JavaMemoryModel: PDA JVM and Linux OS Crash

From: Eliot Moss (
Date: Thu Jul 24 2003 - 07:10:04 EDT

Dear Jonathan -- You have interesting and useful questions, but I think you
have directed them to an inappropriate list for your purposes. The "Java
Memory Model" discussed on this list has to do with what "reorderings" of
memory accesses are allowed by the combination of compiler, JVM, and
hardware, i.e., what ordering relationships are _required_ and which
operations can legally happen in other orders.

While it is possible that your crashes are related to this question, or to
programmers' (or JVM writers?) inappropriate assumptions about these
orders, you pointed to memory consumption (as driven by the allocation and
garbage collection subsystems). This is not part of the subject being
discussed here. You might be interested in the real-time Java spec (among
other things), which has an interest in more predictable / controlled
consumption of memory in limited memory environments. Further, while I
don't know the details, desktops and servers generally provide virtual
memory and strong memory protection. A PDA may provide only real memory,
shared by all applications (AND the OS!), so it is much more likely that
certain patterns of memory consumption could cause the whole system to lock
up. This is probably a bug -- but eliminating all instances this kind of
bug is (I believe) harder than the comparable situation for desktops. But a
lot of desktop OSs will fail miserably if you run out of disk space.

Hope this helps in some small ways at least .... Eliot Moss
J. Eliot B. Moss, Associate Professor www
Director, Arch. and Lang. Impl. Lab. +1-413-545-4206 voice
Department of Computer Science +1-413-695-4226 cell
140 Governor's Drive, Room 372 +1-413-545-1249 fax
University of Massachusetts at Amherst email
Amherst, MA 01003-9264 USA +1-413-545-3733 Priscilla Coe sec'y
JavaMemoryModel mailing list -

This archive was generated by hypermail 2b29 : Thu Oct 13 2005 - 07:00:46 EDT