Problem building R with Intel MKL v10 BLAS

[Michael Braun]

NO

Hi.  I'm not sure if this is an R-help or R-devel problem, so I'm
starting here in the hope that someone can help (and willing to go to
the other list if it's more appropriate).  I think I am following all of
the instructions in the various manuals, but clearly I am missing something.

I have an Intel EM64T Dell with 2 dual-core Xeon processors running Red
Hat EL5.  I would like to build R 2.6.1 with lots of debugging and
profiling options, and link it to the processor-specific Intel MKL blas.
The problem is that after I compile R, and do R CMD config BLAS_LIBS,
the response is
-L/usr/local/lib64/R/lib -lRblas.

This tells me that R is not linked to the Intel BLAS at all.

My config.site file for R is:

#! /bin/sh

R_PAPERSIZE=letter
CFLAGS="-g -O2 -p -pg"
CPPFLAGS="-I/opt/intel/mkl/10.0.1.014/include -I/usr/include
-I/usr/local/include"
LIBnn=lib64
BLAS_LIBS="-L/opt/intel/mkl/10.0.1.014/lib/em64t -Wl,--start-group
-lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so -l -l -l -Wl,
--end-group -lguide -lpthread -lm"

I have set the CONFIG_SITE environment variable to the location of the
config.site.file.
I am doing everything as superuser.

The command I am using for configure is

./configure --disable-R-profiling --with-blas=no

following the instructions in the R-admin file regarding enabling
C-level profiling and linking to the external BLAS libraries referenced
in the config.site file.

The BLAS_LIBS files are different than in the R-admin manual because of
changes in the Intel MKL for version 10.  These libraries, in this
order, were taken from the Intel MKL for Linux User's Guide, chapter 5.

So, still no luck linking to the optimized BLAS.  I'd appreciate any
suggestions.

Thanks,

Michael





--
Michael Braun
Assistant Professor of Marketing
MIT Sloan School of Management
One Amherst St., E40-169
Cambridge, MA 02142
USA

[hidden email]
[phone]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Prof Brian Ripley]

On Thu, 24 Jan 2008, Michael Braun wrote:

> NO
>
> Hi.  I'm not sure if this is an R-help or R-devel problem, so I'm

Definitely R-devel.

> starting here in the hope that someone can help (and willing to go to
> the other list if it's more appropriate).  I think I am following all of
> the instructions in the various manuals, but clearly I am missing something.
>
> I have an Intel EM64T Dell with 2 dual-core Xeon processors running Red
> Hat EL5.  I would like to build R 2.6.1 with lots of debugging and
> profiling options, and link it to the processor-specific Intel MKL blas.
> The problem is that after I compile R, and do R CMD config BLAS_LIBS,
> the response is
> -L/usr/local/lib64/R/lib -lRblas.
>
> This tells me that R is not linked to the Intel BLAS at all.

Not so, because Rblas contains any links to an external BLAS.  Check over
the R-admin manual and note that --enable-BLAS-shlib is the default.
So you need to check

ldd /usr/local/lib64/R/lib/libRblas.so

to see if it linked to MKL.

Check over config.log to see what happened: a common problem is that the
libraries are not known to ld.so (you may need to include
/opt/intel/mkl/10.0.1.014/lib/em64t in LD_LIBRARY_PATH)

> My config.site file for R is:
>
> #! /bin/sh
>
> R_PAPERSIZE=letter
> CFLAGS="-g -O2 -p -pg"
> CPPFLAGS="-I/opt/intel/mkl/10.0.1.014/include -I/usr/include
> -I/usr/local/include"
> LIBnn=lib64
> BLAS_LIBS="-L/opt/intel/mkl/10.0.1.014/lib/em64t -Wl,--start-group
> -lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so -l -l -l -Wl,
> --end-group -lguide -lpthread -lm"
>
> I have set the CONFIG_SITE environment variable to the location of the
> config.site.file.
> I am doing everything as superuser.

What compilers are you using?  That would appear to be using
gcc/gfortran and not the Intel compilers.

