Re: [PATCH v3] printk: Add console owner and waiter logic to load balance console writes

[Sergey Senozhatsky]

On (11/08/17 22:29), Steven Rostedt wrote:
> > On (11/08/17 09:29), Steven Rostedt wrote:
> > > On Wed, 8 Nov 2017 14:19:55 +0900
> > > Sergey Senozhatsky <sergey.senozhatsky.work@xxxxxxxxx> wrote:
> > >   
> > > > the change goes further. I did express some of my concerns during the KS,
> > > > I'll just bring them to the list.
> > > > 
> > > > 
> > > > we now always shift printing from a save - scheduleable - context to
> > > > a potentially unsafe one - atomic. by example:  
> > > 
> > > And vice versa. We are now likely to go from a unscheduleable context
> > > to a schedule one, where before, that didn't exist.  
> > 
> > the existence of "and vice versa" is kinda alarming, isn't it? it's sort
> > of "yes, we can break some things, but we also can improve some things."
> 
> Not really. Because the heuristic is that what calls printk will do the
> printk.

so what we are looking at

   a) we take over printing. can be from safe context to unsafe context
      [well, bad karma]. can be from unsafe context to a safe one. or from
      safe context to another safe context... or from one unsafe context to
      another unsafe context [bad karma again]. we really never know, no
      one does.

      lots of uncertainties - "may be X, may be Y, may be Z". a bigger
      picture: we still can have the same lockup scenarios as we do
      have today.

      and we also bring busy loop with us, so the new console_sem
      owner [regardless its current context] CPU must wait until the
      current console_sem finishes its call_console_drivers(). I
      mentioned it in my another email, you seemed to jump over that
      part. was it irrelevant or wrong?

vs.

   b) we offload to printk_kthread [safe context].


why (a) is better than (b)?

	-ss