[Bug 1593] RTEMS Sequenced Initialization API

[bugzilla-daemon at rtems.org]

https://www.rtems.org/bugzilla/show_bug.cgi?id=1593

--- Comment #9 from Sebastian Huber <sebastian.huber at embedded-brains.de> 2010-07-20 04:19:50 CDT ---
1. You can ignore the files 'c/src/optman/rtems/no-*.c'.  They will be obsolete
with the sysinit API.

2. Do not introduce #if 0 sections.

3. Why are the sysinit functions of each manager globally visible?  They should
be made local.  This avoids the changes in the header file.

4. Do not add empty termination functions.  Using an enum for the sysinit state
was chosen to allow extensions.  Making a possible new state a fatal error
contradicts this approach.  The default case should not lead to a fatal error. 
This simplifies the switch statement to:

switch ( state ) {
  case SYSINIT_INITIALIZE:
    _X_Manager_initialization();
    break;
  default:
    break; 
}

5. Please provide informative comments and not copy&paste stuff that only says
as much as the function signature like this:

/**
 *  @brief _Event_Manager_Sysinit
 *
 *  Input parameters:   
 *    state  -  initialization or temination
 *
 *  Output parameters:  NONE
 */
void _Event_Manager_Sysinit(
  Sysinit_State state
)

What is the documentation value of this comment?  There is no need to document
trivial stuff.  These sysinit functions need no documentation comment.

6. Please provide a third patch for the SIS BSP changes.

7. I guess you used the size of the ELF-file for you comparisons.  The ELF-file
contains a lot of sections and very likely also sections with debug
information.  We are not interested in size changes in these kind of sections. 
You can use the sparc-rtems4.10-size program to compare the sizes of the binary
sections.

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