The only way that the matrix size would affect the compilation is if
somewhere you've got something telling the system how big the matrix is
going to be before it starts.
Most parallel matrix solvers read from a file, either on the head node
which reads the whole thing and then distributes the pieces, or each node
reads its part from the head or local storage. If you have swap available
on each node, you can go much bigger anyway, and a lot of parallel solvers
have an option that lets you restrict their footprint. They will take
longer, but don't swap. Which matrix solver is it? Can you build and run
with smaller test problems?
Damien
> Hi All,
> Iam running LU decomposition algorithm on a 23 node cluster. Each
> node has 2GB of RAM. When iam trying to run the algorithm for
> single precision complex matrix of size 16400*16400,iam getting
> the following compilation errors.
>
> /tmp/ccmtcg1e.o(.text+0xa1): In function `MAIN__':
> : relocation truncated to fit: R_X86_64_32S .bss
> /tmp/ccmtcg1e.o(.text+0x1b9): In function `MAIN__':
> : relocation truncated to fit: R_X86_64_32S .bss
> /tmp/ccmtcg1e.o(.text+0x25b): In function `MAIN__':
> : relocation truncated to fit: R_X86_64_32S .bss
> /tmp/ccmtcg1e.o(.text+0x2f7): In function `MAIN__':
> : relocation truncated to fit: R_X86_64_32S .bss
> /tmp/ccmtcg1e.o(.text+0x59d): In function `MAIN__':
> : relocation truncated to fit: R_X86_64_32S .bss
> /tmp/ccmtcg1e.o(.text+0x71b): In function `MAIN__':
> : relocation truncated to fit: R_X86_64_32S .bss
> /tmp/ccmtcg1e.o(.text+0x788): In function `MAIN__':
> : relocation truncated to fit: R_X86_64_32S .bss
> /usr/lib64/liblamf77mpi.a(bcast_f.o)(.text+0x1a): In function
> `mpi_bcast__':
> : relocation truncated to fit: R_X86_64_PC32 lam_F_handles
> /usr/lib64/liblamf77mpi.a(bcast_f.o)(.text+0x30): In function
> `mpi_bcast__':
> : relocation truncated to fit: R_X86_64_PC32 lam_F_handles
> /usr/lib64/liblamf77mpi.a(bcast_f.o)(.text+0x3d): In function
> `mpi_bcast__':
> : relocation truncated to fit: R_X86_64_PC32 lam_F_bottom
> /usr/lib64/liblamf77mpi.a(crank_f.o)(.text+0x12): In function
> `mpi_comm_rank__':
> : additional relocation overflows omitted from the output
> collect2: ld returned 1 exit status
> mpif77: No such file or directory
>
> I know for such a big matrix the matrix size is
> more than 2GB (16400*16400*8 Bytes = 2.15GB).
> IS this the reason for the above compilation
> errors?
>
> If this is the case,do we have any techniques
> where i can run the algorithm for matrix sizes
> more than 16400*16400. Or the cluster iam
> using is limited for certain problem size
> (16384*16384).
>
> Thanks for your help.
>
> Regards
> Srinivasa Patri
>
>
> _______________________________________________
> This list is archived at http://www.lam-mpi.org/MailArchives/lam/
>
|
