- mpiwg-rma - mailman3.cels.anl.gov

Re: [mpiwg-rma] New RMA-WG meeting slot for 2022
by Thakur, Rajeev 03 Feb '22

03 Feb '22

The calendar invite downloaded from the link below sets it up as weekly, and just above the link below the mail also says weekly. Biweekly is better. -----Original Message----- From: Joseph Schuchart <schuchart(a)icl.utk.edu> Date: Thursday, February 3, 2022 at 10:47 AM To: "Thakur, Rajeev" <thakur(a)anl.gov>, MPI WG Remote Memory Access working group <mpiwg-rma(a)lists.mpi-forum.org> Subject: Re: [mpiwg-rma] New RMA-WG meeting slot for 2022 Rajeev, all, Sorry, just saw your email. We're meeting bi-weekly, next time is next week Thursday at 11am ET then :) Joseph On 2/3/22 10:42, Thakur, Rajeev wrote: > Is this a weekly call, and is there one today? If so, what are the agenda topics? > > Rajeev > > -----Original Message----- > From: mpiwg-rma <mpiwg-rma-bounces(a)lists.mpi-forum.org> on behalf of Joseph Schuchart via mpiwg-rma <mpiwg-rma(a)lists.mpi-forum.org> > Reply-To: MPI WG Remote Memory Access working group <mpiwg-rma(a)lists.mpi-forum.org> > Date: Tuesday, January 25, 2022 at 9:03 AM > To: "mpiwg-rma(a)lists.mpi-forum.org" <mpiwg-rma(a)lists.mpi-forum.org> > Cc: Joseph Schuchart <schuchart(a)icl.utk.edu> > Subject: [mpiwg-rma] New RMA-WG meeting slot for 2022 > > All, > > Thank you for putting your availabilities into the doodle. It appears > that some people had mail issues so getting everyone's vote in took a > little longer. Finding a slot isn't straight-forward as there is always > at least one person with a conflict. After checking back with Nathan and > Bill, I think Thursdays 10am CT / 11am ET / 5pm CET is the best choice. > > Let's have our first meeting of 2022 this week, January 27. Dan went > over the RMA chapter with his Semantic Terms hat on and collected points > that need discussion. I will summarize what Dan sent me so that we can > discuss them during the call. > > Here is dial-in information: > > Please download and import the following iCalendar (.ics) files to your > calendar system. > Weekly: > https://tennessee.zoom.us/meeting/tJcudeyorDwvE9chI15gePR0X-UicQzSd2ll/ics?… > > Join from PC, Mac, Linux, iOS or Android: > https://tennessee.zoom.us/j/93325027593?pwd=amEvNlBZNjlXS0pTaUNKVm1zL3d4Zz09 > Password: 986374 > > Or iPhone one-tap (US Toll): +13017158592,93325027593# or > +13126266799,93325027593# > > Or Telephone: > Dial: > +1 301 715 8592 (US Toll) > +1 312 626 6799 (US Toll) > +1 646 876 9923 (US Toll) > +1 253 215 8782 (US Toll) > +1 346 248 7799 (US Toll) > +1 669 900 6833 (US Toll) > Meeting ID: 933 2502 7593 > International numbers available: https://tennessee.zoom.us/u/adDSqwuNDO > > Looking forward to seeing everyone back on Thursday, > Joseph > _______________________________________________ > mpiwg-rma mailing list > mpiwg-rma(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-rma >

