LAPACK Archives

[Lapack] ATLAS Code-- Routine Invoke Ambiguity!

you certainly have contacted some ATLAS authors with this mailing list but 
the best is that you reask your question to the ATLAS support guys:

On Thu, 16 Mar 2006, Minhaj KHAN wrote:

Dear Author of ATLAS routines

We are trying runtime optimization over ATLAS code.
I want to ask one question regarding routines starting with
ATL_dref... which do not take matrix-order as argument.

when ATL_drefsymm is called for Left+Upper matrix it invokes ATL_drefsymmLU
but when cblas_dsymm is called for Left+Upper matrix with Row Major, it
invokes ATL_drefsymmRL

Does it mean that the routines in ATL_dref (symm, gemm etc.) when
directly called always invoke the one for column major order?

Thanks for the Response.
Best Regards
Minhaj Ahmad
Lapack mailing list

<Prev in Thread] Current Thread [Next in Thread>

For additional information you may use the LAPACK/ScaLAPACK Forum.
Or one of the mailing lists, or