abinit/tests/cpu/tests.cnf

142 lines
5.1 KiB
INI

# Title:
various cpu tests of the sequential version of abinit
#
# Before running RunTests.pl with this file, read README in this directory.
#
# Copyright (C) 2003-2025 ABINIT group (LSi)
# This file is distributed under the terms of the
# GNU General Public License, see ~abinit/COPYING
# or http://www.gnu.org/copyleft/gpl.txt .
# For the initials of contributors, see ~abinit/doc/developers/contributors.txt .
#
# for all the following tests, use unix-style file paths
# they will be converted by function transpath if necessary
# file paths may contain a perl variable (e.g. $PSPS)
#
# The following lines describe the tests to be run. The general syntax of
# these lines is: [testID | end] [command | program] parameters.
#
# The first word is the test identification. All consecutive lines that begin
# with the same word will be executed in that order. "testID" can be 'end' to
# signal the normal end of a tests sequence: multiple 'end's are allowed.
# The second word specifies an ABINIT module, procedure or system command
# to be run. The following system commands are available: copy, erase
# The following keywords refer to ABINIT programs or procedure:
# abinit | seq exercise the sequential version of abinit
# aim exercise the atom-in-molecule Bader analysis
# ana | anaddb exercise anaddb
# band2eps exercise band2eps
# cpuserie run a serie of cpu tests of sequential version of abinit
# cut3d exercise cut3d
# conducti exercise conducti
# lwf exercise lwf
# macroave exercise macroave
# mrg | mrgddb exercise mrgddb
#
# parameters for copy and erase are file or directory names
#
# parameters for mrgddb must comply to the following format:
# [fn1] 'short description' (fn2 fn3 ... fnLast)
# fn2 ... fnLast can reside in the Test directory as normal or g-zipped files
# they will be copied, and unzipped if necessary, to the working directory
# before the test and the copy will be deleted after
#
# parameters for the other programs are either file paths derived from the
# subroutines 'doseqtest' ... in the old Run.pl scripts, or
# options to tailor the flow of results analysis and other operations before
# or after the tests;
# format for file paths is: fnI=path
# format for pseudo-potential file paths is: pspI=path
# format for the GEO file is: GEO=path
# format for the DS3_DOS_AT000x file name is: DS3AT=path
# format for FFT algorithm for a cpu tests serie is: fftalg=ABC
# format for options is: opt=option
# option can be either a standard option of the fldiff script (-easy , -medium , -ridiculous),
# options to tailor the analysis of the results after test completion:
# -nofld don't run fldiff
# -diflog apply diff/fldiff to the .log file
# -DOS apply diff/fldiff to the o_DOS file
# -DS2 apply diff/fldiff to the o_DS2_DOS file
# +xml also apply diff/fldiff to the .cml file
# or the following options to modify the flow of operations during the test:
# -keepfn4
# -istwfk1 for a cpu tests serie
# -istwfk11 for a cpu tests serie
#
# default options to call fldiff are: -ignore -ignoreP
#
# Default file names for the test with Id NNN (1 to 3 characters string):
# 1) mrg/mrgddb
# fn1 = tNNN.ddb.out
# 2) abinit/seq
# fn1 = ../tNNN.in
# fn2 = tNNN.out
# fn3 = tNNNi
# fn4 = tNNNo
# fn5 = tNNN
# fn6-fn9,psp1-psp4 = none
# 4) ana/anaddb
# fn1 = ../tNNN.in
# fn2 = tNNN.out
# fn3 = tNNN.ddb.in
# fn3 can reside in the Test directory as a normal or g-zipped file
# it will be copied, and unzipped if necessary, to the working directory before
# the test and the copy will be deleted after
# fn4 = tNNN.md
# 5) cut3d
# fn1 = ../tNNN.in
# fn2 = tNNN.out
# 6) aim
# fn1 = ../tNNN.in
# fn2 = tNNNi_DEN
# fn3 = tNNN
# fn4-fn5,psp1-psp2 = none
# 7) conducti
# fn1 = ../tNNN.in
# fn2 = tNNN.out
# fn3 = tNNN
# 8) lwf
# fn1 = ../tNNN.in
# fn2 = ../tNNN.moldyn.in
# fn3 = tNNN.out
# fn4 = tNNN.wandata
# 9) band2eps
# fn1 = ../tNNN.in
# fn2 = tNNN.out.eps
# fn3 = ../tNNN.in.freq
# fn4 = ../tNNN.in.displ
# 10) macroave
# fn1 = ../tNNN.in
# fn2 = tNNN_DEN.MAV
# fn3 = tNNN_DEN.PAV
# fn4 = tNNN
# 11) cpuserie [Xnn]
# fn1 = inputX
# fn2 is the Reference file which name is built with the serie letter X [A-D]
# followed by a number; the default value for this number is 3
# fn2 = reportX3
#
# ***********************************************************************
# see the comments of the doserieX subroutine in the script RunTests.pl #
# for explanations of the operations #
# ***********************************************************************
A1 cpuserie fftalg=101 opt=-istwfk11
A2 cpuserie fftalg=200 opt=-istwfk11
A3 cpuserie fftalg=112 opt=-istwfk11
A4 cpuserie fftalg=401 opt=-istwfk11
A5 cpuserie fftalg=412 opt=-istwfk11
# fn2 number will be 6 for series number 6 or higher
A6 cpuserie fn2=reportA6 fftalg=112
A7 cpuserie fn2=reportA6 fftalg=412
B1 cpuserie fftalg=101 opt=-istwfk1
B2 cpuserie fftalg=200 opt=-istwfk1
B3 cpuserie fftalg=112 opt=-istwfk1
B4 cpuserie fftalg=401 opt=-istwfk1
B5 cpuserie fftalg=412 opt=-istwfk1
# fn2 number will be 6 for series number 6 or higher
B6 cpuserie fn2=reportB6 fftalg=112
B7 cpuserie fn2=reportB6 fftalg=412
C3 cpuserie
# Note that case D uses inputB
D3 cpuserie fn1=inputB opt=-istwfk1