Misterious BLACS

phaser75 — Tue, 24 Nov 2009 08:15:29 GMT

Hello.
I use BLACS and ScaLapack to inverse dense matrix.
And I got some troubles.

I set up nprow and npcol for block-cyclic distribution for example 2, 2.
and run this routine

[cpp]nprow = 2;
npcol=2;
Cblacs_get(0,0,&ictxt);
Cblacs_gridinit(&ictxt,"R",nprow,npcol);
Cblacs_gridinfo(ictxt,&nprow,&npcol,&myrow,&mycol);
[/cpp]

The stange thing is that after Cblacs_gridinfo(..) nprow is changed as 0.
Whyis it changed?
Because it is changed to 0, numroc(..) called after that generates error.
Could you give me some ideas?
Thanks in advance.

ps. Is there another method to inverse matrix without BLACS?

Re: Misterious BLACS

Dmitry_K_Intel2 — Tue, 24 Nov 2009 09:02:34 GMT

Quoting - phaser75

Hello.
I use BLACS and ScaLapack to inverse dense matrix.
And I got some troubles.

I set up nprow and npcol for block-cyclic distribution for example 2, 2.
and run this routine

[cpp]nprow = 2;
npcol=2;
Cblacs_get(0,0,&ictxt);
Cblacs_gridinit(&ictxt,"R",nprow,npcol);
Cblacs_gridinfo(ictxt,&nprow,&npcol,&myrow,&mycol);
[/cpp]

Hi phaser75,

I'm afraid that there are no specialists in this thread who could help you with BLACS. This thread is mainly dedicated to Intel MPI Library. You could try Lapack forum first: http://icl.cs.utk.edu/lapack-forum/

Please compare your appplication with an example here: http://www.netlib.org/lapack-dev/old/examples/test_pdgesvd.c.html

Best wishes,
Dmitry

topic Re: Misterious BLACS in Intel® MPI Library

Misterious BLACS

Re: Misterious BLACS