Subject: Re: RCU latency regression in 2.6.16-rc1
From: Lee Revell <rlrevell@joe-job.com>
To: Ingo Molnar <mingo@elte.hu>
Cc: Eric Dumazet <dada1@cosmosbay.com>, dipankar@in.ibm.com,
       paulmck@us.ibm.com, linux-kernel <linux-kernel@vger.kernel.org>,
       Linus Torvalds <torvalds@osdl.org>
In-Reply-To: <20060129075144.GA15056@elte.hu>
References: <20060124213802.GC7139@in.ibm.com>
	 <1138224506.3087.22.camel@mindpipe> <20060126191809.GC6182@us.ibm.com>
	 <1138388123.3131.26.camel@mindpipe> <20060128170302.GB5633@in.ibm.com>
	 <1138471203.2799.13.camel@mindpipe> <1138474283.2799.24.camel@mindpipe>
	 <20060128193412.GH5633@in.ibm.com> <43DBCB62.7030308@cosmosbay.com>
	 <1138520283.2799.103.camel@mindpipe>  <20060129075144.GA15056@elte.hu>
Content-Type: text/plain
Date: Sun, 29 Jan 2006 03:21:41 -0500
Message-Id: <1138522902.2799.111.camel@mindpipe>
Mime-Version: 1.0
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 783
Lines: 20

On Sun, 2006-01-29 at 08:51 +0100, Ingo Molnar wrote:
> 
> well, softirq preemption is not really a drastic step - its biggest 
> problem is that it cannot be included in v2.6.16 ;-) But i agree that
> if a solution can be found to break up a latency path, that is
> preferred.
> 

Agreed.  It's only drastic in the context of my target (meeting a 1ms
soft RT constraint) as this happens to be the one of the only code paths
getting in the way.  Also I'd like to be able to go down to 1ms without
requiring a custom kernel config...

Lee

-
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/