1 0

Re: [mpiwg-rma] New RMA-WG meeting slot for 2022
by Joseph Schuchart 03 Feb '22

03 Feb '22

Rajeev, all, Sorry, just saw your email. We're meeting bi-weekly, next time is next week Thursday at 11am ET then :) Joseph On 2/3/22 10:42, Thakur, Rajeev wrote: > Is this a weekly call, and is there one today? If so, what are the agenda topics? > > Rajeev > > -----Original Message----- > From: mpiwg-rma <mpiwg-rma-bounces(a)lists.mpi-forum.org> on behalf of Joseph Schuchart via mpiwg-rma <mpiwg-rma(a)lists.mpi-forum.org> > Reply-To: MPI WG Remote Memory Access working group <mpiwg-rma(a)lists.mpi-forum.org> > Date: Tuesday, January 25, 2022 at 9:03 AM > To: "mpiwg-rma(a)lists.mpi-forum.org" <mpiwg-rma(a)lists.mpi-forum.org> > Cc: Joseph Schuchart <schuchart(a)icl.utk.edu> > Subject: [mpiwg-rma] New RMA-WG meeting slot for 2022 > > All, > > Thank you for putting your availabilities into the doodle. It appears > that some people had mail issues so getting everyone's vote in took a > little longer. Finding a slot isn't straight-forward as there is always > at least one person with a conflict. After checking back with Nathan and > Bill, I think Thursdays 10am CT / 11am ET / 5pm CET is the best choice. > > Let's have our first meeting of 2022 this week, January 27. Dan went > over the RMA chapter with his Semantic Terms hat on and collected points > that need discussion. I will summarize what Dan sent me so that we can > discuss them during the call. > > Here is dial-in information: > > Please download and import the following iCalendar (.ics) files to your > calendar system. > Weekly: > https://tennessee.zoom.us/meeting/tJcudeyorDwvE9chI15gePR0X-UicQzSd2ll/ics?… > > Join from PC, Mac, Linux, iOS or Android: > https://tennessee.zoom.us/j/93325027593?pwd=amEvNlBZNjlXS0pTaUNKVm1zL3d4Zz09 > Password: 986374 > > Or iPhone one-tap (US Toll): +13017158592,93325027593# or > +13126266799,93325027593# > > Or Telephone: > Dial: > +1 301 715 8592 (US Toll) > +1 312 626 6799 (US Toll) > +1 646 876 9923 (US Toll) > +1 253 215 8782 (US Toll) > +1 346 248 7799 (US Toll) > +1 669 900 6833 (US Toll) > Meeting ID: 933 2502 7593 > International numbers available: https://tennessee.zoom.us/u/adDSqwuNDO > > Looking forward to seeing everyone back on Thursday, > Joseph > _______________________________________________ > mpiwg-rma mailing list > mpiwg-rma(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-rma >

1 0

Re: [mpiwg-rma] New RMA-WG meeting slot for 2022
by Thakur, Rajeev 03 Feb '22

03 Feb '22

Is this a weekly call, and is there one today? If so, what are the agenda topics? Rajeev -----Original Message----- From: mpiwg-rma <mpiwg-rma-bounces(a)lists.mpi-forum.org> on behalf of Joseph Schuchart via mpiwg-rma <mpiwg-rma(a)lists.mpi-forum.org> Reply-To: MPI WG Remote Memory Access working group <mpiwg-rma(a)lists.mpi-forum.org> Date: Tuesday, January 25, 2022 at 9:03 AM To: "mpiwg-rma(a)lists.mpi-forum.org" <mpiwg-rma(a)lists.mpi-forum.org> Cc: Joseph Schuchart <schuchart(a)icl.utk.edu> Subject: [mpiwg-rma] New RMA-WG meeting slot for 2022 All, Thank you for putting your availabilities into the doodle. It appears that some people had mail issues so getting everyone's vote in took a little longer. Finding a slot isn't straight-forward as there is always at least one person with a conflict. After checking back with Nathan and Bill, I think Thursdays 10am CT / 11am ET / 5pm CET is the best choice. Let's have our first meeting of 2022 this week, January 27. Dan went over the RMA chapter with his Semantic Terms hat on and collected points that need discussion. I will summarize what Dan sent me so that we can discuss them during the call. Here is dial-in information: Please download and import the following iCalendar (.ics) files to your calendar system. Weekly: https://tennessee.zoom.us/meeting/tJcudeyorDwvE9chI15gePR0X-UicQzSd2ll/ics?… Join from PC, Mac, Linux, iOS or Android: https://tennessee.zoom.us/j/93325027593?pwd=amEvNlBZNjlXS0pTaUNKVm1zL3d4Zz09 Password: 986374 Or iPhone one-tap (US Toll): +13017158592,93325027593# or +13126266799,93325027593# Or Telephone: Dial: +1 301 715 8592 (US Toll) +1 312 626 6799 (US Toll) +1 646 876 9923 (US Toll) +1 253 215 8782 (US Toll) +1 346 248 7799 (US Toll) +1 669 900 6833 (US Toll) Meeting ID: 933 2502 7593 International numbers available: https://tennessee.zoom.us/u/adDSqwuNDO Looking forward to seeing everyone back on Thursday, Joseph _______________________________________________ mpiwg-rma mailing list mpiwg-rma(a)lists.mpi-forum.org https://lists.mpi-forum.org/mailman/listinfo/mpiwg-rma

1 0

New RMA-WG meeting slot for 2022
by Joseph Schuchart 25 Jan '22

25 Jan '22

All, Thank you for putting your availabilities into the doodle. It appears that some people had mail issues so getting everyone's vote in took a little longer. Finding a slot isn't straight-forward as there is always at least one person with a conflict. After checking back with Nathan and Bill, I think Thursdays 10am CT / 11am ET / 5pm CET is the best choice. Let's have our first meeting of 2022 this week, January 27. Dan went over the RMA chapter with his Semantic Terms hat on and collected points that need discussion. I will summarize what Dan sent me so that we can discuss them during the call. Here is dial-in information: Please download and import the following iCalendar (.ics) files to your calendar system. Weekly: https://tennessee.zoom.us/meeting/tJcudeyorDwvE9chI15gePR0X-UicQzSd2ll/ics?… Join from PC, Mac, Linux, iOS or Android: https://tennessee.zoom.us/j/93325027593?pwd=amEvNlBZNjlXS0pTaUNKVm1zL3d4Zz09 Password: 986374 Or iPhone one-tap (US Toll): +13017158592,93325027593# or +13126266799,93325027593# Or Telephone: Dial: +1 301 715 8592 (US Toll) +1 312 626 6799 (US Toll) +1 646 876 9923 (US Toll) +1 253 215 8782 (US Toll) +1 346 248 7799 (US Toll) +1 669 900 6833 (US Toll) Meeting ID: 933 2502 7593 International numbers available: https://tennessee.zoom.us/u/adDSqwuNDO Looking forward to seeing everyone back on Thursday, Joseph

1 0

Re: [mpiwg-rma] MPI RMA WG 2022 Poll
by Joseph Schuchart 12 Jan '22

12 Jan '22

All, Thanks to those who have filled out the poll already. If you haven't done so already, please do so before the end of Friday this week. It already seems to be tricky to find a slot that fits everyone. Our best options seem to be Monday 11am EST (missing Bill) or Wednesday 11am EST (missing Nathan). Let's see if we can get more votes... Cheers Joseph On 12/31/21 6:22 PM, Joseph Schuchart wrote: > Dear all, > > Happy New Year everyone! With the new year, I would like to see > whether we can find a different slot for the bi-weekly meeting that > works better for those calling in from Europe. I have created a new > doodle poll: > > https://doodle.com/poll/maf44rxd2hx8wf5f > > Please add your availability soon. The dates are placeholders, I'm > still planing to meet every two weeks (weekly if we find it necessary > at some point). If either odd or even weeks work better for you for > some slots, please leave a comment or send me an email. We haven't had > any participants from the US West Coast so I added some earlier slots, > ranging from 9am US Central / 10am Eastern / 4pm CET thru 1pm US > Central / 2pm Easter / 8pm CET (I hope we can find a slot in the > earlier hours). I have left out the Wednesday slots for the Hybrid WG > and the MPI Forum meeting, to avoid creating conflicts. > > Looking forward to resuming the meetings after the break! > Joseph

