[R] BLAS and the Sun Performance Library

Prof Brian Ripley ripley at stats.ox.ac.uk
Wed Jan 9 19:59:16 CET 2002


On Wed, 9 Jan 2002, David Brahm wrote:

> Hi, all,
>
>    We just got the Sun Performance Library for our Suns running Solaris 2.6,
> and I read the relevant section of the "R Installation and Administration
> Manual" about configuring R to use BLAS.  The manual suggests you must use
> Sun's "cc" and "f95" (maybe "f90"?) compilers in order to use the Performance
> Library.  We have "cc" but not "f95".

At least here cc+f77 does not work. libsunperf depends on libraries that
are part of f95 on our system, specifically most of -lfui -lfai -lfai2
-lfsumai -lfprodai -lfminlai -lfmaxlai -lfminvai -lfmaxvai -lfsu, and
FLIBS did not get those from f77, nor does my home machine without f95
installed have them on.  You may be able to work around this.

f90 and f95 are the same thing on our installation.

>    Would we be better off buying "f95", or compiling our own BLAS using ATLAS
> and our usual "gcc"/"g77" Gnu compilers?  Has anyone successfully done the
> latter under Solaris 2.6?

Yes, and it's pretty effective.  I think the Sun Fortran compilers are a
lot better than g77 on Solaris, but unless you are using R for linear
algebra (which perhaps *you* are), the final speed differences are small.
(And if you are and are worried about speed, try a matrix algebra
package.)

-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._



More information about the R-help mailing list