[Wien] Error in nlvdw
Laurence Marks
laurence.marks at gmail.com
Mon Feb 24 05:03:25 CET 2020
Sounds like it was an OOM error [1]. There should be something in the
system logs (e.g. /var/log). Since OOM handling is performed by the kernel
it is not simple to trap this.
[1] e.g. https://en.m.wikipedia.org/wiki/Out_of_memory
_____
Professor Laurence Marks
"Research is to see what everybody else has seen, and to think what nobody
else has thought", Albert Szent-Gyorgi
www.numis.northwestern.edu
On Sun, Feb 23, 2020, 20:34 Gavin Abo <gsabo at crimson.ua.edu> wrote:
> I did some more testing of the nlvdw package compiled with gfortran with
> .outputnlvdw having unit 6. A nlvdw calculation runs fine. It does appear
> to just be due to a deficiency of resources for my system. Further details
> given below should you be interested in them.
>
> In gcc-7.4.0.tar.gz at [1], it can be seen in gcc/fortran/libgfortran.h
> that unit 6 and 5 are preconnected as standard output and standard input,
> respectively, from:
>
> /* Default unit number for preconnected standard input and output. */
> #define GFC_STDIN_UNIT_NUMBER 5
> #define GFC_STDOUT_UNIT_NUMBER 6
> #define GFC_STDERR_UNIT_NUMBER 0
>
> However, as stated at [2]:
>
> *Like any other preconnected unit it may have its connection changed in an
> open statement.*
>
> Indeed, the preconnection of unit 6 as standard output is overridden to
> write to the .outputnlvdw file instead by the open statement in WIEN2k's
> SRC_nlvdw/vdw.F (on line 87 when iunit=6):
>
> open (iunit,file=fname,status=status,form=form)
>
> When I tried running Dr. Chakrabarti's alloy_VOPT.struct at [3] with
> "run_lapw -nlvdw", the .dayfile gave me a "Killed" error and .outputnlvdw
> was empty.
>
> When I ran it, I found it strange that .outputnlvdw was empty. It turned
> out that gfortran buffers file output in memory by default such that the
> output was never written to the file when the program was killed. I have
> found that the environmental variable GFORTRAN_UNBUFFERED_ALL can be set to
> 1 [4]. Using GFORTRAN_UNBUFFERED_ALL can slow down the small sequential
> reads and writes [5] but is obviously better for debugging and also got me
> passed what I previously reported that might be a gfortran bug with
> buffered file output when using unit 6. As using a 'print *,iunit'
> statement to try to debug the above open statement in vdw.F printed to the
> .dayfile that files up to unit 5 were opened but never showed unit 6 and
> the later files from nlvdw.def opening.
>
> The calculation I ran made it to the same spot as Dr. Chakrabarti and
> expect the Ubuntu killed it due to insufficient RAM as Prof. Tran pointed
> out at [6]. The small resource Al calculation ran without errors though as
> shown below.
>
> *alloy_VOPT Calculation*
>
> username at computername:~/wiendata/alloy_VOPT$ ls
> alloy_VOPT.struct
> username at computername:~/wiendata/alloy_VOPT$ init_lapw -b
> ...
> username at computername:~/wiendata/alloy_VOPT$ gedit alloy_VOPT.in0
> username at computername:~/wiendata/alloy_VOPT$ cat alloy_VOPT.in0
> TOT XC_GGA_X_B86_R EC_LDA VC_LDA
> (XC_LDA,XC_PBESOL,XC_WC,XC_MBJ,XC_SCAN)
> NR2V IFFT (R2V)
> 144 144 144 2.00 1 NCON 9 # min IFFT-parameters, enhancement
> factor, iprint, NCON n
> username at computername:~/wiendata/alloy_VOPT$ cp
> $WIENROOT/SRC_templates/case.innlvdw alloy_VOPT.innlvdw
> username at computername:~/wiendata/alloy_VOPT$ gedit alloy_VOPT.innlvdw
> username at computername:~/wiendata/alloy_VOPT$ cat alloy_VOPT.innlvdw
> 1 kernel type
> -1.887 parameters of the kernel
> 25 plane-wave expansion cutoff GMAX
> 0.3 density cutoff rhoc
> T calculation of the potential (T or F)
> 10 plane-wave expansion cutoff GMAXpot for the potential
> username at computername:~/wiendata/alloy_VOPT$ export
> GFORTRAN_UNBUFFERED_ALL=1
> username at computername:~/wiendata/alloy_VOPT$ run_lapw -nlvdw
> hup: Command not found.
