From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from bh-25.webhostbox.net ([208.91.199.152]:49914 "EHLO bh-25.webhostbox.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754152AbbDMNxF (ORCPT ); Mon, 13 Apr 2015 09:53:05 -0400 Received: from mailnull by bh-25.webhostbox.net with sa-checked (Exim 4.82) (envelope-from ) id 1YhenQ-00282r-1O for linux-watchdog@vger.kernel.org; Mon, 13 Apr 2015 13:53:04 +0000 Message-ID: <552BCA3E.5010305@roeck-us.net> Date: Mon, 13 Apr 2015 06:53:02 -0700 From: Guenter Roeck MIME-Version: 1.0 To: Alexander Stein , linux-watchdog@vger.kernel.org Subject: Re: watchdog timeout in sub-seconds? References: <3042537.C3OUI66IzY@ws-stein> In-Reply-To: <3042537.C3OUI66IzY@ws-stein> Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-watchdog-owner@vger.kernel.org List-Id: linux-watchdog@vger.kernel.org On 04/13/2015 06:33 AM, Alexander Stein wrote: > Hi, > > is there a specific reason that the timeout in the linux watchdog drivers is > only provided in seconds resolution? What about timeouts in ms area/precision? > The question would probably be why not. On a loaded system, sub-second watchdog timeouts are quite unreasonable. In other words, you would have to provide a very good argument for sub-second timeouts if you see the need for it. That some hardware may require it is not really an argument here - this is all about user space - kernel interface, not about limitations of specific hardware. Guenter