On Thursday, December 13, 2012 12:51:57 AM UTC+11, Marshall Bockrath-Vandegrift wrote: > > cameron <cdo...@gmail.com <javascript:>> writes: > > > the megamorphic call site hypothesis does sound plausible but I'm > > not sure where the following test fits in. > > ... > > > I was toying with the idea of replacing the EmptyList class with a > > PersistsentList instance to mitigate the problem > > in at least one common case, however it doesn't seem to help. > > If I replace the reverse call in burn with the following code: > > #(reduce conj (list nil) %) > > I get the same slowdown as we see if reverse (equivalent to #(reduce > > conj '() %)) > > Ah, but include your own copy of `conj` and try those two cases. The > existing clojure.core/conj has already been used on multiple types, so > you need a new IFn class with a fresh call site. Here are the numbers I > get when I do that: > Ah, I've also been looking at this in the morning and missed that bit. When I use a copy of conj I get similar results to yourself: w/ list-base : map-ms: 4.0, pmap-ms 0.6, speedup 7.02 w/ empty-list-base : map-ms: 9.9, pmap-ms 20.4, speedup 0.48 w/ list-base : map-ms: 9.4, pmap-ms 20.7, speedup 0.45 w/ vector : map-ms: 4.5, pmap-ms 1.4, speedup 3.28
It does seem that once used on the empty list case conj gets tainted and all future list uses incur a performance penalty. On that basis it would seem reasonable to convert the EmptyList uses in core.lang to a PersistentList instance. Interestingly using a tainted conj with other types doesn't seem to incur the penalty (the last vector in the times above) If I get time I might look at a protocol based version of conj out of interest. Cameron. -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
