Kirk Lapray wrote:
> That's what I am thinking. It looks like those registers are trying
> to be read before the card has locked onto the signal. I only get
> these errors if I tune to channel that doesn't have a digital signal.
> I wonder if the tuner in the HDTV Wonder is able to lock on faster
> than the one in the A180. Since you get these errors when there is
> not a signal, and it doesn't mean anything is wrong, we should
> probably ignore it.
>
> After looking over the nxt2002_readreg_multibyte code again it looks
> like it never even checks for errors. We might just want to do the
> same and remove
Can you just send me a patch? (with your sign-off and short
description) It's a lot easier for me to see what you're changing that
way, and also a much better (and safer) way to apply the change to cvs.
Thanks,
-Mike
>
> In nxt200x_readreg_multibyte change it from:
>
> case NXT2004:
> /* probably not right, but gives correct values */
> attr = 0x02;
> if (reg & 0x80) {
> attr = attr << 1;
> if (reg & 0x04)
> attr = attr >> 1;
> }
>
> /* set multi register length */
> len2 = (attr << 4) | len;
> nxt200x_writebytes(state, 0x34, &len2, 1);
>
> /* toggle the multireg bit*/
> buf = 0x80;
> nxt200x_writebytes(state, 0x21, &buf, 1);
>
> /* read status */
> nxt200x_readbytes(state, 0x21, &buf, 1);
>
> if (buf == 0)
> {
> /* read the actual data */
> for(i = 0; i < len; i++) {
> nxt200x_readbytes(state, 0x36
> + i, &data[i], 1);
> }
> return 0;
> }
> break;
> default:
> return -EINVAL;
> break;
> }
>
> printk(KERN_WARNING "nxt200x: Error reading multireg register
> 0x%02X BUFFER=0x%02X\n",reg,buf);
>
> return 0;
>
> To:
>
> case NXT2004:
> /* probably not right, but gives correct values */
> attr = 0x02;
> if (reg & 0x80) {
> attr = attr << 1;
> if (reg & 0x04)
> attr = attr >> 1;
> }
>
> /* set multi register length */
> len2 = (attr << 4) | len;
> nxt200x_writebytes(state, 0x34, &len2, 1);
>
> /* toggle the multireg bit*/
> buf = 0x80;
> nxt200x_writebytes(state, 0x21, &buf, 1);
>
> /* read the actual data */
> for(i = 0; i < len; i++) {
> nxt200x_readbytes(state, 0x36 + i,
> &data[i], 1);
> }
> return 0;
> break;
> default:
> return -EINVAL;
> break;
> }
>
>
> Kirk
>
> On 10/27/05, *Michael Krufky* <mkrufky@xxxxxxx
> <mailto:mkrufky@xxxxxxx>> wrote:
>
> Kirk Lapray wrote:
>
> > On 10/27/05, *Michael Krufky* <mkrufky@xxxxxxx
> <mailto:mkrufky@xxxxxxx>
> > <mailto:mkrufky@xxxxxxx <mailto:mkrufky@xxxxxxx>>> wrote:
> >
> > Although everything is working well, I am still getting
> these errors
> > when I first run azap:
> >
> > nxt200x: NXT2004 Detected
> > DVB: registering new adapter (saa7133[0]).
> > DVB: registering frontend 0 (Nextwave NXT200X VSB/QAM
> frontend)...
> > nxt2004: Waiting for firmware upload ( dvb-fe-nxt2004.fw)...
> > nxt2004: Waiting for firmware upload(2)...
> > nxt2004: Firmware upload complete
> > nxt200x: Error reading multireg register 0xA6
> > nxt200x: Error reading multireg register 0xE6
> > nxt200x: Error reading multireg register 0xE6
> >
> > Never more than 3 lines in the log... It is tuning and
> delivering
> > my ts
> > stream just fine.
> >
> > I know the nxt2002 driver used dprintk instead of printk in the
> > multireg read and write commands. We could switch it to dprintk and
> > then you would only see these errors if you loaded the driver with
> > debug set.
>
> Kirk -
>
> So... I guess that translates to "those errors don't matter." Would
> that be the correct assessment?
>
> I'm not worried about whether I am *seeing* them or not... I was only
> concerned about whether it was indicating something worth paying
> attention to.
>
> -Mike
>
>
