- 26 Mar, 2024 36 commits
-
-
Charles Keepax authored
Factor out the creation of the SoundWire DAI links into a helper function. No functional change. Reviewed-by:
Bard Liao <yung-chuan.liao@linux.intel.com> Signed-off-by:
Charles Keepax <ckeepax@opensource.cirrus.com> Signed-off-by:
Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Link: https://msgid.link/r/20240326160429.13560-24-pierre-louis.bossart@linux.intel.comSigned-off-by:
Mark Brown <broonie@kernel.org>
-
Charles Keepax authored
The only part left using the old parsing code is now the generation of the actual DAI links. Move this generation over to being based on the new parsing, which allows the removal of the last of the old parsing code. The new DAI link generation is a simple matter of creating a new DAI link for each sof_sdw_dailink struct, and adding a cpu, a codec, and a mapping for each sof_sdw_endpoint contained in that sof_sdw_dailink. Note that the CPUs can be inferred as the endpoint list is iterated, because the endpoints were added into the list sorted by link. Reviewed-by:
-
Charles Keepax authored
append_dai_type should be set to true in any situation that would result in multiple DAI links existing on a single SoundWire bus, because the legacy naming used only the bus number to make things unique. The current code handles this by looking for codecs with multiple dai_info structs and looking for buses that include multiple types of device on them. The first of these assumes that all DAIs on a given device would be in use. The second, with dissimilar aggregation now being supported, isn't really an accurate check either since those devices could be aggregated into a single DAI link. Move the handling for this flag over to the new parsing code and simplify things a little by looking directly for SoundWire links that will contain multiple DAI links to set the flag. These changes should not cause any change in behaviour for any currently supported systems. Reviewed-by:
-
Charles Keepax authored
Trivial move of the handling of ignore_pch_dmic over to the new parsing code. Reviewed-by:
-
Charles Keepax authored
Use the output of the new snd_soc_acpi struct parsing for the purposes of counting the number of SoundWire DAI links and physical devices attached to the SoundWire. These counts are already returned by the parser so those can just be used directly. But the population of the codec_conf structures, is moved from the old parsing code over to the new. As the two parsers currently co-exist it is better to not have them both attempt to modify the same structures. Reviewed-by:
-
Charles Keepax authored
The current machine driver code has a lot of loops parsing through the ACPI structs (snd_soc_acpi_link_adr, snd_soc_acpi_adr_device, snd_soc_acpi_endpoint), this makes it hard to understand exactly what information is being extracted and for what purposes. As well as being slightly inefficient, as the same information is looked up multiple times. There are also some issues with the handling of multiple endpoints on a single device, only the first of the snd_soc_acpi_endpoint structures is currently fully processed by the driver. This means doing things like aggregating the second endpoint on a device with another device are not currently possible. Add new parsing code that will count the devices and endpoints, parse them into an intermediate datastructure, and then use that to create the DAI links. This patch does not actually utilise the results of the parsing, items will be moved across in the following patches. This parsing is based around two new structures which are temporarily allocated whilst parsing. Firstly, sof_sdw_endpoint, which represents a specific endpoint for audio on a device and is more or less directly equivalent to snd_soc_acpi_endpoint. Secondly, sof_sdw_dailink which represents a DAI link and contains a linked list of one or more sof_sdw_endpoints. A single trip through the snd_soc_acpi data structures is used to populate these. One important point to note here is the use of the num field in snd_soc_acpi_endpoint to address sof_sdw_dai_info array in the sof_sdw_codec_info struct. This expects a one to one mapping between endpoints on a device and dai infos. It would be fine for a specific system to not specify an endpoint for all of the dai infos available, but two endpoints mapping to the same dai info would make not sense. Reviewed-by:
-
Charles Keepax authored
Currently the cs42l43 just specifies a single endpoint, as the current machine driver only looks at the first endpoint specified. Future refactoring will process all endpoints, as such proper specification should be added for all the cs42l43 endpoints. Reviewed-by:
-
Charles Keepax authored
The SOF_SDW_NO_AGGREGATION quirk is mostly for debug and no longer works correctly with the current state of the machine driver. Remove it from the code and add an error message if someone uses it. Reviewed-by:
-
Balamurugan C authored
Adding rt711 sdca codec support for arl boards. Reviewed-by:
Chao Song <chao.song@linux.intel.com> Signed-off-by:
Balamurugan C <balamurugan.c@intel.com> Signed-off-by:
-
Charles Keepax authored
In preparation for future refactoring pull out a helper specifically for generating the codec name. Reviewed-by:
-
Charles Keepax authored
Factor out the creation of the BlueTooth DAI links into a helper function. No functional change. Reviewed-by:
-
Charles Keepax authored
Factor out the creation of the HDMI DAI links into a helper function. No functional change. Reviewed-by:
-
Charles Keepax authored
Factor out the creation of the DMIC DAI links into a helper function. No functional change. Reviewed-by:
-
Charles Keepax authored
Factor out the creation of the SSP DAI links into a helper function. No functional change. Reviewed-by:
-
Charles Keepax authored
Rather than open coding a loop to process each bit use for_each_set_bit. Reviewed-by:
-
Charles Keepax authored
Rather than passing around a pointer to the dai_link array and an index into this array, simply pass a pointer to the current dai_link. Also move the DAI link pointer sanity check to the end of the DAI link creation, and change it to a warn on. This check should only be hit if there is a serious bug in the machine driver, so checking it on each iteration is excessive. Reviewed-by:
-
Charles Keepax authored
Move the flags ignore_pch_dmic and append_dai_type into the drivers private structure rather than passing them around between functions. Reviewed-by:
-
Charles Keepax authored
Currently sof_sdw relies on sof_board_helpers to provide get_codec_dai_by_name(), but that is the only function from sof_board_helpers it uses and no other machine driver requires that function. There is no reason for sof_sdw to select SSP_COMMON but more and more functions in sof_board_helpers are gaining dependencies on functions that would require stubs for sof_sdw to build without it. Firstly it was sof_ssp_get_codec_name(), as was fixed in commit c1469c3a ("ASoC: Intel: ssp-common: Add stub for sof_ssp_get_codec_name"), now it is: ERROR: modpost: "sof_ssp_detect_amp_type" [sound/soc/intel/boards/snd-soc-intel-sof-board-helpers.ko] undefined! ERROR: modpost: "sof_ssp_detect_codec_type" [sound/soc/intel/boards/snd-soc-intel-sof-board-helpers.ko] undefined! Rather than adding more stubs, simply move the affected function to the sof_sdw machine driver itself and no longer select SND_SOC_INTEL_SOF_BOARD_HELPERS at all. This should allow work on SSP_COMMON to progress without affecting sof_sdw. Reviewed-by:
-
Charles Keepax authored
The current code evenly distributes the CODEC DAIs across the CPU DAIs in the DAI link, but this is just an assumption about how the devices are connected to the host. All the information about which CODEC is connected to which CPU DAI is contained in the endpoints datastructures and those structures are already parsed to work out which CODECs to include in the DAI link. Simply fill in the mapping from CPUs -> CODECs as we parse through the endpoints structures, this will ensure that mapping matches the physical layout. Reviewed-by:
-
Charles Keepax authored
Rather than passing around a pointer to the codec_conf array, an index into it and a size, simply pass around a pointer to the current codec_conf. This reduces the amount of state moving around. Reviewed-by:
-
Charles Keepax authored
Rather than returning an index simply return a pointer to the located codec info, this simplifies all the callers which only want to access the codec info structure. Also remove the inline specifier the function is fairly large for an inline function, let the compiler decide. And move the function such that it is located with the other find_codec_info_*() functions. Reviewed-by:
-
Charles Keepax authored
Rather than returning an index simply return a pointer to the located codec info, this simplifies all the callers which only want to access the codec info structure. Also remove the inline specifier the function is fairly large for an inline function, let the compiler decide. Reviewed-by:
-
Charles Keepax authored
Rather than returning an index simply return a pointer to the located codec info, this simplifies all the callers which only want to access the codec info structure. Also remove the inline specifier the function is fairly large for an inline function, let the compiler decide. Reviewed-by:
-
Luca Ceresoli authored
Add driver for the internal audio codec of the Rockchip RK3308 SoC. Initially based on the vendor kernel driver [0], with lots of cleanups, fixes, improvements, conversion to DAPM and removal of some features. [0] https://github.com/rockchip-linux/kernel/blob/develop-4.19/sound/soc/codecs/rk3308_codec.cSigned-off-by:
Luca Ceresoli <luca.ceresoli@bootlin.com> Link: https://msgid.link/r/20240305-rk3308-audio-codec-v4-4-312acdbe628f@bootlin.comSigned-off-by:
-
Luca Ceresoli authored
No macro currently allows handling a stereo control that has left and right in the same register and whose minimum register value is not zero. Add one that does that. Note that even though the snd_soc_*_volsw_range() look more appropriate given the _range suffix, they are not suitable because they don't honor the two shift values. The snd_soc_*_volsw() look more generic and are suitable for the task. Signed-off-by:
-
Luca Ceresoli authored
Add device tree bindings document for the internal audio codec of the Rockchip RK3308 SoC. Reviewed-by:
Rob Herring <robh@kernel.org> Signed-off-by:
-
Mark Brown authored
Merge series from Ondřej Jirman <megi@xff.cz>: This series adds support for jack detection to this codec. I used and tested this on Pinephone. It works quite nicely. I tested it against Android headset mic button resistor specification. The patches are a rewritten and debugged version of the original ones from Arnaud Ferraris and Samuel Holland, improved to better handle headset button presses and with more robust plug-in/out event debouncing, and to use set_jack API instead of sniffing the sound card widget names, to detect the type of jack connector.
-
Mark Brown authored
Merge series from Zhang Yi <zhangyi@everest-semi.com>: We changed the configuration related to hibernation. and delete the REG_SUPPLY to cover mute issue.
-
Mark Brown authored
Merge series from Tomi Valkeinen <tomi.valkeinen@ideasonboard.com>: At the moment we cannot instantiate two dmaengine_pcms with the same parent device, as the components will be named the same, leading to conflicts. Add 'name' field to the snd_dmaengine_pcm_config, and use that (if defined) as the component name instead of deriving the component name from the device.
-
Mark Brown authored
