This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [autovect][patch]Loop Versioning for Vectorization


Hello,

> I hope to have a patch to preserve loop-closed-form during our loop-peeling
> within the vectorizer, because we probably don't want to call
> rewrite_into_loop_closed_ssa each time we vectorize a loop (using peeling).

that's almost surely the case.  rewrite_into_loop_closed_ssa is linear
in size of the function, so using it after each loop may cause quadratic
behavior.  It would be possible to make rewrite_into_loop_closed_ssa
more efficient (in fact patches for this already exist), but
optimizations that know about loops should not have much problems with
preserving lcssa.

Zdenek


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]