>
> The command I am using for configure is
>
> ./configure --disable-R-profiling --with-blas=no
>
> following the instructions in the R-admin file regarding enabling
> C-level profiling and linking to the external BLAS libraries referenced
> in the config.site file.
>
> The BLAS_LIBS files are different than in the R-admin manual because of
> changes in the Intel MKL for version 10.  These libraries, in this
> order, were taken from the Intel MKL for Linux User's Guide, chapter 5.
>
> So, still no luck linking to the optimized BLAS.  I'd appreciate any
> suggestions.
>
> Thanks,
>
> Michael
>
>
>
>
>
>

--
Brian D. Ripley,                  [hidden email]
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  [phone] (self)
1 South Parks Road,                     [phone] (PA)
Oxford OX1 3TG, UK                Fax:  [phone]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Hin-Tak Leung-2]

This part "-lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so"
looks wrong - it should be "-lmkl_gf_lp64 -lmkl_gnu_thread -lmkl_core"
without the ".so" part.

I don't know how BLAS_LIBS does it, but when I was linking against
mkl 9, all I did which different from the usual build (diff from the
two rpm spec file I wrote) was 3 changes:

export LDFLAGS=...  -L/opt/intel/mkl/9.1/lib/em64t/
export FFLAGS=... -ff2c
./configure ... --with-blas="-lmkl -lguide -lpthread"

The FFLAGS ezport was needed because of difference between g77 and
gfortran.


Michael Braun wrote:

> NO
>
> Hi.  I'm not sure if this is an R-help or R-devel problem, so I'm
> starting here in the hope that someone can help (and willing to go to
> the other list if it's more appropriate).  I think I am following all of
> the instructions in the various manuals, but clearly I am missing something.
>
> I have an Intel EM64T Dell with 2 dual-core Xeon processors running Red
> Hat EL5.  I would like to build R 2.6.1 with lots of debugging and
> profiling options, and link it to the processor-specific Intel MKL blas.
> The problem is that after I compile R, and do R CMD config BLAS_LIBS,
> the response is
> -L/usr/local/lib64/R/lib -lRblas.
>
> This tells me that R is not linked to the Intel BLAS at all.
>
> My config.site file for R is:
>
> #! /bin/sh
>
> R_PAPERSIZE=letter
> CFLAGS="-g -O2 -p -pg"
> CPPFLAGS="-I/opt/intel/mkl/10.0.1.014/include -I/usr/include
> -I/usr/local/include"
> LIBnn=lib64
> BLAS_LIBS="-L/opt/intel/mkl/10.0.1.014/lib/em64t -Wl,--start-group
> -lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so -l -l -l -Wl,
> --end-group -lguide -lpthread -lm"
>
> I have set the CONFIG_SITE environment variable to the location of the
> config.site.file.
> I am doing everything as superuser.
>
> The command I am using for configure is
>
> ./configure --disable-R-profiling --with-blas=no
>
> following the instructions in the R-admin file regarding enabling
> C-level profiling and linking to the external BLAS libraries referenced
> in the config.site file.
>
> The BLAS_LIBS files are different than in the R-admin manual because of
> changes in the Intel MKL for version 10.  These libraries, in this
> order, were taken from the Intel MKL for Linux User's Guide, chapter 5.
>
> So, still no luck linking to the optimized BLAS.  I'd appreciate any
> suggestions.
>
> Thanks,
>
> Michael
>
>
>
>
>

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Michael Braun]

NO

Thanks for everyone's help.  Unfortunately, still no success.  So I took
the alternate route suggested in section A.3.1.5 of R-admin, and just
created a symbolic link from libRblas.so to .../libmkl_gf_lp64.so.  I
can still multiply 2 matrices together in R, so it looks like this is
working.  Can you propose any other test to be sure?

Should I make a similar link for LAPACK?

I'm still perplexed as to why configure couldn't find the MKL BLAS, but
I suppose any solution that works is a good one.

Thanks again.

Michael




Hin-Tak Leung wrote:

> This part "-lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so"
> looks wrong - it should be "-lmkl_gf_lp64 -lmkl_gnu_thread -lmkl_core"
> without the ".so" part.
>
> I don't know how BLAS_LIBS does it, but when I was linking against
> mkl 9, all I did which different from the usual build (diff from the
> two rpm spec file I wrote) was 3 changes:
>
> export LDFLAGS=...  -L/opt/intel/mkl/9.1/lib/em64t/
> export FFLAGS=... -ff2c
> ./configure ... --with-blas="-lmkl -lguide -lpthread"
>
> The FFLAGS ezport was needed because of difference between g77 and
> gfortran.
>
>
> Michael Braun wrote:
>> NO
>>
>> Hi.  I'm not sure if this is an R-help or R-devel problem, so I'm
>> starting here in the hope that someone can help (and willing to go to
>> the other list if it's more appropriate).  I think I am following all
>> of the instructions in the various manuals, but clearly I am missing
>> something.
>>
>> I have an Intel EM64T Dell with 2 dual-core Xeon processors running
>> Red Hat EL5.  I would like to build R 2.6.1 with lots of debugging
>> and profiling options, and link it to the processor-specific Intel
>> MKL blas. The problem is that after I compile R, and do R CMD config
>> BLAS_LIBS, the response is
>> -L/usr/local/lib64/R/lib -lRblas.
>>
>> This tells me that R is not linked to the Intel BLAS at all.
>>
>> My config.site file for R is:
>>
>> #! /bin/sh
>>
>> R_PAPERSIZE=letter
>> CFLAGS="-g -O2 -p -pg"
>> CPPFLAGS="-I/opt/intel/mkl/10.0.1.014/include -I/usr/include
>> -I/usr/local/include"
>> LIBnn=lib64
>> BLAS_LIBS="-L/opt/intel/mkl/10.0.1.014/lib/em64t -Wl,--start-group
>> -lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so -l -l -l -Wl,
>> --end-group -lguide -lpthread -lm"
>>
>> I have set the CONFIG_SITE environment variable to the location of
>> the config.site.file.
>> I am doing everything as superuser.
>>
>> The command I am using for configure is
>>
>> ./configure --disable-R-profiling --with-blas=no
>>
>> following the instructions in the R-admin file regarding enabling
>> C-level profiling and linking to the external BLAS libraries
>> referenced in the config.site file.
>>
>> The BLAS_LIBS files are different than in the R-admin manual because
>> of changes in the Intel MKL for version 10.  These libraries, in this
>> order, were taken from the Intel MKL for Linux User's Guide, chapter 5.
>>
>> So, still no luck linking to the optimized BLAS.  I'd appreciate any
>> suggestions.
>>
>> Thanks,
>>
>> Michael
>>
>>
>>
>>
>>
>

--
Michael Braun
Assistant Professor of Marketing
MIT Sloan School of Management
One Amherst St., E40-169
Cambridge, MA 02142
USA

[hidden email]
[phone]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Prof Brian Ripley]

On Fri, 25 Jan 2008, Michael Braun wrote:

> Thanks for everyone's help.  Unfortunately, still no success.  So I took the
> alternate route suggested in section A.3.1.5 of R-admin, and just created a
> symbolic link from libRblas.so to .../libmkl_gf_lp64.so.  I can still
> multiply 2 matrices together in R, so it looks like this is working.  Can you
> propose any other test to be sure?

Run 'make check' ... it includes some tests specifically for working
BLAS.