Merge series from Cezary Rojewski <cezary.rojewski@intel.com>: Set of loosely connected patches. Most impactful change is dropping any permisiveness when snd_soc_dapm_add_routes() fails in soc-topology.c To do it safely, disable route checks for all skylake-driver boards. Relevant background: Since commit daa480bd ("ASoC: soc-core: tidyup for snd_soc_dapm_add_routes()") route checks are no longer permissive. Probe failures for Intel boards have been partially addressed by commit a22ae72b ("ASoC: soc-core: isable route checks for legacy devices") and its follow up but only skl_nau88l25_ssm4567.c is patched. The rest of the boards still need fixing. After that, removal of copy-pastas found in ssm4567.c and redundant code in i2s_test.c for avs-boards. Changes in v2: - glk_rt5682_max98357a.c and skl_hda_dsp_generic.c now disable route checks only for the skylake-drvier - asoc now logs failures of snd_soc_dapm_add_routes() in soc-topology.c Amadeusz Sławiński (1): ASoC: Intel: avs: i2s_test: Remove redundant dapm routes Cezary Rojewski (4): ASoC: Intel: Disable route checks for Skylake boards ASoC: topology: Do not ignore route checks when parsing graphs ASoC: Intel: avs: ssm4567: Do not ignore route checks ASoC: Intel: avs: ssm4567: Board cleanup sound/soc/intel/avs/boards/i2s_test.c | 79 ------------------- sound/soc/intel/avs/boards/ssm4567.c | 5 +- sound/soc/intel/boards/bxt_da7219_max98357a.c | 1 + sound/soc/intel/boards/bxt_rt298.c | 1 + sound/soc/intel/boards/glk_rt5682_max98357a.c | 2 + sound/soc/intel/boards/kbl_da7219_max98357a.c | 1 + sound/soc/intel/boards/kbl_da7219_max98927.c | 4 + sound/soc/intel/boards/kbl_rt5660.c | 1 + sound/soc/intel/boards/kbl_rt5663_max98927.c | 2 + .../intel/boards/kbl_rt5663_rt5514_max98927.c | 1 + sound/soc/intel/boards/skl_hda_dsp_generic.c | 2 + .../soc/intel/boards/skl_nau88l25_max98357a.c | 1 + sound/soc/intel/boards/skl_rt286.c | 1 + sound/soc/soc-topology.c | 11 ++- 14 files changed, 27 insertions(+), 85 deletions(-) -- 2.25.1
-
Mark Brown authored
Merge series from Chancel Liu <chancel.liu@nxp.com>: echo /lib/firmware/fw.elf > /sys/class/remoteproc/remoteproc0/firmware (A) echo start > /sys/class/remoteproc/remoteproc0/state (B) echo stop > /sys/class/remoteproc/remoteproc0/state The rpmsg sound card is registered in (A) and unregistered in (B). After "start", imx-audio-rpmsg registers devices for ASoC platform driver and machine driver. Then sound card is registered. After "stop", imx-audio-rpmsg unregisters devices for ASoC platform driver and machine driver. Then sound card is unregistered.
-
Mark Brown authored
Merge series from Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>: This series from Brent Lu adds common helpers and board configurations to reduce the number of quirks.
-
Mark Brown authored
Merge series from AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>: This is the start of a series cleaning up the Mediatek drivers with some preparatory cleanups and improvements.
-
John Watts authored
Clock direction is marked as 'unsigned int' but only two values are currently used in practice. Note these down in the documentation. Signed-off-by:
John Watts <contact@jookia.org> Link: https://msgid.link/r/20240326-dai_mclk_hint-v1-1-653cbd2d78d9@jookia.orgSigned-off-by:
-
Pierre-Louis Bossart authored
We have a couple of duplicate logs and missing information, add __func__ consistently and make sure useful error codes are logged. Signed-off-by:
Péter Ujfalusi <peter.ujfalusi@linux.intel.com> Link: https://msgid.link/r/20240325221925.206507-1-pierre-louis.bossart@linux.intel.comSigned-off-by:
-
Andy Shevchenko authored
The gpio field is not used anymore, remove the leftover. This also fixes the compilation error after the ... Fixes: 9855f05e ("ASoC: fsl: imx-es8328: Switch to using gpiod API") Signed-off-by:
Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://msgid.link/r/20240325191341.3977321-1-andriy.shevchenko@linux.intel.comSigned-off-by:
-
- 25 Mar, 2024 4 commits
-
-
Brent Lu authored
Introduce "cml_rt5682_def" for cml boards which implement headphone codec on SSP0 and speaker amplifiers on SSP1. Reviewed-by:
Brent Lu <brent.lu@intel.com> Signed-off-by:
-
Brent Lu authored
Remove sof_nau8825 board id and use adl_nau8825_def instead since SSP port assignment is the same. Reviewed-by:
-
Brent Lu authored
Remove SOF_NUM_IDISP_HDMI(3) from board quirks since the value is 3 if not defined. Reviewed-by:
-
Brent Lu authored
Remove SOF_NUM_IDISP_HDMI(3) from board quirks since the value is 3 if not defined. Reviewed-by:
-
