Re: timevar: 3/3: use clock_gettime to get wall clock time

[Bruno Haible]

Hi Akim,

> > I don’t know how 'gettime' compares against 'gethrxtime'.
> 
> gethrxtime is exactly what I need, thanks!  It does use gettime as
> a fallback, after having tried better (accuracy and monotony) options.

Very nice!

> Here is my updated proposal.

Looks good, except for the mentioned points 2) - a 'float' with its 24 bits
of precision is surely not enough to hold an 'xtime_t' - and 3).

> > 3) Link dependency: List the link dependencies, like it's done e.g. in
> > the 'clock-time', 'gettime', and 'gethrxtime' modules. Also use these
> > dependencies when linking the test program (in module ‘timevar-tests').
> 
> AFAICT, there are no direct link dependencies in timevar, only indirect
> ones via the modules it depends upon.  What do you mean?

The 'Link' and 'Include' fields in a Gnulib module description are not
evaluated by gnulib-tool. They are purely informative, for the user of the
module.

> there are no direct link dependencies in timevar, only indirect
> ones via the modules it depends upon.  What do you mean?

I mean that in this situation, the user of the module does not want to
perform a recursive search across the dependencies of the module, to
discover the link dependencies. That's your job as author of the module.

If a module 'timevar-tests' existed, its Makefile.am snippet would need to
use the expression listed under 'Link', to link the 'test-timevar' program.

Bruno