The cause of the initial blowup was adding the vector test suite to the
normal cabal file, in both 0O and O2 forms. This was done so that fusion
rules based bugs wouldn't lie in hiding for years.
It sounds like the issues that resulted in built blowup are fixed in 8.0.
My guess is it's a combination of some type class coercion / equality
manipulation getting better plus other improvements in 8.0, both.
On Tuesday, April 5, 2016, Dominic Steinitz
Ryan Newton
writes: Hi all, Is anyone currently working in, or interested in helping with, a new benchmark suite for Haskell? Perhaps, packaging up existing apps and app benchmarks into a new benchmark suite that gives a broad picture of Haskell application performance today?
I am very interested. I recently encountered a serious performance regression from 7.8 to 7.10 which seems to be fixed in 8.0. Now it's not clear whether this was a library change or ghc itself. I suspect the latter but given the performance is better in 8.0, I was not motivated to confirm this.
I am happy to wrap up my example in whatever format but note that it does pull in quite a few libraries.
Dominic. _______________________________________________ ghc-devs mailing list ghc-devs@haskell.org javascript:; http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs