"relocation overflow" and "relocation truncated to fit errors" for Fortran application with g77 compiler
sameer sinha
sinha.sam0201@gmail.com
Thu Jun 5 11:44:00 GMT 2008
My code (FORTRAN 77) successfully compiled on 32 bit Xeon cluster with
8 GB RAM(RHEL platform)
but when
I am using a 64-bit AMD apteron cluster with 8 GB RAM (RHEL platform).
I face "relocation overflow" and "relocation truncated to fit errors"
when I compile big fortran 77 codes using f77/g77 (even with ifort and
pathf95) on it. Though it compiles successfully it cannot assemble or
link the executables. I understand this is because of large arrays or
so. But the executable size is not more than 2GB. so I could not run
my codes with large dimensions.
Can you please tell me what options I can use while compiling thse
kind of codes. Sometimes, ulimit command or -m32 option worked but
when I run the executable it simply dumps the core. Please suggest me
what can I do to get rid of these problems?.
I am getting this error while compilation
[cdacb@master IIAP]$ mpif77 -o step3.x ccea_MPI.F
ccea_MPI.F: In subroutine `twoint':
/tmp/cc5KpcVF.f:3621: warning:
call MPI_RECV(maxinp,1,MPI_INTEGER,isource,1,MPI_COMM_WORLD,
1
/tmp/cc5KpcVF.f:3624: (continued):
call MPI_RECV(VTWOE,maxinp,MPI_DOUBLE_PRECISION,isource,128,
2
Argument #1 of `mpi_recv' is one type at (2) but is some other type at
(1) [info -f g77 M GLOBALS]
/tmp/cc5KpcVF.f:3718: warning:
call MPI_SEND(maxinp,1,MPI_INTEGER,0,1,MPI_COMM_WORLD,ierr3)
1
/tmp/cc5KpcVF.f:3720: (continued):
call MPI_SEND(VTWOE,maxinp,MPI_DOUBLE_PRECISION,0,128,
2
Argument #1 of `mpi_send' is one type at (2) but is some other type at
(1) [info -f g77 M GLOBALS]
ccea_MPI.F: In subroutine `vbar':
/tmp/cc5KpcVF.f:3621: warning:
call MPI_RECV(maxinp,1,MPI_INTEGER,isource,1,MPI_COMM_WORLD,
1
/tmp/cc5KpcVF.f:5085: (continued):
call MPI_RECV(vmpi,maxinp,MPI_DOUBLE_PRECISION,isource,6,
2
Argument #1 of `mpi_recv' is one type at (2) but is some other type at
(1) [info -f g77 M GLOBALS]
/tmp/cc5KpcVF.f:3718: warning:
call MPI_SEND(maxinp,1,MPI_INTEGER,0,1,MPI_COMM_WORLD,ierr3)
1
/tmp/cc5KpcVF.f:5276: (continued):
call MPI_SEND(vmpi,maxinp,MPI_DOUBLE_PRECISION,0,6,
2
Argument #1 of `mpi_send' is one type at (2) but is some other type at
(1) [info -f g77 M GLOBALS]
ccea_MPI.o(.text+0x846): In function `MAIN__':
: relocation truncated to fit: R_X86_64_32S scratch1_
ccea_MPI.o(.text+0x85c): In function `MAIN__':
: relocation truncated to fit: R_X86_64_32S scratch1_
ccea_MPI.o(.text+0x872): In function `MAIN__':
: relocation truncated to fit: R_X86_64_32S scratch2_
ccea_MPI.o(.text+0x888): In function `MAIN__':
: relocation truncated to fit: R_X86_64_32S scratch2_
ccea_MPI.o(.text+0xa31): In function `MAIN__':
: relocation truncated to fit: R_X86_64_PC32 skip3_
ccea_MPI.o(.text+0xbaa): In function `MAIN__':
: relocation truncated to fit: R_X86_64_32S scratch1_
ccea_MPI.o(.text+0xbb3): In function `MAIN__':
: relocation truncated to fit: R_X86_64_32S scratch1_
ccea_MPI.o(.text+0xbcc): In function `MAIN__':
: relocation truncated to fit: R_X86_64_32S scratch1_
ccea_MPI.o(.text+0xfaa): In function `getvbar_':
: relocation truncated to fit: R_X86_64_32S mind_
ccea_MPI.o(.text+0x1111): In function `eamatb_':
: relocation truncated to fit: R_X86_64_PC32 info_
ccea_MPI.o(.text+0x1158): In function `eamatb_':
: additional relocation overflows omitted from the output
collect2: ld returned 1 exit status
--
Sameer Sinha
Member Technical Staff
SSDG,CDAC R&D
Bangalore-38
More information about the Gcc-help
mailing list