Turn off/on clocks when device enters suspend/resume. This
can help saving power.
As a further optimization, we turn off/on mclk only when SAI
is in master mode because otherwise mclk is externally provided.
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Reviewed-by: Viorel Suman <viorel.suman@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Build warning being reported:
sound/soc/fsl/fsl_sai.c: In function 'fsl_sai_remove':
sound/soc/fsl/fsl_sai.c:921:1: warning: no return statement in
function returning non-void [-Wreturn-type]
So this patch just adds a "return 0" to fix it.
Fixes: 812ad463e0 ("ASoC: fsl_sai: Add support for runtime pm")
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Reviewed-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Basically the same actions as for system PM, so make use
of pm_runtime_force_suspend/pm_runtime_force_resume.
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
is_slave_mode defaults to false because sai structure
that contains it is kzalloc'ed.
Anyhow, if we decide to set the following configuration
SAI slave -> SAI master, is_slave_mode will remain set on true
although SAI being master it should be set to false.
Fix this by updating is_slave_mode for each call of
fsl_sai_set_dai_fmt.
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
When stopping audio, ASoC will first stop DMA then CPU DAI.
Sometimes there is a delay between DMA stop and CPU DAI stop, which
triggers an underrun error. Now, because of the delay introduced
by dev_err another underrun error will occur causing a vicious circle
making impossible to stop CPU DAI.
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
Signed-off-by: Daniel Baluta <daniel.baluta@nxp.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Adopt the SPDX license identifier headers to ease license compliance
management.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The FSL SAI can support up to 32 channels using TDM. Report that value so
they can actually be used.
Tested using 8 channels.
Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The 'np' variable is already assigned to 'pdev->dev.of_node', so use
it to improve readability.
No functional change.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
On mx6ul the General Purpose Register 1 (GPR1) contains the following
bits for configuring the direction of the SAI MCLKs:
SAI1_MCLK_DIR, SAI2_MCLK_DIR, SAI3_MCLK_DIR
Introduce the "fsl,sai-mclk-direction-output" optional property to allow
configuring the SAI_MCLK outputs.
Tested on a imx6ul-evk board.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
MX6UL may need to configure the General Purpose Register 1 (GPR1), so
it is better to add a new compatible string to differentiate.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Since msleep() will sleep longer than intended time for values less
than 20ms, this patch allows the use of usleep_range for just 1ms.
usleep_range is a finer precision implementation of msleep and is
designed to be a drop-in replacement for udelay where a precise
sleep/busy-wait is unnecessary.
More details see Documentation/timers/timers-howto.txt.
Signed-off-by: Xiubo Li <lixiubo@cmss.chinamobile.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
After several open/close sai test with ctrl+c, there will be
I/O error. The SAI can't work anymore, can't recover. There
will be no frame clock. With adding the software reset in
trigger stop, the issue can be fixed.
This is a hardware bug/errata and reset is the only option.
According to the reference manual, the software reset doesn't
reset any control register but only internal hardware logics
such as bit clock generator, status flags, and FIFO pointers.
(Our purpose is just to reset the clock generator while the
software reset is the only way to do that.)
Since slave mode doesn't use the clock generator, only apply
the reset procedure to the master mode.
For asynchronous mode, TX will not be reset when RX is still
running. In this case, i can't reproduce this issue.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
using params_width function to simplify code.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Add tdm slots operation support. If tdm slots and slot width have
been configured in machine driver, we should use these values.
Otherwise, using relevant channels and word length to set slots
and slot width.
SAI will generate BCLK depends on sample rate, slots and slot width.
And there may be unused BCLK cycles before each LRCLK transition.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
For SAI master mode, when Tx(Rx) sync with Rx(Tx) clock, Rx(Tx) will
generate bclk and frame clock for Tx(Rx), we should set RCR4(TCR4),
RCR5(TCR5) and RMR(TMR) for playback(capture), or there will be sync
error sometimes.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
In fsl_sai_set_bclk function, we should not set bclk for Tx/Rx Synchronous
with another SAI mode.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
FSL_SAI_TDR register is writable and not readable. According to
regmap_volatile() function, if FSL_SAI_TDR want to be volatile,
it should be readable. So we should remove FSL_SAI_TDR from volatile
register list.
