[PPL-devel] PPL and polyhedra comparisons
Sebastian Pop
sebpop at gmail.com
Mon Aug 25 21:40:20 CEST 2008
On Mon, Aug 25, 2008 at 3:01 AM, Enea Zaffanella <zaffanella at cs.unipr.it> wrote:
> This functionality is indeed ad-hoc for your purposes:
> it assumes some very specific meaning for the space dimensions,
Right. We can just have this sorting function in the cloog-ppl
back-end. I don't understand very well the implications of replacing
this ordering with another one for the generated code, but I did
experimented with some unusual sortings and the code generated did not
looked the same...
> On the other hand, having worked on it, I can fairly say that such an
> inclusion in the PPL would not buy much in terms of efficiency:
Thanks for the detailed improvement descriptions. I will try to get
these implemented.
Sebastian
More information about the PPL-devel
mailing list