>
> Should I make a similar link for LAPACK?
>
> I'm still perplexed as to why configure couldn't find the MKL BLAS, but I
> suppose any solution that works is a good one.
>
> Thanks again.
>
> Michael
>
>
>
>
> Hin-Tak Leung wrote:
>> This part "-lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so"
>> looks wrong - it should be "-lmkl_gf_lp64 -lmkl_gnu_thread -lmkl_core"
>> without the ".so" part.
>>
>> I don't know how BLAS_LIBS does it, but when I was linking against
>> mkl 9, all I did which different from the usual build (diff from the
>> two rpm spec file I wrote) was 3 changes:
>>
>> export LDFLAGS=...  -L/opt/intel/mkl/9.1/lib/em64t/
>> export FFLAGS=... -ff2c
>> ./configure ... --with-blas="-lmkl -lguide -lpthread"
>>
>> The FFLAGS ezport was needed because of difference between g77 and
>> gfortran.
>>
>>
>> Michael Braun wrote:
>>> NO
>>>
>>> Hi.  I'm not sure if this is an R-help or R-devel problem, so I'm starting
>>> here in the hope that someone can help (and willing to go to the other
>>> list if it's more appropriate).  I think I am following all of the
>>> instructions in the various manuals, but clearly I am missing something.
>>>
>>> I have an Intel EM64T Dell with 2 dual-core Xeon processors running Red
>>> Hat EL5.  I would like to build R 2.6.1 with lots of debugging and
>>> profiling options, and link it to the processor-specific Intel MKL blas.
>>> The problem is that after I compile R, and do R CMD config BLAS_LIBS, the
>>> response is
>>> -L/usr/local/lib64/R/lib -lRblas.
>>>
>>> This tells me that R is not linked to the Intel BLAS at all.
>>>
>>> My config.site file for R is:
>>>
>>> #! /bin/sh
>>>
>>> R_PAPERSIZE=letter
>>> CFLAGS="-g -O2 -p -pg"
>>> CPPFLAGS="-I/opt/intel/mkl/10.0.1.014/include -I/usr/include
>>> -I/usr/local/include"
>>> LIBnn=lib64
>>> BLAS_LIBS="-L/opt/intel/mkl/10.0.1.014/lib/em64t -Wl,--start-group
>>> -lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so -l -l -l -Wl,
>>> --end-group -lguide -lpthread -lm"
>>>
>>> I have set the CONFIG_SITE environment variable to the location of the
>>> config.site.file.
>>> I am doing everything as superuser.
>>>
>>> The command I am using for configure is
>>>
>>> ./configure --disable-R-profiling --with-blas=no
>>>
>>> following the instructions in the R-admin file regarding enabling C-level
>>> profiling and linking to the external BLAS libraries referenced in the
>>> config.site file.
>>>
>>> The BLAS_LIBS files are different than in the R-admin manual because of
>>> changes in the Intel MKL for version 10.  These libraries, in this order,
>>> were taken from the Intel MKL for Linux User's Guide, chapter 5.
>>>
>>> So, still no luck linking to the optimized BLAS.  I'd appreciate any
>>> suggestions.
>>>
>>> Thanks,
>>>
>>> Michael
>>>
>>>
>>>
>>>
>>>
>>
>
>
> --
> Michael Braun
> Assistant Professor of Marketing
> MIT Sloan School of Management
> One Amherst St., E40-169
> Cambridge, MA 02142
> USA
>
> [hidden email]
> [phone]
>

--
Brian D. Ripley,                  [hidden email]
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  [phone] (self)
1 South Parks Road,                     [phone] (PA)
Oxford OX1 3TG, UK                Fax:  [phone]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Hin-Tak Leung-2]

Prof Brian Ripley wrote:

> On Fri, 25 Jan 2008, Michael Braun wrote:
>
>> Thanks for everyone's help.  Unfortunately, still no success.  So I
>> took the alternate route suggested in section A.3.1.5 of R-admin, and
>> just created a symbolic link from libRblas.so to
>> .../libmkl_gf_lp64.so.  I can still multiply 2 matrices together in R,
>> so it looks like this is working.  Can you propose any other test to
>> be sure?
>
> Run 'make check' ... it includes some tests specifically for working BLAS.

<snipped>

That reminds me of a discussion I had with our IT guy regarding building
R 2.5.0 against MKL 8. He had MKL 8.0 against 2.4.x previously and it is
actually faster than 9.x ; but he was having trouble upgrading to R
2.5.0 (with MKL 8).

It turns out that R 2.5.0 added a couple of checks for some
double-complex BLAS/LAPACK routines (I can't remember which), which are
missing in MKL8, so it is a fact that R 2.5.0 won't build against MKL 8
- but will build against MKL 9, and once you have built 2.5.0 against
MKL9, you can swap the libraries (somewhat dangerously) to get the speed
of MKL8 back...

It is not unthinkable that some more checks are added in 2.6.x and
MKL 10 is not passing them... one needs the result of config.log to be sure.

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Prof Brian Ripley]

On Mon, 28 Jan 2008, Hin-Tak Leung wrote:

> Prof Brian Ripley wrote:
>> On Fri, 25 Jan 2008, Michael Braun wrote:
>>
>>> Thanks for everyone's help.  Unfortunately, still no success.  So I took
>>> the alternate route suggested in section A.3.1.5 of R-admin, and just
>>> created a symbolic link from libRblas.so to .../libmkl_gf_lp64.so.  I can
>>> still multiply 2 matrices together in R, so it looks like this is working.
>>> Can you propose any other test to be sure?
>>
>> Run 'make check' ... it includes some tests specifically for working BLAS.
> <snipped>
>
> That reminds me of a discussion I had with our IT guy regarding building R
> 2.5.0 against MKL 8. He had MKL 8.0 against 2.4.x previously and it is
> actually faster than 9.x ; but he was having trouble upgrading to R
> 2.5.0 (with MKL 8).
>
> It turns out that R 2.5.0 added a couple of checks for some
> double-complex BLAS/LAPACK routines (I can't remember which), which are
> missing in MKL8, so it is a fact that R 2.5.0 won't build against MKL 8 - but
> will build against MKL 9, and once you have built 2.5.0 against MKL9, you can
> swap the libraries (somewhat dangerously) to get the speed
> of MKL8 back...
>
> It is not unthinkable that some more checks are added in 2.6.x and
> MKL 10 is not passing them... one needs the result of config.log to be sure.

Looking at the svn logs, the last change to the BLAS configure tests was
in Nov 2005.

--
Brian D. Ripley,                  [hidden email]
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  [phone] (self)
1 South Parks Road,                     [phone] (PA)
Oxford OX1 3TG, UK                Fax:  [phone]

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Hin-Tak Leung-2]

Prof Brian Ripley wrote:

> On Mon, 28 Jan 2008, Hin-Tak Leung wrote:
>
>> Prof Brian Ripley wrote:
>>> On Fri, 25 Jan 2008, Michael Braun wrote:
>>>
>>>> Thanks for everyone's help.  Unfortunately, still no success.  So I
>>>> took the alternate route suggested in section A.3.1.5 of R-admin,
>>>> and just created a symbolic link from libRblas.so to
>>>> .../libmkl_gf_lp64.so.  I can still multiply 2 matrices together in
>>>> R, so it looks like this is working. Can you propose any other test
>>>> to be sure?
>>>
>>> Run 'make check' ... it includes some tests specifically for working
>>> BLAS.
>> <snipped>
>>
>> That reminds me of a discussion I had with our IT guy regarding
>> building R 2.5.0 against MKL 8. He had MKL 8.0 against 2.4.x
>> previously and it is
>> actually faster than 9.x ; but he was having trouble upgrading to R
>> 2.5.0 (with MKL 8).
>>
>> It turns out that R 2.5.0 added a couple of checks for some
>> double-complex BLAS/LAPACK routines (I can't remember which), which
>> are missing in MKL8, so it is a fact that R 2.5.0 won't build against
>> MKL 8 - but will build against MKL 9, and once you have built 2.5.0
>> against MKL9, you can swap the libraries (somewhat dangerously) to get
>> the speed
>> of MKL8 back...
>>
>> It is not unthinkable that some more checks are added in 2.6.x and
>> MKL 10 is not passing them... one needs the result of config.log to be
>> sure.
>
> Looking at the svn logs, the last change to the BLAS configure tests was
> in Nov 2005.
>

That would be in the R 2.2 <-> R 2.3 time frame (maybe our IT guys was
having problem jumping from 2.2 to 2.5)... Anyway, I am quite sure about
the R 2.5 + MKL 8 failure as I had looked at the config.log myself then.
To see what's wrong with the original poster's set up one needs to look
at the config.log .

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Ei-ji Nakama]

Hi.

2008/1/25, Michael Braun <[hidden email]>:

> NO
>
> Hi.  I'm not sure if this is an R-help or R-devel problem, so I'm
> starting here in the hope that someone can help (and willing to go to
> the other list if it's more appropriate).  I think I am following all of
> the instructions in the various manuals, but clearly I am missing something.
>
> I have an Intel EM64T Dell with 2 dual-core Xeon processors running Red
> Hat EL5.  I would like to build R 2.6.1 with lots of debugging and
> profiling options, and link it to the processor-specific Intel MKL blas.
> The problem is that after I compile R, and do R CMD config BLAS_LIBS,
> the response is
> -L/usr/local/lib64/R/lib -lRblas.
>
> This tells me that R is not linked to the Intel BLAS at all.
>
> My config.site file for R is:
>
> #! /bin/sh
>
> R_PAPERSIZE=letter
> CFLAGS="-g -O2 -p -pg"
> CPPFLAGS="-I/opt/intel/mkl/10.0.1.014/include -I/usr/include
> -I/usr/local/include"
> LIBnn=lib64
> BLAS_LIBS="-L/opt/intel/mkl/10.0.1.014/lib/em64t -Wl,--start-group
> -lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so -l -l -l -Wl,
> --end-group -lguide -lpthread -lm"

