[PATCH] bttv: Use btv->has_radio rather then the card info

[PATCH] bttv: Use btv->has_radio rather then the card info

[Hans de Goede]

I've been spending some time playing with radio receivers under Linux, and
I noticed that my bttv's radio reception was not working. The patch in the
next message fixes the 1st of 3 problems related to this, can you review this
patch please and also let me know if it is ok to send patch though my tree?

As said there are 3 problems with the radio on my bttv:
1) The problem fixed by the attached patch

2) The audio_mux function in bttv-driver.c does the wrong thing for my
Haupage WinTV + radio + stereo (msp3400) card. Since the radio receiving
is part of the same tuner, and not in a separate chip, the same (default
) msp3400 input should be selected when trying to listen to radio, as when
watching TV.

I could special case my card in audio_mux, but in general it seems
to me that if their is a combined radio/tv tuner, rather then a separate
radio chip (ie the tea5757), the msp3400 input should stay connected
to the default / tuner1 input.

I guess we should go with special casing for now, to avoid regressions,
although I wonder how many people try to use the radio part of any tvcards
they may have.

3) The automatic selection of radio versus tv-mode goes astray pretty quickly,
if I start the radio program from xawtv, which I've patched in git to do
digital loopback of the audio through alsa, so as that I can actual hear
what the radio is doing :) And with a quick hack to fix 2), and then after
that start gtk-v4l to look at the radio controls (only the mute one actually),
then bttv automatically switches from radio to tv mode, not good (tm).

IMHO it would be better to defer the switching to tv mode until something
relevant is actually done to the /dev/video# node (set operations or start
streaming), likewise it would probably be best to defer switching to radio
mode until something relevant (any set operations) is actually done on
/dev/radio#. If you agree I can try to write a patch for this.

Regards,

Hans

[PATCH] bttv: Use btv->has_radio rather then the card info when registering the tuner

[Hans de Goede]

bttv_init_card2() sets btv->has_audio to a *default* value from the tvcards
array and then may update it by reading a card specific eeprom or gpio
detection.

After bttv_init_card2(), bttv_init_tuner(), and it should clearly use
the updated, dynamic has_radio value from btv->has_radio, rather then
the const value in the tvcards array.

This fixes the radio not working on my Hauppauge WinTV.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
---
 drivers/media/video/bt8xx/bttv-cards.c  |    4 ++--
 drivers/media/video/bt8xx/bttv-driver.c |    5 +++++
 2 files changed, 7 insertions(+), 2 deletions(-)

diff --git a/drivers/media/video/bt8xx/bttv-cards.c b/drivers/media/video/bt8xx/bttv-cards.c
index ff2933a..1c030fe 100644
--- a/drivers/media/video/bt8xx/bttv-cards.c
+++ b/drivers/media/video/bt8xx/bttv-cards.c
@@ -3649,7 +3649,7 @@ void __devinit bttv_init_tuner(struct bttv *btv)
 		struct tuner_setup tun_setup;
 
 		/* Load tuner module before issuing tuner config call! */
-		if (bttv_tvcards[btv->c.type].has_radio)
+		if (btv->has_radio)
 			v4l2_i2c_new_subdev(&btv->c.v4l2_dev,
 				&btv->c.i2c_adap, "tuner",
 				0, v4l2_i2c_tuner_addrs(ADDRS_RADIO));
@@ -3664,7 +3664,7 @@ void __devinit bttv_init_tuner(struct bttv *btv)
 		tun_setup.type = btv->tuner_type;
 		tun_setup.addr = addr;
 
-		if (bttv_tvcards[btv->c.type].has_radio)
+		if (btv->has_radio)
 			tun_setup.mode_mask |= T_RADIO;
 
 		bttv_call_all(btv, tuner, s_type_addr, &tun_setup);
diff --git a/drivers/media/video/bt8xx/bttv-driver.c b/drivers/media/video/bt8xx/bttv-driver.c
index a9cfb0f..7ce3aac 100644
--- a/drivers/media/video/bt8xx/bttv-driver.c
+++ b/drivers/media/video/bt8xx/bttv-driver.c
@@ -1181,6 +1181,8 @@ audio_mux(struct bttv *btv, int input, int mute)
 
 	btv->mute = mute;
 	btv->audio = input;
+	
+	printk("input: %d, mute: %d\n", input, mute);
 
 	/* automute */
 	mute = mute || (btv->opt_automute && !signal && !btv->radio_user);
@@ -1220,6 +1222,9 @@ audio_mux(struct bttv *btv, int input, int mute)
 		case TVAUDIO_INPUT_RADIO:
 			in = MSP_INPUT(MSP_IN_SCART2, MSP_IN_TUNER1,
 				    MSP_DSP_IN_SCART, MSP_DSP_IN_SCART);
+				in = MSP_INPUT(MSP_IN_SCART1, MSP_IN_TUNER2, \
+					MSP_DSP_IN_TUNER, MSP_DSP_IN_TUNER);
+				in = MSP_INPUT_DEFAULT;
 			break;
 		case TVAUDIO_INPUT_EXTERN:
 			in = MSP_INPUT(MSP_IN_SCART1, MSP_IN_TUNER1,
-- 
1.7.10

