Xiaolei Wang [Mon, 10 Oct 2022 09:20:14 +0000 (17:20 +0800)]
ASoC: wm8962: Add an event handler for TEMP_HP and TEMP_SPK
In wm8962 driver, the WM8962_ADDITIONAL_CONTROL_4 is used as a volatile
register, but this register mixes a bunch of volatile status bits and a
bunch of non-volatile control bits. The dapm widgets TEMP_HP and
TEMP_SPK leverages the control bits in this register. After the wm8962
probe, the regmap will bet set to cache only mode, then a read error
like below would be triggered when trying to read the initial power
state of the dapm widgets TEMP_HP and TEMP_SPK.
wm8962 0-001a: ASoC: error at soc_component_read_no_lock
on wm8962.0-001a: -16
In order to fix this issue, we add event handler to actually power
up/down these widgets. With this change, we also need to explicitly
power off these widgets in the wm8962 probe since they are enabled
by default.
Mark Brown [Mon, 10 Oct 2022 14:33:47 +0000 (15:33 +0100)]
ASoC: wmxxxx: Revert old "ASoC: wmxxxx: Fix PM disable depth imbalance in wmxxxx_probe"
Merge series from Zhang Qilong <zhangqilong3@huawei.com>:
Both the old and new patch have been applied, it will resulted in
redundant calling of pm_runtime_disable when error returns. We Just
revert the old three patches to fix it.
Fix build errors when CONFIG_GPIOLIB is not enabled:
../sound/soc/codecs/tlv320adc3xxx.c: In function 'adc3xxx_i2c_probe':
../sound/soc/codecs/tlv320adc3xxx.c:1352:28: error: implicit declaration of function 'devm_gpiod_get'; did you mean 'devm_gpio_free'? [-Werror=implicit-function-declaration]
1352 | adc3xxx->rst_pin = devm_gpiod_get(dev, "reset", GPIOD_OUT_LOW);
../sound/soc/codecs/tlv320adc3xxx.c:1352:57: error: 'GPIOD_OUT_LOW' undeclared (first use in this function); did you mean 'GPIOF_INIT_LOW'?
1352 | adc3xxx->rst_pin = devm_gpiod_get(dev, "reset", GPIOD_OUT_LOW);
CC lib/dynamic_debug.o
../sound/soc/codecs/tlv320adc3xxx.c:1400:9: error: implicit declaration of function 'gpiod_set_value_cansleep'; did you mean 'gpio_set_value_cansleep'? [-Werror=implicit-function-declaration]
1400 | gpiod_set_value_cansleep(adc3xxx->rst_pin, 1);
Fixes: e9a3b57efd28 ("ASoC: codec: tlv320adc3xxx: New codec driver") Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Reported-by: kernel test robot <lkp@intel.com> Cc: Mark Brown <broonie@kernel.org> Cc: Liam Girdwood <lgirdwood@gmail.com> Cc: Ricard Wanderlof <ricardw@axis.com> Cc: Jaroslav Kysela <perex@perex.cz> Cc: Takashi Iwai <tiwai@suse.com> Cc: alsa-devel@alsa-project.org Link: https://lore.kernel.org/r/20221006235822.30074-1-rdunlap@infradead.org Signed-off-by: Mark Brown <broonie@kernel.org>
The reverted commit causes the following warnigs:
Runtime PM usage count underflow!
This is due to the fact that the pm_runtime_resume_and_get() is calling
pm_runtime_put_noidle() in case of < 0 return value of
pm_runtime_get_sync() which includes the -EACCES.
The change is wrong as -EACCES is returned in case of 'nested' get_sync()
and it is a valid use of PM runtime.
ASoC: wm5102: Fix PM disable depth imbalance in wm5102_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by calling
pm_runtime_disable when error returns.
ASoC: wm5110: Fix PM disable depth imbalance in wm5110_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by calling
pm_runtime_enable when error returns.
ASoC: wm8997: Fix PM disable depth imbalance in wm8997_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by calling
pm_runtime_disable when error returns.
ASoC: qcom: fix unmet direct dependencies for SND_SOC_QDSP6
SND_SOC_QDSP6 already has COMPILE_TEST so remove that from
SND_SOC_SC8280XP and also add QCOM_APR dependencies to
SND_SOC_SC8280XP like other Qualcomm machine drivers.
