Re: [PATCH 24/32] softirq: Uninline !CONFIG_TRACE_IRQFLAGS __local_bh_disable_ip()

[Frederic Weisbecker]

On Wed, Feb 27, 2019 at 12:14:29PM +0100, Sebastian Andrzej Siewior wrote:
> On 2019-02-12 18:14:15 [+0100], Frederic Weisbecker wrote:
> > diff --git a/include/linux/bottom_half.h b/include/linux/bottom_half.h
> > index 240419382978..ef9e4c752f56 100644
> > --- a/include/linux/bottom_half.h
> > +++ b/include/linux/bottom_half.h
> > @@ -28,17 +28,7 @@ enum
> >  
> >  #define SOFTIRQ_DATA_INIT (SOFTIRQ_ALL_MASK << SOFTIRQ_ENABLED_SHIFT)
> >  
> > -
> > -
> > -#ifdef CONFIG_TRACE_IRQFLAGS
> >  extern void __local_bh_disable_ip(unsigned long ip, unsigned int cnt);
> > -#else
> > -static __always_inline void __local_bh_disable_ip(unsigned long ip, unsigned int cnt)
> > -{
> > -	preempt_count_add(cnt);
> > -	barrier();
> > -}
> > -#endif
> 
> you are aware that a simple local_bh_disable() / spin_lock_bh() gains a
> function call. Let me look further how much code you are going to add
> hereâ

Indeed but as you'll find out, the function is growing beyond reasonable inlining.