Re: IRQ off latency of printk is very high

[Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>]

* Pavel Machek (pavel@ucw.cz) wrote:
> Hi!
> 
> > > > It seems draconian to drain the entire buffer with ints disabled.
> > > > Is it possible to break this up and send out smaller chunks
> > > > at a time?  Maybe by putting a chunk loop in release_console_sem()?
> > > 
> > > Well, I believe someone got
> > > 
> > > DDetetccctted ed 113223 HHzz CPUCPU
> > > 
> > > in his dmesg, and now we have this 'draconian' locking. How can we
> > > prevent mangled messages without it?
> > 
> > The main interest seems to be to protect from mixed printk output
> > between different CPUs in process context. I don't think it would be
> > that bad if interrupts come and output error messages in the middle of a
> > printk, isn't it ?
> > 
> > therefore, could we do something like :
> > 
> > 
> > if (!in_irq())
> >   spin_lock(&logbuf_lock);
> > ...
> > if (!in_irq())
> >   spin_unlock(&logbuf_lock);
> > 
> > ? (yes, this is a crazy idea)
> 
> Two messages in atomic sections on different cpus could still be mixed
> :-). But yes, something like this may be the way to go.

Not in "preempt disable" sections though. Only in interrupt handlers.
But yes, I assume here that messages coming from interrupt handlers can
afford being interleaved.

Mathieu

> 									Pavel
> -- 
> (english) http://www.livejournal.com/~pavelmachek
> (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

-- 
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68