Compiler Release 7.4.0

Clement Mesnier said:

Can you tell us when it will be released ?

The release is planned for the end of July.  That Wiki page needs to be updated.

Clement Mesnier said:

Moreover do you have any details about the "C99 Floating Point Complex Type Support (float _Complex, double _Complex, long double _Complex)." ?

Does that mean that all the functions of C99's complex.h are implemented or not ?

Yes.

Clement Mesnier said:

How is it different that using a double (or __float2_t for C66x) to store the complex number (float real float im) ?

I'm not sure what you mean.  Please give an example of what you are concerned about.

Thanks and regards,

-George

Thanks for the info.

Right now I have complex data stored in memory in one table with real part in even positions and imaginary part in odd positions like :

Re Im Re Im Re Im ...

then I do my calculation (matrix multiplications) with the use of intrinsics

typically : 

double cplx1, cplx2,result

float re_res, im_res

cplx1 = _amemd8(&table1[0]);
cplx2 = _amemd8(&table2[0]);

result = _complex_mpy(cplx1,cplx2);

re_res = -hif(result)
im_res = lof(result) 

What is the interest to use the complex floating point support ?

Would my optimized code take fewer cycles ?

Thanks, CM

Ok thank you for the clarification.