Hi Rduolf, On 2005-07-27, Rudolf Marek wrote: > Uff my bad I did not check this. On the other hand the address can be > changed to virtually any address via CR48, with condition that two lowest > bits cant be changed. I think the two lowest bits can be changed. Their initial value is set at power-up time depending on two pin levels, just like is the case for subclients, but I do believe this can be changed afterwards, just like is the case for subclients too. Not that it really matters... > Should we go for range 0x20 - 0x2f ? No. That's what we did for it87, result is that since the driver was written, only default address 0x2d was seen. I can't think of any reason why someone would use a different address than the four possible power-up default ones, especially since this chip has no other interface IIRC, so the power-up address would have to be used at least once to change the address at CR48. If the power-up address works to do that, there is no reason not to use it afterwards as well. So please stick to 0x2c - 0x2f. If anyone ever needs more, that's exactly what the probe parameter is for (but I'm quite certain it will never happen). Thanks, -- Jean Delvare
