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Link: https://lore.kernel.org/r/326035cb99975361699d9ed748054b08bc06a341.1562842206.git.shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: ssp: BCLK delay parameter
Janusz Jankowski [Mon, 22 Jul 2019 14:14:02 +0000 (09:14 -0500)]
ASoC: SOF: Intel: ssp: BCLK delay parameter

Some codecs require BCLK to be on for some time, before sending
any data. SOF can enable BCLK and then wait for guaranteed time,
before starting DMA on SSP start.

Signed-off-by: Janusz Jankowski <janusz.jankowski@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-22-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: reset link DMA state in prepare
Kai Vehmanen [Mon, 22 Jul 2019 14:13:58 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: reset link DMA state in prepare

When application goes through SUSPEND/STOP->PREPARE->START
cycle, we should always reprogram the DAI link DMA to ensure
it is in sync with the host PCM DMA.

Use same state tracking logic to handle both restart and
system resume flows. Use link_prepared field of
'struct hdac_ext_stream' to store the state, instead of
adding redundant fields to SOF specific structs.

Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-18-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: add a parameter to disable MSI
Guennadi Liakhovetski [Mon, 22 Jul 2019 14:13:57 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: add a parameter to disable MSI

Enabling MSI on HDA can fail, in which case the legacy PCI IRQ mode
will be used. To make testing this mode easier add an "enable_msi"
module parameter, which is only enabled if debugging is enabled too.

Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-17-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: remove duplicated clear WAKESTS
Zhu Yingjiang [Mon, 22 Jul 2019 14:13:56 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: remove duplicated clear WAKESTS

Remove the first clear WAKESTS, only one clear is needed during init
chip.

Signed-off-by: Zhu Yingjiang <yingjiang.zhu@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-16-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: use SOF defined init chip in resume
Zhu Yingjiang [Mon, 22 Jul 2019 14:13:55 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: use SOF defined init chip in resume

Unify resume code by using SOF common function hda_dsp_ctrl_init_chip()
which can handle both HDA and non-HDA cases. Move code to reset
stream-to-link mapping into hda_dsp_ctrl_init_chip().

Signed-off-by: Zhu Yingjiang <yingjiang.zhu@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-15-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: set position buffer in init chip
Zhu Yingjiang [Mon, 22 Jul 2019 14:13:54 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: set position buffer in init chip

Set the HDA stream position buffer during init chip. The position buffer
needs to be set in both HDA codec and nocodec cases. Using SOF defined
function and move it to common code.

Signed-off-by: Zhu Yingjiang <yingjiang.zhu@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-14-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: Enable jack detection
Rander Wang [Mon, 22 Jul 2019 14:13:53 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: Enable jack detection

In commit 7d4f606c50ff ("ALSA: hda - WAKEEN feature enabling for
runtime pm"), legacy HD-A driver sets hda controller in reset mode after
entering runtime-suspend. And when resuming from suspend mode, it checks
hda controller & codec status to detect headphone hotplug event. Now
this patch does the same job in SOF runtime pm functions.

And we need to check all the non-hdmi codecs for some cases like playback
with HDMI or capture with DMIC connected to dsp. In these cases, only
controller is active and codecs are suspended, so codecs can't send
unsolicited event to controller. The jack polling operation will activate
codecs and unsolicited event can work even codecs become suspended later.

Tested on whiskylake with hda codecs.

Signed-off-by: Rander Wang <rander.wang@linux.intel.com>
Signed-off-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-13-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: reduce ifdef usage for hda
Rander Wang [Mon, 22 Jul 2019 14:13:52 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: reduce ifdef usage for hda

Move the code for hda to one point

Signed-off-by: Rander Wang <rander.wang@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-12-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: correct ROM state mask
Keyon Jie [Mon, 22 Jul 2019 14:13:51 +0000 (09:13 -0500)]
ASoC: SOF: Intel: hda: correct ROM state mask

The ROM state is represented by the 24 LSB bits in the ROM status
register, so the mask should be 0xffffff instead of 0xf.

Signed-off-by: Keyon Jie <yang.jie@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-11-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: remove unused state variable in suspend function
Fred Oh [Mon, 22 Jul 2019 14:13:50 +0000 (09:13 -0500)]
ASoC: SOF: remove unused state variable in suspend function

Remove unused and no plan to use variable from suspend function.

Signed-off-by: Fred Oh <fred.oh@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-10-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: loader: Don't ignore SRAM block types
Daniel Baluta [Mon, 22 Jul 2019 14:13:49 +0000 (09:13 -0500)]
ASoC: SOF: loader: Don't ignore SRAM block types

On i.MX8 data/heap/stack is kept in System RAM so
do not ignore SRAM block types received from FW.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-9-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: loader: Use the BAR provided by FW
Daniel Baluta [Mon, 22 Jul 2019 14:13:48 +0000 (09:13 -0500)]
ASoC: SOF: loader: Use the BAR provided by FW

Make sure to use the newly introduced function snd_sof_dsp_get_bar_index
that converts the section type to appropriate BAR index.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-8-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Introduce snd_sof_dsp_get_bar_index ops
Daniel Baluta [Mon, 22 Jul 2019 14:13:47 +0000 (09:13 -0500)]
ASoC: SOF: Introduce snd_sof_dsp_get_bar_index ops

FW encapsulates information about section types (e.g DRAM, IRAM)
inside module block header. This information can be used in order
to correctly load the section to the appropriate place in memory.

SOF Linux driver needs to know for each platform how to map the
section type with the corresponding memory BAR. So, this patch
introduces get_bar_index, a new operation inside snd_sof_dsp_ops.

Intel platforms, usually load all the section in a contiguous memory
area (usually denoted by sdev->mmio_bar) so things are relatively
simple there. Anyhow, on i.MX8 IRAM and DRAM for example are mapped
to distinct BARs.

By default, if no get_bar function is provided the core implementation
will always return sdev->mmio_bar so that there will be no need for
a change to existing Intel code.

Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-7-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: core: increase default IPC timeouts
Kai Vehmanen [Mon, 22 Jul 2019 14:13:46 +0000 (09:13 -0500)]
ASoC: SOF: core: increase default IPC timeouts

Increase the default timeout values for boot (100ms to 2sec) and
IPC message sending (5ms to 500ms). The values should be overridden
with values from platform data.

There is no functional need to have such short timeouts as both boot
and IPC send errors are considered fatal errors. More relaxed timeouts
are convenient when running the driver on top of emulation such as QEMU.

Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-6-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: ipc: use timeout configured at probe
Kai Vehmanen [Mon, 22 Jul 2019 14:13:45 +0000 (09:13 -0500)]
ASoC: SOF: ipc: use timeout configured at probe

Do not hardcode IPC timeout value in ipc.c, but rather use the timeout
value configured during device probe. For platforms that do not override
the IPC timeout, default value TIMEOUT_DEFAULT_IPC_MS has already been
defined in core.c.

Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-5-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: use common code to send PCM_FREE IPC
Kai Vehmanen [Mon, 22 Jul 2019 14:13:44 +0000 (09:13 -0500)]
ASoC: SOF: use common code to send PCM_FREE IPC

Remove duplicated code by using a common helper function
to send the PCM_FREE IPC message to FW.

Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-4-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: reset DMA state in prepare
Kai Vehmanen [Mon, 22 Jul 2019 14:13:43 +0000 (09:13 -0500)]
ASoC: SOF: reset DMA state in prepare

When application goes through SUSPEND/STOP->PREPARE->START
cycle, we should always reprogram the SOF device to start
DMA from a known state so that hw_ptr/appl_ptrs remain valid.
This is expected by ALSA core as it resets the buffer
state as part of prepare (see snd_pcm_do_prepare()).

Fix the issue by forcing reconfiguration of the FW with
STREAM_PCM_PARAMS in prepare(). Use combined logic to handle
prepare and the existing flow to reprogram hw-params after
system suspend.

Without the fix, first call to pcm pointer() will return
an invalid hw_ptr and application may immediately observe XRUN
status, unless "start_threshold" SW parameter is set to maximum
value by the application.

Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-3-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: pci: mark last_busy value at runtime PM init
Pan Xiuli [Mon, 22 Jul 2019 14:13:42 +0000 (09:13 -0500)]
ASoC: SOF: pci: mark last_busy value at runtime PM init

If last_busy value is not set at runtime PM enable, the device will be
suspend immediately after usage counter is 0. Set the last_busy value to
make sure delay is working at first boot up.

