On Thu, Sep 26, 2013 at 5:18 PM, Robert Haas <robertmhaas@gmail.com> wrote: > On Thu, Sep 26, 2013 at 3:55 PM, David Rowley <dgrowleyml@gmail.com> wrote: >> I think I must have forgot to save it before I emailed it... >> >> Here's the correct version. > > Ah ha. Looks better. > > I'm working on getting this committed now. Aside from some stylistic > things, I've found one serious functional bug, which is that you need > to test padding != 0, not padding > 0, when deciding which path to > take. No need to send a new patch, I've already fixed it in my local > copy...
Oops, negative padding. My bad. I was focusing too much on the benchmarks I think.