This patch adds support to WSA8810/WSA8815 Class-D Smart Speaker
Amplifier. This Amplifier is primarily interfaced with SoundWire.
One WSA is used for mono speaker configuration and second one
would give stereo setup.
This patch is tested on SDM845 based DragonBoard DB845c and
Lenovo YOGA C630 Laptop based on SDM850 with WSA8815
speaker amplifiers.
This patch adds bindings for WSA8810/WSA8815 Class-D Smart Speaker
Amplifier. This Amplifier also has a simple thermal sensor for
over temperature and speaker protection.
Commit 62d5ae4cafb7 ("ASoC: max98090: save and restore SHDN when changing
sensitive registers") extended the code for handling many controls by
adding a custom put function to them. That new custom put function
properly handles relations between codec's hardware registers. However
they used card->dapm_mutex to properly serialize those operations. This
in turn triggers a lockdep warning about possible circular dependency.
Fix this by introducing a separate mutex only for serializing the SHDN
hardware register related operations.
This fixes the following lockdep warning observed on Exynos4412-based
Odroid U3 board:
======================================================
WARNING: possible circular locking dependency detected
5.5.0-rc5-next-20200107 #166 Not tainted
------------------------------------------------------
alsactl/1104 is trying to acquire lock: ed0d50f4 (&card->dapm_mutex){+.+.}, at: max98090_shdn_save+0x1c/0x28
but task is already holding lock: edb4b49c (&card->controls_rwsem){++++}, at: snd_ctl_ioctl+0xcc/0xbb8
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
ASoC: max98090: fix incorrect helper in max98090_dapm_put_enum_double()
Commit 62d5ae4cafb7 ("ASoC: max98090: save and restore SHDN when changing
sensitive registers") extended the code for handling "LTENL Mux", "LTENR
Mux", "LBENL Mux" and "LBENR Mux" controls by adding a custom
max98090_dapm_put_enum_double() function to them. However that function
used incorrect helper to get its component object. Fix this by using the
proper snd_soc_dapm_* helper.
This fixes the following NULL pointer exception observed on
Exynos4412-based Odroid U3 board:
8<--- cut here ---
Unable to handle kernel NULL pointer dereference at virtual address 000000b0
pgd = (ptrval)
[000000b0] *pgd=00000000
Internal error: Oops: 5 [#1] PREEMPT SMP ARM
Modules linked in:
CPU: 0 PID: 1104 Comm: alsactl Not tainted 5.5.0-rc5-next-20200107 #166
Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
PC is at __mutex_lock+0x54/0xb18
LR is at ___might_sleep+0x3c/0x2e0
...
Process alsactl (pid: 1104, stack limit = 0x(ptrval))
...
[<c0b49630>] (__mutex_lock) from [<c0b4a110>] (mutex_lock_nested+0x1c/0x24)
[<c0b4a110>] (mutex_lock_nested) from [<c0839b3c>] (max98090_shdn_save+0x1c/0x28)
[<c0839b3c>] (max98090_shdn_save) from [<c083a4f8>] (max98090_dapm_put_enum_double+0x20/0x40)
[<c083a4f8>] (max98090_dapm_put_enum_double) from [<c080d0e8>] (snd_ctl_ioctl+0x190/0xbb8)
[<c080d0e8>] (snd_ctl_ioctl) from [<c02cafec>] (ksys_ioctl+0x470/0xaf8)
[<c02cafec>] (ksys_ioctl) from [<c0101000>] (ret_fast_syscall+0x0/0x28)
...
---[ end trace 0e93f0580f4b9241 ]---
Fixes: 62d5ae4cafb7 ("ASoC: max98090: save and restore SHDN when changing sensitive registers") Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Reviewed-by: Tzung-Bi Shih <tzungbi@google.com> Link: https://lore.kernel.org/r/20200108115007.31095-1-m.szyprowski@samsung.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: max98090: fix incorrect helper in max98090_dapm_put_enum_double()
Commit 62d5ae4cafb7 ("ASoC: max98090: save and restore SHDN when changing
sensitive registers") extended the code for handling "LTENL Mux", "LTENR
Mux", "LBENL Mux" and "LBENR Mux" controls by adding a custom
max98090_dapm_put_enum_double() function to them. However that function
used incorrect helper to get its component object. Fix this by using the
proper snd_soc_dapm_* helper.
This fixes the following NULL pointer exception observed on
Exynos4412-based Odroid U3 board:
8<--- cut here ---
Unable to handle kernel NULL pointer dereference at virtual address 000000b0
pgd = (ptrval)
[000000b0] *pgd=00000000
Internal error: Oops: 5 [#1] PREEMPT SMP ARM
Modules linked in:
CPU: 0 PID: 1104 Comm: alsactl Not tainted 5.5.0-rc5-next-20200107 #166
Hardware name: SAMSUNG EXYNOS (Flattened Device Tree)
PC is at __mutex_lock+0x54/0xb18
LR is at ___might_sleep+0x3c/0x2e0
...
Process alsactl (pid: 1104, stack limit = 0x(ptrval))
...
[<c0b49630>] (__mutex_lock) from [<c0b4a110>] (mutex_lock_nested+0x1c/0x24)
[<c0b4a110>] (mutex_lock_nested) from [<c0839b3c>] (max98090_shdn_save+0x1c/0x28)
[<c0839b3c>] (max98090_shdn_save) from [<c083a4f8>] (max98090_dapm_put_enum_double+0x20/0x40)
[<c083a4f8>] (max98090_dapm_put_enum_double) from [<c080d0e8>] (snd_ctl_ioctl+0x190/0xbb8)
[<c080d0e8>] (snd_ctl_ioctl) from [<c02cafec>] (ksys_ioctl+0x470/0xaf8)
[<c02cafec>] (ksys_ioctl) from [<c0101000>] (ret_fast_syscall+0x0/0x28)
...
---[ end trace 0e93f0580f4b9241 ]---
Fixes: 62d5ae4cafb7 ("ASoC: max98090: save and restore SHDN when changing sensitive registers") Signed-off-by: Marek Szyprowski <m.szyprowski@samsung.com> Reviewed-by: Tzung-Bi Shih <tzungbi@google.com> Link: https://lore.kernel.org/r/20200108115007.31095-1-m.szyprowski@samsung.com Signed-off-by: Mark Brown <broonie@kernel.org>
Wei Yongjun [Wed, 8 Jan 2020 03:59:54 +0000 (03:59 +0000)]
ASoC: amd: acp3x: Fix return value check in acp3x_dai_probe()
In case of error, the function devm_ioremap() returns NULL pointer not
ERR_PTR(). The IS_ERR() test in the return value check should be
replaced with NULL test.
Arnd Bergmann [Tue, 7 Jan 2020 21:48:35 +0000 (22:48 +0100)]
ASoC: gtm601: fix build warning
The driver produces warnings without CONFIG_OF, and makes
no sense without it either:
sound/soc/codecs/gtm601.c:50:34: error: 'bm818_dai' defined but not used [-Werror=unused-variable]
static struct snd_soc_dai_driver bm818_dai = {
^~~~~~~~~
sound/soc/codecs/gtm601.c:32:34: error: 'gtm601_dai' defined but not used [-Werror=unused-variable]
static struct snd_soc_dai_driver gtm601_dai = {
^~~~~~~~~~
ASoC: SOF: pci: change the default firmware path when the community key is used
Since ApolloLake, Intel platforms require signed firmware. On all
Windows platforms the default is to require the Intel production key
be used. But some platforms allow for a community key to be used,
which allows developers to compile/build their own firmware.
In the linux-firmware tree, the default intel/sof path is used for
firmwares signed for the production key, and files signed with the
community key are located in intel/sof/community.
Since we don't have an API to query which key is used on what
platforms, we have to rely on DMI-based quirks.
Developers can bypass this mechanism by setting a kernel 'fw_path'
module parameter. Additional dynamic debug traces are provided to help
debug cases where the wrong file might be used.
Sam McNally [Fri, 3 Jan 2020 01:50:19 +0000 (12:50 +1100)]
ASoC: Intel: sof_rt5682: Ignore the speaker amp when there isn't one.
Some members of the Google_Hatch family include a rt5682 jack codec, but
no speaker amplifier. This uses the same driver (sof_rt5682) as a
combination of rt5682 jack codec and max98357a speaker amplifier. Within
the sof_rt5682 driver, these cases are not currently distinguishable,
relying on a DMI quirk to decide the configuration. This causes an
incorrect configuration when only the rt5682 is present on a
Google_Hatch device.