Signed-off-by: Pan Xiuli <xiuli.pan@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-2-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: pcm3168a: Fix a typo in the name of a constant
Christophe JAILLET [Mon, 22 Jul 2019 21:15:28 +0000 (23:15 +0200)]
ASoC: pcm3168a: Fix a typo in the name of a constant

There is a typo in PCM1368A_MAX_SYSCLK, it should be PCM3168A_MAX_SYSCLK
(1 and 3 switched in 3168)

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Link: https://lore.kernel.org/r/20190722211528.26600-1-christophe.jaillet@wanadoo.fr
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: remove misleading error trace from IRQ thread
Kai Vehmanen [Mon, 22 Jul 2019 14:14:01 +0000 (09:14 -0500)]
ASoC: SOF: Intel: hda: remove misleading error trace from IRQ thread

Downgrade "nothing to do in IRQ thread" message from error to a debug
message in the IPC interrupt handler thread.

The spurious wake-up can happen if a HDA stream interrupt is
raised while the IPC interrupt thread is running. IPC functionality
is not impacted by this condition, so debug is a more appropriate
trace level.

Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20190722141402.7194-21-pierre-louis.bossart@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: qcom: apq8016_sbc: Fix oops with multiple DAI links
Stephan Gerhold [Mon, 22 Jul 2019 13:03:52 +0000 (15:03 +0200)]
ASoC: qcom: apq8016_sbc: Fix oops with multiple DAI links

apq8016_sbc_parse_of() sets up multiple DAI links, depending on the
number of nodes in the device tree. However, at the moment
CPU and platform components are only allocated for the first link.
This causes an oops when more than one link is defined:

Internal error: Oops: 96000044 [#1] SMP
CPU: 0 PID: 1015 Comm: kworker/0:2 Not tainted 5.3.0-rc1 #4
Call trace:
 apq8016_sbc_platform_probe+0x1a8/0x3f0
 platform_drv_probe+0x50/0xa0
...

Move the allocation inside the loop to ensure that each link is
properly initialized.

Fixes: 98b232ca9e0e ("ASoC: qcom: apq8016_sbc: use modern dai_link style")
Signed-off-by: Stephan Gerhold <stephan@gerhold.net>
Acked-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/20190722130352.95874-1-stephan@gerhold.net
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: madera: Read device tree configuration
Charles Keepax [Mon, 22 Jul 2019 13:52:09 +0000 (14:52 +0100)]
ASoC: madera: Read device tree configuration

Read the configuration of the Madera ASoC driver from device tree.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20190722135209.30302-1-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: dapm: fix a memory leak bug
Wenwen Wang [Mon, 22 Jul 2019 13:57:44 +0000 (08:57 -0500)]
ASoC: dapm: fix a memory leak bug

In snd_soc_dapm_new_control_unlocked(), a kernel buffer is allocated in
dapm_cnew_widget() to hold the new dapm widget. Then, different actions are
taken according to the id of the widget, i.e., 'w->id'. If any failure
occurs during this process, snd_soc_dapm_new_control_unlocked() should be
terminated by going to the 'request_failed' label. However, the allocated
kernel buffer is not freed on this code path, leading to a memory leak bug.

To fix the above issue, free the buffer before returning from
snd_soc_dapm_new_control_unlocked() through the 'request_failed' label.

Signed-off-by: Wenwen Wang <wenwen@cs.uga.edu>
Link: https://lore.kernel.org/r/1563803864-2809-1-git-send-email-wang6495@umn.edu
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: sgtl5000: Fix charge pump source assignment
Oleksandr Suvorov [Fri, 19 Jul 2019 10:05:37 +0000 (10:05 +0000)]
ASoC: sgtl5000: Fix charge pump source assignment

If VDDA != VDDIO and any of them is greater than 3.1V, charge pump
source can be assigned automatically [1].

[1] https://www.nxp.com/docs/en/data-sheet/SGTL5000.pdf

Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Reviewed-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@toradex.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Link: https://lore.kernel.org/r/20190719100524.23300-7-oleksandr.suvorov@toradex.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: sgtl5000: Fix of unmute outputs on probe
Oleksandr Suvorov [Fri, 19 Jul 2019 10:05:35 +0000 (10:05 +0000)]
ASoC: sgtl5000: Fix of unmute outputs on probe

To enable "zero cross detect" for ADC/HP, change
HP_ZCD_EN/ADC_ZCD_EN bits only instead of writing the whole
CHIP_ANA_CTRL register.

Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Reviewed-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@toradex.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Link: https://lore.kernel.org/r/20190719100524.23300-6-oleksandr.suvorov@toradex.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: sgtl5000: add ADC mute control
Oleksandr Suvorov [Fri, 19 Jul 2019 10:05:34 +0000 (10:05 +0000)]
ASoC: sgtl5000: add ADC mute control

This control mute/unmute the ADC input of SGTL5000
using its CHIP_ANA_CTRL register.

Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Reviewed-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@toradex.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Link: https://lore.kernel.org/r/20190719100524.23300-5-oleksandr.suvorov@toradex.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: sgtl5000: Fix definition of VAG Ramp Control
Oleksandr Suvorov [Fri, 19 Jul 2019 10:05:33 +0000 (10:05 +0000)]
ASoC: sgtl5000: Fix definition of VAG Ramp Control

SGTL5000_SMALL_POP is a bit mask, not a value. Usage of
correct definition makes device probing code more clear.

Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Reviewed-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@toradex.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Link: https://lore.kernel.org/r/20190719100524.23300-4-oleksandr.suvorov@toradex.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: sgtl5000: Improve VAG power and mute control
Oleksandr Suvorov [Fri, 19 Jul 2019 10:05:31 +0000 (10:05 +0000)]
ASoC: sgtl5000: Improve VAG power and mute control

VAG power control is improved to fit the manual [1]. This patch fixes as
minimum one bug: if customer muxes Headphone to Line-In right after boot,
the VAG power remains off that leads to poor sound quality from line-in.

I.e. after boot:
  - Connect sound source to Line-In jack;
  - Connect headphone to HP jack;
  - Run following commands:
  $ amixer set 'Headphone' 80%
  $ amixer set 'Headphone Mux' LINE_IN

Change VAG power on/off control according to the following algorithm:
  - turn VAG power ON on the 1st incoming event.
  - keep it ON if there is any active VAG consumer (ADC/DAC/HP/Line-In).
  - turn VAG power OFF when there is the latest consumer's pre-down event
    come.
  - always delay after VAG power OFF to avoid pop.
  - delay after VAG power ON if the initiative consumer is Line-In, this
    prevents pop during line-in muxing.

According to the data sheet [1], to avoid any pops/clicks,
the outputs should be muted during input/output
routing changes.

[1] https://www.nxp.com/docs/en/data-sheet/SGTL5000.pdf

Cc: stable@vger.kernel.org
Fixes: 9b34e6cc3bc2 ("ASoC: Add Freescale SGTL5000 codec support")
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Reviewed-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20190719100524.23300-3-oleksandr.suvorov@toradex.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: Define a set of DAPM pre/post-up events
Oleksandr Suvorov [Fri, 19 Jul 2019 10:05:30 +0000 (10:05 +0000)]
ASoC: Define a set of DAPM pre/post-up events

Prepare to use SND_SOC_DAPM_PRE_POST_PMU definition to
reduce coming code size and make it more readable.

Cc: stable@vger.kernel.org
Signed-off-by: Oleksandr Suvorov <oleksandr.suvorov@toradex.com>
Reviewed-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
Reviewed-by: Igor Opaniuk <igor.opaniuk@toradex.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>
Link: https://lore.kernel.org/r/20190719100524.23300-2-oleksandr.suvorov@toradex.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: rockchip-max98090: Remove MICBIAS as supply of input pin IN34
Enric Balletbo i Serra [Fri, 19 Jul 2019 18:05:58 +0000 (20:05 +0200)]
ASoC: rockchip-max98090: Remove MICBIAS as supply of input pin IN34

Commit ec0d23b295b9 ("ASoC: rockchip-max98090: Fix the Headset Mic
route.") moved the MICBIAS widget to supply Headset Mic but forget to
remove the MICBIAS widget to supply IN34 which is not really needed, so
remove that path so we have:

   IN34 <==== Headset MIC <==== MICBIAS

Signed-off-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Link: https://lore.kernel.org/r/20190719180558.11459-1-enric.balletbo@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: cs47l35: Use define for DSPCLK enable bit
Charles Keepax [Mon, 22 Jul 2019 09:35:16 +0000 (10:35 +0100)]
ASoC: cs47l35: Use define for DSPCLK enable bit

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20190722093516.12906-1-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: rt1308: add silence detection and manual PDB control
Shuming Fan [Fri, 19 Jul 2019 06:32:49 +0000 (14:32 +0800)]
ASoC: rt1308: add silence detection and manual PDB control

We enable the silence detection function in initial settings.
PDB control changes to manual mode, hence the driver could
fully control the AMP output on/off.

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Link: https://lore.kernel.org/r/20190719063249.18806-1-shumingf@realtek.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: rt1308: simplify the EFUSE read function
Shuming Fan [Fri, 19 Jul 2019 06:32:35 +0000 (14:32 +0800)]
ASoC: rt1308: simplify the EFUSE read function

The rt1308_efuse function contains many redundant settings.
We remove the redundant settings and the function still works.

