Asoc: qcom: lpass:Update lpaif_dmactl members order
Update the lpaif_dmactl struct members order to match
HDMI reg map members sequence. Separate Interface reg map
as it is used for I2S control but not for HDMI control,
to make use of bulk API, which makes code more readable.
Signed-off-by: V Sujith Kumar Reddy <vsujithk@codeaurora.org> Reviewed-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Signed-off-by: Srinivasa Rao <srivasam@codeaurora.org> Tested-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Reviewed-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Link: https://lore.kernel.org/r/1602134223-2562-5-git-send-email-srivasam@codeaurora.org Signed-off-by: Mark Brown <broonie@kernel.org>
Update dts property read API call with platform get property
by name, as it make code more readable and avoid conflicts
when array of properties to be used.
Signed-off-by: V Sujith Kumar Reddy <vsujithk@codeaurora.org> Reviewed-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Signed-off-by: Srinivasa Rao <srivasam@codeaurora.org> Tested-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Reviewed-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> Link: https://lore.kernel.org/r/1602134223-2562-4-git-send-email-srivasam@codeaurora.org Signed-off-by: Mark Brown <broonie@kernel.org>
Mark Brown [Tue, 6 Oct 2020 14:24:41 +0000 (15:24 +0100)]
Merge series "ASoC: Intel: Remove obsolete solutions and components" from Cezary Rojewski <cezary.rojewski@intel.com>:
Follow up to catpt series as mentioned in:
[PATCH v10 00/14] ASoC: Intel: Catpt - Lynx and Wildcat point
https://www.spinics.net/lists/alsa-devel/msg116440.html
As catpt is a direct replacement to sound/soc/intel/haswell, it leaves a
lot of code redudant. The second legacy solution - baytrail - is
deprecated for a long time by sound/soc/intel/atom with SOF flavor
available too.
This series addresses the redudancy and removes obsolete code. Along
with the legacy solutions, all orphaned components are removed too.
As a consequence, further cleanups are unlocked: sound/soc/intel/skylake
becomes the sole user of processing code found in
sound/soc/intel/common. Those are not part of this series.
Changes in v2:
- just a rebase so patch 04/13 applies cleanly
- left the tags as no actual changes done in between
Cezary Rojewski (13):
ASoC: Intel: Remove haswell solution
ASoC: Intel: Remove max98090 support for baytrail solution
ASoC: Intel: Remove rt5640 support for baytrail solution
ASoC: Intel: Remove baytrail solution
ASoC: Intel: Remove SST ACPI component
ASoC: Intel: Remove SST firmware components
ASoC: Intel: Skylake: Unassign ram_read and read_write ops
ASoC: Intel: Remove unused DSP operations
ASoC: Intel: Remove unused DSP interface fields
ASoC: Intel: Remove SST-legacy specific constants
ASoC: Intel: Make atom components independent of sst-dsp
ASoC: Intel: Remove sst_pdata structure
ASoC: Intel: Remove sst_dsp_get_thread_context
Alex Dewar [Sun, 4 Oct 2020 10:25:36 +0000 (11:25 +0100)]
ASoC: omap-mcbsp: Fix use of uninitialised pointer
Commit 9c34d023dc35 ("ASoC: omap-mcbsp: Re-arrange files for core McBSP
and Sidetone function split"), in rearranging various files, also replaced
calls to platform_get_resource_by_name() + devm_ioremap_resource() with a
single call to devm_platform_ioremap_resource_byname(). However, the
struct resource is needed as we access its members so at present a null
pointer is dereferenced. Fix by doing things the old way.
Cezary Rojewski [Tue, 6 Oct 2020 06:49:07 +0000 (08:49 +0200)]
ASoC: Intel: Remove sst_dsp_get_thread_context
While sst_dsp_get_thread_context() is declared as solution-agnostic, it
is only used by /skylake/ solution. Majority of thread_context field
usages are direct accesses. Improve code cohesiveness and convert to
single usage model.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-14-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:49:05 +0000 (08:49 +0200)]
ASoC: Intel: Make atom components independent of sst-dsp
With sound/soc/intel/haswell and /baytrail gone, registers left within
sst-dsp header are atom-specific. Relocate these to atom internal header
to make atom truely independent of sound/soc/common processing code.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-12-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:49:02 +0000 (08:49 +0200)]
ASoC: Intel: Remove unused DSP operations
sound/soc/intel/common/ declares several helper functions for /intel/
solutions. In practice, differences between these - /haswell/ and
/skylake/ especially - led to many of the helpers being used only by a
single solution. As /skylake/ makes no use of these and /haswell/ and
/baytail/ are no more, remove the unused functions.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-9-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:49:01 +0000 (08:49 +0200)]
ASoC: Intel: Skylake: Unassign ram_read and read_write ops
Skylake driver makes no use of ram_read or ram_write operation so remove
the assignments. This prepares sound/soc/common/sst-dsp* for following
removal of unused DSP operations.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-8-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:49:00 +0000 (08:49 +0200)]
ASoC: Intel: Remove SST firmware components
sst-firmware is host to many image loading over DMA operations. Majority
of code targets sound/soc/intel/haswell solution as /baytrail/ never
switched to DMA-based firmware loading. With /haswell/ removed this code
serves no purpose. Address this redundancy.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-7-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:48:59 +0000 (08:48 +0200)]
ASoC: Intel: Remove SST ACPI component
baytrail and haswell solutions present within sound/soc/intel are the
only users of sst-acpi componenent and with them removed it becomes
redundant so remove it too.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-6-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:48:58 +0000 (08:48 +0200)]
ASoC: Intel: Remove baytrail solution
sound/soc/intel/baytrail is a niche solution which supports limited
number of BYT products - as described by
snd_soc_acpi_intel_baytrail_legacy_machines table. For a long time it's
deprecated in favor of sound/soc/intel/atom solution with SOF providing
support for some products too effectively rendering /baytrail/ redundant.
