Date: Thu, 18 Mar 2004 20:16:42 +0100
Message-ID: <s5hbrmuc6ed.wl@alsa2.suse.de>
From: Takashi Iwai <tiwai@suse.de>
To: Andrew Morton <akpm@osdl.org>
Cc: andrea@suse.de, mjy@geizhals.at, linux-kernel@vger.kernel.org
Subject: Re: CONFIG_PREEMPT and server workloads
In-Reply-To: <20040318110159.321754d8.akpm@osdl.org>
References: <40591EC1.1060204@geizhals.at>
	<20040318060358.GC29530@dualathlon.random>
	<s5hlllycgz3.wl@alsa2.suse.de>
	<20040318110159.321754d8.akpm@osdl.org>
User-Agent: Wanderlust/2.10.1 (Watching The Wheels) SEMI/1.14.5 (Awara-Onsen) FLIM/1.14.5 (Demachiyanagi) APEL/10.6 MULE XEmacs/21.4 (patch 13) (Rational FORTRAN) (i386-suse-linux)
MIME-Version: 1.0 (generated by SEMI 1.14.5 - "Awara-Onsen")
Content-Type: text/plain; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 2242
Lines: 63

At Thu, 18 Mar 2004 11:01:59 -0800,
Andrew Morton wrote:
> 
> Takashi Iwai <tiwai@suse.de> wrote:
> >
> > for ext3, these two spots are relevant.
> > 
> > --- linux-2.6.4-8/fs/jbd/commit.c-dist	2004-03-16 23:00:40.000000000 +0100
> > +++ linux-2.6.4-8/fs/jbd/commit.c	2004-03-18 02:42:41.043448624 +0100
> > @@ -290,6 +290,9 @@ write_out_data_locked:
> >  			commit_transaction->t_sync_datalist = jh;
> >  			break;
> >  		}
> > +
> > +		if (need_resched())
> > +			break;
> >  	} while (jh != last_jh);
> >  
> >  	if (bufs || need_resched()) {
> 
> This one I need to think about.  Perhaps we can remove the yield point a
> few lines above.

yes, i'm afraid that it's also overkill to check this at every time.
perhaps we can optimize it a bit better.  the fact that it imporives
the latency means that there are so many locked buffers or non-dirty
buffers in the list?

> One needs to be really careful with the lock-dropping trick - there are
> weird situations in which the kernel fails to make any forward progress. 
> I've been meaning to do another round of latency tuneups for ages, so I'll
> check this one out, thanks.
> 
> There's also the SMP problem: this CPU could be spinning on a lock with
> need_resched() true, but the other CPU is hanging on the lock for ages
> because its need_resched() is false.

yep, i see a similar problem also in reiserfs's do_journal_end().
it's in lock_kernel().

>  In the 2.4 ll patch I solved that via
> the scary hack of broadcasting a reschedule instruction to all CPUs if an
> rt-prio task just became runnable.  In 2.6-preempt we use
> preempt_spin_lock().
> 
> But in 2.6 non-preempt we have no solution to this, so worst-case
> scheduling latencies on 2.6 SMP CONFIG_PREEMPT=n are high.

hmm, it's bad...

> Last time I looked the worst-case latency is in fact over in the ext3
> checkpoint code.  It's under spinlock and tricky to fix.

BTW, i had the worst latency in sis900's timer handler.
it takes 3ms, and hard to fix, too :-<


Takashi
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/