[CP2K-user] [CP2K:12451] Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL

[Nick Dunn PhD]

I have attached output files with extended output for sopt and popt (we don't have a sdbg built) running a minimal input:

&GLOBAL
   PROJECT Si_bulk8
   PRINT_LEVEL HIGH
&END GLOBAL

I've also attached verbose output for the main calculation in question, both globally and per-processor.

Regards,
- Nick

-------------- next part --------------
A non-text attachment was scrubbed...
Name: sopt_debug.out
Type: application/octet-stream
Size: 213259 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20191121/8be1619b/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: popt_debug.out
Type: application/octet-stream
Size: 52230 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20191121/8be1619b/attachment-0001.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: per_processor_logs.zip
Type: application/zip
Size: 25549 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20191121/8be1619b/attachment.zip>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Ag_bulk-test3.out
Type: application/octet-stream
Size: 9601 bytes
Desc: not available
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20191121/8be1619b/attachment-0002.obj>
-------------- next part --------------

> On Nov 4, 2019, at 4:39 AM, hut... at chem.uzh.ch wrote:
> 
> Hi
> 
> your input file runs without problems on my local build (Trunk version).
> From your output I see that the gradient in the first iteration is
> much to large (7837.47846158 should be 0.75558761). Probably, the MKL
> error is just a follow up problem.
> 
> Can you run different executables (sopt, sdbg, popt) with extended
> output (PRINT_LEVEL HIGH)?
> 
> regards
> 
> Juerg Hutter
> --------------------------------------------------------------
> Juerg Hutter                         Phone : ++41 44 635 4491
> Institut f?r Chemie C                FAX   : ++41 44 635 6838
> Universit?t Z?rich                   E-mail: hut... at chem.uzh.ch
> Winterthurerstrasse 190
> CH-8057 Z?rich, Switzerland
> ---------------------------------------------------------------
> 
> -----cp... at googlegroups.com wrote: -----
> To: "cp2k" <cp... at googlegroups.com>
> From: "Nick Dunn" 
> Sent by: cp... at googlegroups.com
> Date: 10/31/2019 08:28PM
> Subject: [CP2K:12448] Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL
> 
> I am a systems administrator working with a user who is seeing an MKL error while running cp2k. Specifically, the error message is:
> 
> "  Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.
> 
> Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.
> Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.
> 
> Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.
> Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.
> 
> Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.
> Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.
> 
> Intel MKL ERROR: Parameter 4 was incorrect on entry to DLASCL.
> {    1,    1}:  On entry to
> DSTEQR parameter number  -31 had an illegal value
> {    0,    0}:  On entry to
> DSTEQR parameter number  -31 had an illegal value"
> 
> The version of cp2k is 6.1, built against gcc 7.2.0, openmpi 3.0.0, and mkl 2018. The build passes the full set of regtests. I am attaching the build notes, input files, and job script to this message. Any ideas as to what might be going on?
> 
> Regards,
> - Nick  
>  -- 
> You received this message because you are subscribed to the Google Groups "cp2k" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to cp... at googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/9999e1dc-4419-4b27-9384-2778435d7f71%40googlegroups.com.
> 
> 
> [attachment "Si_bulk8.pbs" removed by J?rg Hutter/at/UZH]
> [attachment "Si_bulk8.out" removed by J?rg Hutter/at/UZH]
> [attachment "Si_bulk8.inp" removed by J?rg Hutter/at/UZH]
> [attachment "GTH_POTENTIALS_Si.dms" removed by J?rg Hutter/at/UZH]
> [attachment "BASIS_SET_Si.dms" removed by J?rg Hutter/at/UZH]
> 
> -- 
> You received this message because you are subscribed to a topic in the Google Groups "cp2k" group.
> To unsubscribe from this topic, visit https://groups.google.com/d/topic/cp2k/E8F0SJz62tQ/unsubscribe.
> To unsubscribe from this group and all its topics, send an email to cp... at googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/OF3A1ACFFB.C4B669C1-ONC12584A8.003A878F-C12584A8.003A8791%40lotus.uzh.ch.

-- 
Nicholas Dunn, Ph.D. (he/him/his)
Scientific Computing Consultant
Minnesota Supercomputing Institute
University of Minnesota
Office: (612) 624-4341
dunn... at umn.edu