From sprokuda at dnk.ru Mon Sep 1 08:20:49 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Mon, 1 Sep 2003 10:20:49 +0400 Subject: [Pw_forum] problem with compilation In-Reply-To: <20030831175559.19166.qmail@mail.com> References: <20030831175559.19166.qmail@mail.com> Message-ID: <129986377.20030901102049@dnk.ru> Dear Bachir! What version did you download ? if fftw 3.something - bad idea, use FFTW 2.1.5. I had just the same problem with a couple of weeks ago. As for the rest - ( undercores and manner of compiler to call the function) - see the manual. and a bit more ( the replies were given to me on the same problem) : 1) FFTW is written in C but fortran wrappers are also included. Have you succesfully compiled the FFTW package on your machine? What is the result of " nm /usr/local/lib/libfftw.a | grep f77 "? (replace /usr/local/lib/ with the name of the directory where you installed the libfftw.a library) Different f77 compilers use different name convenctions in order to call routines (lowercase,uppercase, add_one_underscore, add_two_underscores...), the "configure" step of fftw figures out what is your f77 compiler and use g77 if nothing else applies. If you have a fortran compiler which is not in the list checked by the configure script, after the configure step but before actually compiling the library (with the make command), you may need to edit the fftw/config.h file, look for the string /*********************** fortran wrapper configuration *********************/ and set the values appropriated for you fortran compiler. In any case if you succesfully compiled the package, no matter which f77 compiler the configure script found, the "nm /usr/local/lib/libfftw.a | grep f77 " should find something. Then you only need to discover how your fortran compiler likes to call routines (for instance by "nm pwscf.o"), set the fftw/config.h parameters accordingly and repeat the installation 2) First you have to download FFTW-sources (v.2.1.3) from www.fftw.org or www.netlib.org. Please be careful do not download the 3.0 version. It is not incompatible for the code. Then unpack it and go to the directory /your_FFTW_dir/fftw and edit the config.h file. At the end of the file please find lines which look the next ones (as should be). A line should be commented out. /* #undef FFTW_FORTRANIZE_LOWERCASE */ #define FFTW_FORTRANIZE_LOWERCASE_UNDERSCORE 1 /* #undef FFTW_FORTRANIZE_UPPERCASE */ /* #undef FFTW_FORTRANIZE_UPPERCASE_UNDERSCORE */ Then just press "configure", "make", "make install". The last, you should link the fftw-bibrary you are being installed before the MKL one, i.e. -lfftw -lMKL.. Probably the above mentioned procedure is not required for the v.1.2 (PWSCF), but you have to check it. Exactly you do not need the FFTW.3.0 version and you can remove it. ALT Linux is RedHat based one, so it should work good. Your troubles go to FFTW.3.0. Any question, please let me know. regards, Sergey. From sprokuda at dnk.ru Mon Sep 1 15:35:35 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Mon, 1 Sep 2003 17:35:35 +0400 Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <200308312159.27901.giannozz@nest.sns.it> References: <341736498.20030829163926@dnk.ru> <200308312159.27901.giannozz@nest.sns.it> Message-ID: <15319803268.20030901173535@dnk.ru> Dear Paolo , Recieved your advise about "tmp_dir" and "outdir" I've replaced the first with the second. But it didn't help - the problem is the same. But tell me - how it could be that the same input script works for Si, but it doesn't for Ge, while this structures are very similar? Best wishes, Prokudaylo Sergey. > Trying to run the examples I got "FORTRAN STOP 2" error message in > example2, in single mode calculations - what could it be and why ? it could be anything: fortran stop 2 means that fortran execution stopped with error 2 (whatever it means). For me, your scripts stops because of an obsolete variable name in a namelist ("tmp_dir" instead of "outdir"). From sprokuda at dnk.ru Mon Sep 1 15:31:59 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Mon, 1 Sep 2003 17:31:59 +0400 Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <20030829145450.457.qmail@web80512.mail.yahoo.com> References: <20030829145450.457.qmail@web80512.mail.yahoo.com> Message-ID: <1519587105.20030901173159@dnk.ru> Dear Eyvaz, My nscf output file is in attachment. To be honest I do not see there something like report about the convergence is not achieved. What do you meen by error "cdiaghg" - I've found nothing like that in forum - ... probably I was looking not for it not with proper understanding what to look for. Best wishes, Sergey Prokudaylo. -------------- next part -------------- A non-text attachment was scrubbed... Name: ge.nscf_011.in Type: application/octet-stream Size: 845 bytes Desc: not available Url : /pipermail/attachments/20030901/68a9e929/attachment.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: ge.nscf_011.out Type: application/octet-stream Size: 29753 bytes Desc: not available Url : /pipermail/attachments/20030901/68a9e929/attachment-0001.obj From giannozz at nest.sns.it Mon Sep 1 16:00:07 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 1 Sep 2003 16:00:07 +0200 Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <15319803268.20030901173535@dnk.ru> References: <200308312159.27901.giannozz@nest.sns.it> <15319803268.20030901173535@dnk.ru> Message-ID: <200309011600.07689.giannozz@nest.sns.it> On Monday 01 September 2003 15:35, sprokuda at dnk.ru wrote: > [...] But tell me - how it could be that the same input script works > for Si, but it doesn't for Ge, while this structures are very similar? the two scripts are not "the same", they are "very similar", which doesn't imply that both are correct > My nscf output file is in attachment. I don't see anything wrong in it Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From sprokuda at dnk.ru Mon Sep 1 16:08:11 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Mon, 1 Sep 2003 18:08:11 +0400 Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <200309011600.07689.giannozz@nest.sns.it> References: <200308312159.27901.giannozz@nest.sns.it> <15319803268.20030901173535@dnk.ru> <200309011600.07689.giannozz@nest.sns.it> Message-ID: <4821759001.20030901180811@dnk.ru> Dear Paolo! The scripts for Si I took exactly from examples to PWscf, that is why I'm wondering at so big difference in results. Nevertheless Ge is not Si, and probably the energy cuttoff of k-grid mush be changed for Ge in comparison to Si? And ,you see, I have error of the same type running the examples in phonon single mode calculations. Best wishes, Sergey Prokudaylo. From xhongjun at mail.ustc.edu.cn Mon Sep 1 15:51:08 2003 From: xhongjun at mail.ustc.edu.cn (xhongjun) Date: Mon, 01 Sep 2003 21:51:08 +0800 Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <15319803268.20030901173535@dnk.ru> References: <341736498.20030829163926@dnk.ru> <200308312159.27901.giannozz@nest.sns.it> <15319803268.20030901173535@dnk.ru> Message-ID: <3F534ECC.3010204@mail.ustc.edu.cn> Dear sprokuda: I have tried you input file and it runs well. The result is : Diagonalizing the dynamical matrix q = ( 0.000000000 0.100000000 0.100000000 ) ************************************************************************** omega( 1) = 0.728644 [THz] = 24.305102 [cm-1] omega( 2) = 0.899357 [THz] = 29.999510 [cm-1] omega( 3) = 1.296939 [THz] = 43.261503 [cm-1] omega( 4) = 8.777482 [THz] = 292.787245 [cm-1] omega( 5) = 8.797850 [THz] = 293.466635 [cm-1] omega( 6) = 8.798618 [THz] = 293.492246 [cm-1] ************************************************************************** But I don't know why such problem occur in your case. Good luck. sprokuda at dnk.ru wrote: > Dear Paolo , > Recieved your advise about "tmp_dir" and "outdir" I've replaced the > first with the second. But it didn't help - the problem is the same. > But tell me - how it could be that the same input script works for Si, > but it doesn't for Ge, while this structures are very similar? > Best wishes, Prokudaylo Sergey. > > >>Trying to run the examples I got "FORTRAN STOP 2" error message in >>example2, in single mode calculations - what could it be and why ? > > > it could be anything: fortran stop 2 means that fortran execution > stopped with error 2 (whatever it means). For me, your scripts > stops because of an obsolete variable name in a namelist > ("tmp_dir" instead of "outdir"). > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > From giannozz at nest.sns.it Mon Sep 1 16:16:33 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 1 Sep 2003 16:16:33 +0200 Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <4821759001.20030901180811@dnk.ru> References: <200309011600.07689.giannozz@nest.sns.it> <4821759001.20030901180811@dnk.ru> Message-ID: <200309011616.33550.giannozz@nest.sns.it> On Monday 01 September 2003 16:08, sprokuda at dnk.ru wrote: > And ,you see, I have error of the same type running the examples > in phonon single mode calculations. of "which" type ? "fortran stop 2" contains exactly the following bit of information: "something went wrong". What went wrong? Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From eyvaz_isaev at yahoo.com Mon Sep 1 16:59:58 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 1 Sep 2003 07:59:58 -0700 (PDT) Subject: [Pw_forum] Ge(qq0) In-Reply-To: <129986377.20030901102049@dnk.ru> Message-ID: <20030901145958.6181.qmail@web80512.mail.yahoo.com> Dear Sergei, Yes, you are right, there is no "chdiagh" error. Could you please repeat your calculations with Ge -0.125 -0.125 -0.125 Ge 0.125 0.125 0.125 instead of coordinates you used. In this case you have a Real Hamiltonian. Eyvaz. __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From eyvaz_isaev at yahoo.com Mon Sep 1 18:21:50 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Mon, 1 Sep 2003 09:21:50 -0700 (PDT) Subject: [Pw_forum] Germanuim [0qq] problem .... In-Reply-To: <341736498.20030829163926@dnk.ru> Message-ID: <20030901162151.51002.qmail@web80503.mail.yahoo.com> Dear Sergei, I have calculated phonons for Ge at the (0,0.1,0.1) point and have no problem. Just some comments. I am stiil using the 1.1.0 version (I am too lazy to edit my input files). Besides I used the lattice parameter 10.69585a.u.=5.66Ang which is experimental value (see Ashcroft, Mermin) instead of 10.626a.u. you used. I am using AMD/Athlon CPU and self-made LAPACK and ATLAS libraries. So, please pay attention to your configuration (say, the MKL library) and try LAPACK and BLAS instead MKL. Eyvaz. __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From g930102 at ccit.edu.tw Tue Sep 2 06:22:58 2003 From: g930102 at ccit.edu.tw (g930102 at ccit.edu.tw) Date: Tue, 2 Sep 2003 12:22:58 +0800 (CST) Subject: [Pw_forum] Re: [Pw_forum]Germanuim [0qq] problem Message-ID: <200309020423.h824MwHw084487@smtp.ccit.edu.tw> Dear Sergey, I have calculated phonons for Ge at the (0,0.1,0.1) point and have no problem too. I also got the same result as Mr. ( or Dr. ) xhongjun(nice job and quickly). Here I am use the Windows version of PCWscf ver. 1.2.0, I had post my makefile in the previous e-mail some days ago. I think the new version of Intel MKL Library would be acceptable in PWscf. I guess re-run and check the logfile would be help for find out somewhat. Have fun with PWscf! Best wishes, ======================================================= Lu Fu-Fa, Doctoral Student, E-mail: g930102 at ccit.edu.tw Semiconductors Lab., Dept. of Applied Physics Chung Cheng Institute of Technology (CCIT), Taiwan ======================================================= xhongjun's result (xhongjun ) Diagonalizing the dynamical matrix q = ( 0.000000000 0.100000000 0.100000000 ) ********************************************************************** omega( 1) = 0.728644 [THz] = 24.305102 [cm-1] omega( 2) = 0.899357 [THz] = 29.999510 [cm-1] omega( 3) = 1.296939 [THz] = 43.261503 [cm-1] omega( 4) = 8.777482 [THz] = 292.787245 [cm-1] omega( 5) = 8.797850 [THz] = 293.466635 [cm-1] omega( 6) = 8.798618 [THz] = 293.492246 [cm-1] ********************************************************************** Lu Fu-Fa's Result (g930102 at ccit.edu.tw) Diagonalizing the dynamical matrix q = ( 0.000000000 0.100000000 0.100000000 ) ********************************************************************** omega( 1) = 0.728662 [THz] = 24.305726 [cm-1] omega( 2) = 0.899357 [THz] = 29.999534 [cm-1] omega( 3) = 1.296772 [THz] = 43.255956 [cm-1] omega( 4) = 8.777482 [THz] = 292.787243 [cm-1] omega( 5) = 8.797848 [THz] = 293.466564 [cm-1] omega( 6) = 8.798617 [THz] = 293.492210 [cm-1] ********************************************************************** ___________________________________ NOCC, http://nocc.sourceforge.net -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ge.ph_011.in Url: /pipermail/attachments/20030902/d1ae41fc/attachment.txt -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ge.nscf_011.in Url: /pipermail/attachments/20030902/d1ae41fc/attachment-0001.txt -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ge.phG.in Url: /pipermail/attachments/20030902/d1ae41fc/attachment-0002.txt -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: ge.scf.in Url: /pipermail/attachments/20030902/d1ae41fc/attachment-0003.txt From bbouhafs at mail.com Tue Sep 2 13:01:18 2003 From: bbouhafs at mail.com (Bachir Bouhafs) Date: Tue, 02 Sep 2003 12:01:18 +0100 Subject: [Pw_forum] =?iso-8859-1?Q?Re:=A0[Pw=A0forum]=A0problem=A0with=A0compilation?= Message-ID: <20030902110118.55879.qmail@mail.com> Dear Sergey, I still have problem with pwscf compilation, at the end of the output of the make pw command, I get the following message: In function cft_3 undefined reference to 'fftw_f77' undefined reference to 'fftw_f77_create_plan' In function cft_1 undefined reference to 'fftw_f77' undefined reference to 'fftw_f77_create_plan' In function cft_1s undefined reference to 'fftw_f77' undefined reference to 'fftw_f77_create_plan' In function cft_2 undefined reference to 'fftw_f77' undefined reference to 'fftw_f77_create_plan' In function cft_2s undefined reference to 'fftw_f77' undefined reference to 'fftw_f77_create_plan' also In function dscal '__READFILE_NEW_in_RESTART_MODULE' > What version did you download ? if fftw 3.something - bad idea, use > FFTW 2.1.5. I have download the FFTW 2.1.5. version form www.fftw.org > Have you succesfully compiled the FFTW package on your machine? Yes, the compliation was succesfull > What is the result of " nm /usr/local/lib/libfftw.a | grep f77 "? The result is as follows: fftwf77.o: 000001d0 T fftw2d_f77_create_plan__ 00000210 T fftw3d_f77_create_plan__ 00000060 T fftw_f77__ 00000000 T fftw_f77_create_plan__ 00000040 T fftw_f77_destroy_plan__ 000000b0 T fftw_f77_one__ 00000280 T fftwnd_f77__ 00000120 T fftwnd_f77_create_plan__ 00000260 T fftwnd_f77_destroy_plan__ 000002d0 T fftwnd_f77_one__ > (replace /usr/local/lib/ with the name of the directory where you > installed the libfftw.a library) Why may I replace this?? best regards Bachir ----- Original Message ----- From: sprokuda at dnk.ru Date: Mon, 1 Sep 2003 10:20:49 +0400 To: Bachir Bouhafs Subject: Re:?[Pw?forum]?problem?with?compilation > Dear Bachir! > What version did you download ? if fftw 3.something - bad idea, use > FFTW 2.1.5. > I had just the same problem with a couple of weeks ago. > As for the rest - ( undercores and manner of compiler to call the > function) - see the manual. and a bit more ( the replies were given to > me on the same problem) : > 1) > FFTW is written in C but fortran wrappers are also included. > > Have you succesfully compiled the FFTW package on your machine? > What is the result of " nm /usr/local/lib/libfftw.a | grep f77 "? > (replace /usr/local/lib/ with the name of the directory where you > installed the libfftw.a library) > > Different f77 compilers use different name convenctions in order to > call routines (lowercase,uppercase, add_one_underscore, > add_two_underscores...), the "configure" step of fftw figures out what > is your f77 compiler > and use g77 if nothing else applies. > If you have a fortran compiler which is not in the list checked by > the configure script, after the configure step but before actually > compiling the library (with the make command), you may need to edit the > fftw/config.h file, look for the string > /*********************** fortran wrapper configuration > *********************/ > and set the values appropriated for you fortran compiler. > > In any case if you succesfully compiled the package, no matter which f77 > compiler the configure script found, > the "nm /usr/local/lib/libfftw.a | grep f77 " should find something. > Then you only need to discover how your fortran compiler likes to call > routines (for instance by "nm pwscf.o"), set the fftw/config.h > parameters accordingly and repeat the installation > > 2) > First you have to download FFTW-sources > (v.2.1.3) from www.fftw.org or www.netlib.org. Please > be careful do not download the 3.0 version. It is not > incompatible for the code. > > Then unpack it and go to the directory > /your_FFTW_dir/fftw and edit the config.h file. At the > end of the file please find lines which look the next > ones (as should be). A line should be commented out. > > /* #undef FFTW_FORTRANIZE_LOWERCASE */ > #define FFTW_FORTRANIZE_LOWERCASE_UNDERSCORE 1 > /* #undef FFTW_FORTRANIZE_UPPERCASE */ > /* #undef FFTW_FORTRANIZE_UPPERCASE_UNDERSCORE */ > > Then just press "configure", "make", "make install". > The last, you should link the fftw-bibrary you are > being installed before the MKL one, i.e. -lfftw > -lMKL.. > > Probably the above mentioned procedure is not required > for the v.1.2 (PWSCF), but you have to check it. > > Exactly you do not need the FFTW.3.0 version and you > can remove it. > > ALT Linux is RedHat based one, so it should work good. > Your troubles go to FFTW.3.0. > > Any question, please let me know. > > > regards, Sergey. > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum Dr. Bachir Bouhafs Physics Department University of Sidi Bel Abbes 22000 Sidi Bel Abbes, Algeria Mobil:+213 (0)71 18 21 47 Fax:+213 48 54 11 52 E-mail: bbouhafs at mail.com -- __________________________________________________________ Sign-up for your own personalized E-mail at Mail.com http://www.mail.com/?sr=signup CareerBuilder.com has over 400,000 jobs. Be smarter about your job search http://corp.mail.com/careers From sprokuda at dnk.ru Tue Sep 2 13:25:56 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Tue, 2 Sep 2003 15:25:56 +0400 Subject: [Pw_forum] =?ISO-8859-1?B?UmU6IFtQd19mb3J1bV0gUmU6oFtQd6Bmb3J1bV2gcHJvYmxlbaB3aXRo?