[BISECTED] 2.6.36 regression for ALC892

[BISECTED] 2.6.36 regression for ALC892

[Andrey Rahmatullin]

[-- Attachment #1: Type: text/plain, Size: 642 bytes --]

Hello. I have an Intel H55 motherboard (ASUS P7H55-M PRO) with ALC892
sound codec and a 3.5" front panel which doesn't support newer HDA pinout.
So the panel is connected in "AC97" mode and this mode is set in the BIOS
setup. In 2.6.35 and earlier the panel's headphones jack was working. Now
it isn't. I've bisected this down to
ce503f38bdb59c9175a9076215a3ba579fad4e64 (which is present in 2.6.36),
reverting it helps. I've got "Too many connections" messages without this
commit, indeed, but I had no problems (though I use only a pair of
speakers connected to the rear panel and headphones in the front panel).

-- 
WBR, wRAR

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [BISECTED] 2.6.36 regression for ALC892

[Takashi Iwai]

At Fri, 26 Nov 2010 02:18:03 +0500,
Andrey Rahmatullin wrote:
> 
> Hello. I have an Intel H55 motherboard (ASUS P7H55-M PRO) with ALC892
> sound codec and a 3.5" front panel which doesn't support newer HDA pinout.
> So the panel is connected in "AC97" mode and this mode is set in the BIOS
> setup. In 2.6.35 and earlier the panel's headphones jack was working. Now
> it isn't. I've bisected this down to
> ce503f38bdb59c9175a9076215a3ba579fad4e64 (which is present in 2.6.36),
> reverting it helps. I've got "Too many connections" messages without this
> commit, indeed, but I had no problems (though I use only a pair of
> speakers connected to the rear panel and headphones in the front panel).

Obviously your front-panel doesn't suit with the HD-audio, and BIOS
still doesn't set the right pin defaults.  That is, it was a bug of
the driver that made your device working :)  Now, the driver does the
right thing.

A possible fix is to add quirks to override the wrong BIOS setup.
Please give alsa-info.sh outputs with and without the commit to figure
out the real pin configuration.


Takashi

Re: [BISECTED] 2.6.36 regression for ALC892

[Andrey Rahmatullin]

[-- Attachment #1.1: Type: text/plain, Size: 279 bytes --]

On Fri, Nov 26, 2010 at 06:59:23AM +0100, Takashi Iwai wrote:
> A possible fix is to add quirks to override the wrong BIOS setup.
> Please give alsa-info.sh outputs with and without the commit to figure
> out the real pin configuration.
They are attached.

-- 
WBR, wRAR

[-- Attachment #1.2: alsa-info.good.bz2 --]
[-- Type: application/octet-stream, Size: 6954 bytes --]

[-- Attachment #1.3: alsa-info.bad.bz2 --]
[-- Type: application/octet-stream, Size: 7592 bytes --]

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [BISECTED] 2.6.36 regression for ALC892

[Takashi Iwai]

At Fri, 26 Nov 2010 14:12:11 +0500,
Andrey Rahmatullin wrote:
> 
> On Fri, Nov 26, 2010 at 06:59:23AM +0100, Takashi Iwai wrote:
> > A possible fix is to add quirks to override the wrong BIOS setup.
> > Please give alsa-info.sh outputs with and without the commit to figure
> > out the real pin configuration.
> They are attached.

Try the very latest alsa-driver snapshot or sound-2.6 git tree.
You should have an individual "Headphone" volume as well.


Takashi

Re: [BISECTED] 2.6.36 regression for ALC892

[Andrey Rahmatullin]

[-- Attachment #1: Type: text/plain, Size: 303 bytes --]

On Fri, Nov 26, 2010 at 10:28:01AM +0100, Takashi Iwai wrote:
> Try the very latest alsa-driver snapshot or sound-2.6 git tree.
> You should have an individual "Headphone" volume as well.
Yes, that makes difference: I was able to use the headphones after raising
that new volume.

-- 
WBR, wRAR

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [BISECTED] 2.6.36 regression for ALC892

[Takashi Iwai]

At Fri, 26 Nov 2010 19:27:37 +0500,
Andrey Rahmatullin wrote:
> 
> On Fri, Nov 26, 2010 at 10:28:01AM +0100, Takashi Iwai wrote:
> > Try the very latest alsa-driver snapshot or sound-2.6 git tree.
> > You should have an individual "Headphone" volume as well.
> Yes, that makes difference: I was able to use the headphones after raising
> that new volume.

OK, then this should be the patch below.  Care to apply it to 2.6.36
and check whether it gives the same result?

If it works, we'll need to push it to stable tree, too.


thanks,

Takashi

===
>From cc1c452e509aefc28f7ad2deed75bc69d4f915f7 Mon Sep 17 00:00:00 2001
From: David Henningsson <david.henningsson@canonical.com>
Date: Wed, 24 Nov 2010 14:17:47 +0100
Subject: [PATCH] ALSA: HDA: Add an extra DAC for Realtek ALC887-VD

The patch enables ALC887-VD to use the DAC at nid 0x26,
which makes it possible to use this DAC for e g Headphone
volume.

Signed-off-by: David Henningsson <david.henningsson@canonical.com>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
---
 sound/pci/hda/patch_realtek.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)

diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c
index 8f7530f..b0e6b8b 100644
--- a/sound/pci/hda/patch_realtek.c
+++ b/sound/pci/hda/patch_realtek.c
@@ -18997,6 +18997,8 @@ static inline hda_nid_t alc662_mix_to_dac(hda_nid_t nid)
 		return 0x02;
 	else if (nid >= 0x0c && nid <= 0x0e)
 		return nid - 0x0c + 0x02;
+	else if (nid == 0x26) /* ALC887-VD has this DAC too */
+		return 0x25;
 	else
 		return 0;
 }
@@ -19005,7 +19007,7 @@ static inline hda_nid_t alc662_mix_to_dac(hda_nid_t nid)
 static hda_nid_t alc662_dac_to_mix(struct hda_codec *codec, hda_nid_t pin,
 				   hda_nid_t dac)
 {
-	hda_nid_t mix[4];
+	hda_nid_t mix[5];
 	int i, num;
 
 	num = snd_hda_get_connections(codec, pin, mix, ARRAY_SIZE(mix));
-- 
1.7.3.1

Re: [BISECTED] 2.6.36 regression for ALC892

[Andrey Rahmatullin]

[-- Attachment #1: Type: text/plain, Size: 509 bytes --]

On Fri, Nov 26, 2010 at 03:34:03PM +0100, Takashi Iwai wrote:
> > > Try the very latest alsa-driver snapshot or sound-2.6 git tree.
> > > You should have an individual "Headphone" volume as well.
> > Yes, that makes difference: I was able to use the headphones after raising
> > that new volume.
> OK, then this should be the patch below.  Care to apply it to 2.6.36
> and check whether it gives the same result?
I've applied it to 2.6.36 and got a "Headphone" volume, and it works.

-- 
WBR, wRAR

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 490 bytes --]