Tascam's Model 12 is a mixer which can also operate as a USB audio
interface. The audio interface uses explicit feedback but it seems that
it does not correctly handle missing isochronous frames.
When injecting an xrun (or doing anything else that pauses the playback
stream) the feedback rate climbs (for example, at 44,100Hz nominal, I
see a stable rate around 44,099 but xrun injection sees this peak at
around 44,135 in most cases) and glitches are heard in the audio stream
for several seconds - this is significantly worse than the single glitch
expected for an underrun.
While the stream does normally recover and the feedback rate returns to
a stable value, I have seen some occurrences where this does not happen
and the rate continues to increase while no audio is heard from the
output. I have not found a solid reproduction for this.
This misbehaviour can be avoided by totally resetting the stream state
by switching the interface to alt 0 and back before restarting the
playback stream.
Add a new quirk flag which forces the endpoint and interface to be
reconfigured whenever the stream is stopped, and use this for the Tascam
Model 12.
Separate interfaces are used for the playback and capture endpoints, so
resetting the playback interface here will not affect the capture stream
if it is running. While there are two endpoints on the interface,
these are the OUT data endpoint and the IN explicit feedback endpoint
corresponding to it and these are always stopped and started together.
Signed-off-by: John Keeping <john@metanate.com>
Link: https://lore.kernel.org/r/20221129130100.1257904-1-john@metanate.com
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Use link->exit() rather than pdev->remove() to unassign jack during card
unbind procedure so codec link initialization and exit procedures are
symmetrical.
Also, there is no need to perform search for codec dai in suspend_pre()
and resume_post() ourselves. Use snd_soc_card_get_codec_dai() instead.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20221125184032.2565979-2-cezary.rojewski@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
If the device does not come straight from reset, we might receive an IRQ
before we are ready to handle it.
Fixes:
[ 2.334737] Unable to handle kernel read from unreadable memory at virtual address 00000000000001e4
[ 2.522601] Call trace:
[ 2.525040] regmap_read+0x1c/0x80
[ 2.528434] mt8173_afe_irq_handler+0x40/0xf0
...
[ 2.598921] start_kernel+0x338/0x42c
Signed-off-by: Ricardo Ribalda <ribalda@chromium.org>
Fixes: ee0bcaff10 ("ASoC: mediatek: Add AFE platform driver")
Link: https://lore.kernel.org/r/20221128-mt8173-afe-v1-0-70728221628f@chromium.org
Signed-off-by: Mark Brown <broonie@kernel.org>
We found an corner case in RT5640 codec driver which schedules jack work
after system suspend as IRQ was enabled. Due to this, hitting the error
as register access happening after suspend as jack worker thread getting
scheduled in irq handler. The patch disables the irq during the suspend
to prevent the corner case happening.
Signed-off-by: Oder Chiou <oder_chiou@realtek.com>
Reported-by: Mohan Kumar D <mkumard@nvidia.com>
Link: https://lore.kernel.org/r/20221128070825.91215-1-oder_chiou@realtek.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Merge series from Nicolas Frattaroli <frattaroli.nicolas@gmail.com>:
This patchset refactors the Rockchip I2S/TDM driver in order to
support the RK3588 SoC, and then adds the necessary compatible
string to load the driver for it.
Patch 1 rectifies a problem with the bindings where we were too
strict about requiring the rockchip,grf property. Most features
of this audio device don't need access to the GRF to function.
Patch 2 modifies the driver to adjust its behaviour to what the
changed bindings now allow, namely using most things without the
GRF.
Patch 3 and 4 are boring compatible string stuff that enables
RK3588 support. No special data is needed to initialise the
driver for this instance of the I2S/TDM IP.
The DAPM control for PGAx uses the PGA mute bit for
power management. This bit is active high but is set to
non-inverted (ie when powering, it will mute).
The ALSA control "PGA x Capture Switch" uses the active
high PGA_ENx bit, but is set to inverted. So when
enabling this switch, the PGA gets disabled.
To correct the behaviour, invert both these bits.
Signed-off-by: Maarten Zanders <maarten.zanders@mind.be>
Link: https://lore.kernel.org/r/20221028152626.109603-4-maarten.zanders@mind.be
Signed-off-by: Mark Brown <broonie@kernel.org>
"mclk" is retrieved from the configuration and assigned to adau1372->clk.
However adau1372->mclk (==NULL) is used for clk_prepare_enable() and
clk_disable_unprepare() which don't have any effect.
Remove .clk from struct adau1372 and use .mclk throughout.
This change ensures that the input clock is switched on/off when the
bias level is changed.
Signed-off-by: Maarten Zanders <maarten.zanders@mind.be>
Link: https://lore.kernel.org/r/20221028152626.109603-2-maarten.zanders@mind.be
Signed-off-by: Mark Brown <broonie@kernel.org>
Currently the check against the max value for the control is being
applied after the value has had the minimum applied and been masked. But
the max value simply indicates the number of volume levels on an SX
control, and as such should just be applied on the raw value.
Fixes: 97eea946b9 ("ASoC: ops: Check bounds for second channel in snd_soc_put_volsw_sx()")
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20221125162348.1288005-1-ckeepax@opensource.cirrus.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Merge series from Richard Fitzgerald <rf@opensource.cirrus.com>:
Use EXPORT_SYMBOL_NS_GPL() instead of EXPORT_SYMBOL_GPL() and patch the
three drivers that use cs_dsp to add the MODULE_IMPORT_NS().
To make the namespace more specific the KConfig symbol for cs_dsp is
changed from CS_DSP to FW_CS_DSP.
Merge series from Bard Liao <yung-chuan.liao@linux.intel.com>:
The code in drivers/soundwire/intel_init.c is hardware-dependent and the
code does not apply to new generations starting with MeteorLake. Refactor
and clean-up the code to make this intel_init.c hardware-agnostic and
move all hardware-dependencies in the SOF driver using chip descriptors.
Merge series from Richard Fitzgerald <rf@opensource.cirrus.com>:
Writing a firmware control should be returning 1 if the control value
changed, so these two patches add that.
Though this is an ALSA requirement it is also useful for non-ALSA clients
of cs_dsp to know if the control value changed, so the main handling is
implemented in cs_dsp. TLV controls are specifically an ALSA thing so they
are handled specially in wm_adsp.
Simon Trimmer (2):
firmware: cs_dsp: cs_dsp_coeff_write_ctrl() should report changed
ASoC: wm_adsp: Return whether changed when writing controls
drivers/firmware/cirrus/cs_dsp.c | 17 ++++++++++++-----
sound/soc/codecs/wm_adsp.c | 27 ++++++++++++++++++---------
2 files changed, 30 insertions(+), 14 deletions(-)
--
2.30.2
