[Bug 1672] Heap protection

[bugzilla-daemon at rtems.org]

https://www.rtems.org/bugzilla/show_bug.cgi?id=1672

Gedare <giddyup44 at yahoo.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |giddyup44 at yahoo.com

--- Comment #2 from Gedare <giddyup44 at yahoo.com> 2010-08-12 10:22:56 CDT ---
I am not an expert, but I took some time to read through the patch. I like the
idea. :)  I have a few comments and questions, mostly to satisfy my own
curiousity.

General comments:
How does a user plug in new handlers?

heap.h:
  * Currently it is not usual to have inline functions declared in headers in
RTEMS.

  * I see two naming conventions used, Heap_Protection_XXX and Heap_Block_XXX. 
Do these names make sense always? Or should the routines be named
Heap_Block_protection_XXX?  FWIW the code reads fine with the
Heap_Protection_XXX, but if protection is applied at the block level, then it
may make more sense to make protection a 'subpackage' of block.

heapallocate.c:

  152: why the change with the conditional const pointer, when before there was
already a const?

  180: I'm personally not a fan of backwards goto labels.  At least put a
comment there to justify the label and why the code jumps back to restart the
loop.

heap.c:

  161: FIXME should be checked.  Avoid #if 0 in code.  I see this relates
partially to your complaint about Errors in RTEMS.

heapfree.c:

  * Why is the Thread Dispatch disable level being checked?  My intuition is
that heap checking is only activated if the application is freeing its own
allocated space, but it wasn't 100% clear to me.

-- 
Configure bugmail: https://www.rtems.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are watching all bug changes.