On Thu, Dec 22, 2011 at 2:23 PM, Arjan van de Ven <arjan@xxxxxxxxxxxxxxx> wrote: > On 12/22/2011 9:35 AM, Shilimkar, Santosh wrote: > >> Indeed. The SOCs, Arch's which does support low power >> state independently and doesn't need any co-ordination between CPU's >> will continue to work same way as before with this series. > > btw I think you misunderstand; I don't object to a need for something > like this, I did. Thanks for clarification. > I am just very concerned that this may not be possible to be > done in a race-free way. > I agree to those races but may be they are harmless. Also the safe state need not be just WFI and can be bit deeper where the co-ordination between isn't necessary. So that should still not burn the power that much. For simplicity let's assume a two CPU scenario. Ideal scenario: CPU 1 has entered idle and incremented counter for the co-ordinated C state. CPU0 also enter and increments the counter and now the subsystem and interconnect can go down along with CPU cluster. Few of the race conditions will possibly lead to void the above conditions and in that case the counter would reflect that and such a C-state won't be attempted and a safe C-state would be attempted. That should still work fine. Some how this hardware/security restriction is bit stupid and likely going against the existing CPUIDLE design which expect that a CPUIDLE thread are per CPU and it should be independent and local to that CPU. Regards Santosh _______________________________________________ linux-pm mailing list linux-pm@xxxxxxxxxxxxxxxxxxxxxxxxxx https://lists.linuxfoundation.org/mailman/listinfo/linux-pm
