problems with drops <math.h> and '-lm' linkage

[Vladimir Nikolov]

Hi Martin,
thank you for your reply, I have tried to compile my code in the 
"l4env_diet" MODE and everything was okay then!
The library, which I used to link with my drops package, uses internally 
<math.h> functions for doing some FFT and MDCT.
I didn't know about this build-mode, I used always the modes as 
described on the 'bid-spec' tutorial from the L4Env webpage
( http://os.inf.tu-dresden.de/l4env/doc/ ), so thanks for the suggestion!
Regards

Vladimir


Martin Pohlack schrieb:

>Hi Vladimir,
>
>[...]
>
>  
>
>>Does anyone have any solution for this problem?
>>I thought about searching another math library, which does not use 
>>standard C math functions and trying to compile it as a library in a 
>>drops package by including 'lib.mk' in the appropiate Makefile. I am not 
>>really shure, if this would really solve my problem, or just keep me 
>>fighting with other incompatibility problems. Is there no other 
>>possibility to use mathematical functions in any drops application, or 
>>perhaps another linkage concept or configuration?
>>    
>>
>
>I see two ways for you to proceed:
>
>1) Try another mode, e.g. a dietlibc mode.  I built the dietlibc such,
>that a libm is already included.
>
>2) Write a *minimal* example which triggers the problem in the freebsd
>mode, maybe someone can tell more from this.  FYI, the freebsd mode is
>unmaintained now.
>
>Cheers,
>Martin
>  
>