Remove deprecated code from ASoC tree.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-5-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:48:57 +0000 (08:48 +0200)]
ASoC: Intel: Remove rt5640 support for baytrail solution
byt-rt5640 is deprecated in favor of bytcr_rt5640 used by
sound/soc/intel/atom and SOF solutions both. Remove redundant machine
board and all related code.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-4-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:48:56 +0000 (08:48 +0200)]
ASoC: Intel: Remove max98090 support for baytrail solution
byt-max98090 is deprecated in favor of cht-bsw-max98090 used by
sound/soc/intel/atom and SOF solutions both. Remove redundant machine
board and all related code.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-3-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 6 Oct 2020 06:48:55 +0000 (08:48 +0200)]
ASoC: Intel: Remove haswell solution
Newly added catpt solution found in sound/soc/intel/catpt is a direct
replacement to sound/soc/intel/haswell. It covers all features supported
by it and more - by aligning to recommended flows and requirement list
based on Windows driver equivalent. No harm is done to userspace as
catpt - similarly to haswell - loads no extenal topology files while
sharing the exact same ADSP firmware binary.
Given the above, existing haswell code is redundant so remove it.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Acked-by: Liam Girdwood <liam.r.girdwood@intel.com> Link: https://lore.kernel.org/r/20201006064907.16277-2-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Sun, 4 Oct 2020 09:06:09 +0000 (11:06 +0200)]
ASoC: Intel: haswell: Mark FE DAIs as nonatomic
PCM operations for DAI links connected with DSP platform component
involve communication with DSP firmware by IPCs. As IPC protocol may
cause thread to sleep while waiting for a response from DSP, propagate
that information to ALSA core by marking all FE DAIs as nonatomic.
Cezary Rojewski [Sun, 4 Oct 2020 09:06:08 +0000 (11:06 +0200)]
ASoC: Intel: broadwell: Mark FE DAIs as nonatomic
PCM operations for DAI links connected with DSP platform component
involve communication with DSP firmware by IPCs. As IPC protocol may
cause thread to sleep while waiting for a response from DSP, propagate
that information to ALSA core by marking all FE DAIs as nonatomic.
Cezary Rojewski [Sun, 4 Oct 2020 09:06:07 +0000 (11:06 +0200)]
ASoC: Intel: bdw-rt5677: Mark FE DAIs as nonatomic
PCM operations for DAI links connected with DSP platform component
involve communication with DSP firmware by IPCs. As IPC protocol may
cause thread to sleep while waiting for a response from DSP, propagate
that information to ALSA core by marking all FE DAIs as nonatomic.
Cezary Rojewski [Sun, 4 Oct 2020 09:06:06 +0000 (11:06 +0200)]
ASoC: Intel: bdw-rt5650: Mark FE DAIs as nonatomic
PCM operations for DAI links connected with DSP platform component
involve communication with DSP firmware by IPCs. As IPC protocol may
cause thread to sleep while waiting for a response from DSP, propagate
that information to ALSA core by marking all FE DAIs as nonatomic.
Mark Brown [Mon, 5 Oct 2020 14:32:14 +0000 (15:32 +0100)]
Merge series "ASoC: sun8i-codec: support for AIF2 and AIF3" from Samuel Holland <samuel@sholland.org>:
This series adds support the other two AIFs present in the sun8i codec,
which can be used for codec2codec DAI links.
This series first cleans up the DAPM component driver so there is an
organized place to put the new widgets. Then it fills out the DAI
driver, removing assumptions that were made for AIF1 (16 bits, 2
channels, certain clock inversions). Some new logic is required to
handle 3 DAIs and the ADC/DAC sharing the same clock. Finally, it adds
the new DAIs, and hooks them up with DAPM widgets and routes per the
hardware topology.
To minimize the number of patches in this series, related device tree
patches (increasing #sound-dai-cells, adding new DAI links) will be sent
separately.
Samuel Holland (25):
ASoC: sun8i-codec: Set up clock tree at probe time
ASoC: sun8i-codec: Swap module clock/reset dependencies
ASoC: sun8i-codec: Sort DAPM controls, widgets, and routes
ASoC: sun8i-codec: Consistently name DAPM widgets and routes
ASoC: sun8i-codec: Correct DAPM widget types
ASoC: sun8i-codec: Fix AIF widget channel references
ASoC: sun8i-codec: Enable AIF mono/stereo control
ASoC: sun8i-codec: Use snd_soc_dai_get_drvdata
ASoC: sun8i-codec: Prepare to extend the DAI driver
ASoC: sun8i-codec: Program format before clock inversion
ASoC: sun8i-codec: Enable all supported clock inversions
ASoC: sun8i-codec: Program the correct word size
ASoC: sun8i-codec: Round up the LRCK divisor
ASoC: sun8i-codec: Correct the BCLK divisor calculation
ASoC: sun8i-codec: Support the TDM slot binding
ASoC: sun8i-codec: Enforce symmetric DAI parameters
ASoC: sun8i-codec: Enable all supported sample rates
ASoC: sun8i-codec: Automatically set the system sample rate
ASoC: sun8i-codec: Constrain to compatible sample rates
ASoC: sun8i-codec: Protect the clock rate while streams are open
ASoC: sun8i-codec: Require an exact BCLK divisor match
ASoC: sun8i-codec: Enable all supported PCM formats
ASoC: sun8i-codec: Generalize AIF clock control
ASoC: sun8i-codec: Add a DAI, widgets, and routes for AIF2
ASoC: sun8i-codec: Add a DAI, widgets, and routes for AIF3
ASoC: Intel: sof_sdw: add version_id to avoid rt714/rt715 confusion
RT715 and RT714 are essentially the same chip. In addition, there are
two versions, one supporting SoundWire 1.1 and one supporting
SoundWire 1.2 (SDCA).
The previous configurations assumed that RT714 was SDCA-only, which
isn't correct. Add support for the 4 possible combinations to avoid
confusions.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao@linux.intel.com> Reviewed-by: Jack Yu <jack.yu@realtek.com> Reviewed-by: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com> Link: https://lore.kernel.org/r/20201002211902.287692-5-pierre-louis.bossart@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Samuel Holland [Thu, 1 Oct 2020 02:11:30 +0000 (21:11 -0500)]
ASoC: sun8i-codec: Enable AIF mono/stereo control
Each left/right pair of AIF input/output channels can be swapped or
combined. This is useful for sending a mono audio source to both sides
of a stereo sink, or for creating complex mixing scenarios.
Add the support to control this feature from userspace.
Both the left and right side widgets referenced channel 0. This would
unnecessarily power on the right side widget (and its associated path)
when a mono stream was active.
Samuel Holland [Thu, 1 Oct 2020 02:11:28 +0000 (21:11 -0500)]
ASoC: sun8i-codec: Correct DAPM widget types
Whie the aif_in and aif_out widget types are handled exactly the same in
the core DAPM code, a future widget event hook will need the correct
widget type to derive the associated substream. Clean up the widget type
for that reason, and so these widgets will match newly-added widgets for
the other AIFs.
Samuel Holland [Thu, 1 Oct 2020 02:11:27 +0000 (21:11 -0500)]
ASoC: sun8i-codec: Consistently name DAPM widgets and routes
This cleans up the mixer widget names. The AIF1 AD0 Mixer names were
previously wrong -- they do not control the digital side of the ADC. The
DAC mixer widgets were not wrong, but they were verbose and did not
match the naming scheme of the other widgets.
The mixer controls are not renamed because they are exposed to
userspace.
Samuel Holland [Thu, 1 Oct 2020 02:11:26 +0000 (21:11 -0500)]
ASoC: sun8i-codec: Sort DAPM controls, widgets, and routes
Sort the remaining pieces of the DAPM driver so that they are all in the
same order among controls/widgets/routes, and so they roughly match the
register word and bit order of the hardware. This nicely separates the
AIF-related widgets from the ADC/DAC widgets, which allows the AIF
widgets to stay in a logical order as more AIFs are added to the driver.
No widgets are renamed, to ease verification that this commit makes no
functional change.
This matches the module power-up/down sequence from the vendor's driver.
While updating these widgets/routes, reorder them to match the register
and bit layout of the hardware. This puts them in the same place in the
widget and route arrays (previously they were at opposite ends), and it
makes it easier to track which parts of which registers are implemented.
Samuel Holland [Thu, 1 Oct 2020 02:11:24 +0000 (21:11 -0500)]
ASoC: sun8i-codec: Set up clock tree at probe time
The sun8i codec is effectively an on-die variant of the X-Powers AC100
codec. The AC100 can derive its clocks from either of two I2S master
clocks or an internal PLL. For the on-die variant, Allwinner replaced
the codec's own PLL with a connection to SoC's existing PLL_AUDIO, and
they connected both I2S MCLK inputs to the same source -- which happens
to be an integer divider from the same PLL_AUDIO.
So there's actually no clocking flexibility. To run SYSCLK at the
required rate, it must be run straight from the PLL. The only choice is
whether it goes through AIF1CLK or AIF2CLK. Since both run at the same
rate, the only effect of that choice is which field in SYS_SR_CTRL
(AIF1_FS or AIF2_FS) controls the system sample rate.
Since AIFnCLK is required to bring up the corresponding DAI, and AIF1
(connected to the CPU) is used most often, let's use AIF1CLK as the
SYSCLK parent. That means we no longer need to set AIF2_FS.
Since this clock tree never changes, we can program it from the
component probe function, instead of using DAPM widgets. The DAPM
widgets unnecessarily change clock parents when the codec goes in/out
of idle and the supply widgets are powered up/down.
