November 2014 - mpiwg-fortran - mailman3.cels.anl.gov

Re: [MPIWG Fortran] Typos in MPI_WIN_*_FN
by Rolf Rabenseifner 18 Nov '14

18 Nov '14

Dear Junchao Zhang, thank you for your feedback. The Problem was already detected and published in the errata from May 14, 2014, on page 21, lines 19+28: http://www.mpi-forum.org/docs/mpi-3.0/errata-30.pdf The solution is different to your proposal. It is TYPE(MPI_Win) :: oldwin because all callbacks have not any INTENT information. Best regards Rolf ----- Original Message ----- > From: "Junchao Zhang" <jczhang(a)mcs.anl.gov> > To: "MPI-WG Fortran working group" <mpiwg-fortran(a)lists.mpi-forum.org> > Sent: Monday, November 17, 2014 4:09:22 PM > Subject: [MPIWG Fortran] Typos in MPI_WIN_*_FN > > On p.730 of MPI-3.0, line 17 & 25, oldwin should be declared as > > TYPE(MPI_Win), INTENT(IN) :: oldwin > > --Junchao Zhang > > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran -- Dr. Rolf Rabenseifner . . . . . . . . . .. email rabenseifner(a)hlrs.de High Performance Computing Center (HLRS) . phone ++49(0)711/685-65530 University of Stuttgart . . . . . . . . .. fax ++49(0)711 / 685-65832 Head of Dpmt Parallel Computing . . . www.hlrs.de/people/rabenseifner Nobelstr. 19, D-70550 Stuttgart, Germany . . . . (Office: Room 1.307)

1 0

Typos in MPI_WIN_*_FN
by Junchao Zhang 17 Nov '14

17 Nov '14

On p.730 of MPI-3.0, line 17 & 25, oldwin should be declared as TYPE(MPI_Win), INTENT(IN) :: oldwin --Junchao Zhang

1 0

Re: [MPIWG Fortran] Webex meeting
by Jeff Squyres (jsquyres) 11 Nov '14

11 Nov '14

Sigh. Sorry. Friday, December 5, 2014 11am US Eastern time On Nov 11, 2014, at 5:31 PM, Bill Long <longb(a)cray.com> wrote: > Did you mean Dec 5? Dec 4 (Thursday) was not an option in the doodle. Only Monday, Wednesday, and Friday showed up. > > Cheers, > Bill > > On Nov 11, 2014, at 4:17 PM, Jeff Squyres (jsquyres) <jsquyres(a)cisco.com> wrote: > >> Thanks for answering the Doodle. We're setup for 4 Dec, 11am US Eastern (which is 1 business day before the Forum :-) ). >> >> ------------------------------------------------------- >> To join the online meeting (Now from mobile devices!) >> ------------------------------------------------------- >> 1. Go to https://cisco.webex.com/ciscosales/j.php?MTID=m81dbf893282eecc809b5fe13aeb1… >> 2. Enter your name and email address. >> 3. Enter the meeting password: fortran >> 4. Click "Join Now". >> >> See you then. >> >> -- >> Jeff Squyres >> jsquyres(a)cisco.com >> For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/ >> >> _______________________________________________ >> mpiwg-fortran mailing list >> mpiwg-fortran(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran > > Bill Long longb(a)cray.com > Fortran Technical Suport & voice: 651-605-9024 > Bioinformatics Software Development fax: 651-605-9142 > Cray Inc./ Cray Plaza, Suite 210/ 380 Jackson St./ St. Paul, MN 55101 > > > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran -- Jeff Squyres jsquyres(a)cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

