Re: [PATCH] don't use flush_tlb_all in suspend time

[Pavel Machek]

On Ät 18-05-06 16:38:21, Shaohua Li wrote:
> On Thu, 2006-05-18 at 10:31 +0200, Pavel Machek wrote:
> > Hi!
> > 
> > > > > In which case, how's about this?
> > > > 
> > > > Certainly better, I'd say.
> > > > 
> > > > > @@ -420,7 +421,14 @@ void zap_low_mappings (void)
> > > > >  #else
> > > > >  		set_pgd(swapper_pg_dir+i, __pgd(0));
> > > > >  #endif
> > > > > -	if (cpus_weight(cpu_online_map) == 1)
> > > > > +	/*
> > > > > +	 * We can be called at suspend/resume time, with local interrupts
> > > > > +	 * disabled.  But flush_tlb_all() requires that local interrupts be
> > > > > +	 * enabled.
> > > > > +	 *
> > > > > +	 * Happily, the APs are not yet started, so we can use local_flush_tlb()	 * in that case
> > > > > +	 */
> > > > > +	if (num_online_cpus() == 1)
> > > > >  		local_flush_tlb();
> > > > >  	else
> > > > >  		flush_tlb_all();
> > > > 
> > > > But this still scares. It means calling convention is "may enable
> > > > interrupts with >1 cpu, may not with == 1 cpu". 
> > > Below patch should make things clean. How do you think?
> > 
> > Nice...
> > 
> > Could we perhaps reuse swsusp_pg_dir (just make it used for swsusp &
> > suspend-to-ram) to save a bit more code? It is in arch/i386/mm/init.c
> Sure. But it's under CONFIG_SOFTWARE_SUSPEND. 

Yes, #ifdef definitely does need to change there.

> That part needs cleanup I
> think and it's a little strange to me (why should we simply copy
> swapper_pg_dir to swsusp_pg_dir, instead do it in zap_low_mappings?).

?? Sorry, I do not understand. "instead of doing"?
								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/