[RTEMS Project] #3621: Statically initialize object information structures

[RTEMS trac]

#3621: Statically initialize object information structures
-----------------------------+------------------------------
 Reporter:  Sebastian Huber  |       Owner:  Sebastian Huber
     Type:  enhancement      |      Status:  assigned
 Priority:  normal           |   Milestone:  5.1
Component:  score            |     Version:  5
 Severity:  normal           |  Resolution:
 Keywords:                   |  Blocked By:
 Blocking:                   |
-----------------------------+------------------------------

Comment (by Chris Johns):

 Replying to [ticket:3621 Sebastian Huber]:
 > The object controls reside on the heap even for fixed object count
 configuration.  Using a statically allocated array makes it easier to find
 the objects during debugging.

 It is not clear me to what the requirements are these change are being
 based on? Is there some overriding push for everything to be static tables
 every where.

 Static tables for initialisation do solve some issues such as audit-able
 configuration control but are there other use cases where this may not be
 a good fit. I cannot tell. For example statically inflexible kernel
 initialisation and libdl do not sit well together. The demands on the
 kernel configuration from the loadable code can vary and does and if you
 consider libdl as a means to produce "golden images" having the ability to
 vary the configuration is important. I have built systems where the
 bootloader loads the kernel configuration data and then loads the
 application. Can this still be done?

--
Ticket URL: <http://devel.rtems.org/ticket/3621#comment:1>
RTEMS Project <http://www.rtems.org/>
RTEMS Project