Put the all static quirk codes out of patch_realtek.c, split into the
file for each codec model.  For controlling the build of quirk codes,
a new Kconfig, CONFIG_SND_HDA_ENABLE_REALTEK_QUIRKS is introduced.
By setting this off, all quirk codes won't be built, thus you can save
lots of memory.

The codes in patch_realtek.c are also shuffled and more comments are
given, but the contents aren't changed.  This is just a refactoring.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Avoid open-codes.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This patch changes the auto-parser and the auto-mic handling codes to
allow more flexible dynamic ADC-switching with Realtek codecs.

In the new code, the following strategy is taken:

- When a cap-src can't handle all input-sources, either skip it, or
  switch to the ADC-switching mode.  In ADC-switching mode, like the
  former dual-ADC mode for ALC275, it changes ADC on the fly according
  to the current input source.
- When auto-mic is possible, always assign imux.  If the mic pins are
  set statically via a quirk, rebuild imux according to the pins.
  In the auto-mic mode, the driver always changes the imux (although
  the imux isn't exposed as a mixer element).
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The digital out pin on ALC882 may have multiple connections.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Instead of assigning each default hda_pcm_stream pointers, do NULL-checks
and assign default values in alc_build_pcms().
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The only different implmentation was alc880_auto_init_input_src(),
and now it covers this variant, and we can use the single function
for all codecs.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Now with the new code for looking for ADCs and MUXs, we can replace
the whole ADC assignment with the parsed results.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

All alc*_auto_init_analog_input() calls are identical, so let's use
the same function more clearly without aliases.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Parse ADCs and cap-srcs in alc_auto_create_input_ctls() by itself
instead of passing explicitly from the caller.  By this change, all
alc*_auto_create_input_ctls() can be unified to the same calls.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Use the same common code for auto-parsing the output paths and their
initializations, based on the existing ALC662 code, which is smarter
than the old ALC880/2 code.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

When multiple inputs are present on the mixer widget (typically a DAC
and a loopback), mute/unmute both inputs with the corresponding mixer
element.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The initialization of DACs was missing in ALC662 parser code.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

In alc662_auto_fill_dac_nids(), the HP and speaker DACs aren't parsed
when the corresponding pins aren't fixed with single DACs.
Now check these DACs even for non-fixed pins.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

When the dual-adc switching mode is active in Realtek auto-parser,
we need to couple all ADCs as a single capture-volume.  Currently, the
volume control changes only the first ADC, thus others may remain silent.
This patch fixes the problem.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Add the support of ALC269VC codec.
Also delete the unnecessary codec_variant type enum list:
now only three variants (ALC269VA ALC269VB ALC269VC) are needed.

In addition, added some aliases:
 - Add ALC269VB alias name ALC277
 - Add ALC269VC alias name ALC259 ALC281X
 - Add ALC269VC for Lenovo device 0x21f3 name ALC3202
Signed-off-by: NKailang Yang <kailang@realtek.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Since we now route the front DAC via aa-mix widget, adding the aa-mix
to surrounds will result in a mix-up of both front and surround PCM
signals.  For avoiding this, the aa-mix routes have to be disabled
for surround paths.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

When the individual DAC is available for the headphone output, the driver
should create the DAC for its volume control.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Signed-off-by: NTakashi Iwai <tiwai@suse.de>

For VT1718S, the multi-channel path should be like following:
DAC 0-->Mixer 9(index 5)-->Mixer 0(index 1)-->Front Pin;
DAC 1-->Mixer 1(index 0)-->Surround Pin;
DAC 2-->C/LFE Pin;
DAC 3-->Mixer 2(index 0)-->Side Pin;

But current code built Surround and Side path through index 1 of
Mixer 1 and 2. So Adjusting Surround and Side channel amplifier is
invalid. This patch fixes the issue.
Signed-off-by: NLydia Wang <lydiawang@viatech.com.cn>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

For VT1718S, Mixer 9 doesn't expose the connection to DAC 0. So when
building up a 'PCM Playback' amplifier control, it will fail since
getting DAC 0 index of Mixer 9 returned -1. So I added a dac_mixer_idx
to indicated the actual index of DAC 0 to Mixer 9. Following is the
patch and next mail is another.
Signed-off-by: NLydia Wang <lydiawang@viatech.com.cn>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Unmute DAC on front speaker path when Independent HP is enabled.

When to enable Independent HP, the front speaker won't output any sound
for VT1708, VT1708B, VT1708S and VT1702.
I find the via_independent_hp_put() routine will mute DAC 0 path in Mixer 0.
For these codecs, when using Independent HP, there could have two
independent streams, one is from DAC0-->Mixer0-->Front Pin, the other is
from DAC3-->GainSW3-->Side Pin.
So I added a check for DAC-->Mixer path in activate_output_path().

