Hi Ondrej, > i2cdump 0 0x44: I don't know what this one is. Your main device (see below) has no sub-address AFAICT. If you re-run the command some times later, are the values still the same? > (1) a hardware monitoring ASIC (Analog Devices ADM1025, Philips > Semiconductor NE1619, or equivalent) include: > (...) > (2) hardware monitoring and fan control ASIC (Analog Devieces ADM1027, > National Semiconductor LM85, Standart Microsystems SMSC EMC6D101, or > equivalent) include: We already have support for ADM1025, ADM1027, LM85 and EMC6D101, so your chip has to be a NE1619. I took a look at your i2cdump and it matches, I'm 95% sure it's a NE1619. The chip is pin-to-pin compatible with the ADM1025, and register compatible too as far as I can say. I just added detection for the NE1619 to our sensors-detect script. Please grab it there: http://www2.lm-sensors.nu/~lm78/cvs/lm_sensors2/prog/detect/sensors-detect Could you please give it a try? It should detect your chip at 0x2d as a NE1619 with confidence 8, and suggest adm1025 as the driver. Then you could try forcing the adm1025 driver: modprobe adm1025 force_adm1025=0,0x2d Let us know how it works. Since both chips seem to be compatible, I expect it to give very good result. If you confirm that, I'll add the NE1619 ID to the driver so that the force parameter isn't necessary anymore. -- Jean Delvare http://www.ensicaen.ismra.fr/~delvare/
