-
Gabor Melis authored
It's just micro-optimizations: a couple of inline declarations, and a quicker test for group size 1. On CLL and Clisp, also on x86-64, it's about 2% and 5% faster, respectively.
8fe6e63d
It's just micro-optimizations: a couple of inline declarations, and a quicker test for group size 1. On CLL and Clisp, also on x86-64, it's about 2% and 5% faster, respectively.