I strongly suggest adding "-assu buff" to buffer I/O, sometimes much, much faster.<br><br>N.B., it is not obvious to me that the compilation options in your two cases should matter. Except for the known bad versions of recent Intel compilers, buffering I/O and -r8 (or equivalent) they should have little effect.<br>
<br>On Friday, February 3, 2012, Gavin Abo <<a href="mailto:gsabo@crimson.ua.edu">gsabo@crimson.ua.edu</a>> wrote:<br>> First, thanks to everyone that provided feedback on this lapw2 read_vec error.<br>><br>> It seems that the cause of the runtime error was the compiler options. The compiler options that resolved and caused the error in Wien2k version 11 (and version 10) are given below. <br>
><br>> Working Compiler Settings:<br>><br>> O Compiler options: -FR -mp1 -w -prec_div -pc80 -pad -align -DINTEL_VML -traceback<br>> L Linker Flags: $(FOPT) -L/opt/intel/mkl/<a href="http://10.1.0.015/lib/32">10.1.0.015/lib/32</a> -pthread<br>
> P Preprocessor flags '-DParallel'<br>> R R_LIB (LAPACK+BLAS): -lmkl_lapack -lmkl -lguide<br>> RP RP_LIB(SCALAPACK+PBLAS): -lmkl_scalapack -lmkl_blacs -L/opt/local/fftw/lib/ -lfftw_mpi -lfftw $(R_LIBS)<br>
> FP FPOPT(par.comp.options): $(FOPT)<br>> MP MPIRUN commando: mpirun -np _NP_ -machinefile _HOSTS_ _EXEC_<br>><br>> Compiler Settings that Caused the Error:<br>><br>> O Compiler options: -FR -mp1 -w -prec_div -pc80 -pad -ip -DINTEL_VML -traceback<br>
> L Linker Flags: $(FOPT) -L/opt/intel/mkl/<a href="http://10.1.0.015/lib/32">10.1.0.015/lib/32</a> -lpthread<br>> P Preprocessor flags '-DParallel'<br>> R R_LIB (LAPACK+BLAS): $(FOPT) -L/opt/intel/mkl/<a href="http://10.1.0.015/lib/32">10.1.0.015/lib/32</a> -lmkl_lapack -lmkl -lguide -lvml -pthread<br>
> RP RP_LIB(SCALAPACK+PBLAS): -lmkl_scalapack -lmkl_blacs -L/opt/local/fftw/lib/ -lfftw_mpi -lfftw $(R_LIBS)<br>> FP FPOPT(par.comp.options): -FR -mp1 -w -prec_div -pc80 -pad -ip -DINTEL_VML -traceback<br>> MP MPIRUN commando: mpirun -np _NP_ -machinefile _HOSTS_ _EXEC_<br>
><br>> Notes: 64 bit SUSE Linux with 32 bit ifort/mkl. RP_LIB, FPOPT, and MPIRUN should only apply if using "MPI and Scalapack installed" option and with FFTW library.<br>><br>> On 1/30/2012 8:53 AM, Laurence Marks wrote:<br>
><br>> 2012/1/30 Gavin Abo <<a href="mailto:gsabo@crimson.ua.edu">gsabo@crimson.ua.edu</a>>:<br>><br>> Dear Dr. Blaha or developers,<br>><br>> I got the case.vector file from Arqum. When I do a single program run with<br>
> the file on my Debian squeeze IFCE Update 8 AMD processor system, I get the<br>> "double free or corruption" error below. If I do a single program lawp1,<br>> the case.vector file gets rewritten, then I can run lawp2 without any<br>
> problems on my system.<br>><br>> This is probably because there are differences in the detailed<br>> structure of unformatted files from system to system. Hence copying<br>> the vector file to your system is probably not going to work.<br>
><br>> Arqum was using a Wien2k version 10, and the TiC example ran without any<br>> problems. The only system condition that should have changed is the use of<br>> Wien2k version 11. So it seems like there may be some new code in version<br>
> 11 that his compiler does not handle in the same way. So it seems that the<br>> compiler is making an executable that generates a bad case.vector file. Any<br>> ideas on the possible cause in the code and a fix?<br>
><br>> In 99% of cases problems are due to bad input; in a few cases bad<br>> compilation. As has been said by others, so far insufficient<br>> information has been provided for anything except wild guesses.<br>
><br>> Commandline: x lapw2<br>> Program input is: ""<br>><br>> *** glibc detected *** /home/gavin/WIEN2k/lapw2: double free or corruption<br>> (!prev): 0x0000000002a25660 ***<br>> ======= Backtrace: =========<br>
> /lib/libc.so.6(+0x71bd6)[0x2b4b3d7debd6]<br>> /lib/libc.so.6(cfree+0x6c)[0x2b4b3d7e394c]<br>> /home/gavin/WIEN2k/lapw2[0x4c52ec]<br>> /home/gavin/WIEN2k/lapw2[0x40b266]<br>> /home/gavin/WIEN2k/lapw2[0x46b823]<br>
> /home/gavin/WIEN2k/lapw2[0x474a48]<br>> /home/gavin/WIEN2k/lapw2[0x403d6c]<br>> /lib/libc.so.6(__libc_start_main+0xfd)[0x2b4b3d78bc8d]<br>> /home/gavin/WIEN2k/lapw2[0x403c69]<br>> ======= Memory map: ========<br>
> ...(long list of the memory map that I removed here)<br>> Aborted<br>> 0.4u 0.0s 0:00.45 100.0% 0+0k 0+384io 0pf+0w<br>> error: command /home/gavin/WIEN2k/lapw2 lapw2.def failed<br>><br>> Gavin<br>
><br>><br>> On 1/30/2012 1:45 AM, arqum hashmi wrote:<br>><br>> Dear prof. P BLAHA,<br>><br>> yes this error is reproducible and again occured when i run another scf<br>> cycle.<br>><br>> lapw1.def gives this output<br>
> 4,'TiC.klist', 'unknown','formatted',0<br>> 5,'TiC.in1', 'old', 'formatted',0<br>> 6,'TiC.output1','unknown','formatted',0<br>
> 10,'./TiC.vector', 'unknown','unformatted',9000<br>> 11,'TiC.energy', 'unknown','formatted',0<br>> 18,'TiC.vsp', 'old', 'formatted',0<br>
> 19,'TiC.vns', 'unknown','formatted',0<br>> 20,'TiC.struct', 'old', 'formatted',0<br>> 21,'TiC.scf1', 'unknown','formatted',0<br>
> 55,'TiC.vec', 'unknown','formatted',0<br>> 71,'TiC.nsh', 'unknown','formatted',0<br>> 200,'./TiC.storeHinv', 'replace','unformatted',9000<br>
><br>><br>> lapw2.def give this<br>> 2,'TiC.nsh', 'unknown','formatted',0<br>> 3,'TiC.in1', 'unknown','formatted',0<br>> 4,'TiC.inso', 'unknown','formatted',0<br>
> 5,'TiC.in2', 'old', 'formatted',0<br>> 6,'TiC.output2','unknown','formatted',0<br>> 7,'TiC.vorb','unknown','formatted',0<br>> 8,'TiC.clmval','unknown','formatted',0<br>
> 10,'./TiC.vector', 'unknown','unformatted',9000<br>> 11,'TiC.weight', 'unknown','formatted',0<br>> 13,'TiC.recprlist', 'unknown','unformatted',9000<br>
> 14,'TiC.kgen', 'unknown','formatted',0<br>> 15,'TiC.tmp', 'unknown','unformatted',0<br>> 16,'TiC.qtl', 'unknown','formatted',0<br>
> 17,'TiC.weightaver','unknown','formatted',0<br>> 18,'TiC.vsp', 'old', 'formatted',0<br>> 19,'TiC.vns', 'unknown','formatted',0<br>
> 20,'TiC.struct', 'old', 'formatted',0<br>> 21,'TiC.scf2', 'unknown','formatted',0<br>> 22,'TiC.rotlm', 'unknown', 'formatted',0<br>
> 23,'TiC.radwf', 'unknown', 'formatted',0<br>> 24,'TiC.almblm', 'unknown', 'formatted',0<br>> 26,'TiC.weigh', 'unknown','unformatted',0<br>
> 27,'TiC.weighdn', 'unknown','unformatted',0<br>> 29,'TiC.energydn','unknown','formatted',0<br>> 30,'TiC.energy', 'unknown','formatted',0<br>
> 32,'TiC.qdmft', 'unknown', 'formatted',0<br>> 34,'TiC.oubwin', 'unknown', 'formatted',0<br>> 231,'TiC.dmftsym', 'unknown', 'formatted',0<br>
><br>> and also i have enough memory to work on wien2k.<br>><br>><br>> now is this information is enough to solve that what is going wrong.<br>><br>> thanks and Regards<br>> ________________________________<br>
> From: Peter Blaha <<a href="mailto:pblaha@theochem.tuwien.ac.at">pblaha@theochem.tuwien.ac.at</a>><br>> To: A Mailing list for WIEN2k users <<a href="mailto:wien@zeus.theochem.tuwien.ac.at">wien@zeus.theochem.tuwien.ac.at</a>><br>
> Sent: Sunday, January 29, 2012 11:28 PM<br>> Subject: Re: [Wien] urgent help required<br>><br>> We need more information:<br>><br>> Is this error reproducible ? Does it occur again when you run another scf<br>
> cycle (run_lapw) ?<br>><br>> cat lapw1.def<br>> cat lapw2.def<br>><br>> ls -alsrt *vector (or if you are using $SCRATCH: ls -alsrt<br>> $SCRATCH/*.vector )<br>><br>> do you have enough disk space ( df) ?<br>
><br>><br>> Am 30.01.2012 07:03, schrieb arqum hashmi:<br>><br>> Dear wien2k users ,<br>> my system is Intel(R) Xeon(R) CPU E5440 @ 2.83GHz. i used suse linux.<br>> i used ifort compiler 10.1.008 and mkl library is 10.1.0.015.<br>
> i installed the wien2k version 11 succesfully.<br>> after installing succesfully, i run the example of TiC. when i run the Scf<br>> cycle it gives this error below.<br>><br>> LAPW0 END<br>> LAPW1 END<br>
> forrtl: severe (174): SIGSEGV, segmentation fault occurred<br>> Image PC Routine Line Source<br>> lapw2 0812197B Unknown Unknown Unknown<br>> lapw2 0809F9CF read_vec_ 99<br>
> read_vec_tmp_.F<br>> lapw2 0807FB2E l2main_ 632 l2main_tmp_.F<br>> lapw2 0809344B MAIN__ 564 lapw2_tmp_.F<br>> lapw2 0804A491 Unknown Unknown Unknown<br>
> libc.so.6 5620D8AC Unknown Unknown Unknown<br>> lapw2 0804A3D1 Unknown Unknown Unknown<br>><br>> stop error<br>><br>> please tell me why this error occured and whats its solution. i will be<br>
> very thankful to you.<br>> thanks and regards<br>> arqum hashmi<br>><br>><br>><br>><br>><br>><br>><br>><br><br>-- <br>Professor Laurence Marks<br>Department of Materials Science and Engineering<br>
Northwestern University<br><a href="http://www.numis.northwestern.edu" target="_blank">www.numis.northwestern.edu</a> 1-847-491-3996<br>"Research is to see what everybody else has seen, and to think what nobody else has thought"<br>
Albert Szent-Gyorgi<br>