12 Feb, 2010
3 commits
-
The WM2000 is a low power, high quality handset receiver speaker
driver with Wolfson myZone™ Ambient Noise Cancellation (ANC). It
provides enhanced voice communication quality in a noisy environment
if the handset acoustics are designed appropriately.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
Audio on Migo-R cannot work if CONFIG_SH_DMA_API=y, but compilation should not
break anyway.Signed-off-by: Guennadi Liakhovetski
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
This patch expands the omap3beagle sound soc for the
beagle board clone DevKit8000.Signed-off-by: Thomas Weber
Acked-by: Jarkko Nikula
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown
10 Feb, 2010
4 commits
-
Signed-off-by: Paul Menzel
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
The serial interface (TDM/I2S) for the audio block have been
constantly enabled.
Introduce a new DAPM_SUPPLY for handling the AIF_EN bit, so
the interface is only enabled, when there is a need for it.
For example when only the analog loopback is enabled, there
is no need to keep the serial interface active.I have added the persons who contributed to the Voice path
of twl4030 codec driver, so they might have the ability
to test this patch, and send an update for the Voice path,
if it is necessarySigned-off-by: Peter Ujfalusi
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
Enable the bulk regulators at probe time so we can safely disable them
again when going to suspend without confusing the reference counter.Signed-off-by: Daniel Mack
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
This patch adds sound support for Phytec PhyCORE / PhyCARD
modules in AC97 mode.Signed-off-by: Sascha Hauer
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown
06 Feb, 2010
1 commit
-
Pandora's external DAC is connected to VSIM TWL4030 supply, so let's
start switching it too to save more power.
Also DAC got it's own DAPM handler.Signed-off-by: Grazvydas Ignotas
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown
05 Feb, 2010
2 commits
-
Pandora's external DAC is using 256*Fs output from the TWL4030
codec, and TWL4030 needs to have APLL enabled for it's 256*Fs
output to function.Signed-off-by: Grazvydas Ignotas
Acked-by: Peter Ujfalusi
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown
04 Feb, 2010
10 commits
-
The module unloading path had several problems:
- it freed up the private structure twice
- it freed up the codec structure, which was allocated as part
of the private structure
- it did not freed up the reg_cache
- it did not unregistered the dais and the codecSigned-off-by: Peter Ujfalusi
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
The WM8912 is a DAC only device register compatible with the WM8904
CODEC with ADC portions omitted. Support it within the WM8904 driver
based on the configured I2C device name.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
Handle the output PGAs as part of the output powerup since they can
never be powered separately and reorder things so that we remove the
output shorts after both line and headphone outputs have been brought
up, minimising the opportunity for any issues.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
As well as disabling the biases of the CODEC the drop into BIAS_OFF will
also disable all the regulators powering the CODEC, allowing even greater
power savings on appropriately configured systems.Since the regulator API does not currently provide notification when
regulators are disabled we assume that this always happens when we stop
using the regulators. Once 2.6.34 is merged this code can be optimised
to only sync the cache when power was actually removed.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
While the regulators are disabled we cache all register writes.
Currently we assume that the regulator disable actually takes
effect, after the merge with the regulator tree in 2.6.34 the
regulator API will be able to notify us if the power is actually
removed (due to constraints or regulator sharing it may not be).Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
At the minute the regulators are simply enabled for the entire
lifetime of the device.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
Saves a little bit of code duplication.
Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
Add a bit to the CODEC structure indicating if a cache sync is required.
By default this will be set if a cache only write is done to a soc-cache
register cache. This allows us to avoid syncing the cache back after
using cache only writes if there were no changes.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
Currently the soc-cache code will always write to the device, meaning
that we need the device to be powered and active at pretty much all
times the system is active. Allowing cache only writes lays some
groundwork for future enhancements to allow devices to be put into a
full off state when the audio subsystem is idle.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
ctrl_outl() has become void at some point, which breaks compilation of fsi.c.
Make writing functions void, as their output is anyway not evaluated, and use
__raw_writel and __raw_readl instead of deprecated ctrl_outl and ctrl_inl
respectively.Signed-off-by: Guennadi Liakhovetski
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown
02 Feb, 2010
7 commits
-
Unconditionally save the register states when suspending and restore
them again at resume time. Register contents were not preserved over
suspend, and hence the driver takes false assumptions about them.The clock must be enabled to access the register block.
Signed-off-by: Daniel Mack
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
String constants that are continued on subsequent lines with \
are not good.Signed-off-by: Joe Perches
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
This fixes BCLK calculation and removes unnecessary check code.
Signed-off-by: Joonyoung Shim
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
The dependency on MFD_WM8994 rather than I2C went awry.
Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
The WM8994 is a highly integrated ultra-low power hi-fi audio subsystem
designed for smartphones and other portable devices rich in multimedia
features. It provides advanced digital mixing facilities enabling low
power high quality interconnection of CPU, baseband and other audio
sources through flexible digital and analogue routing, and integrates
a class W headphone driver and stereo class D speaker drivers.Signed-off-by: Mark Brown
-
Use a two code correction for optimal performance.
Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
Perform DC servo offset calibration using a series update sequence
rather than startup update sequence, tuning the configuration of the
WM8993 DC servo to make best use of this.Also introduce currently unused data allowing us to correct for
any systematic errors in the DC servo calibration results and an
alternative startup path for the headphone output which performs
better with some chip revisions. The alternative setup sequence is
enabled for WM8993.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood
01 Feb, 2010
2 commits
-
String constants that are continued on subsequent lines with \
are not good.Signed-off-by: Joe Perches
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
In case, if OPCLK is not used, and PLL is used for driving the codec, the
choice of PLL output frequency could result in a needlessly imprecise
system clock frequency. Use an iterative process to select a precise
configuration.Signed-off-by: Guennadi Liakhovetski
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown
29 Jan, 2010
7 commits
-
wm8974 and wm8978 codec drivers control DAC and ADC oversampling rates in their
.set_clkdiv() methods, which is wrong, because these are simple boolean
switches and not clock dividers. Move these bits to sound controls. Also remove
manual configuration of the MCLK divider in wm8978, since it is configured
automatically.Signed-off-by: Guennadi Liakhovetski
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
Tests showed, that bit 6 of the WM8978_POWER_MANAGEMENT_2 register of wm8978
affects codec clocks. Being useless for suspend / resume, it cannot be used in
bias-level control either. Remove this bit handling.Signed-off-by: Guennadi Liakhovetski
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
Add support for audio on sh7722-based Migo-R boards, using SIU and wm8978
codec, recording via external microphone and playback via headphones are
implemented.Signed-off-by: Guennadi Liakhovetski
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
It's more robust when references are provided in control names
rather than numid.Signed-off-by: Jassi Brar
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown -
Commit 761c9d45 (ASoC: Fix build of OMAP sound drivers) changes
CONFIG_MACH_OMAP3517EVM -> CONFIG_SND_OMAP_SOC_OMAP3517EVM in the
Makefile. Whereas the config option defined in Kconfig is
SND_OMAP_SOC_AM3517EVM. Because of this, ASoC driver for AM3517
was not getting compiled.Signed-off-by: Anuj Aggarwal
Signed-off-by: Mark Brown -
Commit e9ff5eb2 (Fixing infinite loop in resume path) uses wrong AIC23
register in resume function because of which register writes happen
on some non-existing registers.Signed-off-by: Anuj Aggarwal
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown
28 Jan, 2010
4 commits
-
Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
Also renames a few things to make volumes and switches match up in
alsamixer.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
The version isn't being updated or used, the kernel revision
tracking is enough.Signed-off-by: Mark Brown
Acked-by: Liam Girdwood -
Change the legacy default register configuration, which left some
internal components on.
Now we have either DAPM, or other ways to control these bits,
so there is no need to enable them by default.The affected parts:
Disable ADCL and ADCR
Disable ARXL2 and ARXR2 analog PGA (playback)
Disable APLL by defaultSigned-off-by: Peter Ujfalusi
Acked-by: Liam Girdwood
Signed-off-by: Mark Brown