[GRASS-dev] Parallel computing for r.sun

[Hamish]

Jan wrote:
> If there is r.los also available for parallel computing we
> will try v.surf.bspline

re. processing processing each of the bspline subregions in
parallel, see this point in the code:

https://trac.osgeo.org/grass/browser/grass/trunk/vector/v.surf.bspline/main.c#L589


note deep openmp pragmas are already in place in the gmath
library calls from that loop, but there's an openmp call to
tell sub functions not to re-parallelize something that is
already broken up into chunks. (or you can just comment out
the ones in the gmath library if nothing else is using them;
they're quite inefficient so far inside the loop)


Hamish