Mark Brown [Fri, 2 Oct 2020 20:05:30 +0000 (21:05 +0100)]
Merge series "Add driver for Microchip S/PDIF RX" from Codrin Ciubotariu <codrin.ciubotariu@microchip.com>:
The Sony/Philips Digital Interface Receiver (SPDIFRX) is a serial port
compliant with the IEC-60958 standard. Among its caracteristics, we
mention the following:
- SPDIF/AES-EBU Compatible Serial Port
- 32 Samples FIFO
- Data Width Configurable to 24 bits, 20 bits or 16 bits
- Packed and Unpacked Data Support for System Memory Optimization
- Line State Events Report and Source of Interrupt
- Line Error Rate Report
- Full Memory Map of 192 bits for Channel 1 and Channel 2 Status and
User Data
- First 32-bit Status A, Status B Change Report and Source of Interrupt
- Line Digital Filter
- Register Write Protection
- Abnormal Software Access and Internal Sequencer Integrity Check Reports
This interface is available in Microchip's SAMA7G5 SoC.
Codrin Ciubotariu (2):
dt-bindings: sound: add DT bindings for Microchip S/PDIF RX Controller
ASoC: mchp-spdifrx: add driver for SPDIF RX
Brent Lu [Wed, 15 Jul 2020 13:01:50 +0000 (21:01 +0800)]
ASoC: hdac_hdmi: remove cancel_work_sync in runtime suspend
A deadlock is identified when there are three contexts running at the
same time:
- a HDMI jack work which is calling snd_soc_dapm_sync().
- user space is calling snd_pcm_release() to close pcm device.
- pm is calling runtime suspend function of HDMI codec driver.
By removing the clear_dapm_works() invocation in the
hdac_hdmi_runtime_suspend() function, the snd_pcm_release() could
always returns from dapm_power_widgets() function call without
blocking the hdac_hdmi_jack_dapm_work() work thread or being blocked
by the hdac_hdmi_runtime_suspend() function. The purpose of the jack
work is to enable/disable the dapm jack pin so it's not necessary to
cancel the work in runtime suspend function which is usually called
when pcm device is closed.
Mark Brown [Fri, 2 Oct 2020 14:49:12 +0000 (15:49 +0100)]
Merge series "ASoC: Intel: Catpt - Lynx and Wildcat point" from Cezary Rojewski <cezary.rojewski@intel.com>:
Implement support for Lynxpoint and Wildcat Point AudioDSP. Catpt
solution deprecates existing sound/soc/intel/haswell which is removed in
the following series.
Due to high range of errors and desynchronization from recommendations
set by Windows solution, re-write came as a lower-cost solution compared
to refactoring /haswell/ with several series of patches.
Series is dependent on linux-spi change:
spi: pxa2xx: Add SSC2 and SSPSP2 SSP registers
https://www.spinics.net/lists/linux-spi/msg23885.html
which has been already merged and is now part of linux-spi tree.
Bulk of series content is device driver core code - everything up to
patch 7/14 - with fs entries and trace macros introduced right after.
While each core patch is shaped in such a way that no unavailable
members are ever called, until patch 14/14 is applied, no code
compilation can occur as no Makefile is present. Once said patch is
added, Makefile and Kconfig are implemented and driver module compiles
as expected.
Special thanks go to Marcin Barlik and Piotr Papierkowski for sharing
their LPT/WPT AudioDSP architecture expertise as well as helping
backtrack its historical background.
My thanks go to Amadeusz Slawinski for reviews and improvements proposed
on and off the internal list. Most of internal diff below is his
contribution.
Krzysztof Hejmowski helped me setup my own Xtensa environment and
recompile LPT/WPT FW binary sources what sped up the development greatly.
This would not have been possible without help from these champions,
especially considering how quickly the catpt was written: 2 weeks
features, 3 weeks optimizations. Thank you.
Userspace-exposed members are compatible with what is exposed by
deprecated solution as well as FW binary being re-used thus no harm is
done. The only visible differences are: the newly added 'Loopback Mute'
kcontrol and volume support extending to quad from stereo.
On top of fixing erros and design flows, catpt also adds module reload,
dynamic SRAM memory allocation during PCM runtime and exposes missing
userspace API: 'Loopback Mute' kcontrol, quad volume controls and sysfs
fw-version entries. Event tracing is provided to ease solution
debugging.
Following are not included in this update and are scheduled as later
addition:
- fw logging
- module (library) support
Note: LPT power up/down sequences might get aligned with WPT once enough
testing is done as capabilities are shared for both DSPs.
Note #2: Both LPT and WPT power up/down sequences may get optimized in
future updates as thanks to help from the Windows team, most of nuances
behind why/what/when in regard to hw registers have been backtracked and
reviewed again.
Link to developer's deep dive message:
https://www.spinics.net/lists/alsa-devel/msg113563.html
Changes in v10:
- reverted DUAL_MONO case relocation from v9
- indented all constants of enum catpt_module_id to the same column
- new newline appended for return path of catpt_dsp_do_send_msg()
Changes in v9:
https://www.spinics.net/lists/alsa-devel/msg116305.html
- fixed newlines in sysfs as requested by Andy, left tags as no other
changes done
- removed volume_map and replaced by simple formulas for volume kcontrol
calculations
- removed redundant parentheses in catpt_get_channel_map() and
relocated DUAL_MONO case
- runtime suspend no longer called during module unload
- removed redundant size checks for catpt_dsp_send_tx() and
catpt_dsp_copy_rx()
Changes in v8:
https://www.spinics.net/lists/alsa-devel/msg116168.html
- updated catpt_arrange_page_table() with GENMASK and U32_MAX usage
- made use of PFN_DOWN() replacing explicit right shitfs by PAGE_SIZE
- made fw hash dumping in catpt_coredump() more readable and removed
hardcodes
- catpt_coredump() dumps fw hash now only if said segment has been found
within fw_info
- shortened _MSECS suffixes to _MS
- IPC structs no longer contain enum members
- simplified definition of catpt_set_dspvol()
Changes in v7:
https://www.spinics.net/lists/alsa-devel/msg116019.html
- fixed licence header for fs.c
- renamed fs.c to sysfs.c to better match its purpose
- added documentation within Documentation/ABI/testing for entries
exposed by catpt
- bin_attribute fw_build replaced by attribute fw_info:
fw_info contains full FW information and after successful handshake,
it's always available (stored in driver data) so no need to invoke
GET_FW_VERSION IPC again, just dump the stored information
- rather than manually creating and removing sysfs files, now makes use
of dev_groups member of struct device_driver
- patch: 10/14 'ASoC: Intel: Select catpt and deprecate haswell' has
been moved to the back of the list: enable catpt after machine boards
have been prepared for it first
- improved readability of several goto labels
Changes in v6:
https://www.spinics.net/lists/alsa-devel/msg115765.html
- reordered and reorganized code for patches 1/13 - 8/13 of v5, so each
patches makes use of no member or function which is unavailable to it.
Series size increased from 13 to 14 patches: addition of base members
e.g.: registers has been split from addition of device.c file which
describes acpi device behavior
Changes in v5:
https://www.spinics.net/lists/alsa-devel/msg115621.html
Basically everything below is result of Andy's review. Thank you Andy
for taking time into this detailed review
- catpt now makes use of common linux/pxa2xx_ssp.h header file, removing
redundant SSP register declarations in the process. As stated in the
opening, this is dependent upon linux-spi change:
spi: pxa2xx: Add SSC2 and SSPSP2 SSP registers
- updated Kconfig by removing DMADEVICES and adding COMPILE_TEST
as optional depends-on
- updated all register macros definitions to be more safe against common
arithmetics when specifying macro's parameters
- removed CONFIG_PM and CONFIG_PM_SLEEP usage in favor of __maybe_unused
- all 'if (ret < 0)' converted to simple 'if (ret)' whenever possible
- fixed erroneous check for platform_device_register_data within
catpt_register_board()
- _SLAVE/_MASTER replaced with more inclusive _CONSUMER/_PROVIDER for
enum catpt_ssp_mode
- catpt_acpi_probe() is now making use of high-level wrappers for
ioremapping and resource assignment, reducing function's code size
- due to improved catpt_acpi_probe() behavior, catpt_acpi_remove() needs
not to cast dma_free_coherent() any longer
- DMA source and destrination maxburst now of value 16, see:
https://www.spinics.net/lists/alsa-devel/msg114394.html
- simplified catpt_dsp_update_lpclock() as list_for_each_entry() is
empty-safe by default
- dropped '_SSP_' from all names of all CATPT_SSP_SSXXX_DEFAULT macros
- catpt_updatel_pci now makes use of linux/pci.h and uapi/linux/pci.h
constants such as: PCI_PM_CTRL_STATE_MASK and PCI_D3hot
Changes in v4:
https://www.spinics.net/lists/alsa-devel/msg113762.html
- fixed compilation with i386 kconfig (conflicting names)
- streamlined naming for SHIM and PCI registers to match SSP ones
(SHIM_REG -> SHIM)
- catpt_component_probe removed and kcontrols again initializzed
statically via snd_kcontrol_new array: this is to remove
kctl->id.device shenanigans
- renamed catpt_set_ctlvol to catpt_set_dspvol - function name wasn't
matching its purpose
Changes in v3:
- fixed IRAM mask usage in lpt_dsp_power_up (dsp.c)
- updated dbg message formatting in catpt_restore_fwimage as suggested
by Andy
- fixed alignment for struct catpt_ssp_device_format
- catpt_set_ctlvol now verifies all-equal scenario based on all
channels rather than just first two as requested by Amadeo
- fixed SPDX for registers.h
Changes in v2:
https://www.spinics.net/lists/alsa-devel/msg113660.html
- fixed SPDX formatting for all header files as well as pcm.c
- fixed size provided to memcpy() in fw_build_read() as reported by Mark
- renamed struct catpt_pdata to struct catpt_spec (cosmetic)
- fixed erroneous path in catpt_load_block: region is properly released
- trace.h events for updating registers have been removed and usages
replaced by dev_dbg (SRAMPGE/ LPCS)
- as requested by Andy, struct resource has replaced struct catpt_mbank
and struct catpt_mregion. This change cascaded into:
- catpt_mbank_size and catpt_mregion_size replaced by resource_size
- catpt_mregion_intersects replaced by resource_overlaps
- all catpt_mbank_ and catpt_mregion_ handlers found in loader.c
(_request, _reserve, _release, _extract, _split, _join) have been
removed
- __request_region and __release_region have been enlisted in their
place
- catpt_mregion_intersecting renamed to catpt_resource_overlapping
- catpt_request_region helper has been provided to deal with -size
based requests
o haven't found direct replacements in resource.c/ ioport.h for
both functions
- catpt_mbank_create and catpt_mbank_remove renamed to catpt_sram_init
and catpt_sram_free respectively
- catpt_sram_init now returns void instead of int and has been
converted to simple initialized. This change ultimately cascaded
into:
o both SRAM banks initialization being moved to catpt_dev_init
from catpt_acpi_probe (device.c)
o catpt_dev::spec is now initialized first, with catpt_dev_init
following it soon after
o catpt_acpi_probe erroneous path has been simplified as SRAM
banks no longer need to be freed
- catpt_sram_free now frees all resources via child -> sibling
enumeration rather than region_list iteration
- catpt_dsp_update_srampge and catpt_dsp_set_srampge now accept new
argument: unsigned long mask. Caused by removal of catpt_mbank -
mask is taken directly from catpt_dev::spec::d/iram_mask
- trace.h events for catpt_mbank and catpt_mregion have been removed
Diff against last drop on internal list:
https://www.spinics.net/lists/alsa-devel/msg113549.html
- replaced spinlock with mutex for mregion allocation and release to
address sleeping in atomic context warnings
- fixed coredump fw_hash dumping
- kcontrol values are now always stored regardless of stream of interest
is running or not
- kcontrol values are now applied after stream is prepared instead of
ignoring what has been set by user initially
- catpt_pdata instances have been renamed from hsw_ and bdw_ to lpt_ and
wpt_ respectively
- reordered Makefile .o(s) (cosmetic)
Randy Dunlap [Thu, 1 Oct 2020 18:35:37 +0000 (11:35 -0700)]
ASoC: qcom: fix SDM845 & QDSP6 dependencies more
Fix a build error and Kconfig warning in sound/soc/qcom/.
ld: sound/soc/qcom/qdsp6/q6afe-clocks.o: in function `q6afe_clock_dev_probe':
q6afe-clocks.c:(.text+0x182): undefined reference to `devm_clk_hw_register'
ld: q6afe-clocks.c:(.text+0x19d): undefined reference to `of_clk_add_hw_provider'
After adding "depends on COMMON_CLK" for SND_SOC_QDSP6, the Kconfig
warning appears because "select" does not honor any "depends on"
clauses, so fix the dependency for SND_SOC_SDM845 also.
Prevent sound/soc/intel/haswell code compile and select catpt instead as
a recommended solution. Userspace-exposed members are compatible with
what is exposed by deprecated solution thus no harm is done. The only
visible difference is the newly added 'Loopback Mute' kcontrol.
ASoC: Intel: bdw-5677: Remove haswell-solution specific code
Remove code specific to sound/soc/intel/haswell. Update BE dai_link
definition to provide seamless transition to catpt solution.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-14-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: Intel: bdw-5650: Remove haswell-solution specific code
Remove code specific to sound/soc/intel/haswell. Update BE dai_link
definition to provide seamless transition to catpt solution.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-13-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: Intel: broadwell: Remove haswell-solution specific code
Remove code specific to sound/soc/intel/haswell. Update BE dai_link
definition to provide seamless transition to catpt solution.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-12-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: Intel: haswell: Remove haswell-solution specific code
Remove code specific to sound/soc/intel/haswell. Update BE dai_link
definition to provide seamless transition to catpt solution.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-11-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Define tracing macros for easy catpt debug. These cover all IPC message
types: requests, replies and notifications.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-9-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Implement ACPI device probing and removal functions as well as handlers
for its PM capabilities. Device probing also takes care of enumerating
ADSP subsystem components.
DSP designed for Lynxpoint and Wildcat Point offers no dynamic topology
i.e. all pipelines are already defined within firmware and host is
relegated to allocing stream for predefined pins. This is represented by
'catpt_topology' member.
Implementation covers all available pin types:
- system playback and capture
- two offload streams
- loopback (reference)
- bluetooth playback and capture
PCM DAI operations differentiate between those pins as some (mainly
offload) are to be handled differently - DSP expects wp updates on each
notify_position notification.
System playback has no volume control capability as it is routed to
mixer stream directly. Other primary streams - capture and two offloads
- offer individual volume controls.
Compared to sound/soc/intel/haswell this configures SSP device format
automatically on pcm creation.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-7-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: Intel: catpt: Firmware loading and context restore
For Lynxpoint and Wildcat Point solution, is it host's responsibility to
allocate SRAM regions and ensure those already taken are not overwritten
with other data until released. Blocks are transferred to SRAM - either
IRAM or DRAM - via DW DMA controller. Once basefw is booted, ownership
of DMA transfer is lost in favour of DSP.
Hosts reponsibilities don't end on initial block allocation and binary
transfer. During Dx transitions host must store FW runtime context from
DRAM before putting AudioDSP subsystem into lower power state. Said
context gets flashed after D0 entry to bring DSP right where it was just
before suspending.
