Known problems This file contains the list of known problems at the time a version of ABINIT is released. Note that bugs discovered later than the release are NOT described here... (In particular, at the time of release, a particular version is not tested already on ALL platforms, only on a subset of 5, usually). For version 5.2.4, list closed on 20061128 after installation and tests for: seq and paral PC/Linux with PGI compiler (~ABINIT/doc/config/build-examples/i686-pgi4.0_sleepy.ac) seq PC/Linux with PGI compiler (~ABINIT/doc/config/build-examples/i686-pgi4.0_dummy.ac) seq and paral PC/Linux with IFC compiler (~ABINIT/doc/config/build-examples/i686-intel8.1_sleepy.ac) seq PC/Linux with IFC compiler (~ABINIT/doc/config/build-examples/i686-intel9.0_dummy.ac) seq PC/Linux with g95 compiler (~ABINIT/doc/config/build-examples/i686-g95_sleepy.ac) seq PC/Linux with pathscale compiler (~ABINIT/doc/config/build-examples/i686-pathscale_sleepy.ac) seq and paral Itanium2 machine with IFC compiler (~ABINIT/doc/config/build-examples/ia64-intel8.1_chpit.ac) seq and paral Compaq/Dec EV67 OSF (~ABINIT/doc/config/build-examples/alphaev67-compaq_deccint.ac) seq and paral Compaq/Dec EV56 Linux (~ABINIT/doc/config/build-examples/alphaev56-compaq_tux.ac) seq SGI Octane IRIX (~ABINIT/doc/config/build-examples/mips-mipspro_spinoza.ac) seq and paral IBM Power 5 (~ABINIT/doc/config/build-examples/powerpc64-ibm_fock.ac) seq IBM Power 4 (~ABINIT/doc/config/build-examples/powerpc-ibm_dirac.ac) seq and paral Mac OS X, xlf compiler (~ABINIT/doc/config/build-examples/powerpc-ibm_max.ac) seq Opteron with Intel compiler (~ABINIT/doc/config/build-examples/opteron-intel9.1_lemaitre.ac) Compilation of Netcdf was not tested, and should NOT be attempted by non-expert users with this ABINITv5.2 Compilation of LibXC was not tested, and should NOT be attempted by non-expert users with this ABINITv5.2 Compilation of XMLF90 was not tested, and should NOT be attempted by non-expert users with this ABINITv5.2 Speed was tested. The present release 5.2.4 is suitable for production runs. All the problems are listed cumulatively. Those specific to version 5.2 are found at the end of this file =============================================== The type of the problem is identified as follows : NUMERICAL ERROR : the physical result is incorrect, but the job does not crash. Moreover, the use of the input variables is legitimate. This is a very dangerous type of problem. There are two subclasses : - INTERNALLY INCOHERENT (the bug can be seen because two sets of input variables should give the same results and do not) - INTRINSIC (the bug does not cause incoherency, it can only be seen when comparing results with those of another code or with experimental results) The latter one is the most dangerous, of course ... CRASH : the code does not produce the result, because it stops unduly. SLOWING DOWN : the physical result is correct, but the code does not go as fast as it should. For example, the reading of a restart file might not be correct, or the SCF convergence might not be as fast as it should. OTHER : some information (not numerical results) might be erroneous ... In each class of problem, one can identify MACHINE-DEPENDENT problems (on some platform, this problem does not occur, not on others) as well as GENERAL problems. Some of the problems listed below have been reported by user, but have not yet been examined for confirmation, in which case they are labelled "TO BE CONFIRMED". ================================================= P0. (Many reports) SLOWING DOWN The iterative procedures (SCF and determination of geometry) still present instabilities, in particular for metallic slabs P2. (XG, 001223 + later) OTHER (TO BE DONE) GENERAL The memory needs are not yet evaluated correctly for the spin-orbit case and non-collinear magnetism case. P8. (GMR, November 2000) SLOWING DOWN GENERAL restartxf for optcell/=0 might have a bug, despite Test_v1#80 Identified in March 2001 by GMR : only happens when restarting from