I think the original i386 did it with a one-bit-per-cycle algorithm,
anything since should be fine. In particular, on a P4 where I just tested,
the bsf seems to be 4 cycles over the whole input set (actually, my whole
loop was 4 cycles per iteration, so 4 cycles is worst-case. I'm assuming
the rest could have been done in parallell).
Linus
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/