On Thu, 2020-09-17 at 13:16 -0400, Mark Pearson wrote: > > As I understand the problem with the configured and actual > > value/performance_level ideas is that if I understand things correctly > > that H* is not the same as M, > > it behaves close-ish to M because of the lower thermal-limits from > > lapmode, but if I understood Benjamin correctly is is not exactly the > > same, so if we were > > to advertise "M" in the actual_performance_level sysfs-attribute then > > that would not really be correct ? > > Just a small clarification - in our case High performance is only for > deskmode. It drops to Medium in lapmode. > Medium mode is slightly different in power rating between lap and desk > mode (e.g on X1Carbon 14.5W on lap, 15W on desk). I haven't really > worried about this in my patch implementation - it's still "medium" Hmm, what I saw on the X1 Carbon 7th Gen was that putting it into performance mode (H) increases the fan speed. And I believe that this was also the case while it was in lap mode. Specifically, I believe I observed that changing dytc_perfmode from M to H caused the fan to turn on (and vice versa). So while CPU power limits are very similar, there still seems to be differences in the active cooling behaviour. Benjamin
Attachment:
signature.asc
Description: This is a digitally signed message part
