A bloke by the name of Jerry Breecher wrote something up on this back in the
late eighties in one of the IEEE journals. I have the article floating about
in paper form. Basically he just used standard old flat profiling data to
reorganise the linking order, compared that to a number of other approaches
(including profiling in the temporal domain) and discovered that it reduced
the resident set size quite nicely for a lot of programs, with the associated
benefits of less paging and possibly better cache locality. Profiler directed
feedback has been around for ages in the MIPS camp (coord & pixie).
Stephen
-- The views expressed above are not those of PGS Tensor."We've heard that a million monkeys at a million keyboards could produce the Complete Works of Shakespeare; now, thanks to the Internet, we know this is not true." Robert Wilensky, University of California
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/