Messages in this thread | | | From | Michael Neuling <> | Subject | Re: [PATCH] sched: Enable arch-specific asym packing option in sched domain | Date | Sat, 05 May 2012 08:51:58 +1000 |
| |
> We need this for Tegra's slight asymmetry with core0.
This would seem to be to be a different problem. I don't think the packing mechanism is what you want.
Shouldn't you be increasing the CPU power of this core, so that all tasks get a fair go on this core? Otherwise you're breaking the fundamental concept of a completely fair scheduler.
Using this packing mechanism you won't get this. Any task that lands on core 0 will get an unfair amount of computation power, compared to tasks that landed on core 1.
Mikey
| |