1 0

Re: [MPIWG Fortran] Webex meeting
by Bill Long 11 Nov '14

11 Nov '14

Did you mean Dec 5? Dec 4 (Thursday) was not an option in the doodle. Only Monday, Wednesday, and Friday showed up. Cheers, Bill On Nov 11, 2014, at 4:17 PM, Jeff Squyres (jsquyres) <jsquyres(a)cisco.com> wrote: > Thanks for answering the Doodle. We're setup for 4 Dec, 11am US Eastern (which is 1 business day before the Forum :-) ). > > ------------------------------------------------------- > To join the online meeting (Now from mobile devices!) > ------------------------------------------------------- > 1. Go to https://cisco.webex.com/ciscosales/j.php?MTID=m81dbf893282eecc809b5fe13aeb1… > 2. Enter your name and email address. > 3. Enter the meeting password: fortran > 4. Click "Join Now". > > See you then. > > -- > Jeff Squyres > jsquyres(a)cisco.com > For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/ > > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran Bill Long longb(a)cray.com Fortran Technical Suport & voice: 651-605-9024 Bioinformatics Software Development fax: 651-605-9142 Cray Inc./ Cray Plaza, Suite 210/ 380 Jackson St./ St. Paul, MN 55101

1 0

Webex meeting
by Jeff Squyres (jsquyres) 11 Nov '14

11 Nov '14

Thanks for answering the Doodle. We're setup for 4 Dec, 11am US Eastern (which is 1 business day before the Forum :-) ). ------------------------------------------------------- To join the online meeting (Now from mobile devices!) ------------------------------------------------------- 1. Go to https://cisco.webex.com/ciscosales/j.php?MTID=m81dbf893282eecc809b5fe13aeb1… 2. Enter your name and email address. 3. Enter the meeting password: fortran 4. Click "Join Now". See you then. -- Jeff Squyres jsquyres(a)cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

1 0

Re: [MPIWG Fortran] Dec mtg: MPI 4: deprecate SIZEOF and mpif.h
by Balaji, Pavan 11 Nov '14

11 Nov '14

> On Nov 11, 2014, at 10:36 AM, Bill Long <longb(a)cray.com> wrote: > > On Nov 11, 2014, at 8:49 AM, Balaji, Pavan <balaji(a)anl.gov> wrote: > >> >> With one minor addition — new functionality will not have deprecated equivalents (at least that’s the model we used with C++). So everything that gets added in MPI-4 will be “use mpi” or “use mpi_f08” only, if this passes. > > OK. That’s what I was hoping. Did you mean “use mpi” *and* “use mpi_f08” only? I would not expect those to be diverging, at least yet. Right. I meant the user will need to use “use mpi” or “use mpi_f08” for MPI-4 functionality. -- Pavan Balaji ✉️ http://www.mcs.anl.gov/~balaji

1 0

Re: [MPIWG Fortran] Dec mtg: MPI 4: deprecate SIZEOF and mpif.h
by Bill Long 11 Nov '14

