Re: [RFC][PATCH 09/11] sched, x86: Kill saved_preempt_count

2015-09-29 Thread Peter Zijlstra
On Tue, Sep 29, 2015 at 11:44:04AM -0400, Steven Rostedt wrote: > On Tue, 29 Sep 2015 11:28:34 +0200 > Peter Zijlstra wrote: > > > With the introduction of the context switch preempt_count invariant, > > and the demise of PREEMPT_ACTIVE, its pointless to save/restore the > > per-cpu preemption co

Re: [RFC][PATCH 09/11] sched, x86: Kill saved_preempt_count

2015-09-29 Thread Steven Rostedt
On Tue, 29 Sep 2015 11:28:34 +0200 Peter Zijlstra wrote: > With the introduction of the context switch preempt_count invariant, > and the demise of PREEMPT_ACTIVE, its pointless to save/restore the > per-cpu preemption count, it must always be 2. > > Signed-off-by: Peter Zijlstra (Intel) Shoul

Re: [RFC][PATCH 09/11] sched, x86: Kill saved_preempt_count

2015-09-29 Thread Thomas Gleixner
On Tue, 29 Sep 2015, Peter Zijlstra wrote: > With the introduction of the context switch preempt_count invariant, > and the demise of PREEMPT_ACTIVE, its pointless to save/restore the > per-cpu preemption count, it must always be 2. Reviewed-by: Thomas Gleixner -- To unsubscribe from this list: