<div dir='auto'>If the number of bands is not prohibitively large you can try to run with the option -nd 1, which disables the usage of scalapack. I also noticed that compiling with elpa parallel diagonalization is more stable. <div dir="auto">I hope this helps</div><div dir="auto">Greetings Pietro</div></div><div class="gmail_extra"><br><div class="gmail_quote">Il 26 apr 2019 5:10 PM, "IORI, Federico" <federico.iori@airliquide.com> ha scritto:<br type="attribution" /><blockquote class="quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Hi all.<div><br /></div><div>I found such error randomly in almost my calculations at the moment  (no matter the size, the number  of nodes or cpus used)</div><div><br /></div><div> <font face="arial, helvetica, sans-serif"><i> Davidson diagonalization with overlap<br />forrtl: severe (71): integer divide by zero<br />Image              PC                Routine            Line        Source<br />pw.x               0000000000CFE5CE  Unknown               Unknown  Unknown<br />libpthread-2.22.s  00002AAAB2C08C10  Unknown               Unknown  Unknown<br />libmkl_scalapack_  00002AAAAB0F713F  pdsbrdb_              Unknown  Unknown<br />libmkl_scalapack_  00002AAAAAF303B0  pdsyrdb_              Unknown  Unknown<br />libmkl_scalapack_  00002AAAAAF1AC6D  mkl_pdsyevd0_         Unknown  Unknown<br />libmkl_scalapack_  00002AAAAAF194CC  mkl_pdsyevdm_         Unknown  Unknown<br />libmkl_scalapack_  00002AAAAAF184E7  pdsyevd_              Unknown  Unknown<br />pw.x               0000000000B63358  dspev_module_mp_p         741  dspev_drv.f90<br />pw.x               0000000000B5C1EC  prdiaghg_                 317  rdiaghg.f90<br />pw.x               0000000000A9D02E  pregterg_                 872  regterg.f90<br />pw.x               00000000006580CF  diag_bands_               346  c_bands.f90<br />pw.x               000000000065633B  c_bands_                  101  c_bands.f90<br />pw.x               000000000040CAF1  electrons_scf_            566  electrons.f90<br />pw.x               0000000000409C04  electrons_                152  electrons.f90<br />pw.x               00000000005764EA  run_pwscf_                133  run_pwscf.f90<br />pw.x               00000000004077B5  MAIN__                     98  pwscf.f90<br />pw.x               000000000040761E  Unknown               Unknown  Unknown<br /><a href="http://libc-2.22.so/">libc-2.22.so</a>       00002AAAB3132725  __libc_start_main     Unknown  Unknown<br />pw.x               0000000000407529  Unknown               Unknown  Unknown  </i></font></div><div><font face="arial, helvetica, sans-serif"><i><br /></i></font></div><div><font face="arial, helvetica, sans-serif">Does someone have an idea where it could come for ? is a matter of scalapack or from the diagonalizatino algoritm of pw ?</font></div><div><font face="arial, helvetica, sans-serif">Or maybe suggest a test to reproduce it ? <br /></font><div><div><br /></div><div>Thanks and sorry to bother again with such issue.</div><div><br /></div>-- <br /><div dir="ltr" data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><span style="background-color:transparent;font-family:'arial';font-size:10pt;font-weight:700;white-space:pre-wrap">Federico IORI</span><br /></div><div><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-size:10pt;font-family:'arial';background-color:transparent;font-style:italic;vertical-align:baseline;white-space:pre-wrap">Computational material scientist</span></p><p style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="background-color:transparent;color:rgb( 61 , 133 , 198 );font-family:'arial';font-size:11pt;font-style:italic;font-weight:700;white-space:pre-wrap">Paris Innovation Campus</span></p><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-size:9pt;font-family:'arial';background-color:transparent;vertical-align:baseline;white-space:pre-wrap"><a href="https://www.google.com/maps/place/Air+Liquide/@48.8297381,2.2016685,11.75z/data=!4m5!3m4!1s0x47e67e61a4fbbdc7:0xca3bea9e80059880!8m2!3d48.7630415!4d2.1333045" style="color:rgb( 17 , 85 , 204 );font-family:'arial' , sans-serif;font-size:12.8px;white-space:normal">1 chemin de la Porte des Loges</a><br style="font-family:'arial' , sans-serif;font-size:12.8px;white-space:normal" /><span style="font-family:'arial' , sans-serif;font-size:12.8px;white-space:normal">Les Loges en Josas – 78354 Jouy en Josas cedex</span><br style="font-family:'arial' , sans-serif;font-size:12.8px;white-space:normal" /><span style="font-family:'arial' , sans-serif;font-size:12.8px;white-space:normal">Mail</span><span style="font-family:'arial' , sans-serif;font-size:12.8px;white-space:normal">: <a href="mailto:federico.iori@airliquide.com">federico.iori@airliquide.com</a></span><br /></span></p><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-size:9pt;font-family:'arial';background-color:transparent;vertical-align:baseline;white-space:pre-wrap">Phone: +33 7 621 605 15</span></p><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><span style="font-size:9pt;font-family:'arial';background-color:transparent;vertical-align:baseline;white-space:pre-wrap"><img src="https://lh5.googleusercontent.com/jWge08q7BtEKOlxuBF_cj804ywzhJmkjw9rEb5dBZLKyUL_HQkVwhcwNi9vtzVTj-EiY-wQhE18VPSACNFaAFCAWvxp6bEwkTDR6-IjGBqp-Y3Afzx-OxN8OHxXYPLS0Dz9tP8fr" style="border:medium none" width="165" height="40" /></span></p><p dir="ltr" style="line-height:1.38;margin-top:0pt;margin-bottom:0pt"><br /></p></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div>
</blockquote></div><br></div>