[R] Using Fortran with MPI, RInside, and calling R functions

[peter dalgaard]

Yeah, well, not much harm done, but once compilers are involved, r-devel is usually preferred over r-help.

-pd

> On 27 Feb 2016, at 21:30 , Erin Hodgess <erinm.hodgess at gmail.com> wrote:
> 
> Sorry...thought it was ok since it uses RInside and Rcpp.
> 
> 
> On Sat, Feb 27, 2016 at 2:15 PM, Jeff Newmiller <jdnewmil at dcn.davis.ca.us>
> wrote:
> 
>> This is off topic here... wrong audience. Read the Posting Guide.
>> --
>> Sent from my phone. Please excuse my brevity.
>> 
>> On February 27, 2016 12:00:23 PM PST, Erin Hodgess <
>> erinm.hodgess at gmail.com> wrote:
>> 
>>> Hello again.
>>> 
>>> This time, I would like to add MPI to my Fortran program.  Here are the
>>> Fortran and C++ codes:
>>> 
>>> program buzzy
>>>   use iso_c_binding
>>>  implicit none
>>>  include '/opt/openmpi/include/mpif.h'
>>> 
>>> 
>>> 
>>> 
>>>  integer :: rank,size,ierror,tag,status(MPI_STATUS_SIZE), i,np
>>>  integer :: argc = 1000000
>>>  real :: x,tot1
>>>  character(len=32) :: argv
>>> 
>>> 
>>>      INTERFACE
>>>         SUBROUTINE R_FUN(argc,argv) bind(C, name="buzzyC")
>>>           use iso_c_binding
>>>        character(kind=c_char), INTENT(INOUT) :: argv
>>>        INTEGER(kind=c_int), INTENT(IN) :: argc
>>> 
>>>      END SUBROUTINE R_FUN
>>>    END INTERFACE
>>> 
>>>      call MPI_INIT(ierror)
>>>      call MPI_COMM_SIZE(MPI_COMM_WORLD,size,ierror)
>>>      call MPI_COMM_RANK(MPI_COMM_WORLD,rank,ierror)
>>> 
>>> 
>>> 
>>> 
>>>        print *, "Fortran Calling RInside",rank
>>>    CALL
>>> R_FUN (argc,argv)
>>>    print *,rank
>>> 
>>>    call MPI_FINALIZE(ierror)
>>> 
>>>  end program buzzy
>>> 
>>> and
>>> 
>>> #include <iostream>
>>> #include <RInside.h>
>>> 
>>> void buzzyC_(int argc,char *argv[]);
>>> 
>>> extern "C" void buzzyC(int argc,char *argv[]) {
>>> 
>>>    // create an embedded R instance
>>>  RInside R(argc,argv);
>>> 
>>>    // convert to string for RInside assignment
>>> 
>>> 
>>>    // eval the string, give R notice
>>>  R.parseEvalQ("cat(mean(rnorm(argc))");
>>> }
>>> 
>>> Now my steps for compiling and linking are the following:
>>> 
>>> erin at erin-Bonobo-Extreme:~$ mpif90 -c buzzy.f90
>>> erin at erin-Bonobo-Extreme:~$ mpic++ buzzyC.cpp -c
>>> -I/home/erin/R/x86_64-pc-linux-gnu-library/3.2/RInside/include
>>> -I/home/erin/R/x86_64-pc-linux-gnu-library/3.2/Rcpp/include
>>> -I/usr/share/R/include -libstdc++
>>> erin at erin-Bonobo-Extreme:~$ mpifort -o fcra buzzy.o buzzyC.o
>>> -L/usr/lib/R/lib -lR
>>> -L
>>> /home/erin/R/x86_64-pc-linux-gnu-library/3.2/RInside/lib -lRInside
>>> -L/home/erin/R/x86_64-pc-linux-gnu-library/3.2/Rcpp/libs/
>>> -Wl,-rpath,/home/erin/R/x86_64-pc-linux-gnu-library/3.2/RInside/lib/
>>> -lRInside
>>> -Wl,-rpath,/home/erin/R/x86_64-pc-linux-gnu-library/3.2/Rcpp/libs/ -lstdc++
>>> /usr/lib/x86_64-linux-gnu/libstdc++.so.6
>>> 
>>> So far so good
>>> But when I run this, disaster strikes:
>>> 
>>> erin at erin-Bonobo-Extreme:~$ mpirun -np 4 ./fcra
>>> Fortran Calling RInside           0
>>> Fortran Calling RInside           1
>>> Fortran Calling RInside           3
>>> Fortran Calling RInside           2
>>> 
>>> Program received signal SIGSEGV: Segmentation fault - invalid memory
>>> reference.
>>> 
>>> Backtrace for this error:
>>> 
>>> Program received signal SIGSEGV: Segmentation fault - invalid memory
>>> reference.
>>> 
>>> Backtrace for this error:
>>> #0  0x7FC59706CE48
>>> #1  0x7FC59706BFD0
>>> #2  0x7FC596AA52EF
>>> #3
>>> 0x7FC596AFB69A
>>> #4  0x7FC597C5E8E8
>>> #5  0x7FC5979671E8
>>> #6  0x7FC5979677A1
>>> #7  0x402A55 in buzzyC
>>> #8  0x402891 in MAIN__ at buzzy.f90:?
>>> #0  0x7F2482294E48
>>> #1  0x7F2482293FD0
>>> #2  0x7F2481CCD2EF
>>> #3  0x7F2481D2369A
>>> #4  0x7F2482E868E8
>>> #5  0x7F2482B8F1E8
>>> #6  0x7F2482B8F7A1
>>> #7  0x402A55 in buzzyC
>>> #8  0x402891 in MAIN__ at buzzy.f90:?
>>> ------------------------------
>>> 
>>> mpirun noticed that process rank 1 with PID 2188 on node
>>> erin-Bonobo-Extreme exited on signal 11 (Segmentation fault).
>>> ------------------------------
>>> 
>>> erin at erin-Bonobo-Extreme:~$
>>> 
>>> Maybe I should be asking:  is this even possible, please?
>>> 
>>> Thanks,
>>> Erin
>>> 
>>> 
> 
> 
> -- 
> Erin Hodgess
> Associate Professor
> Department of Mathematical and Statistics
> University of Houston - Downtown
> mailto: erinm.hodgess at gmail.com
> 
> 	[[alternative HTML version deleted]]
> 
> ______________________________________________
> R-help at r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.

-- 
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Office: A 4.23
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com