On Fri, Aug 17, 2012 at 2:21 PM, Jakub Jelinek <ja...@redhat.com> wrote: > On Fri, Aug 17, 2012 at 02:16:12PM +0200, Richard Guenther wrote: >> Or figure out if or why not the vectorizer does catch this (of course we do >> not enable that with -O2 which we eventually should in a very conservative >> mode). > > It might be helpful if we for the BITMAP_ELEMENT_WORDS == 5 case reordered > indx field after bits, so that bits array is 64-bit aligned.
That's an excellent suggestion! With that change on top of my 5-int bitmap patch, I actually get a decent speed-up for this test case (5%, mostly due to less slow DF). I'm trying this idea with an otherwise unpatched compiler now. Ciao! Steven
