ASoC: wm_adsp: Rename memory fields in wm_adsp_buffer
The wm_adsp_buffer struct is the control header of a circular
buffer used to transfer data from the firmware over the
control interface to an ALSA compressed stream.
The original names of the fields pointing to the data buffer
were based on ADSP2V2 memory layout where they correspond to
{XM, XM, YM}. But this circular buffer could be used on other
types of DSP core that have different memory region types.
Also the names and description of the size fields were not
very clear. The field names and descriptions have been changed
to be generic and not imply any particular memory types.
This patch updates the wm_adsp driver to the new field names.
Marcel Ziswiler [Thu, 18 Oct 2018 11:18:28 +0000 (13:18 +0200)]
ASoC: soc-core: fix trivial checkpatch issues
Fix a few trivial aka cosmetic only checkpatch issues like long lines,
wrong indentations, spurious blanks and newlines, missing newlines,
multi-line comments etc.
Charles Keepax [Thu, 11 Oct 2018 16:28:28 +0000 (17:28 +0100)]
ASoC: dapm: Add support for hw_free on CODEC to CODEC links
Currently, on power down for a CODEC to CODEC DAI link we only call
digital_mute and shutdown. Provide a little more flexibility for drivers
by adding a call to hw_free as well.
R-Car datasheet is indicating that WS output settings of SSICR::SWSP
is inverted on TDM mode from non TDM mode settings.
But, it is meaning that TDM should use 0 here.
Without this patch, sound input/output 1ch will be 2ch, 2ch will be 3ch
..., be jumbled on I2S + TDM settings. This patch fixup it.
This patch is tested on R-Car H3 ulcb-kf board, SSI3/4 TDM sound.
Some SSIs are sharing each pins (= WS/CLK pin for playback/capture).
Then, SSI parent needs control WS/CLK setting for SSI slave.
In such case, SSI parent needs TDM settings if SSI slave is working as
TDM mode. But it is not cared in current driver.
It can't capture TDM sound without this patch if SSIs were pin sharing.
This patch is tested on R-Car H3 ulcb-kf board, SSI3/4 with TDM sound.
ASoC: pcm3168a: add hw constraint for capture channel
LEFT_J / I2S only can use TDM.
commit 594680ea4a394 ("ASoC: pcm3168a: add hw constraint for channel")
commit 3809688980205 ("ASoC: pcm3168a: add HW constraint for non
RIGHT_J") added channel constraint for it, but, it was only for playback.
This patch adds constraint for capture.
Daniel Mack [Wed, 17 Oct 2018 11:37:03 +0000 (13:37 +0200)]
ASoC: sta32x: Add support for XTI clock
The STA32x chips feature an XTI clock input that needs to be stable before
the reset signal is released. Therefore, the chip driver needs to get a
handle to the clock. Instead of relying on other parts of the system to
enable the clock, let the codec driver grab a handle itself.
In order to keep existing boards working, clock support is made optional.
ASoC: sunxi: Add new driver for Allwinner A64 codec's analog path controls
The internal codec on A64 is split into 2 parts. The analog path controls
are routed through an embedded custom register bus accessed through
the PRCM block.
Add an ASoC component driver for it. This should be tied to the codec
audio card as an auxiliary device.
ASoC: dt-binding: Add bindings for Allwinner A64 codec's analog path controls
The internal codec on Allwinner A64 is split into 2 parts. The
analog path controls are routed through an embedded custom register
bus accessed through the PRCM block just as on A23/A33/H3.
Marcus Cooper [Wed, 17 Oct 2018 07:38:05 +0000 (00:38 -0700)]
ASoC: sun4i-i2s: Add compatibility with A64 codec I2S
The I2S block used for the audio codec in the A64 differs from other 3
I2S modules in A64 and isn't compatible with H3. But it is very similar
to what is found in A10(sun4i). However, its TX FIFO is
located at a different address.
Akshu Agrawal [Mon, 15 Oct 2018 06:54:44 +0000 (12:24 +0530)]
ASoC: AMD: Add SND_JACK_LINEOUT jack type
Some 3 pole connectors report impedance greater than threshold of
1000Ohm. Thus, da7219 reports them as LINEOUT.
Adding the SND_JACK_LINEOUT type so that we don't fail to detect
any 3 pole jack type.
Also, changing
SND_JACK_HEADPHONE | SND_JACK_MICROPHONE -> SND_JACK_HEADSET
RIGHT_J only can handle 16bit data bits.
Current driver just errored if user requests non RIGHT_J
+ 16bit combination. But it is not useful for user.
This patch adds HW constraint for it, and avoid
error on such situation.
Daniel Mack [Thu, 11 Oct 2018 18:32:05 +0000 (20:32 +0200)]
ASoC: sta32x: set ->component pointer in private struct
The ESD watchdog code in sta32x_watchdog() dereferences the pointer
which is never assigned.
This is a regression from a1be4cead9b950 ("ASoC: sta32x: Convert to direct
regmap API usage.") which went unnoticed since nobody seems to use that ESD
workaround.
commit fb2815f44a9e ("ASoC: rsnd: add support for 16/24 bit slot widths")
added TDM width check, and return error if it was not 16/24/32 bit.
But it is too strict. This patch uses 32bit same as default.
Arnd Bergmann [Wed, 10 Oct 2018 08:37:13 +0000 (10:37 +0200)]
ASoC: max98988: add I2C dependency
max98988 only builds with I2C support enabled, otherwise we get a build error:
sound/soc/codecs/max98088.c:1789:1: error: data definition has no type or storage class [-Werror]
module_i2c_driver(max98088_i2c_driver);
^~~~~~~~~~~~~~~~~
sound/soc/codecs/max98088.c:1789:1: error: type defaults to 'int' in declaration of 'module_i2c_driver' [-Werror=implicit-int]
sound/soc/codecs/max98088.c:1789:1: error: parameter names (without types) in function declaration [-Werror]
sound/soc/codecs/max98088.c:1780:26: error: 'max98088_i2c_driver' defined but not used [-Werror=unused-variable]
Takashi Iwai [Thu, 4 Oct 2018 18:30:06 +0000 (20:30 +0200)]
ASoC: topology: Use the standard fall-through annotations
As a preparatory patch for the upcoming -Wimplicit-fallthrough
compiler checks, replace with the standard "fall through" annotation.
gcc can't understand the mixed texts, unfortunately.
Takashi Iwai [Thu, 4 Oct 2018 18:30:03 +0000 (20:30 +0200)]
ASoC: pcm186x: Use the standard fall-through annotation
As a preparatory patch for the upcoming -Wimplicit-fallthrough
compiler checks, replace with the standard "fall through" annotation.
Unfortunately gcc doesn't understand the mixed comment lines.
Takashi Iwai [Thu, 4 Oct 2018 18:30:02 +0000 (20:30 +0200)]
ASoC: adau1761: Use the standard fall-through annotation
As a preparatory patch for the upcoming -Wimplicit-fallthrough
compiler checks, replace with the standard "fall through" annotation
at the right place. It has to be put right before the next label.
Daniel Mack [Wed, 3 Oct 2018 19:32:34 +0000 (21:32 +0200)]
ASoC: add fault detect recovery property to DT bindings
The driver already has support for setting the FDRB bit in the CONFA
register through platform data, but there was no property to set it
in the device-tree bindings.
Dan Carpenter [Mon, 1 Oct 2018 16:44:30 +0000 (19:44 +0300)]
ASoC: qdsp6: q6asm-dai: checking NULL vs IS_ERR()
The q6asm_audio_client_alloc() doesn't return NULL, it returns error
pointers.
Fixes: 2a9e92d371db ("ASoC: qdsp6: q6asm: Add q6asm dai driver") Signed-off-by: Dan Carpenter <[email protected]> Signed-off-by: Mark Brown <[email protected]>
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 <[email protected]> Signed-off-by: Mark Brown <[email protected]>
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.
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.
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.
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.
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.
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.
ASoC: AMD: Fix capture unstable in beginning for some runs
alsa_conformance_test -C hw:0,4 -p 1024 --debug
would sometime show:
TIME_DIFF(s) HW_LEVEL READ RATE
0.000095970 1024 1024 10670001.041992
0.042609555 1024 2048 24032.168372
0.021330364 1024 3072 48006.681930
0.021339559 1024 4096 47985.996337
The issue is that in dma pointer function we can have stale value
of the register for current descriptor of channel.
The register retains the number of the last descriptor that
was transferred.
Fix ensures that we report position, 0, till the one period worth of
data is transferred. After one period of data, in handler of period
completion interrupt we update the config and correct value of descriptor
starts reflecting.
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.
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.
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.
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.
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.
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 <[email protected]> Signed-off-by: Yu Zhao <[email protected]> Signed-off-by: Mark Brown <[email protected]>
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.
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.
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.
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.
projects / linux.git / log