[RTEMS Project] #3755: leak in realloc

[RTEMS trac]

#3755: leak in realloc
--------------------------+---------------------
 Reporter:  Jeffrey Hill  |       Owner:  (none)
     Type:  defect        |      Status:  new
 Priority:  normal        |   Milestone:
Component:  score         |     Version:  4.11
 Severity:  normal        |  Resolution:
 Keywords:  realloc leak  |  Blocked By:
 Blocking:                |
--------------------------+---------------------

Comment (by Jeffrey Hill):

 I should add that we see linear 5kB leaked each Lua context create/destroy
 cycle. When a Lua context runs for extended period there isn't a leak.
 Maybe it doesn't look like a fragmentation issue due to the linear 5kB per
 cycle. Admittedly our heapSpace command line diagnostic maybe isn't
 showing info about fragmentation. We create a fresh Lua context when a
 client connects to the EPICS server, and destroy it at client disconnect.

 I did log all of the Lua allocations and frees by pointer value and
 verified that they were balanced using a map in a python script a few
 weeks back. I don't recall now if there were nil pointers in there in that
 mix. Its a good idea, and something to look for, I will have a look when
 spare time arrives :-).

 Also I suppose that a standalone test playing back the various allocations
 requests, shrinking and enlarging existing blocks, simulating the lua
 calls to realloc, but just following input from my log files might
 reproduce the issue standalone?

--
Ticket URL: <http://devel.rtems.org/ticket/3755#comment:9>
RTEMS Project <http://www.rtems.org/>
RTEMS Project