anysee e30 suspend->resume causes wrong profiling of card.

The anysee driver works correctly from cold boot and reinsertion of the device, however, after a suspend resume cycle (S3), the device suddenly is initated as dvb-t as where it was dvb-c before. Yes this is a combo device, so dvb T and C, but why does the profiling in anysee.c not handle this case? Obviously the following snippet produces a false positive on warm boot and resume:

/* Zarlink ZL10353 DVB-T demod inside of Samsung DNOS404ZH103A NIM */

        adap->fe = dvb_attach(zl10353_attach, &anysee_zl10353_config,
                              &adap->dev->i2c_adap);
        if (adap->fe != NULL) {
                state->tuner = DVB_PLL_THOMSON_DTT7579;
                info("mine: case 2");
                return 0;
        }

I've looked through the rest of the code and by no means am I a developer but, isn't the problem that on warm boots the register of the anysee device doesn't hold the right value in combination with a combo device? because in all the other cases when profiling for different kind of device like the e30c the register is put in a different state before probing for the demuxer.

In the meantime I have commented out the above snippet, which results in a works-for-me. But it isn't a nice solution for the average/new linux user wanting to build a htpc with a anysee combo device.

Tested with ubuntu-lucid module, further tested/compiled with the HG repo.

ps I'm new with mailing lists, is this the right place to post for the anysee driver?

Terve Hans,

This is known problem. Actually it is coming from wrong GPIOs / demodulator selection logic. I just guessed those in the time driver was made. Now I have also correct info. Unfortunately I don't even have this device currently... IIRC you can blacklist zl10353 driver as workaround.

regards
Antti

http://palosaari.fi/