testing "magma_sgetrf_gpu" ubuntu 12.04

Open discussion for MAGMA

Re: testing "magma_sgetrf_gpu" ubuntu 12.04

Postby Ammiraglio625 » Fri Nov 15, 2013 8:19 am

Yes, they are there.

And the MAGMA_CL_DIR environment variable is set:

gabriele@gabriele-K52JU:~$ echo $MAGMA_CL_DIR
/home/gabriele/clmagma-1.0.0/interface_opencl


It is getting a nightmare :)
Ammiraglio625
 
Posts: 11
Joined: Fri Nov 01, 2013 6:46 am

Re: testing "magma_sgetrf_gpu" ubuntu 12.04

Postby mgates3 » Tue Nov 19, 2013 3:11 pm

Can you list them, just to humor me?

ls -l $MAGMA_CL_DIR/dinplace_transpose.co
ls -l $MAGMA_CL_DIR/*.co

Otherwise, I can't offer any other suggestions about why it can't find them.
-mark
mgates3
 
Posts: 442
Joined: Fri Jan 06, 2012 2:13 pm

Re: testing "magma_sgetrf_gpu" ubuntu 12.04

Postby Ammiraglio625 » Thu Nov 21, 2013 11:05 am

Sure, thanks a lot :)


gabriele@gabriele-K52JU:~$ ls -l $MAGMA_CL_DIR/*.co
-rw-rw-r-- 1 gabriele gabriele 51948 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/cauxiliary.co
-rw-rw-r-- 1 gabriele gabriele 24948 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/cinplace_transpose.co
-rw-rw-r-- 1 gabriele gabriele 10748 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/clacpy.co
-rw-rw-r-- 1 gabriele gabriele 11076 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/cpermute-v2.co
-rw-rw-r-- 1 gabriele gabriele 10156 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/cswap.co
-rw-rw-r-- 1 gabriele gabriele 12468 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/ctranspose.co
-rw-rw-r-- 1 gabriele gabriele 26344 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/ctranspose-v2.co
-rw-rw-r-- 1 gabriele gabriele 50828 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/sauxiliary.co
-rw-rw-r-- 1 gabriele gabriele 24384 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/sinplace_transpose.co
-rw-rw-r-- 1 gabriele gabriele 10708 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/slacpy.co
-rw-rw-r-- 1 gabriele gabriele 11036 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/spermute-v2.co
-rw-rw-r-- 1 gabriele gabriele 10116 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/sswap.co
-rw-rw-r-- 1 gabriele gabriele 12300 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/stranspose.co
-rw-rw-r-- 1 gabriele gabriele 26016 nov 13 15:10 /home/gabriele/clmagma-1.0.0/interface_opencl/stranspose-v2.co
Ammiraglio625
 
Posts: 11
Joined: Fri Nov 01, 2013 6:46 am

Re: testing "magma_sgetrf_gpu" ubuntu 12.04

Postby mgates3 » Thu Nov 21, 2013 3:49 pm

The d and z versions are missing, which is exactly what it says it can't find. I guess this was an error during compiling. Try running make (or make clean, then make) in the interface_opencl directory, and see if it generates those files or any errors.
-mark
mgates3
 
Posts: 442
Joined: Fri Jan 06, 2012 2:13 pm

Re: testing "magma_sgetrf_gpu" ubuntu 12.04

Postby Ammiraglio625 » Fri Nov 22, 2013 10:02 am

Yes, there are errors :(

----------------------------------------------------------------------------------------------------------------------------------------------