1 0

MPI RMA WG 2022 Poll
by Joseph Schuchart 31 Dec '21

31 Dec '21

Dear all, Happy New Year everyone! With the new year, I would like to see whether we can find a different slot for the bi-weekly meeting that works better for those calling in from Europe. I have created a new doodle poll: https://doodle.com/poll/maf44rxd2hx8wf5f Please add your availability soon. The dates are placeholders, I'm still planing to meet every two weeks (weekly if we find it necessary at some point). If either odd or even weeks work better for you for some slots, please leave a comment or send me an email. We haven't had any participants from the US West Coast so I added some earlier slots, ranging from 9am US Central / 10am Eastern / 4pm CET thru 1pm US Central / 2pm Easter / 8pm CET (I hope we can find a slot in the earlier hours). I have left out the Wednesday slots for the Hybrid WG and the MPI Forum meeting, to avoid creating conflicts. Looking forward to resuming the meetings after the break! Joseph

1 0

No RMA-WG meeting today
by Joseph Schuchart 14 Dec '21

14 Dec '21

All, I am attending a workshop this week so I cannot host the WG meeting (the room should be open if . I will send out a doodle for a potentially new 2022 time slot later this week, to see whether we can find a slot that works better for our European friends. Sorry for the short notice. Cheers Joseph

1 0

Re: [mpiwg-rma] [MPIWG Fortran] "small count" variants of RMA functions
by Jeff Hammond 29 Nov '21

29 Nov '21

I don't think it's worth a plenary. It really comes down to whether the Fortran Wizards (Rolf, Bill, Hubert) accept this. The implementation of this change is easy. Jeff On Mon, Nov 29, 2021 at 4:03 PM Wesley Bland <work(a)wesbland.com> wrote: > As long as we’re having remote meetings, we’re not setting up WG time > during the meeting itself. So each WG is just having its usual con calls on > their own schedule (or in the case of Fortran, not). If you folks want to > throw together a plenary, we can still put that on the agenda though. > > Thanks, > Wes > > On Nov 29, 2021, at 3:45 AM, Jeff Hammond via mpiwg-fortran < > mpiwg-fortran(a)lists.mpi-forum.org> wrote: > > I created https://github.com/mpi-forum/mpi-issues/issues/528. > > Normally, we would have a Fortran WG meeting during the quarterly > meeting. Is there going to be a Fortran WG session at the December meeting? > > Jeff > > On Sat, Nov 27, 2021 at 9:14 PM William Gropp <wgropp(a)illinois.edu> wrote: > >> I agree with Jeff here. This is a good enhancement; I believe the only >> reason that we hadn’t done this before is that we’ve stuck closely to the >> original Fortran look-and-feel - which was pre-Fortran 90. We could have >> done this earlier, and now that there are explicit interfaces that are only >> available in Fortran08, there is no reason not to do this. >> >> Bill >> >> William Gropp >> Director, NCSA >> Thomas M. Siebel Chair