This should also fix below warning:
on x86_64, when QCOM_APR is not set and COMPILE_TEST=y:
ASoC: mediatek: mt8192-mt6359: Set the driver name for the card
The ASoC core automatically populates the driver name field in the card
from the card name if left unset. However, since the driver name can be
at most 16 characters long, wrapping will happen if the card name is
longer, which is the case for the mt8192-mt6359 driver.
Explicitly set the driver name for the card in order to avoid said
wrapping and have a readable driver name exposed to userspace.
Brent Lu [Tue, 13 Sep 2022 07:49:06 +0000 (15:49 +0800)]
ASoC: Intel: sof_rt5682: remove SOF_RT1015_SPEAKER_AMP_100FS flag
This flag could be removed since we now have API to query bclk
fequency setting in the topology. The dai link structure itself also
provides DAI format information instead of figuring it out with fs
number.
Mark Brown [Fri, 30 Sep 2022 08:15:58 +0000 (09:15 +0100)]
Fix PM disable depth imbalance in probe
Merge series from Zhang Qilong <zhangqilong3@huawei.com>:
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by moving
pm_runtime_enable to the endding of probe.
Zhang Qilong (4):
ASoC: wm8997: Fix PM disable depth imbalance in wm8997_probe
ASoC: wm5110: Fix PM disable depth imbalance in wm5110_probe
ASoC: wm5102: Fix PM disable depth imbalance in wm5102_probe
ASoC: mt6660: Fix PM disable depth imbalance in mt6660_i2c_probe
Jaroslav Kysela [Thu, 29 Sep 2022 14:37:54 +0000 (16:37 +0200)]
ASoC: core: clarify the driver name initialization
The driver field in the struct snd_ctl_card_info is a valid
user space identifier. Actually, many ASoC drivers do not care
and let to initialize this field using a standard wrapping method.
Unfortunately, in this way, this field becomes unusable and
unreadable for the drivers with longer card names. Also,
there is a possibility to have clashes (driver field has
only limit of 15 characters).
This change will print an error when the wrapping is used.
The developers of the affected drivers should fix the problem.
Signed-off-by: Jaroslav Kysela <perex@perex.cz> Reviewed-by: Nícolas F. R. A. Prado <nfraprado@collabora.com> Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: mt6660: Fix PM disable depth imbalance in mt6660_i2c_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by moving
pm_runtime_enable to the endding of mt6660_i2c_probe.
ASoC: wm5102: Fix PM disable depth imbalance in wm5102_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by moving
pm_runtime_enable to the endding of wm5102_probe.
ASoC: wm5110: Fix PM disable depth imbalance in wm5110_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by moving
pm_runtime_enable to the endding of wm5110_probe.
ASoC: wm8997: Fix PM disable depth imbalance in wm8997_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by moving
pm_runtime_enable to the endding of wm8997_probe
Audio tuner is used to handle clock drift between 26M and APLL domain.
It's expected when abs(chg_cnt) equals to upper bound, tuner updates pcw
setting automatically, and then abs(chg_cnt) decreases.
In the stress test, we found abs(chg_cnt) possibly equals to 2 at the
unexpected timing. This results in wrong pcw updating.
Finally, abs(chg_cnt) will always be larger than upper bound,
As a result, we update the upper bound to 3 to handle the corner case.
Mark Brown [Tue, 27 Sep 2022 18:58:57 +0000 (19:58 +0100)]
Fix PM disable depth imbalance in stm32 probe
Merge series from Zhang Qilong <zhangqilong3@huawei.com>:
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced. We fix it by moving pm_runtime_enable to the
endding of probe.
ASoC: stm: Fix PM disable depth imbalance in stm32_i2s_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by moving
pm_runtime_enable to the endding of stm32_i2s_probe.
ASoC: stm32: spdifrx: Fix PM disable depth imbalance in stm32_spdifrx_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by moving
pm_runtime_enable to the endding of stm32_spdifrx_probe.
ASoC: stm32: dfsdm: Fix PM disable depth imbalance in stm32_adfsdm_probe
The pm_runtime_enable will increase power disable depth. Thus
a pairing decrement is needed on the error handling path to
keep it balanced according to context. We fix it by moving
pm_runtime_enable to the endding of stm32_adfsdm_probe.