= =?ISO-8859-1?B?oGNvbXBpbGF0aW9u?= In-Reply-To: <20030902110118.55879.qmail@mail.com> References: <20030902110118.55879.qmail@mail.com> Message-ID: <622303547.20030902152556@dnk.ru> Dear Bachir ! Now it is smiple (at least I hope so) I guess you have in you make.sys file : -D"FFTWND_F77_ONE=fftwnd_f77_one" \ -D"FFTW3D_F77_CREATE_PLAN=fftw3d_f77_create_plan" \ -D"FFTW_F77=fftw_f77" \ -D"FFTW_F77_CREATE_PLAN=fftw_f77_create_plan" BUT you should have -D"FFTWND_F77_ONE=fftwnd_f77_one_" \ -D"FFTW3D_F77_CREATE_PLAN=fftw3d_f77_create_plan_" \ -D"FFTW_F77=fftw_f77_" \ -D"FFTW_F77_CREATE_PLAN=fftw_f77_create_plan_" what means you have to add one undercore in the end of deinition. Then: > (replace /usr/local/lib/ with the name of the directory where you > installed the libfftw.a library) Why may I replace this?? if you have there libfftw.a, probably you don't need new fftw. Try to replace it (saved it in some other directory beforehand for backup). One remark: also you compliler might not see this lib, this way you have to say compiler where to look for it by adding in makefiles in directories to PW, PH , PP ...so on in manner like ( for PW ) it was: MODULES = ../Modules/*.o it should be: MODULES = ../Modules/*.o \ /usr/local/lib/*.a Best regards, Sergey. best regards Bachir From giannozz at nest.sns.it Tue Sep 2 15:46:05 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 2 Sep 2003 15:46:05 +0200 Subject: [Pw_forum] =?iso-8859-1?q?problem=A0with=A0compilation?= In-Reply-To: <622303547.20030902152556@dnk.ru> References: <20030902110118.55879.qmail@mail.com> <622303547.20030902152556@dnk.ru> Message-ID: <200309021546.05433.giannozz@nest.sns.it> On Tuesday 02 September 2003 13:25, sprokuda at dnk.ru wrote: > Dear Bachir ! > [...] you have to add one underscore in the end of definition. two underscores: the absoft compiler, unlike many other compilers, does not add an underscore to symbol names Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From sprokuda at dnk.ru Tue Sep 2 15:58:42 2003 From: sprokuda at dnk.ru (sprokuda at dnk.ru) Date: Tue, 2 Sep 2003 17:58:42 +0400 Subject: [Pw_forum] ATLAS question In-Reply-To: <200309021546.05433.giannozz@nest.sns.it> References: <20030902110118.55879.qmail@mail.com> <622303547.20030902152556@dnk.ru> <200309021546.05433.giannozz@nest.sns.it> Message-ID: <50417206.20030902175842@dnk.ru> Dear All, does naybody use ATALA librarty version 3.5.6 ? I'm asking because theres is no dlamcn.f file therein ( which has to be replaced , as said in manual ). Thank you for any comment or if somebody has an older version, with which PWscf was already successfully compiled ? Sergey. From giannozz at nest.sns.it Tue Sep 2 16:09:04 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 2 Sep 2003 16:09:04 +0200 Subject: [Pw_forum] ATLAS question In-Reply-To: <50417206.20030902175842@dnk.ru> References: <20030902110118.55879.qmail@mail.com> <200309021546.05433.giannozz@nest.sns.it> <50417206.20030902175842@dnk.ru> Message-ID: <200309021609.04519.giannozz@nest.sns.it> On Tuesday 02 September 2003 15:58, sprokuda at dnk.ru wrote: > Dear All, > does naybody use ATALA librarty version 3.5.6 ? > I'm asking because theres is no dlamcn.f file therein you mean "dlamch.f" ? it is part of lapack, not of atlas. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From eyvaz_isaev at yahoo.com Tue Sep 2 17:35:02 2003 From: eyvaz_isaev at yahoo.com (Eyvaz Isaev) Date: Tue, 2 Sep 2003 08:35:02 -0700 (PDT) Subject: [Pw_forum] ATLAS question In-Reply-To: <50417206.20030902175842@dnk.ru> Message-ID: <20030902153502.35853.qmail@web80511.mail.yahoo.com> Dear Sergei, Precompied ATLAS is useless for you because during compilation it tests your computer and tries to choose best parameters. You can compile it for yourself. It is very easy to do it. Eyavz. P.S. Thank you distributing my hints. --- sprokuda at dnk.ru wrote: > Dear All, > does naybody use ATALA librarty version 3.5.6 ? > I'm asking because theres is no dlamcn.f file > therein ( which has to > be replaced , as said in manual ). Thank you for any > comment or if > somebody has an older version, with which PWscf was > already > successfully compiled ? > Sergey. > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From bbouhafs at mail.com Tue Sep 2 20:07:05 2003 From: bbouhafs at mail.com (Bachir Bouhafs) Date: Tue, 02 Sep 2003 19:07:05 +0100 Subject: [Pw_forum] =?iso-8859-1?Q?Re:=A0[Pw=A0forum]=A0problem=A0with=A0compilation?= Message-ID: <20030902180706.47462.qmail@mail.com> Dear PWscf users, Thank you for your prompt replies and your help. By adding two underscores in the end of definition in make.sys file, I arrived to eliminate many compilation errors related with fft library, but now I have generated other problems which are mentioned below, just after making make pw. so, what may I change now? another help is welcome Best regards Bachir ------------------------------------------------ /usr/absoft/lib/libfio.a(lowlevel.o): In function `_temp_file': lowlevel.o(.text+0x10): the use of `tempnam' is dangerous, better use `mkstemp' restart.o: In function `__READFILE_NEW_in_RESTART_MODULE': restart.o(.text+0xcda7): undefined reference to `dscal' /usr/local/lib/libfftw.a(rader.o): In function `create_rader_aux': rader.o(.text+0x1ba): undefined reference to `cos' rader.o(.text+0x1e8): undefined reference to `sin' /usr/local/lib/libfftw.a(twiddle.o): In function `fftw_compute_twiddle': twiddle.o(.text+0x7a): undefined reference to `cos' twiddle.o(.text+0x8f): undefined reference to `sin' twiddle.o(.text+0x170): undefined reference to `cos' twiddle.o(.text+0x188): undefined reference to `sin' twiddle.o(.text+0x2c9): undefined reference to `cos' twiddle.o(.text+0x2dc): undefined reference to `sin' collect2: ld returned 1 exit status make[1]: *** [all] Erreur 1 make[1]: Quitte le r?pertoire `/root/pw90/PW' make: *** [pw] Erreur 2 ------------------------------------------------ ----- Original Message ----- From: Paolo Giannozzi Date: Tue, 2 Sep 2003 15:46:05 +0200 To: pw forum at pwscf.org Subject: Re:?[Pw?forum]?problem?with?compilation > On Tuesday 02 September 2003 13:25, sprokuda at dnk.ru wrote: > > Dear Bachir ! > > [...] you have to add one underscore in the end of definition. > > two underscores: the absoft compiler, unlike many other compilers, > does not add an underscore to symbol names > > Paolo > > -- > Paolo Giannozzi e-mail: giannozz at nest.sns.it > Scuola Normale Superiore Phone: +39/050509412 > Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 > I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum Dr. Bachir Bouhafs Physics Department University of Sidi Bel Abbes 22000 Sidi Bel Abbes, Algeria Mobil:+213 (0)71 18 21 47 Fax:+213 48 54 11 52 E-mail: bbouhafs at mail.com -- __________________________________________________________ Sign-up for your own personalized E-mail at Mail.com http://www.mail.com/?sr=signup CareerBuilder.com has over 400,000 jobs. Be smarter about your job search http://corp.mail.com/careers From Steven.Kirk at htu.se Tue Sep 2 20:49:42 2003 From: Steven.Kirk at htu.se (Steven Kirk) Date: Tue, 02 Sep 2003 20:49:42 +0200 Subject: [Pw_forum] AIM routines for PWSCF ? Message-ID: Hello, Does anyone on this list have any AIM (Atoms in Molecules) analysis routines (or even full programs) for PWSCF ? Many thanks in advance for any help ! Regards, Steve Kirk From giannozz at nest.sns.it Tue Sep 2 22:47:47 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 2 Sep 2003 22:47:47 +0200 Subject: [Pw_forum] =?iso-8859-1?q?problem=A0with=A0compilation?= In-Reply-To: <20030902180706.47462.qmail@mail.com> References: <20030902180706.47462.qmail@mail.com> Message-ID: <200309022247.47315.giannozz@nest.sns.it> On Tuesday 02 September 2003 20:07, Bachir Bouhafs wrote: > [...] > /usr/absoft/lib/libfio.a(lowlevel.o): In function `_temp_file': > lowlevel.o(.text+0x10): the use of `tempnam' is dangerous this is irrelevant > [...] restart.o(.text+0xcda7): undefined reference to `dscal' in restart.f90 change "dscal" into "DSCAL". If it doesn't work, add line #include "machine.h" or change to "dscal" to "dscal_" or "dscal__" > [...] rader.o(.text+0x1ba): undefined reference to `cos' > rader.o(.text+0x1e8): undefined reference to `sin' [...] try to add "-lm" to libraries Paolo From degironc at sissa.it Wed Sep 3 10:28:21 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Wed, 03 Sep 2003 10:28:21 +0200 Subject: [Pw_forum] ATLAS question References: <20030902110118.55879.qmail@mail.com> <622303547.20030902152556@dnk.ru> <200309021546.05433.giannozz@nest.sns.it> <50417206.20030902175842@dnk.ru> Message-ID: <3F55A625.5070007@sissa.it> > > >Dear All, >does naybody use ATALA librarty version 3.5.6 ? >I'm asking because theres is no dlamcn.f file therein ( which has to >be replaced , as said in manual ). Thank you for any comment or if >somebody has an older version, with which PWscf was already >successfully compiled ? >Sergey. > > ATLAS contains many but not all routine used by LAPACK. In the ATLAS/doc/LibReadme.txt you can read the following instruction that can be usefull in your case: ************************** GETTING A FULL LAPACK LIB ************************** ATLAS does not provide a full lapack library. However, there is a simple way to get ATLAS to provide its faster LAPACK routines to a full LAPACK library. ATLAS's internal routines are distinct from LAPACK's, so it is safe to compile ATLAS's LAPACK routines directly into a netlib-style LAPACK library. First, obtain the LAPACK src from netlib and build the LAPACK library as normal. Then, in this directory (where you should have a liblapack.a), issue the following commands: mkdir tmp cd tmp ar x ../liblapack.a cp ../liblapack.a ar r ../liblapack.a *.o cd .. rm -rf tmp Just linking in ATLAS's liblapack.a first will not get you the best LAPACK performance, mainly because LAPACK's untuned ILAENV will be used instead of ATLAS's tuned one. EOF Stefano de Gironcoli From breezejd at lsbu.ac.uk Wed Sep 3 13:12:42 2003 From: breezejd at lsbu.ac.uk (Jonathan Breeze) Date: Wed, 03 Sep 2003 12:12:42 +0100 Subject: [Pw_forum] Third order anharmonic dynamical matrix Message-ID: <3F55CCAA.4020902@sbu.ac.uk> Dear PWscf users, Please excuse my earlier post on the above topic. I had been very silly, and not realised that the D3 code only calculated the anharmonic coupling coefficients for zone-centre phonons, $D^3(0,q,-q) Silly me indeed! This leads me onto my next question.... Using the PWSCF D3 code, should it be possible to calculate the coefficients for phonons at arbitraty wave-vector? That is, to calculate D^3(-q1-q2,q1,q2). Looking at the 2n+1 theory, it should be possible, but also looking at the code, it seems formidable. Any ideas on how to the modify the code (if at all required) to be able to do this? Kind regards, Jonathan Breeze From giannozz at nest.sns.it Wed Sep 3 15:02:24 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 3 Sep 2003 15:02:24 +0200 Subject: [Pw_forum] AIM routines for PWSCF ? In-Reply-To: References: Message-ID: <200309031502.24699.giannozz@nest.sns.it> On Tuesday 02 September 2003 20:49, Steven Kirk wrote: > Does anyone on this list have any AIM (Atoms in Molecules) > analysis routines (or even full programs) for PWSCF ? there is a code called "voronoy.x" which divides the space into Voronoy polyedra centered around each atom (that is: it assigns each point in the space to the closest atom) and calculates the charge contained in them. I used this approach for a rough estimate of the charge on each C atom in fullerene systems. Bader's topological analysis of the charge is not presently available. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From goranka.bilalbegovic at zg.hinet.hr Wed Sep 3 18:20:52 2003 From: goranka.bilalbegovic at zg.hinet.hr (Goranka Bilalbegovic) Date: Wed, 3 Sep 2003 18:20:52 +0200 Subject: [Pw_forum] AIM routines for PWSCF ? References: <200309031502.24699.giannozz@nest.sns.it> Message-ID: <000801c37237$bdc248a0$ab3e1dc3@gost> Hello. > > Does anyone on this list have any AIM (Atoms in Molecules) > > analysis routines (or even full programs) for PWSCF ? > > there is a code called "voronoy.x" which divides the space into > Voronoy polyedra centered around each atom (that is: it assigns > each point in the space to the closest atom) and calculates the > charge contained in them. I used this approach for a rough > estimate of the charge on each C atom in fullerene systems. > Bader's topological analysis of the charge is not presently > available. How important is the sentence in the header of voronoy.f90: "Note that this is a very rough way to associate charges to atoms and that it is well defined only if all atoms are of the same type" ? In addition, projwave calculates the Lowdin charges. Perhaps, it is also useful ? Best regards, Goranka From giannozz at nest.sns.it Wed Sep 3 19:23:26 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 3 Sep 2003 19:23:26 +0200 Subject: [Pw_forum] AIM routines for PWSCF ? In-Reply-To: <000801c37237$bdc248a0$ab3e1dc3@gost> References: <200309031502.24699.giannozz@nest.sns.it> <000801c37237$bdc248a0$ab3e1dc3@gost> Message-ID: <200309031923.26079.giannozz@nest.sns.it> On Wednesday 03 September 2003 18:20, Goranka Bilalbegovic wrote: > [...] > How important is the sentence in the header of voronoy.f90: > "Note that this is a very rough way to associate charges to atoms > and that it is well defined only if all atoms are of the same type" ? how important I don't know, but I subscribe to that statement (I wrote it years ago ...) > In addition, projwave calculates the Lowdin charges. > Perhaps, it is also useful ? if one wants just atomic charges, I guess that Lowdin charges (or Mulliken, or whatever recipe) are just fine. There is more than this in AIM methods, though. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From bbouhafs at mail.com Wed Sep 3 20:10:02 2003 From: bbouhafs at mail.com (Bachir Bouhafs) Date: Wed, 03 Sep 2003 19:10:02 +0100 Subject: =?iso-8859-1?Q?Re: [Pw_forum] problem=A0with=A0compilation?= Message-ID: <20030903181002.75902.qmail@mail.com> Dear Paolo Thanks for all, the compilation work well now for all PW, PH, tools, etc... ONLY FOR D3 routines, probelms with call to MPI ROUTINES. best wishes Bachir ----- Original Message ----- From: Paolo Giannozzi Date: Tue, 2 Sep 2003 22:47:47 +0200 To: pw_forum at pwscf.org Subject: Re: [Pw_forum] problem?with?compilation > On Tuesday 02 September 2003 20:07, Bachir Bouhafs wrote: > > [...] > > /usr/absoft/lib/libfio.a(lowlevel.o): In function `_temp_file': > > lowlevel.o(.text+0x10): the use of `tempnam' is dangerous > > this is irrelevant > > > [...] restart.o(.text+0xcda7): undefined reference to `dscal' > > in restart.f90 change "dscal" into "DSCAL". If it doesn't > work, add line > #include "machine.h" > or change to "dscal" to "dscal_" or "dscal__" > > > [...] rader.o(.text+0x1ba): undefined reference to `cos' > > rader.o(.text+0x1e8): undefined reference to `sin' [...] > > try to add "-lm" to libraries > > Paolo > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum Dr. Bachir Bouhafs Physics Department University of Sidi Bel Abbes 22000 Sidi Bel Abbes, Algeria Mobil:+213 (0)71 18 21 47 Fax:+213 48 54 11 52 E-mail: bbouhafs at mail.com -- __________________________________________________________ Sign-up for your own personalized E-mail at Mail.com http://www.mail.com/?sr=signup CareerBuilder.com has over 400,000 jobs. Be smarter about your job search http://corp.mail.com/careers From degironc at sissa.it Thu Sep 4 10:19:47 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Thu, 04 Sep 2003 10:19:47 +0200 Subject: [Pw_forum] AIM routines for PWSCF ? References: <200309031502.24699.giannozz@nest.sns.it> <000801c37237$bdc248a0$ab3e1dc3@gost> Message-ID: <3F56F5A3.6040405@sissa.it> Goranka Bilalbegovic wrote: >How important is the sentence in the header of voronoy.f90: >"Note that this is a very rough way to associate charges to atoms > and that it is well defined only if all atoms are of the same type" ? > > I think it may be very important. If I remember corectly if you calculate the amount of charge contained in the Voronoy polyhedra in an polar semiconductor in the zincblend structure the result is not what you expect: the amount of charge around the cation is LARGER than the one around the anion, basically because anion charge is more extended and spills out of the anion polyhedron into the cation one. The physically sensible result is obtained by the Bader construction, that is however far more complicated and is not presently implemented in PWscf. I think Mulliken or Lowdin charge analysis gives reasonable results. Stefano de Gironcoli >In addition, projwave calculates the Lowdin charges. >Perhaps, it is also useful ? > > > From zchelahi at yahoo.fr Sat Sep 6 16:47:23 2003 From: zchelahi at yahoo.fr (=?iso-8859-1?q?zouaoui=20Chelahi=20Chikr?