in Computer Science >> University of Illinois Urbana-Champaign >> IEEE-CS President-Elect >> >> >> >> >> On Nov 25, 2021, at 4:42 AM, Jeff Hammond via mpiwg-fortran < >> mpiwg-fortran(a)lists.mpi-forum.org> wrote: >> >> RMA absolutely did the right thing with MPI_Aint size and displacement >> arguments. >> >> In C, it is no problem to pass 32b integers to these arguments when AINT >> is 64b, because C type promotion rules just work. >> >> In Fortran, I cannot do this, because an integer literal is not >> compatible with an AINT argument. I find it tedious to have to declare a >> variable just for this, or to explicitly cast with integer(100, >> MPI_ADDRESS_KIND). >> >> Since Fortran does polymorphism right and can have subroutine >> declarations for both default integer and AINT integer scalar arguments, is >> there a good reason not to add the former? >> >> This proposal applies not to just MPI_Win_allocate but any RMA function >> that uses AINT scalar arguments. >> >> Thanks, >> >> Jeff >> >> % mpifort y.F90 >> y.F90:20:60: >> >> 20 | call MPI_Win_allocate(100, 1, MPI_INFO_NULL, comm, XA, WA) >> | 1 >> Error: There is no specific subroutine for the generic 'mpi_win_allocate' >> at (1) >> >> % cat y.F90 >> ! USE mpi_f08 >> ! MPI_Win_allocate(size, disp_unit, info, comm, baseptr, win, ierror) >> ! USE, INTRINSIC :: ISO_C_BINDING, ONLY : C_PTR >> ! INTEGER(KIND=MPI_ADDRESS_KIND), INTENT(IN) :: size >> ! INTEGER, INTENT(IN) :: disp_unit >> ! TYPE(MPI_Info), INTENT(IN) :: info >> ! TYPE(MPI_Comm), INTENT(IN) :: comm >> ! TYPE(C_PTR), INTENT(OUT) :: baseptr >> ! TYPE(MPI_Win), INTENT(OUT) :: win >> ! INTEGER, OPTIONAL, INTENT(OUT) :: ierror >> program main >> use iso_fortran_env >> use mpi_f08 >> implicit none >> integer(kind=MPI_ADDRESS_KIND) :: as = 100 >> type(c_ptr) :: XA >> type(MPI_Win) :: WA >> TYPE(MPI_Comm) :: comm = MPI_COMM_WORLD >> call MPI_Win_allocate(as, 1, MPI_INFO_NULL, comm, XA, WA) >> call MPI_Win_allocate(100, 1, MPI_INFO_NULL, comm, XA, WA) >> end program main >> >> % mpicc y.c && echo SUCCESS >> SUCCESS >> >> % cat y.c >> #include <mpi.h> >> int main(void) >> { >> MPI_Aint as = 100; >> int * XA; >> MPI_Win WA; >> MPI_Win_allocate(as, 1, MPI_INFO_NULL, MPI_COMM_WORLD, XA, &WA); >> MPI_Win_allocate(100, 1, MPI_INFO_NULL, MPI_COMM_WORLD, XA, &WA); >> } >> >> >> -- >> Jeff Hammond >> jeff.science(a)gmail.com >> http://jeffhammond.github.io/ >> <https://urldefense.com/v3/__http://jeffhammond.github.io/__;!!DZ3fjg!rCl1MJ…> >> _______________________________________________ >> mpiwg-fortran mailing list >> mpiwg-fortran(a)lists.mpi-forum.org >> >> https://urldefense.com/v3/__https://lists.mpi-forum.org/mailman/listinfo/mp… >> >> >> >> > > -- > Jeff Hammond > jeff.science(a)gmail.com > http://jeffhammond.github.io/ > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-fortran > > > -- Jeff Hammond jeff.science(a)gmail.com http://jeffhammond.github.io/