Fixes:98e500a12f934 ("ASoC: stm32: dfsdm: add pm_runtime support for audio")
Stefan Binding [Tue, 27 Sep 2022 12:14:40 +0000 (13:14 +0100)]
ASoC: cs42l42: Fallback to headphones for type detect
After tip sense detects a jack insertion, if automatic
type detection, and manual type detection fails, then
fall back to assume the jack connected belongs to
headphones.
Martin Povišer [Tue, 27 Sep 2022 11:34:25 +0000 (13:34 +0200)]
ASoC: apple: mca: Remove stale release of DMA channels
The commit 4ec8179c212f ("ASoC: apple: mca: Postpone requesting of DMA
channels") shuffled around with the requesting and releasing of DMA
channels. It left behind stale release calls from within
apple_mca_release, remove those now.
Martin Povišer [Tue, 27 Sep 2022 11:34:24 +0000 (13:34 +0200)]
ASoC: apple: mca: Trigger, not deassert, the peripheral reset
Replace the deassertion of the peripheral's shared reset with the
triggering of a pulse on it. This is what we should have been using all
along as the platform's custom is not leaving the reset asserted on
unused peripherals.
ASoC: Intel: Skylake: Replace zero-length arrays with DECLARE_FLEX_ARRAY() helper
Zero-length arrays are deprecated and we are moving towards adopting
C99 flexible-array members, instead. So, replace zero-length arrays
declarations in anonymous union with the new DECLARE_FLEX_ARRAY()
helper macro.
This helper allows for flexible-array members in unions.
Marek Vasut [Tue, 27 Sep 2022 00:20:04 +0000 (02:20 +0200)]
ASoC: st,stm32-sai: Document audio OF graph port
It is expected that the SAI subnodes would contain audio OF graph port
with endpoint to link it with the other side of audio link. Document
the port: property.
ASoC: SOF: control.h: Replace zero-length array with DECLARE_FLEX_ARRAY() helper
Zero-length arrays are deprecated and we are moving towards adopting
C99 flexible-array members, instead. So, replace zero-length arrays
declarations in anonymous union with the new DECLARE_FLEX_ARRAY()
helper macro.
This helper allows for flexible-array members in unions.
ASoC: MAINTAINERS: add bindings and APR to Qualcomm Audio entry
Extend the Qualcomm Audio maintainer entry to include sound related
bindings and the Qualcomm APR/GPR (Asynchronous/Generic Packet Router)
IPC driver, which is tightly related to the Audio DSP.
If CONFIG_REGMAP_SLIMBUS is not set,
make ARCH=x86_64 CROSS_COMPILE=x86_64-linux-gnu-,
will be failed, like this:
sound/soc/codecs/wcd934x.o: In function `wcd934x_codec_probe':
wcd934x.c:(.text+0x3310): undefined reference to `__regmap_init_slimbus'
make: *** [vmlinux] Error 1
Add select REGMAP_SLIMBUS to config SND_SOC_WCD934X.
Fixes: a61f3b4f476e ("ASoC: wcd934x: add support to wcd9340/wcd9341 codec") Signed-off-by: Ren Zhijie <renzhijie2@huawei.com> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Link: https://lore.kernel.org/r/20220926074042.13297-1-renzhijie2@huawei.com Signed-off-by: Mark Brown <broonie@kernel.org>
Mark Brown [Fri, 23 Sep 2022 19:46:25 +0000 (20:46 +0100)]
ASoC: SOF: ipc4/Intel: Improve and enable IPC error dump
Merge series from Peter Ujfalusi <peter.ujfalusi@linux.intel.com>:
On Intel platforms the registers for DSP communications are used differently,
the IPC dump information is not correct since important registers are not
printed and existing ones are used a bit differently for IPC4.
As a last step, enable the IPC timeout 'handling' and allow the printout of
the now usefull IPC dump.
ASoC: da7219: Fix an error handling path in da7219_register_dai_clks()
If clk_hw_register() fails, the corresponding clk should not be
unregistered.
To handle errors from loops, clean up partial iterations before doing the
goto. So add a clk_hw_unregister().
Then use a while (--i >= 0) loop in the unwind section.
The driver uses two statically ininitialized struct dma_slave_config,
but only one of them is initialized to zero. Initialize config_be to
zero as well to make sure that no fields are filled with random values.
Let the compiler do this instead of explicitly calling memset() which
makes it easier to read.
Mark Brown [Fri, 23 Sep 2022 16:53:06 +0000 (17:53 +0100)]
ASoC: mediatek: Allow separate handling of headphone and headset mic jack
Merge series from Nícolas F. R. A. Prado <nfraprado@collabora.com>:
This series allows the headphone and headset mic jack status to be
handled separately by userspace on MT8192, MT8195 and MT8186.
Changes based on commit d0508b4f1604 ("ASoC: rk3399_gru_sound: Add DAPM
pins, kcontrols for jack detection"). Found while searching for an
alternative for JackSwitch [1].
Peter Ujfalusi [Fri, 23 Sep 2022 13:36:16 +0000 (16:36 +0300)]
ASoC: SOF: ipc4: Call snd_sof_handle_fw_exception() in case of timeout
It can help debugging IPC timeout issues (like we do with IPC3) if we
dump the IPC and DSP information.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Link: https://lore.kernel.org/r/20220923133616.26267-7-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Peter Ujfalusi [Fri, 23 Sep 2022 13:36:15 +0000 (16:36 +0300)]
ASoC: SOF: Intel: hda: Only dump firmware registers for IPC3
The firmware register dump is IPC3 specific, it is not available for other
IPC versions.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Link: https://lore.kernel.org/r/20220923133616.26267-6-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Peter Ujfalusi [Fri, 23 Sep 2022 13:36:14 +0000 (16:36 +0300)]
ASoC: SOF: Intel: mtl: Print relevant register in ipc_dump
The use of the IPC registers are different between IPC3 and IPC4.
The ipc_dump needs to use different prints depending on the used IPC
protocol.
The existing code was printing registers relevant for IPC3, which is not
even supported on MTL.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Link: https://lore.kernel.org/r/20220923133616.26267-5-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Peter Ujfalusi [Fri, 23 Sep 2022 13:36:13 +0000 (16:36 +0300)]
ASoC: SOF: Intel: skl: Use the ipc4 version of the ipc_dump
The use of the IPC registers are different between IPC3 and IPC4.
The ipc_dump needs to use different prints depending on the used IPC
protocol.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Link: https://lore.kernel.org/r/20220923133616.26267-4-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Peter Ujfalusi [Fri, 23 Sep 2022 13:36:12 +0000 (16:36 +0300)]
ASoC: SOF: Intel: hda: Add separate ops for ipc_dump for IPC4
The use of the IPC registers are different between IPC3 and IPC4.
The ipc_dump needs to use different prints depending on the used IPC
protocol.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Link: https://lore.kernel.org/r/20220923133616.26267-3-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Peter Ujfalusi [Fri, 23 Sep 2022 13:36:11 +0000 (16:36 +0300)]
ASoC: SOF: Intel: cnl: Add separate ops for ipc_dump for IPC4
The use of the IPC registers are different between IPC3 and IPC4.
The ipc_dump needs to use different prints depending on the used IPC
protocol.
Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Rander Wang <rander.wang@intel.com> Link: https://lore.kernel.org/r/20220923133616.26267-2-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
tx_macro_tx_mixer_put() and tx_macro_dec_mode_put() currently returns zero
eventhough it changes the value.
Fix this, so that change notifications are sent correctly.
currently active_decimator[] is unsigned long however we store negative
values when there is no decimator setup -1.
This is first bug, and the second bug is that we do not check if the decimator
is valid before writing to register using decimator as offset in
CDC_TXn_TX_PATH_CTL()
Fix these both by making active_decimator as integer array and adding check
in tx_macro_digital_mute() before accessing CDC_TXn_TX_PATH_CTL() register.
Reset soundwire block on frame sync generation clock reset.
Without this we are hitting read/write timeouts randomly during
runtime pm. Along with this remove a swr_reset redundant flag.
Reset soundwire block on frame sync generation clock reset.
Without this we are hitting read/write timeouts randomly during
runtime pm. Along with this remove a swr_reset redundant flag.
