Ensure that all interrupt handlers are unregistered before the parent
regmap_irq is unregistered.
sdca_irq_cleanup() was only called from the component_remove(). If the
module was loaded and removed without ever being component probed the
FDL interrupts would not be unregistered and this would hit a WARN
when devm called regmap_del_irq_chip() during the removal of the
parent IRQ.
Fixes: 4e53116437e9 ("ASoC: SDCA: Fix errors in IRQ cleanup")
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://patch.msgid.link/20260408093835.2881486-5-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
IRQs are enabled through sdca_irq_populate() from component probe
using devm_request_threaded_irq(), this however means the IRQs can
persist if the sound card is torn down. Some of the IRQ handlers
store references to the card and the kcontrols which can then
fail. Some detail of the crash was explained in [1].
Generally it is not advised to use devm outside of bus probe, so
the code is updated to not use devm. The IRQ requests are not moved
to bus probe time as it makes passing the snd_soc_component into
the IRQs very awkward and would the require a second step once the
component is available, so it is simpler to just register the IRQs
at this point, even though that necessitates some manual cleanup.
Link: https://lore.kernel.org/linux-sound/20260310183829.2907805-1-gaggery.tsai@intel.com/ [1]
Fixes: b126394d9ec6 ("ASoC: SDCA: Generic interrupt support")
Reported-by: Gaggery Tsai <gaggery.tsai@intel.com>
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://patch.msgid.link/20260316141449.2950215-1-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
The number of the initialization table may exceed 2048. Therefore, this patch removes the limitation and allows the driver to allocate memory dynamically based on the size of the initialization table. Signed-off-by: Shuming Fan <shumingf@realtek.com> Reviewed-by: Charles Keepax <ckeepax@opensource.cirrus.com> Link: https://patch.msgid.link/20260325092017.3221640-1-shumingf@realtek.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: Fixes for v7.0 A moderately large pile of fixes, though none of them are super major, plus a few new quirks and device IDs.
Move the ASP register patches to a separate struct and apply this from the ASP DAI probe() function so that the registers are only patched if the DAI is part of a DAI link. Some systems use the ASP as a special-purpose interconnect and on these systems the ASP registers are configured by a third party (the firmware, the BIOS, or another device using the amp's secondary host control interface). If the machine driver does not hook up the ASP DAI then the ASP registers must be omitted from the patch to prevent overwriting the third party configuration. If the machine driver includes the ASP DAI in a DAI link, this implies that the machine driver and higher components (such as alsa-ucm) are taking ownership of the ASP. In this case the ASP registers are patched to known defaults and the machine driver should configure the ASP. Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com> Link: https://patch.msgid.link/20260226110137.1664562-1-rf@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
On TAS2781, if the Speaker calibrated impedance is lower than default value hard-coded inside the TAS2781, it will cuase vol lower than normal. In order to fix this issue, the parameter of SineGainI need updating. Signed-off-by: Shenghao Ding <shenghao-ding@ti.com> Tested-by: Matthew Schwartz <matthew.schwartz@linux.dev> Link: https://patch.msgid.link/20260227144641.1243-1-shenghao-ding@ti.com Signed-off-by: Takashi Iwai <tiwai@suse.de>
ASoC: Updates for v7.0 This release is almost all abut driers, there's very little core work here, although some of that driver work is in more generic areas like SDCA and SOF: - Generic SDCA support for reporting jack events. - Continuing platform support, cleanup and feature improements for the AMD, Intel, Qualcomm and SOF code. - Platform description improvements for the Cirrus drivers. - Support for NXP i.MX952, Realtek RT1320 and RT5575, and Sophogo CV1800B. We also pulled in one small SPI API update and some more substantial regmap work (cache description improvements) for use in drivers.
Add support for using the state of pins on the amplifier to indicate the type of speaker fitted. Previously, where there were alternate speaker vendors, this was indicated using host CPU GPIOs. Some new Dell models use spare pins on the CS35L63 as GPIOs for the speaker ID detection. Cirrus-specific SDCA Disco properties provide a list of the pins to be used, and pull-up/down settings for the pads. This list is ordered, MSbit to LSbit. The code to set the firmware filename has been modified to check for using chip pins for speaker ID. The entire block of code to set firmware name has been moved out of cs35l56_component_probe() into its own function to make it easier to KUnit test. Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com> Link: https://patch.msgid.link/20260205164838.1611295-2-rf@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
Merge series from Charles Keepax <ckeepax@opensource.cirrus.com>: A bit of a mixed bag of minor misc fixes, improve handling of volatile SDCA Controls, make some minor bug fixes to jack detect, improve the cache syncing by adding some more defaults, and improve some FDL error messages.
Some of the SDCA Controls have a defined reset value in the specification. Update the parsing to add these specification defined values into the regmap defaults array. This will reduce the number of registers that are synchronised on a cache sync. Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Link: https://patch.msgid.link/20260204125944.1134011-6-ckeepax@opensource.cirrus.com Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.dev> Signed-off-by: Mark Brown <broonie@kernel.org>