Ok, I did some preliminary testing on i386 (not at home with amd64 system, sorry). The code I proposed is on average 15% faster. The latency saved isn't very much except when 1M or more of these operations are performed. The total time saved is about 6 milliseconds per million calls. Are these the numbers you are talking about?
Patrick Baggett, Figgle Software This message posted from opensolaris.org _______________________________________________ opensolaris-code mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/opensolaris-code