Load and restore procedures are finalized with SRAM power gating and
adequate clock level selection. This power gates unused EBBs and clock
speed effectively reducing power consumption.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-6-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Implement dsp lifecycle functions such as core RESET and STALL,
SRAM power control and LP clock selection. This also adds functions for
handling transport over DW DMA controller.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-5-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Declare global and stream IPC message handlers for all known message
types.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-4-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Implement IRQ handlers for immediate and delayed replies and
notifications. Communication is synchronous and allows for serialization
of maximum one message at a time.
DSP may respond with ADSP_PENDING status for a request - known as
delayed reply - and when situation occurs, framework keeps the lock and
awaits upcoming response through IPCD channel which is handled in
bottom-half. Immediate replies spawn no BH at all as their processing is
very short.
Declare base structures, registers and extension routines for the catpt
solution.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200929141247.8058-2-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Cezary Rojewski [Tue, 25 Aug 2020 20:17:43 +0000 (22:17 +0200)]
spi: pxa2xx: Add SSC2 and SSPSP2 SSP registers
Update list of SSP registers with SSC2 and SSPSP2. These registers are
utilized by LPT/WPT AudioDSP architecture.
While SSC2 shares the same offset (0x40) as SSACDD, description of this
register for SSP device present on mentioned AudioDSP is different so
define separate constant to avoid any ambiguity.
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Link: https://lore.kernel.org/r/20200825201743.4926-1-cezary.rojewski@intel.com Signed-off-by: Mark Brown <broonie@kernel.org>
Tang Bin [Wed, 26 Aug 2020 15:09:18 +0000 (23:09 +0800)]
ASoC: fsl_spdif: Fix unnecessary check in fsl_spdif_probe()
The function fsl_spdif_probe() is only called with an openfirmware
platform device. Therefore there is no need to check that the passed
in device is NULL.
ASoC: soc-pcm: add missing ret=0 at soc_pcm_open()
commit 140a4532cdb8c ("ASoC: soc-pcm: add soc_pcm_clean() and call it
from soc_pcm_open/close()") switched to use soc_pcm_clean() at
soc_pcm_open().
But it removed "return 0", and missing "ret = 0",
because of it, it always return -EINVAL eventhough no error.
This patch adds missing "ret = 0" for success case.
Reported-by: Marek Szyprowski <m.szyprowski@samsung.com> Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx@renesas.com> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com> Link: https://lore.kernel.org/r/87ft6ya65z.wl-kuninori.morimoto.gx@renesas.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: soc-pcm: ignore un-needed mutex_unlock() case on soc_pcm_open()
commit 140a4532cdb8c ("ASoC: soc-pcm: add soc_pcm_clean() and call it
from soc_pcm_open/close()") switch to call soc_pcm_clean() on
soc_pcm_open() when rollback case.
But, it uses "goto err" (A) *before* mutex_lock() (B) when error of
snd_soc_pcm_component_pm_runtime_get().
The mutex_unlock() (C) is not needed in such case. This patch fix it.
static int soc_pcm_open(...)
{
...
ret = snd_soc_pcm_component_pm_runtime_get(rtd, substream);
if (ret < 0)
(A) goto err;
Since commit 95e0e07e710e ("ASoC: atmel-pcm: use generic dmaengine
framework"), the driver is using dmaengine and is not using any definition
from include/linux/platform_data/dma-atmel.h, stop including it.
Since commit d5fab59cab18 ("spi: atmel: trivial: remove unused fields in
DMA structure"), the driver is not using any definitions from
linux/platform_data/dma-atmel.h, stop including it.
This kind of duplicated code can be a hotbed of bugs,
thus, this patch-set share soc_pcm_close() and rollback.
v1 -> v2
- indicate more detail background/logic on git-log
Link: https://lore.kernel.org/r/87wo2oku0m.wl-kuninori.morimoto.gx@renesas.com Link: https://lore.kernel.org/r/87wo1kvozz.wl-kuninori.morimoto.gx@renesas.com
Kuninori Morimoto (7):
ASoC: soc-dai: add mark for snd_soc_dai_startup/shutdown()
ASoC: soc-link: add mark for snd_soc_link_startup/shutdown()
ASoC: soc-component: add mark for soc_pcm_components_open/close()
ASoC: soc-component: add mark for
snd_soc_pcm_component_pm_runtime_get/put()
ASoC: soc-pcm: add soc_pcm_clean() and call it from
soc_pcm_open/close()
ASoC: soc-pcm: remove unneeded dev_err() for snd_soc_dai_startup()
ASoC: soc-pcm: remove unneeded dev_err() for
snd_soc_component_module/open()
regmap: add support to regmap_field_bulk_alloc/free apis
Usage of regmap_field_alloc becomes much overhead when number of fields
exceed more than 3.
QCOM LPASS driver has extensively converted to use regmap_fields.
Using new bulk api to allocate fields makes it much more cleaner code to read!
Lucas Tanure [Mon, 28 Sep 2020 11:18:21 +0000 (12:18 +0100)]
ASoC: cs4234: Add support for Cirrus Logic CS4234 codec
The CS4234 is a highly versatile CODEC that combines 4 channels of
high performance analog to digital conversion, 4 channels of high
performance digital to analog conversion for audio, and 1 channel of
digital to analog conversion to provide a nondelayed audio reference
signal to an external Class H tracking power supply.
DAC5 is only supported as a 5th audio channel. Tracking Power Supply
mode is not currently supported by the driver.
In DSP_A mode the slots for DAC1-4 and optionally DAC5 can be set.
