Ok, I'm abandoning this revision. I'm going to rename the functions so the names better suit the specialized functionality and implement different versions for PMR. There is clearly a lot of interest in changes to the power of 2 / prime resizing of `__hash_table` but I think a better time to handle this all will be when I'm writing libc++ performance tests. As Howard mentioned the performance of this code is critical and I don't want to make any changes without doing the proper performance testing legwork.
Thank's for all the input it has been invaluable. http://reviews.llvm.org/D4948 _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