11 Nov '14

On Nov 11, 2014, at 8:49 AM, Balaji, Pavan <balaji(a)anl.gov> wrote: > > With one minor addition — new functionality will not have deprecated equivalents (at least that’s the model we used with C++). So everything that gets added in MPI-4 will be “use mpi” or “use mpi_f08” only, if this passes. OK. That’s what I was hoping. Did you mean “use mpi” *and* “use mpi_f08” only? I would not expect those to be diverging, at least yet. Cheers, Bill > > — Pavan > >> On Nov 11, 2014, at 8:09 AM, William Gropp <wgropp(a)illinois.edu> wrote: >> >> Deprecate in MPI uses the same definition that the Fortran standards have used - discouraged from use and a candidate for deletion in a future standard. Thus, until formally deleted, it must be implemented and appear in mpif.h (as mpif.h is now defined). >> >> Bill >> >> On Nov 11, 2014, at 8:01 AM, Bill Long <longb(a)cray.com> wrote: >> >>> Seems reasonable to me. >>> >>> Are there plans for a Webex meeting for the Fortran working group before the December MPI meeting in San Jose? This is a “vacation challenged” time of the year - I’m scheduled to be in the office only 3 days between December 5 and 31. >>> >>> How does MPI define “deprecate”? For example, would any new feature/function be excluded from mpif.h? Personally, I think that would be a good idea - it simplifies maintenance of the distribution and emphasizes that mpif.h is for legacy use only. But I can imagine other opinions. >>> >>> Cheers, >>> Bill >>> >>> >>> >>> On Nov 11, 2014, at 6:22 AM, Jeff Squyres (jsquyres) <jsquyres(a)cisco.com> wrote: >>> >>>> I won't have time to do any prep in detail until after SC and the subsequent US thanksgiving holiday, but are these 2 topics that we want to discuss in December? >>>> >>>> Sent from my phone. No type good. >>>> _______________________________________________ >>>> mpiwg-fortran mailing list >>>> mpiwg-fortran(a)lists.mpi-forum.org >>>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran >>> >>> Bill Long longb(a)cray.com >>> Fortran Technical Suport & voice: 651-605-9024 >>> Bioinformatics Software Development fax: 651-605-9142 >>> Cray Inc./ Cray Plaza, Suite 210/ 380 Jackson St./ St. Paul, MN 55101 >>> >>> >>> _______________________________________________ >>> mpiwg-fortran mailing list >>> mpiwg-fortran(a)lists.mpi-forum.org >>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran >> >> _______________________________________________ >> mpiwg-fortran mailing list >> mpiwg-fortran(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran > > -- > Pavan Balaji ✉️ > http://www.mcs.anl.gov/~balaji > > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran Bill Long longb(a)cray.com Fortran Technical Suport & voice: 651-605-9024 Bioinformatics Software Development fax: 651-605-9142 Cray Inc./ Cray Plaza, Suite 210/ 380 Jackson St./ St. Paul, MN 55101

1 0

Re: [MPIWG Fortran] Dec mtg: MPI 4: deprecate SIZEOF and mpif.h
by Jeff Squyres (jsquyres) 11 Nov '14

11 Nov '14

Yes, we should probably have a webex to discuss first (this is not the firs time discussing these issues, but a cache-refresh would be a good idea). We span Europe-to-California, so I'm thinking we really only have one or two possible time slots a day. Here's a doodle for a few days in the week of Dec 1: http://doodle.com/965sgk95nrza8r8q On Nov 11, 2014, at 9:01 AM, Bill Long <longb(a)cray.com> wrote: > Seems reasonable to me. > > Are there plans for a Webex meeting for the Fortran working group before the December MPI meeting in San Jose? This is a “vacation challenged” time of the year - I’m scheduled to be in the office only 3 days between December 5 and 31. > > How does MPI define “deprecate”? For example, would any new feature/function be excluded from mpif.h? Personally, I think that would be a good idea - it simplifies maintenance of the distribution and emphasizes that mpif.h is for legacy use only. But I can imagine other opinions. > > Cheers, > Bill > > > > On Nov 11, 2014, at 6:22 AM, Jeff Squyres (jsquyres) <jsquyres(a)cisco.com> wrote: > >> I won't have time to do any prep in detail until after SC and the subsequent US thanksgiving holiday, but are these 2 topics that we want to discuss in December? >> >> Sent from my phone. No type good. >> _______________________________________________ >> mpiwg-fortran mailing list >> mpiwg-fortran(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran > > Bill Long longb(a)cray.com > Fortran Technical Suport & voice: 651-605-9024 > Bioinformatics Software Development fax: 651-605-9142 > Cray Inc./ Cray Plaza, Suite 210/ 380 Jackson St./ St. Paul, MN 55101 > > > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran -- Jeff Squyres jsquyres(a)cisco.com For corporate legal information go to: http://www.cisco.com/web/about/doing_business/legal/cri/

1 0

Re: [MPIWG Fortran] Dec mtg: MPI 4: deprecate SIZEOF and mpif.h
by Balaji, Pavan 11 Nov '14

