From mboxrd@z Thu Jan 1 00:00:00 1970 From: Stephen Hemminger Subject: Re: [RFC] RTT metrics scaling Date: Fri, 27 Jun 2008 22:30:07 -0700 Message-ID: <20080627223007.02eb6622@speedy> References: <20080620162234.1c2103ee@extreme> <20080627.195013.184418536.davem@davemloft.net> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: shemminger@vyatta.com, netdev@vger.kernel.org To: David Miller Return-path: Received: from mail.vyatta.com ([216.93.170.194]:59809 "EHLO mail.vyatta.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751071AbYF1FaU (ORCPT ); Sat, 28 Jun 2008 01:30:20 -0400 In-Reply-To: <20080627.195013.184418536.davem@davemloft.net> Sender: netdev-owner@vger.kernel.org List-ID: On Fri, 27 Jun 2008 19:50:13 -0700 (PDT) David Miller wrote: > From: Stephen Hemminger > Date: Fri, 20 Jun 2008 16:22:34 -0700 > > > Some of the metrics (RTT, RTTVAR and RTAX_RTO_MIN) are stored in kernel > > units (jiffies) and this leaks out through the netlink API to user space > > where the units for jiffies are unknown. > > > > This patches changes the kernel to convert to/from milliseconds. This changes > > the ABI, but milliseconds seemed like the most natural unit for these parameters. > > > > This patch is for comment only, just to show how it would be implemented. > > I will update it after testing more. > > No real objection. > > But aren't we approaching microsecond RTT's soon? :-) (yes, I know in > TCP we min clamp at 200ms RTT, but that clamp we probably want to > remove soon) We could put in microseconds, but since it is computer off of jiffies (ie clock ticks), the resolution wouldn't be worth it. > If you don't think that's worth worrying about, I'll apply this patch. Hold off it needs more testing.