[Help-gsl] gsl_blas speedup

[Riccardo Lucchese]

Hi,

I'd like to use gsl in my project (ingegeristic physics simulations, GPL3),
 especially blas futures.

I noticed that the performance of some calls is too poor to deal with
realtime constraints.

ie. gsl_blas_ddot, which checks for the vectors sizes and then performs one more
call to the blas implementor function.

I'd like to volunteer to work on this, but I don't know if this work is welcome
in the library (as it adds a bit of complexity).

Mainly I was thinking to had two switchs to be defined in the user
application like:

- GSL_BLAS_NOSIZECHECK, turns on/off the vectors sizes check.
  problems: gsl_blas functions will than always return GSL_SUCCESS, and probably
             crash when vectors don't respect the constraints.

- GSL_BLAS_FORCEINLINE, forces inlining both the gsl functions and the the blas
   implementation functions.
  problems: when defined, linking against ATLAS or other blas libs
just has no effect

Anyway on calls like ddot this can be 2 times faster than now, and as
in my tests
simple functions like ddot don't take any advantage in using sse ecc..
(especially
due to the calls overhead).

Maybe I'm all wrong :) ? Any other ideas?

Thanks,
Riccardo