a different, non-zero, optcell. Should be corrected, but documented, now. P9. (Massimilio Stengel, October 2000) CRASH MACHINE-DEPENDENT : Compaq/DEC Problem with non-linear XC correction, it seems P15. (XG, 010103) OTHER GENERAL Should enforce coding rule related to length of file P117. (Laurent Pizzagalli, 010622) CRASH Observed on PC AMD, atomic relaxation (ionmov==7), no cell optimisation, stops in xfpack complaining about ndim. See mail 010622. P118. (XG, 010710) OTHER GENERAL (d/dk by finite differences) Test_v3 #5 . Sensitivity to phase choice. P121. (GMR, 010801) NUMERICAL RESULT MACHINE_DEPENDENT (Compaq ES40, Turing) The use of fourdp , with fftalg 200 or 1xx gives complex conjugate results. This was seen in the process of merging the GW code with ABINIT. To be examined in detail. P204. (XG, 011020) OTHER (TO BE DONE) MACHINE_DEPENDENT Test_v3#14 : the output file is machine-dependent, because the spin-phase of spin-degenerate wavefunctions has not been fixed... Or, is there something more fundamental ?? P205. (XG, 011020) SLOWING DOWN MACHINE_DEPENDENT ?? On a Intel/Linux PIII 900 MHz, with PGI compiler, slowing down in getghc, due to the use of datastructures... For example, 18% of the total time in Test_v3/t10.out ... Argh ... even more in Test_v2/t30.out ... P220. (XG, 020114) OTHER GENERAL Refs files in Test_physics should be updated. P230. (MTijssens, 020306) TIMING PROBLEM MACHINE-DEPENDENT (P6, with IFC compiler v5.0) The wall clock timing of one routine can be completely wrong. Seems to come from the date_and_time intrinsics, called by timein.f . Previously, it was even causing crash, in the headwr.f routine. P250. (LSi, 020521) CRASH MACHINE-DEPENDENT (DOS/Windows) Access violation Test_v2#7,26,30,90,92,96 and Tutorial#55,56 P302. (XG, 020305; DDKlug, 020313) NUMERICAL ERROR MACHINE_DEPENDENT (IBM 44P; SGI) RF GGA : Test_v3#8,16,18,60,61,62,66 Problem with Response Function GGA tests. The accuracy of the RF GGA Test_v3#16 is very bad on these machines : the acoustic modes have large frequencies... P331. (XG, 020528) SLOWING DOWN MACHINE_DEPENDENT (MACOSX) Test_v2 #53 and 54 The computation of the dielectric matrix is numerically incorrect. Fortunately, this dielectric matrix is only used to precondition the SCF cycle. P332. (XG,020528) OTHER MACHINE_DEPENDENT (MACOSX) Test_v3 #31 and 32 The Fermi energy is not printed out correctly. P349. (XG021016) OTHER LIKELY GENERAL Test_v3 #77. This test is not yet portable : accuracy of fldiff was already tuned, but still problems of portability. P350. (XG030225) OTHER GENERAL psp*cc.f the fifth derivative is not computed. P402. (XG021227) NUMERICAL ERROR LIKELY GENERAL Lattice Wannier Function only Test_v3#96 and 97 are still undergoing changes P404. (XG030103) NUMERICAL ERROR LIKELY GENERAL fftalg 401 works, but gives a slightly different answer than usual for Tfast#10. To test this, replace the default value in indefo.f P405. (XG030106) CONVERGENCE PROBLEM LIKELY GENERAL Test_v3#15 : with very well non-SCF converged wavefunction (nnsclo=2), the total energy at intermediate steps is lower than the final energy ! P406. (XG030106) CONVERGENCE PROBLEM LIKELY GENERAL Test_v3#85 : set iscf to 5 in the RF, one sees a very bad convergence... P407. (XG030108) NUMERICAL ACCURACY INTEL with IFC compiler, EV67 under OSF, EV56 under Linux Test_v3#14,77,85 with diffs compared to refs, that are a bit too large. P411. (XG030515) PORTABILITY PROBLEM INTEL with IFC compiler, EV67 under OSF, EV56 under Linux Test_v4#40 differences compared to refs are a bit large Test_v4#42 sign of some eigenvectors might not be portable P416. (XG030522) CRASH Seen on Intel/PC with PGI compiler, in parallel Test_v1#75,80 do not work Problem to read XFhistory part of a wavefunction file. P41.9. (XG030719) PORTABILITY PROBLEM GENERAL Test_v4#37 is not portable : the spin decomposition of the spinor wavefunctions is not gauge invariant. This test has been disabled. P42.7. (DH030929) OTHER Should examine the d2sym3.f routine, giving extra zeros in the output file. See mail from DHamman, on 2003-9-29. P43.3. (XG040218) PORTABILITY of AUTOMATIC TESTS On P6 with INTEL compiler, DEC EV67, DEC EV56 Test_v4#54,67,68,71,75,77,78 and Tutorial#61-69 : Accuracy criterion to be adjusted, or portability improved. P43.12. (DRH040227) NUMERICAL ERRORS On a PC with IFC compiler version 5.0.1 Test_v4#77 Erroneous polarization P43.14. (XG040309) PORTABILITY On SGI "Spinoza" Test_v4#54 , anaddb analysis of eigenvectors gives a different sign. P44.2. (DHamann20040527) NUMERICAL INACCURACY For the treatment of pseudopotentials with pspcod=6 (FHI), the non-linear XC core charge is not treated consistently in psp6cc.f and in the rest of the code. See Don's mail to the developers of 20040527 See also input variable optnlxccc . P44.12 (MM20040901) PORTABILITY / NUMERICAL ERROR IFC v7.0 From Mikami-san Test_v4#62 : completely erroneous ?! P45.5. (XG050220) OTHER : PARALLEL TESTING Test_v4#90 (CASINO output) crash in parallel (test_paral on grumpy). P45.6. (XG050225) NUMERICAL PROBLEM (with ionmov=12) Test_v4#97 has one strange output P45.8. (XG050227) SCRIPT PROBLEM Too much errors observed when using the script parents : should change the rules or the script P45.14. (XG050619) Libxc from Miguel is still to be tested. P45.16. (N. Choudhury 050107) In case of a very high symmetry unit cell, with atoms placed such that the space group is of low symmetry, there is a problem to recognize the space group. This is illustrated by Test_v4#98. P45.17 (B. Andriyevsky 050405) Problem of normalisation of the epsilon'' from the "conducti" code ?! A factor of pi (3.1416) might be missing. See his mail of 5 April 2005 to the forum. See also other mails exchanged on the forum. P45.18 (D Hamann 050510) NUMERICAL ERROR With finite-difference ddk calculation See his mail of 10 May, 2005. (Excerpt : ... The problem is that 4.4.4 and 4.5.2 WILL run drf.in and produce incorrect results, all related to the dataset 3 finite-differend ddk calculation. The results are only somewhat incorrect, not obvious garbage. Here is a small sample, Born charges from "2nd-order matrix" in the .out files: Correct (from erf.out): 1 1 1 4 -4.1367712586 0.0000000000 1 1 2 4 -0.0863319602 0.0000000000 1 1 3 4 0.0706059161 0.0000000000 Incorrect (from drf.out): 1 1 1 4 -3.4782061716 0.0000000000 1 1 2 4 -0.0832639192 0.0000000000 1 1 3 4 0.0560776887 0.0000000000 Putting in some istwfk printing, I found that setting berryopt/=0 now forces istwfk=1 for dataset 3. The problem apparently is that dataset 2 has some istwfk>1, and apparently dataset 3 is getting incorrect WFK's from dataset 2 and using them. Why it doesn't re-converge them first isn't clear. Setting istwfk=1 in dataset 2 (erf.in) fixes the problem and gets correct results. (My hack also had the effect of setting istwfk=1 in dataset 2 as well as 3.) Curiously, removing the dataset 2 calculation altogether and feeding the kptopt1=1 WFK's from dataset 1 directly to the finite-difference ddk calculation (nrf.in) also fixes the problem, despite the fact that dataset 1 has some istwfk>1. Clearly something bad is going on in WFK readin, but only in certain cases. I've never looked at this part of the code, and I'm afraid I can't divert any more time now towards tracking this down. P46.1. (XG050701) PORTABILITY PROBLEM Test tutorial Finite electric field alas_relaxdion.in (the last one) does not work on my PC : it stops before the end. P46.4. (XG050714) NUMERICAL INACCURACY / PORTABILITY Test_v2#13,15,22 Relaxed ion dielectric tensor varies quite a lot between PC/IFC8.1 and PC/PGI4.0.2 P46.6. (XG050716) MISC. Variables nqptdm and qptdm should be substituted with nqpt and qpt. P46.7. (XG050716) PORTABILITY DEC EV67 (Deccint) afterscfloop.F90 only compiles with -O1 . dyfnl3.F90 only compiles with -O1 . P46.8. (XG050716 - modified XG050723) PORTABILITY Itanium 2 (Chpit) Test_v4#30, initialisation differs widely from other machines ?! P46.15. (XG051002) (PORTABILITY) Tutorial#optic_1 and optic_3, on SGI (Spinoza) and IBM 44P (Dirac). Some of the printed residuals are 10-310 . Other results seem fine. P46.19. (XG051003) (PORTABILITY) Test_v4#66 on IBM 44P (Dirac) Strain response not accurate in the DDB P46.21. (XG051017) Memory leaks Using the g95 compiler with debugging option, makefile_macros.g95_sleepy_debug (-g) Identified remaining memory leaks in Test_v2#01 : hdr not cleaned (init in hdr_init routine) Test_v3#30 : hdr not cleaned (init in hdr_io routine) Test_v4#50 : in sortph.F90 P50.1 (XG051130) OTHER : PARALLEL CASE in tests on sleepy with PGI compiler, make tests_abinip has the following problems : tutorial#ffield_6 numerical result is wrong tutorial#gw_7 The diff analysis cannot be done : the number of lines to be analysed differ. v1#75 and 80 did not read (x,f)history v2#81,82 Berry phase with positive berryopt is not parallelized v3#03,04,05 Berry phase with positive berryopt is not parallelized v4#78 Fermi energy is wrong (finite electric field calculation) v4#90 outqmc parallelism is not presently supported P50.2 (XG051203) (PORTABILITY) On a PC/Linux, with g95 compiler. g95 in sequential is OK, parallel compilation is OK, but parallel execution fails : -P-0000 leave_test : synchronization done... chkinp: machine precision is 2.2204460492503131E-16 chkinp: Checking input parameters for consistency. -P-0000 -P-0000 ================================================================================ Actually, this problem is already present in v4.6.5 . Parallelisation with g95 never worked (this was also checked on intermediary patches leading to v4.6.5). P50.3 (XG051203, confirmed XG060212, adapted XG060223, XG060311) (PORTABILITY) On a PC/linux, With pathscale compiler. The binaries cannot be linked properly when MPI is enabled. One has to use enable_mpi="no" in the ~/.abinit/build/*.ac file . (error as in v5.0) Likely due to an error in the script util/build/wrap-mpif90 . The error message is /usr/bin/ld: skipping incompatible /usr/local/mpich-1.2.4-pathscale-2.1-64/lib/libmpichf90.a when searching for -lmpichf90 /usr/bin/ld: cannot find -lmpichf90 (error as in v5.1.1) make[3]: Entering directory `/home/gonze/ABINIT/ABINITv5.1.1/abinit-5.1.1_pathscale/src/01managempi' /opt/pathscale/bin/pathf90 -m64 -DHAVE_CONFIG_H -I. -I. -I../.. -I../defs -DMPI -O2 -march=opteron -msse2 -extend-source -fno-second-underscore -extend-source -fno-second-underscore -c -o lib01managempip_a-defs_xderive.o `test -f 'defs_xderive.F90' || echo './'`defs_xderive.F90 module defs_xderive ^ pathf90-855 pathf90: ERROR DEFS_XDERIVE, File = defs_xderive.F90, Line = 48, Column = 8 The compiler has detected errors in module "DEFS_XDERIVE". No module information file will be created for this module. include 'mpif.h' pathf90-63 pathf90: ERROR XDERIVEWRITE_INT, File = defs_xderive.F90, Line = 154 Cannot open INCLUDE file "mpif.h". ... (error as in v5.1.2) /usr/local/mpi-pathscale/bin/mpif90 -O2 -march=opteron -msse2 -extend-source -fno-second-underscore -o abinis abinit.o ../../src/11drive/lib11drive.a ../../src/08seqpar/lib08seqpars.a ../../src/07suscep/lib07suscep.a ../../src/06response/lib06response.a ../../src/06geomoptim/lib06geomoptim.a ../../src/05gw/lib05gw.a ../../src/05common/lib05common.a ../../src/04occeig/lib04occeig.a ../../src/04iowfdenpot/lib04iowfdenpot.a ../../src/04wfs/lib04wfs.a ../../src/03ionetcdf/lib03ionetcdf.a ../../src/03iovars/lib03iovars.a ../../src/03paw/lib03paw.a ../../src/03recipspace/lib03recipspace.a ../../src/03xc/lib03xc.a ../../src/03xml/lib03xml.a ../../src/03nonlocal/lib03nonlocal.a ../../src/02nlstrain/lib02nlstrain.a ../../src/02ffts/lib02ffts.a ../../src/02psp/lib02psp.a ../../src/02geometry/lib02geometry.a ../../src/02parser/lib02parser.a ../../src/02spacepar/lib02spacepars.a ../../src/lib01fftnew/liblib01fftnews.a ../../src/01contract/lib01contract.a ../../src/01managempi/lib01managempis.a ../../src/01util/lib01util.a ../../src/00basis/lib00basis.a ../../src/defs/libdefs.a -L../../lib/numeric -lnumeric -L../../lib/numericf90 -lnumericf90 -L../../lib/lapack -llapack -L../../lib/blas -lblas /usr/bin/ld: skipping incompatible /usr/local/mpich-1.2.4-pathscale-2.1-64/lib/libmpichf90nc.a when searching for -lmpichf90nc /usr/bin/ld: skipping incompatible /usr/local/mpich-1.2.4-pathscale-2.1-64/lib/libmpichf90nc.a when searching for -lmpichf90nc /usr/bin/ld: cannot find -lmpichf90nc collect2: ld returned 1 exit status P50.4 (moved to P51.7 and P51.8) P50.5 (XG051206) (PORTABILITY) Deccint alphaev67 tutorial#nlo_2 The numerical accuracy is quite bad (sixth digit). P50.6 (closed) P50.7 (XG051209) NQXC and XMLF90 libraries In order to work properly, these libraries need the AC_FC_WRAPPERS line to be uncommented in configure.ac . However, in that case, the config does not work for the PGI compiler. P50.8 (closed) P50.9 (closed) P50.10 (closed) P50.11 (closed) P50.12 (closed) P50.13 (closed) P50.14 (closed) P50.16 (XG061215) DOC Many documentation files should be updated still ... Although most of the files put on the Web (.html) are OK. In particular : doc/developers/dirs_and_files is to be updated P51.1 (XG060221) (PORTABILITY) The tests Tutorial#paw1* present too large variations of precision (sometimes at a few digit level). The number of lines produced by fldiff is quite large. P51.2 (closed) P51.3 (closed) P51.4 (XG060222) (PORTABILITY) On chpit (an itanium II - ia64 machine). Problem with the Optic utility. Test_v4#57 is wrong, at the level of the second digit. Tutorial#optic_2 is wrong, at the level of the second digit. P51.5 (closed) P51.6 (closed) P51.7 (closed) P51.8 (closed) P51.9 (closed) P51.10 (XG060311) SLOWING DOWN On tux, in parallel, some routines (inwffil, vtorho3:SYNCHRO) take a huge amount of time. Speed should be checked systematically before this version can be put in production. P51.11 (XG060312) PORTABILITY PAW : numerical results seem quite sensitive, for the all series tutorial/paw1_* P51.12 (closed) P51.13 (closed) P51.14 (closed) P51.15 (XG060520) In tests dirs, execute make tests_fast start=04 stop=04 Problem : the log file is t03.log, the diff file is numbered 03 as well ... ?! This only happens for 04 , e.g. make tests_fast start=03 stop=03 is fine make tests_fast start=05 stop=05 is fine !!! P51.16 (MTorrent060505) In the specific setting of directories used by MTorrent, Problem Tv3#68 and Tv3#69 does not work, because the CML file is not at the correct place P51.17 (closed) P51.18 (closed, see P51.5) P51.19 (closed) P51.20 (closed) P51.21 (closed) P51.22 (MGiantomassi060627) The new value of rprimd is not stored in _WFK file . So, restart cannot benefit from modifications of acell. See his mail of June 27, 2006. See also P8. P52.1 (XG060719) PORTABILITY / UPDATE Test_v2#15 : fluctuations beyond allowed tolerance, Ref file not updated from 5.1 (all platforms) Test_v2#48 : fluctuations beyond allowed tolerance, Ref file not updated from 5.1 (all platforms) Tutorial#elphon_5 : too inaccurate results, Ref file not updated from 5.1 (all platforms) Test_v4#48 : Wannier functions calculations, under g95 P52.2 (closed) P52.3 (closed) P52.4 (XG060720) PORTABILITY Test_paral#M (GW in parallel) crash on Decci (Compaq/Dec EV67), as well as on max (PowerPC Mac OS X xlf). (this problem does not happen on Tux EV56 under Linux) P52.5 (XG060809) PORTABILITY On chpit (an itanium II - ia64 machine) Tutorial#paw1_6 gives errors on the order of 1.0d-4, too large ... P52.6 (XG060809) PORTABILITY On chpit (an itanium II - ia64 machine) Test_v3#05 is completely wrong . d/dk by finite difference P52.7 (XG060809) PORTABILITY On chpit (an itanium II - ia64 machine) Test_v4#68 ANADDB , the three (small) eigenvalues are rather inaccurate, with a change of sign P52.8 (XG060813) PORTABILITY Tutorial ffield_5 on Fock (IBM Power5 4-procs) One of the SCF loops needs one more step -> comparison fails. But result is correct P52.9 (closed) P52.10 (closed) P52.11 (closed) P52.12 (XG061009) PORTABILITY Tv5#1 crash on Decci and Tux Compaq/Dec EV56 and EV67. P52.13 (XG061009) PORTABILITY Tparal#M fails on Fock (IBM Power 5).