For CML, the jack codec is used as the primary key when matching,
with a possible speaker amplifier described in quirk_data. The two cases
of interest are the second and third 10EC5682 entries in
snd_soc_acpi_intel_cml_machines[]. The second entry matches the
combination of rt5682 and max98357a, resulting in the quirk_data field
in the snd_soc_acpi_mach being non-null, pointing at
max98357a_spk_codecs, the snd_soc_acpi_codecs for the matched speaker
amplifier. The third entry matches just the rt5682, resulting in a null
quirk_data.
The sof_rt5682 driver's DMI data matching identifies that a speaker
amplifier is present for all Google_Hatch family devices. Detect cases
where there is no speaker amplifier by checking for a null quirk_data in
the snd_soc_acpi_mach and remove the speaker amplifier bit in that case.
ASoC: amd: Added ACP3x system resume and runtime pm
When system wide suspend happens, ACP will be powered off
and when system resumes,for audio usecase to continue,all
the runtime configuration data needs to be programmed again.
Added resume pm call back to ACP pm ops and also added runtime
PM operations for ACP3x PCM platform device.
Device will enter into D3 state when there is no activity
on audio I2S lines.
This patch adds I2S SP support in ACP PCM DMA and DAI.
Added I2S support in DMA and DAI probe,its hw_params handling
its open and close functionalities.
This enables to open and close on the SP instance for
playback and capture.
ASoC: PCM DMA driver should only have dma ops.
So Removed all DAI related functionality.Refactoring
the PCM DMA diver code.Added new file containing only DAI ops
ASoC: cs47l92: Simplify error handling code in 'cs47l92_probe()'
If 'madera_init_bus_error_irq()' fails,
'wm_adsp2_remove(&cs47l92->core.adsp[0])' will be called twice.
Once in the 'if' block, and once in the error handling path.
This is harmless, but one of this call can be axed.
ASoC: qcom: dt-bindings: Add compatible for DB845c and Lenovo Yoga
This patch adds compatible strings for DB845c and Lenovo Yoga C630
soundcard. Based on this compatible strings common machine driver
will be in better position to setup board specific configuration.
ASoC: wcd934x: add support to wcd9340/wcd9341 codec
Qualcomm WCD9340/WCD9341 Codec is a standalone Hi-Fi audio codec IC,
It supports both I2S/I2C and SLIMbus audio interfaces.
On slimbus interface it supports two data lanes; 16 Tx ports
and 8 Rx ports. It has Five DACs and seven dedicated interpolators,
Seven (six audio ADCs, and one VBAT ADC), Multibutton headset
control (MBHC), Active noise cancellation, Sidetone paths,
MAD (mic activity detection) and codec processing engine.
It supports Class-H differential earpiece out and stereo single
ended headphones out.
This codec also has integrated SoundWire controller.
This patchset adds very basic support for playback and capture
via the interpolators and ADC respectively.
YueHaibing [Tue, 24 Dec 2019 14:02:37 +0000 (22:02 +0800)]
ASoC: Intel: kbl_da7219_max98357a: remove unused variable 'constraints_16000' and 'ch_mono'
sound/soc/intel/boards/kbl_da7219_max98357a.c:343:48:
warning: constraints_16000 defined but not used [-Wunused-const-variable=]
sound/soc/intel/boards/kbl_da7219_max98357a.c:348:27:
warning: ch_mono defined but not used [-Wunused-const-variable=]
Daniel Baluta [Fri, 20 Dec 2019 17:05:31 +0000 (11:05 -0600)]
ASoC: SOF: imx8: Fix dsp_box offset
dsp_box is used to keep DSP initiated messages. The value of dsp_offset
is set by the DSP with the first message, so we need a way to bootstrap
it in order to get the first message.
We do this by setting the correct default dsp_box offset which on i.MX8
is not zero.
Very interesting is why it has worked until now.
On i.MX8, DSP communicates with ARM core using a shared SDRAM memory
area. Actually, there are two shared areas:
* SDRAM0 - starting at 0x92400000, size 0x800000
* SDRAM1 - starting at 0x92C00000, size 0x800000
SDRAM0 keeps the data sections, starting with .rodata. By chance
fw_ready structure was placed at the beginning of .rodata.
dsp_box_base is defined as SDRAM0 + dsp_box_offset and it is placed
at the beginning of SDRAM1 (dsp_box_offset should be 0x800000). But
because it is zero initialized by default it points to SDRAM0 where
by chance the fw_ready was placed in the SOF firmware.
Anyhow, SOF commit 7466bee378dd811b ("clk: make freq arrays constant")
fw_ready is no longer at the beginning of SDRAM0 and everything shows
how lucky we were until now.
Fix this by properly setting the default dsp_box offset.
Dragos Tarcatu [Wed, 4 Dec 2019 21:04:47 +0000 (15:04 -0600)]
ASoC: topology: Prevent use-after-free in snd_soc_get_pcm_runtime()
remove_link() is currently calling snd_soc_remove_dai_link() after
it has already freed the memory for the link name. But this is later
read from snd_soc_get_pcm_runtime() causing a KASAN use-after-free
warning. Reorder the cleanups to fix this issue.
Olivier Moysan [Wed, 4 Dec 2019 15:43:33 +0000 (16:43 +0100)]
ASoC: stm32: spdifrx: fix input pin state management
Changing input state in iec capture control is not safe,
as the pin state may be changed concurrently by ASoC
framework.
Remove pin state handling in iec capture control.
Note: This introduces a restriction on capture control,
when pin sleep state is defined in device tree. In this case
channel status can be captured only when an audio stream
capture is active.
Olivier Moysan [Wed, 4 Dec 2019 15:43:31 +0000 (16:43 +0100)]
ASoC: stm32: spdifrx: fix inconsistent lock state
In current spdifrx driver locks may be requested as follows:
- request lock on iec capture control, when starting synchronization.
- request lock in interrupt context, when spdifrx stop is called
from IRQ handler.
Take lock with IRQs disabled, to avoid the possible deadlock.
Takashi Iwai [Wed, 4 Dec 2019 15:14:54 +0000 (16:14 +0100)]
ASoC: core: Fix access to uninitialized list heads
The error path of soc_new_pcm_runtime() invokes soc_free_pcm_runtime()
that may cause a few problems. First off, it calls list_del() for
rtd->list that may not be initialized. Similarly,
snd_soc_pcm_component_free() traverses over the component list that
may not be initialized, either. Such access to the uninitialized list
head would lead to either a BUG_ON() or a memory corruption.
This patch fixes the access to uninitialized list heads by
initializing the list heads properly at the beginning before those
error paths.
Daniel Baluta [Wed, 4 Dec 2019 15:13:33 +0000 (17:13 +0200)]
ASoC: soc-core: Set dpcm_playback / dpcm_capture
When converting a normal link to a DPCM link we need
to set dpcm_playback / dpcm_capture otherwise playback/capture
streams will not be created resulting in errors like this:
[ 36.039111] sai1-wm8960-hifi: ASoC: no backend playback stream
Colin Ian King [Wed, 4 Dec 2019 12:48:16 +0000 (12:48 +0000)]
ASoC: SOF: imx8: fix memory allocation failure check on priv->pd_dev
The memory allocation failure check for priv->pd_dev is incorrectly
pointer checking priv instead of priv->pd_dev. Fix this.
Addresses-Coverity: ("Logically dead code") Fixes: 202acc565a1f ("ASoC: SOF: imx: Add i.MX8 HW support") Signed-off-by: Colin Ian King <colin.king@canonical.com> Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com> Acked-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20191204124816.1415359-1-colin.king@canonical.com Signed-off-by: Mark Brown <broonie@kernel.org>
The current structures are not well designed. We include Xtensa
information from the ACPI and PCI levels, but at the Kconfig/module
level everything Xtensa related is included at the sof/intel level.
Move the arch_ops under ops so that Xtensa is hidden in the DSP ops,
with a structure that follows the Kconfig/module partition.
Jerome Brunet [Wed, 18 Dec 2019 17:24:20 +0000 (18:24 +0100)]
ASoC: meson: axg-fifo: relax period size constraints
Now that the fifo depths and thresholds are properly in the axg-fifo
driver, we can relax the constraints on period. As long as the period is a
multiple of the fifo burst size (8 bytes) things should be OK.
Jerome Brunet [Wed, 18 Dec 2019 17:24:19 +0000 (18:24 +0100)]
ASoC: meson: axg-fifo: improve depth handling
Let the fifo driver parse the fifo depth from DT. Eventually all DT should
have this property. Until it is actually the case, default to 256 bytes if
the property is missing. 256 bytes is the size of the smallest fifo on the
supported SoCs.
