[PATCH 03/17] score: Add SYSTEM_STATE_FATAL

Chris Johns chrisj at rtems.org
Tue Feb 18 21:49:34 UTC 2014


On 18/02/2014 6:59 pm, Sebastian Huber wrote:
> On 2014-02-18 08:55, Chris Johns wrote:
>> On 18/02/2014 6:08 pm, Sebastian Huber wrote:
>>> On 2014-02-17 22:41, Chris Johns wrote:
>>>> On 18/02/2014 2:51 am, Sebastian Huber wrote:
>>>>> Merge systems states SYSTEM_STATE_SHUTDOWN and SYSTEM_STATE_FAILED
>>>>> into
>>>>> new system state SYSTEM_STATE_FATAL.  This reflects that all system
>>>>> termination paths end up in _Internal_error_Occurred().
>>>>
>>>> Does this mean non-error termination happens here ? If so is the
>>>> naming suitable ?
>>>
>>> Yes, all (non-error and error) termination paths end up here.  The name
>>> reflects the user extension name that is invoked here.  Since this name
>>> is part of the public API we should not change it.
>>>
>>
>> Should the current function be renamed and the API interface functoin
>> wrap the
>> name changed function ? It is confusing to end up in an error handler
>> when
>> there has been no error.
>
> What about the name SYSTEM_STATE_TERMINAL?

Terminal has the overloaded meaning of a display terminal.

>
> What about renaming _Internal_error_Occurred() to _Terminate()?
>

Yes _Terminate is fine or even _System_Terminate ? This means the state 
could be SYSTEM_STATE_TERMINATED.

Chris



More information about the devel mailing list