| Age | Commit message (Collapse) | Author |
|---|
|
MCPS free was being done from PGA context which will free up MCPS
for only last modules in a pipe and not the rest causing MCPS
leak and eventual audio loss due to no "free" MCPS.
This needs to be freed for every module while cleaning up the
modules, so move the check to
skl_tplg_mixer_dapm_post_pmd_event()
Signed-off-by: Mohan Krishna Velaga <mohan.krishnax.velaga@intel.com>
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Smatch warns that we dereferenced substream before check, so fix
this by initializing ebus after the check
sound/soc/intel/skylake/skl-pcm.c:802 skl_get_position()
warn: variable dereferenced before check 'substream->runtime'
Reported by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
This patch fixes the below warning reported by Dan by invoking
skl_sst_dsp_cleanup() in cleanup path on error and not bailing out
sound/soc/intel/skylake/skl-sst.c:270 skl_sst_dsp_init()
info: ignoring unreachable code.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
The size of the pointer to a data structure to send is erroneously
passed to sst_ipc_tx_message_wait() as its tx_bytes argument. It should
be given the size of the pointed skl_ipc_dxstate_info structure instead.
Coincidentally, both the pointer and the structure have the same size of
8 bytes on a 64 bit machine, which "masks" the issue. Compiling for 32
bit reveals the issue more clearly.
Fix the typo for correctness, and to make the code robust to future
evolutions of the skl_ipc_dxstate_info structure size.
This fixes the following coccicheck error:
sound/soc/intel/skylake/skl-sst-ipc.c:641:8-14: ERROR: application of sizeof to pointer
Signed-off-by: Vincent Stehlé <vincent.stehle@laposte.net>
Cc: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Cc: Jeeja KP <jeeja.kp@intel.com>
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Mark Brown <broonie@kernel.org>
Cc: trivial@kernel.org
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
SSP0 FMT uses 24 bits so fix to the value to 24 bits
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
The driver runtime behaviour is fine but in suspend, we missed
setting the DSP to suspend and also missed resuming DSP on
resume.
Fix this by having common SKL suspend and resume routines which
power up/down links, suspend/resume DSP and other common
routines, and call these routines from both runtime as well as
system PM handlers
Signed-off-by: Jayachandran B <jayachandran.b@intel.com>
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Widget FW topology private data already has the information
on the channel map, ch_cfg and interleaving. This patch removes
the calculation of channel_map in driver and reads the value
directly from widget private data.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
DMIC NHLT entry is sample rate agnostic, so ignore the rate
checks for DMIC type
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
The DSP FW specifies loadable modules using GUIDs so add support to
specify the GUIDs from topology
Signed-off-by: Hardik T Shah <hardik.t.shah@intel.com>
Signed-off-by: Omair M Abdullah <omair.m.abdullah@intel.com>
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
This patch updates the topology interface structure alignment and
also updates the Sample interleaving defines
Signed-off-by: Hardik T Shah <hardik.t.shah@intel.com>
Signed-off-by: Omair M Abdullah <omair.m.abdullah@intel.com>
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
The module pin formats are considered homogeneous, but some
modules can have different pcm formats on different pins, like
reference signal for a module.
This patch add support for configuration of each pin of module
and allows us to specify if pins and homogeneous or heterogeneous
Signed-off-by: Hardik T Shah <hardik.t.shah@intel.com>
Signed-off-by: Omair M Abdullah <omair.m.abdullah@intel.com>
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Subhransu S. Prusty <subhransu.s.prusty@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
ASoC core already checks if BE is active. If BE is active,
hw_params callback is ignored.
This patch removes the redundant check in driver for copier
widget power check in update be hw_params.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
While rigourous testing of SKL drivers, we noticed underuns and
overuns and on debug realized that we need to change driver
handling of FE pipe startup and shutdown
We need to start DMA and then run pipe together and not split
these up. Similarly while stopping we should stop pipe and then
DMA in a sequence.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
In SKL topology routes, some paths can be connected by a widget
which are not a DSP FW widget and virtual with respect to
firmware. In these case when module has to bind, then the
virtual DSP modules needs to skipped till a actual DSP module is
found which connects the pipelines.
So we need to walk the graph and find a widget which is real in
nature. This patch adds that support and splits
skl_tplg_pga_dapm_pre_pmu_event() fn with parsing code to
skl_tplg_bind_sinks() fn and call that recursively as well as
while parsing
The patch moves code a bit while splitting so diffstat doesn't
tell real picture
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
in_pin and out_pin list for a module has the information about
the module that are bound together. So we can directly look at
pin information of module for binding and unbind.
As a result the preinitialized dapm_path_last we had is removed
and code and memory optimzed.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Bps calculation is not correct as this needs to be based on valid
bit depth. 16 bit fmt bit depth is 16 bit and for 24 and 32 bit
as it is container size This patch fixes the bps.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
For supporting multiple dynamic pins, module state check is
incorrect. In case of unbind, module state need to be changed to
uninit if all pins in the module is is unbind state.
