Intel pstate

[Colin Ian King]

On 13/04/14 01:16, Tim Gardner wrote:
> Should we revisit our decision to disable pstate in light of support
> from thermald ?
> 
> https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1188647/comments/14
> 
> Colin - was your work on pstate power consumption ever published in an
> open forum? I am referring to a Dec 20, 2013 email from you entitled
> "Intel-pstate driver power measurements".
> 
> rtg

My concern was that some i7 machines we're overheating with pstate
enabled, hence the drive to get thermald into Trusty.

However, I wanted to take a more conservative view on thermald since it
is still maturing and we went through some iterations to iron out quite
a few issues found in static analysis, the security review and testing
during Q1 of 2014. Since this was a relatively new hunk of C++ I wanted
to give it more time to get some thorough testing on various systems out
there, so I was reluctant to make it default with intel-pstate enabled
for the LTS.  It is however, my aim to enable intel-pstate and with
thermald for T+1.

Colin