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Link: https://lore.kernel.org/r/20190719063235.18757-1-shumingf@realtek.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: tlv320aic31xx: suppress error message for EPROBE_DEFER
Lucas Stach [Fri, 19 Jul 2019 14:36:37 +0000 (16:36 +0200)]
ASoC: tlv320aic31xx: suppress error message for EPROBE_DEFER

Both the supplies and reset GPIO might need a probe deferral for the
resource to be available. Don't print a error message in that case, as
it is a normal operating condition.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Acked-by: Andrew F. Davis <afd@ti.com>
Link: https://lore.kernel.org/r/20190719143637.2018-1-l.stach@pengutronix.de
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: max98383: fix i2c probe failure
fengchunguo [Mon, 15 Jul 2019 09:55:56 +0000 (17:55 +0800)]
ASoC: max98383: fix i2c probe failure

Added reset_gpio configuration for i2c probe successfully.
If not,i2c address can't be found rightly.

Error information:
max98373 3-0031: Failed to read: 0x21FF

Fixed:
[3.761299@3] max98373 3-0031: MAX98373 revisionID: 0x43
[3.828911@3] asoc-aml-card auge_sound: max98373-aif1 <-> TDM-B mapping ok

Signed-off-by: fengchunguo <chunguo.feng@amlogic.com>
Link: https://lore.kernel.org/r/20190715095556.1614-1-chunguo.feng@amlogic.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: fsl_sai: mark regmap as fast_io
Lucas Stach [Wed, 17 Jul 2019 10:51:56 +0000 (12:51 +0200)]
ASoC: fsl_sai: mark regmap as fast_io

The regmap is only ever used to access MMIO registers, so it's fair
to say that register access is fast.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Link: https://lore.kernel.org/r/20190717105156.15721-1-l.stach@pengutronix.de
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: fsl_sai: derive TX FIFO watermark from FIFO depth
Lucas Stach [Wed, 17 Jul 2019 10:56:34 +0000 (12:56 +0200)]
ASoC: fsl_sai: derive TX FIFO watermark from FIFO depth

The DMA request schould be triggered as soon as the FIFO has space
for another burst. As different versions of the SAI block have
different FIFO sizes, the watrmark level needs to be derived from
version specific data.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Tested-by: Angus Ainslie <angus@akkea.ca>
Reviewed-by: Angus Ainslie <angus@akkea.ca>
Link: https://lore.kernel.org/r/20190717105635.18514-3-l.stach@pengutronix.de
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: fsl_sai: add of_match data
Lucas Stach [Wed, 17 Jul 2019 10:56:33 +0000 (12:56 +0200)]
ASoC: fsl_sai: add of_match data

New revisions of the SAI IP block have even more differences that need
be taken into account by the driver. To avoid sprinking compatible
checks all over the driver move the current differences into of_match_data.

Signed-off-by: Lucas Stach <l.stach@pengutronix.de>
Tested-by: Angus Ainslie <angus@akkea.ca>
Reviewed-by: Angus Ainslie <angus@akkea.ca>
Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com>
Link: https://lore.kernel.org/r/20190717105635.18514-2-l.stach@pengutronix.de
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: mediatek: mt8183: make headset codec optional
Tzung-Bi Shih [Tue, 16 Jul 2019 03:24:17 +0000 (11:24 +0800)]
ASoC: mediatek: mt8183: make headset codec optional

Make headset codec optional because some variant machines may not
have an audio jack.

Signed-off-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20190716032417.19015-1-tzungbi@google.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: wcd9335: add irqflag IRQF_ONESHOT flag
Hariprasad Kelam [Wed, 10 Jul 2019 02:16:27 +0000 (07:46 +0530)]
ASoC: wcd9335: add irqflag IRQF_ONESHOT flag

Add IRQF_ONESHOT to ensure "Interrupt is not reenabled after the hardirq
handler finished".

fixes below issue reported by coccicheck

sound/soc/codecs/wcd9335.c:4068:8-33: ERROR: Threaded IRQ with no
primary handler requested without IRQF_ONESHOT

Signed-off-by: Hariprasad Kelam <hariprasad.kelam@gmail.com>
Link: https://lore.kernel.org/r/20190710021627.GA13396@hari-Inspiron-1545
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: bcm: cygnus-pcm: Unneeded variable: "ret".
Hariprasad Kelam [Tue, 9 Jul 2019 18:42:37 +0000 (00:12 +0530)]
ASoC: bcm: cygnus-pcm: Unneeded variable: "ret".

