Commit graph

13,657 commits

Author SHA1 Message Date
Takashi Iwai
16ffe32c77 ALSA: hda - Fix line-out jack handling with STAC/IDT codec
When the line-out jack is plugged/unplugged, the driver needs to check
the headphone plug, not only the line-out jack itself.  Otherwise the
headphone or the speaker may be wrongly muted/unmuted.

As a result, both STAC_HP_EVENT and STAC_LO_EVENT need to call the
same function, stac92xx_hp_detect().

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-04 13:40:54 +02:00
Mark Brown
27ded041f0 ASoC: Factor out 7 bit register 9 bit data SPI write
This converts all the Wolfson drivers using this format (the only devices
that do) except WM8753 to use it.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-08-03 16:59:16 +01:00
Mark Brown
8d50e447d1 ASoC: Factor out I/O for Wolfson 8 bit data 16 bit register CODECs
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-08-03 16:59:16 +01:00
Mark Brown
afa2f1066e ASoC: Factor out I2C 8 bit address 16 bit data I/O
As part of this refactoring the type of the CODEC hw_read operation
is changed to match the regular read operation.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-08-03 16:59:15 +01:00
Mark Brown
7084a42b96 ASoC: Add I/O control bus information to factored out cache setup
While writes tend to be able to use a fairly bus independant format to
do the writes reads are all bus specific. To allow us to factor out
this code include the bus type as a parameter when setting up the
cache.

Initially just use this to factor out hw_write_t for I2C.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-08-03 16:59:09 +01:00
Takashi Iwai
15cfa2b3db ALSA: hda - Fix line-out jack detection
The commit fefd67f31e
    ALSA: hda - Add line-out jack detection on IDT/STAC codecs
enabled wrong pins for jack detections.  Fixed to the correct ones.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 14:23:33 +02:00
Lubomir Rintel
51840409b6 ALSA: pcm - Tell user that stream to be rewound is suspended
Return STRPIPE instead of EBADF when userspace attempts to rewind
of forward a stream that was suspended in meanwhile, so that it
can be recovered by snd_pcm_recover().

This was causing Pulseaudio to unload the ALSA sink module under a race
condition when it attempted to rewind the stream right after resume from
suspend, before writing to the stream which would cause it to revive the
stream otherwise. Tested to work with Pulseaudio patched to attempt to
snd_pcm_recover() upon receiving an error from snd_pcm_rewind().

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:57:00 +02:00
Takashi Iwai
e44d4e4cee Merge branch 'fix/hda' into topic/hda 2009-08-03 08:37:17 +02:00
Wu Fengguang
559059b27f ALSA: hda: add IbexPeak/Clarkdale HDMI model with static cvt/pin number
The new IbexPeak HDMI codec has 3 pin nodes and 2 converter nodes.
Here we assume only the first ones will be used.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:35:05 +02:00
Roel Kluin
4b35d2ca23 ALSA: hda - Read buffer overflow
Check whether index is within bounds before testing the element.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:34:06 +02:00
Chengu Wang
84d3dc200f ALSA: hda: Correct EAPD for Dell Inspiron 1525
The commit 24918b61b5 statically changes
the model from dell-bios to dell-3stack to solve the sound decreasing
regression (http://lkml.org/lkml/2008/9/12/203), however it leads to another
problem that the 2nd headphone jack doesn't work
(https://bugtrack.alsa-project.org/alsa-bug/view.php?id=3987). So I think
the commit 249**2dc is just a workaround. I would like to give a true solution
here.

The datasheet for STAC9228 says, GPIO2 is the same pin as VOL DOWN, and
the EAPD pin is GPIO0. This is why the sound decreases if we set EAPD as
GPIO2. This patch changes EAPD to GPIO0 to solve the problem.

Signed-off-by: Chengu Wang <wangchengu@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:30:56 +02:00
Wu Fengguang
e310bb0646 ALSA: hda: warn on spurious response
To help disclose hardware bugs.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:27:53 +02:00
Wu Fengguang
feb273404f ALSA: hda: remember last command for each codec
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:27:39 +02:00
Wu Fengguang
c32649feb4 ALSA: hda: read CORBWP inside reg_lock
This converts the last CORBWP access outside of reg_lock.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:26:55 +02:00
Wu Fengguang
cdb1fbf231 ALSA: hda: take reg_lock in azx_init_cmd_io/azx_free_cmd_io
Just for safety.  azx_init_cmd_io() and azx_free_cmd_io() may be
called when switching to single command mode.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:26:42 +02:00
Wu Fengguang
a678cdee25 ALSA: hda: take cmd_mutex in probe_codec()
Now that each codec will have its own module, it is possible
for the user to load one codec while another one is running.

So cmd_mutex would be a safe addition to probe_codec().

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:26:23 +02:00
Wu Fengguang
deadff1665 ALSA: hda: track CIRB/CORB command/response states for each codec
Recently we hit a bug in our dev board, whose HDMI codec#3 may emit
redundant/spurious responses, which were then taken as responses to
command for another onboard Realtek codec#2, and mess up both codecs.

Extend the azx_rb.cmds and azx_rb.res to array and track each codec's
commands/responses separately. This helps keep good codec safe from
broken ones.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:26:13 +02:00
Takashi Iwai
ce577e8cf5 ALSA: hda - Fix quirk for Toshiba Satellite A135-S4527
Use model=lenovo instead of model=dallas for Toshiba Satellite A135-S4527
with ALC861-VD codec.

Reference: Novell bnc#526325
	https://bugzilla.novell.com/show_bug.cgi?id=526325

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-08-03 08:23:52 +02:00
Janusz Krzysztofik
6d7f68a1ea ASoC: add support for Amstrad E3 (Delta) machine
This patch adds machine support for Amstrad E3 (Delta) videophone to ASoC.

Created and tested against linux-2.6.31-rc3.
Applies and works with linux-omap-2.6 commit
7c5cb7862d32cb344be7831d466535d5255e35ac as well.

Depends on:
1) latest version of the CX20442 codec driver that exposes v253_ops
   structure[1],
2) patch 2/3 form this series: TTY: Add definition of a new line
   discipline required by Amstrad E3 (Delta) ASoC driver[2].

CPU DAI parameters best matching the codec DAI has been selected out
empirically for best user experience.

Board specific audio function control (with related DAPM widgets) has been
modeled after empirically discovered codec capabilities.

Unlike other ASoC machine drivers, this one makes use of a codec provided line
discipline that is required for talking to a modem chip that can control the
codec behavoiur. As the line discipline operations must call board specific
bits as well, the machine driver registers its own line discipline ops, not
the codec provided, and then calls those codec provided from inside its own
callbacks.
If some kind of a glue, like a bus over a tty, exsited that could help in
runtime detection of a modem (bus adapter) over a more generic line discipline
(bus driver)[3], the line discipline code could be probably designed in a
more generic way.

In order to work at all, this driver requires a working McBSP1. On OMAP1510
based machines (not sure if other OMAP1 variants as well), where McBSP1 is a
DSP public peripheral, that means the kernel must provide basic DSP support,
ie. omap_dsp_init(), in order to power up the DSP. This used to be included in
linux-omap-2.6 tree up to commit 2512fd29db4eb09e82d182596304c7aaf76d2c5c.
Without that, the driver would not work, ie. not shift in/out any bits over
the CPU DAI[4]. This limitation is not board, but CPU specific, and may apply
to other code that makes use of McBSP1/McBSP3 on affected machines. I provide
an extra patch (4/3) as a temporary solution.

To work correctly in playback mode, this driver requires my prevoiusly
submitted patch that corrects pcm pointer calculation for OMAP1510 based
machines[5] (already included in linux-2.6.31-rc3).

To support codec controls, this driver requires my previously submitted patch
that adds support for modem found on Amstrad Delta[6].

[1] http://mailman.alsa-project.org/pipermail/alsa-devel/2009-July/019780.html
[2] http://www.spinics.net/lists/linux-serial/msg01862.html
[3] http://www.spinics.net/lists/linux-serial/msg01856.html
[4] http://www.spinics.net/lists/linux-omap/msg15114.html
[5] http://mailman.alsa-project.org/pipermail/alsa-devel/2009-June/018950.html
[6] http://www.spinics.net/lists/linux-omap/msg15432.html

Credits to:
Mark Underwood - for his initial, omap-alsa based sound driver for
this machine,
Mark Brown - for his help, patience and excellent subsytem maintainer support.

Signed-off-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-31 22:38:44 +01:00
Janusz Krzysztofik
ad120dae12 ASoC: CX20442: push down machine independent line discipline bits
This corrected patch adds machine independent line discipline code, prevoiusly
exsiting inside my Amstrad Delta ASoC machine dirver, to the Conexant CX20442
codec driver. The code can be used as a standalone line discipline, or as a
set of codec specific functions called from machine's line discipline
callbacks. Anyway, the line discipline itself must be registered by a machine
driver.

Applies on top of the followup to my initial driver version:
http://mailman.alsa-project.org/pipermail/alsa-devel/2009-July/019757.html

Suggested by ASoC manintainer Mark Brown <broonie@opensource.wolfsonmicro.com>

Signed-off-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-31 22:38:44 +01:00
Lars-Peter Clausen
b8e22c1fe3 ASoC: jack: Fix race in snd_soc_jack_add_gpios
The irq can fire as soon as it has been requested, thus all fields accessed
from within the irq handler must be initialized prior to requesting the irq.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-31 22:38:43 +01:00
Mark Brown
77ee09c67e ASoC: Allow CODECs to flag invalid registers
This helps CODECs with sparse register maps work better with the
register cache display interface.

Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-31 18:54:48 +01:00
Takashi Iwai
ec86fe5209 Merge branch 'fix/oss' into for-linus
* fix/oss:
  sound: mpu401.c: Buffer overflow
  sound: aedsp16: Buffer overflow
2009-07-31 10:17:45 +02:00
Takashi Iwai
d62e345f14 Merge branch 'fix/misc' into for-linus
* fix/misc:
  ALSA: sound/aoa: Add kmalloc NULL tests
2009-07-31 10:17:44 +02:00
Takashi Iwai
6280b61af5 Merge branch 'fix/hda' into for-linus
* fix/hda:
  ALSA: hda - Increase PCM stream name buf in patch_realtek.c
  ALSA: hda: fix out-of-bound hdmi_eld.sad[] write
  ALSA: hda - Add quirk for Dell Studio 1555
2009-07-31 10:17:42 +02:00
Julia Lawall
f065fabc86 ALSA: sound/aoa: Add kmalloc NULL tests
Check that the result of kzalloc is not NULL before a dereference.

The semantic match that finds this problem is as follows:
(http://www.emn.fr/x-info/coccinelle/)

// <smpl>
@@
expression *x;
identifier f;
constant char *C;
@@

x = \(kmalloc\|kcalloc\|kzalloc\)(...);
... when != x == NULL
    when != x != NULL
    when != (x || ...)
(
kfree(x)
|
f(...,C,...,x,...)
|
*f(...,x,...)
|
*x->f
)
// </smpl>

Signed-off-by: Julia Lawall <julia@diku.dk>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-31 10:14:58 +02:00
Takashi Iwai
aa563af763 ALSA: hda - Increase PCM stream name buf in patch_realtek.c
The name buf with size 16 is too short for some codec names, e.g.
truncated like "ALC861-VD Analo".  Now the size is doubled.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-31 10:06:34 +02:00
Takashi Iwai
03cb2dafcb Merge branch 'topic/hda-cirrus' into topic/hda 2009-07-30 18:09:04 +02:00
Takashi Iwai
d195658bd7 Merge branch 'fix/hda' into topic/hda 2009-07-30 18:08:54 +02:00
Takashi Iwai
fefd67f31e ALSA: hda - Add line-out jack detection on IDT/STAC codecs
Add the automatic mute of speakers via line-out jack plugging on
STAC/IDT codecs.  The feature is enabled when the HP detect is present.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-30 18:08:46 +02:00
Stelian Pop
3a38516750 ALSA: hda - Enable HP output with Macbook Pro 5, 5
The patch below, to be applied on the latest sound-unstable-2.6.git,
enables headphones output on my MacBookPro 5,5, together with the
automuting feature.

Here is the exact soundcard id:
	Vendor Id: 0x10134206
	Subsystem Id: 0x106b4d00
	Revision Id: 0x100301

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-30 14:57:25 +02:00
Mark Brown
a1daf67d72 Merge branch 'gta02-audio' into for-2.6.32 2009-07-30 13:21:38 +01:00
Takashi Iwai
5207e10ed4 ALSA: hda - Integrate Digital Input Source to Input Source
STAC/IDT codecs provide both "Input Source" and "Digital Input Source"
controls to choose the analog input source and the digital input source.
But this is far user-unfriendly.

This patch merges the input source selections into one "Input Source"
control.  To have separate digital and analog input source controls,
you can pass "separate_dmux = 1 " hint string.

At the same time, this patch gets rid of analog mixer stuff that was
already disabled in previous patches.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-30 13:09:08 +02:00
Takashi Iwai
bf677bd8fb ALSA: hda - Fix typos of Capture controls.
The commit 6479c63188
    ALSA: hda - Create Capture controls dynamically
introduced typos of "Capture".  Fixed now.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-30 09:24:29 +02:00
Wu Fengguang
6732bd0d15 ALSA: hda: add HP automute support to Intel ALC889/ALC889A models
It auto mutes all 8-channel outputs at rear panel when
the front panel headphone is connected.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-30 09:19:14 +02:00
Wu Fengguang
dd7714c922 ALSA: hda: add 2-channel mode to Intel ALC889/ALC889A models
This 2-channel mode is useful in that it will broadcast
a 2-channel audio stream to all front/side/... ports.

Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-30 08:53:56 +02:00
Barry Song
3a39f832a5 ASoC: Fix checkpatch issues and typos of ad1938 codec and bf5xx-tdm dai
1. fix "line over 80 characters" checkpatch warnings
2. ‘DMA_nnBIT_MASK’ is deprecated, use DMA_BIT_MASK instead
3. fix typos

Signed-off-by: Barry Song <21cnbao@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-29 21:31:53 +01:00
Lars-Peter Clausen
82c4362ee3 ASoC: neo1973_gta02_wm8753: Replace deprecated s3c_gpio calls with gpiolib
With the s3c platform has implementing gpiolib support the s3c_gpio api has been
deprecated.
This patch gets rid of all s3c_gpio calls and replaces them by using gpiolib.

Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-29 21:29:33 +01:00
Lars-Peter Clausen
69331fbdee ASoC: neo1973_gta02_wm8753: Replace snd_soc_cnew with snd_soc_add_controls.
Signed-off-by: Lars-Peter Clausen <lars@metafoo.de>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-29 21:29:33 +01:00
Takashi Iwai
71443b0b74 ALSA: hda - No analog mix input source as default for IDT92HD71bxx
The analog mix is disabled now as default (unless "analog_mixer" hint
is given), so it shoudn't appear in the digital input source as well.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-29 18:41:29 +02:00
Takashi Iwai
15b4f296fc ALSA: hda - Add missing DMUX initialization for auto-mic with STAC/IDT
Added the missing initialization of DMUX connection (to analog input)
for auto-mic mode with STAC/IDT codecs.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-29 16:32:55 +02:00
Takashi Iwai
26a2798053 ALSA: hda - Remove static connection in IDT 92HD71bxx
We don't need any more static connection to the port F (which is often
used for docking stations) since its connection is done dynamically via
DAC assignment now.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-29 16:30:14 +02:00
Roel Kluin
a987004fbc sound: mpu401.c: Buffer overflow
mpu_synth_info[m].name is a char[30], and the minimum length of the data
written by sprintf is 31 bytes including terminating null.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-29 14:41:24 +02:00
Roel Kluin
c45ec06c74 sound: aedsp16: Buffer overflow
DSPVersion is declared as char[3], but the sprintf writes at least 4 bytes
including terminating null.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-29 14:37:12 +02:00
Roel Kluin
78735cffc2 ALSA: hda: fix out-of-bound hdmi_eld.sad[] write
e->sad[] is declared with size ELD_MAX_SAD=16, but the guard
allows range 0-31.

Signed-off-by: Roel Kluin <roel.kluin@gmail.com>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-29 14:35:20 +02:00
Takashi Iwai
3d21d3f7e7 ALSA: hda - Support auto-mic switching with IDT/STAC codec
Support the automatic mic-switching with some devices with IDT/STAC
codecs.  The condition is that the device has only two inputs, one
for an external mic and one for an internal mic.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-29 14:32:56 +02:00
Takashi Iwai
62558ce157 ALSA: hda - Avoid overwrite of jack events with STAC/IDT
Since only one event can be associated to a (pin) widget, it's safer
to avoid the multiple mapping.  This patch fixes the behavior of the
STAC/IDT codec driver.

Now stac_get_event() doesn't take the type argument but simply returns
the first hit element.  Then enable_pin_detect() checks the validity
of the type, and returns non-zero only if a valid entry.  The caller
can call stac_issue_unsol_event() after checking the return value.

Signed-off-by: Takashi Iwai <tiwai@suse.de>
2009-07-29 14:28:37 +02:00
Barry Song
c8489c3ed3 ASoC: board driver to connect bf5xx with ad1938
Signed-off-by: Barry Song <21cnbao@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-28 22:16:02 +01:00
Barry Song
01e2ab207c ASoC: blackfin I2S(TDM mode) CPU DAI driver
The I2S DAI driver for blackfin SPORT, but works in TDM mode.
I2S is not a special case of TDM with only left and right two slots for
SPORT interface. I2S coordinates with TDM in SPORT, but not a part of
TDM. TDM require different hardware configuration with I2S, not only
different slot number.  One is "Stereo Serial Operation" mode of SPORT,
the other one is "Multichannel Operation" mode. They are incompatible
at the same time.
Hardware and DMA description and data transfer flow are much different
for I2S and TDM. Merging them as a whole will be very ugly and difficult
to maintain.
So we don't define a new DAI type, but give two DAI instances for standard
I2S and TDM, both in I2S-family DAI type. The TDM instance still uses the
I2S-family DAI type.

Signed-off-by: Barry Song <21cnbao@gmail.com>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-28 22:15:27 +01:00
Janusz Krzysztofik
b84eab08a6 ASoC: CX20442: fix issues pointed out by subsystem maintainer
The patch fixes some checkpatch identified issues and adds a comment about
line discipline interaction to my driver code, as requested by Mark on my
inital submission (thank you Mark for applying my imperfect patch anyway).
It also fixes MODULE_ALIAS mismatch as used in my machine driver.

Signed-off-by: Janusz Krzysztofik <jkrzyszt@tis.icnet.pl>
Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com>
2009-07-28 22:15:23 +01:00