> grep: *scf1*: No such file or directory
> grep: lapw2*.error: No such file or directory
>
> > stop error
> username at computername:~/wiendata/alloy_VOPT$ cat alloy_VOPT.dayfile
>
> Calculating alloy_VOPT in /home/username/wiendata/alloy_VOPT
> on computername with PID 26433
> using WIEN2k_19.1 (Release 25/6/2019) in /home/username/WIEN2k
>
>
> start (Sun Feb 23 17:47:33 MST 2020) with nlvdw (40/99 to go)
>
> cycle 1 (Sun Feb 23 17:47:33 MST 2020) (40/99 to go)
>
> > nlvdw (17:47:33) Killed
> 710.0u 58.6s 1:18:02.53 16.4% 0+0k 7490168+40io 262257pf+0w
> error: command /home/username/WIEN2k/nlvdw nlvdw.def failed
>
> > stop error
> username at computername:~/wiendata/alloy_VOPT$ cat alloy_VOPT.outputnlvdw
> kernel type = 1 (M. Dion et al., PRL 92, 246401 (2004))
> parameter Z_ab of kernel = -1.88700000
> gmax = 25.0
> density cutoff rhoc = 0.300E+00
> the NL-vdW potential is calculated with gmax_pot = 10.0
>
> n_max, m_max, p_max = 101 101 101
> ifft1, ifft2, ifft3 (for proc myid 0) = 203 203 203
> ifft1*ifft2*ifft3 (for proc myid 0) = 8365427
> Number of G-vectors (for proc myid 0) = 3099627
>
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
> % %
> % You are using vdW-DF, which was implemented by the Thonhauser group. %
> % Please cite the following two papers that made this development %
> % possible and the two reviews that describe the various versions: %
> % %
> % T. Thonhauser et al., PRL 115, 136402 (2015). %
> % T. Thonhauser et al., PRB 76, 125112 (2007). %
> % K. Berland et al., Rep. Prog. Phys. 78, 066501 (2015). %
> % D.C. Langreth et al., J. Phys.: Condens. Matter 21, 084203 (2009). %
> % %
> % %
> % If you are calculating the stress with vdW-DF, please also cite: %
> % %
> % R. Sabatini et al., J. Phys.: Condens. Matter 24, 424209 (2012). %
> % %
> %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
>
> parameters of the kernel table:
> Nq = 30, lambda = 0.111292E+01, q(1) = 0.100000E-04, q_cut = 0.100000E+02
> Nr_points = 2000, r_max = 100.0
> q_mesh =
> 0.00001000 0.05312929 0.11224701 0.17804050
> 0.25126365 0.33275542 0.42344952 0.52438515
> 0.63671875 0.76173753 0.90087384 1.05572188
> 1.22805595 1.41985071 1.63330352 1.87086022
> 2.13524270 2.42948008 2.75694394 3.12138629
> 3.52698255 3.97838044 4.48075151 5.03985262
> 5.66208887 6.35459089 7.12529230 7.98302412
> 8.93761444 10.00000000
>
> *Al Calculation*
>
> username at computername:~/wiendata/Al$ lsb_release -a
> No LSB modules are available.
> Distributor ID: Ubuntu
> Description: Ubuntu 18.04.4 LTS
> Release: 18.04
> Codename: bionic
> username at computername:~/wiendata/Al$ cat $WIENROOT/WIEN2k_VERSION
> WIEN2k_19.1 (Release 25/6/2019)
> username at computername:~/wiendata/Al$ dpkg -l gfortran
> Desired=Unknown/Install/Remove/Purge/Hold
> |
> Status=Not/Inst/Conf-files/Unpacked/halF-conf/Half-inst/trig-aWait/Trig-pend
> |/ Err?=(none)/Reinst-required (Status,Err: uppercase=bad)
> ||/ Name Version Architecture Description
>
> +++-==============-============-============-=================================
> ii gfortran 4:7.4.0-1ubu amd64 GNU Fortran 95 compiler
> username at computername:~/wiendata/Al$ ls
> Al.struct
> username at computername:~/wiendata/Al$ init_lapw -b
> ...
> username at computername:~/wiendata/Al$ cp
> $WIENROOT/SRC_templates/case.innlvdw Al.innlvdw
>
> For rev-vdW-DF2:
>
> username at computername:~/wiendata/Al$ cat Al.innldvdw
> 1 kernel type
> -1.887 parameters of the kernel
> 25 plane-wave expansion cutoff GMAX
> 0.3 density cutoff rhoc
> T calculation of the potential (T or F)
> 10 plane-wave expansion cutoff GMAXpot for the potential
> username at computername:~/wiendata/Al$ cat Al.in0
> TOT XC_GGA_X_B86_R EC_LDA VC_LDA
> (XC_LDA,XC_PBESOL,XC_WC,XC_MBJ,XC_SCAN)
> NR2V IFFT (R2V)
> 30 30 30 2.00 1 NCON 9 # min IFFT-parameters, enhancement
> factor, iprint, NCON n
> username at computername:~/wiendata/Al$ run_lapw -nlvdw
> ...
> in cycle 6 ETEST: .0003618600000000 CTEST: .0358926
> hup: Command not found.
> STOP NLVDW END
> STOP LAPW0 END
> STOP LAPW1 END
> STOP LAPW2 END
> STOP CORE END
> STOP MIXER END
> ec cc and fc_conv 1 1 1
>
> > stop
> [1] http://mirrors.ocf.berkeley.edu/gnu/gcc/gcc-7.4.0/
> <https://urldefense.proofpoint.com/v2/url?u=http-3A__mirrors.ocf.berkeley.edu_gnu_gcc_gcc-2D7.4.0_&d=DwMD-g&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=eQjcCVoSqogB8I3gXkAPxUwcxtbS-Hm9O0xfNFd6WRs&e=>
> [2]
> https://stackoverflow.com/questions/44867775/fortran-code-wont-write-to-file
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__stackoverflow.com_questions_44867775_fortran-2Dcode-2Dwont-2Dwrite-2Dto-2Dfile&d=DwMD-g&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=8O22qnbrrW4on-g8vm9tQF3ZLIu6czlxzoALhypQAUw&e=>
> [3]
> https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19663.html
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__www.mail-2Darchive.com_wien-40zeus.theochem.tuwien.ac.at_msg19663.html&d=DwMD-g&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=Dy6-d-gkdrebvx4aGxrJPOUH4Lwr_YECpmw9N-j-i4c&e=>
> [4]
> https://stackoverflow.com/questions/1533717/how-do-i-flush-output-to-file-after-each-write-with-a-gfortran-fortran-90-progra
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__stackoverflow.com_questions_1533717_how-2Ddo-2Di-2Dflush-2Doutput-2Dto-2Dfile-2Dafter-2Deach-2Dwrite-2Dwith-2Da-2Dgfortran-2Dfortran-2D90-2Dprogra&d=DwMD-g&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=HANLf_ggjL_sl0evR5wupdtim92P4VEV6F3C-fnDrYA&e=>
> [5]
> https://gcc.gnu.org/onlinedocs/gcc-7.4.0/gfortran/GFORTRAN_005fUNBUFFERED_005fALL.html
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__gcc.gnu.org_onlinedocs_gcc-2D7.4.0_gfortran_GFORTRAN-5F005fUNBUFFERED-5F005fALL.html&d=DwMD-g&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=Pr19bWzscoLNBm8DX_bG5L4qmVPcsjFI1Im6bSSzF5o&e=>
> [6]
> https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg19667.html
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__www.mail-2Darchive.com_wien-40zeus.theochem.tuwien.ac.at_msg19667.html&d=DwMD-g&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=AJaOsduDgoCJ_mlaxFFNZc592oqJQVEMBJuGnlQ9k0I&e=>
>
>
> On 2/17/2020 3:16 AM, Peter Blaha wrote:
>
> But in nlvdw we do NOT write to the screen (except in case of some
> errors).
>
> In any case, please change nlvdw.def unit 6 to 66 and run nlvdw nlved.def
> to verify if this is the source of the problem.
> It should print all output on the screen then.
>
>
> On 2/16/20 6:39 PM, Gavin Abo wrote:
>
> Perhaps it's just my system but the WIEN2k 19.1 nlvdw, which has been
> compiled with gfortran (gcc version 7.4.0 under Ubuntu 18.04.4 LTS), is
> behaving strangely. It seems to get hung up trying to open
> $file.outputnlvdw in SCR_nlvdw/vdw.F. In x_lapw, I see $file.outputnlvdw
> defined as unit 6.
>
>
> It might be that use of unit 6 in the nlvdw package is problematic when
> compiled with gfortran similar to what we saw for SRC_symmetry:
>
>
> https://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/msg16674.html
> <https://urldefense.proofpoint.com/v2/url?u=https-3A__www.mail-2Darchive.com_wien-40zeus.theochem.tuwien.ac.at_msg16674.html&d=DwMD-g&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=ZpweoysBkNubrAwKeoEjwtQ2-CRv6cT5AF6PJ1Stqko&e=>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
>
> https://urldefense.proofpoint.com/v2/url?u=http-3A__zeus.theochem.tuwien.ac.at_mailman_listinfo_wien&d=DwICAg&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=3ktO9qKFs01QXllahJ0fc5ALkn3_T6u6i1NdRT3bjfc&e=
> SEARCH the MAILING-LIST at:
> https://urldefense.proofpoint.com/v2/url?u=http-3A__www.mail-2Darchive.com_wien-40zeus.theochem.tuwien.ac.at_index.html&d=DwICAg&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=U_T4PL6jwANfAy4rnxTj8IUxm818jnvqKFdqWLwmqg0&m=2ndQWR6tn-8Au2xV9qWpEGZCDtC_VAKsykRUdroZfec&s=a5AHOSatTO1NiSyl3Z9dK2uczJXbUph24-zJzAqJGLI&e=
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://zeus.theochem.tuwien.ac.at/pipermail/wien/attachments/20200223/49e9e82b/attachment.html>
More information about the Wien
mailing list