hda driver, correct bug in model "auto"

PATCH In review Nicolas GRAZIANO 2 participants 6 messages Jul 19, 2005 – Jul 29, 2005

Share & tools

Nicolas GRAZIANO

author

2005-07-19 21:34 UTC #0 lore raw

PATCH 1/1 hda driver, correct bug in model "auto"

- Correct some index variable inversion in patch_cmedia.c
- Correct the channels order in model autodetection (not sure of this part, I need validation)

Signed-off-by Nicolas Graziano <[email protected]>

Index: pci/hda/hda_codec.c
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/pci/hda/hda_codec.c,v retrieving revision 1.20

hda/hda_codec.c +9 -1

diff -u -r1.20 hda_codec.c
@@ -1804,12 +1804,20 @@
 			}
 
 	/* Swap surround and CLFE: the association order is front/CLFE/surr/back */
-	if (cfg->line_outs >= 3) {
+	if (cfg->line_outs == 3) {
 		nid = cfg->line_out_pins[1];
 		cfg->line_out_pins[1] = cfg->line_out_pins[2];
 		cfg->line_out_pins[2] = nid;
 	}
 
+	if (cfg->line_outs >= 4) {
+		nid = cfg->line_out_pins[1];
+		cfg->line_out_pins[1] = cfg->line_out_pins[3];
+		cfg->line_out_pins[3] = cfg->line_out_pins[2];
+		cfg->line_out_pins[2] = nid;
+	}
+
+
 	return 0;
 }

Index: pci/hda/patch_cmedia.c
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/pci/hda/patch_cmedia.c,v retrieving revision 1.10

hda/patch_cmedia.c +2 -3

diff -u -r1.10 patch_cmedia.c
@@ -408,7 +408,7 @@
 		/* search for an empty channel */
 		for (j = 0; j < cfg->line_outs; j++) {
 			if (! assigned[j]) {
-				spec->dac_nids[i] = i + 0x03;
+				spec->dac_nids[i] = j + 0x03;
 				assigned[j] = 1;
 				break;
 			}

@@ -444,11 +444,10 @@
 			len = snd_hda_get_connections(codec, nid, conn, 4);
 			for (k = 0; k < len; k++)
 				if (conn[k] == spec->dac_nids[i]) {
-					spec->multi_init[j].param = j;
+					spec->multi_init[j].param = k;
 					break;
 				}
 			j++;
-			break;
 		}
 	}
 	return 0;

Takashi Iwai

2005-07-27 09:37 UTC #1 lore raw

At Tue, 19 Jul 2005 23:34:20 +0200,

Quoted 3 lines · Nicolas GRAZIANO

- Correct the channels order in model autodetection (not sure of this part, I need validation)

I think this is not correct (although I'm not 100% sure, too). The association order is front/clfe/surround/side. (The comment in hda_codec.c is wrong.)

I'll apply the fix for c-media now.

Thanks,

Takashi

-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click

Nicolas GRAZIANO

author

2005-07-27 18:32 UTC #2 lore raw

Quoted 4 lines · Takashi Iwai

At Tue, 19 Jul 2005 23:34:20 +0200,
Nicolas GRAZIANO wrote:

Quoted 11 lines

Quoted 4 lines
- Correct the channels order in model autodetection (not sure of this part, I need validation)
I think this is not correct (although I'm not 100% sure, too). The association order is front/clfe/surround/side. (The comment in hda_codec.c is wrong.)

I make the modification because I was thinking than the order was front/clfe/side/back (aka surround ?) and the text on back panel of my motherboard (ASUS P5AD2) make me more confident (with my modification the channels are at the right place). Can we check with an other card/codec ?

an another question : the model full_dig correspond to the position on back panel of my motherboard, can we make the selection of model automatic ?

lspci -vn for my card :

0000:00:1b.0 0403: 8086:2668 (rev 03)

        Subsystem: 1043:813d

thanks in advance,
Nicolas

Takashi Iwai

2005-07-28 09:07 UTC #3 lore raw

At Wed, 27 Jul 2005 20:32:24 +0200,

Quoted 18 lines · Nicolas GRAZIANO

Quoted 4 lines · Takashi Iwai
At Tue, 19 Jul 2005 23:34:20 +0200,
Nicolas GRAZIANO wrote:
Quoted 11 lines
Quoted 4 lines
- Correct the channels order in model autodetection (not sure of this part, I need validation)
I think this is not correct (although I'm not 100% sure, too). The association order is front/clfe/surround/side. (The comment in hda_codec.c is wrong.)

Quoted 5 lines

I make the modification because I was thinking than the order was front/clfe/side/back (aka surround ?) and the text on back panel of my motherboard (ASUS P5AD2) make me more confident (with my modification the channels are at the right place). Can we check with an other card/codec ?

Hmm, interesting. So, on your board, the association 3 (4th) corresponds to a surround jack?

Quoted 2 lines

an another question : the model full_dig correspond to the position on back panel of my motherboard, can we make the selection of model automatic ?

Sure, just add a hda_board_config entry for the corresponding pci subids. See patch_realtek.c.

Takashi

-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf

Nicolas GRAZIANO

author

2005-07-28 18:32 UTC #4 lore raw

Quoted 31 lines · Takashi Iwai

At Wed, 27 Jul 2005 20:32:24 +0200,
Quoted 19 lines · Nicolas GRAZIANO
Quoted 10 lines · Takashi Iwai
At Tue, 19 Jul 2005 23:34:20 +0200,
Quoted 6 lines · Nicolas GRAZIANO
- Correct the channels order in model autodetection (not sure of this part, I need validation)
Quoted 6 lines
I think this is not correct (although I'm not 100% sure, too). The association order is front/clfe/surround/side. (The comment in hda_codec.c is wrong.)
Quoted 7 lines
I make the modification because I was thinking than the order was front/clfe/side/back (aka surround ?) and the text on back panel of my motherboard (ASUS P5AD2) make me more confident (with my modification the channels are at the right place). Can we check with an other card/codec ?

Quoted 5 lines

Hmm, interesting. So, on your board, the association 3 (4th) corresponds to a surround jack?

I am not sure to understand what you mean.
But perhaps the following information will help.
On the back pannel the inscription are :
Green ()-> Front
Orange ()-> Ctr Bass
Black ()-> Side SPK
Grey ()-> Rear SPK

Extract of the /proc/asound/Intel/codec#0 order by the sequence number (last byte of Pincap):

Node 0x0b [Pin Complex] wcaps 0x400581: Stereo
Pincap 0x0833f: IN OUT HP
Pin Default 0x01014010: [Jack] Line Out at Ext Rear
Conn = 1/8, Color = Green

Node 0x1f [Pin Complex] wcaps 0x400181: Stereo
Pincap 0x0833f: IN OUT HP
Pin Default 0x01016012: [Jack] Line Out at Ext Rear
Conn = 1/8, Color = Orange

Node 0x0e [Pin Complex] wcaps 0x400581: Stereo
Pincap 0x081033f: IN OUT HP
Pin Default 0x01011014: [Jack] Line Out at Ext Rear
Conn = 1/8, Color = Black

Node 0x20 [Pin Complex] wcaps 0x400181: Stereo
Pincap 0x0833f: IN OUT HP
Pin Default 0x01012016: [Jack] Line Out at Ext Rear
Conn = 1/8, Color = Grey

with my modification :

$speaker-test -D hw:0,0 -c8
0 - Front Left --> Green Jack
4 - Center --> Orange left
1 - Front Right --> Green Jack
7 - Side Right --> Black Jack
3 - Rear Right --> Grey Jack
2 - Rear Left --> Grey Jack
6 - Side Left --> Black Jack
5 - LFE --> nothing

without my modification (actual cvs code):

$speaker-test -D hw:0,0 -c8
0 - Front Left --> Green Jack
4 - Center --> black Jack
1 - Front Right --> Grenn Jack
7 - Side Right --> nothing
3 - Rear Right --> orange Jack
2 - Rear Left --> orange Jack
6 - Side Left --> Grey Jack
5 - LFE --> black Jack

I hope this information will help.

Quoted 5 lines

Sure, just add a hda_board_config entry for the corresponding pci subids. See patch_realtek.c.

I try the following path for my ASUS P5AD2 and it's working. thanks

Nicolas

Index: pci/hda/patch_cmedia.c
===================================================================
RCS file: /cvsroot/alsa/alsa-kernel/pci/hda/patch_cmedia.c,v retrieving revision 1.11

hda/patch_cmedia.c 28 Jul 2005 18:22:45 -0000 renamed from hda/patch_cmedia.c 27 Jul 2005 16:25:08 -0000 1.11 +1 -0

diff -u -r1.11 patch_cmedia.c
@@ -647,6 +647,8 @@
        { .modelname = "min_fp", .config = CMI_MIN_FP },
        { .modelname = "full", .config = CMI_FULL },
        { .modelname = "full_dig", .config = CMI_FULL_DIG },
+       { .pci_subvendor = 0x1043, .pci_subdevice = 0x813d, .config =

CMI_FULL_DIG },
+
{ .modelname = "allout", .config = CMI_ALLOUT }, { .modelname = "auto", .config = CMI_AUTO }, {} /* terminator */

Takashi Iwai

2005-07-29 10:49 UTC #5 lore raw

At Thu, 28 Jul 2005 20:32:41 +0200,

Quoted 47 lines · Nicolas GRAZIANO

Quoted 31 lines · Takashi Iwai
At Wed, 27 Jul 2005 20:32:24 +0200,
Quoted 19 lines · Nicolas GRAZIANO
Quoted 10 lines · Takashi Iwai
At Tue, 19 Jul 2005 23:34:20 +0200,
Quoted 6 lines · Nicolas GRAZIANO
- Correct the channels order in model autodetection (not sure of this part, I need validation)
Quoted 6 lines
I think this is not correct (although I'm not 100% sure, too). The association order is front/clfe/surround/side. (The comment in hda_codec.c is wrong.)
Quoted 7 lines
I make the modification because I was thinking than the order was front/clfe/side/back (aka surround ?) and the text on back panel of my motherboard (ASUS P5AD2) make me more confident (with my modification the channels are at the right place). Can we check with an other card/codec ?
Quoted 5 lines
Hmm, interesting. So, on your board, the association 3 (4th) corresponds to a surround jack?
I am not sure to understand what you mean.
But perhaps the following information will help.
On the back pannel the inscription are :
Green ()-> Front
Orange ()-> Ctr Bass
Black ()-> Side SPK
Grey ()-> Rear SPK

Quoted 2 lines

Extract of the /proc/asound/Intel/codec#0 order by the sequence number (last byte of Pincap):

OK, then you seem right. It's front/clfe/side/rear. OK, let's fix that.

Quoted 6 lines

Quoted 5 lines
Sure, just add a hda_board_config entry for the corresponding pci subids. See patch_realtek.c.
I try the following path for my ASUS P5AD2 and it's working.

Will apply this, too.

thanks,

Takashi