I have an AMD 64x2 Debian(etch)
$ gcc-4.2 -v
Using built-in specs.
Target: x86_64-linux-gnu
Configured with: ../gcc-4.2.2/configure -v
--enable-languages=c,c++,fortran,objc,obj-c++,treelang
--prefix=/usr/local/gcc-4.2.2 --enable-shared --disable-multilib
--with-system-zlib --without-included-gettext --enable-threads=posix
--enable-nls --program-suffix=-4.2 --enable-__cxa_atexit
--enable-clocale=gnu --enable-libstdcxx-debug --enable-mpfr
--enable-checking=release x86_64-linux-gnu
Thread model: posix
gcc version 4.2.2

MKL_LIB_PATH=/opt/intel/mkl/10.0.1.014/lib/em64t
MKL="   -L${MKL_LIB_PATH}                               \
        -Wl,--start-group                               \
                ${MKL_LIB_PATH}/libmkl_gf_lp64.a        \
                ${MKL_LIB_PATH}/libmkl_gnu_thread.a     \
                ${MKL_LIB_PATH}/libmkl_core.a           \
        -Wl,--end-group                                 \
        -liomp5 -lguide -lpthread -lgomp"

./configure CC=gcc-4.2\
        CXX=g++-4.2\
        F77=gfortran-4.2\
        FC=gfortran-4.2\
        --with-lapack="$MKL" --with-blas="$MKL"

mkl_core seemed to want to cause libiomp5. dgemm gave a funny result
in matrix of 1000x1000 if I did not link with real libiomp5.

--
EI-JI Nakama  <[hidden email]>
"\u4e2d\u9593\u6804\u6cbb"  <[hidden email]>

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Michael Redmond]

Ei-ji Nakama <nakama <at> ki.rim.or.jp> writes:

>
> Hi.
>
> 2008/1/25, Michael Braun <braunm <at> mit.edu>:
> > NO
> >
> > Hi.  I'm not sure if this is an R-help or R-devel problem, so I'm
> > starting here in the hope that someone can help (and willing to go to
> > the other list if it's more appropriate).  I think I am following all of
> > the instructions in the various manuals, but clearly I am missing something.
> >
> > I have an Intel EM64T Dell with 2 dual-core Xeon processors running Red
> > Hat EL5.  I would like to build R 2.6.1 with lots of debugging and
> > profiling options, and link it to the processor-specific Intel MKL blas.
> > The problem is that after I compile R, and do R CMD config BLAS_LIBS,
> > the response is
> > -L/usr/local/lib64/R/lib -lRblas.
> >
> > This tells me that R is not linked to the Intel BLAS at all.
> >
> > My config.site file for R is:
> >
> > #! /bin/sh
> >
> > R_PAPERSIZE=letter
> > CFLAGS="-g -O2 -p -pg"
> > CPPFLAGS="-I/opt/intel/mkl/10.0.1.014/include -I/usr/include
> > -I/usr/local/include"
> > LIBnn=lib64
> > BLAS_LIBS="-L/opt/intel/mkl/10.0.1.014/lib/em64t -Wl,--start-group
> > -lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so -l -l -l -Wl,
> > --end-group -lguide -lpthread -lm"
>
> I have an AMD 64x2 Debian(etch)
> $ gcc-4.2 -v
> Using built-in specs.
> Target: x86_64-linux-gnu
> Configured with: ../gcc-4.2.2/configure -v
> --enable-languages=c,c++,fortran,objc,obj-c++,treelang
> --prefix=/usr/local/gcc-4.2.2 --enable-shared --disable-multilib
> --with-system-zlib --without-included-gettext --enable-threads=posix
> --enable-nls --program-suffix=-4.2 --enable-__cxa_atexit
> --enable-clocale=gnu --enable-libstdcxx-debug --enable-mpfr
> --enable-checking=release x86_64-linux-gnu
> Thread model: posix
> gcc version 4.2.2
>
> MKL_LIB_PATH=/opt/intel/mkl/10.0.1.014/lib/em64t
> MKL="   -L${MKL_LIB_PATH}                               \
>         -Wl,--start-group                               \
>                 ${MKL_LIB_PATH}/libmkl_gf_lp64.a        \
>                 ${MKL_LIB_PATH}/libmkl_gnu_thread.a     \
>                 ${MKL_LIB_PATH}/libmkl_core.a           \
>         -Wl,--end-group                                 \
>         -liomp5 -lguide -lpthread -lgomp"
>
> ./configure CC=gcc-4.2\
>         CXX=g++-4.2\
>         F77=gfortran-4.2\
>         FC=gfortran-4.2\
>         --with-lapack="$MKL" --with-blas="$MKL"
>
> mkl_core seemed to want to cause libiomp5. dgemm gave a funny result
> in matrix of 1000x1000 if I did not link with real libiomp5.
>