1 0

Re: [mpiwg-rma] [MPIWG Fortran] "small count" variants of RMA functions
by Wesley Bland 29 Nov '21

29 Nov '21

As long as we’re having remote meetings, we’re not setting up WG time during the meeting itself. So each WG is just having its usual con calls on their own schedule (or in the case of Fortran, not). If you folks want to throw together a plenary, we can still put that on the agenda though. Thanks, Wes > On Nov 29, 2021, at 3:45 AM, Jeff Hammond via mpiwg-fortran <mpiwg-fortran(a)lists.mpi-forum.org> wrote: > > I created https://github.com/mpi-forum/mpi-issues/issues/528 <https://github.com/mpi-forum/mpi-issues/issues/528>. > > Normally, we would have a Fortran WG meeting during the quarterly meeting. Is there going to be a Fortran WG session at the December meeting? > > Jeff > > On Sat, Nov 27, 2021 at 9:14 PM William Gropp <wgropp(a)illinois.edu <mailto:[email protected]>> wrote: > I agree with Jeff here. This is a good enhancement; I believe the only reason that we hadn’t done this before is that we’ve stuck closely to the original Fortran look-and-feel - which was pre-Fortran 90. We could have done this earlier, and now that there are explicit interfaces that are only available in Fortran08, there is no reason not to do this. > > Bill > > William Gropp > Director, NCSA > Thomas M. Siebel Chair in Computer Science > University of Illinois Urbana-Champaign > IEEE-CS President-Elect > > > > >> On Nov 25, 2021, at 4:42 AM, Jeff Hammond via mpiwg-fortran <mpiwg-fortran(a)lists.mpi-forum.org <mailto:[email protected]>> wrote: >> >> RMA absolutely did the right thing with MPI_Aint size and displacement arguments. >> >> In C, it is no problem to pass 32b integers to these arguments when AINT is 64b, because C type promotion rules just work. >> >> In Fortran, I cannot do this, because an integer literal is not compatible with an AINT argument. I find it tedious to have to declare a variable just for this, or to explicitly cast with integer(100,MPI_ADDRESS_KIND). >> >> Since Fortran does polymorphism right and can have subroutine declarations for both default integer and AINT integer scalar arguments, is there a good reason not to add the former? >> >> This proposal applies not to just MPI_Win_allocate but any RMA function that uses AINT scalar arguments. >> >> Thanks, >> >> Jeff >> >> % mpifort y.F90 >> y.F90:20:60: >> >> 20 | call MPI_Win_allocate(100, 1, MPI_INFO_NULL, comm, XA, WA) >> | 1 >> Error: There is no specific subroutine for the generic 'mpi_win_allocate' at (1) >> >> % cat y.F90 >> ! USE mpi_f08 >> ! MPI_Win_allocate(size, disp_unit, info, comm, baseptr, win, ierror) >> ! USE, INTRINSIC :: ISO_C_BINDING, ONLY : C_PTR >> ! INTEGER(KIND=MPI_ADDRESS_KIND), INTENT(IN) :: size >> ! INTEGER, INTENT(IN) :: disp_unit >> ! TYPE(MPI_Info), INTENT(IN) :: info >> ! TYPE(MPI_Comm), INTENT(IN) :: comm >> ! TYPE(C_PTR), INTENT(OUT) :: baseptr >> ! TYPE(MPI_Win), INTENT(OUT) :: win >> ! INTEGER, OPTIONAL, INTENT(OUT) :: ierror >> program main >> use iso_fortran_env >> use mpi_f08 >> implicit none >> integer(kind=MPI_ADDRESS_KIND) :: as = 100 >> type(c_ptr) :: XA >> type(MPI_Win) :: WA >> TYPE(MPI_Comm) :: comm = MPI_COMM_WORLD >> call MPI_Win_allocate(as, 1, MPI_INFO_NULL, comm, XA, WA) >> call MPI_Win_allocate(100, 1, MPI_INFO_NULL, comm, XA, WA) >> end program main >> >> % mpicc y.c && echo SUCCESS >> SUCCESS >> >> % cat y.c >> #include <mpi.h> >> int main(void) >> { >> MPI_Aint as = 100; >> int * XA; >> MPI_Win WA; >> MPI_Win_allocate(as, 1, MPI_INFO_NULL, MPI_COMM_WORLD, XA, &WA); >> MPI_Win_allocate(100, 1, MPI_INFO_NULL, MPI_COMM_WORLD, XA, &WA); >> } >> >> >> -- >> Jeff Hammond >> jeff.science(a)gmail.com <mailto:[email protected]> >> http://jeffhammond.github.io/ <https://urldefense.com/v3/__http://jeffhammond.github.io/__;!!DZ3fjg!rCl1MJ…>_______________________________________________ >> mpiwg-fortran mailing list >> mpiwg-fortran(a)lists.mpi-forum.org <mailto:[email protected]> >> https://urldefense.com/v3/__https://lists.mpi-forum.org/mailman/listinfo/mp… <https://urldefense.com/v3/__https://lists.mpi-forum.org/mailman/listinfo/mp…> > > > > -- > Jeff Hammond > jeff.science(a)gmail.com <mailto:[email protected]> > http://jeffhammond.github.io/ <http://jeffhammond.github.io/>_______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org <mailto:[email protected]> > https://lists.mpi-forum.org/mailman/listinfo/mpiwg-fortran <https://lists.mpi-forum.org/mailman/listinfo/mpiwg-fortran>

1 0

Re: [mpiwg-rma] [MPIWG Fortran] "small count" variants of RMA functions
by Jeff Hammond 29 Nov '21

29 Nov '21

I created https://github.com/mpi-forum/mpi-issues/issues/528. Normally, we would have a Fortran WG meeting during the quarterly meeting. Is there going to be a Fortran WG session at the December meeting? Jeff On Sat, Nov 27, 2021 at 9:14 PM William Gropp <wgropp(a)illinois.edu> wrote: > I agree with Jeff here. This is a good enhancement; I believe the only > reason that we hadn’t done this before is that we’ve stuck closely to the > original Fortran look-and-feel - which was pre-Fortran 90. We could have > done this earlier, and now that there are explicit interfaces that are only > available in Fortran08, there is no reason not to do this. > > Bill > > William Gropp > Director, NCSA > Thomas M. Siebel Chair in Computer Science > University of Illinois Urbana-Champaign > IEEE-CS President-Elect > > > > > On Nov 25, 2021, at 4:42 AM, Jeff Hammond via mpiwg-fortran < > mpiwg-fortran(a)lists.mpi-forum.org> wrote: > > RMA absolutely did the right thing with MPI_Aint size and displacement > arguments. > > In C, it is no problem to pass 32b integers to these arguments when AINT > is 64b, because C type promotion rules just work. > > In Fortran, I cannot do this, because an integer literal is not compatible > with an AINT argument. I find it tedious to have to declare a variable > just for this, or to explicitly cast with integer(100,MPI_ADDRESS_KIND). > > Since Fortran does polymorphism right and can have subroutine declarations > for both default integer and AINT integer scalar arguments, is there a good > reason not to add the former? > > This proposal applies not to just MPI_Win_allocate but any RMA function > that uses AINT scalar arguments. > > Thanks, > > Jeff > > % mpifort y.F90 > y.F90:20:60: > > 20 | call MPI_Win_allocate(100, 1, MPI_INFO_NULL, comm, XA, WA) > | 1 > Error: There is no specific subroutine for the generic 'mpi_win_allocate' > at (1) > > % cat y.F90 > ! USE mpi_f08 > ! MPI_Win_allocate(size, disp_unit, info, comm, baseptr, win, ierror) > ! USE, INTRINSIC :: ISO_C_BINDING, ONLY : C_PTR > ! INTEGER(KIND=MPI_ADDRESS_KIND), INTENT(IN) :: size > ! INTEGER, INTENT(IN) :: disp_unit > ! TYPE(MPI_Info), INTENT(IN) :: info > ! TYPE(MPI_Comm), INTENT(IN) :: comm > ! TYPE(C_PTR), INTENT(OUT) :: baseptr > ! TYPE(MPI_Win), INTENT(OUT) :: win > ! INTEGER, OPTIONAL, INTENT(OUT) :: ierror > program main > use iso_fortran_env > use mpi_f08 > implicit none > integer(kind=MPI_ADDRESS_KIND) :: as = 100 > type(c_ptr) :: XA > type(MPI_Win) :: WA > TYPE(MPI_Comm) :: comm = MPI_COMM_WORLD > call MPI_Win_allocate(as, 1, MPI_INFO_NULL, comm, XA, WA) > call MPI_Win_allocate(100, 1, MPI_INFO_NULL, comm, XA, WA) > end program main > > % mpicc y.c && echo SUCCESS > SUCCESS > > % cat y.c > #include <mpi.h> > int main(void) > { > MPI_Aint as = 100; > int * XA; > MPI_Win WA; > MPI_Win_allocate(as, 1, MPI_INFO_NULL, MPI_COMM_WORLD, XA, &WA); > MPI_Win_allocate(100, 1, MPI_INFO_NULL, MPI_COMM_WORLD, XA, &WA); > } > > > -- > Jeff Hammond > jeff.science(a)gmail.com > http://jeffhammond.github.io/ > <https://urldefense.com/v3/__http://jeffhammond.github.io/__;!!DZ3fjg!rCl1MJ…> > _______________________________________________ > mpiwg-fortran mailing list > mpiwg-fortran(a)lists.mpi-forum.org > > https://urldefense.com/v3/__https://lists.mpi-forum.org/mailman/listinfo/mp… > > > -- Jeff Hammond jeff.science(a)gmail.com http://jeffhammond.github.io/

1 0