The attached patch implements delinearization of array accesses in the 
Fortran front end, something that has been discussed for a long time.

I've been asked to try to get this patch committed on the OG10 branch 
since it is blocking some further optimization work with Graphite for 
OpenACC kernels regions.  I have a mainline version of this patch as 
well that I can send to anyone interested in trying it out, but TBH, I 
don't think this is ready for mainline yet.  The current status is that 
there are still two gfortran tests that are regressing 
(gfortran.dg/graphite/id-9.f and 
gfortran.dg/vect/fast-math-mgrid-resid.f), and while it's been confirmed 
that this helps with Graphite optimizations as intended, we haven't yet 
run any benchmarks to confirm that it doesn't make other things slower. 
(It might, for instance, be appropriate to only delinearize when 
Graphite optimizations are also enabled.)

Anyway, I thought that for the OG10 branch I ought to also put this out 
for review/objection instead of just pushing it immediately.  Unless 
somebody explicitly approves the patch meanwhile, I will probably hold 
off until after the first of the year before pushing it, since many of 
us are going to be away until then anyway.

-Sandra