1.3 Magma installation with MKL

Open discussion for MAGMA

1.3 Magma installation with MKL

Postby jscott26 » Tue Jun 18, 2013 6:31 pm

Hi,

I am attempting to install Magma 1.3 with MKL 11.0 on RH Linux 6.0 machine with two GTX Titans. I need to use 64-bit integers because I will be doing Cholesky decomp on very large systems. After changing "typedef int magma_int_t" --> "typedef long magma_int_t", I recieve the following error:

connection_mgpu.cpp: In function ‘int magma_buildconnection_mgpu(magma_int_t (*)[10], magma_int_t*, magma_int_t)’:
connection_mgpu.cpp:65: error: cannot convert ‘magma_int_t*’ to ‘int*’ for argument ‘1’ to ‘cudaError_t cudaDeviceCanAccessPeer(int*, int, int)’
make[1]: *** [connection_mgpu.o] Error 1
make[1]: Leaving directory `/home/jscott/Downloads/magma-1.3.0/control'
make: *** [libmagma_control] Error 2

Here is my make.inc:

#//////////////////////////////////////////////////////////////////////////////
# -- MAGMA (version 1.3.0) --
# Univ. of Tennessee, Knoxville
# Univ. of California, Berkeley
# Univ. of Colorado, Denver
# November 2012
#//////////////////////////////////////////////////////////////////////////////

# GPU_TARGET specifies for which GPU you want to compile MAGMA:
# "Tesla" (NVIDIA compute capability 1.x cards)
# "Fermi" (NVIDIA compute capability 2.x cards)
# "Kepler" (NVIDIA compute capability 3.x cards)
# See http://developer.nvidia.com/cuda-gpus

GPU_TARGET = Kepler

CC = gcc
NVCC = nvcc
FORT = gfortran

ARCH = ar
ARCHFLAGS = cr
RANLIB = ranlib

# CUDA 5 needs -DCUBLAS_GFORTRAN or appropriate setting to compile cuda/src/fortran.c bindings
OPTS = -O3 -DADD_ -DCUBLAS_GFORTRAN -DMKL_ILP64 -m64 -I$(MKLROOT)/include
F77OPTS = -O3 -DADD_
FOPTS = -O3 -DADD_ -x f95-cpp-input
NVOPTS = -O3 -DADD_ --compiler-options -fno-strict-aliasing -DUNIX
LDOPTS = -fPIC -Xlinker -zmuldefs

# gcc with MKL 10.3, Intel threads
LIB = -L$(MKLROOT)/lib/intel64 -lmkl_intel_ilp64 -lmkl_intel_thread -lmkl_core -liomp5 -lpthread -lm -lcublas -lcudart -lm -fopenmp

# define these in your environment
# for MKL run, e.g.: /opt/sw/intel/2011.2.137/mkl/bin/mklvars.sh intel64
MKLROOT = /usr/local/composer_xe_2013.2.146/mkl
CUDADIR = /usr/local/cuda

LIBDIR = -L$(MKLROOT)/lib/intel64 \
-L$(CUDADIR)/lib64
INC = -I$(CUDADIR)/include

-----

Any help is appreciated. Thank you so much!

Best,

Justin
jscott26
 
Posts: 1
Joined: Tue Jun 18, 2013 6:18 pm

Re: 1.3 Magma installation with MKL

Postby mgates3 » Wed Jun 19, 2013 7:53 pm

That is a bug in the control/connection_mgpu.cpp file. I think you just need to change this:
magma_int_t samecomplex=-1;
to this:
int samecomplex=-1;

-mark
mgates3
 
Posts: 442
Joined: Fri Jan 06, 2012 2:13 pm


Return to User discussion

Who is online

Users browsing this forum: No registered users and 2 guests

cron