svn commit: r313533 - stable/10/usr.sbin/watchdogd

[Ian Lepore]

On Fri, 2017-02-10 at 06:53 +0000, Xin LI wrote:
> Author: delphij
> Date: Fri Feb 10 06:53:48 2017
> New Revision: 313533
> URL: https://svnweb.freebsd.org/changeset/base/313533
> 
> Log:
>   MFC r274583: Default to use 10 seconds as nap interval instead of
> 1.
> 
> Modified:
>   stable/10/usr.sbin/watchdogd/watchdogd.8
>   stable/10/usr.sbin/watchdogd/watchdogd.c
> 

I didn't see this change when it happened on -current, but my burning
question now is:  how will this work when the timeout is set to values
less than 10 seconds, which is the only option on some hardware?

-- Ian


> Modified: stable/10/usr.sbin/watchdogd/watchdogd.8
> =====================================================================
> =========
> --- stable/10/usr.sbin/watchdogd/watchdogd.8	Fri Feb 10
> 06:34:52 2017	(r313532)
> +++ stable/10/usr.sbin/watchdogd/watchdogd.8	Fri Feb 10
> 06:53:48 2017	(r313533)
> @@ -80,7 +80,7 @@ reboot if there are problems with the sc
>  The
>  .Fl s Ar sleep
>  argument can be used to control the sleep period between each
> execution
> -of the check and defaults to one second.
> +of the check and defaults to 10 seconds.
>  .Pp
>  The
>  .Fl t Ar timeout
> 
> Modified: stable/10/usr.sbin/watchdogd/watchdogd.c
> =====================================================================
> =========
> --- stable/10/usr.sbin/watchdogd/watchdogd.c	Fri Feb 10
> 06:34:52 2017	(r313532)
> +++ stable/10/usr.sbin/watchdogd/watchdogd.c	Fri Feb 10
> 06:53:48 2017	(r313533)
> @@ -80,7 +80,7 @@ static u_int timeout = WD_TO_128SEC;
>  static u_int exit_timeout = WD_TO_NEVER;
>  static u_int pretimeout = 0;
>  static u_int timeout_sec;
> -static u_int nap = 1;
> +static u_int nap = 10;
>  static int passive = 0;
>  static int is_daemon = 0;
>  static int is_dry_run = 0;  /* do not arm the watchdog, only
>