Well I'd thought I'd share this information with ya. It seems some people (including me) were baffled as to why some systems running phenoms had only 2 p-states in linux but in windows it would show more. After a lot of finger pointing (ACPI guys pointed to BIOS devels and they pointed to MB manufacturers who pointed back to the kernel guys.....).
Anyhow here is the skinny on p-states from Joachim Deguara @ AMD.
So there you go, straight from AMD.
Anyhow here is the skinny on p-states from Joachim Deguara @ AMD.
This is really a bug in the windows software than in Linux. There are only two
P-States for this property to be exported by the BIOS and that is being done
correctly looking at your logs. However the window tool shows other
information which is not correct. I will follow up with our windows people but
this bug can be closed as it is correct behavior under Linux.
I tried the tool version 1.2.3 on my system and only saw it only switch between
the two speeds. One thing that might be confusing is it lists a third speed
which is that of the North-Bridge and can be ignored as Linux does use
North-Bridge P-States but does not output that information in dmesg.
P-States for this property to be exported by the BIOS and that is being done
correctly looking at your logs. However the window tool shows other
information which is not correct. I will follow up with our windows people but
this bug can be closed as it is correct behavior under Linux.
I tried the tool version 1.2.3 on my system and only saw it only switch between
the two speeds. One thing that might be confusing is it lists a third speed
which is that of the North-Bridge and can be ignored as Linux does use
North-Bridge P-States but does not output that information in dmesg.
So there you go, straight from AMD.
Comment