Hi Alexandre, > Here is the latest patch for adm1031 support on 2.6 kernel. I've tried your even latest patch as downloaded from your website. Works great except one problem: Lowering fan div doesn't correctly preserve fan min (most likely an 8-bit reg overflow). For exemple: # echo 2 > fan1_div # echo 2000 > fan1_min # echo 1 > fan1_div fan1_div will now read 8437 (instead of 2678, the lowest value for div 1) I guess you forgot to port the change from the 2.4 driver. I've also noticed something strange. That's not a bug, but I'd like to know if you noticed the same. It looks like temperature comparisons are done on the integer part only. For example, if the high limit set to 45 degrees, and measured temperature is 45.5 degrees, no alarm will raise. I'll trigger at 46.0 degrees. I think this isn't what most chips do. I'll check with the ADM1032, but I'm almost sure... Maybe we could slightly shift the hardware limit and the user-space visible limit so as to correct that difference? -- Jean Delvare http://khali.linux-fr.org/