If the flat cache don't have default register map, when do regcache_sync
operation, the non volatile and writable registers will be synchronised
to 0. FSL_SAI_TDR reigster will be written a 0 and cause channel swap.
So add default register map for flat cache, and such register will not
be written.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
When using the Rx clock for both, transmitter and receiver, the
transmitter needs to be set to synchronous with receiver.
This reverts 855675f6e6 ("ASoC: fsl_sai: Set SYNC bit of TCR2 to
Asynchronous Mode"), which, judiging from the commit log, seems to
mixed up between the two synchronous modes: The boolean
sai->synchronous[TX] is indicating wheather the SAI should work in
Rx synchronous mode (sync Tx with Rx), hence if the value is true,
the SYNC field of TCR2 needs to be set to 0x1 ("Synchronous with
receiver").
Signed-off-by: Stefan Agner <stefan@agner.ch>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Change them to #ifdef as CONFIG_PM_SLEEP might not be defined at all.
Reported-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The rate constraints in this driver are shared between all device
instances. It should not be (and is not) modified at runtime, so make them
const.
While we are at it also change the type of the rates array from u32 to
unsigned int. While both resolve to the same type, u32 is usually used to
empathize that the value is for a 32-bit hardware register or similar which
makes it slightly confusing in this context.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
For i.MX6 SoloX, there is a mode of the SoC to shutdown all power source of
modules during system suspend and resume procedure. Thus, SAI needs to save
all the values of registers before the system suspend and restore them after
the system resume.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
This platform driver has a OF device ID table but the OF module
alias information is not created so module autoloading won't work.
Signed-off-by: Luis de Bethencourt <luis@debethencourt.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
As the ssi is not the only cpu dai, there are esai, spdif, sai.
and imx_pcm_dma can be used by all of them. Especially ESAI need
a larger DMA buffer size. So Add dedicated DMA buffer for each cpu
dai.
Signed-off-by: Shengjiu Wang <shengjiu.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Acked-by: Timur Tabi <timur@tabi.org>
Signed-off-by: Mark Brown <broonie@kernel.org>
Normally we don't support 12kHz, 24kHz in audio driver, alsa didn't
have formal definition of 12kHz, 24kHz, but alsa supply a way to
support these sample rates. And add 176.4kHz and 192kHz support.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
When sai works on master mode, set its bit clock and frame clock.
SAI has 4 MCLK source, bus clock, MCLK1, MCLK2 and MCLK3. fsl_sai_set_bclk
will select proper MCLK source, then calculate and set the bit clock divider.
After fsl_sai_set_bclk, enable the selected mclk in hw_params(), and add
hw_free() to disable the mclk.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Whether residue can be reported or not is not a property of the audio
controller but of the DMA controller. The FLAG_NO_RESIDUE was initially
added when the DMAengine framework had no support for describing the residue
reporting capabilities of the controller. Support for this was added quite a
while ago and recently the DMAengine framework started to complain if a
driver does not describe its capabilities and a lot of patches have been
merged that add support for this where it was missing. So it should be safe
to assume that driver on actively used platforms properly implement the DMA
capabilities API.
This patch makes the FLAG_NO_RESIDUE internal and no longer allows audio
controller drivers to manually set the flag. If a DMA driver against
expectations does not support reporting its capabilities for now the generic
DMAengine PCM driver will now emit a warning and simply assume that residue
reporting is not supported. In the future this might be changed to aborting
with an error.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
Currently the error message uses 'np->full_name' which leads to a very verbose
log that contains the full path of the node.
We can have a concise log by using pdev->name instead.
Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
The 'big-endian-data' property is originally used to indicate whether the
LSB firstly or MSB firstly will be transmitted to the CODEC or received
from the CODEC, and there has nothing relation to the memory data.
Generally, if the audio data in big endian format, which will be using the
bytes reversion, Here this can only be used to bits reversion.
So using the 'lsb-first' instead of 'big-endian-data' can make the code
to be readable easier and more easy to understand what this property is
used to do.
This property used for configuring whether the LSB or the MSB is transmitted
first for the fifo data.
Signed-off-by: Xiubo Li <Li.Xiubo@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
There is one design rule according to SAI's reference manual:
If the transmitter bit clock and frame sync are to be used by both transmitter
and receiver, the transmitter must be configured for asynchronous operation
and the receiver for synchronous operation.
And SYNC of TCR2 is a 2-width control bit:
00 Asynchronous mode.
01 Synchronous with receiver.
10 Synchronous with another SAI transmitter.
11 Synchronous with another SAI receiver.
So the driver should have set SYNC bit of TCR2 to 0x0, and meanwhile set SYNC
bit of RCR2 to 0x1 (Synchronous with transmitter).
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
The previous patch (ASoC: fsl_sai: Add asynchronous mode support) added
new Device Tree bindings for Asynchronous and Synchronous modes support.
However, these two shall not be present at the same time.
So this patch just simply makes them exclusive so as to avoid incorrect
Device Tree binding usage.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
SAI supports these operation modes:
1) asynchronous mode
Both Tx and Rx are set to be asynchronous.
2) synchronous mode (Rx sync with Tx)
Tx is set to be asynchronous, Rx is set to be synchronous.
3) synchronous mode (Tx sync with Rx)
Rx is set to be asynchronous, Tx is set to be synchronous.
4) synchronous mode (Tx/Rx sync with another SAI's Tx)
5) synchronous mode (Tx/Rx sync with another SAI's Rx)
* 4) and 5) are beyond this patch because they are related with another SAI.
As the initial version of this SAI driver, it supported 2) as default while
the others were totally missing.
So this patch just adds supports for 1) and 3).
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
There is one design rule according to SAI's reference manual:
If the transmitter bit clock and frame sync are to be used by both transmitter
and receiver, the transmitter must be configured for asynchronous operation
and the receiver for synchronous operation.
And SYNC of TCR2 is a 2-width control bit:
00 Asynchronous mode.
01 Synchronous with receiver.
10 Synchronous with another SAI transmitter.
11 Synchronous with another SAI receiver.
So the driver should have set SYNC bit of TCR2 to 0x0, and meanwhile set SYNC
bit of RCR2 to 0x1 (Synchronous with transmitter).
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
This patch adds software reset code in dai_probe() so as to make a true init
by clearing SAI's internal logic, including the bit clock generation, status
flags, and FIFO pointers.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
DPCM needs extra dapm routes in the machine driver to route audio
between Front-End and Back-End. In order to differ the stream names
in the route map from CODECs, we here add specific stream names to
SAI driver so that we can implement ASRC via DPCM to it.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
The previous enable flow:
1, Enable TE&RE (SAI starts to consume tx FIFO and feed rx FIFO)
2, Mask IRQ of Tx/Rx to enable its interrupt.
3, Enable DMA request of Tx/Rx.
As this flow would enable DMA request later than TERE, the Tx FIFO
would be easily emptied into underrun while Rx FIFO would be easily
stuffed into overrun due to the delayed DMA transfering.
This issue happened merely occational before the patch 'ASoC: fsl_sai:
Reset FIFOs after disabling TE/RE' because there were useless data
remaining in the FIFO for the gap. However, it manifested after FIFO
reset's implemented.
After this patch, the new flow:
1, Enable DMA request of Tx/Rx.
2, Enable TE&RE (SAI starts to consume tx FIFO and feed rx FIFO)
3, Mask IRQ of Tx/Rx to enable its interrupt.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
TE/RE bit of T/RCSR will remain set untill the current frame is physically
finished. The FIFO reset operation should wait this bit's totally cleared
rather than ignoring its status which might cause TE/RE disabling failed.
This patch adds delay and timeout to wait for its completion before FIFO
reset.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
For trigger start, we don't need to check if it's the first time to
enable TE/RE or second time. It doesn't hurt to enable them any way,
which in the meantime can reduce race condition for TE/RE enabling.
For trigger stop, we will definitely clear FRDE of current direction.
Thus the driver only needs to read the opposite one's.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
In the rx irq handling part, we should clear the flags in RCSR not TCSR.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
SAI will not clear their FIFOs after disabling TE/RE. Therfore, the driver
should take care the task so as not to let useless data remain in the FIFO.
Signed-off-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@linaro.org>
