FLOATING POINT with rtems

[Joel Sherrill]

On 11/29/2013 7:35 AM, Luca Cinquepalmi wrote:
> So what is the reason of my error?
> 
> NB: it is not a compilation error, but it only occurs when SW is 
> executed on my target HW.

It is not a compilation error but likely a mismatch in compiler
flags between the way RTEMS was built (assuming no FPU and
only soft-float) and the way your application code was built
(assuming an FPU).

RTEMS was told there is no FPU and thus the FPU enable bit
will never be set.

FWIW you never said whether you got the tools from Aeroflex
Gaisler (RCC) or not. They tend to build with soft-float as
I recall.

--joel

> 
> Il 29/11/2013 14:23, Sebastian Huber ha scritto:
>> On 2013-11-29 14:19, Luca Cinquepalmi wrote:
>>> But -msoft-float is used when you wnat to simulate FP coprocessor. It 
>>> is a SW
>>> library.
>>> In my case I have a real FP coprocessor, so I don't need to use it. 
>>> Do you agree?
>>
>> Yes, and you must compile the BSP without the -msoft-float option, 
>> otherwise there will be no support for hardware floating-point 
>> operations at all.
>>
> 
> _______________________________________________
> rtems-users mailing list
> rtems-users at rtems.org
> http://www.rtems.org/mailman/listinfo/rtems-users
> 



-- 
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