Re: [patch 16/18] init: Introduce SYSTEM_BOOTING_UP/SMP states

2017-05-15 Thread Steven Rostedt
On Sun, 14 May 2017 20:27:32 +0200 Thomas Gleixner wrote: > might_sleep() debugging should be active right after the scheduler starts > working and smp_processor_id() debugging right before the first non boot > cpu is brought up. > > Add two new states which allow to enable those checks earlier

Re: [patch 16/18] init: Introduce SYSTEM_BOOTING_UP/SMP states

2017-05-15 Thread Juergen Gross
On 14/05/17 20:27, Thomas Gleixner wrote: > might_sleep() debugging should be active right after the scheduler starts > working and smp_processor_id() debugging right before the first non boot > cpu is brought up. > > Add two new states which allow to enable those checks earlier and add them > to

[patch 16/18] init: Introduce SYSTEM_BOOTING_UP/SMP states

2017-05-14 Thread Thomas Gleixner
might_sleep() debugging should be active right after the scheduler starts working and smp_processor_id() debugging right before the first non boot cpu is brought up. Add two new states which allow to enable those checks earlier and add them to the xen do_poweroff() function. No functional change.