Date: Fri, 14 Oct 2005 06:56:15 +0200
From: Ingo Molnar <mingo@elte.hu>
To: Fernando Lopez-Lezcano <nando@ccrma.Stanford.EDU>
Cc: linux-kernel@vger.kernel.org, Thomas Gleixner <tglx@linutronix.de>,
       Steven Rostedt <rostedt@goodmis.org>, dwalker@mvista.com,
       david singleton <dsingleton@mvista.com>
Subject: Re: 2.6.14-rc4-rt1
Message-ID: <20051014045615.GC13595@elte.hu>
References: <20051011111454.GA15504@elte.hu> <1129064151.5324.6.camel@cmn3.stanford.edu> <20051012061455.GA16586@elte.hu> <20051012071037.GA19018@elte.hu> <1129242595.4623.14.camel@cmn3.stanford.edu> <1129256936.11036.4.camel@cmn3.stanford.edu>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1129256936.11036.4.camel@cmn3.stanford.edu>
User-Agent: Mutt/1.4.2.1i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 927
Lines: 37


* Fernando Lopez-Lezcano <nando@ccrma.Stanford.EDU> wrote:

> #!/bin/bash
> while true ; do
>     START=`date +"%s"`
>     sleep 10
>     END=`date +"%s"`
>     let DIFF=END-START
>     echo "$DIFF" >>time
>     echo "---"
> done
> 
> I'm attaching what I found when I got back.

> 1
> 10
> 6
> 0
> 0

could you try:

	strace -o log sleep 10

and wait for a failure, and send us the log? Is it perhaps nanosleep 
unexpectedly returning with -EAGAIN or -512? There's a transient 
nanosleep failure that happens on really fast boxes, which we havent 
gotten to the bottom yet. That problem is very sporadic, but maybe your 
box is just too fast and triggers it more likely :-)

	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/