On Sun, May 14, 2017 at 12:33:36AM +0200, Ævar Arnfjörð Bjarmason wrote: > The tip of that branch are a bunch of "TODO perldelta: *" commits > where I don't know what should be said, but think that something > should be clarified/fixed by others who actually know what these > entries are referring to. I've made some further changes to your TODO bits and some of the previous fixes for them with v5.26.0-RC1-49-g8928e7d > Several other ops may now also be faster in boolean context. > > + Could we be a bit less vague here? Arithmetic ops? stuff like > + print, what? I've removed that sentence because it's essentially wrong. > Speed up C<AV> and C<HV> clearing/undeffing. > > + Does this mean that `undef %h` and `undef @a` are faster, or is it > + for some reason only the internal XS stuff? If it's the former > + (which I assume) let's use the more readable forms someone not > + deep into XS will know about. I've removed your fix for this because that relatively small speedup was dwarved by a later change to list assignment that was getting 33% speedups. Instead I've expanded the following section: > =item * > > Better optimise array and hash assignment -- Indomitable in retreat, invincible in advance, insufferable in victory -- Churchill on MontgomeryThread Previous | Thread Next