(j3.2006) an intrinsic for SORT()

[Bill Long]

> On Feb 6, 2018, at 1:45 PM, Van Snyder <van.snyder at jpl.nasa.gov> wrote:
> 
> On Tue, 2018-02-06 at 18:52 +0000, Bill Long wrote:
> 
>> We have other, similar, intrinsics.  MATMUL, for example.
> 
> If I write MATMUL(TRANSPOSE(A),B) will most processors create a temp for
> TRANSPOSE(A), or will they do what DGEMM does and handle the TRANSPOSE
> as a flag that changes how the multiply is done?

I suppose it depends on the compiler.  At default compiler optimization options, for one compiler I have available,  this source

subroutine test (a, b, c, n)
 
  real(8) :: a(n,n), b(n,n), c(n,n)

  c = matmul(transpose(a),b)

end subroutine test



results in this code for  line 5:

    5.       if ( n > 0 ) then
    5.           _substr($W0,1,1) = 't'
    5.           _substr($W1,1,1) = 'n'
    5.          $W2 = n
    5.          $W3 = n
    5.          $W4 = n
    5.          $W5 = 1.0
    5.          $W6 = n
    5.          $W7 = n
    5.          $W8 = 0.0
    5.          $W9 = n
    5.          _dgemm(  _substr($W0,1,1),  _substr($W1,1,1), $W2, $W3, $W4, $W5, a(1, 1), $W6, b(1
, 1), $W7, $W8, c(1, 1), $W9 )
    5.       endif

Cheers,
Bill



> 

Bill Long                                                                       longb at cray.com
Principal Engineer, Fortran Technical Support &   voice:  651-605-9024
Bioinformatics Software Development                      fax:  651-605-9143
Cray Inc./ 2131 Lindau Lane/  Suite 1000/  Bloomington, MN  55425