This patch fixes below issues reported by coccicheck

sound/soc/bcm/cygnus-pcm.c:642:5-8: Unneeded variable: "ret". Return "0"
on line 650
sound/soc/bcm/cygnus-pcm.c:671:5-8: Unneeded variable: "ret". Return "0"
on line 696

We cannot change return type of these functions as they are callback
functions of snd_pcm_ops

Signed-off-by: Hariprasad Kelam <hariprasad.kelam@gmail.com>
Link: https://lore.kernel.org/r/20190709184236.GA7873@hari-Inspiron-1545
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agosound: soc: codecs: mt6358: change return type of mt6358_codec_init_reg
Hariprasad Kelam [Tue, 9 Jul 2019 18:25:43 +0000 (23:55 +0530)]
sound: soc: codecs: mt6358: change return type of mt6358_codec_init_reg

As mt6358_codec_init_reg function always returns 0 , change return type
from int to void.

fixes below issue reported by coccicheck
sound/soc/codecs/mt6358.c:2260:5-8: Unneeded variable: "ret". Return "0"
on line 2289

Signed-off-by: Hariprasad Kelam <hariprasad.kelam@gmail.com>
Acked-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20190709182543.GA6611@hari-Inspiron-1545
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: fsl_esai: Wrap some operations to be functions
Shengjiu Wang [Thu, 11 Jul 2019 10:49:45 +0000 (18:49 +0800)]
ASoC: fsl_esai: Wrap some operations to be functions

Extract the operation to be functions, to improve the
readability.

In this patch, fsl_esai_hw_init, fsl_esai_register_restore,
fsl_esai_trigger_start and fsl_esai_trigger_stop are
extracted.

Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Link: https://lore.kernel.org/r/804d7e75ae7e06a913479912b578b3538ca7cd3f.1562842206.git.shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: rt5677: remove superfluous set
Curtis Malainey [Thu, 11 Jul 2019 00:28:38 +0000 (17:28 -0700)]
ASoC: rt5677: remove superfluous set

We define this variable with the same value, no need to set it twice

Signed-off-by: Curtis Malainey <cujomalainey@chromium.org>
Link: https://lore.kernel.org/r/20190711002838.35363-1-cujomalainey@chromium.org
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: codecs: ad193x: Use regmap_multi_reg_write() when initializing
Codrin Ciubotariu [Wed, 10 Jul 2019 10:51:19 +0000 (13:51 +0300)]
ASoC: codecs: ad193x: Use regmap_multi_reg_write() when initializing

Using regmap_multi_reg_write() when we set the default values for our
registers makes the code smaller and easier to read.

Suggested-by: Tzung-Bi Shih <tzungbi@google.com>
Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Reviewed-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20190710105119.22987-1-codrin.ciubotariu@microchip.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: wcd9335: Fix misuse of GENMASK macro
Joe Perches [Wed, 10 Jul 2019 05:04:25 +0000 (22:04 -0700)]
ASoC: wcd9335: Fix misuse of GENMASK macro

Arguments are supposed to be ordered high then low.

Signed-off-by: Joe Perches <joe@perches.com>
Link: https://lore.kernel.org/r/92e31a9f321fe731d428ec3ec9d4654ea8a16d1b.1562734889.git.joe@perches.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: Intel: hda: Make hdac_device device-managed
Ranjani Sridharan [Wed, 26 Jun 2019 07:04:50 +0000 (00:04 -0700)]
ASoC: SOF: Intel: hda: Make hdac_device device-managed

snd_hdac_ext_bus_device_exit() has been recently modified
to no longer free the hdac device. SOF allocates memory for
hdac_device and hda_hda_priv with kzalloc. Make them
device-managed instead so that they will be freed when the
SOF driver is unloaded.

Because of the above change, hda_codec is device-managed and
it will be freed when the ASoC device is removed. Freeing
the codec in snd_hda_codec_dev_release() leads to kernel
panic while unloading and reloading the ASoC driver. So,
avoid freeing the hda_codec for ASoC driver. This is done in
the same patch to avoid bisect failure.

Signed-off-by: Libin Yang <libin.yang@intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20190626070450.7229-1-ranjani.sridharan@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: SOF: use __u32 instead of uint32_t in uapi headers
Masahiro Yamada [Sun, 21 Jul 2019 14:23:08 +0000 (23:23 +0900)]
ASoC: SOF: use __u32 instead of uint32_t in uapi headers

When CONFIG_UAPI_HEADER_TEST=y, exported headers are compile-tested to
make sure they can be included from user-space.

Currently, header.h and fw.h are excluded from the test coverage.
To make them join the compile-test, we need to fix the build errors
attached below.

For a case like this, we decided to use __u{8,16,32,64} variable types
in this discussion:

  https://lkml.org/lkml/2019/6/5/18

Build log:

  CC      usr/include/sound/sof/header.h.s
  CC      usr/include/sound/sof/fw.h.s
In file included from <command-line>:32:0:
./usr/include/sound/sof/header.h:19:2: error: unknown type name ‘uint32_t’
  uint32_t magic;  /**< 'S', 'O', 'F', '\0' */
  ^~~~~~~~
./usr/include/sound/sof/header.h:20:2: error: unknown type name ‘uint32_t’
  uint32_t type;  /**< component specific type */
  ^~~~~~~~
./usr/include/sound/sof/header.h:21:2: error: unknown type name ‘uint32_t’
  uint32_t size;  /**< size in bytes of data excl. this struct */
  ^~~~~~~~
./usr/include/sound/sof/header.h:22:2: error: unknown type name ‘uint32_t’
  uint32_t abi;  /**< SOF ABI version */
  ^~~~~~~~
./usr/include/sound/sof/header.h:23:2: error: unknown type name ‘uint32_t’
  uint32_t reserved[4]; /**< reserved for future use */
  ^~~~~~~~
./usr/include/sound/sof/header.h:24:2: error: unknown type name ‘uint32_t’
  uint32_t data[0]; /**< Component data - opaque to core */
  ^~~~~~~~
In file included from <command-line>:32:0:
./usr/include/sound/sof/fw.h:49:2: error: unknown type name ‘uint32_t’
  uint32_t size;  /* bytes minus this header */
  ^~~~~~~~
./usr/include/sound/sof/fw.h:50:2: error: unknown type name ‘uint32_t’
  uint32_t offset; /* offset from base */
  ^~~~~~~~
./usr/include/sound/sof/fw.h:64:2: error: unknown type name ‘uint32_t’
  uint32_t size;  /* bytes minus this header */
  ^~~~~~~~
./usr/include/sound/sof/fw.h:65:2: error: unknown type name ‘uint32_t’
  uint32_t num_blocks; /* number of blocks */
  ^~~~~~~~
./usr/include/sound/sof/fw.h:73:2: error: unknown type name ‘uint32_t’
  uint32_t file_size; /* size of file minus this header */
  ^~~~~~~~
./usr/include/sound/sof/fw.h:74:2: error: unknown type name ‘uint32_t’
  uint32_t num_modules; /* number of modules */
  ^~~~~~~~
./usr/include/sound/sof/fw.h:75:2: error: unknown type name ‘uint32_t’
  uint32_t abi;  /* version of header format */
  ^~~~~~~~

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Link: https://lore.kernel.org/r/20190721142308.30306-1-yamada.masahiro@socionext.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoSoC: rockchip: rockchip_max98090: Enable MICBIAS for headset keypress detection
Enric Balletbo i Serra [Fri, 19 Jul 2019 17:39:29 +0000 (19:39 +0200)]
SoC: rockchip: rockchip_max98090: Enable MICBIAS for headset keypress detection

The TS3A227E says that the headset keypress detection needs the MICBIAS
power in order to report the key events to ensure proper operation
The headset keypress detection needs the MICBIAS power in order to report
the key events all the time as long as MIC is present. So MICBIAS pin
is forced on when a MICROPHONE is detected.

On Veyron Minnie I observed that if the MICBIAS power is not present and
the key press detection is activated (just because it is enabled when you
insert a headset), it randomly reports a keypress on insert.
E.g. (KEY_PLAYPAUSE)

 Event: (SW_HEADPHONE_INSERT), value 1
 Event: (SW_MICROPHONE_INSERT), value 1
 Event: -------------- SYN_REPORT ------------
 Event: (KEY_PLAYPAUSE), value 1

Userspace thinks that KEY_PLAYPAUSE is pressed and produces the annoying
effect that the media player starts a play/pause loop.

Note that, although most of the time the key reported is the one
associated with BTN_0, not always this is true. On my tests I also saw
different keys reported

Signed-off-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
Link: https://lore.kernel.org/r/20190719173929.24065-1-enric.balletbo@collabora.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: cs42xx8: Fix MFREQ selection issue for async mode
Shengjiu Wang [Tue, 16 Jul 2019 09:45:47 +0000 (17:45 +0800)]
ASoC: cs42xx8: Fix MFREQ selection issue for async mode

When sample rate of TX is different with sample rate of RX in
async mode, the MFreq selection will be wrong.