=) Date: Sat, 6 Sep 2003 16:47:23 +0200 (CEST) Subject: [Pw_forum] email Message-ID: <20030906144723.93296.qmail@web40507.mail.yahoo.com> my email is zchelahi at yahoo.fr --------------------------------- Do You Yahoo!? -- Une adresse @yahoo.fr gratuite et en fran?ais ! Testez le nouveau Yahoo! Mail -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030906/d1017911/attachment.htm From proffess at yandex.ru Sun Sep 14 10:42:27 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Sun, 14 Sep 2003 12:42:27 +0400 (MSD) Subject: [Pw_forum] Makefile for HP-PA RISC machines In-Reply-To: <3F37463E.7070700@sissa.it> References: <3EE86CDE.000001.22552@pantene.yandex.ru> <200306121547.02106.giannozz@nest.sns.it> <3F372EC0.000004.05643@camay.yandex.ru> <3F37463E.7070700@sissa.it> Message-ID: <3F6429F3.000002.10863@tide.yandex.ru> Dear PWscf users, I would like to anounce the make.sys files for HP-PA RISC machines under OS=HP-UX (>=11.0) in serial and parallel versions. I suggest that files will work for HP machines on Itanium platform under OS=HP-UX or True64 Unix. I suggest you have a HP f90 compiler (version >=2.4) and HP C compiler (any version), LAPACK and BLAS. I use the HP vendor math library MLIB that contain the LAPACK, BLAS and FFTW, but I don't know how to use FFTW from MLIB such CXML. I use the sorce fftw library. The main part of these files have written by Dr. Eyvaz Isaev (Moscow, Russia), the 64-bit version of these files have written by DIECKHANS,JOEY (HP-Cupertino,ex1)from HP corporation. ================================================================================== #Serial version CPPFLAGS = -I$(OSHOME)/include -DORIGIN -DFFTW #Parallel version #CPPFLAGS = -DPARA -D__MPI -I$(OSHOME)/include -DORIGIN -DFFTW # Fortran compiler: # F90 = f90 #Serial Version FFLAGS = +O2 -w +cpp=yes +Odataprefetch +Onolimit +r8 +U77 +DA2.0 +DS2.0 F90FLAGS = $(FFLAGS) $(CPPFLAGS) # This is needed to tell the compiler where modules are #Serial version #MODULEFLAG= -I$(OSHOME)/Modules -I$(OSHOME)/PW -I$(OSHOME)/PH # Parallel version #MODULEFLAG= -I$(OSHOME)/Modules -I$(OSHOME)/PW -I$(OSHOME)/PH -I/opt/mpi/include # Libraries: # Serial Version LIBS = L/opt/mlib/lib/pa2.0 -lveclib -llapack -L/usr/local/lib/fftw-2.1.3/lib -lfftw -lm #Parallel version # Libraries: # LIBS = -L/opt/mpi/lib/pa2.0 -lmpi -ldmpi -lmpio -lmtmpi -L/opt/mlib/lib/pa2.0 -lveclib -llapack -L/usr/local/lib/fftw-2.1.3/lib -lfftw -lm # Loader: LD=$(F90) LFLAGS = +U77 +DA2.0 +DS2.0 +O2 -lpthread $(LIBS) # # ar: # AR = ar ARFLAGS = ruv ================================================================================== Please, note if you want to compile PWscf in 64-bit mode, you must add in FFLAGS "+DA2.0W +noppu" and use the 64-bit version of MLIB. You must recompile fftw libvrary in 64-bit mode. Firstly, you must change the configure file in fftw. You should add: ac_f77_mangle_type=lowercase before the line echo "$ac_t""$ac_f77_mangle_type" 1>&6 and compile the fftw as usually. Please, note, there is one version of MPI that requires the argument "-stdio=i" in comman line for reading from standart output (thanks HP MPI team). The question for Authors of PWscf. The cite from HP Fortran manual:"If your program does not need to run in 64-bit mode, there is no benefit to compiling it in 64-bit mode. In fact, the executable program may run slower than if compiled in 32-bit mode." Is it neccesary to compile PW in 64-bit mode? I know it gives the all memory that computer has. Hope it helps Best wishes, Sergey Lisenkov From likedew at phys.ksu.edu Sun Sep 14 21:42:03 2003 From: likedew at phys.ksu.edu (Hong, SamPyo) Date: Sun, 14 Sep 2003 14:42:03 -0500 Subject: [Pw_forum] Error during phonon calculation Message-ID: Hi, During phonon calculation, I got this error. I don't understand what these errors are about. I appreicate any of your adivces ! Sampyo ------------------------------------------------------------------------ ------------------ iter # 12 total cpu time : 41698.2 secs av.it.: 10.9 thresh= .402E-05 alpha_mix = .100 |ddv_scf|^2 = NaNQ kpoint 1 ibnd 82 linter: root not converged -NaNQ kpoint 2 ibnd 82 linter: root not converged -NaNQ kpoint 3 ibnd 82 linter: root not converged -NaNQ kpoint 4 ibnd 81 linter: root not converged -NaNQ kpoint 5 ibnd 80 linter: root not converged -NaNQ kpoint 6 ibnd 80 linter: root not converged -NaNQ kpoint 7 ibnd 80 linter: root not converged -NaNQ Pert. # 1: Fermi energy shift (Ryd) = -NaNQ -NaNQ DSWAP : 2538-2001 The number of elements (ARG NO. 1) in a vector must be greater than or equal to zero. DSWAP : 2538-2099 End of input argument error reporting. For more information, refer to Engineering and Scientific Subroutine Library Guide and Reference (SA22-7272). DSWAP : 2538-2604 Execution terminating due to error count for error number 2099. DSWAP : 2538-2605 Message summary: 2001 - 1 DSWAP : 2538-2605 ---------------------------------------------------------------------- -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030914/1ef4871e/attachment.htm From giannozz at nest.sns.it Sun Sep 14 21:56:02 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 14 Sep 2003 21:56:02 +0200 (CEST) Subject: [Pw_forum] Error during phonon calculation In-Reply-To: Message-ID: On Sun, 14 Sep 2003, Hong, SamPyo wrote: > During phonon calculation, I got this error. I don't understand what > these errors are about. [...] > DSWAP : 2538-2001 > The number of elements (ARG NO. 1) in a vector must be greater than > or equal to zero. Very strange. Does it happen for a specific case ? Is the same executable working for other cases ? Does it happen in serial execution, in parallel execution, on just one machine, on all machines you tried, ... ? Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From giannozz at nest.sns.it Sun Sep 14 22:00:26 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sun, 14 Sep 2003 22:00:26 +0200 (CEST) Subject: [Pw_forum] Makefile for HP-PA RISC machines In-Reply-To: <3F6429F3.000002.10863@tide.yandex.ru> Message-ID: On Sun, 14 Sep 2003, Sergei Lisenkov wrote: > The question for Authors of PWscf. The cite from HP Fortran manual:"If > your program does not need to run in 64-bit mode, there is no benefit to > compiling it in 64-bit mode. In fact, the executable program may run > slower than if compiled in 32-bit mode." Is it neccesary to compile PW > in 64-bit mode? I know it gives the all memory that computer has. if you need more than 2Gb RAM (per process, I think, on parallel machines), you need to compile in 64-bit mode. Otherwise, you don't. I think you seldom need it. Thank you for contributing the Makefile Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From likedew at phys.ksu.edu Mon Sep 15 22:09:05 2003 From: likedew at phys.ksu.edu (Hong, SamPyo) Date: Mon, 15 Sep 2003 15:09:05 -0500 Subject: [Pw_forum] Error during phonon calculation Message-ID: Thank you for your questions ! The executable ph.x used has been used successfully for other calculations. Theses error occurred in parallel execution but I don't know if they are related to just one node or all machines. When I repeat the problematic calculation under exact same condition, the error above happens in different stages of the calculation compared to the previous one, for examples, different representation, different mode, and different iteration number. Sampyo -----Original Message----- From: Paolo Giannozzi [mailto:giannozz at nest.sns.it] Sent: Sunday, September 14, 2003 2:56 PM To: pw_forum at pwscf.org Subject: Re: [Pw_forum] Error during phonon calculation On Sun, 14 Sep 2003, Hong, SamPyo wrote: > During phonon calculation, I got this error. I don't understand what > these errors are about. [...] > DSWAP : 2538-2001 > The number of elements (ARG NO. 1) in a vector must be greater than or > equal to zero. Very strange. Does it happen for a specific case ? Is the same executable working for other cases ? Does it happen in serial execution, in parallel execution, on just one machine, on all machines you tried, ... ? Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 _______________________________________________ Pw_forum mailing list Pw_forum at pwscf.org http://www.democritos.it/mailman/listinfo/pw_forum From giannozz at nest.sns.it Mon Sep 15 22:40:25 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Mon, 15 Sep 2003 22:40:25 +0200 Subject: [Pw_forum] Error during phonon calculation In-Reply-To: References: Message-ID: <200309152240.25337.giannozz@nest.sns.it> On Monday 15 September 2003 22:09, Hong, SamPyo wrote: > The executable ph.x used has been used successfully for other > calculations. > These errors occurred in parallel execution but I don't know if they are > related to just one node or all machines. When I repeat the problematic > calculation under exact same condition, the error above happens in > different stages of the calculation compared to the previous one, for > examples, different representation, different mode, and different > iteration number. it looks like some array index going out of bounds. You might try to recompile everything with debugging and boundary check (-g -C options for IBM). Meanwhile: could you please post your input data (and output files) ? Paolo From proffess at yandex.ru Tue Sep 16 07:53:53 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Tue, 16 Sep 2003 09:53:53 +0400 (MSD) Subject: [Pw_forum] Fw: Compiling error of Pw-1.3 Message-ID: <3F66A571.000008.14217@tide.yandex.ru> Dear PW users, I have error with compiling the new version of PWscf-1.3 on Compaq Alpha clusters using Compaq Compiler: ..... mpif90 -O -fpe0 -real_size 64 -align dcommons -align records -free -cpp -D__ALPHA -D__PARA -D__MPI -DDXML -I/home/proffess/PWscf-1.3/include -I/home/proffess/PWscf-1.3/Modules -I/home/proffess/PWscf-1.3/PW -I/home/proffess/PWscf-1.3/PH -c cft3.f90 f90: Error: cft3.f90, line 79: This name does not have a type, and must have an explicit type. [DFFTP] mc = dfftp%ismap (i) -------------^ make[1]: *** [cft3.o] Error 1 make[1]: Leaving directory `/home/proffess/PWscf-1.3/PW' Is anybody know how to solve it? Thanks, Sergey Lisenkov --- End of forwarded message -- -- ??? ????????! ????? ??????????? ???????? ?????? - ?????? "12 ???????"! ???????! www.12stulyev.ru From roma at cea.fr Tue Sep 16 08:54:52 2003 From: roma at cea.fr (Guido Roma) Date: Tue, 16 Sep 2003 08:54:52 +0200 Subject: [Pw_forum] Fw: Compiling error of Pw-1.3 References: <3F66A571.000008.14217@tide.yandex.ru> Message-ID: <3F66B3BC.939B8E3D@cea.fr> Sergei Lisenkov wrote: > > Dear PW users, > > I have error with compiling the new version of PWscf-1.3 on Compaq Alpha clusters using Compaq Compiler: > > ...... > mpif90 -O -fpe0 -real_size 64 -align dcommons -align records -free -cpp -D__ALPHA -D__PARA -D__MPI -DDXML -I/home/proffess/PWscf-1.3/include -I/home/proffess/PWscf-1.3/Modules -I/home/proffess/PWscf-1.3/PW -I/home/proffess/PWscf-1.3/PH -c cft3.f90 > f90: Error: cft3.f90, line 79: This name does not have a type, and must have an explicit type. [DFFTP] > mc = dfftp%ismap (i) Hello, I think that the problems comes from the following lines: #if defined __FFT_MODULE_DRV use fft_scalar, only : cft_1z, cft_2xy use sticks, only: dfftp #endif which should better look like this: #if defined __FFT_MODULE_DRV use fft_scalar, only : cft_1z, cft_2xy #endif use sticks, only: dfftp can you confirm, Paolo? I compiled it with no problems on HP/Compaq (former DEC/alpha) cluster, but using internal FFTW: CPPFLAGS = -D__ALPHA -D__PARA -D__MPI -D__FFTW -D__USE_INTERNAL_FFTW \ -I$(OSHOME)/include -I./ I don't know if it works in parallel with DXML (I don't see any cft_dxml.f90 or similar ...). Guido -- Guido Roma -- CEA-Saclay - DEN/DMN/SRMP Bat.520/130 Phone: [+33]-1-69085738 -- Fax: ...6867 -- Mobile: [+33]-6-20069085 From likedew at phys.ksu.edu Mon Sep 15 22:51:04 2003 From: likedew at phys.ksu.edu (Hong, SamPyo) Date: Mon, 15 Sep 2003 15:51:04 -0500 Subject: [Pw_forum] Error during phonon calculation Message-ID: Hi, This is the input file: &inputph tr2_ph=1.0d-14, alpha_mix(1)=0.1, prefix='cni', amass(1)=12.0107,amass(2)=58.6934, tmp_dir='./tmp/', fildyn='cni.dynG', / 0.0 0.0 0.0 Thanks. Sampyo -------------- next part -------------- A non-text attachment was scrubbed... Name: cniG.87002.out Type: application/octet-stream Size: 56931 bytes Desc: cniG.87002.out Url : /pipermail/attachments/20030915/adf7a494/attachment.obj From degironc at sissa.it Tue Sep 16 09:40:07 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Tue, 16 Sep 2003 09:40:07 +0200 Subject: [Pw_forum] Error during phonon calculation References: Message-ID: <3F66BE57.7020200@sissa.it> Strange errors, maybe also the one you encountered, may appear in parallel calculations if some garbage is left in the outdir directory and is mis-interpreted by the phonon code (I dont know why but will try to find out). Make sure you run the pw+ph codes using a clean directory as your outdir. Stefano de Gironcoli Hong, SamPyo wrote: >Thank you for your questions ! > >The executable ph.x used has been used successfully for other >calculations. >Theses error occurred in parallel execution but I don't know if they are >related to just one node or all machines. When I repeat the problematic >calculation under exact same condition, the error above happens in >different stages of the calculation compared to the previous one, for >examples, different representation, different mode, and different >iteration number. > >Sampyo > > From proffess at yandex.ru Tue Sep 16 10:26:21 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Tue, 16 Sep 2003 12:26:21 +0400 (MSD) Subject: [Pw_forum] Fw: Compiling error of Pw-1.3 In-Reply-To: <3F66B3BC.939B8E3D@cea.fr> References: <3F66A571.000008.14217@tide.yandex.ru> <3F66B3BC.939B8E3D@cea.fr> Message-ID: <3F66C92D.00000C.26967@colgate.yandex.ru> Dear Guido, Thanks for your help. But I found problems on linking phase: ..... ../Modules/fft_scalar.o: In function `$fft_scalar$cft_1z': :179: undefined reference to `create_plan_1d__' :179: undefined reference to `create_plan_1d__' :179: undefined reference to `create_plan_1d__' :180: undefined reference to `create_plan_1d__' :180: undefined reference to `create_plan_1d__' ../Modules/fft_scalar.o::180: more undefined references to `create_plan_1d__' follow ../Modules/fft_scalar.o: In function `$fft_scalar$cfft3d': :563: undefined reference to `destroy_plan_3d__' :563: undefined reference to `destroy_plan_3d__' :563: undefined reference to `destroy_plan_3d__' :564: undefined reference to `destroy_plan_3d__' :564: undefined reference to `destroy_plan_3d__' ../Modules/fft_scalar.o::564: more undefined references to `destroy_plan_3d__' follow ../Modules/fft_scalar.o: In function `$fft_scalar$cfft3d': :565: undefined reference to `create_plan_3d__' :565: undefined reference to `create_plan_3d__' :565: undefined reference to `create_plan_3d__' :566: undefined reference to `create_plan_3d__' :566: undefined reference to `create_plan_3d__' ../Modules/fft_scalar.o::566: more undefined references to `create_plan_3d__' follow ../Modules/fft_scalar.o: In function `$fft_scalar$cfft3d': :594: undefined reference to `fftw_inplace_drv_3d__' :594: undefined reference to `fftw_inplace_drv_3d__' :594: undefined reference to `fftw_inplace_drv_3d__' :601: undefined reference to `fftw_inplace_drv_3d__' :601: undefined reference to `fftw_inplace_drv_3d__' ../Modules/fft_scalar.o::601: more undefined references to `fftw_inplace_drv_3d__' follow ../Modules/fft_scalar.o: In function `$fft_scalar$cfft3ds': :738: undefined reference to `create_plan_1d__' :738: undefined reference to `create_plan_1d__' :738: undefined reference to `create_plan_1d__' :739: undefined reference to `create_plan_1d__' :739: undefined reference to `create_plan_1d__' ../Modules/fft_scalar.o::739: more undefined references to `create_plan_1d__' follow ../Modules/fft_scalar.o: In function `$fft_scalar$cfft3ds': :794: undefined reference to `fftw_inplace_drv_1d__' :794: undefined reference to `fftw_inplace_drv_1d__' :794: undefined reference to `fftw_inplace_drv_1d__' :815: undefined reference to `fftw_inplace_drv_1d__' :815: undefined reference to `fftw_inplace_drv_1d__' ../Modules/fft_scalar.o::815: more undefined references to `fftw_inplace_drv_1d__' follow ../Modules/fft_scalar.o: In function `$fft_scalar$cft_b': :993: undefined reference to `create_plan_1d__' :993: undefined reference to `create_plan_1d__' :993: undefined reference to `create_plan_1d__' :995: undefined reference to `destroy_plan_2d__' :995: undefined reference to `destroy_plan_2d__' :995: undefined reference to `destroy_plan_2d__' :996: undefined reference to `create_plan_2d__' :996: undefined reference to `create_plan_2d__' :996: undefined reference to `create_plan_2d__' :1034: undefined reference to `fftw_inplace_drv_1d__' :1034: undefined reference to `fftw_inplace_drv_1d__' :1034: undefined reference to `fftw_inplace_drv_1d__' :1035: undefined reference to `fftw_inplace_drv_2d__' :1035: undefined reference to `fftw_inplace_drv_2d__' :1035: undefined reference to `fftw_inplace_drv_2d__' collect2: ld returned 1 exit status fort: Severe: Failed while trying to link. make[1]: *** [memory] Error 1 make[1]: Leaving directory `/home/proffess/PWscf-1.3/PW' make: *** [pw] Error 2 What is problem? Sergey From roma at cea.fr Tue Sep 16 11:21:22 2003 From: roma at cea.fr (Guido Roma) Date: Tue, 16 Sep 2003 11:21:22 +0200 Subject: [Pw_forum] Fw: Compiling error of Pw-1.3 References: <3F66A571.000008.14217@tide.yandex.ru> <3F66B3BC.939B8E3D@cea.fr> <3F66C92D.00000C.26967@colgate.yandex.ru> Message-ID: <3F66D612.7D9BFA78@cea.fr> Sergei Lisenkov wrote: > > Dear Guido, > > Thanks for your help. But I found problems on linking phase: Hmmm, it looks like the dependencies are not correct ... I don't know, did you, by chance, modify the make.sys without rerunning configure (i.e. moduledep.x)? Guido > > ...... > .../Modules/fft_scalar.o: In function `$fft_scalar$cft_1z': > :179: undefined reference to `create_plan_1d__' > :179: undefined reference to `create_plan_1d__' > :179: undefined reference to `create_plan_1d__' > :180: undefined reference to `create_plan_1d__' > :180: undefined reference to `create_plan_1d__' > .../Modules/fft_scalar.o::180: more undefined references to `create_plan_1d__' follow > .../Modules/fft_scalar.o: In function `$fft_scalar$cfft3d': > :563: undefined reference to `destroy_plan_3d__' > :563: undefined reference to `destroy_plan_3d__' > :563: undefined reference to `destroy_plan_3d__' > :564: undefined reference to `destroy_plan_3d__' > :564: undefined reference to `destroy_plan_3d__' > .../Modules/fft_scalar.o::564: more undefined references to `destroy_plan_3d__' follow > .../Modules/fft_scalar.o: In function `$fft_scalar$cfft3d': > :565: undefined reference to `create_plan_3d__' > :565: undefined reference to `create_plan_3d__' > :565: undefined reference to `create_plan_3d__' > :566: undefined reference to `create_plan_3d__' > :566: undefined reference to `create_plan_3d__' > .../Modules/fft_scalar.o::566: more undefined references to `create_plan_3d__' follow > .../Modules/fft_scalar.o: In function `$fft_scalar$cfft3d': > :594: undefined reference to `fftw_inplace_drv_3d__' > :594: undefined reference to `fftw_inplace_drv_3d__' > :594: undefined reference to `fftw_inplace_drv_3d__' > :601: undefined reference to `fftw_inplace_drv_3d__' > :601: undefined reference to `fftw_inplace_drv_3d__' > .../Modules/fft_scalar.o::601: more undefined references to `fftw_inplace_drv_3d__' follow > .../Modules/fft_scalar.o: In function `$fft_scalar$cfft3ds': > :738: undefined reference to `create_plan_1d__' > :738: undefined reference to `create_plan_1d__' > :738: undefined reference to `create_plan_1d__' > :739: undefined reference to `create_plan_1d__' > :739: undefined reference to `create_plan_1d__' > .../Modules/fft_scalar.o::739: more undefined references to `create_plan_1d__' follow > .../Modules/fft_scalar.o: In function `$fft_scalar$cfft3ds': > :794: undefined reference to `fftw_inplace_drv_1d__' > :794: undefined reference to `fftw_inplace_drv_1d__' > :794: undefined reference to `fftw_inplace_drv_1d__' > :815: undefined reference to `fftw_inplace_drv_1d__' > :815: undefined reference to `fftw_inplace_drv_1d__' > .../Modules/fft_scalar.o::815: more undefined references to `fftw_inplace_drv_1d__' follow > .../Modules/fft_scalar.o: In function `$fft_scalar$cft_b': > :993: undefined reference to `create_plan_1d__' > :993: undefined reference to `create_plan_1d__' > :993: undefined reference to `create_plan_1d__' > :995: undefined reference to `destroy_plan_2d__' > :995: undefined reference to `destroy_plan_2d__' > :995: undefined reference to `destroy_plan_2d__' > :996: undefined reference to `create_plan_2d__' > :996: undefined reference to `create_plan_2d__' > :996: undefined reference to `create_plan_2d__' > :1034: undefined reference to `fftw_inplace_drv_1d__' > :1034: undefined reference to `fftw_inplace_drv_1d__' > :1034: undefined reference to `fftw_inplace_drv_1d__' > :1035: undefined reference to `fftw_inplace_drv_2d__' > :1035: undefined reference to `fftw_inplace_drv_2d__' > :1035: undefined reference to `fftw_inplace_drv_2d__' > collect2: ld returned 1 exit status > fort: Severe: Failed while trying to link. > make[1]: *** [memory] Error 1 > make[1]: Leaving directory `/home/proffess/PWscf-1.3/PW' > make: *** [pw] Error 2 > > What is problem? > > Sergey > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum -- Guido Roma -- CEA-Saclay - DEN/DMN/SRMP Bat.520/130 Phone: [+33]-1-69085738 -- Fax: ...6867 -- Mobile: [+33]-6-20069085 From giannozz at nest.sns.it Tue Sep 16 15:06:32 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 16 Sep 2003 15:06:32 +0200 Subject: [Pw_forum] Fw: Compiling error of Pw-1.3 In-Reply-To: <3F66C92D.00000C.26967@colgate.yandex.ru> References: <3F66A571.000008.14217@tide.yandex.ru> <3F66B3BC.939B8E3D@cea.fr> <3F66C92D.00000C.26967@colgate.yandex.ru> Message-ID: <200309161506.32580.giannozz@nest.sns.it> On Tuesday 16 September 2003 10:26, Sergei Lisenkov wrote: >[...] But I found problems on linking phase: > ../Modules/fft_scalar.o: In function `$fft_scalar$cft_1z': > :179: undefined reference to `create_plan_1d__' [...] funny, I was convinced that problems with underscores were solved once and for all ... apparently your compiler is adding two underscores to symbols. This is not supposed to happen on Alpha machines with Compaq compiler. BTW: in the makefile for HP PA-RISC (version 1.3), line: LIBS = L/opt/mlib/lib/pa2.0 -lveclib -llapack $(FFTW_LIB) -lm should be LIBS = -L/opt/mlib/lib/pa2.0 -lveclib -llapack $(FFTW_LIB) -lm and line: LFLAGS = +U77 +DA2.0 +DS2.0 +O2 -lpthread $(LIBS) should be: LDFLAGS = $(OSHOME)/flib/ptools.a $(OSHOME)/flib/flib.a $(OSHOME)/clib/clib.a +U77 +DA2.0 +DS2.0 +O2 -lpthread $(LIBS) Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From giannozz at nest.sns.it Tue Sep 16 15:15:49 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Tue, 16 Sep 2003 15:15:49 +0200 Subject: [Pw_forum] Fw: Compiling error of Pw-1.3 In-Reply-To: <3F66B3BC.939B8E3D@cea.fr> References: <3F66A571.000008.14217@tide.yandex.ru> <3F66B3BC.939B8E3D@cea.fr> Message-ID: <200309161515.49438.giannozz@nest.sns.it> On Tuesday 16 September 2003 08:54, Guido Roma wrote: > > I think that the problems comes from the following lines: > > #if defined __FFT_MODULE_DRV > use fft_scalar, only : cft_1z, cft_2xy > use sticks, only: dfftp > #endif > > which should better look like this: > > #if defined __FFT_MODULE_DRV > use fft_scalar, only : cft_1z, cft_2xy > #endif > use sticks, only: dfftp > > can you confirm, Paolo? I cannot verify your proposal, but it looks reasonable ... > I don't know if it works in parallel with DXML it doesn't. A driver for fft routines contained in the dxml/cxml/whatever they are called now/ libraries exists only for the serial case. Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From zhufhou at yahoo.com Wed Sep 17 03:53:41 2003 From: zhufhou at yahoo.com (Zhu-feng Hou) Date: Tue, 16 Sep 2003 18:53:41 -0700 (PDT) Subject: [Pw_forum] The makefile for FPMD is absent in fpmd.tgz Message-ID: <20030917015341.22442.qmail@web20603.mail.yahoo.com> Dear pwscf users: I have downloaded the 'fpmd.tgz' from the site(http://www.democritos.it/download/sw/fpmd.tgz). After I decompressed it, in the directory of 'FPMD' I have not found the makefile for 'fpmd'. Who can tell me where the makefile for 'FPMD' is. Thanks alot for your kindly help! __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From proffess at yandex.ru Wed Sep 17 07:34:55 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Wed, 17 Sep 2003 09:34:55 +0400 (MSD) Subject: [Pw_forum] The makefile for FPMD is absent in fpmd.tgz In-Reply-To: <20030917015341.22442.qmail@web20603.mail.yahoo.com> References: <20030917015341.22442.qmail@web20603.mail.yahoo.com> Message-ID: <3F67F27F.00000B.04235@ariel.yandex.ru> You should download the pwscf code that contains the Makefiles for FPMD. Simply type "./configure your_platform", next modify the file "make.sys" with your local configuration ant type "mkae fpmd". Good luck, Sergey From giannozz at nest.sns.it Wed Sep 17 09:24:20 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 17 Sep 2003 09:24:20 +0200 Subject: [Pw_forum] The makefile for FPMD is absent in fpmd.tgz In-Reply-To: <20030917015341.22442.qmail@web20603.mail.yahoo.com> References: <20030917015341.22442.qmail@web20603.mail.yahoo.com> Message-ID: <200309170924.20269.giannozz@nest.sns.it> On Wednesday 17 September 2003 03:53, Zhu-feng Hou wrote: > I have downloaded the 'fpmd.tgz' from the > site(http://www.democritos.it/download/sw/fpmd.tgz). > After I decompressed it, in the directory of 'FPMD' I > have not found the makefile for 'fpmd'. please ?!? > Who can tell me where the makefile for 'FPMD' is. in subdirectory install/ Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From zhufhou at yahoo.com Wed Sep 17 11:45:01 2003 From: zhufhou at yahoo.com (Zhu-feng Hou) Date: Wed, 17 Sep 2003 02:45:01 -0700 (PDT) Subject: [Pw_forum] The makefile for FPMD is absent in fpmd.tgz In-Reply-To: <200309170924.20269.giannozz@nest.sns.it> Message-ID: <20030917094501.52996.qmail@web20604.mail.yahoo.com> Dear pwscf user: In my apprehension, the makefiles in the subdirectory '/install' are just for all program. If one type "./configure platform", it just creat the local configuration(writen in the files: Makefile, make.sys and make.rules). It can't create the particular 'Makefile' for 'fpmd', which should be in the subdirectory 'FPMD'. If one take some minutes to browse the subdirectories of 'CPV', 'PW', 'PH','PP','D3' and 'Gamma', in each subdirectory there is a particular 'Makefile' for the program. However in the subdirectory 'FPMD', the particular 'Makefile' for 'fpmd' is missing. After 'configure platform' and 'make fpmd', there would occur some errores as follows: -------------- ( cd FPMD; make all ) make[1]: Entering directory `/home/xxx/pwscf1.3/FPMD' make[1]: *** No rule to make target `all'. Stop. make[1]: Leaving directory `/home/xxx/pwscf1.3/FPMD' make: *** [fpmd] Error 1 ------------------- I think that this error just coming from a 'makefile' missing in the subdirectory 'FPMD'. Thanks for your help! __________________________________ Do you Yahoo!? Yahoo! SiteBuilder - Free, easy-to-use web site design software http://sitebuilder.yahoo.com From giannozz at nest.sns.it Wed Sep 17 11:55:42 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 17 Sep 2003 11:55:42 +0200 Subject: [Pw_forum] The makefile for FPMD is absent in fpmd.tgz In-Reply-To: <20030917094501.52996.qmail@web20604.mail.yahoo.com> References: <20030917094501.52996.qmail@web20604.mail.yahoo.com> Message-ID: <200309171155.42209.giannozz@nest.sns.it> On Wednesday 17 September 2003 11:45, Zhu-feng Hou wrote: > I think that this error just coming from a 'makefile' > missing in the subdirectory 'FPMD'. oops, you are right, the file on Democritos web site for FPMD contains only the Makefile for CP ... Here it is Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 -------------- next part -------------- A non-text attachment was scrubbed... Name: Makefile Type: text/x-makefile Size: 1954 bytes Desc: not available Url : /pipermail/attachments/20030917/1f804231/attachment.bin From proffess at yandex.ru Wed Sep 17 13:19:32 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Wed, 17 Sep 2003 15:19:32 +0400 (MSD) Subject: [Pw_forum] Fw: Compiling error Of PWscf-1.3 Message-ID: <3F684344.000003.10252@camay.yandex.ru> Dear Paolo, I tried to use local copy of FFTW, but I got the next on my Alpha machine. mpicc -O -D__ALPHA -D__PARA -D__MPI -D__FFTW -D__USE_INTERNAL_FFTW -I/home/proffess/PWscf-1.3/include -I./ -c fft_stick.c cc: Error: fftw.c, line 159: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_1; -------^ cc: Error: fftw.c, line 160: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_2; -------^ cc: Error: fftw.c, line 161: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_3; -------^ cc: Error: fftw.c, line 162: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_4; -------^ cc: Error: fftw.c, line 163: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_5; -------^ cc: Error: fftw.c, line 164: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_6; -------^ cc: Error: fftw.c, line 165: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_7; -------^ cc: Error: fftw.c, line 166: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_8; -------^ cc: Error: fftw.c, line 167: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_9; -------^ cc: Error: fftw.c, line 168: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_10; -------^ cc: Error: fftw.c, line 169: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_11; -------^ cc: Error: fftw.c, line 170: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_12; -------^ cc: Error: fftw.c, line 171: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_13; -------^ cc: Error: fftw.c, line 172: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_14; -------^ cc: Error: fftw.c, line 173: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_15; -------^ cc: Error: fftw.c, line 174: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_16; -------^ cc: Error: fftw.c, line 175: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_32; -------^ cc: Error: fftw.c, line 176: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_64; -------^ cc: Error: fftw.c, line 178: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_1; -------^ cc: Error: fftw.c, line 179: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_2; -------^ cc: Error: fftw.c, line 180: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_3; -------^ cc: Error: fftw.c, line 181: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_4; -------^ cc: Error: fftw.c, line 182: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_5; -------^ cc: Error: fftw.c, line 183: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_6; -------^ cc: Error: fftw.c, line 184: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_7; -------^ cc: Error: fftw.c, line 185: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_8; -------^ cc: Error: fftw.c, line 186: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_9; -------^ cc: Error: fftw.c, line 187: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_10; -------^ cc: Error: fftw.c, line 188: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_11; -------^ cc: Error: fftw.c, line 189: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_12; -------^ cc: Error: fftw.c, line 190: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_13; -------^ cc: Severe: More than 30 errors were encountered in the course of compilation. (toomanyerr) make: *** [fft_stick.o] Error 1 Any comments? Very strange, but I haven't problem with underscores with compiling previuos versions. Sergey From giannozz at nest.sns.it Wed Sep 17 15:38:26 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 17 Sep 2003 15:38:26 +0200 Subject: [Pw_forum] Fw: Compiling error Of PWscf-1.3 In-Reply-To: <3F684344.000003.10252@camay.yandex.ru> References: <3F684344.000003.10252@camay.yandex.ru> Message-ID: <200309171538.26578.giannozz@nest.sns.it> On Wednesday 17 September 2003 13:19, Sergei Lisenkov wrote: > I tried to use local copy of FFTW, but I got the next on my Alpha machine. > > mpicc -O -D__ALPHA -D__PARA -D__MPI -D__FFTW -D__USE_INTERNAL_FFTW > -I/home/proffess/PWscf-1.3/include -I./ -c fft_stick.c cc: Error: fftw.c, > line 159: In this declaration, "notw_codelet" appears to be used as if it > named a type, but there is no declared type of that name visible. > (typedefnotdef) extern notw_codelet fftw_no_twiddle_1; try to use "cc" instead of "mpicc" -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From lucre at ms1.hinet.net Thu Sep 18 06:37:31 2003 From: lucre at ms1.hinet.net (lucre at ms1.hinet.net) Date: Thu, 18 Sep 2003 12:37:31 +0800 (CST) Subject: [Pw_forum] Ga 3d core electrons as valence electrons (fhi2upf) Message-ID: <200309180437.MAA09095@msr83.hinet.net> An HTML attachment was scrubbed... URL: /pipermail/attachments/20030918/cbca12a2/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: zbGaN3d_ABINIT.GIF Type: image/gif Size: 6564 bytes Desc: not available Url : /pipermail/attachments/20030918/cbca12a2/attachment.gif -------------- next part -------------- A non-text attachment was scrubbed... Name: zbGaN3d_PWscf.GIF Type: image/gif Size: 6137 bytes Desc: not available Url : /pipermail/attachments/20030918/cbca12a2/attachment-0001.gif From lucre at ms1.hinet.net Thu Sep 18 07:38:41 2003 From: lucre at ms1.hinet.net (Lu