gabriele@gabriele-K52JU:~/clmagma-1.0.0/interface_opencl$ make
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c interface.cpp -o interface.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c strerror.cpp -o strerror.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c CL_MAGMA_RT.cpp -o CL_MAGMA_RT.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c magmablas_z.cpp -o magmablas_z.o
magmablas_z.cpp: In function ‘magma_err_t magma_zsetmatrix(magma_int_t, magma_int_t, const magmaDoubleComplex*, size_t, magma_int_t, magmaDoubleComplex_ptr, size_t, magma_int_t, magma_queue_t)’:
magmablas_z.cpp:46:68: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_z.cpp: In function ‘magma_err_t magma_zgetmatrix(magma_int_t, magma_int_t, magmaDoubleComplex_const_ptr, size_t, magma_int_t, magmaDoubleComplex*, size_t, magma_int_t, magma_queue_t)’:
magmablas_z.cpp:92:68: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_z.cpp: In function ‘magma_err_t magma_zsetmatrix_async(magma_int_t, magma_int_t, const magmaDoubleComplex*, size_t, magma_int_t, magmaDoubleComplex_ptr, size_t, magma_int_t, magma_queue_t, _cl_event**)’:
magmablas_z.cpp:190:68: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_z.cpp: In function ‘magma_err_t magma_zgetmatrix_async(magma_int_t, magma_int_t, magmaDoubleComplex_const_ptr, size_t, magma_int_t, magmaDoubleComplex*, size_t, magma_int_t, magma_queue_t, _cl_event**)’:
magmablas_z.cpp:210:68: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_z.cpp: In function ‘magma_err_t magma_zcopymatrix(magma_int_t, magma_int_t, magmaDoubleComplex_const_ptr, size_t, magma_int_t, magmaDoubleComplex_ptr, size_t, magma_int_t, magma_queue_t)’:
magmablas_z.cpp:230:68: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c zpermute.cpp -o zpermute.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c zinplace_transpose.cpp -o zinplace_transpose.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c ztranspose.cpp -o ztranspose.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c ztranspose2.cpp -o ztranspose2.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c zauxiliary.cpp -o zauxiliary.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c zlacpy.cpp -o zlacpy.o
zlacpy.cpp: In function ‘void magmablas_zlacpy(magma_uplo_t, magma_int_t, magma_int_t, magmaDoubleComplex_ptr, size_t, magma_int_t, magmaDoubleComplex_ptr, size_t, magma_int_t, magma_queue_t)’:
zlacpy.cpp:67:51: warning: narrowing conversion of ‘(((m / 64) + ((((unsigned int)m) & 63u) != 0u)) * 64)’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c zswap.cpp -o zswap.o
zswap.cpp: In function ‘void magmablas_zswap(magma_int_t, magmaDoubleComplex_ptr, size_t, magma_int_t, magmaDoubleComplex_ptr, size_t, magma_int_t, magma_queue_t)’:
zswap.cpp:40:38: warning: narrowing conversion of ‘blocksize’ from ‘int’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c magmablas_c.cpp -o magmablas_c.o
magmablas_c.cpp: In function ‘magma_err_t magma_csetmatrix(magma_int_t, magma_int_t, const magmaFloatComplex*, size_t, magma_int_t, magmaFloatComplex_ptr, size_t, magma_int_t, magma_queue_t)’:
magmablas_c.cpp:46:67: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_c.cpp: In function ‘magma_err_t magma_cgetmatrix(magma_int_t, magma_int_t, magmaFloatComplex_const_ptr, size_t, magma_int_t, magmaFloatComplex*, size_t, magma_int_t, magma_queue_t)’:
magmablas_c.cpp:92:67: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_c.cpp: In function ‘magma_err_t magma_csetmatrix_async(magma_int_t, magma_int_t, const magmaFloatComplex*, size_t, magma_int_t, magmaFloatComplex_ptr, size_t, magma_int_t, magma_queue_t, _cl_event**)’:
magmablas_c.cpp:190:67: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_c.cpp: In function ‘magma_err_t magma_cgetmatrix_async(magma_int_t, magma_int_t, magmaFloatComplex_const_ptr, size_t, magma_int_t, magmaFloatComplex*, size_t, magma_int_t, magma_queue_t, _cl_event**)’:
magmablas_c.cpp:210:67: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_c.cpp: In function ‘magma_err_t magma_ccopymatrix(magma_int_t, magma_int_t, magmaFloatComplex_const_ptr, size_t, magma_int_t, magmaFloatComplex_ptr, size_t, magma_int_t, magma_queue_t)’:
magmablas_c.cpp:230:67: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c cpermute.cpp -o cpermute.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c cinplace_transpose.cpp -o cinplace_transpose.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c ctranspose.cpp -o ctranspose.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c ctranspose2.cpp -o ctranspose2.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c cauxiliary.cpp -o cauxiliary.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c clacpy.cpp -o clacpy.o
clacpy.cpp: In function ‘void magmablas_clacpy(magma_uplo_t, magma_int_t, magma_int_t, magmaFloatComplex_ptr, size_t, magma_int_t, magmaFloatComplex_ptr, size_t, magma_int_t, magma_queue_t)’:
clacpy.cpp:67:51: warning: narrowing conversion of ‘(((m / 64) + ((((unsigned int)m) & 63u) != 0u)) * 64)’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c cswap.cpp -o cswap.o
cswap.cpp: In function ‘void magmablas_cswap(magma_int_t, magmaFloatComplex_ptr, size_t, magma_int_t, magmaFloatComplex_ptr, size_t, magma_int_t, magma_queue_t)’:
cswap.cpp:40:38: warning: narrowing conversion of ‘blocksize’ from ‘int’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c magmablas_d.cpp -o magmablas_d.o
magmablas_d.cpp: In function ‘magma_err_t magma_dsetmatrix(magma_int_t, magma_int_t, const double*, size_t, magma_int_t, magmaDouble_ptr, size_t, magma_int_t, magma_queue_t)’:
magmablas_d.cpp:46:56: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_d.cpp: In function ‘magma_err_t magma_dgetmatrix(magma_int_t, magma_int_t, magmaDouble_const_ptr, size_t, magma_int_t, double*, size_t, magma_int_t, magma_queue_t)’:
magmablas_d.cpp:92:56: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_d.cpp: In function ‘magma_err_t magma_dsetmatrix_async(magma_int_t, magma_int_t, const double*, size_t, magma_int_t, magmaDouble_ptr, size_t, magma_int_t, magma_queue_t, _cl_event**)’:
magmablas_d.cpp:190:56: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_d.cpp: In function ‘magma_err_t magma_dgetmatrix_async(magma_int_t, magma_int_t, magmaDouble_const_ptr, size_t, magma_int_t, double*, size_t, magma_int_t, magma_queue_t, _cl_event**)’:
magmablas_d.cpp:210:56: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_d.cpp: In function ‘magma_err_t magma_dcopymatrix(magma_int_t, magma_int_t, magmaDouble_const_ptr, size_t, magma_int_t, magmaDouble_ptr, size_t, magma_int_t, magma_queue_t)’:
magmablas_d.cpp:230:56: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c dpermute.cpp -o dpermute.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c dinplace_transpose.cpp -o dinplace_transpose.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c dtranspose.cpp -o dtranspose.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c dtranspose2.cpp -o dtranspose2.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c dauxiliary.cpp -o dauxiliary.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c dlacpy.cpp -o dlacpy.o
dlacpy.cpp: In function ‘void magmablas_dlacpy(magma_uplo_t, magma_int_t, magma_int_t, magmaDouble_ptr, size_t, magma_int_t, magmaDouble_ptr, size_t, magma_int_t, magma_queue_t)’:
dlacpy.cpp:67:51: warning: narrowing conversion of ‘(((m / 64) + ((((unsigned int)m) & 63u) != 0u)) * 64)’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c dswap.cpp -o dswap.o
dswap.cpp: In function ‘void magmablas_dswap(magma_int_t, magmaDouble_ptr, size_t, magma_int_t, magmaDouble_ptr, size_t, magma_int_t, magma_queue_t)’:
dswap.cpp:40:38: warning: narrowing conversion of ‘blocksize’ from ‘int’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c magmablas_s.cpp -o magmablas_s.o
magmablas_s.cpp: In function ‘magma_err_t magma_ssetmatrix(magma_int_t, magma_int_t, const float*, size_t, magma_int_t, magmaFloat_ptr, size_t, magma_int_t, magma_queue_t)’:
magmablas_s.cpp:46:55: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_s.cpp: In function ‘magma_err_t magma_sgetmatrix(magma_int_t, magma_int_t, magmaFloat_const_ptr, size_t, magma_int_t, float*, size_t, magma_int_t, magma_queue_t)’:
magmablas_s.cpp:92:55: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_s.cpp: In function ‘magma_err_t magma_ssetmatrix_async(magma_int_t, magma_int_t, const float*, size_t, magma_int_t, magmaFloat_ptr, size_t, magma_int_t, magma_queue_t, _cl_event**)’:
magmablas_s.cpp:190:55: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_s.cpp: In function ‘magma_err_t magma_sgetmatrix_async(magma_int_t, magma_int_t, magmaFloat_const_ptr, size_t, magma_int_t, float*, size_t, magma_int_t, magma_queue_t, _cl_event**)’:
magmablas_s.cpp:210:55: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
magmablas_s.cpp: In function ‘magma_err_t magma_scopymatrix(magma_int_t, magma_int_t, magmaFloat_const_ptr, size_t, magma_int_t, magmaFloat_ptr, size_t, magma_int_t, magma_queue_t)’:
magmablas_s.cpp:230:55: warning: narrowing conversion of ‘n’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c spermute.cpp -o spermute.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c sinplace_transpose.cpp -o sinplace_transpose.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c stranspose.cpp -o stranspose.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c stranspose2.cpp -o stranspose2.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c sauxiliary.cpp -o sauxiliary.o
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c slacpy.cpp -o slacpy.o
slacpy.cpp: In function ‘void magmablas_slacpy(magma_uplo_t, magma_int_t, magma_int_t, magmaFloat_ptr, size_t, magma_int_t, magmaFloat_ptr, size_t, magma_int_t, magma_queue_t)’:
slacpy.cpp:67:51: warning: narrowing conversion of ‘(((m / 64) + ((((unsigned int)m) & 63u) != 0u)) * 64)’ from ‘magma_int_t {aka int}’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c sswap.cpp -o sswap.o
sswap.cpp: In function ‘void magmablas_sswap(magma_int_t, magmaFloat_ptr, size_t, magma_int_t, magmaFloat_ptr, size_t, magma_int_t, magma_queue_t)’:
sswap.cpp:40:38: warning: narrowing conversion of ‘blocksize’ from ‘int’ to ‘size_t {aka long unsigned int}’ inside { } is ill-formed in C++11 [-Wnarrowing]
ar cr ../lib/libclmagma.a interface.o strerror.o CL_MAGMA_RT.o magmablas_z.o zpermute.o zinplace_transpose.o ztranspose.o ztranspose2.o zauxiliary.o zlacpy.o zswap.o zinplace_transpose.cl zpermute-v2.cl ztranspose.cl ztranspose-v2.cl zauxiliary.cl zlacpy.cl zswap.cl magmablas_c.o cpermute.o cinplace_transpose.o ctranspose.o ctranspose2.o cauxiliary.o clacpy.o cswap.o cinplace_transpose.cl cpermute-v2.cl ctranspose.cl ctranspose-v2.cl cauxiliary.cl clacpy.cl cswap.cl magmablas_d.o dpermute.o dinplace_transpose.o dtranspose.o dtranspose2.o dauxiliary.o dlacpy.o dswap.o dinplace_transpose.cl dpermute-v2.cl dtranspose.cl dtranspose-v2.cl dauxiliary.cl dlacpy.cl dswap.cl magmablas_s.o spermute.o sinplace_transpose.o stranspose.o stranspose2.o sauxiliary.o slacpy.o sswap.o sinplace_transpose.cl spermute-v2.cl stranspose.cl stranspose-v2.cl sauxiliary.cl slacpy.cl sswap.cl
ranlib ../lib/libclmagma.a
g++ -O0 -DADD_ -g -Wall -DHAVE_clAmdBlas -DHAVE_CBLAS -I/opt/clAmdBlas-1.10.321/include -I/opt/AMDAPP/include -I../include -I../control -c clcompile.cpp -o clcompile.o
g++ -fPIC -Xlinker -zmuldefs clcompile.o -o clcompile \
-L../lib -lclmagma -L/opt/acml5.3.0/gfortran64_int64/lib -L/opt/acml5.3.0/gfortran64_mp_int64/lib -L/usr/lib/atlas-base/lib -L/opt/clAmdBlas-1.10.321/lib64 -L/opt/intel/opencl-1.2-3.1.1.11385/lib64 -L/usr/lib/gcc/x86_64-linux-gnu/4.7 -L/usr/lib -L/usr/lib/gcc/x86_64-linux-gnu/4.6.3 -lacml -lacml_mp -lcblas -latlas -lf77blas -llapack_atlas -lclAmdBlas -lOpenCL -lgfortran -lm -lf2c -lstdc++
./clcompile sinplace_transpose.cl
Initializing...
Device: Cedar
./clcompile spermute-v2.cl
Initializing...
Device: Cedar
./clcompile stranspose.cl
Initializing...
Device: Cedar
./clcompile stranspose-v2.cl
Initializing...
Device: Cedar
./clcompile sauxiliary.cl
Initializing...
Device: Cedar
./clcompile slacpy.cl
Initializing...
Device: Cedar
./clcompile sswap.cl
Initializing...
Device: Cedar
./clcompile dinplace_transpose.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile dpermute-v2.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile dtranspose.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile dtranspose-v2.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile dauxiliary.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile dlacpy.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile dswap.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile cinplace_transpose.cl
Initializing...
Device: Cedar
./clcompile cpermute-v2.cl
Initializing...
Device: Cedar
./clcompile ctranspose.cl
Initializing...
Device: Cedar
./clcompile ctranspose-v2.cl
Initializing...
Device: Cedar
./clcompile cauxiliary.cl
Initializing...
Device: Cedar
./clcompile clacpy.cl
Initializing...
Device: Cedar
./clcompile cswap.cl
Initializing...
Device: Cedar
./clcompile zinplace_transpose.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile zpermute-v2.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile ztranspose.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile ztranspose-v2.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile zauxiliary.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile zlacpy.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp
./clcompile zswap.cl
Initializing...
Device: Cedar
clBuildProgram error at 716 in CL_MAGMA_RT.cpp


-----------------------------------------------------------------------------------------------------------------------------------------------------
Ammiraglio625
 
Posts: 11
Joined: Fri Nov 01, 2013 6:46 am

Re: testing "magma_sgetrf_gpu" ubuntu 12.04

Postby mgates3 » Mon Nov 25, 2013 2:29 pm

What GPU do you have? Does it support double precision? If not, that would explain why it can't compile those double precision and double-complex kernels. Unfortunately at the moment you will have to comment all those routines out of the CL_MAGMA_RT.cpp file. We don't have a mechanism to build only the single-precision routines.
-mark
mgates3
 
Posts: 442
Joined: Fri Jan 06, 2012 2:13 pm

Re: testing "magma_sgetrf_gpu" ubuntu 12.04

Postby Ammiraglio625 » Sun Dec 01, 2013 2:06 pm

Yep, my gpu does not support double!

Once I've solved this problem, I've still try to run testing_sgetrf_gpu.cpp, but I get this errors


-------------- Build: Debug in magmatest ---------------

Compiling: testing_sgetrf_gpu.cpp
/home/gabriele/Scrivania/TEST MAGMA/magmatest/testing_sgetrf_gpu.cpp: In function ‘int main(int, char**)’:
/home/gabriele/Scrivania/TEST MAGMA/magmatest/testing_sgetrf_gpu.cpp:148:5: error: invalid conversion from ‘float**’ to ‘void**’ [-fpermissive]
In file included from /home/gabriele/Scrivania/TEST MAGMA/magmatest/testing_sgetrf_gpu.cpp:19:0:
/opt/clmagma-1.0.0/include/magma.h:44:1: error: initializing argument 1 of ‘magma_err_t magma_malloc(void**, size_t)’ [-fpermissive]
Process terminated with status 1 (0 minutes, 0 seconds)
2 errors, 0 warnings

------------------------------------------------------------- here the code (from testing_sgetrf_gpu.cpp ) which trig the errors

OK TESTING_MALLOC( ipiv, magma_int_t, min_mn);
OK TESTING_MALLOC( h_A, float, n2 );
OK TESTING_MALLOC_HOST( h_R, float, n2 );
ERROR - >>>>>>>>>>> TESTING_MALLOC_DEV( d_A, float, ldda*N );



Thanks in advance
Ammiraglio625
 
Posts: 11
Joined: Fri Nov 01, 2013 6:46 am

Re: testing "magma_sgetrf_gpu" ubuntu 12.04

Postby mgates3 » Mon Dec 02, 2013 11:04 am

It appears you need to add a (void**) cast. Really, it would be easiest if you compiled using the provided makefiles first, to verify that everything does in fact compile. Then when you port to a different compile system, you can debug the differences between your compile system and the provided makefiles much easier. Probably your compile system is using a different compiler flag that is more strict about some C++ feature, but again without seeing the actual compile command (gcc -c -o ... ) it is impossible to tell.
-mark
mgates3
 
Posts: 442
Joined: Fri Jan 06, 2012 2:13 pm

Previous

Return to User discussion

Who is online

Users browsing this forum: Bing [Bot], Google [Bot] and 1 guest

cron