Re: [PATCH] bttv: Use btv->has_radio rather then the card info

[Hans Verkuil]

On Sun May 20 2012 13:28:11 Hans de Goede wrote:
> I've been spending some time playing with radio receivers under Linux, and
> I noticed that my bttv's radio reception was not working. The patch in the
> next message fixes the 1st of 3 problems related to this, can you review this
> patch please and also let me know if it is ok to send patch though my tree?
> 
> As said there are 3 problems with the radio on my bttv:
> 1) The problem fixed by the attached patch
> 
> 2) The audio_mux function in bttv-driver.c does the wrong thing for my
> Haupage WinTV + radio + stereo (msp3400) card. Since the radio receiving
> is part of the same tuner, and not in a separate chip, the same (default
> ) msp3400 input should be selected when trying to listen to radio, as when
> watching TV.
> 
> I could special case my card in audio_mux, but in general it seems
> to me that if their is a combined radio/tv tuner, rather then a separate
> radio chip (ie the tea5757), the msp3400 input should stay connected
> to the default / tuner1 input.
> 
> I guess we should go with special casing for now, to avoid regressions,
> although I wonder how many people try to use the radio part of any tvcards
> they may have.
> 
> 3) The automatic selection of radio versus tv-mode goes astray pretty quickly,
> if I start the radio program from xawtv, which I've patched in git to do
> digital loopback of the audio through alsa, so as that I can actual hear
> what the radio is doing :) And with a quick hack to fix 2), and then after
> that start gtk-v4l to look at the radio controls (only the mute one actually),
> then bttv automatically switches from radio to tv mode, not good (tm).

This was the behavior for a long time: opening the radio node was sufficient
to switch to radio mode, and opening a video node would switch to video.

That behavior has been marked obsolete and in the feature-removal document it
is slated for removal. Unfortunately I haven't had time to do any of that work.

The only time you should switch between radio and tv tuner is with S_TUNER,
S_FREQUENCY or S_HW_FREQ_SEEK, issued from the corresponding radio or video node.

Trying to switch to radio mode if streaming is in progress should obviously
result in EBUSY.

> IMHO it would be better to defer the switching to tv mode until something
> relevant is actually done to the /dev/video# node (set operations or start
> streaming), likewise it would probably be best to defer switching to radio
> mode until something relevant (any set operations) is actually done on
> /dev/radio#. If you agree I can try to write a patch for this.

It would be nice if bttv was fixed. There are loads more issues with bttv,
that's going to be a big project at one time.

Regards,

	Hans

Re: [PATCH] bttv: Use btv->has_radio rather then the card info when registering the tuner

[Andrew Benham]

On 20/05/12 12:28, Hans de Goede wrote:

> After bttv_init_card2(), bttv_init_tuner(), and it should clearly use
> the updated, dynamic has_radio value from btv->has_radio, rather then
> the const value in the tvcards array.
> 
> This fixes the radio not working on my Hauppauge WinTV.
> 
> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
> ---

Thanks for this.  The radio on my Hauppauge WinTV last worked with the
2.6.38 kernel.
Since April 2012 there's been no analogue TV in the London area, and as
the radio part of the card didn't work I was about to throw it away :-(
But a quick rebuild of the module and I have radio again.  Many thanks.

> diff --git a/drivers/media/video/bt8xx/bttv-cards.c b/drivers/media/video/bt8xx/bttv-cards.c
> index ff2933a..1c030fe 100644
> --- a/drivers/media/video/bt8xx/bttv-cards.c
> +++ b/drivers/media/video/bt8xx/bttv-cards.c
> @@ -3649,7 +3649,7 @@ void __devinit bttv_init_tuner(struct bttv *btv)
>  		struct tuner_setup tun_setup;
>  
>  		/* Load tuner module before issuing tuner config call! */
> -		if (bttv_tvcards[btv->c.type].has_radio)
> +		if (btv->has_radio)
>  			v4l2_i2c_new_subdev(&btv->c.v4l2_dev,
>  				&btv->c.i2c_adap, "tuner",
>  				0, v4l2_i2c_tuner_addrs(ADDRS_RADIO));
> @@ -3664,7 +3664,7 @@ void __devinit bttv_init_tuner(struct bttv *btv)
>  		tun_setup.type = btv->tuner_type;
>  		tun_setup.addr = addr;
>  
> -		if (bttv_tvcards[btv->c.type].has_radio)
> +		if (btv->has_radio)
>  			tun_setup.mode_mask |= T_RADIO;
>  
>  		bttv_call_all(btv, tuner, s_type_addr, &tun_setup);

-- 
Andrew Benham     Southgate, London N14, United Kingdom

The gates in my computer are AND OR and NOT, not "Bill"