11 Nov '14

With one minor addition — new functionality will not have deprecated equivalents (at least that’s the model we used with C++). So everything that gets added in MPI-4 will be “use mpi” or “use mpi_f08” only, if this passes. — Pavan > On Nov 11, 2014, at 8:09 AM, William Gropp <wgropp(a)illinois.edu> wrote: > > Deprecate in MPI uses the same definition that the Fortran standards have used - discouraged from use and a candidate for deletion in a future standard. Thus, until formally deleted, it must be implemented and appear in mpif.h (as mpif.h is now defined). > > Bill > > On Nov 11, 2014, at 8:01 AM, Bill Long <longb(a)cray.com> wrote: > >> Seems reasonable to me. >> >> Are there plans for a Webex meeting for the Fortran working group before the December MPI meeting in San Jose? This is a “vacation challenged” time of the year - I’m scheduled to be in the office only 3 days between December 5 and 31. >> >> How does MPI define “deprecate”? For example, would any new feature/function be excluded from mpif.h? Personally, I think that would be a good idea - it simplifies maintenance of the distribution and emphasizes that mpif.h is for legacy use only. But I can imagine other opinions. >> >> Cheers, >> Bill >> >> >> >> On Nov 11, 2014, at 6:22 AM, Jeff Squyres (jsquyres) <jsquyres(a)cisco.com> wrote: >> >>> I won't have time to do any prep in detail until after SC and the subsequent US thanksgiving holiday, but are these 2 topics that we want to discuss in December? >>> >>> Sent from my phone. No type good. >>> _______________________________________________ >>> mpiwg-fortran mailing list >>> mpiwg-fortran(a)lists.mpi-forum.org >>> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran >> >> Bill Long longb(a)cray.com >> Fortran Technical Suport & voice: 651-605-9024 >> Bioinformatics Software Development fax: 651-605-9142 >> Cray Inc./ Cray Plaza, Suite 210/ 380 Jackson St./ St. Paul, MN 55101 >> >> >> _______________________________________________ >> mpiwg-fortran mailing list >> mpiwg-fortran(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran > > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran -- Pavan Balaji ✉️ http://www.mcs.anl.gov/~balaji

1 0

Re: [MPIWG Fortran] Dec mtg: MPI 4: deprecate SIZEOF and mpif.h
by William Gropp 11 Nov '14

11 Nov '14

Deprecate in MPI uses the same definition that the Fortran standards have used - discouraged from use and a candidate for deletion in a future standard. Thus, until formally deleted, it must be implemented and appear in mpif.h (as mpif.h is now defined). Bill On Nov 11, 2014, at 8:01 AM, Bill Long <longb(a)cray.com> wrote: > Seems reasonable to me. > > Are there plans for a Webex meeting for the Fortran working group before the December MPI meeting in San Jose? This is a “vacation challenged” time of the year - I’m scheduled to be in the office only 3 days between December 5 and 31. > > How does MPI define “deprecate”? For example, would any new feature/function be excluded from mpif.h? Personally, I think that would be a good idea - it simplifies maintenance of the distribution and emphasizes that mpif.h is for legacy use only. But I can imagine other opinions. > > Cheers, > Bill > > > > On Nov 11, 2014, at 6:22 AM, Jeff Squyres (jsquyres) <jsquyres(a)cisco.com> wrote: > >> I won't have time to do any prep in detail until after SC and the subsequent US thanksgiving holiday, but are these 2 topics that we want to discuss in December? >> >> Sent from my phone. No type good. >> _______________________________________________ >> mpiwg-fortran mailing list >> mpiwg-fortran(a)lists.mpi-forum.org >> http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran > > Bill Long longb(a)cray.com > Fortran Technical Suport & voice: 651-605-9024 > Bioinformatics Software Development fax: 651-605-9142 > Cray Inc./ Cray Plaza, Suite 210/ 380 Jackson St./ St. Paul, MN 55101 > > > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > http://lists.mpi-forum.org/mailman/listinfo.cgi/mpiwg-fortran

1 0