Re: [patch 14/16] sched: make __update_entity_runnable_avg() fast

2012-08-28 Thread Paul Turner
Applied, Thanks. -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/

Re: [patch 14/16] sched: make __update_entity_runnable_avg() fast

2012-08-24 Thread Namhyung Kim
On Thu, 23 Aug 2012 07:14:36 -0700, > From: Paul Turner > > __update_entity_runnable_avg forms the core of maintaining an entity's > runnable > load average. In this function we charge the accumulated run-time since last > update and handle appropriate decay. In some cases, e.g. a waking task,

[patch 14/16] sched: make __update_entity_runnable_avg() fast

2012-08-23 Thread pjt
From: Paul Turner __update_entity_runnable_avg forms the core of maintaining an entity's runnable load average. In this function we charge the accumulated run-time since last update and handle appropriate decay. In some cases, e.g. a waking task, this time interval may be much larger than our p

Re: [PATCH 14/16] sched: make __update_entity_runnable_avg() fast

2012-07-12 Thread Peter Zijlstra
On Wed, 2012-07-11 at 17:15 -0700, Paul Turner wrote: > convergence > 345> 47765 Ah, 345 is much saner indeed! > And for posterity, a simple generator so that I don't lose it again: > #include > #include > > #define SRR(x, y) (((x) + (1UL << ((y) - 1))) >> (y)) > #define N 32 > #define WMULT_S

Re: [PATCH 14/16] sched: make __update_entity_runnable_avg() fast

2012-07-11 Thread Paul Turner
So I've been trying to dig up the little proglets that originally computed this stuff, since some specific adjustments were made but for the life of me[*] I cannot find it, so I am stuck trying to reverse engineer it like you :-). [*] Including some over-night greps on all my source trees. The sho

Re: [PATCH 14/16] sched: make __update_entity_runnable_avg() fast

2012-07-10 Thread Peter Zijlstra
On Mon, 2012-07-09 at 13:18 -0700, Benjamin Segall wrote: > Correct, the sum is 1..n. The maximum was chosen as 516/46742 because > that is the point when the approximation of loop + fixed point math > being used reaches a maximum, even if the ideal y^n series would cap out > slightly differently.

Re: [PATCH 14/16] sched: make __update_entity_runnable_avg() fast

2012-07-09 Thread Benjamin Segall
Correct, the sum is 1..n. The maximum was chosen as 516/46742 because that is the point when the approximation of loop + fixed point math being used reaches a maximum, even if the ideal y^n series would cap out slightly differently. -- To unsubscribe from this list: send the line "unsubscribe linux