I'm sure that p is the number of processes in this example and therefore dividing by p is correct, but j always without +1 (because j is a process number) and k always with +1 (because indix is starting with 1 in Fortran). According to official Postscript version of the standard: Page 117 lines 44-45 division by p instead of m is necessary. Line 44 without +1, Line 45 with +1. Page 118 line 45 with +1. Page 120 lines 33 with +1. About Page 117, lines 47-48, I'm not sure. I propose to put this into MPI 2.2, because this example should be tested before correcting it. Best regards Rolf On Fri, 8 Feb 2008 12:06:26 -0300 "Lisandro Dalcin" <dalcinl_at_[hidden]> wrote:
Can someone take a look at the second example in:
http://www.mpi-forum.org/docs/mpi-20-html/node124.htm
and also to the example in:
http://www.mpi-forum.org/docs/mpi-20-html/node125.htm
I believe in both cases the lines saying
j = map(i)/p k = MOD(map(i),p)
should read
j = map(i)/m k = MOD(map(i),m)
Additionally, in the first example again in:
http://www.mpi-forum.org/docs/mpi-20-html/node124.htm
in the part where origin and target indices are computed, there are two lines reading:
oindex(total(j) + count(j)) = i tindex(total(j) + count(j)) = k
Should'nt it read like this (for indices being zero-based)?
oindex(total(j) + count(j)) = i - 1 tindex(total(j) + count(j)) = k - 1
Regards,
-- Lisandro Dalcín --------------- Centro Internacional de Métodos Computacionales en Ingeniería (CIMEC) Instituto de Desarrollo Tecnológico para la Industria Química (INTEC) Consejo Nacional de Investigaciones Científicas y Técnicas (CONICET) PTLC - Güemes 3450, (3000) Santa Fe, Argentina Tel/Fax: +54-(0)342-451.1594
_______________________________________________ mpi-forum mailing list mpi-forum_at_[hidden] http://lists.cs.uiuc.edu/mailman/listinfo/mpi-forum
Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner_at_[hidden] High Performance Computing Center (HLRS) . phone ++49(0)711/685-65530 University of Stuttgart . . . . . . . . .. fax ++49(0)711 / 685-65832 Head of Dpmt Parallel Computing . . . www.hlrs.de/people/rabenseifner Nobelstr. 19, D-70550 Stuttgart, Germany . (Office: Allmandring 30)