When creating topology templates and overriding data in specific test
cases it should be done with cpu_to_le32 macro, so we operate on correct
data on all architectures, as topology parser use le32_to_cpu to parse
data from structures.
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Amadeusz Sławiński <amadeuszx.slawinski@linux.intel.com>
Link: https://lore.kernel.org/r/20210202163123.3942040-1-amadeuszx.slawinski@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
DT properties "dmic-mode" and "mic-type-X" are optional. Reduces the
log verbosity and changes the message a bit to avoid misleading.
Signed-off-by: Tzung-Bi Shih <tzungbi@google.com>
Link: https://lore.kernel.org/r/20210202033557.1621029-1-tzungbi@google.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Audio Graph Card based Tegra driver is only useful on NVIDIA Tegra SoCs.
Hence add a dependency on SND_SOC_TEGRA, to prevent asking the user
about this driver when configuring a kernel without Tegra sound support.
Wrap all Tegra sound config options inside a big if/endif block, instead
of just adding the dependency to the single config option that does not
have it yet, to preventing similar future mistakes.
Fixes: 202e2f7745 ("ASoC: tegra: Add audio graph based card driver")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Acked-by: Sameer Pujar <spujar@nvidia.com>
Link: https://lore.kernel.org/r/20210129125915.2652952-1-geert+renesas@glider.be
Signed-off-by: Mark Brown <broonie@kernel.org>
microblaze-linux-gcc (GCC) 9.3.0 complains about missing __ffssi2
symbol while using __builtin_ffs at runtime.
This is because arch/h8300 is compiled with -fno-builtin option.
so fallback and use kernel ffs() instead to all the arch builds happy!
Fixes: 1da0b9899a ("ASoC: soc-component: add snd_soc_component_read/write_field()")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20210129100539.23459-1-srinivas.kandagatla@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Series to refactor the DSP core management:
- move tracking of powered up DSP cores to common SOF code
- add logic filter unnecessary power actions
- modify existing implementations to use common code
whenever DSP cores are powered, so the state in common
code stays in sync
Bard Liao (5):
ASoC: SOF: Intel: hda: use snd_sof_dsp_core_power_up/down API
ASoC: SOF: Intel: hda-loader: keep init cores alive
ASoC: SOF: update dsp core power status in common APIs
ASoC: SOF: Filter out unneeded core power up/downs
ASoC: SOF: intel: hda-loader: use snd_sof_dsp_core_power_down/up APIs
sound/soc/sof/intel/hda-dsp.c | 2 +-
sound/soc/sof/intel/hda-loader.c | 9 +++++----
sound/soc/sof/intel/hda.c | 2 +-
sound/soc/sof/loader.c | 6 ------
sound/soc/sof/ops.h | 24 ++++++++++++++++++------
sound/soc/sof/pm.c | 1 -
sound/soc/sof/topology.c | 8 --------
7 files changed, 25 insertions(+), 27 deletions(-)
base-commit: e32df14235
--
2.29.2
The snd-soc-acpi-intel-match has duplicate module tags for all
platforms separately. Remove all but one and save some storage
space and cleanup modinfo output.
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Rander Wang <rander.wang@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Link: https://lore.kernel.org/r/20210128105751.1049837-1-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
To manage enabled_cores_mask flag, we should always use snd_sof_dsp_
core_power_down/up APIs to power up/down dsp cores. The APIs do
a little bit more than the original functions, but it is harmless.
Suggested-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20210128093850.1041387-6-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Exclude cores that are already powered on/off correctly. This allows to
simplify dsp_power_up/down() implementations and avoid unexpected error.
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20210128093850.1041387-5-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Only manage enabled_cores_mask in common snd_sof_dsp_core_power_up/down
APIs to ensure it stays in sync with actual DSP core state.
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20210128093850.1041387-4-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
init_core_mask should be the available cores mask after fw boot. So we
should keep not core 0 but init cores alive.
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20210128093850.1041387-3-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
To implement common logic in SOF core, core power up/down flows should
use common SOF API and not directly use low-level platform specific
helper functions.
Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
Reviewed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20210128093850.1041387-2-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
ADL will use sof-adl-s.ri if it is ADL-S platform. So let's use
the default_fw_filename in pdata->desc for the ADL FW filename.
Signed-off-by: Libin Yang <libin.yang@intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Bard Liao <bard.liao@intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20210125070500.807474-3-kai.vehmanen@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
It's often the case that we would write or read a particular field
in register. With the current soc_component apis, reading a particular
field in register would involve first read the register and then
perform shift operations.
Ex:
to read from a field mask of 0xf0
val = snd_soc_component_read(component, reg);
field = ((val & 0xf0) >> 0x4);
This is sometimes prone to errors and code become less readable!
With this new api we could just do
field = snd_soc_component_read_field(component, reg, 0xf0);
this makes it bit simple, easy to write and less error prone!
This also applies to writing!
There are various places in kernel which provides such field interfaces
however soc_component seems to be missing this.
This patch is inspired by FIELD_GET/FIELD_PREP macros in include/linux/bitfield.h
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20210126171749.1863-1-srinivas.kandagatla@linaro.org
Signed-off-by: Mark Brown <broonie@kernel.org>
Enabling DMI L1 for capture streams could result in xruns during
pause/release. As pause/release is not a valid scenario for trace,
we can safely enable DMI L1 for it.
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/20210127020737.1088960-3-ranjani.sridharan@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
DMI L1 entry is currently disabled whenever any capture stream is
opened to prevent xruns during pause/release. But, in
order to maximise power savings for the wake-on-voice usecase,
DMI L1 entry should be enabled for D0i3-compatible capture streams.
Introduce a new field, flags in struct sof_intel_hda_stream
that stores whether a stream is dmi_l1_compatible. All playback streams,
and D0i3-compatible capture streams are DMI L1 compatible.
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Link: https://lore.kernel.org/r/20210127020737.1088960-2-ranjani.sridharan@linux.intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Assert hardware resets before clocks are enabled and then de-assert them
after clocks are enabled. This brings hardware into a predictable state.
Tested-by: Peter Geis <pgwipeout@gmail.com> # Ouya T30 audio works
Tested-by: Matt Merhar <mattmerhar@protonmail.com> # Ouya T30 boot-tested
Tested-by: Dmitry Osipenko <digetx@gmail.com> # Nexus7 T30 audio works
Tested-by: Nicolas Chauvet <kwizart@gmail.com> # TK1 boot-tested
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/20210120003154.26749-7-digetx@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
AHUB driver misses D_AUDIO and APBIF resets. CPU hangs on trying to
access hardware if resets aren't de-asserted. This problem is currently
masked by the tegra-clk driver which implicitly de-asserts the resets when
the corresponding clocks are enabled. Soon the implicit de-assertion will
be gone from the tegra-clk driver, thus we need to fix the AHUB driver.
Add the missing resets to the driver.
Tested-by: Peter Geis <pgwipeout@gmail.com> # Ouya T30 audio works
Tested-by: Matt Merhar <mattmerhar@protonmail.com> # Ouya T30 boot-tested
Tested-by: Dmitry Osipenko <digetx@gmail.com> # Nexus7 T30 audio works
Tested-by: Nicolas Chauvet <kwizart@gmail.com> # TK1 boot-tested
Signed-off-by: Dmitry Osipenko <digetx@gmail.com>
Link: https://lore.kernel.org/r/20210120003154.26749-5-digetx@gmail.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Topology API exposes just 2 function calls, to load and unload topology.
This adds sanity checks to make sure that it behaves well when some of
parameters are set incoorectly or not needed.
This makes developer live easier by failing early instead of proceeding
on and then failing in unexpected ways.
As loading and unloading topology usually happens one the overhead of
additional checks should be negligible.
Amadeusz Sławiński (2):
ASoC: topology: Ensure that needed parameters are set
ASoC: topology: Check if ops is set before dereference
sound/soc/soc-topology.c | 22 +++++++++++++++-------
1 file changed, 15 insertions(+), 7 deletions(-)
--
2.25.1
This series adds unit tests for ASoC topology.
First fix problems found when developing and running test cases and
then add tests implementation.
Tests themselves are quite simple and just call
snd_soc_tplg_component_load() with various parameters and check the
result. Tests themselves are described in more detail in commits
adding them.
Goal is to expand the amount of test cases in following patches.
Prerequisity for this patchset are 2 patches which have already been
sent:
https://lore.kernel.org/alsa-devel/20210114163602.911205-1-amadeuszx.slawinski@linux.intel.com/T/#t
Description on how typical test case itself works:
In order to load topology we need to have 3 things:
card, codec component & platform component.
In typical test case we register card and platform component and bind
to dummy codec. There are of course execeptions, when we want to
test behaviour of topology API when component or card is missing.
Note that this is bit different from typical scenario (in SOF and skylake
drivers) where card is registered by machine driver and component by
platform driver, as we register both when setting up test.
If you check the test case most of them have similar architecture of:
1.
/* run test */
ret = snd_soc_register_card(&kunit_comp->card);
if (ret != 0 && ret != -EPROBE_DEFER)
KUNIT_FAIL(test, "Failed to register card");
2.
ret = snd_soc_component_initialize(&kunit_comp->comp, &test_component, test_dev);
KUNIT_EXPECT_EQ(test, 0, ret);
3.
ret = snd_soc_add_component(&kunit_comp->comp, NULL, 0);
KUNIT_EXPECT_EQ(test, 0, ret);
Ad. 1.
First we register card, which in most tests returns -EPROBE_DEFER
(from snd_soc_bind_card()), as platform component is not yet created.
I test for both 0 and -EPROBE_DEFER, as it makes it easier to reshuffle
this code around if needed and there is one test case which does it in
different order.
Ad. 2.
Then we initialize platform component with structure pointing at proper
probe function, which calls snd_soc_tplg_component_load() with test
parameters and checks expected result.
Ad. 3.
And then in follow up we call snd_soc_add_component() which creates
platform component for us and calls snd_soc_try_rebind_card() which
if everything is bound properly calls previously set probe function.
Amadeusz Sławiński (5):
ASoC: topology: Properly unregister DAI on removal
Revert "ASoC: soc-devres: add devm_snd_soc_register_dai()"
ASoC: topology: KUnit: Add KUnit tests passing various arguments to
snd_soc_tplg_component_load
ASoC: topology: KUnit: Add KUnit tests passing empty topology with
variants to snd_soc_tplg_component_load
ASoC: topology: KUnit: Add KUnit tests passing topology with PCM to
snd_soc_tplg_component_load
include/sound/soc.h | 4 -
sound/soc/Kconfig | 17 +
sound/soc/Makefile | 5 +
sound/soc/soc-devres.c | 37 --
sound/soc/soc-topology-test.c | 843 ++++++++++++++++++++++++++++++++++
sound/soc/soc-topology.c | 9 +-
6 files changed, 870 insertions(+), 45 deletions(-)
create mode 100644 sound/soc/soc-topology-test.c
--
2.25.1
The application circuit shall provide MICVDD power.
In default, the codec driver doesn't need to enable LDO2.
In case, a board wants to use VBAT for micbias,
it should add a DAPM route which IN1P connects with LDO2 in the machine driver.
e.g. { "IN1P", NULL, "LDO2" },
Signed-off-by: Shuming Fan <shumingf@realtek.com>
Link: https://lore.kernel.org/r/20210121100353.6402-1-shumingf@realtek.com
Signed-off-by: Mark Brown <broonie@kernel.org>
soc_pcm_params_symmetry() checks rate/channel/sample_bits state.
These are very similar but different, thus, it needs to have very
verbose code.
This patch use macro for it and make code more simple.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/878s8un6si.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
soc_pcm_apply_symmetry() want to call snd_pcm_hw_constraint_single()
for rate/channel/sample_bits, but, it needs many condition checks.
These are very similar but different, thus, it needs to have very
verbose code.
This patch use macro for it and make code more simple.
Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Link: https://lore.kernel.org/r/87a6tan6sm.wl-kuninori.morimoto.gx@renesas.com
Signed-off-by: Mark Brown <broonie@kernel.org>
