New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Install/Make problem #179
Comments
Hi @deakkon , Sorry, this is a bug in sgemv_t.s. I have fixed it on develop branch. You can try it. I think your virtual machin is 64-bit Linux.
Meanwhile, I think OpenBLAS supports your CPU. Thank you Xianyi |
Hi, thank you for the fast response. I downloaded the develop branch in a zip file. Running make 2>&1 | tee log produces an error once again. Debug output follows. nm: '/lib/libpthread.so*': No such file gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=strsm_kernel_LN -DASMFNAME=strsm_kernel_LN_ -DNAME=strsm_kernel_LN_ -DCNAME=strsm_kernel_LN -DCHAR_NAME="strsm_kernel_LN_" -DCHAR_CNAME="strsm_kernel_LN" -I.. -UDOUBLE -UCOMPLEX -DTRSMKERNEL -UCOMPLEX -UDOUBLE -DUPPER -DLN -UCONJ ../kernel/x86_64/ -o strsm_kernel_LN.o gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=cgemm_kernel_l -DASMFNAME=cgemm_kernel_l_ -DNAME=cgemm_kernel_l_ -DCNAME=cgemm_kernel_l -DCHAR_NAME="cgemm_kernel_l_" -DCHAR_CNAME="cgemm_kernel_l" -I.. -UDOUBLE -DCOMPLEX -c -UDOUBLE -DCOMPLEX -DCN ../kernel/x86_64/ -o cgemm_kernel_l.o gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=ctrsm_kernel_LN -DASMFNAME=ctrsm_kernel_LN_ -DNAME=ctrsm_kernel_LN_ -DCNAME=ctrsm_kernel_LN -DCHAR_NAME="ctrsm_kernel_LN_" -DCHAR_CNAME="ctrsm_kernel_LN" -I.. -UDOUBLE -DCOMPLEX -DTRSMKERNEL -DCOMPLEX -UDOUBLE -DUPPER -DLN -UCONJ ../kernel/x86_64/ -o ctrsm_kernel_LN.o gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=cgemm3m_kernel -DASMFNAME=cgemm3m_kernel_ -DNAME=cgemm3m_kernel_ -DCNAME=cgemm3m_kernel -DCHAR_NAME="cgemm3m_kernel_" -DCHAR_CNAME="cgemm3m_kernel" -I.. -UDOUBLE -DCOMPLEX -c -UDOUBLE -DCOMPLEX -DNN ../kernel/x86_64/ -o cgemm3m_kernel.o |
Hi @deakkon , Please try, make clean Then, provide these files, nohup.out, config.h and Makefile.conf. You can upload the file to https://gist.github.com/ Xianyi |
Hi Xianyi, I made a public gist, per your instructions, available at https://gist.github.com/4529925 Best, |
Hi J, Because your CPU is very old, OpenBLAS/GotoBLAS didn't support in on x86_64. OpenBLAS/GotoBLAS supports your CPU on x86 32bit. make TARGET=BANIS BINARY=32 Xianyi |
Hi, once again, it failed to build anything and exited with an error. here are the 3 specified files, made after running make TARGET=BANIS BINARY=32 in OpenBLAS dir. https://gist.github.com/4530790 I think the servers area few years old but they are not that old. Could it be that it's a vritual Ubuntu system I am trying to install the files? J. |
Hi @deakkon , Could you install 32-bit Linux? Or install gcc multilib as following
|
Yep, installed both libs (gcc-multilib, gfortran-multilib) and reran I looks like all went well this time. Thanks for the helps along the way. OpenBLAS build complete. OS ... Linux To install the library, you can run "make PREFIX=/path/to/your/installation install". |
Hi, If you want to link 32bit OpenBLAS on 64bit Linux, please use -m32 option with gcc & gfortran.
