Re: [PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-10 Thread Sudeep Holla
On 10/11/16 10:28, Vincent Guittot wrote: On 9 November 2016 at 19:48, Sudeep Holla wrote: On 09/11/16 18:39, Lorenzo Pieralisi wrote: [..] I would just init the enter_freeze() pointer and be done with that, adding code to check whether the idle back-end enables

Re: [PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-10 Thread Sudeep Holla
On 10/11/16 10:28, Vincent Guittot wrote: On 9 November 2016 at 19:48, Sudeep Holla wrote: On 09/11/16 18:39, Lorenzo Pieralisi wrote: [..] I would just init the enter_freeze() pointer and be done with that, adding code to check whether the idle back-end enables IRQs when it enters

Re: [PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-10 Thread Vincent Guittot
On 9 November 2016 at 19:48, Sudeep Holla wrote: > > > On 09/11/16 18:39, Lorenzo Pieralisi wrote: >> >> On Wed, Nov 09, 2016 at 05:43:30PM +, Sudeep Holla wrote: >>> >>> enter_freeze() callback is expected atleast to do the same as enter() >>> but it has to guarantee

Re: [PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-10 Thread Vincent Guittot
On 9 November 2016 at 19:48, Sudeep Holla wrote: > > > On 09/11/16 18:39, Lorenzo Pieralisi wrote: >> >> On Wed, Nov 09, 2016 at 05:43:30PM +, Sudeep Holla wrote: >>> >>> enter_freeze() callback is expected atleast to do the same as enter() >>> but it has to guarantee that interrupts aren't

Re: [PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-09 Thread Sudeep Holla
On 09/11/16 18:39, Lorenzo Pieralisi wrote: On Wed, Nov 09, 2016 at 05:43:30PM +, Sudeep Holla wrote: enter_freeze() callback is expected atleast to do the same as enter() but it has to guarantee that interrupts aren't enabled at any point in its execution, as the tick is frozen. CPUs

Re: [PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-09 Thread Sudeep Holla
On 09/11/16 18:39, Lorenzo Pieralisi wrote: On Wed, Nov 09, 2016 at 05:43:30PM +, Sudeep Holla wrote: enter_freeze() callback is expected atleast to do the same as enter() but it has to guarantee that interrupts aren't enabled at any point in its execution, as the tick is frozen. CPUs

Re: [PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-09 Thread Lorenzo Pieralisi
On Wed, Nov 09, 2016 at 05:43:30PM +, Sudeep Holla wrote: > enter_freeze() callback is expected atleast to do the same as enter() > but it has to guarantee that interrupts aren't enabled at any point > in its execution, as the tick is frozen. > > CPUs execute ->enter_freeze with the local

Re: [PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-09 Thread Lorenzo Pieralisi
On Wed, Nov 09, 2016 at 05:43:30PM +, Sudeep Holla wrote: > enter_freeze() callback is expected atleast to do the same as enter() > but it has to guarantee that interrupts aren't enabled at any point > in its execution, as the tick is frozen. > > CPUs execute ->enter_freeze with the local

[PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-09 Thread Sudeep Holla
enter_freeze() callback is expected atleast to do the same as enter() but it has to guarantee that interrupts aren't enabled at any point in its execution, as the tick is frozen. CPUs execute ->enter_freeze with the local tick or entire timekeeping suspended, so it must not re-enable interrupts

[PATCH] drivers: cpuidle: assign enter_freeze to same as enter callback function

2016-11-09 Thread Sudeep Holla
enter_freeze() callback is expected atleast to do the same as enter() but it has to guarantee that interrupts aren't enabled at any point in its execution, as the tick is frozen. CPUs execute ->enter_freeze with the local tick or entire timekeeping suspended, so it must not re-enable interrupts