For example, sysclk = 24.576MHz, TX rate = 96000Hz, RX rate = 48000Hz.
Then ratio of TX = 256, ratio of RX = 512, For MFreq is shared by TX
and RX instance, the correct value of MFreq is 2 for both TX and RX.

But original method will cause MFreq = 0 for TX, MFreq = 2 for RX.
If TX is started after RX, RX will be impacted, RX work abnormal with
MFreq = 0.

This patch is to select proper MFreq value according to TX rate and
RX rate.

Fixes: 0c516b4ff85c ("ASoC: cs42xx8: Add codec driver support for CS42448/CS42888")
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Link: https://lore.kernel.org/r/20190716094547.46787-1-shengjiu.wang@nxp.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoASoC: dapm: Fix handling of custom_stop_condition on DAPM graph walks
Charles Keepax [Thu, 18 Jul 2019 08:43:33 +0000 (09:43 +0100)]
ASoC: dapm: Fix handling of custom_stop_condition on DAPM graph walks

DPCM uses snd_soc_dapm_dai_get_connected_widgets to build a
list of the widgets connected to a specific front end DAI so it
can search through this list for available back end DAIs. The
custom_stop_condition was added to is_connected_ep to facilitate this
list not containing more widgets than is necessary. Doing so both
speeds up the DPCM handling as less widgets need to be searched and
avoids issues with CODEC to CODEC links as these would be confused
with back end DAIs if they appeared in the list of available widgets.

custom_stop_condition was implemented by aborting the graph walk
when the condition is triggered, however there is an issue with this
approach. Whilst walking the graph is_connected_ep should update the
endpoints cache on each widget, if the walk is aborted the number
of attached end points is unknown for that sub-graph. When the stop
condition triggered, the original patch ignored the triggering widget
and returned zero connected end points; a later patch updated this
to set the triggering widget's cache to 1 and return that. Both of
these approaches result in inaccurate values being stored in various
end point caches as the values propagate back through the graph,
which can result in later issues with widgets powering/not powering
unexpectedly.

As the original goal was to reduce the size of the widget list passed
to the DPCM code, the simplest solution is to limit the functionality
of the custom_stop_condition to the widget list. This means the rest
of the graph will still be processed resulting in correct end point
caches, but only widgets up to the stop condition will be added to the
returned widget list.

Fixes: 6742064aef7f ("ASoC: dapm: support user-defined stop condition in dai_get_connected_widgets")
Fixes: 5fdd022c2026 ("ASoC: dpcm: play nice with CODEC<->CODEC links")
Fixes: 09464974eaa8 ("ASoC: dapm: Fix to return correct path list in is_connected_ep.")
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20190718084333.15598-1-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
5 years agoLinus 5.3-rc1
Linus Torvalds [Sun, 21 Jul 2019 21:05:38 +0000 (14:05 -0700)]
Linus 5.3-rc1

5 years agoMerge tag 'devicetree-fixes-for-5.3' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 21 Jul 2019 17:28:39 +0000 (10:28 -0700)]
Merge tag 'devicetree-fixes-for-5.3' of git://git./linux/kernel/git/robh/linux

Pull Devicetree fixes from Rob Herring:
 "Fix several warnings/errors in validation of binding schemas"

* tag 'devicetree-fixes-for-5.3' of git://git.kernel.org/pub/scm/linux/kernel/git/robh/linux:
  dt-bindings: pinctrl: stm32: Fix missing 'clocks' property in examples
  dt-bindings: iio: ad7124: Fix dtc warnings in example
  dt-bindings: iio: avia-hx711: Fix avdd-supply typo in example
  dt-bindings: pinctrl: aspeed: Fix AST2500 example errors
  dt-bindings: pinctrl: aspeed: Fix 'compatible' schema errors
  dt-bindings: riscv: Limit cpus schema to only check RiscV 'cpu' nodes
  dt-bindings: Ensure child nodes are of type 'object'

5 years agoMerge branch 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Linus Torvalds [Sun, 21 Jul 2019 17:09:43 +0000 (10:09 -0700)]
Merge branch 'work.misc' of git://git./linux/kernel/git/viro/vfs

Pull vfs documentation typo fix from Al Viro.

* 'work.misc' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
  typo fix: it's d_make_root, not d_make_inode...