|
Hi, like I said, I managed to install OpenBLAS but can't seem to be able to link it to numpy, following these instructions http://osdf.github.com/blog/numpyscipy-with-openblas-for-ubuntu-1204-second-try.html I was able to build numpy from source but #imports #logging #!/usr/bin/env python try: print "version:", numpy.version x = numpy.random.random((1000,1000)) setup = "import numpy; x = numpy.random.random((1000,1000))" t = timeit.Timer("numpy.dot(x, x.T)", setup=setup) returns "slow blas" meaning that _dotblas was not compiled/installed (as far as I understood it, without _dotblas the blas lib's don't actually work to their fullest potential; or am i wrong?) Not really sure how to link 32bit OpenBLAS on 64bit Linux; can you please elaborate? Are there any instructions that you can recommend for linking OpenBLAS with numpy/scipy with _dotblas support? Man, this is really taking quite a lot of time. I think i managed to get to the end of the Internet regarding this issue :D Anyway, thank you once again for your help! Best, |
Hi @deakkon , After google, I found http://stackoverflow.com/questions/10478802/compiling-32-bit-python-modules-e-g-numpy-on-64-bit-linux I think you can set CFLAGS=-m32 LDFLAGS=-m32. You may need install 32bit python. Xianyi |
Numpy (and ScPy) both build nicely with OpenBLAS. (When you build a combinded BLAS/LAPACK) All tests do pass, however OpenBLAS does not optimize/thread LAPACK functions (e.g. eigenvalue decomp.). The BLAS part is very fast with OpenBLAS. |
http://stackoverflow.com/a/10537366 Could this be a nice workaround? |
Hi @deakkon , Sorry, I didn't try it. I don't know. Xianyi |
In the end I solved the problem (as building/linking/... was unsuccessful in one of the stages of the process) by installing EPD that has Intel MKL support embedded. I am sorry that i didn't manage to try out OpenBLAS so far but I will definitely give it a try once I have a bit more time on my hands to play around with different setups. |
I also encountered the same question " ar: sgemm_kernel.o: No such file or directory", if there any final solution now? thank you very much |
Hi @cxfneo , Sorry for the delay. What's your OS, CPU, and OpenBLAS version? Did you use virtual machine? Xianyi |
I am facing the exact same problem, see hashdist/hashstack#204. It's a virtual machine on cloud.sagemath.org.
(It has 10 processors, I only pasted the first one.)
So following the advice from this issue, I used the
and it failed immediately with this output: https://gist.github.com/certik/9955293 I used openblas 0.2.8. |
Hi @certik , This error related to the mismatch between You need to set TARGET explicitly as following.
|
@xianyi, I just tried it and it compiles without an error! Thanks a lot for this tip and thanks for your help. I really appreciate it. I wouldn't be able to figure this out easily. |
Hi,
I am trying to install OpenBLAS (the last version available through git) but I can't seem to pull it off.
The environment is a virtual machine withUbuntu 12.04 with Intel based cpu, family : 6, model: 13 (Dothan, based on http://en.wikipedia.org/wiki/List_of_Intel_microprocessors). The processor version (Dothan) is not on Supported List in TargetList.txt. The Ubuntu installed in these virtual machines is 12.04 (32bit version, at least that's what I have been told)
root@shroomz:/usr/local/lib/OpenBLAS# lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit -> not really sure which one I have
Byte Order: Little Endian
CPU(s): 4
On-line CPU(s) list: 0-3
Thread(s) per core: 1
Core(s) per socket: 1
Socket(s): 4
NUMA node(s): 1
Vendor ID: GenuineIntel
CPU family: 6
Model: 13
Stepping: 3
CPU MHz: 1995.053
BogoMIPS: 3990.10
Hypervisor vendor: KVM
Virtualization type: full
L1d cache: 32K
L1i cache: 32K
L2 cache: 4096K
NUMA node0 CPU(s): 0-3
Version checkup:
root@shroomz:/usr/local/lib/OpenBLAS# lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 12.04.1 LTS
Release: 12.04
Codename: precise
root@shroomz:/usr/local/lib/OpenBLAS# uname -a
Linux shroomz 3.2.0-35-generic #55-Ubuntu SMP Wed Dec 5 17:42:16 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
The set of errors I get when running make from root OpenBLAS dir:
root@shroomz:/usr/local/lib/OpenBLAS# make
nm: '/lib/libpthread.so*': No such file
getarch_2nd.c: In function âmainâ:
getarch_2nd.c:37:2: warning: format â%ldâ expects argument of type âlong intâ, but argument 2 has type âintâ [-Wformat]
ln -fs libopenblas_baniasp-r0.2.5.a libopenblas.a
for d in interface driver/level2 driver/level3 driver/others kernel ;
do if test -d $d; then
make -j 4 -C $d libs || exit 1 ;
fi;
done
make[1]: Entering directory `/usr/local/lib/OpenBLAS/interface'
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=saxpy -DASMFNAME=saxpy_ -DNAME=saxpy_ -DCNAME=saxpy -DCHAR_NAME="saxpy_" -DCHAR_CNAME="saxpy" -I.. -UDOUBLE -UCOMPLEX -c axpy.c -o saxpy.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sswap -DASMFNAME=sswap_ -DNAME=sswap_ -DCNAME=sswap -DCHAR_NAME="sswap_" -DCHAR_CNAME="sswap" -I.. -UDOUBLE -UCOMPLEX -c swap.c -o sswap.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=scopy -DASMFNAME=scopy_ -DNAME=scopy_ -DCNAME=scopy -DCHAR_NAME="scopy_" -DCHAR_CNAME="scopy" -I.. -UDOUBLE -UCOMPLEX -c copy.c -o scopy.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sscal -DASMFNAME=sscal_ -DNAME=sscal_ -DCNAME=sscal -DCHAR_NAME="sscal_" -DCHAR_CNAME="sscal" -I.. -UDOUBLE -UCOMPLEX -c scal.c -o sscal.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sdot -DASMFNAME=sdot_ -DNAME=sdot_ -DCNAME=sdot -DCHAR_NAME="sdot_" -DCHAR_CNAME="sdot" -I.. -UDOUBLE -UCOMPLEX -c dot.c -o sdot.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sdsdot -DASMFNAME=sdsdot_ -DNAME=sdsdot_ -DCNAME=sdsdot -DCHAR_NAME="sdsdot_" -DCHAR_CNAME="sdsdot" -I.. -UDOUBLE -UCOMPLEX -c sdsdot.c -o sdsdot.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=dsdot -DASMFNAME=dsdot_ -DNAME=dsdot_ -DCNAME=dsdot -DCHAR_NAME="dsdot_" -DCHAR_CNAME="dsdot" -I.. -UDOUBLE -UCOMPLEX -c dsdot.c -o dsdot.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sasum -DASMFNAME=sasum_ -DNAME=sasum_ -DCNAME=sasum -DCHAR_NAME="sasum_" -DCHAR_CNAME="sasum" -I.. -UDOUBLE -UCOMPLEX -c asum.c -o sasum.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=snrm2 -DASMFNAME=snrm2_ -DNAME=snrm2_ -DCNAME=snrm2 -DCHAR_NAME="snrm2_" -DCHAR_CNAME="snrm2" -I.. -UDOUBLE -UCOMPLEX -c nrm2.c -o snrm2.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=smax -DASMFNAME=smax_ -DNAME=smax_ -DCNAME=smax -DCHAR_NAME="smax_" -DCHAR_CNAME="smax" -I.. -UDOUBLE -UCOMPLEX -c -UUSE_ABS -UUSE_MIN max.c -o smax.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=samax -DASMFNAME=samax_ -DNAME=samax_ -DCNAME=samax -DCHAR_NAME="samax_" -DCHAR_CNAME="samax" -I.. -UDOUBLE -UCOMPLEX -c -DUSE_ABS -UUSE_MIN max.c -o samax.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=ismax -DASMFNAME=ismax_ -DNAME=ismax_ -DCNAME=ismax -DCHAR_NAME="ismax_" -DCHAR_CNAME="ismax" -I.. -UDOUBLE -UCOMPLEX -c -UUSE_ABS -UUSE_MIN imax.c -o ismax.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=isamax -DASMFNAME=isamax_ -DNAME=isamax_ -DCNAME=isamax -DCHAR_NAME="isamax_" -DCHAR_CNAME="isamax" -I.. -UDOUBLE -UCOMPLEX -c -DUSE_ABS -UUSE_MIN imax.c -o isamax.o
followd by a lot of similar output and to be finished with
make[1]: Leaving directory
/usr/local/lib/OpenBLAS/driver/others' make[1]: Entering directory