---
According to Intel, the V10 MKL is modular and therefore requires that all the
needed modules be on the library path in the correct order. This differs from V9
in that the single BLAS and LAPACK libraries could be directly substituted
(linked from) libRblas.so and libRlapack.so. To build, I needed to add the paths
into the  config.sh and make sure all the tests worked during R build. These are
the config.sh entries. I think the result is similar to what is noted above,
though I use the intel thread rather than the gnu thread. I tried both, and
don't recall that it made a difference:

--with-blas="-L/.../intel/mkl/10.0.1.014/lib/em64t -lRblas -lmkl_intel_thread
-lguide -lmkl_lapack -lmkl_core -lpthread" \
--with-lapack="-L/.../intel/mkl/10.0.1.014/lib/em64t -lRlapack
-lmkl_intel_thread -lguide -lmkl_lapack -lmkl_core -lpthread" \

where ... is the absolute path to your Intel MKL installation.

Hope that helps...

MJR

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Michael Redmond]

Michael Braun <braunm <at> MIT.EDU> writes:

>
> NO
>
> Hi.  I'm not sure if this is an R-help or R-devel problem, so I'm
> starting here in the hope that someone can help (and willing to go to
> the other list if it's more appropriate).  I think I am following all of
> the instructions in the various manuals, but clearly I am missing something.
>
> I have an Intel EM64T Dell with 2 dual-core Xeon processors running Red
> Hat EL5.  I would like to build R 2.6.1 with lots of debugging and
> profiling options, and link it to the processor-specific Intel MKL blas.
> The problem is that after I compile R, and do R CMD config BLAS_LIBS,
> the response is
> -L/usr/local/lib64/R/lib -lRblas.
>
> This tells me that R is not linked to the Intel BLAS at all.
>
> My config.site file for R is:
>
> #! /bin/sh
>
> R_PAPERSIZE=letter
> CFLAGS="-g -O2 -p -pg"
> CPPFLAGS="-I/opt/intel/mkl/10.0.1.014/include -I/usr/include
> -I/usr/local/include"
> LIBnn=lib64
> BLAS_LIBS="-L/opt/intel/mkl/10.0.1.014/lib/em64t -Wl,--start-group
> -lmkl_gf_lp64.so -lmkl_gnu_thread.so -lmkl_core.so -l -l -l -Wl,
> --end-group -lguide -lpthread -lm"
>
> I have set the CONFIG_SITE environment variable to the location of the
> config.site.file.
> I am doing everything as superuser.
>
> The command I am using for configure is
>
> ./configure --disable-R-profiling --with-blas=no
>
> following the instructions in the R-admin file regarding enabling
> C-level profiling and linking to the external BLAS libraries referenced
> in the config.site file.
>
> The BLAS_LIBS files are different than in the R-admin manual because of
> changes in the Intel MKL for version 10.  These libraries, in this
> order, were taken from the Intel MKL for Linux User's Guide, chapter 5.
>
> So, still no luck linking to the optimized BLAS.  I'd appreciate any
> suggestions.
>
> Thanks,
>
> Michael
>

---
I now recall a similar problem with my install of MKL V10. Here is the info:

I got the 10.X, layered library version of MKL 10.X working on R-2.6.1a. I
set the full path information for the R configuration blas and lapack
option:

--with-blas="-L/.../intel/mkl/10.0.011/lib/em64t -lRblas -lmkl_sequential
-lmkl_lapack -lmkl_core -lpthread"
--with-lapack="-L/.../intel/mkl/10.0.011/lib/em64t -lRlapack -lmkl_sequential
-lmkl_lapack -lmkl_core -lpthread"

In this case, libRblas.so and libRlapack.so are both linked to
libmkl_gf_lp64.so in the em64t directory. There must be some internal error
in the R configuration and make processing, because it would not pick up the
primary path specification for the mkl libraries location. Instead, I was
able to define:

*******critical step to get around a possible R bug*******
sh-3.00$ export LD_RUN_PATH=/.../intel/mkl/10.0.011/lib/em64t

After that, I was able to perform the R configure with no errors. Then the
make also worked with no errors. The BLAS_LIBS and LAPACK_LIBS variables
were properly defined, and the path to the MKL libraries was properly
defined in ldpaths:

sh-3.00$ /.../R-2.6.1a/bin/R CMD config BLAS_LIBS
-L/.../intel/mkl/10.0.011/lib/em64t -lRblas -lmkl_sequential -lmkl_lapack
-lmkl_core -lpthread
sh-3.00$ /.../R-2.6.1a/bin/R CMD config LAPACK_LIBS
-L/.../R-2.6.1a/intel/mkl/10.0.011/lib/em64t -lRlapack -lmkl_sequential
-lmkl_lapack -lmkl_core -lpthread
sh-3.00$

power02(43)% more ldpaths
...
${R_LD_LIBRARY_PATH=${R_HOME}/lib:/s/gcc-4.2.1/lib64:
/.../intel/mkl/10.0.011/lib/em64t:/usr/X11R6/lib64:
/.../tcl-8.4.2/i386_rh72/lib:/.../XFree86-4.2.1/i386_rh72/lib}

(... are the various local paths to the objects of interest)

Hope that helps (maybe more than my other post)
MJR

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Ei-ji Nakama]

hi.

2008/2/5, M Redmond <[hidden email]>:
> --with-blas="-L/.../intel/mkl/10.0.1.014/lib/em64t -lRblas -lmkl_intel_thread
> -lguide -lmkl_lapack -lmkl_core -lpthread" \
> --with-lapack="-L/.../intel/mkl/10.0.1.014/lib/em64t -lRlapack
> -lmkl_intel_thread -lguide -lmkl_lapack -lmkl_core -lpthread" \

Did libRblas and libRlapack make it with manual operation?

I see.../opt/intel/mkl/10.0.1.014/doc/userguide.pdf(Building Custom
Shared Objects)

awk '/F77_SYMBOL\([a-z]+\)\(\)/{FS="[()\t]";if(length($2)){print $2
"_"}}' m4/R.m4 > blas_list
MKL_ROOT=/opt/intel/mkl/10.0.1.014
make -f ${MKL_ROOT}/tools/builder/makefile \
            em64t export=blas_list xerbra=xerbla.o \
            name=/usr/local/lib/libmkl_myblas MKL_ROOT=${MKL_ROOT}

However, zdotu did not move even if I linked at this point in time....

--
EI-JI Nakama  <[hidden email]>
"\u4e2d\u9593\u6804\u6cbb"  <[hidden email]>

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel

[Ei-ji Nakama]

hi.

2008/2/5, M Redmond <[hidden email]>:
> --with-blas="-L/.../intel/mkl/10.0.1.014/lib/em64t -lRblas -lmkl_intel_thread
> -lguide -lmkl_lapack -lmkl_core -lpthread" \
> --with-lapack="-L/.../intel/mkl/10.0.1.014/lib/em64t -lRlapack
> -lmkl_intel_thread -lguide -lmkl_lapack -lmkl_core -lpthread" \

Did libRblas and libRlapack make it with manual operation?

I see.../opt/intel/mkl/10.0.1.014/doc/userguide.pdf(Building Custom
Shared Objects)

awk '/F77_SYMBOL\([a-z]+\)\(\)/{FS="[()\t]";if(length($2)){print $2
"_"}}' m4/R.m4 > blas_list
MKL_ROOT=/opt/intel/mkl/10.0.1.014
make -f ${MKL_ROOT}/tools/builder/makefile \
            em64t export=blas_list xerbra=xerbla.o \
            name=/usr/local/lib/libmkl_myblas MKL_ROOT=${MKL_ROOT}

However, zdotu did not move even if I linked at this point in time....

--
EI-JI Nakama  <[hidden email]>
"\u4e2d\u9593\u6804\u6cbb"  <[hidden email]>

______________________________________________
[hidden email] mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel