Re: regression : saa7134 with Pinnacle PCTV 50i (analog) can not tune anymore

[hermann pitton <hermann-pitton@arcor.de>]

Am Dienstag, den 07.07.2009, 08:44 +0200 schrieb
eric.paturage@orange.fr:
> > 
> > Hi Eric,
> > 
> > on a first short look only at lsmod and your kernel .config,
> > I can see tda8290 and tda827x are not loaded and you have nothing
> > conflicting in your kernel .config.
> > 
> > Must be a bug in the build system and I need to get a recent 2.6.30 I
> > guess, since else not reproducible.
> > 
> > Did you try to load the tda827x manually too?
> > If it is not successfully tuned it shows that black.
> > 
> > Try to load all tuner modules before the saa7134 and on a second look
> > try also to disable the IR for now.
> > 
> > Cheers,
> > Hermann
> > 
> > 
> > 
> 
> Hi Hermann 
> 
> Enclosed is dmesg when doing a modprobe saa7134  , before doing that tuner modules tea5767 tda8290 tda18271 tda827x
> were manualy loaded . I add also the output of lsmod (after that operation) .
> Still no changes . 
> 
> I don't like these  lines (are they "normal"? ) :
> i2c-adapter i2c-1: Invalid 7-bit address 0x7a
> saa7133[0]: i2c xfer: < 8e >
> input: i2c IR (Pinnacle PCTV) as /class/input/input4
> ir-kbd-i2c: i2c IR (Pinnacle PCTV) detected at i2c-1/1-0047/ir0 [saa7133[0]]
> saa7133[0]: i2c xfer: < 8f ERROR: ARB_LOST
> saa7133[0]: i2c xfer: < 84 ERROR: NO_DEVICE
> saa7133[0]: i2c xfer: < 86 ERROR: ARB_LOST
> saa7133[0]: i2c xfer: < 94 ERROR: NO_DEVICE                   
> saa7133[0]: i2c xfer: < 96 ERROR: ARB_LOST                    
> saa7133[0]: i2c xfer: < c0 ERROR: NO_DEVICE                   
> saa7133[0]: i2c xfer: < c2 ERROR: NO_DEVICE                   
> saa7133[0]: i2c xfer: < c4 ERROR: NO_DEVICE                   
> saa7133[0]: i2c xfer: < c6 ERROR: NO_DEVICE                   
> saa7133[0]: i2c xfer: < c8 ERROR: NO_DEVICE                   
> saa7133[0]/core: hwinit2   
> 
> These lines do not appear in the dmesg of modprobe saa7134 in the vanilla kernel 2.6.30.1
> (which works fine ) 
> 
> I enclose for reference dmesg of vanillla kernel 2.6.30.1 when doing modprobe saa7134
> 
> regards 
> 

Hi Eric,

yes, arbitration lost on i2c is an error condition.

As far I know we did not change the bus speed or anything, but some
cards need and i2c quirk to work correctly with the clients.

Mike recently changed the old quirk with good reasons and it was widely
tested, also by me, without any negative effect seen.

Maybe your card is a rare case needing the old quirk.

You could try to change the quirk in saa7134-i2c.c

static int saa7134_i2c_xfer(struct i2c_adapter *i2c_adap,
			    struct i2c_msg *msgs, int num)
{
	struct saa7134_dev *dev = i2c_adap->algo_data;
	enum i2c_status status;
	unsigned char data;
	int addr,rc,i,byte;

	status = i2c_get_status(dev);
	if (!i2c_is_idle(status))
		if (!i2c_reset(dev))
			return -EIO;

	d2printk("start xfer\n");
	d1printk(KERN_DEBUG "%s: i2c xfer:",dev->name);
	for (i = 0; i < num; i++) {
		if (!(msgs[i].flags & I2C_M_NOSTART) || 0 == i) {
			/* send address */
			d2printk("send address\n");
			addr  = msgs[i].addr << 1;
			if (msgs[i].flags & I2C_M_RD)
				addr |= 1;
			if (i > 0 && msgs[i].flags & I2C_M_RD && msgs[i].addr != 0x40) {
				/* workaround for a saa7134 i2c bug
				 * needed to talk to the mt352 demux
				 * thanks to pinnacle for the hint */
				int quirk = 0xfe;    <--------------------------------------
				d1printk(" [%02x quirk]",quirk);
				i2c_send_byte(dev,START,quirk);
				i2c_recv_byte(dev);
			}

back to 0xfd.

Cheers,
Hermann