Add missing pci_set_drv to NULL call on error.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Link: https://lore.kernel.org/r/20231009115437.99976-6-maarten.lankhorst@linux.intel.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

The hda_codec_i915_init() errors are ignored in
hda_init() so it can never return -EPROBE_DEFER.

Fix this before we move the call to hda_init() from the
deferred probe to early probe.

While at it, also fix error handling when hda_dsp_ctrl_get_caps
fails.
Suggested-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Acked-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20231009115437.99976-5-maarten.lankhorst@linux.intel.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

This patch moves the initial parts of the probe to the probe_early()
callback, which provides a much faster decision on whether the SOF
driver shall deal with a specific platform or yield to other Intel
drivers.

This is a limited functionality change, the bigger change is to move
the i915/Xe initialization to the probe_early().
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Acked-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Acked-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20231009115437.99976-4-maarten.lankhorst@linux.intel.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

The existing DSP probe may be handled in a workqueue to allow for
extra time, typically for the i915 request_module and HDAudio codec
handling.

With the upcoming changes for i915/Xe driver relying on the
-EPROBE_DEFER mechanism, we need to have a first pass of the probe
which cannot be pushed to a workqueue. Introduce 2 new optional
callbacks.

probe_early is called before the workqueue runs. remove_late may be
called from the workqueue if load is unsuccesful, but will otherwise
be called on module unload.
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Acked-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Acked-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Link: https://lore.kernel.org/r/20231009115437.99976-3-maarten.lankhorst@linux.intel.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

In an effort to not call sof_ops_free twice, we stopped running it when
probe was aborted.

Check the result of cancel_work_sync to see if this was the case.

Fixes: 31bb7bd9 ("ASoC: SOF: core: Only call sof_ops_free() on remove if the probe was successful")
Cc: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Acked-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Acked-by: Peter Ujfalusi <peter.ujfalusi@linux.intel.com>
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://lore.kernel.org/r/20231009115437.99976-2-maarten.lankhorst@linux.intel.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

For applying HD-audio EPROBE_DEFER series cleanly.
Signed-off-by: Takashi Iwai <tiwai@suse.de>

Merge tag 'asoc-fix-v6.6-rc6' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

ASoC: Fixes for v6.6

A fairly large set of fixes here but all driver specific, the biggest
block is Johan's work shaking out issues with device setup and teardown
for the wcd938x driver which is a relatively large but clearly broken
down set of changes.

There is one core helper function added as part of a fix for wsa-macro.

Enable Gnd switch to improve stability when Jack insert event
occurs, and then disable Gnd switch after Jack type detection
is finished.
Signed-off-by: David Rau <David.Rau.opensource@dm.renesas.com>
Link: https://lore.kernel.org/r/20231017021258.5929-1-David.Rau.opensource@dm.renesas.comSigned-off-by: Mark Brown <broonie@kernel.org>

ASUS platform Headset Mic was disable by default.
Assigned verb table for Mic pin will enable it.
Signed-off-by: Kailang Yang <kailang@realtek.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/1155d914c20c40569f56d36c79254879@realtek.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

Enables the SPI-connected Cirrus amp and the required pins
for headset mic detection.

As of BIOS version 313 it is still necessary to modify the
ACPI table to add the related _DSD properties:
  https://gist.github.com/Flex1911/1bce378645fc95a5743671bd5deabfc8Signed-off-by: Artem Borisov <dedsa2002@gmail.com>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20231014075044.17474-1-dedsa2002@gmail.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

This HP Laptop uses ALC236 codec with COEF 0x07 controlling the
mute LED. Enable existing quirk for this device.
Signed-off-by: Luka Guzenko <l.guzenko@web.de>
Cc: <stable@vger.kernel.org>
Link: https://lore.kernel.org/r/20231016221328.1521674-1-l.guzenko@web.deSigned-off-by: Takashi Iwai <tiwai@suse.de>

Commit d6d6c513 ("ASoC: dwc: Use ops to get platform data")
converted the DesignWare I2S driver to use a DT specific function to
obtain platform data but this breaks at least non-DT systems such as
AMD.  Revert it.

Fixes: d6d6c513 ("ASoC: dwc: Use ops to get platform data")
Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20231013-asoc-fix-dwc-v1-1-63211bb746b9@kernel.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Correctly log failures of reset via I2C.
Signed-off-by: Roy Chateau <roy.chateau@mep-info.com>
Link: https://lore.kernel.org/r/20231013110239.473123-1-roy.chateau@mep-info.comSigned-off-by: Mark Brown <broonie@kernel.org>

The RT5650 should enable a power setting for button detection to avoid the wrong result.
Signed-off-by: Shuming Fan <shumingf@realtek.com>
Link: https://lore.kernel.org/r/20231013094525.715518-1-shumingf@realtek.comSigned-off-by: Mark Brown <broonie@kernel.org>

The call to gpiod_set_value_cansleep() in cs42l42_sdw_update_status()
needs the header file gpio/consumer.h to be included.

This was introduced by commit 2d066c6a ("ASoC: cs42l42: Avoid stale
SoundWire ATTACH after hard reset")

and caused error:
    sound/soc/codecs/cs42l42-sdw.c:374:4: error: implicit declaration of
    function ‘gpiod_set_value_cansleep’;
    did you mean gpio_set_value_cansleep’?
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Fixes: 2d066c6a ("ASoC: cs42l42: Avoid stale SoundWire ATTACH after hard reset")
Link: https://lore.kernel.org/r/20231011134853.20059-1-rf@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

Due to an error in the datasheet the bias sense values currently don't
match the hardware. Whilst this is a change to the binding no devices
have yet shipped so updating the binding will not cause any issues.
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20231010102425.3662364-2-ckeepax@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

Due to an error in the datasheet the bias sense values currently don't
match the hardware. Whilst this is a change to the binding no devices
have yet shipped so updating the binding will not cause any issues.
Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
Link: https://lore.kernel.org/r/20231010102425.3662364-1-ckeepax@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

The ASP1 DOUT line must be defaulted to be high-impedance when
it is not actually transmitting data for an active channel.

In non-SoundWire modes ASP1 will usually be shared by multiple
amps so each amp must only drive the line during the slot for
an enabled TX channel.

In SoundWire mode a custom firmware can use ASP1 as a secondary
chip-to-chip audio link or as GPIO. It should be defaulted to
high-impedance since by default the purpose of this pin is not
known.

Backport note:
On kernel versions before 6.6 the cs35l56->base.regmap argument
to regmap_set_bits() must be changed to cs35l56->regmap.
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Fixes: e4961125 ("ASoC: cs35l56: Add driver for Cirrus Logic CS35L56")
Link: https://lore.kernel.org/r/20231009153412.30380-1-rf@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

Merge series from Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>:

Some codec drivers compare widget names with strcmp, ignoring the component
name prefix.  If prefix is used, the comparisons start failing.
Add a helper to fix the issue.

Merge series from Johan Hovold <johan+linaro@kernel.org>:

The wcd938x codec driver happily ignores error handling, something which
has bitten us in the past when we hit a probe deferral:

	https://lore.kernel.org/lkml/20230705123018.30903-1-johan+linaro@kernel.org/

Fix up the remaining probe and component bind paths that left resources
allocated and registered after errors to avoid similar future issues.

Free the "priv" pointer before returning the error code.

Fixes: 90eb6b59 ("ASoC: pxa-ssp: add support for an external clock in devicetree")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Link: https://lore.kernel.org/r/84ac2313-1420-471a-b2cb-3269a2e12a7c@moroto.mountainSigned-off-by: Mark Brown <broonie@kernel.org>

Fix cs35l56_patch() to call reinit_completion() to reinitialize
the completion object.

It was incorrectly using init_completion().
Signed-off-by: Richard Fitzgerald <rf@opensource.cirrus.com>
Fixes: e4961125 ("ASoC: cs35l56: Add driver for Cirrus Logic CS35L56")
Link: https://lore.kernel.org/r/20231006164405.253796-1-rf@opensource.cirrus.comSigned-off-by: Mark Brown <broonie@kernel.org>

Make sure to balance the runtime PM operations, including the disable
count, on probe errors and on driver unbind.

Fixes: 16572522 ("ASoC: codecs: wcd938x-sdw: add SoundWire driver")
Cc: stable@vger.kernel.org      # 5.14
Cc: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Link: https://lore.kernel.org/r/20231003155558.27079-8-johan+linaro@kernel.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Make sure to deregister the component when the driver is being unbound
and before the underlying device-managed resources are freed.

Fixes: 16572522 ("ASoC: codecs: wcd938x-sdw: add SoundWire driver")
Cc: stable@vger.kernel.org      # 5.14
Cc: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Link: https://lore.kernel.org/r/20231003155558.27079-7-johan+linaro@kernel.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Make sure to balance the runtime PM operations, including the disable
count, on driver unbind.

Fixes: 16572522 ("ASoC: codecs: wcd938x-sdw: add SoundWire driver")
Cc: stable@vger.kernel.org      # 5.14
Cc: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Link: https://lore.kernel.org/r/20231003155558.27079-6-johan+linaro@kernel.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Make sure to disable and free the regulators on probe errors and on
driver unbind.

Fixes: 16572522 ("ASoC: codecs: wcd938x-sdw: add SoundWire driver")
Cc: stable@vger.kernel.org      # 5.14
Cc: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Link: https://lore.kernel.org/r/20231003155558.27079-5-johan+linaro@kernel.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Add the missing code to release resources on bind errors, including the
references taken by wcd938x_sdw_device_get() which also need to be
dropped on unbind().

Fixes: 16572522 ("ASoC: codecs: wcd938x-sdw: add SoundWire driver")
Cc: stable@vger.kernel.org      # 5.14
Cc: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Link: https://lore.kernel.org/r/20231003155558.27079-4-johan+linaro@kernel.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Make sure to deregister the component before tearing down the resources
it depends on during unbind().

Fixes: 16572522 ("ASoC: codecs: wcd938x-sdw: add SoundWire driver")
Cc: stable@vger.kernel.org      # 5.14
Cc: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Link: https://lore.kernel.org/r/20231003155558.27079-3-johan+linaro@kernel.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Drop the bogus error handling for a soundwire device backcast during
bind() that cannot fail.

Fixes: 16572522 ("ASoC: codecs: wcd938x-sdw: add SoundWire driver")
Cc: stable@vger.kernel.org      # 5.14
Cc: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Signed-off-by: Johan Hovold <johan+linaro@kernel.org>
Link: https://lore.kernel.org/r/20231003155558.27079-2-johan+linaro@kernel.orgSigned-off-by: Mark Brown <broonie@kernel.org>

When comparing widget names in wsa_macro_spk_boost_event(), consider
also the component's name prefix.  Otherwise the WSA codec won't have
proper mixer setup resulting in no sound playback through speakers.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20231003155710.821315-3-krzysztof.kozlowski@linaro.orgSigned-off-by: Mark Brown <broonie@kernel.org>

Some drivers use one event callback for multiple widgets but still need
to perform a bit different actions based on actual widget.  This is done
by comparing widget name, however drivers tend to miss possible name
prefix.  Add a helper to solve common mistakes.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20231003155710.821315-2-krzysztof.kozlowski@linaro.orgSigned-off-by: Mark Brown <broonie@kernel.org>

I own an external usb Webcam, model NexiGo N930AF, which had low mic volume and
inconsistent sound quality. Video works as expected.