ASoC: Intel: sof_da7219_mx98360a: Access num_codecs through dai_link
After commit 3989ade2d1e7 ("ASoC: soc.h: remove num_cpus/codecs"), the
following build error occurs:
sound/soc/intel/boards/sof_da7219_max98373.c:198:27: error: no member named 'num_codecs' in 'struct snd_soc_pcm_runtime'
for (j = 0; j < runtime->num_codecs; j++) {
~~~~~~~ ^
1 error generated.
This conversion was missed by the aforementioned change. Do it now to
fix the build error.
Fixes: 3989ade2d1e7 ("ASoC: soc.h: remove num_cpus/codecs") Signed-off-by: Nathan Chancellor <nathan@kernel.org> Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20220922153752.336193-1-nathan@kernel.org Signed-off-by: Mark Brown <broonie@kernel.org>
Gaosheng Cui [Fri, 23 Sep 2022 09:03:55 +0000 (17:03 +0800)]
ASoC: fsl: Remove unused inline function imx_pcm_dma_params_init_data()
The imx_pcm_dma_params_init_data() are no longer used since
commit c31da0b196f9 ("ASoC: imx-ssi: Remove unused driver"),
and the function is used to initialize some members of
"struct imx_dma_data", it's more readable to assign the value
directly, imx_pcm_dma_params_init_data is useless, so remove it.
ASoC: mediatek: mt8186-rt5682: Expose individual headset jack pins
The rt5682 codec is able to distinguish between two event types:
headphone insertion/removal and headset microphone insertion/removal.
However, currently, the mt8186-rt5682 driver exposes a single kcontrol
for the headset jack, so userspace isn't able to differentiate between
the two events.
Add a definition for the headset jack pins, so that a separate jack
kcontrol is created for each one, allowing userspace to track and handle
them individually.
ASoC: mediatek: mt8186-da7219: Expose individual headset jack pins
The da7219 codec is able to distinguish between two event types:
headphone insertion/removal and headset microphone insertion/removal.
However, currently, the mt8186-da7219 driver exposes a single kcontrol
for the headset jack, so userspace isn't able to differentiate between
the two events.
Add a definition for the headset jack pins, so that a separate jack
kcontrol is created for each one, allowing userspace to track and handle
them individually.
ASoC: mediatek: mt8186-da7219: Add headset widgets with switches
Add DAPM widgets for headphones and headset microphone, with matching
switches, to allow toggling these paths based on the jack connection
status.
Note that differently from others (mt8192, mt8195 and mt8186-rt5682),
the widget here is named "Headphones" (with an 's'), since "Headphone
Switch" was already registered by da7219.
ASoC: mediatek: mt8195: Expose individual headset jack pins
The rt5682 codec is able to distinguish between two event types:
headphone insertion/removal and headset microphone insertion/removal.
However, currently, the mt8195 ASoC driver exposes a single kcontrol
for the headset jack, so userspace isn't able to differentiate between
the two events.
Add a definition for the headset jack pins, so that a separate jack
kcontrol is created for each one, allowing userspace to track and handle
them individually.
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Link: https://lore.kernel.org/r/20220922235951.252532-3-nfraprado@collabora.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: mediatek: mt8192-mt6359: Expose individual headset jack pins
The rt5682 codec is able to distinguish between two event types:
headphone insertion/removal and headset microphone insertion/removal.
However, currently, the mt8192-mt6359 driver exposes a single kcontrol
for the headset jack, so userspace isn't able to differentiate between
the two events.
Add a definition for the headset jack pins, so that a separate jack
kcontrol is created for each one, allowing userspace to track and handle
them individually.
Signed-off-by: Nícolas F. R. A. Prado <nfraprado@collabora.com> Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com> Link: https://lore.kernel.org/r/20220922235951.252532-2-nfraprado@collabora.com Signed-off-by: Mark Brown <broonie@kernel.org>
Peter Ujfalusi [Wed, 21 Sep 2022 11:27:51 +0000 (14:27 +0300)]
ASoC: SOF: ipc4-topology: Free the ida when IPC fails in sof_ipc4_widget_setup()
The allocated ida needs to be freed up if the IPC message fails since
next time when we try again to set up the widget we are going to try to
allocate another ID and given enough tries, we are going to run out of
unique IDs.
Fixes: 711d0427c713 ("ASoC: SOF: ipc4-topology: move ida allocate/free to widget_setup/free") Signed-off-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20220921112751.9253-1-peter.ujfalusi@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
projects / thirdparty / linux.git / log
