FYI,

I'm factoring out the code that parses the link-map table from 
frysk.proc.dead into frysk.solib; once there it will become possible to 
construct a MemoryMap[] from the in-memory table of both live and dead 
processes.

Being able to construct the MemoryMap[] from the in-memory table of a 
live process will let us work-around a kernel bug where, during 
termination, /proc/$$/maps stops being available.

Andrew