On 08/08/15 15:20, Guenter Roeck wrote: > >> >> We are running RedHawk5.4.11, 2.6.31.13, based on RHEL5.4 x86, on an >> IBM HS23. >> Executing sensors showed no output at all, no sensors detected. >> I looked at coretemp.c and saw that the HS23 CPU, 32nm, CPUID= 0X2D, >> was not supported and in fact coretemp was reporting that in > /var/log/messages. >> I added 0x2D to coretemp.c and rebuilt the kernel to see if we could > even read >> a sensor. That worked and now there is a temperature reported for each >> CPU. Additionally, you don't need to patch and recompile the whole kernel for just one driver - you can build your patched coretemp driver out of tree. >> However, with the "workaround" I don't have any confidence in the > accuracy >> of the output of course. Is there a patch I can apply to the kernel > source >> that will update coretemp.c and related source code? > > The upstream driver has been rewritten pretty much completely since then. > I had a quick look, but back-porting the latest version to 2.6.31 > would be a major effort. > > I think the best you can do is to stick with what you did. > > Guenter > _______________________________________________ lm-sensors mailing list lm-sensors@xxxxxxxxxxxxxx http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
