timespec wrapper

[Joel Sherrill]

On 08/25/2012 05:41 AM, Krzysztof Mięsowicz wrote:
> Hi,
>
> I have written wrapper for directives useful in manipulating 
> timespecs. Please, review attached patch and comment it.
>
Let's go completely to lower case on the names and make them all inline.
These are methods and I suspect they may all end up with real bodies as
error checking gets added.

The Score routines they are based on do not validate the parameters in 
anyway
because they are running in a "trusted" environment.

Other comments:

+ Do not use multiple blank lines in a row
+ Update the copyright notice to 2012 and your name.

There needs to be a chapter in the User's Guide (doc/user) and
there is a simple shell script in doc/new_chapters which can
be hacked to generate the starting point.

> FWIW I did it because it wasn't directly available in RTEMS API and 
> when I run coverage analysis for i386/pc386 it seems that this 
> directives is almost unused and their availability in API can be 
> useful for someone.
>
Just so this is publicly recorded. He did it at my request. struct 
timespec is
POSIX and doing math on it is common.  These are useful as helpers to 
applications
and much easier to completely test when they are exposed.
> Krzysztof