To handle module state correctly add pin state and use pin
state check to set module state correctly.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
If link type is HDA, NHLT blob is null, as NHLT defines non HDA
links only. So we should ignore blob query for HDA links.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
We have both I2S and hda codec support in the driver. codec_mask
check is relevant only for hda codec and some boards may have
only I2S Codec, so removed probe error in case no hda codec is
found and update the log to info as it may not be error.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
In case of 32 bit, the FE update params returns error as it falls
thru to default case. This patch adds 32 bit depth handling in
update FE params.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Return value from skl_tplg_be_update_params() is ignored. But if the
blob is null then the hw_params needs to return error.
This patch fixes the issue by not ignoring return value from
skl_tplg_be_update_params().
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
To get the FE copier module, the check to ignore non DSP widgets
was wrong. This path corrects the check to ignore non DSP widget.
Signed-off-by: Jeeja KP <jeeja.kp@intel.com>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Add missing initialiser for control_set variable
in img_prl_out_set_fmt
Signed-off-by: Damien.Horsley <Damien.Horsley@imgtec.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Add driver for Imagination Technologies SPDIF output
controller
Signed-off-by: Damien.Horsley <Damien.Horsley@imgtec.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Add driver for Imagination Technologies SDPIF input
controller
Signed-off-by: Damien.Horsley <Damien.Horsley@imgtec.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Add driver for Imagination Technologies parallel output
controller
Signed-off-by: Damien.Horsley <Damien.Horsley@imgtec.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Add driver for Imagination Technologies I2S output
controller
Signed-off-by: Damien.Horsley <Damien.Horsley@imgtec.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Add driver for Imagination Technologies I2S input
controller
Signed-off-by: Damien.Horsley <Damien.Horsley@imgtec.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
using macro for default register map
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
STL and STR registers are writable and not readable. So they are non volatile.
Remove them from volatile list, and add default register value for them.
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>
|
|
ETDR and TX0~5 registers are writable and not readable. So they are non
volatile. Remove them from volatile list, and add default register value for
them.
Signed-off-by: Zidan Wang <zidan.wang@freescale.com>
Acked-by: Nicolin Chen <nicoleotsuka@gmail.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
During suspend/resume, there is a flow that if a driver does not support
SNDRV_PCM_INFO_RESUME, it will fail at snd_pcm_resume(), and user space
can then issue SNDRV_PCM_IOCTL_PREPARE to let audio continue to play.
However, in dpcm_be_dai_prepare() it only allows BEs to be prepared
in state SND_SOC_DPCM_STATE_HW_PARAMS or SND_SOC_DPCM_STATE_STOP.
The BE state will then stay in SND_SOC_DPCM_STATE_SUSPEND, consequently
dpcm_be_dai_shutdown() is skipped in the end of playback and
be_substream->runtime is not cleared while this runtime is actually freed
by snd_pcm_detach_substream(). If another suspend comes, a NULL pointer
dereference will happen in snd_pcm_suspend() when accessing
BE substream's runtime.
Signed-off-by: Koro Chen <koro.chen@mediatek.com>
Acked-by: Liam Girdwood <liam.r.girdwood@linux.intel.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Spotted by kbuild bot:
sound/soc/soc-ops.c:415: warning: No description found for parameter 'ucontrol'
sound/soc/soc-ops.c:415: warning: Excess function parameter 'uinfo' description in 'snd_soc_put_volsw_sx'
Reported-by: kbuild test robot <lkp@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Signed-off-by: Axel Lin <axel.lin@ingics.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
This patch adds support for the Cirrus Logic CS47L24 and WM1831 codecs.
Signed-off-by: Richard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
The SOC_xxx_DECL() macros already include 'const' so there's
no need to put a const in the source where they are used.
Signed-off-by: Richard Fitzgerald <rf@opensource.wolfsonmicro.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
|
|
Commit 939f325f4a0f ("usb: add usb_endpoint_maxp() macro") and commit
29cc88979a88 ("USB: use usb_endpoint_maxp() instead of le16_to_cpu()")
introduced a new helper macro. This trivial patch convert remaining
users found in ua101 driver.
Signed-off-by: Cheah Kok Cheong <thrust73@gmail.com>
Acked-by: Clemens Ladisch <clemens@ladisch.de>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
|
|
One of the many faults of the QinHeng CH345 USB MIDI interface chip is
that it does not handle received SysEx messages correctly -- every second
event packet has a wrong code index number, which is the one from the last
seen message, instead of 4. For example, the two messages "FE F0 01 02 03
04 05 06 07 08 09 0A 0B 0C 0D 0E F7" result in the following event
packets:
correct: CH345:
0F FE 00 00 0F FE 00 00
04 F0 01 02 04 F0 01 02
04 03 04 05 0F 03 04 05
04 06 07 08 04 06 07 08
04 09 0A 0B 0F 09 0A 0B
04 0C 0D 0E 04 0C 0D 0E
05 F7 00 00 05 F7 00 00
A class-compliant driver must interpret an event packet with CIN 15 as
having a single data byte, so the other two bytes would be ignored. The
message received by the host would then be missing two bytes out of six;
in this example, "F0 01 02 03 06 07 08 09 0C 0D 0E F7".
These corrupted SysEx event packages contain only data bytes, while the
CH345 uses event packets with a correct CIN value only for messages with
a status byte, so it is possible to distinguish between these two cases by
checking for the presence of this status byte.
(Other bugs in the CH345's input handling, such as the corruption resulting
from running status, cannot be worked around.)
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Cc: stable@vger.kernel.org
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
The CH345 USB MIDI chip has two output ports. However, they are
multiplexed through one pin, and the number of ports cannot be reduced
even for hardware that implements only one connector, so for those
devices, data sent to either port ends up on the same hardware output.
This becomes a problem when both ports are used at the same time, as
longer MIDI commands (such as SysEx messages) are likely to be
interrupted by messages from the other port, and thus to get lost.
It would not be possible for the driver to detect how many ports the
device actually has, except that in practice, _all_ devices built with
the CH345 have only one port. So we can just ignore the device's
descriptors, and hardcode one output port.
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Cc: stable@vger.kernel.org
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
Signed-off-by: Clemens Ladisch <clemens@ladisch.de>
Cc: <stable@vger.kernel.org>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
Use the new dmaengine_synchronize() function to make sure that all complete
callbacks have finished running before the runtime data, which is accessed
in the completed callback, is freed.
This fixes a long standing use-after-free race condition that has been
observed on some systems.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Reviewed-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Vinod Koul <vinod.koul@intel.com>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
Pull sound fixes from Takashi Iwai:
"Here are a collection of small fixes tha have been gathered for
4.4-rc1. The only significant changes are those in PCI drivers
Kconfig, to use "depends on" instead of "select" for CONFIG_ZONE_DMA.
A reverse select is often more user-friendly, but in this case, it
makes hard to manage with the conflict with ZONE_DEVICE, so changed in
such a way for now.
Others are all small fixes and quirks: an error check in soundcore
reigster_chrdev(), HD-audio HDMI/DP phantom jack fix, Intel Broxton DP
quirk, USB-audio DSD device quirk, some constifications, etc"
* tag 'sound-fix-4.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound:
ALSA: pci: depend on ZONE_DMA
ALSA: hda - Simplify phantom jack handling for HDMI/DP
ALSA: hda/hdmi - apply Skylake fix-ups to Broxton display codec
ALSA: ctxfi: constify rsc ops structures
ALSA: usb: Add native DSD support for Aune X1S
ALSA: oxfw: add an comment to Kconfig for TASCAM FireOne
sound: fix check for error condition of register_chrdev()
|
|
The counter is incremented/decremented in critical section protected with
mutex. Therefore, no need to use atomic_t.
This commit changes the type to unsigned int.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
Currently, critical section is protected by mutex in functions of
fireworks_stream.c. Callers increments/decrements substreams counter
before calling the functions. Moving mutex to the callers code allows
to change type of the substeram counter from atomic_t to unsigned int.
This commit is a preparation for obsoleting usage of atomic_t for
substream counter.
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
Commit a471fcde8c2c ("ALSA: dice: fix detection of Weiss devices") adds
a quirk of Weiss models. According to users' reports, Loud models also
have the similar quirk. They have 0x10 in the category field.
This commit adds support for Mackie Onyx Blackbird and Onyx-i series.
As long as I know, Dice-based models produced by
Focusrite/Alesis/PreSonus/M-Audio/TC Electronic have default value (0x04)
in their category field, thus it may be reasonable to add a condition
statement for Loud models, instead of removing the check of category value.
Reported-by: Rouge Etienne <erouge.externe@m6.fr>
Reported-by: Etilem <contact@etilem.net>
Signed-off-by: Takashi Sakamoto <o-takashi@sakamocchi.jp>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
Dell Latitude E6440 (1028:05bd) needs the same fixup as applied to
other Latitude E7xxx models for the click noise due to the recent
power-saving changes.
Bugzilla: http://bugzilla.opensuse.org/show_bug.cgi?id=954876
Cc: <stable@vger.kernel.org> # v4.1+
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
There are several sound drivers that 'select ZONE_DMA'. This is
backwards as ZONE_DMA is an architecture capability exported to drivers.
Switch the polarity of the dependency to disable these drivers when the
architecture does not support ZONE_DMA. This was discovered in the
context of testing/enabling devm_memremap_pages() which depends on
ZONE_DEVICE. ZONE_DEVICE in turn depends on !ZONE_DMA.
Reported-by: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
|
The HDMI codec parser may create a phantom jack, but the helper
function snd_hda_jack_add_kctl() treats always as a normal jack. This
is superfluous as the jack query is executed at each time the jack
sync is performed.
Since the HDMI codec parser is the only caller of this function, it's
easier to change back this directly calling the original
__snd_hda_jack_add_kctl() with phantom_jack parameter.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
|
