From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andy Shevchenko Date: Fri, 07 Jul 2017 14:15:53 +0300 Subject: [U-Boot] [PATCH v1 2/3] wdt: Timeout better to be in microseconds In-Reply-To: References: <20170705174408.72891-1-andriy.shevchenko@linux.intel.com> <20170705174408.72891-2-andriy.shevchenko@linux.intel.com> Message-ID: <1499426153.22624.299.camel@linux.intel.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit To: u-boot@lists.denx.de On Thu, 2017-07-06 at 21:59 -0600, Simon Glass wrote: > Hi Andy, > > On 5 July 2017 at 11:44, Andy Shevchenko > wrote: > > Timeout in some abstract ticks is not what we are applying to get > > deterministic behaviour. > > A tick is always milliseconds in U-Boot, as I understand it. You see, there is a confusion. I would like to see units somewhere there, to make it clear. > > > > > Convert name to show explicitly that we are using microseconds (for > > watchdog it's more than precise). > > Do you want microseconds, or is milliseconds enough accuracy? I have a > hard time imagining a case where a microsecond watchdog timeout is > useful. For me ticks sounded like processor cycles (nanosecond-ish). So, milliseconds are better. Since I have not much time, feel free to drop, modify, etc. Thanks for review. > > > > > No functional change intended. > > > > Signed-off-by: Andy Shevchenko > > --- > >  drivers/watchdog/wdt-uclass.c | 4 ++-- > >  include/wdt.h                 | 8 ++++---- > >  2 files changed, 6 insertions(+), 6 deletions(-) > > > > Regards, > Simon -- Andy Shevchenko Intel Finland Oy