/usr/local/lib/OpenBLAS/kernel'gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=samax_k -DASMFNAME=samax_k_ -DNAME=samax_k_ -DCNAME=samax_k -DCHAR_NAME="samax_k_" -DCHAR_CNAME="samax_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -DUSE_ABS -UUSE_MIN ../kernel/x86_64/amax_sse.S -o samax_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=samin_k -DASMFNAME=samin_k_ -DNAME=samin_k_ -DCNAME=samin_k -DCHAR_NAME="samin_k_" -DCHAR_CNAME="samin_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -DUSE_ABS -DUSE_MIN ../kernel/x86_64/amax_sse.S -o samin_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=smax_k -DASMFNAME=smax_k_ -DNAME=smax_k_ -DCNAME=smax_k -DCHAR_NAME="smax_k_" -DCHAR_CNAME="smax_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -UUSE_ABS -UUSE_MIN ../kernel/x86_64/amax_sse.S -o smax_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=smin_k -DASMFNAME=smin_k_ -DNAME=smin_k_ -DCNAME=smin_k -DCHAR_NAME="smin_k_" -DCHAR_CNAME="smin_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -UUSE_ABS -DUSE_MIN ../kernel/x86_64/amax_sse.S -o smin_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=isamax_k -DASMFNAME=isamax_k_ -DNAME=isamax_k_ -DCNAME=isamax_k -DCHAR_NAME="isamax_k_" -DCHAR_CNAME="isamax_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -DUSE_ABS -UUSE_MIN ../kernel/x86_64/iamax_sse.S -o isamax_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=isamin_k -DASMFNAME=isamin_k_ -DNAME=isamin_k_ -DCNAME=isamin_k -DCHAR_NAME="isamin_k_" -DCHAR_CNAME="isamin_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -DUSE_ABS -DUSE_MIN ../kernel/x86_64/iamax_sse.S -o isamin_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=ismax_k -DASMFNAME=ismax_k_ -DNAME=ismax_k_ -DCNAME=ismax_k -DCHAR_NAME="ismax_k_" -DCHAR_CNAME="ismax_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -UUSE_ABS -UUSE_MIN ../kernel/x86_64/iamax_sse.S -o ismax_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=ismin_k -DASMFNAME=ismin_k_ -DNAME=ismin_k_ -DCNAME=ismin_k -DCHAR_NAME="ismin_k_" -DCHAR_CNAME="ismin_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -UUSE_ABS -DUSE_MIN ../kernel/x86_64/iamax_sse.S -o ismin_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sasum_k -DASMFNAME=sasum_k_ -DNAME=sasum_k_ -DCNAME=sasum_k -DCHAR_NAME="sasum_k_" -DCHAR_CNAME="sasum_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE ../kernel/x86_64/asum_sse.S -o sasum_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=saxpy_k -DASMFNAME=saxpy_k_ -DNAME=saxpy_k_ -DCNAME=saxpy_k -DCHAR_NAME="saxpy_k_" -DCHAR_CNAME="saxpy_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UCOMPLEX -UDOUBLE ../kernel/x86_64/axpy_sse.S -o saxpy_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=scopy_k -DASMFNAME=scopy_k_ -DNAME=scopy_k_ -DCNAME=scopy_k -DCHAR_NAME="scopy_k_" -DCHAR_CNAME="scopy_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -UCOMPLEX -DC_INTERFACE ../kernel/x86_64/copy_sse.S -o scopy_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sdot_k -DASMFNAME=sdot_k_ -DNAME=sdot_k_ -DCNAME=sdot_k -DCHAR_NAME="sdot_k_" -DCHAR_CNAME="sdot_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE ../kernel/x86_64/dot_sse.S -o sdot_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sdsdot_k -DASMFNAME=sdsdot_k_ -DNAME=sdsdot_k_ -DCNAME=sdsdot_k -DCHAR_NAME="sdsdot_k_" -DCHAR_CNAME="sdsdot_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE ../kernel/x86_64/dot_sse.S -o sdsdot_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=dsdot_k -DASMFNAME=dsdot_k_ -DNAME=dsdot_k_ -DCNAME=dsdot_k -DCHAR_NAME="dsdot_k_" -DCHAR_CNAME="dsdot_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -DDSDOT ../kernel/x86_64/dot_sse.S -o dsdot_k.o
gcc -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=snrm2_k -DASMFNAME=snrm2_k_ -DNAME=snrm2_k_ -DCNAME=snrm2_k -DCHAR_NAME="snrm2_k_" -DCHAR_CNAME="snrm2_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -c -UDOUBLE ../kernel/x86_64/nrm2_sse.S -o snrm2_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=srot_k -DASMFNAME=srot_k_ -DNAME=srot_k_ -DCNAME=srot_k -DCHAR_NAME="srot_k_" -DCHAR_CNAME="srot_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UCOMPLEX -UDOUBLE ../kernel/x86_64/rot_sse.S -o srot_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sscal_k -DASMFNAME=sscal_k_ -DNAME=sscal_k_ -DCNAME=sscal_k -DCHAR_NAME="sscal_k_" -DCHAR_CNAME="sscal_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE ../kernel/x86_64/scal_sse.S -o sscal_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sswap_k -DASMFNAME=sswap_k_ -DNAME=sswap_k_ -DCNAME=sswap_k -DCHAR_NAME="sswap_k_" -DCHAR_CNAME="sswap_k" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE ../kernel/x86_64/swap_sse.S -o sswap_k.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sgemv_n -DASMFNAME=sgemv_n_ -DNAME=sgemv_n_ -DCNAME=sgemv_n -DCHAR_NAME="sgemv_n_" -DCHAR_CNAME="sgemv_n" -I.. -UDOUBLE -UCOMPLEX -UDOUBLE -UCOMPLEX -UTRANS ../kernel/x86_64/sgemv_n.S -o sgemv_n.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=sgemv_t -DASMFNAME=sgemv_t_ -DNAME=sgemv_t_ -DCNAME=sgemv_t -DCHAR_NAME="sgemv_t_" -DCHAR_CNAME="sgemv_t" -I.. -UDOUBLE -UCOMPLEX -UDOUBLE -UCOMPLEX -DTRANS ../kernel/x86_64/sgemv_t.S -o sgemv_t.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=ssymv_U -DASMFNAME=ssymv_U_ -DNAME=ssymv_U_ -DCNAME=ssymv_U -DCHAR_NAME="ssymv_U_" -DCHAR_CNAME="ssymv_U" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -ULOWER ../kernel/x86_64/symv_U_sse.S -o ssymv_U.o
gcc -c -O2 -DEXPRECISION -m128bit-long-double -Wall -m64 -DF_INTERFACE_GFORT -fPIC -DNO_LAPACK -DNO_LAPACKE -DSMP_SERVER -DMAX_CPU_NUMBER=4 -DASMNAME=ssymv_L -DASMFNAME=ssymv_L_ -DNAME=ssymv_L_ -DCNAME=ssymv_L -DCHAR_NAME="ssymv_L_" -DCHAR_CNAME="ssymv_L" -I.. -UDOUBLE -UCOMPLEX -UCOMPLEX -UDOUBLE -DLOWER ../kernel/x86_64/symv_L_sse.S -o ssymv_L.o
../kernel/x86_64/sgemv_t.S: Assembler messages:
../kernel/x86_64/sgemv_t.S:1: Error: junk at end of line, first unrecognized character is
2' ../kernel/x86_64/sgemv_t.S:2: Error: junk at end of line, first unrecognized character is
1'make[1]: *** [sgemv_t.o] Error 1
make[1]: *** Waiting for unfinished jobs....
make[1]: Leaving directory `/usr/local/lib/OpenBLAS/kernel'
make: *** [libs] Error 1
root@shroomz:/usr/local/lib/OpenBLAS#
I have tired different approaches to this:
https://github.com/xianyi/OpenBLAS/blob/develop/GotoBLAS_02QuickInstall.txt
http://osdf.github.com/blog/numpyscipy-with-openblas-for-ubuntu-1204.html
http://osdf.github.com/blog/numpyscipy-with-openblas-for-ubuntu-1204-second-try.html
and some that i cant even remember.
I am trying to include OpenBLAS to work with numpy/scipy as I am implementing gensim funcionality for my PhD data analysis and these libraries speed up the calculation up to 15x (based on gensim docs).
Any suggestion woul be appreciated; is it bacause of unsupported CPU, is there something wrong with the config directions or something else...
Thank you in advance for your answer. I have been googleing it for the last 2 days and have by now run out of other options.
Best,
Jurica
UPDATE:
Managed to properly install numpy after following http://stackoverflow.com/questions/12249089/how-to-use-numpy-with-openblas-instead-of-atlas-in-ubuntu
print numpy.show_config() returns:
So i guess I have BLAS support for numpy but that's not the library I was aiming for (I would like to try it with OpenBLAS). Since I can't make/compile OpenBLAS from source I am not sure how to proceed.
The text was updated successfully, but these errors were encountered: