[regression] probably in mixer.c code

[regression] probably in mixer.c code

[Alexey Fisher]

Hallo all, Takashi,

there is one regression in current linux main tree:

cval->res quirks do not work any more. See usb/mixer.c:1150
Are there was some work in this area or may be some one bisected this 
bug down?

Regards,
Alexey.

Re: [regression] probably in mixer.c code

[Takashi Iwai]

At Tue, 08 Nov 2011 09:49:13 +0100,
Alexey Fisher wrote:
> 
> Hallo all, Takashi,
> 
> there is one regression in current linux main tree:
> 
> cval->res quirks do not work any more. See usb/mixer.c:1150
> Are there was some work in this area or may be some one bisected this 
> bug down?

Are you referring to the device 046d:09a4, which sets the mixer range
manually?  If yes, does the patch below work?


Takashi

---
diff --git a/sound/usb/mixer.c b/sound/usb/mixer.c
index 60f65ac..92feff1 100644
--- a/sound/usb/mixer.c
+++ b/sound/usb/mixer.c
@@ -765,6 +765,28 @@ static void usb_mixer_elem_free(struct snd_kcontrol *kctl)
  * interface to ALSA control for feature/mixer units
  */
 
+/* convert from USB descriptions to dB scale
+ * USB descriptions contain the dB scale in 1/256 dB unit
+ * while ALSA TLV contains in 1/100 dB unit
+ */
+static int convert_cval_to_dB(struct usb_mixer_elem_info *cval)
+{
+	cval->dBmin = (convert_signed_value(cval, cval->min) * 100) / 256;
+	cval->dBmax = (convert_signed_value(cval, cval->max) * 100) / 256;
+	if (cval->dBmin > cval->dBmax) {
+		/* something is wrong; assume it's either from/to 0dB */
+		if (cval->dBmin < 0)
+			cval->dBmax = 0;
+		else if (cval->dBmin > 0)
+			cval->dBmin = 0;
+		if (cval->dBmin > cval->dBmax) {
+			/* totally crap, return an error */
+			return -EINVAL;
+		}
+	}
+	return 0;
+}
+
 /*
  * retrieve the minimum and maximum values for the specified control
  */
@@ -844,24 +866,7 @@ static int get_min_max(struct usb_mixer_elem_info *cval, int default_min)
 		cval->initialized = 1;
 	}
 
-	/* USB descriptions contain the dB scale in 1/256 dB unit
-	 * while ALSA TLV contains in 1/100 dB unit
-	 */
-	cval->dBmin = (convert_signed_value(cval, cval->min) * 100) / 256;
-	cval->dBmax = (convert_signed_value(cval, cval->max) * 100) / 256;
-	if (cval->dBmin > cval->dBmax) {
-		/* something is wrong; assume it's either from/to 0dB */
-		if (cval->dBmin < 0)
-			cval->dBmax = 0;
-		else if (cval->dBmin > 0)
-			cval->dBmin = 0;
-		if (cval->dBmin > cval->dBmax) {
-			/* totally crap, return an error */
-			return -EINVAL;
-		}
-	}
-
-	return 0;
+	return convert_cval_to_dB(cval);
 }
 
 
@@ -1134,6 +1139,7 @@ static void build_feature_ctl(struct mixer_build *state, void *raw_desc,
 			snd_printk(KERN_INFO
 				 "set volume quirk for UDA1321/N101 chip\n");
 			cval->max = -256;
+			convert_cval_to_dB(cval);
 		}
 		break;
 
@@ -1144,6 +1150,8 @@ static void build_feature_ctl(struct mixer_build *state, void *raw_desc,
 			cval->min = 6080;
 			cval->max = 8768;
 			cval->res = 192;
+			cval->initialized = 1;
+			convert_cval_to_dB(cval);
 		}
 		break;
 

Re: [regression] probably in mixer.c code

[Alexey Fisher]

Am 08.11.2011 10:59, schrieb Takashi Iwai:
> At Tue, 08 Nov 2011 09:49:13 +0100,
> Alexey Fisher wrote:
>>
>> Hallo all, Takashi,
>>
>> there is one regression in current linux main tree:
>>
>> cval->res quirks do not work any more. See usb/mixer.c:1150
>> Are there was some work in this area or may be some one bisected this
>> bug down?
>
> Are you referring to the device 046d:09a4, which sets the mixer range
> manually?  If yes, does the patch below work?
>


No, i referring to devices 046d:0991, 046d:081d, i don't have access to 
046d:09a4.

This part is broken for me:
         case USB_ID(0x046d, 0x0991):
         /* Most audio usb devices lie about volume resolution.
          * Most Logitech webcams have res = 384.
          * Proboly there is some logitech magic behind this number --fishor
          */
                 if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
                         snd_printk(KERN_INFO
                                 "set resolution quirk: cval->res = 384\n");
                         cval->res = 384;
                 }
                 break;

Re: [regression] probably in mixer.c code

[Takashi Iwai]

At Tue, 08 Nov 2011 12:25:06 +0100,
Alexey Fisher wrote:
> 
> Am 08.11.2011 10:59, schrieb Takashi Iwai:
> > At Tue, 08 Nov 2011 09:49:13 +0100,
> > Alexey Fisher wrote:
> >>
> >> Hallo all, Takashi,
> >>
> >> there is one regression in current linux main tree:
> >>
> >> cval->res quirks do not work any more. See usb/mixer.c:1150
> >> Are there was some work in this area or may be some one bisected this
> >> bug down?
> >
> > Are you referring to the device 046d:09a4, which sets the mixer range
> > manually?  If yes, does the patch below work?
> >
> 
> 
> No, i referring to devices 046d:0991, 046d:081d, i don't have access to 
> 046d:09a4.
> 
> This part is broken for me:
>          case USB_ID(0x046d, 0x0991):
>          /* Most audio usb devices lie about volume resolution.
>           * Most Logitech webcams have res = 384.
>           * Proboly there is some logitech magic behind this number --fishor
>           */
>                  if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
>                          snd_printk(KERN_INFO
>                                  "set resolution quirk: cval->res = 384\n");
>                          cval->res = 384;
>                  }
>                  break;

OK, so _what_ is actually broken?  Do you see the message but
cval->res is overridden later?


Takashi

Re: [regression] probably in mixer.c code

[Alexey Fisher]

Am 08.11.2011 13:11, schrieb Takashi Iwai:
> At Tue, 08 Nov 2011 12:25:06 +0100,
> Alexey Fisher wrote:
>>
>> Am 08.11.2011 10:59, schrieb Takashi Iwai:
>>> At Tue, 08 Nov 2011 09:49:13 +0100,
>>> Alexey Fisher wrote:
>>>>
>>>> Hallo all, Takashi,
>>>>
>>>> there is one regression in current linux main tree:
>>>>
>>>> cval->res quirks do not work any more. See usb/mixer.c:1150
>>>> Are there was some work in this area or may be some one bisected this
>>>> bug down?
>>>
>>> Are you referring to the device 046d:09a4, which sets the mixer range
>>> manually?  If yes, does the patch below work?
>>>
>>
>>
>> No, i referring to devices 046d:0991, 046d:081d, i don't have access to
>> 046d:09a4.
>>
>> This part is broken for me:
>>           case USB_ID(0x046d, 0x0991):
>>           /* Most audio usb devices lie about volume resolution.
>>            * Most Logitech webcams have res = 384.
>>            * Proboly there is some logitech magic behind this number --fishor
>>            */
>>                   if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
>>                           snd_printk(KERN_INFO
>>                                   "set resolution quirk: cval->res = 384\n");
>>                           cval->res = 384;
>>                   }
>>                   break;
>
> OK, so _what_ is actually broken?  Do you see the message but
> cval->res is overridden later?

yes.

Re: [regression] probably in mixer.c code

[Takashi Iwai]

At Tue, 08 Nov 2011 13:39:38 +0100,
Alexey Fisher wrote:
> 
> Am 08.11.2011 13:11, schrieb Takashi Iwai:
> > At Tue, 08 Nov 2011 12:25:06 +0100,
> > Alexey Fisher wrote:
> >>
> >> Am 08.11.2011 10:59, schrieb Takashi Iwai:
> >>> At Tue, 08 Nov 2011 09:49:13 +0100,
> >>> Alexey Fisher wrote:
> >>>>
> >>>> Hallo all, Takashi,
> >>>>
> >>>> there is one regression in current linux main tree:
> >>>>
> >>>> cval->res quirks do not work any more. See usb/mixer.c:1150
> >>>> Are there was some work in this area or may be some one bisected this
> >>>> bug down?
> >>>
> >>> Are you referring to the device 046d:09a4, which sets the mixer range
> >>> manually?  If yes, does the patch below work?
> >>>
> >>
> >>
> >> No, i referring to devices 046d:0991, 046d:081d, i don't have access to
> >> 046d:09a4.
> >>
> >> This part is broken for me:
> >>           case USB_ID(0x046d, 0x0991):
> >>           /* Most audio usb devices lie about volume resolution.
> >>            * Most Logitech webcams have res = 384.
> >>            * Proboly there is some logitech magic behind this number --fishor
> >>            */
> >>                   if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
> >>                           snd_printk(KERN_INFO
> >>                                   "set resolution quirk: cval->res = 384\n");
> >>                           cval->res = 384;
> >>                   }
> >>                   break;
> >
> > OK, so _what_ is actually broken?  Do you see the message but
> > cval->res is overridden later?
> 
> yes.

To which value?


Takashi

Re: [regression] probably in mixer.c code

[Takashi Iwai]

At Tue, 08 Nov 2011 14:05:32 +0100,
Takashi Iwai wrote:
> 
> At Tue, 08 Nov 2011 13:39:38 +0100,
> Alexey Fisher wrote:
> > 
> > Am 08.11.2011 13:11, schrieb Takashi Iwai:
> > > At Tue, 08 Nov 2011 12:25:06 +0100,
> > > Alexey Fisher wrote:
> > >>
> > >> Am 08.11.2011 10:59, schrieb Takashi Iwai:
> > >>> At Tue, 08 Nov 2011 09:49:13 +0100,
> > >>> Alexey Fisher wrote:
> > >>>>
> > >>>> Hallo all, Takashi,
> > >>>>
> > >>>> there is one regression in current linux main tree:
> > >>>>
> > >>>> cval->res quirks do not work any more. See usb/mixer.c:1150
> > >>>> Are there was some work in this area or may be some one bisected this
> > >>>> bug down?
> > >>>
> > >>> Are you referring to the device 046d:09a4, which sets the mixer range
> > >>> manually?  If yes, does the patch below work?
> > >>>
> > >>
> > >>
> > >> No, i referring to devices 046d:0991, 046d:081d, i don't have access to
> > >> 046d:09a4.
> > >>
> > >> This part is broken for me:
> > >>           case USB_ID(0x046d, 0x0991):
> > >>           /* Most audio usb devices lie about volume resolution.
> > >>            * Most Logitech webcams have res = 384.
> > >>            * Proboly there is some logitech magic behind this number --fishor
> > >>            */
> > >>                   if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
> > >>                           snd_printk(KERN_INFO
> > >>                                   "set resolution quirk: cval->res = 384\n");
> > >>                           cval->res = 384;
> > >>                   }
> > >>                   break;
> > >
> > > OK, so _what_ is actually broken?  Do you see the message but
> > > cval->res is overridden later?
> > 
> > yes.
> 
> To which value?

Also you should have got a message like "0:1: cannot get min/max
values for ..." when built with CONFIG_SND_DEBUG.  Check whether you
get it actually.


Takashi

Re: [regression] probably in mixer.c code

[Alexey Fisher]

Am 08.11.2011 14:09, schrieb Takashi Iwai:
> At Tue, 08 Nov 2011 14:05:32 +0100,
> Takashi Iwai wrote:
>>
>> At Tue, 08 Nov 2011 13:39:38 +0100,
>> Alexey Fisher wrote:
>>>
>>> Am 08.11.2011 13:11, schrieb Takashi Iwai:
>>>> At Tue, 08 Nov 2011 12:25:06 +0100,
>>>> Alexey Fisher wrote:
>>>>>
>>>>> Am 08.11.2011 10:59, schrieb Takashi Iwai:
>>>>>> At Tue, 08 Nov 2011 09:49:13 +0100,
>>>>>> Alexey Fisher wrote:
>>>>>>>
>>>>>>> Hallo all, Takashi,
>>>>>>>
>>>>>>> there is one regression in current linux main tree:
>>>>>>>
>>>>>>> cval->res quirks do not work any more. See usb/mixer.c:1150
>>>>>>> Are there was some work in this area or may be some one bisected this
>>>>>>> bug down?
>>>>>>
>>>>>> Are you referring to the device 046d:09a4, which sets the mixer range
>>>>>> manually?  If yes, does the patch below work?
>>>>>>
>>>>>
>>>>>
>>>>> No, i referring to devices 046d:0991, 046d:081d, i don't have access to
>>>>> 046d:09a4.
>>>>>
>>>>> This part is broken for me:
>>>>>            case USB_ID(0x046d, 0x0991):
>>>>>            /* Most audio usb devices lie about volume resolution.
>>>>>             * Most Logitech webcams have res = 384.
>>>>>             * Proboly there is some logitech magic behind this number --fishor
>>>>>             */
>>>>>                    if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
>>>>>                            snd_printk(KERN_INFO
>>>>>                                    "set resolution quirk: cval->res = 384\n");
>>>>>                            cval->res = 384;
>>>>>                    }
>>>>>                    break;
>>>>
>>>> OK, so _what_ is actually broken?  Do you see the message but
>>>> cval->res is overridden later?
>>>
>>> yes.
>>
>> To which value?
>
> Also you should have got a message like "0:1: cannot get min/max
> values for ..." when built with CONFIG_SND_DEBUG.  Check whether you
> get it actually.

