warning flags was Re: rename issue

[Joel Sherrill]

Ralf Corsepius wrote:
> Joel Sherrill wrote:
>   
>> Chris Johns wrote:
>>     
>>> Ralf Corsepius wrote:
>>>       
>> IMO BSPs should not add specific warnings but we can't
>> just delete these
>>     
> I do not agree with this.
>
> There is nothing wrong with locally using obscure warning flags in 
> testing, but there should not be any room for such flags in CVS.
>
> These things should simply be removed and die!
>
>   
I must not be clear this Monday morning.  I am not
advocating leaving them in the BSP specific flags.
Just evaluating if they are useful in finding questionable
code and if so, adding them to the standard set.
>> -- we need to decide if they become part
>> of the standard set.
>>     
> -Wall and nothing else, or even no warning flags at all.
>
> Otherwise we end up in swamp of flags and toolchain compatibility issues.
>
>   
Warnings (and other static checks) can be very useful.
Any time they find questionable or broken code, it is one
less thing you have to debug.

I don't believe every warning flag is useful but some of the
ones this BSP turned on are showing up real questionable
type abuse.

I am actually concerned about these:

-mstructure-size-boundary=8
-mapcs-frame

which show up across a number of ARM BSPs.  Either they
match the default or potentially create incompatabilities
with the libraries.
> Ralf
>   


-- 
Joel Sherrill, Ph.D.             Director of Research & Development
joel.sherrill at OARcorp.com        On-Line Applications Research
Ask me about RTEMS: a free RTOS  Huntsville AL 35805
   Support Available             (256) 722-9985