(snip)
[  +0.047857] usb 5-1: new high-speed USB device number 2 using xhci_hcd
[  +0.003406] usb 5-1: New USB device found, idVendor=1bcf, idProduct=2283, bcdDevice=12.17
[  +0.000007] usb 5-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[  +0.000004] usb 5-1: Product: NexiGo N930AF FHD Webcam
[  +0.000003] usb 5-1: Manufacturer: SHENZHEN AONI ELECTRONIC CO., LTD
[  +0.000004] usb 5-1: SerialNumber: 20201217011
[  +0.003900] usb 5-1: Found UVC 1.00 device NexiGo N930AF FHD Webcam (1bcf:2283)
[  +0.025726] usb 5-1: 3:1: cannot get usb sound sample rate freq at ep 0x86
[  +0.071482] usb 5-1: 3:2: cannot get usb sound sample rate freq at ep 0x86
[  +0.004679] usb 5-1: 3:3: cannot get usb sound sample rate freq at ep 0x86
[  +0.051607] usb 5-1: Warning! Unlikely big volume range (=4096), cval->res is probably wrong.
[  +0.000005] usb 5-1: [7] FU [Mic Capture Volume] ch = 1, val = 0/4096/1

Set up quirk cval->res to 16 for 256 levels,
Set GET_SAMPLE_RATE quirk flag to stop trying to get the sample rate.
Confirmed that happened anyway later due to the backoff mechanism, after 3 failures

All audio stream on device interfaces share the same values,
apart from wMaxPacketSize and tSamFreq :

(snip)
Interface Descriptor:
      bLength                 9
      bDescriptorType         4
      bInterfaceNumber        3
      bAlternateSetting       3
      bNumEndpoints           1
      bInterfaceClass         1 Audio
      bInterfaceSubClass      2 Streaming
      bInterfaceProtocol      0
      iInterface              0
      AudioStreaming Interface Descriptor:
        bLength                 7
        bDescriptorType        36
        bDescriptorSubtype      1 (AS_GENERAL)
        bTerminalLink           8
        bDelay                  1 frames
        wFormatTag         0x0001 PCM
      AudioStreaming Interface Descriptor:
        bLength                11
        bDescriptorType        36
        bDescriptorSubtype      2 (FORMAT_TYPE)
        bFormatType             1 (FORMAT_TYPE_I)
        bNrChannels             1
        bSubframeSize           2
        bBitResolution         16
        bSamFreqType            1 Discrete
        tSamFreq[ 0]        44100
      Endpoint Descriptor:
        bLength                 9
        bDescriptorType         5
        bEndpointAddress     0x86  EP 6 IN
        bmAttributes            5
          Transfer Type            Isochronous
          Synch Type               Asynchronous
          Usage Type               Data
        wMaxPacketSize     0x005c  1x 92 bytes
        bInterval               4
        bRefresh                0
        bSynchAddress           0
        AudioStreaming Endpoint Descriptor:
          bLength                 7
          bDescriptorType        37
          bDescriptorSubtype      1 (EP_GENERAL)
          bmAttributes         0x01
            Sampling Frequency
          bLockDelayUnits         0 Undefined
          wLockDelay         0x0000
(snip)

Based on the usb data about manufacturer, SPCA2281B3 is the most likely controller IC
Manufacturer does not provide link for datasheet nor detailed specs.
No way to confirm if the firmware supports any other way of getting the sample rate.

Testing patch provides consistent good sound recording quality and volume range.

(snip)
[  +0.045764] usb 5-1: new high-speed USB device number 2 using xhci_hcd
[  +0.106290] usb 5-1: New USB device found, idVendor=1bcf, idProduct=2283, bcdDevice=12.17
[  +0.000006] usb 5-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[  +0.000004] usb 5-1: Product: NexiGo N930AF FHD Webcam
[  +0.000003] usb 5-1: Manufacturer: SHENZHEN AONI ELECTRONIC CO., LTD
[  +0.000004] usb 5-1: SerialNumber: 20201217011
[  +0.043700] usb 5-1: set resolution quirk: cval->res = 16
[  +0.002585] usb 5-1: Found UVC 1.00 device NexiGo N930AF FHD Webcam (1bcf:2283)
Signed-off-by: Christos Skevis <xristos.thes@gmail.com>
Link: https://lore.kernel.org/r/20231006155330.399393-1-xristos.thes@gmail.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

It has been confirmed that all devices in the Focusrite Clarett USB
series work the same as the devices in the Clarett+ series. Add the
missing PIDs to enable support for the Clarett 2Pre and 4Pre USB.
Signed-off-by: Geoffrey D. Bennett <g@b4.vu>
Link: https://lore.kernel.org/r/ZSFB8EVTG1PK1eq/@m.b4.vuSigned-off-by: Takashi Iwai <tiwai@suse.de>

AudioDSP firmware is the one who kicks SDxFIFOS calculation when a
stream is decoupled mode. During firmware bring up procedure, there is
no firmware running and the code-loading stream is always a decoupled
one. So, there is none to trigger the calculation and we end up with
false-positive timeout (-110) messages.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Acked-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20231006102857.749143-4-cezary.rojewski@intel.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

SDxFIFOS register indicates the fifo size directly. There is no need to
modify the value after reading the register.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20231006102857.749143-3-cezary.rojewski@intel.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

While AudioDSP drivers assign streams exclusively of HOST or LINK type,
nothing blocks a user to attempt to assign a COUPLED stream. As
supplied substream instance may be a stub, what is the case when
code-loading, such scenario ends with null-ptr-deref.
Signed-off-by: Cezary Rojewski <cezary.rojewski@intel.com>
Link: https://lore.kernel.org/r/20231006102857.749143-2-cezary.rojewski@intel.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

Merge tag 'asoc-fix-v6.6-rc5' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus

ASoC: More fixes for v6.6

Some additional fixes for v6.6, some fairly unremarkable driver specific
ones and a couple of minor core fixes for error handling and improved
logging.

asm/prom.h should not be included directly as it no longer contains
anything drivers need. Drivers should include of.h and/or other headers
which were getting implicitly included.
Signed-off-by: Rob Herring <robh@kernel.org>
Link: https://lore.kernel.org/r/20231003163209.770750-1-robh@kernel.orgSigned-off-by: Takashi Iwai <tiwai@suse.de>

Add new control element 'PCM Slave Access Mode' which shows the access
mode (interleaved/non-interleaved) for the PCM playing device. Add
corresponding control change notification calls.
Signed-off-by: Ivan Orlov <ivan.orlov0322@gmail.com>
Reviewed-by: Jaroslav Kysela <perex@perex.cz>
Link: https://lore.kernel.org/r/20230927113555.14877-2-ivan.orlov0322@gmail.comSigned-off-by: Takashi Iwai <tiwai@suse.de>

The current version of the loopback driver supports interleaved access
mode only. This patch introduces support for the non-interleaved
access mode.

When in the interleaved mode, the 'copy_play_buf' function copies data
from the playback to the capture buffer using one memcpy call. This call
copies samples for multiple, interleaved channels.

In the non-interleaved mode we have multiple channel buffers, so we have
to perform multiple memcpy calls to copy samples channel after channel.

Add new function called 'copy_play_buf_part_n', which copies a part of
each channel buffer from playback to capture. Modify the 'copy_play_buf'
to use the corresponding memory copy function(just memcpy /
copy_play_buf_part_n) depending on the access mode.
Signed-off-by: Ivan Orlov <ivan.orlov0322@gmail.com>
Reviewed-by: Jaroslav Kysela <perex@perex.cz>
Link: https://lore.kernel.org/r/20230927113555.14877-1-ivan.orlov0322@gmail.comSigned-off-by: Takashi Iwai <tiwai@suse.de>