Hi, sorry for not noticing this thread before! > > 0x11 indeed writes to 0x11. > > This reflects somehow how echo into the /sys filesystem reacts. > > Indeed. So this is not a bug in the driver, but rather the hardware > misbehaving. I'll ask out contact at Nuvoton about this. Same behavior here with another Asus board (Maximus II Formula - W83667HG-A). A few words about the topic W83667HG vs ATK0110. I'm currently using the atk0110 driver [1] and I'm happy with it since I don't need a fine grain control for my fans and the default hw control looks pretty good to me. As you know, the Asus software for Windows indeed uses the ACPI interface, so I suggest the Asus owners to move to the ATK0110 driver. However I'd like to implement or to see implemented the fan profile switching for the Q-Fan mode, allowing minimum fan control but still sufficient for most users. I had a quick look at the driver code and at my dsdt and it seems easy enought. I'd like to put attention on another aspect of many recent board from Asus, the power management, aka EPU. Today I gave another try to Everest on Windows and made a sensors dump (attached). As you can see from the bus scan many chips are present, the Winbond, the two unidentified ICs at 0x38 and 0x40, and, among others, an IC at 0x20. Accordingly with a scan from the previous model of this board, the Maxiums Formula (Sensor Type W83627DHG + W83791D + ADT7475 + ADP3228 (ISA 290h, SMB 2Ch/2Eh/20h)) and the Everest output, it is the ASP0800, a custom VRM control chip from AD, providing (I suppose) the current CPU wattage reading. I don't know the right indexes for this information but it's somehow possibile. I think it's interesting for the Asus owners to add this reading as well, if possible, isn't it? It would be great to receive further informations from Asus in order to have some power tuning capabilities like the Windows 6-Engine software provides, but it's out of the hwmon scope and need good contacts at Asus. [1] Well, to tell the true I would have no choice since most readings are provided by two unknown ICs. http://lists.lm-sensors.org/pipermail/lm-sensors/2008-September/024163.html -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: everest.txt Url: http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20090713/6c670ae9/attachment-0001.txt