It sets the res value to 1.
I get fallowing message after i plug in the webcam:

[  132.878150] usb 1-3.2: new high-speed USB device number 8 using ehci_hcd
[  133.087255] usb 1-3.2: New USB device found, idVendor=046d, 
idProduct=0991
[  133.087259] usb 1-3.2: New USB device strings: Mfr=0, Product=0, 
SerialNumber=2
[  133.087261] usb 1-3.2: SerialNumber: 9671DCEE
[  133.088428] uvcvideo: Found UVC 1.00 device <unnamed> (046d:0991)
[  133.122400] input: UVC Camera (046d:0991) as 
/devices/pci0000:00/0000:00:1a.7/usb1/1-3/1-3.2/1-3.2:1.0/input/input15
[  133.488506] ALSA mixer.c:817 5:2: cannot get min/max values for 
control 2 (id 5)
[  133.488524] ALSA mixer.c:1168 set resolution quirk: cval->res = 384
[  133.488527] ALSA mixer.c:817 5:2: cannot get min/max values for 
control 2 (id 5)
[  133.492187] ALSA mixer.c:817 5:2: cannot get min/max values for 
control 2 (id 5)


With amixer i get this:
amixer -c U0x46d0x991
Simple mixer control 'Mic',0
   Capabilities: cvolume cvolume-joined cswitch cswitch-joined penum
   Capture channels: Mono
   Limits: Capture 0 - 3072
   Mono: Capture 2749 [89%] [28.73dB] [on]


But Limits should be 0-7 instead of 0-3072

Re: [regression] probably in mixer.c code

[Takashi Iwai]

At Tue, 08 Nov 2011 15:32:10 +0100,
Alexey Fisher wrote:
> 
> Am 08.11.2011 14:09, schrieb Takashi Iwai:
> > At Tue, 08 Nov 2011 14:05:32 +0100,
> > Takashi Iwai wrote:
> >>
> >> At Tue, 08 Nov 2011 13:39:38 +0100,
> >> Alexey Fisher wrote:
> >>>
> >>> Am 08.11.2011 13:11, schrieb Takashi Iwai:
> >>>> At Tue, 08 Nov 2011 12:25:06 +0100,
> >>>> Alexey Fisher wrote:
> >>>>>
> >>>>> Am 08.11.2011 10:59, schrieb Takashi Iwai:
> >>>>>> At Tue, 08 Nov 2011 09:49:13 +0100,
> >>>>>> Alexey Fisher wrote:
> >>>>>>>
> >>>>>>> Hallo all, Takashi,
> >>>>>>>
> >>>>>>> there is one regression in current linux main tree:
> >>>>>>>
> >>>>>>> cval->res quirks do not work any more. See usb/mixer.c:1150
> >>>>>>> Are there was some work in this area or may be some one bisected this
> >>>>>>> bug down?
> >>>>>>
> >>>>>> Are you referring to the device 046d:09a4, which sets the mixer range
> >>>>>> manually?  If yes, does the patch below work?
> >>>>>>
> >>>>>
> >>>>>
> >>>>> No, i referring to devices 046d:0991, 046d:081d, i don't have access to
> >>>>> 046d:09a4.
> >>>>>
> >>>>> This part is broken for me:
> >>>>>            case USB_ID(0x046d, 0x0991):
> >>>>>            /* Most audio usb devices lie about volume resolution.
> >>>>>             * Most Logitech webcams have res = 384.
> >>>>>             * Proboly there is some logitech magic behind this number --fishor
> >>>>>             */
> >>>>>                    if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
> >>>>>                            snd_printk(KERN_INFO
> >>>>>                                    "set resolution quirk: cval->res = 384\n");
> >>>>>                            cval->res = 384;
> >>>>>                    }
> >>>>>                    break;
> >>>>
> >>>> OK, so _what_ is actually broken?  Do you see the message but
> >>>> cval->res is overridden later?
> >>>
> >>> yes.
> >>
> >> To which value?
> >
> > Also you should have got a message like "0:1: cannot get min/max
> > values for ..." when built with CONFIG_SND_DEBUG.  Check whether you
> > get it actually.
> 
> It sets the res value to 1.
> I get fallowing message after i plug in the webcam:
> 
> [  132.878150] usb 1-3.2: new high-speed USB device number 8 using ehci_hcd
> [  133.087255] usb 1-3.2: New USB device found, idVendor=046d, 
> idProduct=0991
> [  133.087259] usb 1-3.2: New USB device strings: Mfr=0, Product=0, 
> SerialNumber=2
> [  133.087261] usb 1-3.2: SerialNumber: 9671DCEE
> [  133.088428] uvcvideo: Found UVC 1.00 device <unnamed> (046d:0991)
> [  133.122400] input: UVC Camera (046d:0991) as 
> /devices/pci0000:00/0000:00:1a.7/usb1/1-3/1-3.2/1-3.2:1.0/input/input15
> [  133.488506] ALSA mixer.c:817 5:2: cannot get min/max values for 
> control 2 (id 5)
> [  133.488524] ALSA mixer.c:1168 set resolution quirk: cval->res = 384
> [  133.488527] ALSA mixer.c:817 5:2: cannot get min/max values for 
> control 2 (id 5)
> [  133.492187] ALSA mixer.c:817 5:2: cannot get min/max values for 
> control 2 (id 5)
> 
> 
> With amixer i get this:
> amixer -c U0x46d0x991
> Simple mixer control 'Mic',0
>    Capabilities: cvolume cvolume-joined cswitch cswitch-joined penum
>    Capture channels: Mono
>    Limits: Capture 0 - 3072
>    Mono: Capture 2749 [89%] [28.73dB] [on]
> 
> 
> But Limits should be 0-7 instead of 0-3072

OK, then the problem is that the first few reads of min/max values
failed, but cval->res quirk is applied only for the first read.

Does the patch below fix?


Takashi

---
diff --git a/sound/usb/mixer.c b/sound/usb/mixer.c
index 60f65ac..c5444e0 100644
--- a/sound/usb/mixer.c
+++ b/sound/usb/mixer.c
@@ -765,10 +765,60 @@ static void usb_mixer_elem_free(struct snd_kcontrol *kctl)
  * interface to ALSA control for feature/mixer units
  */
 
+/* volume control quirks */
+static void volume_control_quirks(struct usb_mixer_elem_info *cval,
+				  struct snd_kcontrol *kctl)
+{
+	switch (cval->mixer->chip->usb_id) {
+	case USB_ID(0x0471, 0x0101):
+	case USB_ID(0x0471, 0x0104):
+	case USB_ID(0x0471, 0x0105):
+	case USB_ID(0x0672, 0x1041):
+	/* quirk for UDA1321/N101.
+	 * note that detection between firmware 2.1.1.7 (N101)
+	 * and later 2.1.1.21 is not very clear from datasheets.
+	 * I hope that the min value is -15360 for newer firmware --jk
+	 */
+		if (!strcmp(kctl->id.name, "PCM Playback Volume") &&
+		    cval->min == -15616) {
+			snd_printk(KERN_INFO
+				 "set volume quirk for UDA1321/N101 chip\n");
+			cval->max = -256;
+		}
+		break;
+
+	case USB_ID(0x046d, 0x09a4):
+		if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
+			snd_printk(KERN_INFO
+				"set volume quirk for QuickCam E3500\n");
+			cval->min = 6080;
+			cval->max = 8768;
+			cval->res = 192;
+		}
+		break;
+
+	case USB_ID(0x046d, 0x0808):
+	case USB_ID(0x046d, 0x0809):
+	case USB_ID(0x046d, 0x0991):
+	/* Most audio usb devices lie about volume resolution.
+	 * Most Logitech webcams have res = 384.
+	 * Proboly there is some logitech magic behind this number --fishor
+	 */
+		if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
+			snd_printk(KERN_INFO
+				"set resolution quirk: cval->res = 384\n");
+			cval->res = 384;
+		}
+		break;
+
+	}
+}
+
 /*
  * retrieve the minimum and maximum values for the specified control
  */
-static int get_min_max(struct usb_mixer_elem_info *cval, int default_min)
+static int get_min_max_with_quirks(struct usb_mixer_elem_info *cval,
+				   int default_min, struct snd_kcontrol *kctl)
 {
 	/* for failsafe */
 	cval->min = default_min;
@@ -844,6 +894,9 @@ static int get_min_max(struct usb_mixer_elem_info *cval, int default_min)
 		cval->initialized = 1;
 	}
 
+	if (kctl)
+		volume_control_quirks(cval, kctl);
+
 	/* USB descriptions contain the dB scale in 1/256 dB unit
 	 * while ALSA TLV contains in 1/100 dB unit
 	 */
@@ -864,6 +917,7 @@ static int get_min_max(struct usb_mixer_elem_info *cval, int default_min)
 	return 0;
 }
 
+#define get_min_max(cval, def)	get_min_max_with_quirks(cval, def, NULL)
 
 /* get a feature/mixer unit info */
 static int mixer_ctl_feature_info(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_info *uinfo)
@@ -882,7 +936,7 @@ static int mixer_ctl_feature_info(struct snd_kcontrol *kcontrol, struct snd_ctl_
 		uinfo->value.integer.max = 1;
 	} else {
 		if (!cval->initialized) {
-			get_min_max(cval, 0);
+			get_min_max_with_quirks(cval, 0, kcontrol);
 			if (cval->initialized && cval->dBmin >= cval->dBmax) {
 				kcontrol->vd[0].access &= 
 					~(SNDRV_CTL_ELEM_ACCESS_TLV_READ |
@@ -1045,9 +1099,6 @@ static void build_feature_ctl(struct mixer_build *state, void *raw_desc,
 		cval->ch_readonly = readonly_mask;
 	}
 
-	/* get min/max values */
-	get_min_max(cval, 0);
-
 	/* if all channels in the mask are marked read-only, make the control
 	 * read-only. set_cur_mix_value() will check the mask again and won't
 	 * issue write commands to read-only channels. */
@@ -1069,6 +1120,9 @@ static void build_feature_ctl(struct mixer_build *state, void *raw_desc,
 		len = snd_usb_copy_string_desc(state, nameid,
 				kctl->id.name, sizeof(kctl->id.name));
 
+	/* get min/max values */
+	get_min_max_with_quirks(cval, 0, kctl);
+
 	switch (control) {
 	case UAC_FU_MUTE:
 	case UAC_FU_VOLUME:
@@ -1118,51 +1172,6 @@ static void build_feature_ctl(struct mixer_build *state, void *raw_desc,
 		break;
 	}
 
-	/* volume control quirks */
-	switch (state->chip->usb_id) {
-	case USB_ID(0x0471, 0x0101):
-	case USB_ID(0x0471, 0x0104):
-	case USB_ID(0x0471, 0x0105):
-	case USB_ID(0x0672, 0x1041):
-	/* quirk for UDA1321/N101.
-	 * note that detection between firmware 2.1.1.7 (N101)
-	 * and later 2.1.1.21 is not very clear from datasheets.
-	 * I hope that the min value is -15360 for newer firmware --jk
-	 */
-		if (!strcmp(kctl->id.name, "PCM Playback Volume") &&
-		    cval->min == -15616) {
-			snd_printk(KERN_INFO
-				 "set volume quirk for UDA1321/N101 chip\n");
-			cval->max = -256;
-		}
-		break;
-
-	case USB_ID(0x046d, 0x09a4):
-		if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
-			snd_printk(KERN_INFO
-				"set volume quirk for QuickCam E3500\n");
-			cval->min = 6080;
-			cval->max = 8768;
-			cval->res = 192;
-		}
-		break;
-
-	case USB_ID(0x046d, 0x0808):
-	case USB_ID(0x046d, 0x0809):
-	case USB_ID(0x046d, 0x0991):
-	/* Most audio usb devices lie about volume resolution.
-	 * Most Logitech webcams have res = 384.
-	 * Proboly there is some logitech magic behind this number --fishor
-	 */
-		if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
-			snd_printk(KERN_INFO
-				"set resolution quirk: cval->res = 384\n");
-			cval->res = 384;
-		}
-		break;
-
-	}
-
 	range = (cval->max - cval->min) / cval->res;
 	/* Are there devices with volume range more than 255? I use a bit more
 	 * to be sure. 384 is a resolution magic number found on Logitech

Re: [regression] probably in mixer.c code

[Alexey Fisher]

Am 08.11.2011 15:54, schrieb Takashi Iwai:
> At Tue, 08 Nov 2011 15:32:10 +0100,
> Alexey Fisher wrote:
>>
>> Am 08.11.2011 14:09, schrieb Takashi Iwai:
>>> At Tue, 08 Nov 2011 14:05:32 +0100,
>>> Takashi Iwai wrote:
>>>>
>>>> At Tue, 08 Nov 2011 13:39:38 +0100,
>>>> Alexey Fisher wrote:
>>>>>
>>>>> Am 08.11.2011 13:11, schrieb Takashi Iwai:
>>>>>> At Tue, 08 Nov 2011 12:25:06 +0100,
>>>>>> Alexey Fisher wrote:
>>>>>>>
>>>>>>> Am 08.11.2011 10:59, schrieb Takashi Iwai:
>>>>>>>> At Tue, 08 Nov 2011 09:49:13 +0100,
>>>>>>>> Alexey Fisher wrote:
>>>>>>>>>
>>>>>>>>> Hallo all, Takashi,
>>>>>>>>>
>>>>>>>>> there is one regression in current linux main tree:
>>>>>>>>>
>>>>>>>>> cval->res quirks do not work any more. See usb/mixer.c:1150
>>>>>>>>> Are there was some work in this area or may be some one bisected this
>>>>>>>>> bug down?
>>>>>>>>
>>>>>>>> Are you referring to the device 046d:09a4, which sets the mixer range
>>>>>>>> manually?  If yes, does the patch below work?
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> No, i referring to devices 046d:0991, 046d:081d, i don't have access to
>>>>>>> 046d:09a4.
>>>>>>>
>>>>>>> This part is broken for me:
>>>>>>>             case USB_ID(0x046d, 0x0991):
>>>>>>>             /* Most audio usb devices lie about volume resolution.
>>>>>>>              * Most Logitech webcams have res = 384.
>>>>>>>              * Proboly there is some logitech magic behind this number --fishor
>>>>>>>              */
>>>>>>>                     if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
>>>>>>>                             snd_printk(KERN_INFO
>>>>>>>                                     "set resolution quirk: cval->res = 384\n");
>>>>>>>                             cval->res = 384;
>>>>>>>                     }
>>>>>>>                     break;
>>>>>>
>>>>>> OK, so _what_ is actually broken?  Do you see the message but
>>>>>> cval->res is overridden later?
>>>>>
>>>>> yes.
>>>>
>>>> To which value?
>>>
>>> Also you should have got a message like "0:1: cannot get min/max
>>> values for ..." when built with CONFIG_SND_DEBUG.  Check whether you
>>> get it actually.
>>
>> It sets the res value to 1.
>> I get fallowing message after i plug in the webcam:
>>
>> [  132.878150] usb 1-3.2: new high-speed USB device number 8 using ehci_hcd
>> [  133.087255] usb 1-3.2: New USB device found, idVendor=046d,
>> idProduct=0991
>> [  133.087259] usb 1-3.2: New USB device strings: Mfr=0, Product=0,
>> SerialNumber=2
>> [  133.087261] usb 1-3.2: SerialNumber: 9671DCEE
>> [  133.088428] uvcvideo: Found UVC 1.00 device<unnamed>  (046d:0991)
>> [  133.122400] input: UVC Camera (046d:0991) as
>> /devices/pci0000:00/0000:00:1a.7/usb1/1-3/1-3.2/1-3.2:1.0/input/input15
>> [  133.488506] ALSA mixer.c:817 5:2: cannot get min/max values for
>> control 2 (id 5)
>> [  133.488524] ALSA mixer.c:1168 set resolution quirk: cval->res = 384
>> [  133.488527] ALSA mixer.c:817 5:2: cannot get min/max values for
>> control 2 (id 5)
>> [  133.492187] ALSA mixer.c:817 5:2: cannot get min/max values for
>> control 2 (id 5)
>>
>>
>> With amixer i get this:
>> amixer -c U0x46d0x991
>> Simple mixer control 'Mic',0
>>     Capabilities: cvolume cvolume-joined cswitch cswitch-joined penum
>>     Capture channels: Mono
>>     Limits: Capture 0 - 3072
>>     Mono: Capture 2749 [89%] [28.73dB] [on]
>>
>>
>> But Limits should be 0-7 instead of 0-3072
>
> OK, then the problem is that the first few reads of min/max values
> failed, but cval->res quirk is applied only for the first read.
>
> Does the patch below fix?

Yes. Thanks a lot!

Re: [regression] probably in mixer.c code

[Takashi Iwai]

At Tue, 08 Nov 2011 17:46:57 +0100,
Alexey Fisher wrote:
> 
> Am 08.11.2011 15:54, schrieb Takashi Iwai:
> > At Tue, 08 Nov 2011 15:32:10 +0100,
> > Alexey Fisher wrote:
> >>
> >> Am 08.11.2011 14:09, schrieb Takashi Iwai:
> >>> At Tue, 08 Nov 2011 14:05:32 +0100,
> >>> Takashi Iwai wrote:
> >>>>
> >>>> At Tue, 08 Nov 2011 13:39:38 +0100,
> >>>> Alexey Fisher wrote:
> >>>>>
> >>>>> Am 08.11.2011 13:11, schrieb Takashi Iwai:
> >>>>>> At Tue, 08 Nov 2011 12:25:06 +0100,
> >>>>>> Alexey Fisher wrote:
> >>>>>>>
> >>>>>>> Am 08.11.2011 10:59, schrieb Takashi Iwai:
> >>>>>>>> At Tue, 08 Nov 2011 09:49:13 +0100,
> >>>>>>>> Alexey Fisher wrote:
> >>>>>>>>>
> >>>>>>>>> Hallo all, Takashi,
> >>>>>>>>>
> >>>>>>>>> there is one regression in current linux main tree:
> >>>>>>>>>
> >>>>>>>>> cval->res quirks do not work any more. See usb/mixer.c:1150
> >>>>>>>>> Are there was some work in this area or may be some one bisected this
> >>>>>>>>> bug down?
> >>>>>>>>
> >>>>>>>> Are you referring to the device 046d:09a4, which sets the mixer range
> >>>>>>>> manually?  If yes, does the patch below work?
> >>>>>>>>
> >>>>>>>
> >>>>>>>
> >>>>>>> No, i referring to devices 046d:0991, 046d:081d, i don't have access to
> >>>>>>> 046d:09a4.
> >>>>>>>
> >>>>>>> This part is broken for me:
> >>>>>>>             case USB_ID(0x046d, 0x0991):
> >>>>>>>             /* Most audio usb devices lie about volume resolution.
> >>>>>>>              * Most Logitech webcams have res = 384.
> >>>>>>>              * Proboly there is some logitech magic behind this number --fishor
> >>>>>>>              */
> >>>>>>>                     if (!strcmp(kctl->id.name, "Mic Capture Volume")) {
> >>>>>>>                             snd_printk(KERN_INFO
> >>>>>>>                                     "set resolution quirk: cval->res = 384\n");
> >>>>>>>                             cval->res = 384;
> >>>>>>>                     }
> >>>>>>>                     break;
> >>>>>>
> >>>>>> OK, so _what_ is actually broken?  Do you see the message but
> >>>>>> cval->res is overridden later?
> >>>>>
> >>>>> yes.
> >>>>
> >>>> To which value?
> >>>
> >>> Also you should have got a message like "0:1: cannot get min/max
> >>> values for ..." when built with CONFIG_SND_DEBUG.  Check whether you
> >>> get it actually.
> >>
> >> It sets the res value to 1.
> >> I get fallowing message after i plug in the webcam:
> >>
> >> [  132.878150] usb 1-3.2: new high-speed USB device number 8 using ehci_hcd
> >> [  133.087255] usb 1-3.2: New USB device found, idVendor=046d,
> >> idProduct=0991
> >> [  133.087259] usb 1-3.2: New USB device strings: Mfr=0, Product=0,
> >> SerialNumber=2
> >> [  133.087261] usb 1-3.2: SerialNumber: 9671DCEE
> >> [  133.088428] uvcvideo: Found UVC 1.00 device<unnamed>  (046d:0991)
> >> [  133.122400] input: UVC Camera (046d:0991) as
> >> /devices/pci0000:00/0000:00:1a.7/usb1/1-3/1-3.2/1-3.2:1.0/input/input15
> >> [  133.488506] ALSA mixer.c:817 5:2: cannot get min/max values for
> >> control 2 (id 5)
> >> [  133.488524] ALSA mixer.c:1168 set resolution quirk: cval->res = 384
> >> [  133.488527] ALSA mixer.c:817 5:2: cannot get min/max values for
> >> control 2 (id 5)
> >> [  133.492187] ALSA mixer.c:817 5:2: cannot get min/max values for
> >> control 2 (id 5)
> >>
> >>
> >> With amixer i get this:
> >> amixer -c U0x46d0x991
> >> Simple mixer control 'Mic',0
> >>     Capabilities: cvolume cvolume-joined cswitch cswitch-joined penum
> >>     Capture channels: Mono
> >>     Limits: Capture 0 - 3072
> >>     Mono: Capture 2749 [89%] [28.73dB] [on]
> >>
> >>
> >> But Limits should be 0-7 instead of 0-3072
> >
> > OK, then the problem is that the first few reads of min/max values
> > failed, but cval->res quirk is applied only for the first read.
> >
> > Does the patch below fix?
> 
> Yes. Thanks a lot!

Thanks for a quick test.  Now I queued the patch up for the next pull
request.


Takashi