java - Why would jmap -permstat report more than MaxPermSize bytes used? -

i've taken jmap -permstat output of jvm (reported version 24.76-b04), reports following totals:

total = 5190 76930 1031431696 n/a alive=1, dead=5189 n/a

which should number of entries, classes using information, bytes used metadata , liveness information.

now, wonder why report 1031431696 bytes, shy of gigabyte when startup vm -xx:maxpermsize=256m. can shed light on how number calculated?

don't know if it's relevant using rhino ~3k entries being definingclassloaders.

i have looked bit @ code jmap -permstat implementation. bytes used value estimation based on sizes different types of data classloader loads (methods, fields, interfaces etc).

the top level method size calculation loaded class sun.jvm.hotspot.tools.permstat.computesize:

private long computesize(instanceklass k) {   long size = 0l;    // instanceklass object   size += k.getobjectsize();    // constant pool   constantpool cp = k.getconstants();   size += cp.getobjectsize();   size += objectsize(cp.getcache());   size += objectsize(cp.gettags());    // interfaces   size += arraysize(k.getlocalinterfaces());   size += arraysize(k.gettransitiveinterfaces());    // inner classes   size += objectsize(k.getinnerclasses());    // fields   size += objectsize(k.getfields());    // methods   objarray methods = k.getmethods();   int nmethods = (int) methods.getlength();   if (nmethods != 0l) {      size += methods.getobjectsize();      (int = 0; < nmethods; ++i) {         method m = (method) methods.getobjat(i);         size += m.getobjectsize();         size += objectsize(m.getconstmethod());      }   }    // methodordering - int array records original   // ordering of methods in class file   size += arraysize(k.getmethodordering());    return size; } 

my guess estimation fails horribly in case. reason fails hard without checking object size calculations in detail jvm version in combination characteristics of loaded classes in application.

also, troubleshooting guide java se 6 hotspot vm mentions value approximation.

bottom line, take bytes used value grain of salt.