On the supported SoC, fifo A is usually bigger than the other ones. With
depth known, we can improve the usage of the fifo and adapt the setup of
request threshold.
Jerome Brunet [Wed, 18 Dec 2019 17:24:17 +0000 (18:24 +0100)]
ASoC: meson: axg-fifo: fix fifo threshold setup
On TODDR sm1, the fifo threshold register field is slightly different
compared to the other SoCs. This leads to the fifo A being flushed to
memory every 8kB. If the period is smaller than that, several periods
are pushed to memory and notified at once. This is not ideal.
Fix the register field update. With this, the fifos are flushed every
128B. We could still do better, like adapt the threshold depending on
the period size, but at least it consistent across the different
SoC/fifos
Dragos Tarcatu [Wed, 18 Dec 2019 00:05:18 +0000 (18:05 -0600)]
ASoC: topology: Prevent use-after-free in snd_soc_get_pcm_runtime()
remove_link() is currently calling snd_soc_remove_pcm_runtime() after
it has already freed the memory for the link name. But this is later
read from snd_soc_get_pcm_runtime() causing a KASAN use-after-free
warning. Reorder the cleanups to fix this issue.
Jerome Brunet [Wed, 18 Dec 2019 17:24:18 +0000 (18:24 +0100)]
ASoC: meson: axg-fifo: add fifo depth to the bindings documentation
Add a new property with the depth of the fifo in bytes. This is useful
since some instance of the fifo, even on the same SoC, may have different
depth. The depth is useful is set some parameters of the fifo.
Guido Roncarolo [Wed, 18 Dec 2019 00:26:15 +0000 (18:26 -0600)]
ASoC: SOF: imx: Describe SAI parameters to be sent to DSP
Introduce sof_ipc_dai_sai_params to keep information that
we get from topology and we send to DSP FW.
For the moment it is identical to ESAI one but it will
evolve shortly independently
Fix typo in Kconfig dependencies. The correct dependency
for HDMI is SND_SOC_SOF_HDA_AUDIO_CODEC.
Reported-by: Yong Zhi <yong.zhi@intel.com> Fixes: e3d8f8ae5b1e ("ASoC: Intel: boards: make common HDMI driver the default for SOF") Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20191218002616.7652-7-pierre-louis.bossart@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Kai Vehmanen [Wed, 18 Dec 2019 00:26:13 +0000 (18:26 -0600)]
ASoC: SOF: Intel: drop HDA codec upon probe failure
In case a HDA codec probe fails, do not raise error immediately,
but instead remove the codec from bus->codec_mask and continue
probe for other codecs.
This allows for more robust behaviour in cases where one codec
in the system is faulty. SOF driver load can still proceed with
the codecs that can be probed successfully. Probe may still
fail if suitable machine driver is not found, but in many
cases the generic HDA machine driver can operate with a subset
of codecs.
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20191218002616.7652-6-pierre-louis.bossart@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Karol Trzcinski [Wed, 18 Dec 2019 00:26:10 +0000 (18:26 -0600)]
ASoC: SOF: define struct with compiler name and version
Add compiler information structure sof_ipc_cc_version.
Add new enum value in sof_ipc_ext_data for new structure.
This struct will be used to show more information about firmware
in host system. It will be helpful during debugging.
Add a state machine for FW boot to track the
different stages of FW boot and replace the boot_complete
field with fw_state field in struct snd_sof_dev.
This will be used to determine the actions to be performed
during system suspend.
One of the main motivations for adding this change is the
fact that errors during the top-level SOF device probe cannot
be propagated and therefore suspending the SOF device normally
during system suspend could potentially run into errors.
For example, with the current flow, if the FW boot failed
for some reason and the system suspends, the SOF device
suspend could fail because the CTX_SAVE IPC would be attempted
even though the FW never really booted successfully causing it
to time out. Another scenario that the state machine fixes
is when the runtime suspend for the SOF device fails and
the DSP is powered down nevertheless, the CTX_SAVE IPC during
system suspend would timeout because the DSP is already
powered down.
Reviewed-by: Curtis Malainey <cujomalainey@chromium.org> Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com> Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://lore.kernel.org/r/20191218002616.7652-2-pierre-louis.bossart@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: SOF: Intel: hda: hda-dai: fix oops on hda_link .hw_free
When the PCM_PARAM IPC fails while configuring the FE, the kernel
oopses in the HDaudio link DMA .hw_free operation. The root cause is a
NULL dma_data since the BE .hw_params was never called by the SOC
core.
This error can also happen if the HDaudio link DMA configuration IPC
fails in the BE .hw_params.
This patches makes sure the dma_data is properly saved in .hw_params,
and tested before being use in hw_free.
Kai Vehmanen [Wed, 18 Dec 2019 00:05:16 +0000 (18:05 -0600)]
ASoC: SOF: fix fault at driver unload after failed probe
If sof_machine_check() fails during driver probe, the IPC
state is not initialized and this will lead to a NULL
dereference at driver unload. Example log is as follows:
Cezary Rojewski [Tue, 17 Dec 2019 09:58:50 +0000 (10:58 +0100)]
ASoC: compress: Add pm_runtime support
For some devices, components need to be powered-up before stream startup
sequence commences. Update soc_compr_open to provide such functionality.
Based on soc_pcm_open. Adjust soc_compr_free accordingly to power down
components once compress stream is closed.
ASoC: soc-core: support snd_soc_dai_link_component for codec_conf
To find codec_conf component, it is using dev_name, of_node.
But, we already has this kind of finding component method by
snd_soc_dai_link_component, and snd_soc_is_matching_component().
We shouldn't have duplicate implementation to do same things.
This patch adds snd_soc_dai_link_component support to find
codec_conf component.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Link: https://lore.kernel.org/r/87lfrh59kj.wl-kuninori.morimoto.gx@renesas.com Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Mark Brown <broonie@kernel.org>
Takashi Iwai [Wed, 11 Dec 2019 17:20:19 +0000 (18:20 +0100)]
ASoC: SOF: Use managed buffer allocation
Clean up the drivers with the new managed buffer allocation API.
The superfluous snd_pcm_lib_malloc_pages() and
snd_pcm_lib_free_pages() calls are dropped.
Cc: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Cc: Ranjani Sridharan <ranjani.sridharan@linux.intel.com> Signed-off-by: Takashi Iwai <tiwai@suse.de> Link: https://lore.kernel.org/r/20191211172019.23206-2-tiwai@suse.de Signed-off-by: Mark Brown <broonie@kernel.org>
Takashi Iwai [Wed, 11 Dec 2019 17:20:18 +0000 (18:20 +0100)]
ASoC: mediatek: Use managed buffer allocation
Clean up the drivers with the new managed buffer allocation API.
The superfluous snd_pcm_lib_malloc_pages() and
snd_pcm_lib_free_pages() calls are dropped, as well as the superfluous
snd_pcm_lib_preallocate_free_for_all() call. As of the result,
hw_free and pcm_destruct ops became empty and got removed.
Due to power rail dependencies, the SoundWire Master driver cannot
make decisions on its own when entering pm runtime suspend.
Add quirk mask for each link, so that the SOF parent driver can inform
the SoundWire master driver of the desired behavior:
a) leave clock on
b) power-off instead of clock stop
c) power-off if all devices cannot generate wakes
d) force bus reset on clock restart
Note that for now the interface with the SOF driver relies on a single
mask for all links. If needed, the interface might be modified at a
later point to provide more freedom. The code at the lower level does
not assume any commonality between links.
soundwire: intel: add mutex for shared SHIM register access
Some of the Intel SoundWire SHIM registers contain fields for
different links. Without protection, the master drivers for the
different links will access these shared registers, leading to invalid
configurations and timeouts (specifically when changing CPA/SPA
power-related registers and polling for the changes to be applied).
A mutex is added to make sure all rmw access to those registers are
serialized.
Bard Liao [Thu, 12 Dec 2019 01:45:04 +0000 (19:45 -0600)]
soundwire: intel: add link_list to handle interrupts with a single thread
In MSI mode, the use of separate handlers and threads for the Intel
IPC, stream and SoundWire shared interrupt leads to timeouts and lost
interrupts.
The solution is to merge all interrupt handling across all links with
a single thread function. The use of a linked list enables this thread
function to walk through all contexts and figure out which link needs
attention.
projects / mirror_ubuntu-hirsute-kernel.git / log