On Sun, 13 Apr 2003, Mark M. Hoffman wrote: > I think (but I'm not certain) that most of the problems with > Asus support are for actual AS99127. As written, ASB100 works > well enough for me. I've heard some Athlon owners found another > register that gives better CPU temps, but I haven't seen a patch > for it. Hi, I wrote a small patch which reports the same values that xmbmon does. It breaks other AS99127 chips right now, but it works on my A7V8X. The reason for this is because the 0x17 register holds the actual temp in Celcius, which doesn't work well with the TEMP_FROM_REG macros (the >> 7 zeroes the value). So I added a conditional which multiplies the raw value by 10 instead. After that, the correct value is reported. I chose temp3 to modify instead of temp2 because temp2 matches the temp3 value that xmbmon reports (chipset). Maybe someone else with more knowledge of lm-sensors than I can clean this up for insertion into upstream. Also note that the limit and hysteresis values are incorrect. A reference chart I found at http://home.kabelfoon.nl/~bertrik/asb100/asb100.html listed 0x18 and 0x19 as possible locations for them, but a w83781d_read_value returns a 0 for those locations. You can find the patch at http://tynian.net/asb100/ -- Brian Almeida | http://tynian.net | bma at tynian.net Debian Developer | http://www.debian.org | bma at debian.org
