Forums

IAR compiler and C++ code

Started by wish_akuze wish_akuze March 28, 2006
Hello
   
  I am running a C++ code in the IAR Embedded Workbench and i am having a
problem with these lines :
   
  float *temp1= new float[N];
  float *temp2= new float[N];
  ..............
   
  delete[] temp1;
 delete[] temp2;
   
  Are these declaration supported with this compiler or is there any other way
of  doing this?
   
  Moreover as i cannot use printf statement in IAR, i have tried to create a
file pointer where the data could be written, but there is error (FILE *Fp) it
says that FILE is undefined??Is there any one whose familiar with this?
   
  Thanks in advance!
   

		
---------------------------------
Blab-away for as little as 1/min. Make  PC-to-Phone Calls using Yahoo!
Messenger with Voice.




Beginning Microcontrollers with the MSP430

Hi!

wish_akuze wish_akuze wrote:

>   I am running a C++ code in the IAR Embedded
Workbench and i am having a problem with these lines :
>    
>   float *temp1= new float[N];
>   float *temp2= new float[N];
>   ..............
>    
>   delete[] temp1;
>  delete[] temp2;
>    
>   Are these declaration supported with this compiler or is there any other
way of  doing this?

Yes, the IAR compiler can handle C++ (well, actually, Embedded C++). 
However, you must manually specify the language in Project -> Options -> 
C/C++ Compiler -> Language.


>   Moreover as i cannot use printf statement in
IAR, i have tried to create a file pointer where the data could be written, but
there is error (FILE *Fp) it says that FILE is undefined??Is there any one whose
familiar with this?

You have to select the "full" version of the DLib library. Nethier
CLib 
or the "normal" version of DLib have support for "FILE *".

However, all libraries has support for plain printf:s.

     -- Anders Lindgren, IAR Systems

-- 
Disclaimer: Opinions expressed in this posting are strictly my own and
not necessarily those of my employer.


Hello,
   
  After enabling the "full DLIB", the code was compiled with no
errors! However when i added these file headers and  disabled the watchodog
timer:
   
  #define MSP430
#ifdef MSP430
#include "msp430x16x.h"
#endif
  ........
  #ifdef MSP430
WDTCTL = WDTPW + WDTHOLD ; /* Stop watchdog timer*/
#endif
   
  I got a bunch of  25 errors! at the defined function prototype:
  void calculate(int N, float *x,float *y);
  and the errors are such as: incomplete type not allowed, and expected brakets
after int N, and float *x, and float*y....etc? Is there anything wrong using
these pointers??
   
  I don't understand why i got these errors. I have ran the same code in visual
studio C++6 and it was ok, and of course i did not need to add the msp430
headers!
   
  The other thing is that the compiler is complaining about the the function's
prototype and the functions itself : it says the variable calculate has been
already defined!!
   
  Any ideas??
   
  Thanks in advance

   
   
  Anders Lindgren <Anders.lindgren@Ande...> wrote:
  
You have to select the "full" version of the DLib library. Nethier
CLib 
or the "normal" version of DLib have support for "FILE *".

However, all libraries has support for plain printf:s.

     -- Anders Lindgren, IAR Systems

-- 
Disclaimer: Opinions expressed in this posting are strictly my own and
not necessarily those of my employer.



.





      
---------------------------------
  . 

    
---------------------------------
  



		
---------------------------------
New Yahoo! Messenger with Voice. Call regular phones from your PC and save big.




wish_akuze wish_akuze wrote:
> Hello,
> 
> After enabling the "full DLIB", the code was compiled with no
errors!
> However when i added these file headers and  disabled the watchodog
> timer:
> 
> #define MSP430
 >
 > #ifdef MSP430
 > #include "msp430x16x.h"
 > #endif ........
 >
> #ifdef MSP430
 >   WDTCTL = WDTPW + WDTHOLD ; /* Stop watchdog timer*/
> #endif 

> I got a bunch of  25 errors! at the defined
function prototype: void
> calculate(int N, float *x,float *y); and the errors are such as:
> incomplete type not allowed, and expected brakets after int N, and
> float *x, and float*y....etc? Is there anything wrong using these
> pointers??

That line looks ok to me -- maybe N is defined as a preprocessor macro, 
which fools the compiler. If you are unsure about a construction, try to 
make a very small example that contains your construction, if it works 
then the problem must be elsewhere.

Another thing that could be the problem is if you're using so-called 
"variable length arrays", that is local arrays whose size is not 
constant, e.g:

    void test(int n)
    {
       float fs[n];
       ...
    }

Unfortunately we, like most microprocessor C implementations, do not 
support them.


> I don't understand why i got these errors. I have
ran the same code
> in visual studio C++6 and it was ok, and of course i did not need to
> add the msp430 headers!
> 
> The other thing is that the compiler is complaining about the the
> function's prototype and the functions itself : it says the variable
> calculate has been already defined!!
> 
> Any ideas??

None, not without actually seeing the code in question.

     -- Anders Lindgren, IAR Systems
-- 
Disclaimer: Opinions expressed in this posting are strictly my own and
not necessarily those of my employer.