Yes, of course. Your benchmark is even more appropriate to show the overhead, which consists in creating, populating and destroying a hash (call with hashref) vs. populating a hash already allocated on the subroutine's pad (call with list).
Might be small, but might sum up. Call it a micro-de-optimization, but I see no value in making perl slower for a questionable gain and a loss of readability.
In reply to Re^4: Preferred technique for named subroutine parameters?
by shmem
in thread Preferred technique for named subroutine parameters?
by Anonymous Monk
| For: | Use: | ||
| & | & | ||
| < | < | ||
| > | > | ||
| [ | [ | ||
| ] | ] |