openwrt/staging/blogic.git
6 years agoASoC: atmel: add SND_SOC_I2C_AND_SPI dependency
Arnd Bergmann [Wed, 26 Sep 2018 12:59:40 +0000 (14:59 +0200)]
ASoC: atmel: add SND_SOC_I2C_AND_SPI dependency

Selecting SND_SOC_WM8731 is only allowed when all its dependencies
are already there:

WARNING: unmet direct dependencies detected for SND_SOC_WM8731
  Depends on [m]: SOUND [=y] && !UML && SND [=y] && SND_SOC [=y] && SND_SOC_I2C_AND_SPI [=m]
  Selected by [y]:
  - SND_SOC_MIKROE_PROTO [=y] && SOUND [=y] && !UML && SND [=y] && SND_SOC [=y] && SND_ATMEL_SOC [=y] && OF [=y]
  Selected by [m]:
  - SND_AT91_SOC_SAM9X5_WM8731 [=m] && SOUND [=y] && !UML && SND [=y] && SND_SOC [=y] && SND_ATMEL_SOC [=y] && (ARCH_AT91 [=y] || COMPILE_TEST [=y]) && ATMEL_SSC [=y] && SND_SOC_I2C_AND_SPI [=m]
  - SND_SOC_ALL_CODECS [=m] && SOUND [=y] && !UML && SND [=y] && SND_SOC [=y] && COMPILE_TEST [=y] && SND_SOC_I2C_AND_SPI [=m]

Fixes: a45f8853a5f9 ("ASoC: Add driver for PROTO Audio CODEC (with a WM8731)")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Reviewed-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: pxa: avoid AC97_BUS build warning
Arnd Bergmann [Wed, 26 Sep 2018 12:46:35 +0000 (14:46 +0200)]
ASoC: pxa: avoid AC97_BUS build warning

Selecting AC97_BUS_NEW from SND_PXA2XX_SOC_AC97 leads to a Kconfig
warning if any other driver selects AC97_BUS:

WARNING: unmet direct dependencies detected for AC97_BUS_COMPAT
  Depends on [n]: SOUND [=y] && !UML && SND [=y] && AC97_BUS_NEW [=y] && !AC97_BUS [=y]
  Selected by [y]:
  - SND_SOC_WM9713 [=y] && SOUND [=y] && !UML && SND [=y] && SND_SOC [=y] && AC97_BUS_NEW [=y]

I don't know if that combination is supposed to work.
Assuming it is not, this adds a dependency on all users
for PXA to avoids the combination.

Fixes: 1c8bc7b3de5e ("ASoC: pxa: switch to new ac97 bus support")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: soc-utils: Rename dummy_dma_ops to snd_dummy_dma_ops
Matthias Kaehlcke [Tue, 25 Sep 2018 18:09:14 +0000 (11:09 -0700)]
ASoC: soc-utils: Rename dummy_dma_ops to snd_dummy_dma_ops

The symbols 'dummy_dma_ops' is declared with different data types by
sound/soc/soc-utils.c and arch/arm64/include/asm/dma-mapping.h. This
leads to conflicts when soc-utils.c (indirectly) includes dma-mapping.h:

sound/soc/soc-utils.c:282:33: error: conflicting types for 'dummy_dma_ops'
  static const struct snd_pcm_ops dummy_dma_ops = {
                                  ^
...
arch/arm64/include/asm/dma-mapping.h:27:33: note: previous declaration of 'dummy_dma_ops' was here
  extern const struct dma_map_ops dummy_dma_ops;
                                  ^

Rename the symbol in soc-utils.c to 'snd_dummy_dma_ops' to avoid the
conflict.

Signed-off-by: Matthias Kaehlcke <mka@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: max98088: add OF support
Andreas Färber [Tue, 25 Sep 2018 14:23:49 +0000 (16:23 +0200)]
ASoC: max98088: add OF support

MAX98088 is an older version of the MAX98089 device.

Signed-off-by: Andreas Färber <afaerber@suse.de>
[m.felsch@pengutronix.de: add CONFIG_OF compile switch]
[m.felsch@pengutronix.de: adapt commit message]
Signed-off-by: Marco Felsch <m.felsch@pengutronix.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: dt-bindings: add max98088 audio codec
Andreas Färber [Tue, 25 Sep 2018 14:23:48 +0000 (16:23 +0200)]
ASoC: dt-bindings: add max98088 audio codec

This patch adds the bindings for maxim max98088/9 audio codec.

Signed-off-by: Andreas Färber <afaerber@suse.de>
[m.felsch@pengutronix.de: adapt commit message]
[m.felsch@pengutronix.de: adapt formatting]
Signed-off-by: Marco Felsch <m.felsch@pengutronix.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: cs4265: Add a MIC pre. route
Matt Flax [Tue, 25 Sep 2018 06:40:18 +0000 (16:40 +1000)]
ASoC: cs4265: Add a MIC pre. route

The cs4265 driver is missing a microphone preamp enable.
This patch enables/disables the microphone preamp when mic
selection is made using the kcontrol.

Signed-off-by: Matt Flax <flatmax@flatmax.org>
Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_component_dais() macro
Kuninori Morimoto [Fri, 21 Sep 2018 05:23:17 +0000 (05:23 +0000)]
ASoC: add for_each_component_dais() macro

To be more readable code, this patch adds
new for_each_component_dais() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_component() macro
Kuninori Morimoto [Fri, 21 Sep 2018 05:23:01 +0000 (05:23 +0000)]
ASoC: add for_each_component() macro

To be more readable code, this patch adds
new for_each_component() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: fixup SSI clock during suspend/resume modes
Dmytro Prokopchuk [Fri, 21 Sep 2018 04:59:59 +0000 (04:59 +0000)]
ASoC: rsnd: fixup SSI clock during suspend/resume modes

Prepare <-> Cleanup functions pair has balanced calls.
But in case of suspend mode no call to rsnd_soc_dai_shutdown()
function, so cleanup isn't called. OTOH during resume mode
function rsnd_soc_dai_prepare() is called, but calling
rsnd_ssi_prepare() is skipped (rsnd_status_update() returns zero,
bacause was not cleanup before).
We need to call rsnd_ssi_prepare(), because it enables SSI clocks
by calling rsnd_ssi_master_clk_start().

This patch allows to call prepare/cleanup functions always.

Signed-off-by: Dmytro Prokopchuk <dmytro.prokopchuk@globallogic.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
[kuninori: adjusted to upstream]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: qcom: qdsp6: remove duplicated include from q6adm.c
zhong jiang [Fri, 21 Sep 2018 10:24:58 +0000 (18:24 +0800)]
ASoC: qcom: qdsp6: remove duplicated include from q6adm.c

We include wait.h twice in q6adm.c. it is unnecessary. hence remove
it.  Further, order the include files as alphabet.

Signed-off-by: zhong jiang <zhongjiang@huawei.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: tidyup for_each_card_prelinks() dai_link
Kuninori Morimoto [Fri, 21 Sep 2018 07:46:28 +0000 (07:46 +0000)]
ASoC: tidyup for_each_card_prelinks() dai_link

commit 7fe072b4df5d0 ("ASoC: add for_each_card_prelinks() macro")
added new for_each_card_prelinks() macro, but it had typo.
This patch fixup it

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5663: Remove the boost volume in the beginning of playback
Oder Chiou [Wed, 19 Sep 2018 01:56:47 +0000 (09:56 +0800)]
ASoC: rt5663: Remove the boost volume in the beginning of playback

The patch removes the boost volume in the beginning of playback while the
DAC volume set to lower.

Signed-off-by: Oder Chiou <oder_chiou@realtek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_dpcm_be() macro
Kuninori Morimoto [Tue, 18 Sep 2018 01:31:09 +0000 (01:31 +0000)]
ASoC: add for_each_dpcm_be() macro

To be more readable code, this patch adds
new for_each_dpcm_be() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_dpcm_fe() macro
Kuninori Morimoto [Tue, 18 Sep 2018 01:30:54 +0000 (01:30 +0000)]
ASoC: add for_each_dpcm_fe() macro

To be more readable code, this patch adds
new for_each_dpcm_fe() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_comp_order() macro
Kuninori Morimoto [Tue, 18 Sep 2018 01:30:41 +0000 (01:30 +0000)]
ASoC: add for_each_comp_order() macro

To be more readable code, this patch adds
new for_each_comp_order() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_card_components() macro
Kuninori Morimoto [Tue, 18 Sep 2018 01:29:55 +0000 (01:29 +0000)]
ASoC: add for_each_card_components() macro

To be more readable code, this patch adds
new for_each_card_components() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_card_rtds() macro
Kuninori Morimoto [Tue, 18 Sep 2018 01:29:35 +0000 (01:29 +0000)]
ASoC: add for_each_card_rtds() macro

To be more readable code, this patch adds
new for_each_card_rtds() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_card_links() macro
Kuninori Morimoto [Tue, 18 Sep 2018 01:29:16 +0000 (01:29 +0000)]
ASoC: add for_each_card_links() macro

To be more readable code, this patch adds
new for_each_card_links() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_card_prelinks() macro
Kuninori Morimoto [Tue, 18 Sep 2018 01:28:49 +0000 (01:28 +0000)]
ASoC: add for_each_card_prelinks() macro

To be more readable code, this patch adds
new for_each_card_prelinks() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rename for_each_rtd_codec_dai_reverse to rollback
Kuninori Morimoto [Tue, 18 Sep 2018 01:28:30 +0000 (01:28 +0000)]
ASoC: rename for_each_rtd_codec_dai_reverse to rollback

commit 0b7990e38971 ("ASoC: add for_each_rtd_codec_dai() macro")
added for_each_rtd_codec_dai_reverse(). but _rollback() is better
naming than _reverse(). This patch rename it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: convert for_each_rtd_codec_dai() for missing part
Kuninori Morimoto [Tue, 18 Sep 2018 01:28:04 +0000 (01:28 +0000)]
ASoC: convert for_each_rtd_codec_dai() for missing part

commit 0b7990e38971 ("ASoC: add for_each_rtd_codec_dai() macro")
added for_each_rtd_codec_dai(), but it didn't convert few loop
which is not using "rtd". This patch fixup it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: q6afe: dt-bindings: Update input range for qcom,sd-lines
Rohit kumar [Wed, 19 Sep 2018 06:11:21 +0000 (11:41 +0530)]
ASoC: q6afe: dt-bindings: Update input range for qcom,sd-lines

Input to qcom,sd-lines should be between 0 and 3 instead of
1 to 4 as 0 corresponds to BIT(0) which is MI2S_SD0 line.
Bit 1 to 3 corresponds to SD1 to SD3 lines respectively.
Updated documentation for the same.

Signed-off-by: Rohit kumar <rohitkr@codeaurora.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5677-spi: Drop unused GPIO include
Linus Walleij [Tue, 18 Sep 2018 19:11:57 +0000 (12:11 -0700)]
ASoC: rt5677-spi: Drop unused GPIO include

This SPI driver does not use the legacy GPIO header so
just delete it.

Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5682: Minor code modification
Shuming Fan [Tue, 18 Sep 2018 11:51:53 +0000 (19:51 +0800)]
ASoC: rt5682: Minor code modification

Minor code changes are:
 - improve the readability in patch list
 - add i2c remove function
 - regmap_register_patch changes to regmap_multi_reg_write

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5682: Fix the boost volume at the begining of playback
Shuming Fan [Tue, 18 Sep 2018 11:51:38 +0000 (19:51 +0800)]
ASoC: rt5682: Fix the boost volume at the begining of playback

This patch fixed the boost volume at the begining of playback
while DAC volume set to lower level.

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5682: Update calibration function
Shuming Fan [Tue, 18 Sep 2018 11:50:38 +0000 (19:50 +0800)]
ASoC: rt5682: Update calibration function

The ADC/DAC path should open while calibration process.

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoMerge branch 'asoc-4.19' into asoc-4.20 for rt5682 deps.
Mark Brown [Tue, 18 Sep 2018 17:30:14 +0000 (10:30 -0700)]
Merge branch 'asoc-4.19' into asoc-4.20 for rt5682 deps.

6 years agoASoC: rt5682: Remove HP volume control
Shuming Fan [Tue, 18 Sep 2018 11:51:24 +0000 (19:51 +0800)]
ASoC: rt5682: Remove HP volume control

This patch removed Headphone Playback Volume control.
Due to codec settings, we don't want the user to change HP analog gain.
The user could use DAC1 Playback Volume control to
change playback volume.

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5682: Improve HP performance
Shuming Fan [Tue, 18 Sep 2018 11:51:08 +0000 (19:51 +0800)]
ASoC: rt5682: Improve HP performance

We change the settings while HP power-up for better performance.

Signed-off-by: Shuming Fan <shumingf@realtek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: remove redundant include
zhong jiang [Tue, 18 Sep 2018 08:16:24 +0000 (16:16 +0800)]
ASoC: remove redundant include

module.h already contained moduleparam.h,  so it is safe to remove
the redundant include.

The issue is detected with the help of Coccinelle.

Signed-off-by: zhong jiang <zhongjiang@huawei.com>
Acked-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5514-spi: Get the period_bytes in the copy work to make sure the value correctly
Oder Chiou [Mon, 17 Sep 2018 11:03:09 +0000 (19:03 +0800)]
ASoC: rt5514-spi: Get the period_bytes in the copy work to make sure the value correctly

The value of period_bytes will get the zero before the hw_params() is not
run completely. Move the function snd_pcm_lib_period_bytes() to copy work,
and make sure that is not zero.

Signed-off-by: Oder Chiou <oder_chiou@realtek.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: soc-core: add snd_soc_is_matching_component()
Kuninori Morimoto [Tue, 11 Sep 2018 06:51:45 +0000 (06:51 +0000)]
ASoC: soc-core: add snd_soc_is_matching_component()

To find (CPU/)Codec/Platform, we need to find component first
(= on CPU/Codec/Platform), and find DAI from it (= CPU/Codec).
These are similar operation but difficult to be simple,
and has many duplicate code to finding component.
This patch adds new snd_soc_is_matching_component(),
and reduce duplicate codes.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: soc-core: manage platform name under snd_soc_init_platform()
Kuninori Morimoto [Tue, 11 Sep 2018 06:51:14 +0000 (06:51 +0000)]
ASoC: soc-core: manage platform name under snd_soc_init_platform()

Now "platform" is controlled by snd_soc_dai_link_component,
thus its "name" can be initialized in snd_soc_init_platform(),
instead of soc_bind_dai_link() local.
This patch do it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: qdsp6: q6asm-dai: remove duplicated include from q6asm-dai.c
YueHaibing [Fri, 14 Sep 2018 01:36:04 +0000 (01:36 +0000)]
ASoC: qdsp6: q6asm-dai: remove duplicated include from q6asm-dai.c

Remove duplicated include.

Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Acked-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: hisilicon: fix fall-through annotations
Gustavo A. R. Silva [Thu, 13 Sep 2018 19:11:07 +0000 (14:11 -0500)]
ASoC: hisilicon: fix fall-through annotations

Replace "fallthru" with a proper "fall through" annotation.

This fix is part of the ongoing efforts to enabling
-Wimplicit-fallthrough

Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: es8328: Fix fall-through annotations
Gustavo A. R. Silva [Thu, 13 Sep 2018 19:08:15 +0000 (14:08 -0500)]
ASoC: es8328: Fix fall-through annotations

Replace "fallthru" with a proper "fall through" annotation.

This fix is part of the ongoing efforts to enabling
-Wimplicit-fallthrough

Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoALSA: hda: Fix implicit definition of pci_iomap() on SH
Mark Brown [Wed, 12 Sep 2018 11:31:32 +0000 (12:31 +0100)]
ALSA: hda: Fix implicit definition of pci_iomap() on SH

Include asm/io.h directly so we've got a definition of pci_iomap(), the
current set of includes do this implicitly on most architectures but not
on SH.

Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agosound: don't call skl_init_chip() to reset intel skl soc
Yu Zhao [Tue, 11 Sep 2018 21:15:16 +0000 (15:15 -0600)]
sound: don't call skl_init_chip() to reset intel skl soc

Internally, skl_init_chip() calls snd_hdac_bus_init_chip() which
1) sets bus->chip_init to prevent multiple entrances before device
is stopped; 2) enables interrupt.

We shouldn't use it for the purpose of resetting device only because
1) when we really want to initialize device, we won't be able to do
so; 2) we are ready to handle interrupt yet, and kernel crashes when
interrupt comes in.

Rename azx_reset() to snd_hdac_bus_reset_link(), and use it to reset
device properly.

Fixes: 60767abcea3d ("ASoC: Intel: Skylake: Reset the controller in probe")
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Yu Zhao <yuzhao@google.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agosound: enable interrupt after dma buffer initialization
Yu Zhao [Tue, 11 Sep 2018 21:14:04 +0000 (15:14 -0600)]
sound: enable interrupt after dma buffer initialization

In snd_hdac_bus_init_chip(), we enable interrupt before
snd_hdac_bus_init_cmd_io() initializing dma buffers. If irq has
been acquired and irq handler uses the dma buffer, kernel may crash
when interrupt comes in.

Fix the problem by postponing enabling irq after dma buffer
initialization. And warn once on null dma buffer pointer during the
initialization.

Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Yu Zhao <yuzhao@google.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoRevert "ASoC: Intel: Skylake: Acquire irq after RIRB allocation"
Yu Zhao [Tue, 11 Sep 2018 21:12:46 +0000 (15:12 -0600)]
Revert "ASoC: Intel: Skylake: Acquire irq after RIRB allocation"

This reverts commit 12eeeb4f4733bbc4481d01df35933fc15beb8b19.

The patch doesn't fix accessing memory with null pointer in
skl_interrupt().

There are two problems: 1) skl_init_chip() is called twice, before
and after dma buffer is allocate. The first call sets bus->chip_init
which prevents the second from initializing bus->corb.buf and
rirb.buf from bus->rb.area. 2) snd_hdac_bus_init_chip() enables
interrupt before snd_hdac_bus_init_cmd_io() initializing dma buffers.
There is a small window which skl_interrupt() can be called if irq
has been acquired. If so, it crashes when using null dma buffer
pointers.

Will fix the problems in the following patches. Also attaching the
crash for future reference.

[   16.949148] general protection fault: 0000 [#1] PREEMPT SMP KASAN PTI
<snipped>
[   16.950903] Call Trace:
[   16.950906]  <IRQ>
[   16.950918]  skl_interrupt+0x19e/0x2d6 [snd_soc_skl]
[   16.950926]  ? dma_supported+0xb5/0xb5 [snd_soc_skl]
[   16.950933]  __handle_irq_event_percpu+0x27a/0x6c8
[   16.950937]  ? __irq_wake_thread+0x1d1/0x1d1
[   16.950942]  ? __do_softirq+0x57a/0x69e
[   16.950944]  handle_irq_event_percpu+0x95/0x1ba
[   16.950948]  ? _raw_spin_unlock+0x65/0xdc
[   16.950951]  ? __handle_irq_event_percpu+0x6c8/0x6c8
[   16.950953]  ? _raw_spin_unlock+0x65/0xdc
[   16.950957]  ? time_cpufreq_notifier+0x483/0x483
[   16.950959]  handle_irq_event+0x89/0x123
[   16.950962]  handle_fasteoi_irq+0x16f/0x425
[   16.950965]  handle_irq+0x1fe/0x28e
[   16.950969]  do_IRQ+0x6e/0x12e
[   16.950972]  common_interrupt+0x7a/0x7a
[   16.950974]  </IRQ>
<snipped>
[   16.951031] RIP: snd_hdac_bus_update_rirb+0x19b/0x4cf [snd_hda_core] RSP: ffff88015c807c08
[   16.951036] ---[ end trace 58bf9ece1775bc92 ]---

Fixes: 2eeeb4f4733b ("ASoC: Intel: Skylake: Acquire irq after RIRB allocation")
Signed-off-by: Yu Zhao <yuzhao@google.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: soc-core: remove dai->driver NULL check
Kuninori Morimoto [Tue, 11 Sep 2018 06:50:27 +0000 (15:50 +0900)]
ASoC: soc-core: remove dai->driver NULL check

It is strange if it has "dai" but doesn't have "dai->driver".
And more over "dai->driver->xxx" is used everywhere without
"dai->driver" pointer NULL checking.
It got Oops already if "dai->driver" was NULL.
Let's remove un-needed "dai->driver" NULL check.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: core: add support to card rebind
Srinivas Kandagatla [Wed, 12 Sep 2018 09:15:00 +0000 (10:15 +0100)]
ASoC: core: add support to card rebind

Current behaviour of ASoC core w.r.t to component removal is that it
unregisters dependent sound card totally. There is no support to
rebind the card if the component comes back.
Typical use case is DSP restart or kernel modules itself.

With this patch, core now maintains list of cards that are unbind due to
any of its depended components are removed and card not unregistered yet.
This list is cleared when the card is rebind successfully or when the
card is unregistered from machine driver.

This list of unbind cards are tried to bind once again after every new
component is successfully added, giving a fair chance for card bind
to be successful.

Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: remove unneeded static set .owner field in platform_driver
zhong jiang [Wed, 12 Sep 2018 03:41:49 +0000 (11:41 +0800)]
ASoC: remove unneeded static set .owner field in platform_driver

platform_driver_register will set the .owner field. So it is safe
to remove the redundant assignment.

The issue is detected with the help of Coccinelle.

Signed-off-by: zhong jiang <zhongjiang@huawei.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rockchip: add missing INTERLEAVED PCM attribute
Katsuhiro Suzuki [Mon, 10 Sep 2018 16:39:32 +0000 (01:39 +0900)]
ASoC: rockchip: add missing INTERLEAVED PCM attribute

This patch adds SNDRV_PCM_INFO_INTERLEAVED into PCM hardware info.

Signed-off-by: Katsuhiro Suzuki <katsuhiro@katsuster.net>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: soc-core: remove unused num_dai_links
Kuninori Morimoto [Tue, 11 Sep 2018 06:59:01 +0000 (06:59 +0000)]
ASoC: soc-core: remove unused num_dai_links

ALSA SoC is counting card->dai_link_list user,
but no-one is using it.
Let's remove it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: soc-core: avoid nested code on soc_remove_dai()
Kuninori Morimoto [Tue, 11 Sep 2018 06:54:26 +0000 (06:54 +0000)]
ASoC: soc-core: avoid nested code on soc_remove_dai()

Nested code is not readable.
This patch avoid it on soc_remove_dai().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5668: remove empty rt5668_i2c_remove()
Kuninori Morimoto [Tue, 11 Sep 2018 07:02:04 +0000 (07:02 +0000)]
ASoC: rt5668: remove empty rt5668_i2c_remove()

rt5668_i2c_remove() is empty, and no longer needed.
Let's remove it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: Fix UBSAN warning at snd_soc_get/put_volsw_sx()
Rohit kumar [Tue, 11 Sep 2018 09:29:21 +0000 (14:59 +0530)]
ASoC: Fix UBSAN warning at snd_soc_get/put_volsw_sx()

In functions snd_soc_get_volsw_sx() or snd_soc_put_volsw_sx(),
if the result of (min + max) is negative, then fls() returns
signed integer with value as 32. This leads to signed integer
overflow as complete operation is considered as signed integer.

UBSAN: Undefined behaviour in sound/soc/soc-ops.c:382:50
signed integer overflow:
-2147483648 - 1 cannot be represented in type 'int'
Call trace:
[<ffffff852f746fe4>] __dump_stack lib/dump_stack.c:15 [inline]
[<ffffff852f746fe4>] dump_stack+0xec/0x158 lib/dump_stack.c:51
[<ffffff852f7b5f3c>] ubsan_epilogue+0x18/0x50 lib/ubsan.c:164
[<ffffff852f7b6840>] handle_overflow+0xf8/0x130 lib/ubsan.c:195
[<ffffff852f7b68f0>] __ubsan_handle_sub_overflow+0x34/0x44 lib/ubsan.c:211
[<ffffff85307971a0>] snd_soc_get_volsw_sx+0x1a8/0x1f8 sound/soc/soc-ops.c:382

Typecast the operation to unsigned int to fix the issue.

Signed-off-by: Rohit kumar <rohitkr@codeaurora.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoMerge branch 'asoc-4.19' into asoc-4.20 Cirrus conflict
Mark Brown [Mon, 10 Sep 2018 17:55:12 +0000 (18:55 +0100)]
Merge branch 'asoc-4.19' into asoc-4.20 Cirrus conflict

6 years agoASoC: pxa: switch to new ac97 bus support
Robert Jarzmik [Sat, 25 Aug 2018 08:46:18 +0000 (10:46 +0200)]
ASoC: pxa: switch to new ac97 bus support

Switch to the new ac97 bus support in sound/ac97 instead of the legacy
snd_ac97 one.

Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: AMD: Fix simultaneous playback and capture on different channel
Akshu Agrawal [Mon, 10 Sep 2018 17:20:27 +0000 (22:50 +0530)]
ASoC: AMD: Fix simultaneous playback and capture on different channel

If capture and playback are started on different channel (I2S/BT)
there is a possibilty that channel information passed from machine driver
is overwritten before the configuration is done in dma driver.
Example:
113.597588: cz_max_startup: ---playback sets BT channel
113.597694: cz_dmic1_startup: ---capture sets I2S channel
113.597979: acp_dma_hw_params: ---configures capture for I2S channel
113.598114: acp_dma_hw_params: ---configures playback for I2S channel

This is fixed by having 2 separate instance for playback and capture.

Signed-off-by: Akshu Agrawal <akshu.agrawal@amd.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: cs4265: fix MMTLR Data switch control
Sébastien Szymanski [Thu, 6 Sep 2018 09:16:00 +0000 (11:16 +0200)]
ASoC: cs4265: fix MMTLR Data switch control

The MMTLR bit is in the CS4265_SPDIF_CTL2 register at address 0x12 bit 0
and not at address 0x0 bit 1. Fix this.

Signed-off-by: Sébastien Szymanski <sebastien.szymanski@armadeus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org
6 years agoASoC: dapm: Add missing return value check for snd_soc_dapm_new_dai
Charles Keepax [Mon, 10 Sep 2018 14:28:39 +0000 (15:28 +0100)]
ASoC: dapm: Add missing return value check for snd_soc_dapm_new_dai

snd_soc_dapm_new_dai may return an error pointer and currently this
isn't checked for in dapm_connect_dai_link_widgets. Add code to check
the return value and not add routes in that case.

Fixes: 778ff5bb8689 ("ASoC: dapm: Move connection of CODEC to CODEC DAIs")
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: skl-topology: Use kmemdup to replace kzalloc + memcpy
zhong jiang [Sat, 8 Sep 2018 08:36:20 +0000 (16:36 +0800)]
ASoC: skl-topology: Use kmemdup to replace kzalloc + memcpy

kmemdup has implemented the function that kzalloc() + memcpy() will
do. and we prefer to kmemdup rather than the open coded implementation.

Signed-off-by: zhong jiang <zhongjiang@huawei.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: q6core: Use kmemdup to replace kzalloc + memcpy
zhong jiang [Sat, 8 Sep 2018 08:36:19 +0000 (16:36 +0800)]
ASoC: q6core: Use kmemdup to replace kzalloc + memcpy

kmemdup has implemented the function that kzalloc() + memcpy() will
do. and we prefer to use the kmemdup rather than the open coded implementation.

Signed-off-by: zhong jiang <zhongjiang@huawei.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: txx9: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:03:53 +0000 (01:03 +0000)]
ASoC: txx9: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: sh: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:03:25 +0000 (01:03 +0000)]
ASoC: sh: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: omap: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:02:54 +0000 (01:02 +0000)]
ASoC: omap: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: nuc900: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:02:38 +0000 (01:02 +0000)]
ASoC: nuc900: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: tscs454: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:01:50 +0000 (01:01 +0000)]
ASoC: tscs454: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rt5668: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:01:34 +0000 (01:01 +0000)]
ASoC: rt5668: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: hdac_hda: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:01:19 +0000 (01:01 +0000)]
ASoC: hdac_hda: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: bcm: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:00:49 +0000 (01:00 +0000)]
ASoC: bcm: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: atmel_ssc_dai: use devm_snd_soc_register_component()
Kuninori Morimoto [Fri, 7 Sep 2018 01:00:15 +0000 (01:00 +0000)]
ASoC: atmel_ssc_dai: use devm_snd_soc_register_component()

Now we have devm_snd_soc_register_component().
Let's use it instead of snd_soc_register_component().

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: AMD: Ensure reset bit is cleared before configuring
Akshu Agrawal [Mon, 10 Sep 2018 08:06:30 +0000 (13:36 +0530)]
ASoC: AMD: Ensure reset bit is cleared before configuring

HW register descriptions says:
"DMA Channel Reset...Software must confirm that this bit is
cleared before reprogramming any of the channel configuration registers."
There could be cases where dma stop errored out leaving dma channel
in reset state. We need to ensure that before the start of another dma,
channel is out of the reset state.

Signed-off-by: Akshu Agrawal <akshu.agrawal@amd.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: qdsp6: q6asm-dai: clean up a return
Dan Carpenter [Mon, 10 Sep 2018 08:40:48 +0000 (11:40 +0300)]
ASoC: qdsp6: q6asm-dai: clean up a return

Smatch complains that if both "psubstream" and "csubstream" are NULL
then "ret" is uninitialized.  That probably can't happen, but it's
cleaner to just return zero anyway so let's do that.

Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: dapm: Fix a couple uninitialized ret variables
Dan Carpenter [Fri, 7 Sep 2018 19:40:33 +0000 (22:40 +0300)]
ASoC: dapm: Fix a couple uninitialized ret variables

Smatch complains that these variables could be uninitialized.  The first
one in snd_soc_dai_link_event() is probably a false positive, because
probably we know the lists are not empty.  I would normally ignore the
warning, but GCC complains here as well so I just silenced the warning.
The "ret" in snd_soc_dapm_new_dai() does need to be initialized or it
leads to a bogus dereference in the caller.

Fixes: 3bbf5d34fd4a ("ASoC: dapm: Move error handling to snd_soc_dapm_new_control_unlocked")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: max98373: usleep_range() needs include/delay.h
Grant Grundler [Fri, 7 Sep 2018 00:27:28 +0000 (17:27 -0700)]
ASoC: max98373: usleep_range() needs include/delay.h

Commit ca917f9fe1a0fab added use of usleep_range() but not
the corresponding "include <linux/delay.h>". The result is
with Chrome OS won't build because warnings are forced
to be errors:
mnt/host/source/src/third_party/kernel/v4.4/sound/soc/codecs/max98373.c:734:2: error: implicit declaration of function 'usleep_range' [-Werror,-Wimplicit-function-declaration]
        usleep_range(10000, 11000);
        ^

Including delay.h "fixes" this.

Signed-off-by: Grant Grundler <grundler@chromium.org>
Reviewed-by: Benson Leung <bleung@chromium.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: Intel: hdac_hdmi: Limit sampling rates at dai creation
Yong Zhi [Tue, 7 Aug 2018 17:19:16 +0000 (12:19 -0500)]
ASoC: Intel: hdac_hdmi: Limit sampling rates at dai creation

Playback of 44.1Khz contents with HDMI plugged returns
"Invalid pipe config" because HDMI paths in the FW
topology are configured to operate at 48Khz.

This patch filters out sampling rates not supported
at hdac_hdmi_create_dais() to let user space SRC
to do the converting.

Signed-off-by: Yong Zhi <yong.zhi@intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: dapm: Avoid uninitialised variable warning
Charles Keepax [Thu, 6 Sep 2018 16:41:55 +0000 (17:41 +0100)]
ASoC: dapm: Avoid uninitialised variable warning

Commit 4a75aae17b2a ("ASoC: dapm: Add support for multi-CODEC
CODEC to CODEC links") adds loops that iterate over multiple
CODECs in snd_soc_dai_link_event. This also introduced a compiler
warning for a potentially uninitialised variable in the case
no CODECs are present. This should never be the case as the
DAI link must by definition contain at least 1 CODEC however
probably best to avoid the compiler warning by initialising ret
to zero.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: Add device tree binding for r8a77990
Hiroyuki Yokoyama [Thu, 6 Sep 2018 13:28:33 +0000 (22:28 +0900)]
ASoC: rsnd: Add device tree binding for r8a77990

This patch adds the device tree binding of the r8a77990 SoC.

Signed-off-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Yoshihiro Kaneko <ykaneko0929@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: sgtl5000: avoid division by zero if lo_vag is zero
Colin Ian King [Thu, 6 Sep 2018 10:41:52 +0000 (11:41 +0100)]
ASoC: sgtl5000: avoid division by zero if lo_vag is zero

In the case where lo_vag <= SGTL5000_LINE_OUT_GND_BASE, lo_vag
is set to zero and later vol_quot is computed by dividing by
lo_vag causing a division by zero error.  Fix this by avoiding
a zero division and set vol_quot to zero in this specific case
so that the lowest setting for i is correctly set.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: don't use %p for dev_dbg()
Kuninori Morimoto [Thu, 6 Sep 2018 03:22:01 +0000 (03:22 +0000)]
ASoC: rsnd: don't use %p for dev_dbg()

rsnd driver sometimes want to know which address is used when debugging.
But it will indicate "(____ptrval____)" if it used "%p" on dev_dbg().
Let's use "%pa" or "%px" for it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: don't fallback to PIO mode when -EPROBE_DEFER
Kuninori Morimoto [Thu, 6 Sep 2018 03:21:47 +0000 (03:21 +0000)]
ASoC: rsnd: don't fallback to PIO mode when -EPROBE_DEFER

Current rsnd driver will fallback to PIO mode if it can't get DMA
handler. But, DMA might return -EPROBE_DEFER when probe timing.
This driver always fallback to PIO mode especially from
commit ac6bbf0cdf4206c ("iommu: Remove IOMMU_OF_DECLARE") because
of this reason.

The DMA driver will be probed later, but sound driver might be
probed as PIO mode in such case. This patch fixup this issue.
Then, -EPROBE_DEFER is not error. Thus, let's don't indicate error
message in such case.
And it needs to call rsnd_adg_remove() individually if probe failed,
because it registers clk which should be unregister.

Maybe PIO fallback feature itself is not needed,
but let's keep it so far.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: adg: care clock-frequency size
Kuninori Morimoto [Thu, 6 Sep 2018 03:21:33 +0000 (03:21 +0000)]
ASoC: rsnd: adg: care clock-frequency size

ADG has buffer over flow bug if DT has more than 3 clock-frequency.
This patch fixup this issue, and uses first 2 values.

clock-frequency = <x y>; /* this is OK */
clock-frequency = <x y z>; /* this is NG */

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: gen: use tab instead of white-space
Kuninori Morimoto [Thu, 6 Sep 2018 03:21:16 +0000 (03:21 +0000)]
ASoC: rsnd: gen: use tab instead of white-space

commit 8c9d75033340 ("ASoC: rsnd: ssiu: Support BUSIF
other than BUSIF0") added new SSIU registers.
But it is using white-space for it.
This patch fixup it to use tab.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: hdac_hdmi: remove redundant check for !port condition
Colin Ian King [Thu, 6 Sep 2018 09:39:01 +0000 (10:39 +0100)]
ASoC: hdac_hdmi: remove redundant check for !port condition

The !port check is redundant as it being performed in the following
check. Remove it.

Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: dapm: Move CODEC to CODEC params from the widget to the runtime
Charles Keepax [Wed, 5 Sep 2018 14:21:02 +0000 (15:21 +0100)]
ASoC: dapm: Move CODEC to CODEC params from the widget to the runtime

Larger CODECs may contain many several hundred widgets and which set of
parameters is selected only needs to be recorded on a per DAI basis. As
such move the selected CODEC to CODEC link params to be stored in the
runtime rather than the DAPM widget, to save some memory.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: dapm: Add support for multi-CODEC CODEC to CODEC links
Charles Keepax [Wed, 5 Sep 2018 14:21:01 +0000 (15:21 +0100)]
ASoC: dapm: Add support for multi-CODEC CODEC to CODEC links

Currently multi-CODEC is not supported on CODEC to CODEC links.
There are common applications where this would be useful, such
as connecting two mono amplifiers to an audio CODEC. Adding
support simply requires an update of snd_soc_dai_link_event
to loop over the attached CODEC DAIs.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: dapm: Move connection of CODEC to CODEC DAIs
Charles Keepax [Wed, 5 Sep 2018 14:21:00 +0000 (15:21 +0100)]
ASoC: dapm: Move connection of CODEC to CODEC DAIs

Currently, snd_soc_dapm_connect_dai_link_widgets connects up the routes
representing normal DAIs, however CODEC to CODEC links are hooked up
through separate infrastructure in soc_link_dai_widgets. Improve the
consistency of the code by using snd_soc_dapm_connect_dai_link for both
types of DAIs.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: dapm: Cosmetic tidy up of snd_soc_dapm_new_control
Charles Keepax [Wed, 5 Sep 2018 14:20:59 +0000 (15:20 +0100)]
ASoC: dapm: Cosmetic tidy up of snd_soc_dapm_new_control

Move the function snd_soc_dapm_new_control to be next to
snd_soc_dapm_new_controls and add some kernel doc for it.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: dapm: Move error handling to snd_soc_dapm_new_control_unlocked
Charles Keepax [Wed, 5 Sep 2018 14:20:58 +0000 (15:20 +0100)]
ASoC: dapm: Move error handling to snd_soc_dapm_new_control_unlocked

Currently DAPM has a lot of similar code to handle errors from
snd_soc_dapm_new_control_unlocked, and much of this code does
not really accurately reflect what the function returns.

Firstly, most places will check for a return value of
-EPROBE_DEFER and silence any error messages in that case. The
one notable exception here being dapm_kcontrol_data_alloc
which does currently print any error messages in the case
of snd_soc_dapm_new_control_unlocked returning NULL or an
error. Additionally the error prints being silenced in these
case are redundant as snd_soc_dapm_new_control_unlocked can
only return -EPROBE_DEFER or NULL when failing.

Secondly, most places will treat a return value of NULL as
an -ENOMEM.  This is not correct either since any error except
EPROBE_DEFER will cause a return value of NULL from
snd_soc_dapm_new_control_unlocked.

Centralise this handling and the error messages within
snd_soc_dapm_new_control_unlocked and update the callers
to simply check IS_ERR and return. Note that this update is
slightly simpler in the case of dapm_kcontrol_data_alloc where
that is fairly close to the handling that was already in place.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: uniphier: change status to orphan
Katsuhiro Suzuki [Wed, 5 Sep 2018 08:31:37 +0000 (17:31 +0900)]
ASoC: uniphier: change status to orphan

Since I'm leaving from Socionext, I'll unable to access specification
documents of this hardware (these are not public). So change the
state to orphan until someone will maintain this driver.

Signed-off-by: Katsuhiro Suzuki <suzuki.katsuhiro@socionext.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_rtd_codec_dai() macro
Kuninori Morimoto [Mon, 3 Sep 2018 02:12:56 +0000 (02:12 +0000)]
ASoC: add for_each_rtd_codec_dai() macro

ALSA SoC snd_soc_pcm_runtime has snd_soc_dai array for codec_dai.
To be more readable code, this patch adds
new for_each_rtd_codec_dai() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: add for_each_link_codecs() macro
Kuninori Morimoto [Mon, 3 Sep 2018 02:12:40 +0000 (02:12 +0000)]
ASoC: add for_each_link_codecs() macro

ALSA SoC snd_soc_dai_link has snd_soc_dai_link_component array
for codecs.
To be more readable code, this patch adds
new for_each_link_codecs() macro, and replace existing code to it.

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: tlv320aic31xx: Add overflow detection support
Andrew F. Davis [Tue, 4 Sep 2018 15:36:17 +0000 (10:36 -0500)]
ASoC: tlv320aic31xx: Add overflow detection support

Similar to short circuit detection, when the ADC/DAC is saturated and
overflows poor audio quality can result and should be reported to the
user. This device support Automatic Dynamic Range Compression (DRC)
to reduce this but it is not enabled currently in this driver.

Signed-off-by: Andrew F. Davis <afd@ti.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: core: Don't schedule DAPM work if already in target state
Jon Hunter [Fri, 17 Aug 2018 15:35:43 +0000 (16:35 +0100)]
ASoC: core: Don't schedule DAPM work if already in target state

When dapm_power_widgets() is called, the dapm_pre_sequence_async() and
dapm_post_sequence_async() functions are scheduled for all DAPM contexts
(apart from the card DAPM context) regardless of whether the DAPM
context is already in the desired state. The overhead of this is not
insignificant and the more DAPM contexts there are the more overhead
there is.

For example, on the Tegra124 Jetson TK1, when profiling the time taken
to execute the dapm_power_widgets() the following times were observed.

  Times for function dapm_power_widgets() are (us):
     Min 23, Ave 190, Max 434, Count 39

Here 'Count' is the number of times that dapm_power_widgets() has been
called. Please note that the above time were measured using ktime_get()
to log the time on entry and exit from dapm_power_widgets(). So it
should be noted that these times may not be purely the time take to
execute this function if it is preempted. However, after applying this
patch and measuring the time taken to execute dapm_power_widgets() again
a significant improvement is seen as shown below.

  Times for function dapm_power_widgets() are (us):
     Min 4, Ave 16, Max 82, Count 39

Therefore, optimise the dapm_power_widgets() function by only scheduling
the dapm_pre/post_sequence_async() work if the DAPM context is not in
the desired state.

Signed-off-by: Jon Hunter <jonathanh@nvidia.com>
Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: mikroe-proto: dt-bindings: add DT bindings for PROTO board
Codrin Ciubotariu [Fri, 31 Aug 2018 17:14:36 +0000 (20:14 +0300)]
ASoC: mikroe-proto: dt-bindings: add DT bindings for PROTO board

DT binding documentation for this new ASoC driver.

Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: Add driver for PROTO Audio CODEC (with a WM8731)
Codrin Ciubotariu [Fri, 31 Aug 2018 17:14:35 +0000 (20:14 +0300)]
ASoC: Add driver for PROTO Audio CODEC (with a WM8731)

Add support for the MikroElektronika PROTO audio codec board.

URL to the audio chip:
http://www.mikroe.com/add-on-boards/audio-voice/audio-codec-proto/

Signed-off-by: Florian Meier <florian.meier@koalo.de>
Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: merge .nolock_start and .prepare
Kuninori Morimoto [Mon, 3 Sep 2018 07:09:17 +0000 (07:09 +0000)]
ASoC: rsnd: merge .nolock_start and .prepare

Main purpose of .nolock_start is we need to call
some function without spinlock.
OTOH we have .prepare which main purpose is
called under atomic context.
Then, it is called without spinlock.

In summary, our main callback init/quit, and start/stop
are called under "atomic context and with spinlock".
And some function need to be called under
"non-atomic context or without spinlock".
Let's merge .nolock_start and prepare to be more clear code.
Then, let's rename nolock_stop to cleanup

Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoMerge branch 'asoc-4.19' into asoc-4.20 for rcar dep
Mark Brown [Mon, 3 Sep 2018 13:37:28 +0000 (14:37 +0100)]
Merge branch 'asoc-4.19' into asoc-4.20 for rcar dep

6 years agoASoC: rsnd: ssiu: Support to init different BUSIF instance
Jiada Wang [Mon, 3 Sep 2018 07:08:37 +0000 (07:08 +0000)]
ASoC: rsnd: ssiu: Support to init different BUSIF instance

Currently ssiu's .init is only called once during audio stream.
But SSIU with different BUSIF, shall be initialized each time,
even they are used in the same audio stream.

This patch introduces ssiu_status for BUSIF0 to BUSIF7 in rsnd_ssiu,
to make sure same .init for different BUSIF can always be executed.

To avoid the first stopped stream to stop the whole SSIU,
which may still has other BUSIF instance running, use usrcnt to count
the usage of SSIU, only the last user of SSIU can stop the whole SSIU.

Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[Kuninori: tidyup for upstream]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: ssiu: Support BUSIF other than BUSIF0
Jiada Wang [Mon, 3 Sep 2018 07:08:20 +0000 (07:08 +0000)]
ASoC: rsnd: ssiu: Support BUSIF other than BUSIF0

Currently only BUSIF0 is supported by SSIU, all register setting
is done only for BUSIF.

Since BUSIF1 ~ BUSIF7 has been supported, so also support
these BUSIF from SSIU.

One note is that we can't support SSI9-4/5/6/7 so far,
because its address is out of calculation rule.

Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[Kuninori: tidyup for upstream]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoc: rsnd: dma: Calculate PDMACHCRE with consider of BUSIF
Jiada Wang [Mon, 3 Sep 2018 07:08:00 +0000 (07:08 +0000)]
ASoc: rsnd: dma: Calculate PDMACHCRE with consider of BUSIF

PDMACHCR setting for SSI only considers BUSIF0 so far.
But BUSIF1 ~ BUSIF7 also maybe used, in the future.

This patch updates table gen2_id_table_ssiu, to also consider
BUSIF number used by SSI.

Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[kuninori: adjust to upstreaming]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoc: rsnd: dma: Calculate dma address with consider of BUSIF
Jiada Wang [Mon, 3 Sep 2018 07:07:43 +0000 (07:07 +0000)]
ASoc: rsnd: dma: Calculate dma address with consider of BUSIF

DMA address calculated by rsnd_dma_addr() only considers BUSIF0 so far.
But BUSIF1 ~ BUSIF7 also maybe used, in the future.

This patch updates DMA address calculations, to also consider
BUSIF number used by SSI.

One note is that we can't support SSI9-4/5/6/7 so far,
because its address is out of calculation rule.

Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[kuninori: adjust to upstreaming]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: ssi: Check runtime channel number rather than hw_params
Jiada Wang [Mon, 3 Sep 2018 07:07:26 +0000 (07:07 +0000)]
ASoC: rsnd: ssi: Check runtime channel number rather than hw_params

The number of channel handled by SSI maybe differs from the one set
in hw_params, currently SSI checks hw_params's channel number,
and constrains to use same channel number, when it is being
used by multiple clients.

This patch corrects to check runtime channel number rather
than channel number set in hw_params.

Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[kuninori: adjust to upstreaming]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: ssi: Fix issue in dma data address assignment
Jiada Wang [Mon, 3 Sep 2018 07:07:07 +0000 (07:07 +0000)]
ASoC: rsnd: ssi: Fix issue in dma data address assignment

Same SSI device may be used in different dai links,
by only having one dma struct in rsnd_ssi, after the first
instance's dma config be initilized, the following instances
can no longer configure dma, this causes issue, when their
dma data address are different from the first instance.

Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[Kuninori: tidyup for upstream]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: remove is_play parameter from hw_rule function
Jiada Wang [Mon, 3 Sep 2018 07:06:50 +0000 (07:06 +0000)]
ASoC: rsnd: remove is_play parameter from hw_rule function

Currently rsnd_dai_stream *io is set to either &rdai->playback or
&rdai->capture based on whether it is a playback or capture stream,
in __rsnd_soc_hw_rule_* functions, but this is not necessary,
rsnd_dai_stream *io handler can be get from rule->private.

This patch removes 'is_play' parameter from hw_rule function.

Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[Kuninori: tidyup for upstream]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: add support for 8 bit S8 format
Dragos Tarcatu [Mon, 3 Sep 2018 07:06:29 +0000 (07:06 +0000)]
ASoC: rsnd: add support for 8 bit S8 format

This patch adds support for SNDRV_PCM_FMTBIT_S8 format.

Signed-off-by: Dragos Tarcatu <dragos_tarcatu@mentor.com>
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[Kuninori: tidyup for upstream]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
6 years agoASoC: rsnd: add support for the DSP_A/DSP_B formats
Dragos Tarcatu [Mon, 3 Sep 2018 07:06:01 +0000 (07:06 +0000)]
ASoC: rsnd: add support for the DSP_A/DSP_B formats

Signed-off-by: Dragos Tarcatu <dragos_tarcatu@mentor.com>
Signed-off-by: Jiada Wang <jiada_wang@mentor.com>
Signed-off-by: Timo Wischer <twischer@de.adit-jv.com>
[Kuninori: tidyup for upstream]
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Tested-by: Hiroyuki Yokoyama <hiroyuki.yokoyama.vx@renesas.com>
Signed-off-by: Mark Brown <broonie@kernel.org>