Re: [gfortran] Implementation SYSTEM_CLOCK intrinsic subroutine

[Tobias Schlüter <tobias dot schlueter at physik dot uni-muenchen dot de>]

Helge Avlesen wrote:
> Hi,
> great that you look at this routine - a possible enhancement could be
> the non portable assembly snippets
>
> http://www.ncsa.uiuc.edu/UserInfo/Resources/Hardware/IA32LinuxCluster/Doc/timing.html#linuxasm
>
> for linux. similar hacks excist for most systems, and the routine
> could default to some unix function if not. this could give roughly
> nano second resolution on gigaherz machines for those who wants to
> time e.g. single loops. in the ultimate implementation the user could
> adjust the resolution of the clock via an environment variable to
> avoid overwrapping for longer measurements.

This is intriguing, maybe you could open a PR, so that we remember this? 
I don't think it should be very high on our priority list right now, but 
it seems to be interesting enough to incorporate this when the compiler 
has become decently stable.

- Tobi