The codec always claims 4 slots for DAC1-4 and these must be in the
same nibble of the mask. The codec has a fixed mapping for ADC slots.
In I2S/LJ modes the codec has a fixed mapping for DAC1-4 and ADC1-4.
DAC5 is not available in these modes.
The MCLK source must be preset to a valid frequency before probe()
because it must be running all the time the codec is out of reset.
The VA_SEL bit will be set automatically to 3.3v or 5v during probe()
based on the reported voltage of the regulator supplying VA.
Signed-off-by: Lucas Tanure <tanureal@opensource.cirrus.com> Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com> Link: https://lore.kernel.org/r/20200928111821.26967-2-rf@opensource.cirrus.com Signed-off-by: Mark Brown <broonie@kernel.org>
ASoC: dt-bindings: sgtl5000: Add common clock properties
Add common properties appearing in DTSes (assigned-clocks and similar)
to fix dtbs_check warnings like:
arch/arm64/boot/dts/freescale/imx8mq-librem5-devkit.dt.yaml: audio-codec@a:
'assigned-clock-parents', 'assigned-clock-rates', 'assigned-clocks' do not match any of the regexes: 'pinctrl-[0-9]+'
ASoC: soc-pcm: remove unneeded dev_err() for snd_soc_component_module/open()
snd_soc_component_module_get(), snd_soc_component_open() itself will
indicate error message, thus, soc_pcm_components_open() don't need to
handle it.
This patch removes these.
Now, 1) to 5) are handled.
This patch adds new soc_pcm_clean() and call it from
soc_pcm_open() as rollback, and from soc_pcm_close() as
normal close handler.
One note here is that it don't need to call snd_soc_runtime_deactivate()
when rollback case, because it will be called without
snd_soc_runtime_activate().
It also don't need to call snd_soc_dapm_stream_stop() when rollback case.
The idea of having bit-flag or counter is not enough for this purpose.
For example if one DAI is used for 2xPlaybacks for some reasons,
and if 1st Playback was succeeded but 2nd Playback was failed,
2nd Playback rollback doesn't need to call shutdown.
But it has succeeded bit-flag or counter via 1st Playback,
thus, 2nd Playback rollback will call unneeded shutdown.
And 1st Playback's necessary shutdown will not be called,
because bit-flag or counter was cleared by wrong 2nd Playback rollback.
To avoid such case, this patch marks substream pointer when get() was
succeeded. If rollback needed, it will check rollback flag and marked
substream pointer.
One note here is that it cares *current* get() only now.
but we might want to check *whole* marked substream in the future.
This patch is using macro named "push/pop", so that it can be easily
update.
This patch is for 3) snd_soc_component_module_get/put()
4) snd_soc_component_open/close().
The idea of having bit-flag or counter is not enough for this purpose.
For example if one DAI is used for 2xPlaybacks for some reasons,
and if 1st Playback was succeeded but 2nd Playback was failed,
2nd Playback rollback doesn't need to call shutdown.
But it has succeeded bit-flag or counter via 1st Playback,
thus, 2nd Playback rollback will call unneeded shutdown.
And 1st Playback's necessary shutdown will not be called,
because bit-flag or counter was cleared by wrong 2nd Playback rollback.
To avoid such case, this patch marks substream pointer when open() was
succeeded. If rollback needed, it will check rollback flag and marked
substream pointer.
One note here is that it cares *current* open() only now.
but we might want to check *whole* marked substream in the future.
This patch is using macro named "push/pop", so that it can be easily
update.
This patch is for 2) snd_soc_link_startup/shutdown().
The idea of having bit-flag or counter is not enough for this purpose.
For example if one DAI is used for 2xPlaybacks for some reasons,
and if 1st Playback was succeeded but 2nd Playback was failed,
2nd Playback rollback doesn't need to call shutdown.
But it has succeeded bit-flag or counter via 1st Playback,
thus, 2nd Playback rollback will call unneeded shutdown.
And 1st Playback's necessary shutdown will not be called,
because bit-flag or counter was cleared by wrong 2nd Playback rollback.
To avoid such case, this patch marks substream pointer when startup() was
succeeded. If rollback needed, it will check rollback flag and marked
substream pointer.
One note here is that it cares *current* startup() only now.
but we might want to check *whole* marked substream in the future.
This patch is using macro named "push/pop", so that it can be easily
update.
This patch is for 1) snd_soc_dai_startup/shutdown().
The idea of having bit-flag or counter is not enough for this purpose.
For example if one DAI is used for 2xPlaybacks for some reasons,
and if 1st Playback was succeeded but 2nd Playback was failed,
2nd Playback rollback doesn't need to call shutdown.
But it has succeeded bit-flag or counter via 1st Playback,
thus, 2nd Playback rollback will call unneeded shutdown.
And 1st Playback's necessary shutdown will not be called,
because bit-flag or counter was cleared by wrong 2nd Playback rollback.
To avoid such case, this patch marks substream pointer when startup() was
succeeded. If rollback needed, it will check rollback flag and marked
substream pointer.
One note here is that it cares *current* startup() only now.
but we might want to check *whole* marked substream in the future.
This patch is using macro named "push/pop", so that it can be easily
update.
projects / mirror_ubuntu-kernels.git / log