From mboxrd@z Thu Jan 1 00:00:00 1970 Return-path: Received: from bh-25.webhostbox.net ([208.91.199.152]:55952 "EHLO bh-25.webhostbox.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752348AbbERNPV (ORCPT ); Mon, 18 May 2015 09:15:21 -0400 Message-ID: <5559E5E6.2000508@roeck-us.net> Date: Mon, 18 May 2015 06:15:18 -0700 From: Guenter Roeck MIME-Version: 1.0 To: Nick & John , linux-watchdog@vger.kernel.org Subject: Re: [PATCH] watchdog: via_wdt: Partial fix for programming error and support additional southbridges References: In-Reply-To: Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-watchdog-owner@vger.kernel.org List-Id: linux-watchdog@vger.kernel.org On 05/18/2015 03:50 AM, Nick & John wrote: > This patch fixes a programming error by limiting WDT_TIMEOUT_MAX to 255. > That is done to avoid invalid settings. Putting a value bigger than 255 > results in non working watchdog and reset is not performed. We are trying > to fit a 10bit value to an 8bit area! To support 1023 seconds MMIO addr+5 > bit1:0 must be used to store the last(higher) two bits of count number. > Strangely no error was logged at the system when using values bigger than > 255, the driver just ignores the time count. > While I have no idea why, the watchdog in this driver is pinged using a soft timer, and it is always pinged every 0.5 seconds. So reducing the maximum timeout is not necessary; in fact the driver could support a higher timeout. The real fix, if it is really necessary, would be to either store the upper bits as suggested above, or to just program the chip to a timeout of 255 seconds if the desired timeout is larger than 255. > Also it adds support for VT8235 and VT8237 series southbridges. Tested with > VT8235M, VT8237 and VT8237S boards. > This should be a separate patch; see Documentation/SubmittingPatches, chapter 3. Thanks, Guenter