LGTM! On Jun 19, 2014, at 9:55 AM, Mark Heffernan <[email protected]> wrote:
> Thanks for the comments. Tyler, I incorporated your suggested rewrite of the > unrolling paragraph. Hopefully it is clearer now. > > I also split the vectorization/interleaving pragmas and the unrolling pragmas > into separate sections. > > Please take another look. > > ================ > Comment at: docs/LanguageExtensions.rst:1791 > @@ +1790,3 @@ > +Loop unrolling replicates the body of a loop increasing the loop size > +and reducing the loop count. Loop control overhead can be reduced or > +eliminated, and additional ILP can be exposed. The unroller selects > ---------------- > Eli Bendersky wrote: >> Maybe s/loop count/loop trip count/ for clarify? There's another instance >> later on > Rewrote this bit. > > ================ > Comment at: docs/LanguageExtensions.rst:1793 > @@ +1792,3 @@ > +eliminated, and additional ILP can be exposed. The unroller selects > +an unroll count based on a limit on the growth of code size and > +whether the loop can be unrolled completely. > ---------------- > Eli Bendersky wrote: >> I find this sentence a bit confusing in the context of unroll pragmas which >> tell the unroller exactly how much we want to unroll. OTOH this sentence >> seems to be describing the unrolling heuristic? > Rewrote this bit. > > http://reviews.llvm.org/D4198 > > _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
