[Lapack] Question

[scolnik]

I am looking for the most efficient implementation of a routine in C for 
computing the euclidean norm of a vector taking into account speed and accuracy.
Of course I have been using BLAS in C and Fortran, but to check possible 
under/overflow conditions insumes CPU time and to use extended precision also. 

Therefore the questions is:

Which is in your experience the best approach ?

Thank you very much in advance

Hugo D.Scolnik Ph.D.
Full professor
Computer Sciences Department
University of Buenos Aires
-------------- next part --------------
An HTML attachment was scrubbed...
URL: 
http://lists.cs.utk.edu/private/lapack/attachments/20061002/80d7f3fa/attachment.html