If current path is DAC-->Mixer, no need to mute DAC index in Mixer.
In fact, to change connection of Headphone pin or Mux connected with HP
is enough.
Signed-off-by: NLydia Wang <lydiawang@viatech.com.cn>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

A missing initialization resulted in wrong DAC assignments in
ALC662 (and other) auto-parsers.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

This patch adds the re-implementation of Independent-HP mode in the
case where the DAC is shared between HP and side-channel streams.
Now the driver tries to parse the output-path using the pre-parsed
side-channel DAC for the independent HP output, too.

When a playback PCM stream is opened with this shared mode, the
Independent-HP mixer switch can't be changed for avoiding the conflict,
thus it returns -EBUSY error.

One remaining unintuitive issue is that the DAC volume is still
controlled as "Side" volume although it's shared by both independent-HP
and side streams.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

The char can be unsigned on some architectures.  Since the code checks
the negative values, they should be declared as signed char explicitly.

  sound/pci/rme9652/hdspm.c:5449: warning: comparison is always false due to limited range of data type
  sound/pci/rme9652/hdspm.c:5462: warning: comparison is always false due to limited range of data type
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Fix the wrongly converted short values:
  sound/pci/cs5535audio/cs5535audio_pcm.c:152: warning: large integer implicitly truncated to unsigned type
  sound/pci/cs5535audio/cs5535audio_pcm.c:160: warning: large integer implicitly truncated to unsigned type
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

I noticed that the last character of the ELD monitor name is lost,
this fixes the issue.

This fix should be confirming to the HDA spec, and works together with
the DRM part of the ELD patch.

The HDA spec does not mention that Monitor_Name_String is an '\0'
ending string, and it allows NML to be 1, which is only valid when MNL
does not count the possible ending '\0'.
Signed-off-by: NWu Fengguang <fengguang.wu@intel.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Move codec init verb which is only applicatable for VT1708.

I've found the root cause that jack plugged in can't be detected.
The verb in vt1708_init_verbs is used to power down jack detect circuit.
This verb is only applicable to VT1708. vt1708 didn't implement jack
detect function in hardware, so we should shut down this function to
avoid noise. But for other codecs, hardware implement jack detect
function. If sending this verb during initialization, jack detect will
be invalid. So I move this verb from via_parse_auto_config() to
patch_vt1708().
Signed-off-by: NLydia Wang <lydiawang@viatech.com.cn>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

sound/pci/hda/patch_cmedia.c: In function ‘cmi9880_fill_multi_init’:
sound/pci/hda/patch_cmedia.c:401:15: warning: unused variable ‘len’
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Let's use auto-parser as default now.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Create snd_hda_get_conn_index() helper function for obtaining the
connection index of the widget.  Replaced the similar codes used in
several codec-drivers with this common helper.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Fix a issue to enable unsolicited response to line-out pins.
Signed-off-by: NLydia Wang <lydiawang@viatech.com.cn>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Since we're not using the new auto parser as a fallback yet,
add it manually as a quirk.
Signed-off-by: NDavid Henningsson <david.henningsson@canonical.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Conexant ID 506c was found on Acer Aspire 3830TG. As users report
no playback, sending to stable should be safe.

Cc: stable@kernel.org
BugLink: https://bugs.launchpad.net/bugs/783582
Reported-by: andROOM
Signed-off-by: NDavid Henningsson <david.henningsson@canonical.com>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Use static inline for dummy function to fix the warnings like below
  sound/pci/hda/patch_sigmatel.c: In function ‘stac92xx_init’:
  sound/pci/hda/patch_sigmatel.c:4387:3: warning: statement with no effect
  sound/pci/hda/patch_sigmatel.c: In function ‘stac92xx_resume’:
  sound/pci/hda/patch_sigmatel.c:4927:3: warning: statement with no effect
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

to fix build problems when it is disabled.
Signed-off-by: NStephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

More similar fixes like previous commits: handle the exceptional case
like ALC267 where no volume amp is found in ADC widget but in the
capsrc widget instead.

Also minor checks for avoiding possible erros: no connection-select
when the pin has a single selection, and add beep verbs only when the
0x1d is used for beep.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

ALC259 seems to provide an invalid widget capability for the input-src
selector widget.  The widget shows the input-amp while it's a selector,
and this confuses the current ALC882 initialization code that is used
for ALC259, too.  For fixing this, check the amp capability and handle
the connection selection individually.

Also, ALC259 has no mute bit in DAC volume, so we need to initialize
it as ZERO instead of MUTE.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

ALC269 and compatible codecs have the output volume in DACs, thus we
can't use the ALC880's code as is.  Fixed by checking the amp caps and
picking up the right widget for initialization.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Look through the known NIDs that may have EAPD capabilities and turn
on/off them appropriately instead of checking the individual vendor ids.

This will also avoid the forgotten entries of newly added codec ids
in future.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>

Fix the regression introduced by the commit
1f0f4b80
  ALSA: hda - Reduce static init verbs for Realtek auto-parsers

The input amps of mixer widgets should be unmuted as default (as
usually they have no assigned mixer switches).

More fixes in this commit are, however, for ALC260: ALC260 codec can
have multiple output mixers connnected to a single DAC althouh the
driver didn't pick up them properly.
Signed-off-by: NTakashi Iwai <tiwai@suse.de>