Frederic) Date: Thu, 18 Sep 2003 13:38:41 +0800 Subject: [Pw_forum] Ga 3d core electrons as valence electrons (fhi2upf) Message-ID: <001f01c37da7$21ff8470$ef3310ac@phys> Dear PWscf users, I am very sorry for previous E-mail format which is delivered by WebMail. I have generated Ga 3d PP with FHI98PP, and check by band structure of zb GaN with ABINIT. It seems ok. But when I use the same PP files to transforms as UPF format, check it with PWscf, I got the wrong band structure. I have seen before in ABINIT, caused by impropriety of lloc value. I guess it's same problem. It is not the problem of Ecut(ecutwfc). I can't find any variable to set lloc in UPF format. May be it is locate somewhere else? If someone has experiences to treat Ga 3d as valence electrons with PWscf, Would you please to share your some recipes or knowhow ? Thank you! ======================================================= Lu Fu-Fa, Doctoral Student, E-mail: g930102 at ccit.edu.tw Semiconductors Lab., Dept. of Applied Physics Chung Cheng Institute of Technology (CCIT), Taiwan ======================================================= Here is my process of FHI2UPF $ ../bin/fhi2upf input PP file in FHI format > 31-Ga-3d-fhil3.cpi Pseudopotential successfully read Atom name > Ga l local > 2 DFT > PW91 Wavefunction # 1: label, occupancy > 4S, 2 Wavefunction # 2: label, occupancy > 4P, 1 Wavefunction # 3: label, occupancy > 3D, 10 Wavefunction # 4: label, occupancy > 4F, 0 Pseudopotential successfully converted Output PP file in UPF format : 31-Ga-3d-fhil3.cpi.UPF -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030918/db839815/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: zbGaN3d_PWscf.GIF Type: image/gif Size: 6137 bytes Desc: not available Url : /pipermail/attachments/20030918/db839815/attachment.gif -------------- next part -------------- A non-text attachment was scrubbed... Name: zbGaN3d_ABINIT.GIF Type: image/gif Size: 6564 bytes Desc: not available Url : /pipermail/attachments/20030918/db839815/attachment-0001.gif From francesco.antoniella at aquila.infn.it Thu Sep 18 14:16:15 2003 From: francesco.antoniella at aquila.infn.it (Francesco Antoniella) Date: 18 Sep 2003 14:16:15 +0200 Subject: [Pw_forum] Ga 3d core electrons as valence electrons (fhi2upf) In-Reply-To: <001f01c37da7$21ff8470$ef3310ac@phys> References: <001f01c37da7$21ff8470$ef3310ac@phys> Message-ID: <1063887385.2292.11.camel@altair> Il gio, 2003-09-18 alle 07:38, Lu Frederic ha scritto: > Dear PWscf users, > > I am very sorry for previous E-mail format which is delivered by WebMail. > > I have generated Ga 3d PP with FHI98PP, and check by band structure of > zb GaN with ABINIT. It seems ok. But when I use the same PP files to > transforms as UPF format, check it with PWscf, I got the wrong band structure. Are you using nonlinear core correction? This is neglected in the conversion so if you used it in the pseudo generation you will get a somewath different pseudo than expected. > > I have seen before in ABINIT, caused by impropriety of lloc value. I guess it's > same problem. It is not the problem of Ecut(ecutwfc). > You will not find any lloc in UPF simply because this info is hidden inside the conversion. You must supply it by hand in the input of the converter. > I can't find any variable to set lloc in UPF format. May be it is locate somewhere else? > > If someone has experiences to treat Ga 3d as valence electrons with PWscf, > Would you please to share your some recipes or knowhow ? > From yanming_ma at hotmail.com Thu Sep 18 15:52:47 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Thu, 18 Sep 2003 21:52:47 +0800 Subject: [Pw_forum] Ga 3d core electrons as valence electrons (fhi2upf) Message-ID: Hi, Actually, you do not need to convert UPF format at this moment. After you generate the PP using FHI, you can simply convert to PWSCF old PP format using PPc.x. But be sure, no core correction included. Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030918/93dba39e/attachment.htm From pmodak at magnum.barc.ernet.in Fri Sep 19 08:52:10 2003 From: pmodak at magnum.barc.ernet.in (Paritosh Modak) Date: Fri, 19 Sep 2003 12:22:10 +0530 (IST) Subject: [Pw_forum] Re:pw-1.3.0 not compiling in pc_abs (fwd) Message-ID: ---------- Forwarded message ---------- Date: Fri, 19 Sep 2003 12:07:46 +0530 (IST) From: Paritosh Modak To: pw_users at pwscf.org Subject: Re:pw-1.3.0 not compiling in pc_abs Dear pw_users, I have down loaded the recent 1.3 version of PWSCF code.When I untar the package I found pw_examples subdirectory is not there. Then I tried to compile in pc_abs but failed. It gives the error "when declared implicit none, then the variables should be defined explicitly. P. Modak On Wed, 17 Sep 2003 pw_users-request at pwscf.org wrote: > Send Pw_users mailing list submissions to > pw_users at pwscf.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://www.democritos.it/mailman/listinfo/pw_users > or, via email, send a message with subject or body 'help' to > pw_users-request at pwscf.org > > You can reach the person managing the list at > pw_users-admin at pwscf.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Pw_users digest..." > > > Today's Topics: > > 1. New PWSCF version (Paolo Giannozzi) > > --__--__-- > > Message: 1 > From: Paolo Giannozzi > Organization: Scuola Normale Superiore di Pisa > To: pw_users at pwscf.org > Date: Tue, 16 Sep 2003 14:47:12 +0200 > Subject: [Pw_users] New PWSCF version > Reply-To: pw_users at pwscf.org > > A new version (1.3) of the PWscf package is available for > download on the web site http://www.pwscf.org. The new > version contains many changes, improvements, bug fixes. > Everybody who is using the PWscf package is encouraged > to upgrade and to report problems to the mailing list. > > This release is another step forward towards full integration > and interoperability of PWscf with the Car-Parrinello codes > FPMD and CP. The entire updated package is available for > download from http://www.democritos.it > > Major changes: > - various bug fixes > - various compilation/installation problems fixed > - unified installation procedure for all codes > - external libraries are no longer required > (but it may still be convenient to use them) > - new features: > - Berry phase calculation of polarization > (contributed by Vanderbilt group in Rutgers University) > - Calculation of projected DOS > - Possibility to add an external electric field > - Dipolar correction (for surface calculation) > > Incompatibilities with previous versions: > - a make.sys file must be re-generated for your installation > (do not just re-use those from v.1.2 !) > - files produced by preceding versions may be unreadable > > -- > Paolo Giannozzi e-mail: giannozz at nest.sns.it > Scuola Normale Superiore Phone: +39/050509412 > Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 > I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 > > > > --__--__-- > > _______________________________________________ > Pw_users mailing list > Pw_users at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_users > > > End of Pw_users Digest > From proffess at yandex.ru Tue Sep 16 07:25:58 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Tue, 16 Sep 2003 09:25:58 +0400 (MSD) Subject: [Pw_forum] Compiling error of Pw-1.3 In-Reply-To: <3F37463E.7070700@sissa.it> References: <3EE86CDE.000001.22552@pantene.yandex.ru> <200306121547.02106.giannozz@nest.sns.it> <3F372EC0.000004.05643@camay.yandex.ru> <3F37463E.7070700@sissa.it> Message-ID: <3F669EE6.000009.31663@pantene.yandex.ru> Dear PW users, I have error with compiling the new version of PWscf-1.3 on Compaq Alpha clusters using Compaq Compiler: ..... mpif90 -O -fpe0 -real_size 64 -align dcommons -align records -free -cpp -D__ALPHA -D__PARA -D__MPI -DDXML -I/home/proffess/PWscf-1.3/include -I/home/proffess/PWscf-1.3/Modules -I/home/proffess/PWscf-1.3/PW -I/home/proffess/PWscf-1.3/PH -c cft3.f90 f90: Error: cft3.f90, line 79: This name does not have a type, and must have an explicit type. [DFFTP] mc = dfftp%ismap (i) -------------^ make[1]: *** [cft3.o] Error 1 make[1]: Leaving directory `/home/proffess/PWscf-1.3/PW' Is anybody know how to solve it? Thanks, Sergey Lisenkov From proffess at yandex.ru Wed Sep 17 13:08:25 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Wed, 17 Sep 2003 15:08:25 +0400 (MSD) Subject: [Pw_forum] Compiling error Of PWscf-1.3 In-Reply-To: <3F37463E.7070700@sissa.it> References: <3EE86CDE.000001.22552@pantene.yandex.ru> <200306121547.02106.giannozz@nest.sns.it> <3F372EC0.000004.05643@camay.yandex.ru> <3F37463E.7070700@sissa.it> Message-ID: <3F6840A9.000001.16530@pantene.yandex.ru> Dear Paolo, I tried to use local copy of FFTW, but I got the next on my Alpha machine. mpicc -O -D__ALPHA -D__PARA -D__MPI -D__FFTW -D__USE_INTERNAL_FFTW -I/home/proffess/PWscf-1.3/include -I./ -c fft_stick.c cc: Error: fftw.c, line 159: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_1; -------^ cc: Error: fftw.c, line 160: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_2; -------^ cc: Error: fftw.c, line 161: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_3; -------^ cc: Error: fftw.c, line 162: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_4; -------^ cc: Error: fftw.c, line 163: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_5; -------^ cc: Error: fftw.c, line 164: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_6; -------^ cc: Error: fftw.c, line 165: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_7; -------^ cc: Error: fftw.c, line 166: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_8; -------^ cc: Error: fftw.c, line 167: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_9; -------^ cc: Error: fftw.c, line 168: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_10; -------^ cc: Error: fftw.c, line 169: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_11; -------^ cc: Error: fftw.c, line 170: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_12; -------^ cc: Error: fftw.c, line 171: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_13; -------^ cc: Error: fftw.c, line 172: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_14; -------^ cc: Error: fftw.c, line 173: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_15; -------^ cc: Error: fftw.c, line 174: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_16; -------^ cc: Error: fftw.c, line 175: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_32; -------^ cc: Error: fftw.c, line 176: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftw_no_twiddle_64; -------^ cc: Error: fftw.c, line 178: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_1; -------^ cc: Error: fftw.c, line 179: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_2; -------^ cc: Error: fftw.c, line 180: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_3; -------^ cc: Error: fftw.c, line 181: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_4; -------^ cc: Error: fftw.c, line 182: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_5; -------^ cc: Error: fftw.c, line 183: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_6; -------^ cc: Error: fftw.c, line 184: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_7; -------^ cc: Error: fftw.c, line 185: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_8; -------^ cc: Error: fftw.c, line 186: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_9; -------^ cc: Error: fftw.c, line 187: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_10; -------^ cc: Error: fftw.c, line 188: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_11; -------^ cc: Error: fftw.c, line 189: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_12; -------^ cc: Error: fftw.c, line 190: In this declaration, "notw_codelet" appears to be used as if it named a type, but there is no declared type of that name visible. (typedefnotdef) extern notw_codelet fftwi_no_twiddle_13; -------^ cc: Severe: More than 30 errors were encountered in the course of compilation. (toomanyerr) make: *** [fft_stick.o] Error 1 Any comments? Very strange, but I haven't problem with underscores with compiling previuos versions. Sergey From felipe at titan.ipicyt.edu.mx Fri Sep 19 17:01:02 2003 From: felipe at titan.ipicyt.edu.mx (FELIPE VALENCIA) Date: Fri, 19 Sep 2003 10:01:02 -0500 (CDT) Subject: [Pw_forum] Re:pw-1.3.0 not compiling in pc_abs (fwd) In-Reply-To: Message-ID: Dear Paritosh, Concernig the pw_examples directories: examples come within a different tarball -available also at the download section of the web page- , they are not included in the main distribution. By the other hand, several compilation problems with the new distro have been reported in the past two years... ?does your error message have something to do with the cft3.o module? If this is the case you may find usefull previous discussions on that field. Felipe > Dear pw_users, > I have down loaded the recent 1.3 version of PWSCF code.When > I untar the package I found pw_examples subdirectory is not there. Then I > tried to compile in pc_abs but failed. It gives the error "when declared > implicit none, then the variables should be defined explicitly. > > P. Modak > > On Wed, 17 Sep 2003 pw_users-request at pwscf.org wrote: > > > Send Pw_users mailing list submissions to > > pw_users at pwscf.org > > > > To subscribe or unsubscribe via the World Wide Web, visit > > http://www.democritos.it/mailman/listinfo/pw_users > > or, via email, send a message with subject or body 'help' to > > pw_users-request at pwscf.org > > > > You can reach the person managing the list at > > pw_users-admin at pwscf.org > > > > When replying, please edit your Subject line so it is more specific > > than "Re: Contents of Pw_users digest..." > > > > > > Today's Topics: > > > > 1. New PWSCF version (Paolo Giannozzi) > > > > --__--__-- > > > > Message: 1 > > From: Paolo Giannozzi > > Organization: Scuola Normale Superiore di Pisa > > To: pw_users at pwscf.org > > Date: Tue, 16 Sep 2003 14:47:12 +0200 > > Subject: [Pw_users] New PWSCF version > > Reply-To: pw_users at pwscf.org > > > > A new version (1.3) of the PWscf package is available for > > download on the web site http://www.pwscf.org. The new > > version contains many changes, improvements, bug fixes. > > Everybody who is using the PWscf package is encouraged > > to upgrade and to report problems to the mailing list. > > > > This release is another step forward towards full integration > > and interoperability of PWscf with the Car-Parrinello codes > > FPMD and CP. The entire updated package is available for > > download from http://www.democritos.it > > > > Major changes: > > - various bug fixes > > - various compilation/installation problems fixed > > - unified installation procedure for all codes > > - external libraries are no longer required > > (but it may still be convenient to use them) > > - new features: > > - Berry phase calculation of polarization > > (contributed by Vanderbilt group in Rutgers University) > > - Calculation of projected DOS > > - Possibility to add an external electric field > > - Dipolar correction (for surface calculation) > > > > Incompatibilities with previous versions: > > - a make.sys file must be re-generated for your installation > > (do not just re-use those from v.1.2 !) > > - files produced by preceding versions may be unreadable > > > > -- > > Paolo Giannozzi e-mail: giannozz at nest.sns.it > > Scuola Normale Superiore Phone: +39/050509412 > > Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 > > I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 > > > > > > > > --__--__-- > > > > _______________________________________________ > > Pw_users mailing list > > Pw_users at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_users > > > > > > End of Pw_users Digest > > > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From michele.lazzeri at lmcp.jussieu.fr Fri Sep 19 18:01:46 2003 From: michele.lazzeri at lmcp.jussieu.fr (Michele Lazzeri) Date: Fri, 19 Sep 2003 18:01:46 +0200 Subject: [Pw_forum] J.Breeze on:Third order anharmonic dynamical matrix Message-ID: <3F6B286A.F3D02836@lmcp.jussieu.fr> >Using the PWSCF D3 code, should it be possible to calculate the >coefficients for phonons at arbitraty wave-vector? >That is, to calculate D^3(-q1-q2,q1,q2). >Jonathan Breeze Dear Jonathan, As you know the code calculates D^3(0,-q, q). More in general, to compute D^3(-q1-q2,q1,q2) is possible, but the modification to be done to the code could be tedious. In alternative, you could use a super-cell approach: If you want to compute D^3(Q,Q-q, Q+q), you can compute D^3(0,-q,q) in a super-cell such that the gamma-point of the super-cell is equivalent to the Q point of the unit-cell. An example: if you want to compute the broadening of a phonon at the border zone, you can use a crystal obtained doubling the unit-cell in the appropriate direction. The phonon you are interested in is a phonon at Gamma for this double-cell. So, if you compute the D3 matrix at gamma, for this double-cell, you obtain all the information that you need. For the moment, this is the most practical thing you can do if you are interested in high-symmetry points. This procedure cannot be applied to any point Q of the Brillouin Zone, and, depending on the point, can be computationally expensive. Michele Lazzeri -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030919/35df39d4/attachment.htm From felipe at titan.ipicyt.edu.mx Fri Sep 19 19:54:54 2003 From: felipe at titan.ipicyt.edu.mx (FELIPE VALENCIA) Date: Fri, 19 Sep 2003 12:54:54 -0500 (CDT) Subject: [Pw_forum] Cray compilation problems report. In-Reply-To: <3F6840A9.000001.16530@pantene.yandex.ru> Message-ID: Hi there, I have been trying to compile the latest version of pwscf on a cray machine, and have found some minor (apparentely) problems. First, when compiling with the cray scilib library, I encountered the same "dfftp undeclared"/"something like that" problem that has been reported by other guys in the mailist, and used the Guido Roma proposal to solve it. It is remarkable that this error dissappears when compiling against the included fftw. Then, several references to double precision routines remained in the distro (I guess some precompiling options are still lacking): The symbol `ZDOTC' referenced in relocatable object `bp_c_phase.o:C_PHASE' is not defined. cld-404 cld: WARNING The symbol `IZAMAX' referenced in relocatable object `bp_zgefa.o:ZGEFA' is not defined. cld-404 cld: WARNING The symbol `ZSCAL' referenced in relocatable object `bp_zgefa.o:ZGEFA' is not defined. cld-404 cld: WARNING The symbol `ZAXPY' referenced in relocatable object `bp_zgefa.o:ZGEFA' is not defined. cld-404 cld: WARNING The symbol `ZSWAP' referenced in relocatable object `bp_zgedi.o:ZGEDI' is not defined. cld-404 cld: WARNING The symbol `DSPEV' referenced in relocatable object `../Modules/ptoolkit.o:DSPEV_DRV_in_PARALLEL_TOOLKIT' is not defined. cld-404 cld: WARNING The symbol `ZHPEV' referenced in relocatable object `../Modules/ptoolkit.o:CDIAGONALIZE_in_PARALLEL_TOOLKIT' is not defined. By now, I have solved those problems by linking with a file including the double to single precision equivalents ( equiv (DSCAL) =SSCAL , and the like) but I guess the precompiling approach would be better ;-) Finally I found a reference to GETARG (a routine that is not included in each Cray T3E machine and, particulary, was not included in the machine i was compiling) in the startup.f90 file ... This step was quite simple and I just correct it by changing a bit the startup.f90 file: #ifdef __T3E do iiarg=1,nargs-1 call PXFGETARG (iiarg, np,ilen,ierr) if (trim(np) == '-npool' .or. trim(np) == '-npools' ) then call PXFGETARG (iiarg+1, np,ilen,ierr) read (np,*) npool end if end do #else do iiarg=1,nargs-1 call GETARG (iiarg, np) if (trim(np) == '-npool' .or. trim(np) == '-npools' ) then call GETARG (iiarg+1, np) read (np,*) npool end if end do #endif I used a similar approach in plotrho.f90 , plotband.f90, Well, I just tested the pw.x program with one simple example and it seems to work fine. Good Luck, Felipe From giannozz at nest.sns.it Sat Sep 20 20:47:04 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sat, 20 Sep 2003 20:47:04 +0200 Subject: [Pw_forum] Cray compilation problems report. In-Reply-To: References: Message-ID: <200309202047.05499.giannozz@nest.sns.it> On Friday 19 September 2003 19:54, FELIPE VALENCIA wrote: > I have been trying to compile the latest version of pwscf on a cray > machine, and have found some minor (apparentely) problems. thank you for reporting problems in an understandable way, and solutions for them. Most of the problems you mention have been meanwhile fixed and the remaining will be soon. An additional one: in Modules/stick_bas.f90, you need to change INTEGER, INTENT(OUT) :: ... to INTEGER, INTENT(INOUT) :: ... in lines 423-425. Paolo From yanming_ma at hotmail.com Sat Sep 20 15:17:39 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Sat, 20 Sep 2003 21:17:39 +0800 Subject: [Pw_forum] Question for electron phonon coulping Message-ID: Dear All, I have a question concerning the electron-phonon coupling. I increase the parameter "nsig' to 40 in elphon.f90. So in my case I have 40 Gaussian Broadening steps. The calculated lambda(q) is negative number until 35 step, i.e., Gaussian Broadening: 0.35 Ry. From xhongjun at mail.ustc.edu.cn Sat Sep 20 15:29:00 2003 From: xhongjun at mail.ustc.edu.cn (xhongjun) Date: Sat, 20 Sep 2003 21:29:00 +0800 Subject: [Pw_forum] Question for electron phonon coulping In-Reply-To: References: Message-ID: <3F6C561C.7030008@mail.ustc.edu.cn> Hello ma Yanming: Please decrease the Gaussian Broadening value and try again. May it works. Regards. xianghjun ma Yanming wrote: > > Dear All, > > I have a question concerning the electron-phonon coupling. > I increase the parameter "nsig' to 40 in elphon.f90. > So in my case I have 40 Gaussian Broadening steps. The calculated > lambda(q) is negative number until 35 step, i.e., Gaussian Broadening: > 0.35 Ry. > >> From step 36, the lambda(q) comes positive and increase gradully. Until > > step 40 it continue increasing. It seems that it still hasn't converged. > I am not clear the principle for this Gaussian broadening procedures. In > my case, the Fermi energy is around 19.47 eV. I am wondering if it is > reasonable if I take the positive number (in the final 40 step) as final > lambda value. But at the step 40, the calculated lambda is at fermi > energy 18.9 eV which is far away from 19.47eV. > > Any idea on the principle for Gaussian broadening procedures. Can I take > the positive lambda after 40 step Gaussian broadening and by 35 step > negative number? > > THanks > > > > Yanming Ma PhD > Steacie Institute for Molecular Sciences, > National Research Councils of Canada. > 100 Sussex > K1A 0R6 > > _________________________________________________________________ > ?????????????? MSN Messenger: http://messenger.msn.com/cn > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > > -- =================================================== HongJun Xiang Ph.D. Candidate Open Laboratory of Bond Selective Chemistry, University of Science and Technology of China, Hefei, Anhui, 230026, People's Republic of China Tel.: 86-551-3603748,3603418 Fax.: 86-551-3602969 =================================================== From yanming_ma at hotmail.com Sun Sep 21 00:41:01 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Sun, 21 Sep 2003 06:41:01 +0800 Subject: [Pw_forum] Question for electron phonon coulping Message-ID: Dear Hongjun, Thanks for your reply. My Gaussian Broadening value is 0.01Ry. You mean I should decrease this number? Best Regards Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 >From: xhongjun >Reply-To: pw_forum at pwscf.org >To: pw_forum at pwscf.org >Subject: Re: [Pw_forum] Question for electron phonon coulping >Date: Sat, 20 Sep 2003 21:29:00 +0800 > >Hello ma Yanming: > Please decrease the Gaussian Broadening value and try again. >May it works. >Regards. >xianghjun > >ma Yanming wrote: > > > > Dear All, > > > > I have a question concerning the electron-phonon coupling. > > I increase the parameter "nsig' to 40 in elphon.f90. > > So in my case I have 40 Gaussian Broadening steps. The calculated > > lambda(q) is negative number until 35 step, i.e., Gaussian Broadening: > > 0.35 Ry. > > > >> From step 36, the lambda(q) comes positive and increase gradully. Until > > > > step 40 it continue increasing. It seems that it still hasn't converged. > > I am not clear the principle for this Gaussian broadening procedures. In > > my case, the Fermi energy is around 19.47 eV. I am wondering if it is > > reasonable if I take the positive number (in the final 40 step) as final > > lambda value. But at the step 40, the calculated lambda is at fermi > > energy 18.9 eV which is far away from 19.47eV. > > > > Any idea on the principle for Gaussian broadening procedures. Can I take > > the positive lambda after 40 step Gaussian broadening and by 35 step > > negative number? > > > > THanks > > > > > > > > Yanming Ma PhD > > Steacie Institute for Molecular Sciences, > > National Research Councils of Canada. > > 100 Sussex > > K1A 0R6 > > > > _________________________________________________________________ > > ?????????????? MSN Messenger: http://messenger.msn.com/cn > > _______________________________________________ > > Pw_forum mailing list > > Pw_forum at pwscf.org > > http://www.democritos.it/mailman/listinfo/pw_forum > > > > > > >-- >=================================================== > HongJun Xiang > Ph.D. Candidate > Open Laboratory of Bond Selective Chemistry, > University of Science and Technology of China, > Hefei, Anhui, 230026, > People's Republic of China > Tel.: 86-551-3603748,3603418 > Fax.: 86-551-3602969 >=================================================== > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum _________________________________________________________________ ?????????????? MSN Messenger: http://messenger.msn.com/cn From lucre at ms1.hinet.net Mon Sep 22 04:56:08 2003 From: lucre at ms1.hinet.net (Lu Frederic) Date: Mon, 22 Sep 2003 10:56:08 +0800 Subject: [Pw_forum] Re: Ga 3d core electrons as valence electrons (fhi2upf) Message-ID: <019401c380b5$22cd9480$ef3310ac@phys> Dear Dr. Ma and PWscf users, I have check all of PWscf after version 1.0, I can't find any information (include source codes and Makefile) with respect to PPc.x. Would you please provides more information in detail. By the way, I have done PP files using fhi98PP with no core correction included in my previous work, and the result just you have seen. Thank you! Lu Fu-Fa, 09/22/2003 From ferretti.andrea at unimore.it Mon Sep 22 08:59:24 2003 From: ferretti.andrea at unimore.it (Andrea Ferretti) Date: Mon, 22 Sep 2003 08:59:24 +0200 (CEST) Subject: [Pw_forum] Re: Ga 3d core electrons as valence electrons (fhi2upf) In-Reply-To: <019401c380b5$22cd9480$ef3310ac@phys> Message-ID: dear all, PPc.x is the old name of the code doing the conversion form FHI to PWSCF pseudopotential format.... at the moment (I think from the prevoius version) that code has been integrated in the "fhi2upf.f90" converter you can find in the directory upftools. regards andrea -- Andrea Ferretti INFM National Research Center on nanoStructures and bioSystems at Surfaces (S3) Dipartimento di Fisica, Universita' di Modena e Reggio Emilia Via Campi 213/A I-41100 Modena, Italy Tel: +39 059 2055283 Fax: +39 059 367488 E-mail: ferretti.andrea at unimore.it URL: http://www.nanoscience.unimo.it On Mon, 22 Sep 2003, Lu Frederic wrote: > Dear Dr. Ma and PWscf users, > > I have check all of PWscf after version 1.0, I can't find any information > (include source codes and Makefile) with respect to PPc.x. > Would you please provides more information in detail. > > By the way, I have done PP files using fhi98PP with no core correction > included > in my previous work, and the result just you have seen. > > Thank you! > > > Lu Fu-Fa, 09/22/2003 > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From alice.ruini at unimore.it Mon Sep 22 09:41:41 2003 From: alice.ruini at unimore.it (alice.ruini at unimore.it) Date: Mon, 22 Sep 2003 09:41:41 +0200 Subject: [Pw_forum] wow! In-Reply-To: Message-ID: che bello vedere la tua autorevole risposta sul forum pw!!!!!! mi potresti ridire l'ampiezza di banda dell'meh per piacere? per il sst sono veramente messa da panico, perche` questo we c'era troppo bel tempo! comunque ora mi metto SOLO su quello.... cioa, a -- Alice Ruini Tel: +39 059 2055300 INFM-S^3 and Dipartimento di Fisica FAX: +39 059 2055235 Universita` di Modena e Reggio E. E-mail: alice.ruini at unimore.it Via Campi 213/A URL: http://www.nanoscience.unimo.it/ I-41100 Modena, Italy http://www.s3.infm.it/ From degironc at sissa.it Mon Sep 22 14:54:04 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Mon, 22 Sep 2003 14:54:04 +0200 Subject: [Pw_forum] Question for electron phonon coulping References: Message-ID: <3F6EF0EC.7070807@sissa.it> One is interested in the limit for sigma going to zero with infinit number of k points. All other numbers are approximations to this results. How can you judge whether they are good approximations ? As usual by checking convergence. Do a calculation with a given set of k points and calculate el-ph as a function of sigma Do again the same calculation increasing the number of k-points and the same set of values for sigma. You will notice that for sufficiently large values of sigma the two curves will agrees. This means that the k-points mesh is sufficiently accurate for sigmas larger than that but may not be good for smaller values. Repeating the calculation for even larger set of k-points (same set of sigma) you will see that from a value of sigma that is smaller that the previous one the most accurate calculation (the one using more k-points) will agree with the second-best (and with the first, less accurate, calculation when these last two agree) Trust only the curve in the region in which the calculation is converged with respect to k-ponts summation. For sufficiently large set of k-points the SAFE region will approach enough the limit of sigma going to zero that you can stop the procedure. Stefano de Gironcoli ma Yanming wrote: > > Dear All, > > I have a question concerning the electron-phonon coupling. > I increase the parameter "nsig' to 40 in elphon.f90. > So in my case I have 40 Gaussian Broadening steps. The calculated > lambda(q) is negative number until 35 step, i.e., Gaussian Broadening: > 0.35 Ry. > >> From step 36, the lambda(q) comes positive and increase gradully. Until > > step 40 it continue increasing. It seems that it still hasn't > converged. I am not clear the principle for this Gaussian broadening > procedures. In my case, the Fermi energy is around 19.47 eV. I am > wondering if it is reasonable if I take the positive number (in the > final 40 step) as final lambda value. But at the step 40, the > calculated lambda is at fermi energy 18.9 eV which is far away from > 19.47eV. > > Any idea on the principle for Gaussian broadening procedures. Can I > take the positive lambda after 40 step Gaussian broadening and by 35 > step negative number? > > THanks > From yanming_ma at hotmail.com Mon Sep 22 17:12:13 2003 From: yanming_ma at hotmail.com (Yanming ma) Date: Mon, 22 Sep 2003 23:12:13 +0800 Subject: [Pw_forum] Re: Ga 3d core electrons as valence electrons (fhi2upf) Message-ID: Hi Yes, the PPc.x was replaced by the fhi2upf in upftools. I have a quick look on the fhi2upf. Actually, it is similar with PPc.x. One thing you should be careful is to set the local potential. This is decided during your ghost test in FHI PP generation. Others is quite straightforwad. good luck Yanming Ma PhD Steacie Institute for Molecular Sciences, National Research Councils of Canada. 100 Sussex K1A 0R6 >From: "Lu Frederic" >Reply-To: pw_forum at pwscf.org >To: "PW_forum" >Subject: [Pw_forum] Re: Ga 3d core electrons as valence electrons (fhi2upf) >Date: Mon, 22 Sep 2003 10:56:08 +0800 > >Dear Dr. Ma and PWscf users, > >I have check all of PWscf after version 1.0, I can't find any information >(include source codes and Makefile) with respect to PPc.x. >Would you please provides more information in detail. > >By the way, I have done PP files using fhi98PP with no core correction >included > in my previous work, and the result just you have seen. > >Thank you! > > >Lu Fu-Fa, 09/22/2003 > >_______________________________________________ >Pw_forum mailing list >Pw_forum at pwscf.org >http://www.democritos.it/mailman/listinfo/pw_forum _________________________________________________________________ Add MSN 8 Internet Software to your existing Internet access and enjoy patented spam protection and more. Sign up now! http://join.msn.com/?page=dept/byoa From yanming_ma at hotmail.com Mon Sep 22 20:18:21 2003 From: yanming_ma at hotmail.com (Yanming ma) Date: Tue, 23 Sep 2003 02:18:21 +0800 Subject: [Pw_forum] Question for electron phonon coulping Message-ID: Dear Stefano de Gironcoli, Thanks very much for your excellent comments. I am not sure if I can understand 100 percent of your comments . I try to follow your comments in order to make sure I can do the calculation in a right way. >Do a calculation with a given set of k points and calculate el-ph as a >function of sigma In practise, I choose a 8x8x8 k mesh and a q point (say 0.5 0.5 0.0). I set nsig =4, 8, 12, 16, 20, 24...respectively. I calculate the lambdas at this q point with this k mesh and with different nsigs. Concerning the lambda value, For example, for nsig=4 and 8, the final lambda values are choosen as the final step gaussian broadening (step 4 for nsig=4, step 8 for nsig=8). Then I can get a curve for nsig.vs.lambda at this q point. >Do again the same calculation increasing the number of k-points and the >same set of values for sigma. Repeat the calculation with a larger k mesh, for example, 12x12x12. >You will notice that for sufficiently large values of sigma the two >curves will agrees. This means that >the k-points mesh is sufficiently accurate for sigmas larger than that >but may not be good for smaller values. I compare the two curves and get a reliable nsig value, which the two curves start to agree. >Repeating the calculation for even larger set of k-points (same set of >sigma) you will see that from a value >of sigma that is smaller than the previous one the most accurate >calculation (the one using more k-points) >will agree with the second-best (and with the first, less accurate, >calculation when these last two agree) if I repeat the calculation with more denser k mesh (say 16x16x16), I will get a smaller reliable nsig by comparing the two more denser k mesh. >Trust only the curve in the region in which the calculation is converged >with respect to k-ponts summation. In practise, how can I decide the calculation is converged with respect to k points summation? At a q point, with the same nsig, The calculated lambdas are same within some errors for two K points mesh. Does this mean convergence? >For sufficiently large set of k-points the SAFE region will approach >enough the limit of sigma going to zero >that you can stop the procedure. > I can not understand this paragraph. Sigma going to zero...? Any comments? Thanks in advance. Yanming Ma _________________________________________________________________ Instant message during games with MSN Messenger 6.0. Download it now FREE! http://msnmessenger-download.com From degironc at sissa.it Tue Sep 23 11:41:14 2003 From: degironc at sissa.it (Stefano de Gironcoli) Date: Tue, 23 Sep 2003 11:41:14 +0200 Subject: [Pw_forum] Question for electron phonon coulping References: Message-ID: <3F70153A.6030305@sissa.it> Yanming ma wrote: >> Do a calculation with a given set of k points and calculate el-ph as a >> function of sigma > > > In practise, > I choose a 8x8x8 k mesh and a q point (say 0.5 0.5 0.0). I set nsig > =4, 8, 12, 16, 20, 24...respectively. I calculate the lambdas at this > q point with this k mesh and with different nsigs. Concerning the > lambda value, For example, for nsig=4 and 8, the final lambda values > are choosen as the final step gaussian broadening (step 4 for nsig=4, > step 8 for nsig=8). > Then I can get a curve for nsig.vs.lambda at this q point. You do not need to do a separate calculation for each value of nsig. with the original value of nsig=10 and the step size of 0.01 (ryd) in a sigle run you get results for degauss1=0.01, 0.02, 0.03,,..., 0.10 (ryd). The result of step 4 for nsig=4 is identical to step 4 for nsig=8,12,16,20,24 or any N>4 ! If you increase nsig to 24 (leaving the step size equal to 0.01) you can reach degauss1=0.24 if you like but, remember, one is interested in the limit of degauss1 going to zero. NOTICE that I call degauss1 the value of the smearing used in the calculation of el-ph interaction which will differ in general from degauss used in the scf/phonon calculation. > > Repeat the calculation with a larger k mesh, for example, 12x12x12. > > I compare the two curves and get a reliable nsig value, which the two > curves start to agree. > > if I repeat the calculation with more denser k mesh (say 16x16x16), I > will get a smaller reliable nsig by comparing the two more denser k mesh. > >> Trust only the curve in the region in which the calculation is converged >> with respect to k-ponts summation. > > In practise, how can I decide the calculation is converged with > respect to k points summation? > At a q point, with the same nsig, The calculated lambdas are same > within some errors for two K points mesh. Does this mean convergence? When for a given value of degauss1 (a given step in the nsig loop) the calculated lambda does not change anymore with increasing density of the K-points mesh then you have reached convergence for that value of degauss1. > >> For sufficiently large set of k-points the SAFE region will approach >> enough the limit of sigma going to zero >> that you can stop the procedure. >> > > I can not understand this paragraph. Sigma going to zero...? > SAFE meant converged with respect to k-point mesh. The idea is that for sufficiently large set of K-points even the smallest values of degauss1are converged and one can use these values as approximations of the degauss1 going to zero limit. How close one needs to go to the limit depends on the shape of the lambda vs degauss1curve. Let's say 0.01-0.03 ryd or larger if the curve is sufficiently flat (horizontal). As an example of this idea in a different context I include a figure where convergence with respect to k-points and degauss is computed for the total energy of lead. In this figure one sees that for ngauss=1 the "converged" courve is sufficiently flat that you can usea value of degauss as large as 0.05 ryd being within a franction of mryd from the exact limit (degauss=0 and infinite number of k-points). ngauss=0 converge to the same limit but with a stronger dependence on degauss. This is the reason why in the calculation of lambda ngauss1 is always chosen equal to 1, although for el-ph interaction the dependence on degauss1 might be less nice. So, do the calculation for several set of k points and a given value of nsig (24 if you like). This will give you an entire curve lambda(degauss1) for each set of k-points. Increse the k-point grid until the region of converged degauss1 extent sufficiently close to zero that you feel safe in extrapolating that value to zero. How much you need to go will depend on the shape of the function: whether it is very flat as the ngauss=1 case in the figure or with a large and changing slope as the negauss=0 case. Once you are satisfied with the convergence of lambda at a given value of phonon wavevector q, then you still need to check convergence with respect to number of q-points (phonon wavevectors) you include in the calculation. So you should compute el-ph interaction for a larger set of q-points and REPEAT the above procedure because the convergence properties will in general depend on the q-points.... ... and so on an so forth ... Stefano de Gironcoli -------------- next part -------------- A non-text attachment was scrubbed... Name: EN_VS_SIGMA.ps Type: application/postscript Size: 7073 bytes Desc: not available Url : /pipermail/attachments/20030923/68af8f37/attachment.ps From yanming_ma at hotmail.com Tue Sep 23 16:41:55 2003 From: yanming_ma at hotmail.com (ma Yanming) Date: Tue, 23 Sep 2003 22:41:55 +0800 Subject: [Pw_forum] Question for electron phonon coulping Message-ID: Dear Stefano de Gironcoli, Actually, I have been struggling the electron-phonon coupling for a long time, say half a year. Up to now, I would like to say I can do things correctly by your excellent and detailed explaination on this stuff. Thanks very very ... much. One final thing I may ask you again is the q points choosing. I would like to choose Monkhorst-pack q points. If I use 1616x16 Monkhorst-pack q points, there are,say, 500 q points can be choosen. I only can choose more or less 50 of them randomly by considering the computer soucre . But for a anisotropic system, it may have a risk. If I use 4x4x4 Monkhorst-pack q points, there are only, say, 50 q points in the whole first BZ. I can choose all of them to compute. This procedure may cover all of the BZ. Then I check the convergence. Is this procedure reasonable? Yanming Ma _________________________________________________________________ ???? MSN Explorer: http://explorer.msn.com/lccn From mutombo at fzu.cz Wed Sep 24 17:51:12 2003 From: mutombo at fzu.cz (mutombo) Date: Wed, 24 Sep 2003 17:51:12 +0200 Subject: [Pw_forum] compilation error References: <3F70153A.6030305@sissa.it> Message-ID: <3F71BD6F.E32AF67E@fzu.cz> Dear List, I am getting the following error when trying to compile the "pw" module of the newest version of the PWSCF code. My system is alpha linux (compiler :fort). Does anyone have an idea on how to solve this? Best wishes, Pingo ../Modules/berry_phase.o: In function `$berry_phase$ln_setup': :40: undefined reference to `ln_alloc__' :40: undefined reference to `ln_alloc__' :40: undefined reference to `ln_alloc__' :42: undefined reference to `ln_set__' :42: undefined reference to `ln_set__' :42: undefined reference to `ln_set__' :44: undefined reference to `ln_activate__' :44: undefined reference to `ln_activate__' :44: undefined reference to `ln_activate__' ../Modules/berry_phase.o: In function `$berry_phase$ln_closeup': :51: undefined reference to `ln_dealloc__' :51: undefined reference to `ln_dealloc__' :51: undefined reference to `ln_dealloc__' ../Modules/berry_phase.o: In function `$berry_phase$indi_of_ig': :198: undefined reference to `ln_ind__' :198: undefined reference to `ln_ind__' :198: undefined reference to `ln_ind__' :201: undefined reference to `ln_ind__' :201: undefined reference to `ln_ind__' ../Modules/berry_phase.o::201: more undefined references to `ln_ind__' follow collect2: ld returned 1 exit status fort: Severe: Failed while trying to link. make[1]: *** [memory] Error 1 make[1]: Leaving directory `/home/crystal/newpwscf/PW' make: *** [pw] Error 2ear List, From roma at cea.fr Wed Sep 24 18:13:25 2003 From: roma at cea.fr (Guido Roma) Date: Wed, 24 Sep 2003 18:13:25 +0200 Subject: [Pw_forum] compilation error References: <3F70153A.6030305@sissa.it> <3F71BD6F.E32AF67E@fzu.cz> Message-ID: <3F71C2A5.18E6CDC9@cea.fr> mutombo wrote: > > Dear List, > > I am getting the following error when trying to compile the "pw" module of > the newest version of the PWSCF code. My system is alpha linux (compiler > :fort). > Does anyone have an idea on how to solve this? > Best wishes, > Pingo > > .../Modules/berry_phase.o: In function `$berry_phase$ln_setup': > :40: undefined reference to `ln_alloc__' > :40: undefined reference to `ln_alloc__' > :40: undefined reference to `ln_alloc__' Hello, do you link with $(OSHOME)/clib/clib.a ? If not then add it to the LDFLAGS in install/Make.alphalinux (or the way you call it). If it still doesn't solve the problem, it has perhaps something to do with underscores, or capital letters. Guido -- Guido Roma -- CEA-Saclay - DEN/DMN/SRMP Bat.520/130 Phone: [+33]-1-69085738 -- Fax: ...6867 -- Mobile: [+33]-6-20069085 From giannozz at nest.sns.it Wed Sep 24 19:28:21 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 24 Sep 2003 19:28:21 +0200 Subject: [Pw_forum] compilation error In-Reply-To: <3F71BD6F.E32AF67E@fzu.cz> References: <3F70153A.6030305@sissa.it> <3F71BD6F.E32AF67E@fzu.cz> Message-ID: <200309241928.21576.giannozz@nest.sns.it> Ciao Carlo, a proposito dell'errore segnalato dal Pingo Mutombo: noto che Alpha Linux fa una cosa differente in cp.h rispetto ad Alpha e basta. Come mai ? Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From giannozz at nest.sns.it Wed Sep 24 19:32:15 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Wed, 24 Sep 2003 19:32:15 +0200 Subject: [Pw_forum] compilation error In-Reply-To: <200309241928.21576.giannozz@nest.sns.it> References: <3F71BD6F.E32AF67E@fzu.cz> <200309241928.21576.giannozz@nest.sns.it> Message-ID: <200309241932.16000.giannozz@nest.sns.it> Oops...sorry for my previous message, it wasn't intended for the mailing list Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From proffess at yandex.ru Thu Sep 25 10:26:52 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Thu, 25 Sep 2003 12:26:52 +0400 (MSD) Subject: [Pw_forum] compilation error In-Reply-To: <200309241928.21576.giannozz@nest.sns.it> References: <3F70153A.6030305@sissa.it> <3F71BD6F.E32AF67E@fzu.cz> <200309241928.21576.giannozz@nest.sns.it> Message-ID: <3F72A6CC.000001.22496@tide.yandex.ru> Dear Mutombo, I had this problem some times ago. My system also Compaq Alpha under Linux. You shoul add in CPPFLAGS the "-D__LINUX". I have some questions to ask: what fftw library are you using - cxml or fftw? I can compile new version only in serial mode with cxml library. The parallel version could not compile, because i got in files cft1s.o and cft3s.o references:cft_1__ and cft_2__. Good luck, Sergey From g930102 at ccit.edu.tw Thu Sep 25 11:58:00 2003 From: g930102 at ccit.edu.tw (Lu Fu-Fa) Date: Thu, 25 Sep 2003 17:58:00 +0800 Subject: [Pw_forum] Re: Ga 3d core electrons as valence electrons (fhi2upf) Message-ID: <003e01c3834b$84b4e7a0$1670848c@InAlN> Dear PWscf users, I have solved the problem about PP of Ga 3d core electrons as valence electrons generated by fhi2upf(fhi98PP) using PWscf. I should be noticed that the ouput of PW what they released at the beginning. The fhi2upf/PWscf only suit for lloc = 0, so I change my recipes of elements(Ga, N) in fhi98PP, just use the defaults. That's all. I have made the test of band structures for zb GaN and wz GaN. It is compatible with ABINIT, but differ in lloc setting. ======================================================= Lu Fu-Fa, Doctoral Student, E-mail: g930102 at ccit.edu.tw Semiconductors Lab., Dept. of Applied Physics Chung Cheng Institute of Technology (CCIT), Taiwan ======================================================= There are some information might be help for someone who needs. I wish it is no garbage, no hiding. Here is my process of fhi2upf. $ ../bin/fhi2upf input PP file in FHI format > 31-Ga-3d_tm.cpi Pseudopotential successfully read Atom name > Ga l local > 0 DFT > PW91 Wavefunction # 1: label, occupancy > 4S, 2 Wavefunction # 2: label, occupancy > 4P, 1 Wavefunction # 3: label, occupancy > 3D, 10 Pseudopotential successfully converted Output PP file in UPF format : 31-Ga-3d_tm.cpi.UPF $ ../bin/fhi2upf input PP file in FHI format > 07-N_tm.cpi Pseudopotential successfully read Atom name > N l local > 0 DFT > PW91 Wavefunction # 1: label, occupancy > 2S, 2 Wavefunction # 2: label, occupancy > 2P, 3 Wavefunction # 3: label, occupancy > 3D, 0 Pseudopotential successfully converted Output PP file in UPF format : 07-N_tm.cpi.UPF ======================================================= Here are Ga 3d and N elements default input files(fhi98PP) from FHI. 31-Ga-3d_tm.ini 31.00 5 3 8 0.00 : z nc nv iexc rnlc 1 0 2.00 : n l f 2 0 2.00 2 1 6.00 3 0 2.00 3 1 6.00 3 2 10.00 4 0 2.00 4 1 1.00 2 t : lmax s_pp_def ======================================================= 07-N_tm.ini 7.00 1 2 8 0.00 : z nc nv iexc rnlc 1 0 2.00 : n l f 2 0 2.00 2 1 3.00 2 t : lmax s_pp_def ======================================================= From mutombo at fzu.cz Thu Sep 25 12:17:04 2003 From: mutombo at fzu.cz (mutombo) Date: Thu, 25 Sep 2003 12:17:04 +0200 Subject: [Pw_forum] compilation error References: <3F70153A.6030305@sissa.it> <3F71BD6F.E32AF67E@fzu.cz> <200309241928.21576.giannozz@nest.sns.it> <3F72A6CC.000001.22496@tide.yandex.ru> Message-ID: <3F72C0A0.DA72F59D@fzu.cz> Dear Guido, Paolo, Sergey. I am still getting the same error even after editing the" make.sys" according to your suggestions. I do use both the CXML and fftw(version 2.1.3) libraries. Best wishes, Pingo Sergei Lisenkov wrote: > Dear Mutombo, > > I had this problem some times ago. My system also Compaq Alpha under Linux. You shoul add in CPPFLAGS the "-D__LINUX". > > I have some questions to ask: what fftw library are you using - cxml or fftw? I can compile new version only in serial mode with cxml library. The parallel version could not compile, because i got in files cft1s.o and cft3s.o references:cft_1__ and cft_2__. > > Good luck, > Sergey > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From tvy at otf.fti.udmurtia.su Thu Sep 25 13:45:07 2003 From: tvy at otf.fti.udmurtia.su (Victor Trubitsin) Date: Thu, 25 Sep 2003 16:45:07 +0500 Subject: [Pw_forum] Re: [Pw_fforum] make PW error In-Reply-To: <3F71C2A5.18E6CDC9@cea.fr> References: <3F71BD6F.E32AF67E@fzu.cz> <3F71C2A5.18E6CDC9@cea.fr> Message-ID: Dear PWscf users, I am getting the following error when trying to make the "pw" module of the newest version of the PWSCF code. My system is pc_ifc ( Linux, Intel, mkl) ############################# $ make pw ( cd Modules; make all ) ....... ...... make[1]: enter in the directory `/home/tvy/PW1.3/instsours/PW' ifc -o memory.x memory.o pwcom.o para.o becmod.o clean_pw.o ..... ..... vpack.o w0gauss.o w1gauss.o wfcinit.o wgauss.o which_dft.o write_config_to_file.o write_ns.o wsweight.o ylmr2.o ../Modules/*.o /home/tvy/PW1.3/instsours/flib/ptools.a /home/tvy/PW1.3/instsours/flib/flib.a /home/tvy/PW1.3/instsours/clib/clib.a -Vaxlib /opt/intel/mkl/lib/32/libmkl_lapack.a /opt/intel/mkl/lib/32/libmkl_p3.a -lguide -lpthread ld: Warning: size of symbol `MODULE.reciprocal_vectors_1' changed from 212 to 556 in ../Modules/recvec.o make[1]: *** [memory] Error 1 make[1]: Exit from directory `/home/tvy/PW1.3/instsours/PW' make: *** [pw] Error 2 ############################## Does anyone have an idea on how to solve this? Best wishes, V. Trubitsin From proffess at yandex.ru Thu Sep 25 14:28:51 2003 From: proffess at yandex.ru (Sergei Lisenkov) Date: Thu, 25 Sep 2003 16:28:51 +0400 (MSD) Subject: [Pw_forum] compilation error In-Reply-To: <3F72C0A0.DA72F59D@fzu.cz> References: <3F70153A.6030305@sissa.it> <3F71BD6F.E32AF67E@fzu.cz> <200309241928.21576.giannozz@nest.sns.it> <3F72A6CC.000001.22496@tide.yandex.ru> <3F72C0A0.DA72F59D@fzu.cz> Message-ID: <3F72DF83.00000D.10623@soapbox.yandex.ru> Dear Pingo, Could you send me your make.sys file? What version of "fort" compiler are you using? Best wishes, Sergey From giannozz at nest.sns.it Thu Sep 25 19:03:19 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 25 Sep 2003 19:03:19 +0200 Subject: [Pw_forum] compilation error In-Reply-To: <3F72C0A0.DA72F59D@fzu.cz> References: <3F72A6CC.000001.22496@tide.yandex.ru> <3F72C0A0.DA72F59D@fzu.cz> Message-ID: <200309251903.19620.giannozz@nest.sns.it> On Thursday 25 September 2003 12:17, mutombo wrote: > Dear Guido, Paolo, Sergey. > > I am still getting the same error even after editing the" make.sys" > according to your suggestions. since I didn't suggest anything, I am doing it now: did you remove the C library (clib/clib.a) and all objects (clib/*.o) after changing the make.sys file and before recompiling ? -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From giannozz at nest.sns.it Thu Sep 25 19:05:51 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 25 Sep 2003 19:05:51 +0200 Subject: [Pw_forum] Re: [Pw_fforum] make PW error In-Reply-To: References: <3F71C2A5.18E6CDC9@cea.fr> Message-ID: <200309251905.51100.giannozz@nest.sns.it> On Thursday 25 September 2003 13:45, Victor Trubitsin wrote: > [...] ld: Warning: size of symbol `MODULE.reciprocal_vectors_1' > changed from 212 to 556 in ../Modules/recvec.o this means that you have two conflicting versions of module "reciprocal_vectors" lingering (why, I have no idea): remove all *.o, *.d, *.mod, retry -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From giannozz at nest.sns.it Thu Sep 25 19:07:09 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 25 Sep 2003 19:07:09 +0200 Subject: [Pw_forum] Re: Ga 3d core electrons as valence electrons (fhi2upf) In-Reply-To: <003e01c3834b$84b4e7a0$1670848c@InAlN> References: <003e01c3834b$84b4e7a0$1670848c@InAlN> Message-ID: <200309251907.09170.giannozz@nest.sns.it> On Thursday 25 September 2003 11:58, Lu Fu-Fa wrote: > I should be noticed that the ouput of PW what they released at the > beginning. The fhi2upf/PWscf only suit for lloc = 0, so I change my > recipes of elements(Ga, N) in fhi98PP, just use the defaults. That's all. sorry, I don't understand: is the converter working properly? Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From giannozz at nest.sns.it Thu Sep 25 19:08:58 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Thu, 25 Sep 2003 19:08:58 +0200 Subject: [Pw_forum] compilation error In-Reply-To: <3F72A6CC.000001.22496@tide.yandex.ru> References: <200309241928.21576.giannozz@nest.sns.it> <3F72A6CC.000001.22496@tide.yandex.ru> Message-ID: <200309251908.58931.giannozz@nest.sns.it> On Thursday 25 September 2003 10:26, Sergei Lisenkov wrote: > I have some questions to ask: what fftw library are you using - cxml or > fftw? I can compile new version only in serial mode with cxml library. > The parallel version could not compile, because i got in files cft1s.o and > cft3s.o references:cft_1__ and cft_2__. there is no driver for FFT in cxml library in the parallel case, so you definitely cannot compile the parallel version with cxml library Paolo -- Paolo Giannozzi e-mail: giannozz at nest.sns.it Scuola Normale Superiore Phone: +39/050509412 Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 From lucre at ms1.hinet.net Fri Sep 26 00:44:31 2003 From: lucre at ms1.hinet.net (Lu Frederic) Date: Fri, 26 Sep 2003 06:44:31 +0800 Subject: [Pw_forum] Re: Ga 3d core electrons as valence electrons (fhi2upf) Message-ID: <000601c383b6$9b2e5fa0$ef3310ac@phys> Dear Dr. Paolo and PWscf users, >> I should be noticed that the ouput of PW what they released at the >> beginning. The fhi2upf/PWscf only suit for lloc = 0, so I change my >> recipes of elements(Ga, N) in fhi98PP, just use the defaults. That's all. >sorry, I don't understand: is the converter working properly? Just as I have wrote in previous E-mail, I have solved that problem. The converter(fhi2upf) is working properly, if someone follows my procedure. I do never mean the converter(fhi2upf) is not working properly. I don't know how to use the converter(fhi2upf) at the beginning, there is no any document, so I ask for help. But no one provides his process in detail. I think no one has experiences to treat Ga 3d as valence electrons with fhi2upf at this moment(I am really surprised). So I share my process, how to use and setting in fhi2upf and fhi98PP, and the key just lloc=0 simply. Best wishes, Lu Fu-Fa, 09/26/2003 From mutombo at fzu.cz Fri Sep 26 15:14:44 2003 From: mutombo at fzu.cz (mutombo) Date: Fri, 26 Sep 2003 15:14:44 +0200 Subject: [Pw_forum] compilation error References: <3F72A6CC.000001.22496@tide.yandex.ru> <3F72C0A0.DA72F59D@fzu.cz> <200309251903.19620.giannozz@nest.sns.it> Message-ID: <3F743BC4.A99FAFAC@fzu.cz> Dear Paolo, I could not respond in time because I was not at work. Unfortunatenaly I am still getting the same error after removing the C library (clib/clib.a) and all objects (clib/*.o) before recompiling. Best wishes, Pingo > > On Thursday 25 September 2003 12:17, mutombo wrote: > > Dear Guido, Paolo, Sergey. > > > > I am still getting the same error even after editing the" make.sys" > > according to your suggestions. > > since I didn't suggest anything, I am doing it now: did you remove > the C library (clib/clib.a) and all objects (clib/*.o) after changing the > make.sys file and before recompiling ? > > -- > Paolo Giannozzi e-mail: giannozz at nest.sns.it > Scuola Normale Superiore Phone: +39/050509412 > Piazza dei Cavalieri 7 Fax: +39/050509417, 050563513 > I-56126 Pisa, Italy Office: Lab. NEST, Via della Faggiola 19 > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From tvynew at otf.pti.udm.ru Fri Sep 26 15:18:49 2003 From: tvynew at otf.pti.udm.ru (Victor Trubitsin) Date: Fri, 26 Sep 2003 18:18:49 +0500 Subject: [Pw_forum] Re: [Pw_fforum] make PW error In-Reply-To: <200309251905.51100.giannozz@nest.sns.it> References: <200309251905.51100.giannozz@nest.sns.it> Message-ID: 25 ???????? 2003 22:05, ?? ????????: > On Thursday 25 September 2003 13:45, Victor Trubitsin wrote: > > [...] ld: Warning: size of symbol `MODULE.reciprocal_vectors_1' > > changed from 212 to 556 in ../Modules/recvec.o > > this means that you have two conflicting versions of module > "reciprocal_vectors" lingering (why, I have no idea): > remove all *.o, *.d, *.mod, retry Hm..., I have not two versions of this module! I removed al l : *o,.d.*.mod and recompiled and again got the same error. V.Trubitsin From mutombo at fzu.cz Fri Sep 26 16:37:55 2003 From: mutombo at fzu.cz (mutombo) Date: Fri, 26 Sep 2003 16:37:55 +0200 Subject: [Pw_forum] compilation error References: <200309241928.21576.giannozz@nest.sns.it> <3F72A6CC.000001.22496@tide.yandex.ru> <200309251908.58931.giannozz@nest.sns.it> Message-ID: <3F744F43.B8DD8C88@fzu.cz> Hello, I just wanted to say that I succeeded to compile the PW module. What I needed was to use -DDXML in CPPFLAGS but -CXML in LIBS. Many thanks for all your advice, Pingo From giannozz at nest.sns.it Sat Sep 27 22:59:08 2003 From: giannozz at nest.sns.it (Paolo Giannozzi) Date: Sat, 27 Sep 2003 22:59:08 +0200 Subject: [Pw_forum] make PW error In-Reply-To: References: <200309251905.51100.giannozz@nest.sns.it> Message-ID: <200309272259.08184.giannozz@nest.sns.it> On Friday 26 September 2003 15:18, Victor Trubitsin wrote: > Hm..., I have not two versions of this module! I removed all : > *o,.d.*.mod and recompiled and again got the same error. does this happen with version 7 of intel compiler (ifc)? try the very latest release, or an earlier version Paolo From shein at ihim.uran.ru Mon Sep 29 03:49:11 2003 From: shein at ihim.uran.ru (Igor R. Shein) Date: Mon, 29 Sep 2003 07:49:11 +0600 Subject: [Pw_forum] About PWscf 1.3 version Message-ID: <003901c3862b$e17542e0$e58813c3@boa> Dear colleagues, I am very glad, that you managed to correct errors in makefile which were in version 1.2. I compiled your programs in ifc 7.1 c libraries MKL 6.0 and have corrected the some syntactic fortran errors. Almost all tests from your list at me have successfully come to the end. I have remained is very pleased with your program. But when I have made on the basis of these tests my test program for calculation of band structure and DOS's for MgB2 (hexagonal structure), at me had problems. I send these files as an investment in this letter. As I send also pseudo-potentials. I have questions: 1. Why is not present p-and d - state at Mg? 2. How correctly to construct band structure for hexagonal lattice? Very much I hope, that you will find a little time to answer my questions. Yours faithfully, Igor R. Shein Institute of Solid State Chemistry Russian Academy of Sciences-Ural Division 620219 Yekaterinburg GSP-145 Russia shein at ihim.uran.ru -------------- next part -------------- An HTML attachment was scrubbed... URL: /pipermail/attachments/20030929/47dcfb91/attachment.htm -------------- next part -------------- A non-text attachment was scrubbed... Name: results.tar.bz2 Type: application/octet-stream Size: 238320 bytes Desc: not available Url : /pipermail/attachments/20030929/47dcfb91/attachment.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: run1.dat Type: application/octet-stream Size: 7066 bytes Desc: not available Url : /pipermail/attachments/20030929/47dcfb91/attachment-0001.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: Mg.pz-n-vbc.UPF Type: application/octet-stream Size: 30209 bytes Desc: not available Url : /pipermail/attachments/20030929/47dcfb91/attachment-0002.obj -------------- next part -------------- A non-text attachment was scrubbed... Name: B.pz-vbc.UPF Type: application/octet-stream Size: 22160 bytes Desc: not available Url : /pipermail/attachments/20030929/47dcfb91/attachment-0003.obj From hslee at phya.snu.ac.kr Mon Sep 29 13:21:13 2003 From: hslee at phya.snu.ac.kr (Lee Ho-Sik) Date: Mon, 29 Sep 2003 20:21:13 +0900 (KST) Subject: [Pw_forum] Projwfc.x inputfile Message-ID: Dear pwscf users, I am trying to use projwfc.x with pwscf 1.3.0. However, I got the following messages. Projwave.f90 say that inputpp files require only prefix and outdir. How can I get the proper results? %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% from read_file : error # 29 problem reading file .pwscf.pun %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% stopping ... Hosik From hslee at phya.snu.ac.kr Mon Sep 29 13:52:52 2003 From: hslee at phya.snu.ac.kr (Lee Ho-Sik) Date: Mon, 29 Sep 2003 20:52:52 +0900 (KST) Subject: [Pw_forum] Projwfc.x inputfile In-Reply-To: Message-ID: Dear pwscf users, It was a mistake. My ourdir was '.'. The correct one is './'. I am sorry for bothering you :) Hosik On Mon, 29 Sep 2003, Lee Ho-Sik wrote: > Dear pwscf users, > > I am trying to use projwfc.x with pwscf 1.3.0. However, I got the > following messages. Projwave.f90 say that inputpp files require only > prefix and outdir. How can I get the proper results? > > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > from read_file : error # 29 > problem reading file .pwscf.pun > > %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% > stopping ... > > > Hosik > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum > From nagano at atrc.mhi.co.jp Tue Sep 30 05:23:40 2003 From: nagano at atrc.mhi.co.jp (Ichiro Nagano) Date: Tue, 30 Sep 2003 12:23:40 +0900 Subject: [Pw_forum] dielectric functions in the region of GHz to THz Message-ID: <1c5301c38702$47e9e0e0$0a130190@kibanad.artc.mhi.co.jp> Dear PWscf users, It may be a trivial thing, I would try to calculate dielectric functions in the region of GHz to THz. Can we perform the calculation with PWscf? Thanks for any information, Ichiro Nagano ******************************************************************** Advanced Technology Research Center, Mitsubishi Heavy Industries, Ltd. 8-1, Sachiura, 1-Chome, Kanazawa-ku, Yokohama, 236-8515, Japan Ichiro Nagano E-mail: TEL: + 81 (45) 775-2437 (EXT 9065) FAX: + 81 (45) 771-3879 From baroni at sissa.it Tue Sep 30 08:54:21 2003 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 30 Sep 2003 08:54:21 +0200 Subject: [Pw_forum] dielectric functions in the region of GHz to THz In-Reply-To: <1c5301c38702$47e9e0e0$0a130190@kibanad.artc.mhi.co.jp> References: <1c5301c38702$47e9e0e0$0a130190@kibanad.artc.mhi.co.jp> Message-ID: <3F79289D.4050704@sissa.it> Well, it depends on the material. If the relevant elementary excitations which determine the dielectric constant in that spectral region, then PWscf+PHONON will certainly do. You'll have to calculate zone-center phonons + effective charges + static electronic dielectric constant, and look into a standard textbook how to obtain the diectric constant from the data so produced (it is very easy). For the time being, PWscf is unable to deal with any electronic contribution to the dynamic response. (it can calculate static response, though). Hope this helps. SB Ichiro Nagano wrote: > Dear PWscf users, > > It may be a trivial thing, > I would try to calculate dielectric functions in the region of GHz to THz. > Can we perform the calculation with PWscf? > > Thanks for any information, > > Ichiro Nagano > > ******************************************************************** > Advanced Technology Research Center, Mitsubishi Heavy Industries, Ltd. > 8-1, Sachiura, 1-Chome, Kanazawa-ku, Yokohama, 236-8515, Japan > Ichiro Nagano > E-mail: > TEL: + 81 (45) 775-2437 (EXT 9065) > FAX: + 81 (45) 771-3879 > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From baroni at sissa.it Tue Sep 30 09:02:30 2003 From: baroni at sissa.it (Stefano Baroni) Date: Tue, 30 Sep 2003 09:02:30 +0200 Subject: [Pw_forum] dielectric functions in the region of GHz to THz In-Reply-To: <3F79289D.4050704@sissa.it> References: <1c5301c38702$47e9e0e0$0a130190@kibanad.artc.mhi.co.jp> <3F79289D.4050704@sissa.it> Message-ID: <3F792A86.6030004@sissa.it> Sorry for the wrong (incomplete) post, what I actually wanted to say is: Stefano Baroni wrote: > Well, it depends on the material. If the relevant elementary excitations > which determine the dielectric constant in that spectral region, >>> ARE PHONONS, > then > PWscf+PHONON will certainly do. You'll have to calculate zone-center > phonons + effective charges + static electronic dielectric constant, and > look into a standard textbook how to obtain the diectric >>> FUNCTION > from > the data so produced (it is very easy). For the time being, PWscf is > unable to deal with any electronic contribution to the dynamic response. > (it can calculate static response, though). Hope this helps. SB > > Ichiro Nagano wrote: > >>Dear PWscf users, >> >>It may be a trivial thing, >>I would try to calculate dielectric functions in the region of GHz to THz. >>Can we perform the calculation with PWscf? >> >>Thanks for any information, >> >>Ichiro Nagano >> >>******************************************************************** >>Advanced Technology Research Center, Mitsubishi Heavy Industries, Ltd. >>8-1, Sachiura, 1-Chome, Kanazawa-ku, Yokohama, 236-8515, Japan >>Ichiro Nagano >>E-mail: >>TEL: + 81 (45) 775-2437 (EXT 9065) >>FAX: + 81 (45) 771-3879 >>_______________________________________________ >>Pw_forum mailing list >>Pw_forum at pwscf.org >>http://www.democritos.it/mailman/listinfo/pw_forum > > > _______________________________________________ > Pw_forum mailing list > Pw_forum at pwscf.org > http://www.democritos.it/mailman/listinfo/pw_forum From breezejd at lsbu.ac.uk Tue Sep 30 18:27:58 2003 From: breezejd at lsbu.ac.uk (Jonathan Breeze) Date: Tue, 30 Sep 2003 17:27:58 +0100 Subject: [Pw_forum] J.Breeze on:Third order anharmonic dynamical matrix References: <3F6B286A.F3D02836@lmcp.jussieu.fr> Message-ID: <3F79AF0E.9050102@lsbu.ac.uk> Dear Michele, Thank you for your reply. It was very useful. I had used the super-cell approach for the frozen phonon method for comparison with the D3 calculation. So, hopefully it should be relatively easy to apply your suggestion of zone folding. Out of interest, I recently read a paper by G. Deinzer, G. Birner and D. Strauch (Phys. Rev. B 67, 144304, 2003) which reports on calculations of D^3(-q1-q2,q1,q2). The form of the calculation was similar to that reported by Debarnardi and Baroni. (Solid State Comm. 91, 813, 1994), using the 2n+1 theorem. The main differences are having to calculate the electronic term for pairs of q values, a more involved Fourier transformation of the anharmonic coefficients and a slightly more complex Ewald transformation for the ionic part. Anyway, thanks again for your advice. Jonathan Breeze Michele Lazzeri wrote: >>Using the PWSCF D3 code, should it be possible to calculate the >>coefficients for phonons at arbitraty wave-vector? >>That is, to calculate D^3(-q1-q2,q1,q2). > >>Jonathan Breeze > > Dear Jonathan, > As you know the code calculates D^3(0,-q, q). > More in general, to compute D^3(-q1-q2,q1,q2) is possible, but the > modification to be done to the code could be tedious. > > In alternative, you could use a super-cell approach: > If you want to compute D^3(Q,Q-q, Q+q), you can compute > D^3(0,-q,q) in a super-cell such that the gamma-point of the > super-cell is equivalent to the Q point of the unit-cell. > An example: if you want to compute the broadening of a phonon > at the border zone, you can use a crystal obtained doubling the unit-cell > in the appropriate direction. > The phonon you are interested in is a phonon at Gamma for this > double-cell. > So, if you compute the D3 matrix at gamma, for this double-cell, you > obtain > all the information that you need. > > For the moment, this is the most practical thing you can do if you are > interested in high-symmetry points. > This procedure cannot be applied to any point Q of the Brillouin Zone, > and, depending on the point, can be computationally expensive. > > Michele Lazzeri > > > > -- Jonathan Breeze Research Fellow Centre for Physical Electronics and Materials London South Bank University 103 Borough Road London SE1 0AA Tel: +44(0)20 7815 7582 Fax: +44(0)20 7815 7599