Date: Thu, 29 Oct 2009 10:09:57 +0100
From: Ingo Molnar <mingo@elte.hu>
To: Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
Cc: Roland Dreier <rdreier@cisco.com>, linux-kernel@vger.kernel.org,
       x86@kernel.org, Andi Kleen <ak@linux.intel.com>,
       Thomas Gleixner <tglx@linutronix.de>, Ingo Molnar <mingo@redhat.com>,
       "H. Peter Anvin" <hpa@zytor.com>, Mike Travis <travis@sgi.com>
Subject: Re: [PATCH] x86, mce: short output of MCE banks ownership
	information
Message-ID: <20091029090957.GD22963@elte.hu>
References: <4AE90E43.2030709@jp.fujitsu.com> <adaaaza6exw.fsf@cisco.com> <4AE93B42.5040605@jp.fujitsu.com> <20091029073408.GA18976@elte.hu> <4AE951F4.7080406@jp.fujitsu.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <4AE951F4.7080406@jp.fujitsu.com>
User-Agent: Mutt/1.5.19 (2009-01-05)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1994
Lines: 43


* Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com> wrote:

> Ingo Molnar wrote:
> > * Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com> wrote:
> >> Maybe there would be more desirable ways, but I think that "compress 
> >> messages shorter to bear heavy repeating" will be a good way at this 
> >> time.
> > 
> > We really want to only print out relevant information. 128 lines of 
> > identical output is not relevant. (or 2x 64 lines, or 4096 lines of 
> > identical output)
> > 
> > So we only want to print MCE setup messages on the boot CPU. That gives 
> > us 90% of the benefits already: we see the rough structure of the 
> > hardware, and if the bootup has a problem with MCE initialization we get 
> > relevant printouts that helps debugging.
> > 
> > Now, it's certainly true that with things like MCE bank sharing the MCE 
> > setup output from different CPUs might not be identical all the time - 
> > but the information is represented in other (topology) info anyway. (and 
> > if not it wasnt all that important to begin with)
> > 
> > For non-boot CPUs we can perhaps add a mce=verbose (default-disabled) 
> > mode of bootup that allows all CPUs to be printed - should there be any 
> > problem with MCE details only visible on non-boot CPUs. (unlikely)
> 
> How about having a kind of "boot=quiet_ap" which suppress boot-time 
> messages for all non-boot CPUs, rather than "mce=verbose" which only 
> suppress one (or, with thermal message, two) line?  I don't not 
> understand why this problem would be solved only by removing the lines 
> from mce subsystem.

The idea is to suppress bootup messages by default - and to allow opt-in 
verbose printing, in the (rare) case of having to debug bootup problems 
that involve APs.

	Ingo
--
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/