> I don't know, cpu cycles are cpu cycles...they're good for anything aren't > they? If you reduce per-frame-code-load then that frees time up for other tasks > like disk io. Or am I confused about this?
You are confused about that. Most modern processors are memory bandwidth limited. That's what killed RISC. If you inline code, you save on the frame setup, but it costs memory bandwidth. So, you have to ask whether there will be more idled instructions, while waiting for the processor instruction cache to be filled with inline code, or instructions that are constructively doing the frame thing out of cache (because the subroutine itself is probably in cache). _______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel
