The playback DAI is connected to the DSP and the DSP might be sourcing
signals from the playback stream. Add a DAPM route between the two to make
sure that the playback DAI is powered up, when the DSP is active.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de> Signed-off-by: Alexandru Ardelean <alexandru.ardelean@analog.com> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: pxa: make SND_PXA_SOC_SSP depend on PLAT_PXA
For the moment, we can't enable CONFIG_SND_PXA_SOC_SSP unless we are
building for ARM PXA or MMP:
WARNING: unmet direct dependencies detected for PXA_SSP
Depends on [n]: PLAT_PXA [=n]
Selected by [y]:
- SND_PXA_SOC_SSP [=y] && SOUND [=y] && !UML && SND [=y] && SND_SOC [=y]
This adds an explicit dependency for it.
Fixes: 0a94cf345740 ("ASoC: pxa: make SND_PXA2XX_SOC_I2S selectable") Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: nau8824: use 64-bit arithmetic instead of 32-bit
Add suffix ULL to constant 256 in order to give the compiler complete
information about the proper arithmetic to use.
Notice that such constant is used in a context that expects an
expression of type u64 (64 bits, unsigned) and the following
expression is currently being evaluated using 32-bit arithmetic:
256 * fs * 2 * mclk_src_scaling[i].param
Addresses-Coverity-ID: 1432039 ("Unintentional integer overflow") Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Mack [Mon, 2 Jul 2018 15:11:00 +0000 (17:11 +0200)]
ASoC: pxa-ssp: add support for an external clock in devicetree
Allow setting a clock called 'extclk' in the device of the ssp-dai
device. If specified, this clock will be set to the mclk rate from the
DAI's .set_sysclk() callback. The DAI will also configure itself to
use that external clock.
Signed-off-by: Daniel Mack <daniel@zonque.org> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: fsl_spdif: Use 64-bit arithmetic instead of 32-bit
Add suffix ULL to constant 64 in order to give the compiler complete
information about the proper arithmetic to use.
Notice that such constant is used in a context that expects an
expression of type u64 (64 bits, unsigned) and the following
expression is currently being evaluated using 32-bit arithmetic:
rate[index] * txclk_df * 64
Addresses-Coverity-ID: 1222129 ("Unintentional integer overflow") Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Acked-by: Nicolin Chen <nicoleotsuka@gmail.com> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: nau8825: use 64-bit arithmetic instead of 32-bit
Add suffix ULL to constant 256 in order to give the compiler complete
information about the proper arithmetic to use.
Notice that such constant is used in a context that expects an
expression of type u64 (64 bits, unsigned) and the following
expression is currently being evaluated using 32-bit arithmetic:
256 * fs * 2 * mclk_src_scaling[i].param
Addresses-Coverity-ID: 1339616 ("Unintentional integer overflow") Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: core: Allow topology to override machine driver FE DAI link config.
Machine drivers statically define a number of DAI links that currently
cannot be changed or removed by topology. This means PCMs and platform
components cannot be changed by topology at runtime AND machine drivers
are tightly coupled to topology.
This patch allows topology to override the machine driver DAI link config
in order to reuse machine drivers with different topologies and platform
components. The patch supports :-
1) create new FE PCMs with a topology defined PCM ID.
2) destroy existing static FE PCMs
3) change the platform component driver.
4) assign any new HW params fixups.
5) assign a new card name prefix to differentiate this topology to userspace.
The patch requires no changes to the machine drivers, but does add some
platform component flags that the platform component driver can assign
before loading topologies.
Signed-off-by: Liam Girdwood <liam.r.girdwood@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
The generated clock (gclk) driver is able to set aclk as its parent and
change its rate alone, if needed. This means that our driver no longer
needs to configure aclk and we can let gclk select and configure its
clock source.
Signed-off-by: Codrin Ciubotariu <codrin.ciubotariu@microchip.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Drop AVDD in favor of PVDD to match the names used in the datasheet
and only claim PVDD on the es7154. The es7134 and es7144 don't have
a separate supply for the digital I/O.
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Kurtz [Mon, 2 Jul 2018 21:19:55 +0000 (15:19 -0600)]
ASoC: AMD: Simplify trigger handler
Now that the I2S channel names are fixed, and DMA data flow order is
consistent (ch1 then ch2), we can simplify channel start order:
start the upstream channel and then the downstream channel for both
playback and capture cases.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Kurtz [Mon, 2 Jul 2018 21:19:51 +0000 (15:19 -0600)]
ASoC: AMD: Reset bytescount when starting transaction
The pointer() callback gets its value by reading the I2S BYTE_COUNT
register. This is a 64-bit runnning transaction counter. If a
transaction was aborted in the middle of a sample buffer, the counter will
stop counting on a number divisible by the buffer size. Since we actually
use it as a pointer into an aligned buffer, however, we do want to ensure
that it always starts at a number divisible by the buffer size when
starting a transaction, hence we reset it whenever starting a transaction.
To accomplish this, it wasn't necessary to zero bytescount at the
termination of each transaction, so remove this unnecessary code.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Kurtz [Mon, 2 Jul 2018 21:19:54 +0000 (15:19 -0600)]
ASoC: AMD: Do not generate interrups for every captured sample
On capture, audio data is first copied from I2S to ACP memory, and then
from ACP to SYSRAM. The I2S_TO_ACP_DMA interrupt fires on every sample
transferred from I2S to ACP memory. That is it fires ~48000 times per
second when capturing @ 48 kHz. Since we don't do anything on this
interrupt anyway, disable it to save quite a few unnecessary interrupts.
The real "work" (calling snd_pcm_period_elapsed()) is done when transfer
from ACP to SYSRAM is complete.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Kurtz [Mon, 2 Jul 2018 21:19:53 +0000 (15:19 -0600)]
ASoC: AMD: Fix Capture DMA channel names
On capture, audio data is first copied from I2S to ACP memory, and then
to SYSRAM. For each step the channel number increases, so the names in
the driver were wrong.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Kurtz [Mon, 2 Jul 2018 21:19:52 +0000 (15:19 -0600)]
ASoC: AMD: Always subtract bytescount
It is always correct to subtract out the starting bytescount value. Even
in the case of 2^64 byte rollover (292 Million Years in the future
@ 48000 Hz) the math still works out.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Kurtz [Mon, 2 Jul 2018 21:19:50 +0000 (15:19 -0600)]
ASoC: AMD: Always stop ch2 first
Commit 6b116dfb4633a ("ASoC: AMD: make channel 1 dma as circular") made
both channels circular, so this comment and logic no longer applies. Always
stop ch2 (the channel closest to the output) before ch1. This ensures
that the downstream circular DMA channel does not continue to play/capture
repeated samples after the upstream circular DMA channel has already
stopped.
Signed-off-by: Daniel Kurtz <djkurtz@chromium.org> Signed-off-by: Mark Brown <broonie@kernel.org>
We are currently using power saving mode for button detection.
However, it will impact the headset recording performance.
This patch will switch button detection to normal mode in capture
and switch to power saving mode in the end of capture.
Signed-off-by: Bard Liao <bardliao@realtek.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Mack [Fri, 29 Jun 2018 12:59:40 +0000 (14:59 +0200)]
ASoC: pxa-ssp: remove .set_pll() and .set_clkdiv() callbacks
The .set_pll() and .set_clkdiv() callbacks are considered legacy and should
not be used anymore. In order to support PXA boards on DT platforms, remove
them and let the code figure out the correct dividers and PLL base
frequencies itself.
Signed-off-by: Daniel Mack <daniel@zonque.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Mack [Sat, 30 Jun 2018 20:24:33 +0000 (22:24 +0200)]
ASoC: pxa: select SND_PXA2XX_LIB for drivers that depend on it
Commit d767d3ce5c48b ("ASoC: pxa: provide PCM ops for ssp, i2s and ac97
components") created a build-time dependency to SND_PXA2XX_LIB but
missed to reflect that in Kconfig.
Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Daniel Mack <daniel@zonque.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Mack [Wed, 27 Jun 2018 19:33:57 +0000 (21:33 +0200)]
ASoC: pxa: provide PCM ops for ssp, i2s and ac97 components
Now that the functions are now available through pxa2xx-lib, hook them up
to pxa-sspi, pxa-ac97 and pxa-i2s. This allows DT platforms to use the DAIs
without a platform driver.
Signed-off-by: Daniel Mack <daniel@zonque.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Daniel Mack [Wed, 27 Jun 2018 19:33:53 +0000 (21:33 +0200)]
ASoC: fold pxa2xx-pcm into its only user, pxa2xx-ac97
Now that the PXA SSP bits are ported over to generic DMA, the pxa2xx-pcm
code only has a single user left. This patch folds the remaining bits into
its only user and removes the unnecessary glue layer along with its header
file.
The include dependency to linux/dma/pxa-dma.h is also gone now.
Signed-off-by: Daniel Mack <daniel@zonque.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Robert Jarzmik [Thu, 28 Jun 2018 20:08:37 +0000 (22:08 +0200)]
ASoC: pxa: remove the dmaengine compat need
As the pxa architecture switched towards the dmaengine slave map, the
old compatibility mechanism to acquire the dma requestor line number and
priority are not needed anymore.
This patch simplifies the dma resource acquisition, using the more
generic function dma_request_slave_channel().
Signed-off-by: Robert Jarzmik <robert.jarzmik@free.fr> Signed-off-by: Daniel Mack <daniel@zonque.org> Signed-off-by: Mark Brown <broonie@kernel.org>
Jerome Brunet [Wed, 27 Jun 2018 09:48:18 +0000 (11:48 +0200)]
ASoC: dpcm: extend channel merging to the backend cpu dai
Extend dpcm_merge_chan to also check backend cpu dai channels
capabilities. Apply the same policy as soc_pcm_init_runtime_hw() for
multicodec links and only check cpu dai in this case.
Cc: Jiada Wang <jiada_wang@mentor.com> Signed-off-by: Jerome Brunet <jbrunet@baylibre.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Jerome Brunet [Tue, 26 Jun 2018 10:07:25 +0000 (12:07 +0200)]
ASoC: dpcm: improve runtime update predictability
As it is, dpcm_runtime_update() performs the old path and new path
update of a frontend before going on to the next frontend DAI.
Depending the order of the FEs within the rtd list, the result of
the update might be different.
For example:
* Frontend A connected to backend C, with a 48kHz playback
* Frontend B connected to backend D, with a 44.1kHz playback
* FE A appears before FE B in the rtd list of the card.
If we reparent BE C to FE B (disconnecting BE D):
* old path update of FE A will run first, and BE C will get hw_free()
and shutdown()
* new path update of FE B will run after and BE C, which is stopped,
so it will be configured at 44.1kHz, as expected
If we reparent BE D to FE A (disconnecting BE C):
* new path update of FE A will run first but since BE D is still running
at 44.1kHz, it won't be reconfigured (no call to startup() or
hw_params())
* old path update of FE B runs after, nothing happens
* In this case, we end up with a BE playing at 44.1kHz a stream which is
supposed to be played at 48Khz (too slow)
To improve this situation, this patch performs all the FE old paths update
before going on to update the new paths. With this, the result should
no longer depend on the order of the FE within the card rtd list.
Please note that there might be a small performance penalty since
dpcm_process_paths() is called twice per stream direction.
Signed-off-by: Jerome Brunet <jbrunet@baylibre.com> Signed-off-by: Mark Brown <broonie@kernel.org>
As we get more entries in the DMI quirk table it is nice to have some
sort of ordering in the table, sort it alphabetically.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
One some models (Chuwi Vi8 Plus, Chuwi Hi8 Pro) the headphone output has
left and right swapped. This can be fixed in with special mixer settings
in the UCM profile, bit this requires these devices loading a different
UCM profile.
This commit adds a BYT_RT5651_HP_LR_SWAPPED quirk for this and postfixes
the longname with "-hp-swapped" if set, so that a different UCM profile
will be loaded.
We can safely do this without causing regressions (UCM profile not found
due to the longname change) as the UCM profiles are not in upstream
alsa-lib yet.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Now that the headset-mic is always IN3 there is no reason to have
the headset-mic mapping in the long-name.
This commit simplifies the long name to "bytcr-rt5651-<intmic-map>-mic".
We can safely do this without causing regressions (UCM profile not found
due to the longname change) as the UCM profiles are not in upstream
alsa-lib yet.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
The initial bytcr_rt5651 machine driver commit mapped IN2 as the headset
mic. In retrospect this is not correct as all known boards have the headset
mic on IN3.
This commit fixes the original DMIC mapping to correctly have the headset
mic on IN3.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
The initial bytcr_rt5651 machine driver commit mapped IN2 as the headset
mic. In retrospect this is not correct as all known boards have the headset
mic on IN3. To workaround this special IN?_HS_IN3 mappings were added.
This commit fixes the original IN1 mapping to correctly have the headset
mic on IN3, moves all users of the IN1_HS_IN3 mapping over to the fixed
IN1_MAP and drops the now no longer needed IN1_HS_IN3 mapping.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
BYT_RT5651_IN2_MAP was introduced in commit 39712db878a4 ("SoC: intel: byt:
Introduce new custom IN2 map"), uses in commit 2fe30129b0a6 ("ASoC: intel:
byt: Enable IN2 map quirk for a KIANO laptop"), only to be replaced by a
new BYT_RT5651_IN1_IN2_MAP quirk in commit ea261bd02a67 ("ASoC: intel:
byt: Introduce new map for dual mics") quickly afterwards, because the
KIANO laptop has 2 internal mics on IN1 and IN2 and the headset mic is
not in IN1 where the BYT_RT5651_IN2_MAP maps it, but on IN3.
Now that the KIANO quirk entry uses BYT_RT5651_IN1_IN2_MAP, there are no
users of BYT_RT5651_IN2_MAP left. This makes sense since the headset mic
seems to always be connected to IN3, so BYT_RT5651_IN2_MAP is not useful.
To deal with BYT_RT5651_IN2_MAP wrongly mapping the headset mic to IN1,
BYT_RT5651_IN2_HS_IN3_MAP was added in commit f026e0631780 ("ASoC: Intel:
bytcr_rt5651: Add new IN2_HS_IN3 input map and a quirk using it"). This
was based on the assumption then some devices have the internal mic
connected to IN2 only. Further testing has shown that this is wrong and the
internal mic is always connected to IN1 and sometimes to both IN1 and IN2.
TL;DR: Both BYT_RT5651_IN2_MAP and BYT_RT5651_IN2_HS_IN3_MAP are based on
on wrong assumptions from the past and are no longer useful now, so they
can both be removed.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Hans de Goede [Sun, 24 Jun 2018 14:06:28 +0000 (16:06 +0200)]
ASoC: Intel: bytcr_rt5651: Fix IN1_IN2_MAP quirk not being logged
Fix the quirk logging code not logging the IN1_IN2_MAP quirk.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Hans de Goede [Sun, 24 Jun 2018 14:06:27 +0000 (16:06 +0200)]
ASoC: Intel: bytcr_rt5651: Change default input map from in2 to in1
Further testing on all 6 model x86 tablets with a rt5651 which I have
access to for testing has shown that their single (mono) microphone is
connected to both IN1 *and* IN2.
The previous default mapping of IN2 was based on testing on the same 6
tablets, where the internal mic works fine with a mapping of IN2. But it
works fine too with a mapping of IN1.
This commit changes the default input mapping to to use IN1 instead of
IN2, to match the mapping used for the other mono devices in the DMI quirk
table. So that we need less different mappings.
The same change is made to the Chuwi Vi8 Plus quirks, which is one of the
6 models tested.
This is a preparation patch for simplifying the maps in a follow-up commit.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Almost all boards use the mclk and use the same jack-detect settings, add
a BYT_RT5651_DEFAULT_QUIRKS define for this.
This shaves of some lines and makes it easier to see which settings are
unique to a certain model.
Signed-off-by: Hans de Goede <hdegoede@redhat.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: qdsp6: q6afe: use of_platform_populate/depopulate()
Now that the child nodes have there own compatible strings,
Use of_platform_populate/depopulate() instead of less common
of_platform_device_create()/destroy().
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Acked-by: Niklas Cassel <niklas.cassel@linaro.org> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: qdsp6: q6asm: use of_platform_populate/depopulate()
Now that the child nodes have there own compatible strings,
Use of_platform_populate/depopulate() instead of less common
of_platform_device_create()/destroy().
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Acked-by: Niklas Cassel <niklas.cassel@linaro.org> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: qdsp6: q6adm: use of_platform_populate/depopulate()
Now that the child nodes have there own compatible strings,
Use of_platform_populate/depopulate() instead of less common
of_platform_device_create()/destroy().
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Acked-by: Niklas Cassel <niklas.cassel@linaro.org> Signed-off-by: Mark Brown <broonie@kernel.org>
In its current shape, the driver just ignores errors returned by
regulator_get() at component_probe(). This doesn't hurt on Amstrad
Delta board as long as it registers the codec device at late_initcall,
when the regulator which depends on basic-mmio-gpio device (probed as
late as at dev_initcall) is already available. Otherwise the driver
may end up trying to control a codec which is not powered up.
Remove that dependency on initialization order by handling the error.
If the regulator is not yet available and -ENODEV is returned, convert
it to -EPROBE_DEFER to get another chance.
Signed-off-by: Janusz Krzysztofik <jmkrzyszt@gmail.com> Signed-off-by: Mark Brown <broonie@kernel.org>
projects / mirror_ubuntu-hirsute-kernel.git / log