Hello, Viresh.
On Fri, Mar 29, 2013 at 12:57:23PM +0530, Viresh Kumar wrote:
On 28 March 2013 23:43, Tejun Heo tj@kernel.org wrote:
Viresh, would it be difficult to make another measurement of the same workload with the said workqueues converted to unbound? I think that would at least provide a nice reference point.
My heart is shattered after getting the results :)
Cluster A15 Energy Cluster A7 Energy Total
Without this patchset (Energy in Joules):
0.322866 2.619137 2.942003
With this patchset (Energy in Joules):
0.182721 2.443745 2.626466
With WQ Unbound (Energy in Joules):
0.2289042 2.2514632 2.4803674
So the scheduler does know what it's doing after all, which is a nice news. Given the result, the best thing to do would be somehow marking these workqueues as unbound for powersaving?
Thanks!