Julian Seward be18f74393 findSb: gradually rearrange the superblock list to bring frequently
accessed blocks closer to the front.  This speeds up malloc/free
intensive programs because evidently those searches cause a lot of
cache misses (so cachegrind tells us).  For perf/heap.c on P4
Northwood, this halves the run-time (!) from 85.8 to 42.9 seconds.
For "real" code (start/exit ktuberling) there is a small but
worthwhile performance gain, of about 2 seconds out of 95.



git-svn-id: svn://svn.valgrind.org/valgrind/trunk@5365
2005-12-17 20:37:36 +00:00
..
2005-11-11 13:46:52 +00:00
2005-12-08 03:24:37 +00:00
2005-08-25 22:55:19 +00:00
2005-11-10 15:03:26 +00:00
2005-07-08 18:23:40 +00:00
2005-08-14 22:14:55 +00:00
2005-12-09 19:30:02 +00:00
2005-08-25 22:55:19 +00:00