Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-11-07 Thread Tru Huynh
On Mon, Nov 06, 2017 at 04:41:08PM -0800, Guyen Gn wrote:
> Getting back to this thread after a while, is this the name of package?:
> rpmdevtools.noarch 8.3-5.el7   Server
> Red Hat Enterprise Linux Server release 7.3 (Maipo)
> 
> I see it is included in the RH7.3

That's not the devtoolset...

> Thanks.,
> 
AFAIK, RHEL is now as 7.4 unless you have an specific subscription to
stay on a point release.

You can get newer GCC toolchain with the Red Hat Developer Toolset 
https://developers.redhat.com/products/developertoolset/overview/

The latest version devtoolset-7-toolchain provides gcc version 7.2.
https://www.softwarecollections.org/en/scls/rhscl/devtoolset-7/

RHEL:
sudo yum-config-manager --enable rhel-server-rhscl-7-rpms
sudo yum install devtoolset-7

CentOS:
yum install centos-release-scl-rh && yum install devtoolset-7

of course, you can try devtoolset-4/devtoolset-6 to match your needs.

Cheers

Tru
-- 
Dr Tru Huynh | mailto:t...@pasteur.fr | tel/fax +33 1 45 68 87 37/19
https://research.pasteur.fr/en/team/structural-bioinformatics/
Institut Pasteur, 25-28 rue du Docteur Roux, 75724 Paris CEDEX 15 France  
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.


Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-11-06 Thread Mark Abraham
Hi,

I suggest using the installation procedure on the link I already provided.

Mark

On Tue, 7 Nov 2017 01:41 Guyen Gn  wrote:

> Getting back to this thread after a while, is this the name of package?:
> rpmdevtools.noarch 8.3-5.el7   Server
> Red Hat Enterprise Linux Server release 7.3 (Maipo)
>
> I see it is included in the RH7.3
> Thanks.,
>
>
> On Wed, Aug 2, 2017 at 6:10 PM, Mark Abraham 
> wrote:
>
> > Hi,
> >
> > Usually this means, like it says, that the linker isn't able to do the
> > required job. I don't know whether the stock RHEL 7.2 linker can, but I
> > imagine that's the issue. Usually using a devtoolset e.g.
> > https://www.softwarecollections.org/en/scls/rhscl/devtoolset-4/ is the
> way
> > to get a machine with useful tools on it rather than one with "stable" =
> > "old" tools, particularly when you want to use the latest hardware...
> >
> > Mark
> >
> > On Thu, Aug 3, 2017 at 1:28 AM Guyen Gn  wrote:
> >
> > > Any update on this one??
> > >
> > > On Fri, Jul 14, 2017 at 1:30 PM, Guyen Gn  wrote:
> > >
> > > > Any ideas on this one? Thanks!
> > > >
> > > > On Mon, Jul 3, 2017 at 9:33 PM, Guyen Gn  wrote:
> > > >
> > > >> Still did not work. change xMIC to xCore and log is below (if it was
> > > >> other way around, then originally it must have not worked).
> > > >> Script file is all oneline.
> > > >>
> > > >> -- Performing Test C_xCORE_AVX512_FLAG_WORKS
> > > >> -- Performing Test C_xCORE_AVX512_FLAG_WORKS - Failed
> > > >> -- Performing Test C_mavx512f_mfma_FLAG_WORKS
> > > >> -- Performing Test C_mavx512f_mfma_FLAG_WORKS - Failed
> > > >> -- Performing Test C_mavx512f_FLAG_WORKS
> > > >> -- Performing Test C_mavx512f_FLAG_WORKS - Failed
> > > >> -- Performing Test C_arch_AVX_FLAG_WORKS
> > > >> -- Performing Test C_arch_AVX_FLAG_WORKS - Failed
> > > >> -- Performing Test C_hgnu_FLAG_WORKS
> > > >> -- Performing Test C_hgnu_FLAG_WORKS - Failed
> > > >> -- Performing Test C_FLAG_WORKS
> > > >> -- Performing Test C_FLAG_WORKS - Success
> > > >> -- Performing Test C_COMPILE_WORKS
> > > >> -- Performing Test C_COMPILE_WORKS - Failed
> > > >> -- Compiler flag was valid, but executable did not build - perhaps
> > > update
> > > >> the binutils package
> > > >> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS
> > > >> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS - Failed
> > > >> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS
> > > >> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS - Failed
> > > >> -- Performing Test CXX_mavx512f_FLAG_WORKS
> > > >> -- Performing Test CXX_mavx512f_FLAG_WORKS - Failed
> > > >> -- Performing Test CXX_arch_AVX_FLAG_WORKS
> > > >> -- Performing Test CXX_arch_AVX_FLAG_WORKS - Failed
> > > >> -- Performing Test CXX_hgnu_FLAG_WORKS
> > > >> -- Performing Test CXX_hgnu_FLAG_WORKS - Failed
> > > >> -- Performing Test CXX_FLAG_WORKS
> > > >> -- Performing Test CXX_FLAG_WORKS - Success
> > > >> -- Performing Test CXX_COMPILE_WORKS
> > > >> -- Performing Test CXX_COMPILE_WORKS - Failed
> > > >> -- Compiler flag was valid, but executable did not build - perhaps
> > > update
> > > >> the binutils package
> > > >> CMake Error at cmake/gmxManageSimd.cmake:92 (message):
> > > >>   Found a compiler flag for AVX 512F support, but some other problem
> > > >> exists.
> > > >>   Update your assembler and/or linker, e.g.  in the binutils package
> > of
> > > >> your
> > > >>   distribution.
> > > >> Call Stack (most recent call first):
> > > >>   cmake/gmxManageSimd.cmake:314 (gmx_give_fatal_error_when_sim
> > > >> d_support_not_found)
> > > >>   CMakeLists.txt:666 (gmx_manage_simd)
> > > >>
> > > >>
> > > >> -- Configuring incomplete, errors occurred!
> > > >>
> > > >>
> > > >> On Mon, Jul 3, 2017 at 2:44 PM, Mark Abraham <
> > mark.j.abra...@gmail.com>
> > > >> wrote:
> > > >>
> > > >>> Hi,
> > > >>>
> > > >>> On Mon, Jul 3, 2017 at 10:50 PM Guyen Gn 
> wrote:
> > > >>>
> > > >>> > Sorry for long wait on response, basically i followed the
> > instruction
> > > >>> at
> > > >>> > nvidia website:
> > > >>> >
> > > >>> > http://www.nvidia.com/object/gromacs-installation.html
> > > >>> >
> > > >>> > I instaled GPU driver versoin 369, as well as CUDA, gcc, cmake.
> > After
> > > >>> that
> > > >>> > it tells to run following:
> > > >>> >
> > > >>> > CC=gcc CXX=g++ cmake  -DGMX_OPENMP=ON
> > -DGMX_GPU=ON -
> > > >>> > DGPU_DEPLOYMENT_KIT_ROOT_DIR= -DGMX_BUILD_OWN_FFTW=ON -
> > > >>> > DGMX_PREFER_STATIC_LIBS=ON -DCMAKE_BUILD_TYPE=Release -
> > > >>> > DCMAKE_INSTALL_PREFIX=
> > > >>> >
> > > >>> > I tweaked as follows to fit my environment I setup (basically
> just
> > > >>> > installation source and run directory:
> > > >>> > CC=gcc CXX=g++ cmake /root/gromacs/gromacs-2016 -DGMX_OPENMP=ON
> > > >>> > -DGMX_GPU=ON -DGPU_DEPLOYMENT_KIT_ROOT_DIR=/usr/local/cuda
> > > >>> > -DGMX_BUILD_OWN_FFTW=ON -DGMX_PREFER_STATIC_LIBS=ON
> > > >>> > -DCMAKE_BUILD_TYPE=Release
> > > >>> > -DCMAKE_INSTALL_PREFIX=/root/gromacs/gromacs-2016-run inside
> > build.sh

Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-11-06 Thread Guyen Gn
Getting back to this thread after a while, is this the name of package?:
rpmdevtools.noarch 8.3-5.el7   Server
Red Hat Enterprise Linux Server release 7.3 (Maipo)

I see it is included in the RH7.3
Thanks.,


On Wed, Aug 2, 2017 at 6:10 PM, Mark Abraham 
wrote:

> Hi,
>
> Usually this means, like it says, that the linker isn't able to do the
> required job. I don't know whether the stock RHEL 7.2 linker can, but I
> imagine that's the issue. Usually using a devtoolset e.g.
> https://www.softwarecollections.org/en/scls/rhscl/devtoolset-4/ is the way
> to get a machine with useful tools on it rather than one with "stable" =
> "old" tools, particularly when you want to use the latest hardware...
>
> Mark
>
> On Thu, Aug 3, 2017 at 1:28 AM Guyen Gn  wrote:
>
> > Any update on this one??
> >
> > On Fri, Jul 14, 2017 at 1:30 PM, Guyen Gn  wrote:
> >
> > > Any ideas on this one? Thanks!
> > >
> > > On Mon, Jul 3, 2017 at 9:33 PM, Guyen Gn  wrote:
> > >
> > >> Still did not work. change xMIC to xCore and log is below (if it was
> > >> other way around, then originally it must have not worked).
> > >> Script file is all oneline.
> > >>
> > >> -- Performing Test C_xCORE_AVX512_FLAG_WORKS
> > >> -- Performing Test C_xCORE_AVX512_FLAG_WORKS - Failed
> > >> -- Performing Test C_mavx512f_mfma_FLAG_WORKS
> > >> -- Performing Test C_mavx512f_mfma_FLAG_WORKS - Failed
> > >> -- Performing Test C_mavx512f_FLAG_WORKS
> > >> -- Performing Test C_mavx512f_FLAG_WORKS - Failed
> > >> -- Performing Test C_arch_AVX_FLAG_WORKS
> > >> -- Performing Test C_arch_AVX_FLAG_WORKS - Failed
> > >> -- Performing Test C_hgnu_FLAG_WORKS
> > >> -- Performing Test C_hgnu_FLAG_WORKS - Failed
> > >> -- Performing Test C_FLAG_WORKS
> > >> -- Performing Test C_FLAG_WORKS - Success
> > >> -- Performing Test C_COMPILE_WORKS
> > >> -- Performing Test C_COMPILE_WORKS - Failed
> > >> -- Compiler flag was valid, but executable did not build - perhaps
> > update
> > >> the binutils package
> > >> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS
> > >> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS - Failed
> > >> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS
> > >> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS - Failed
> > >> -- Performing Test CXX_mavx512f_FLAG_WORKS
> > >> -- Performing Test CXX_mavx512f_FLAG_WORKS - Failed
> > >> -- Performing Test CXX_arch_AVX_FLAG_WORKS
> > >> -- Performing Test CXX_arch_AVX_FLAG_WORKS - Failed
> > >> -- Performing Test CXX_hgnu_FLAG_WORKS
> > >> -- Performing Test CXX_hgnu_FLAG_WORKS - Failed
> > >> -- Performing Test CXX_FLAG_WORKS
> > >> -- Performing Test CXX_FLAG_WORKS - Success
> > >> -- Performing Test CXX_COMPILE_WORKS
> > >> -- Performing Test CXX_COMPILE_WORKS - Failed
> > >> -- Compiler flag was valid, but executable did not build - perhaps
> > update
> > >> the binutils package
> > >> CMake Error at cmake/gmxManageSimd.cmake:92 (message):
> > >>   Found a compiler flag for AVX 512F support, but some other problem
> > >> exists.
> > >>   Update your assembler and/or linker, e.g.  in the binutils package
> of
> > >> your
> > >>   distribution.
> > >> Call Stack (most recent call first):
> > >>   cmake/gmxManageSimd.cmake:314 (gmx_give_fatal_error_when_sim
> > >> d_support_not_found)
> > >>   CMakeLists.txt:666 (gmx_manage_simd)
> > >>
> > >>
> > >> -- Configuring incomplete, errors occurred!
> > >>
> > >>
> > >> On Mon, Jul 3, 2017 at 2:44 PM, Mark Abraham <
> mark.j.abra...@gmail.com>
> > >> wrote:
> > >>
> > >>> Hi,
> > >>>
> > >>> On Mon, Jul 3, 2017 at 10:50 PM Guyen Gn  wrote:
> > >>>
> > >>> > Sorry for long wait on response, basically i followed the
> instruction
> > >>> at
> > >>> > nvidia website:
> > >>> >
> > >>> > http://www.nvidia.com/object/gromacs-installation.html
> > >>> >
> > >>> > I instaled GPU driver versoin 369, as well as CUDA, gcc, cmake.
> After
> > >>> that
> > >>> > it tells to run following:
> > >>> >
> > >>> > CC=gcc CXX=g++ cmake  -DGMX_OPENMP=ON
> -DGMX_GPU=ON -
> > >>> > DGPU_DEPLOYMENT_KIT_ROOT_DIR= -DGMX_BUILD_OWN_FFTW=ON -
> > >>> > DGMX_PREFER_STATIC_LIBS=ON -DCMAKE_BUILD_TYPE=Release -
> > >>> > DCMAKE_INSTALL_PREFIX=
> > >>> >
> > >>> > I tweaked as follows to fit my environment I setup (basically just
> > >>> > installation source and run directory:
> > >>> > CC=gcc CXX=g++ cmake /root/gromacs/gromacs-2016 -DGMX_OPENMP=ON
> > >>> > -DGMX_GPU=ON -DGPU_DEPLOYMENT_KIT_ROOT_DIR=/usr/local/cuda
> > >>> > -DGMX_BUILD_OWN_FFTW=ON -DGMX_PREFER_STATIC_LIBS=ON
> > >>> > -DCMAKE_BUILD_TYPE=Release
> > >>> > -DCMAKE_INSTALL_PREFIX=/root/gromacs/gromacs-2016-run inside
> build.sh
> > >>> shell
> > >>> > script.
> > >>> >
> > >>> > This is the full log, I highlighted toward the end section of code
> > >>> where it
> > >>> > started to fail. It appears something related to AVX. Thanks.,
> > >>> >
> > >>> > [root@localhost gromacs-2016]# ./build.sh
> > >>> > CMake Error: The source directory "/root/gromacs/gromacs-2016/-"
> does
> > >>> not
> > >>>

Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-08-02 Thread Mark Abraham
Hi,

Usually this means, like it says, that the linker isn't able to do the
required job. I don't know whether the stock RHEL 7.2 linker can, but I
imagine that's the issue. Usually using a devtoolset e.g.
https://www.softwarecollections.org/en/scls/rhscl/devtoolset-4/ is the way
to get a machine with useful tools on it rather than one with "stable" =
"old" tools, particularly when you want to use the latest hardware...

Mark

On Thu, Aug 3, 2017 at 1:28 AM Guyen Gn  wrote:

> Any update on this one??
>
> On Fri, Jul 14, 2017 at 1:30 PM, Guyen Gn  wrote:
>
> > Any ideas on this one? Thanks!
> >
> > On Mon, Jul 3, 2017 at 9:33 PM, Guyen Gn  wrote:
> >
> >> Still did not work. change xMIC to xCore and log is below (if it was
> >> other way around, then originally it must have not worked).
> >> Script file is all oneline.
> >>
> >> -- Performing Test C_xCORE_AVX512_FLAG_WORKS
> >> -- Performing Test C_xCORE_AVX512_FLAG_WORKS - Failed
> >> -- Performing Test C_mavx512f_mfma_FLAG_WORKS
> >> -- Performing Test C_mavx512f_mfma_FLAG_WORKS - Failed
> >> -- Performing Test C_mavx512f_FLAG_WORKS
> >> -- Performing Test C_mavx512f_FLAG_WORKS - Failed
> >> -- Performing Test C_arch_AVX_FLAG_WORKS
> >> -- Performing Test C_arch_AVX_FLAG_WORKS - Failed
> >> -- Performing Test C_hgnu_FLAG_WORKS
> >> -- Performing Test C_hgnu_FLAG_WORKS - Failed
> >> -- Performing Test C_FLAG_WORKS
> >> -- Performing Test C_FLAG_WORKS - Success
> >> -- Performing Test C_COMPILE_WORKS
> >> -- Performing Test C_COMPILE_WORKS - Failed
> >> -- Compiler flag was valid, but executable did not build - perhaps
> update
> >> the binutils package
> >> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS
> >> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS - Failed
> >> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS
> >> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS - Failed
> >> -- Performing Test CXX_mavx512f_FLAG_WORKS
> >> -- Performing Test CXX_mavx512f_FLAG_WORKS - Failed
> >> -- Performing Test CXX_arch_AVX_FLAG_WORKS
> >> -- Performing Test CXX_arch_AVX_FLAG_WORKS - Failed
> >> -- Performing Test CXX_hgnu_FLAG_WORKS
> >> -- Performing Test CXX_hgnu_FLAG_WORKS - Failed
> >> -- Performing Test CXX_FLAG_WORKS
> >> -- Performing Test CXX_FLAG_WORKS - Success
> >> -- Performing Test CXX_COMPILE_WORKS
> >> -- Performing Test CXX_COMPILE_WORKS - Failed
> >> -- Compiler flag was valid, but executable did not build - perhaps
> update
> >> the binutils package
> >> CMake Error at cmake/gmxManageSimd.cmake:92 (message):
> >>   Found a compiler flag for AVX 512F support, but some other problem
> >> exists.
> >>   Update your assembler and/or linker, e.g.  in the binutils package of
> >> your
> >>   distribution.
> >> Call Stack (most recent call first):
> >>   cmake/gmxManageSimd.cmake:314 (gmx_give_fatal_error_when_sim
> >> d_support_not_found)
> >>   CMakeLists.txt:666 (gmx_manage_simd)
> >>
> >>
> >> -- Configuring incomplete, errors occurred!
> >>
> >>
> >> On Mon, Jul 3, 2017 at 2:44 PM, Mark Abraham 
> >> wrote:
> >>
> >>> Hi,
> >>>
> >>> On Mon, Jul 3, 2017 at 10:50 PM Guyen Gn  wrote:
> >>>
> >>> > Sorry for long wait on response, basically i followed the instruction
> >>> at
> >>> > nvidia website:
> >>> >
> >>> > http://www.nvidia.com/object/gromacs-installation.html
> >>> >
> >>> > I instaled GPU driver versoin 369, as well as CUDA, gcc, cmake. After
> >>> that
> >>> > it tells to run following:
> >>> >
> >>> > CC=gcc CXX=g++ cmake  -DGMX_OPENMP=ON -DGMX_GPU=ON -
> >>> > DGPU_DEPLOYMENT_KIT_ROOT_DIR= -DGMX_BUILD_OWN_FFTW=ON -
> >>> > DGMX_PREFER_STATIC_LIBS=ON -DCMAKE_BUILD_TYPE=Release -
> >>> > DCMAKE_INSTALL_PREFIX=
> >>> >
> >>> > I tweaked as follows to fit my environment I setup (basically just
> >>> > installation source and run directory:
> >>> > CC=gcc CXX=g++ cmake /root/gromacs/gromacs-2016 -DGMX_OPENMP=ON
> >>> > -DGMX_GPU=ON -DGPU_DEPLOYMENT_KIT_ROOT_DIR=/usr/local/cuda
> >>> > -DGMX_BUILD_OWN_FFTW=ON -DGMX_PREFER_STATIC_LIBS=ON
> >>> > -DCMAKE_BUILD_TYPE=Release
> >>> > -DCMAKE_INSTALL_PREFIX=/root/gromacs/gromacs-2016-run inside build.sh
> >>> shell
> >>> > script.
> >>> >
> >>> > This is the full log, I highlighted toward the end section of code
> >>> where it
> >>> > started to fail. It appears something related to AVX. Thanks.,
> >>> >
> >>> > [root@localhost gromacs-2016]# ./build.sh
> >>> > CMake Error: The source directory "/root/gromacs/gromacs-2016/-" does
> >>> not
> >>> > exist.
> >>> > Specify --help for usage, or press the help button on the CMake GUI.
> >>> > ./build.sh: line 2: -DGMX_BUILD_OWN_FFTW=ON: command not found
> >>> > ./build.sh: line 3: -DCMAKE_BUILD_TYPE=Release: command not found
> >>> >
> >>>
> >>> All the above is because your script needed editing to handle newlines
> >>> from
> >>> your copy-paste.
> >>>
> >>>
> >>> > [root@localhost gromacs-2016]# nano -w build.sh
> >>> > [root@localhost gromacs-2016]# pwd
> >>> > /root/gromacs/gromacs-2016
> >>> > [root@localhost gromacs-2016]# mkdir ~/gromacs/gromac

Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-08-02 Thread Guyen Gn
Any update on this one??

On Fri, Jul 14, 2017 at 1:30 PM, Guyen Gn  wrote:

> Any ideas on this one? Thanks!
>
> On Mon, Jul 3, 2017 at 9:33 PM, Guyen Gn  wrote:
>
>> Still did not work. change xMIC to xCore and log is below (if it was
>> other way around, then originally it must have not worked).
>> Script file is all oneline.
>>
>> -- Performing Test C_xCORE_AVX512_FLAG_WORKS
>> -- Performing Test C_xCORE_AVX512_FLAG_WORKS - Failed
>> -- Performing Test C_mavx512f_mfma_FLAG_WORKS
>> -- Performing Test C_mavx512f_mfma_FLAG_WORKS - Failed
>> -- Performing Test C_mavx512f_FLAG_WORKS
>> -- Performing Test C_mavx512f_FLAG_WORKS - Failed
>> -- Performing Test C_arch_AVX_FLAG_WORKS
>> -- Performing Test C_arch_AVX_FLAG_WORKS - Failed
>> -- Performing Test C_hgnu_FLAG_WORKS
>> -- Performing Test C_hgnu_FLAG_WORKS - Failed
>> -- Performing Test C_FLAG_WORKS
>> -- Performing Test C_FLAG_WORKS - Success
>> -- Performing Test C_COMPILE_WORKS
>> -- Performing Test C_COMPILE_WORKS - Failed
>> -- Compiler flag was valid, but executable did not build - perhaps update
>> the binutils package
>> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS
>> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS - Failed
>> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS
>> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS - Failed
>> -- Performing Test CXX_mavx512f_FLAG_WORKS
>> -- Performing Test CXX_mavx512f_FLAG_WORKS - Failed
>> -- Performing Test CXX_arch_AVX_FLAG_WORKS
>> -- Performing Test CXX_arch_AVX_FLAG_WORKS - Failed
>> -- Performing Test CXX_hgnu_FLAG_WORKS
>> -- Performing Test CXX_hgnu_FLAG_WORKS - Failed
>> -- Performing Test CXX_FLAG_WORKS
>> -- Performing Test CXX_FLAG_WORKS - Success
>> -- Performing Test CXX_COMPILE_WORKS
>> -- Performing Test CXX_COMPILE_WORKS - Failed
>> -- Compiler flag was valid, but executable did not build - perhaps update
>> the binutils package
>> CMake Error at cmake/gmxManageSimd.cmake:92 (message):
>>   Found a compiler flag for AVX 512F support, but some other problem
>> exists.
>>   Update your assembler and/or linker, e.g.  in the binutils package of
>> your
>>   distribution.
>> Call Stack (most recent call first):
>>   cmake/gmxManageSimd.cmake:314 (gmx_give_fatal_error_when_sim
>> d_support_not_found)
>>   CMakeLists.txt:666 (gmx_manage_simd)
>>
>>
>> -- Configuring incomplete, errors occurred!
>>
>>
>> On Mon, Jul 3, 2017 at 2:44 PM, Mark Abraham 
>> wrote:
>>
>>> Hi,
>>>
>>> On Mon, Jul 3, 2017 at 10:50 PM Guyen Gn  wrote:
>>>
>>> > Sorry for long wait on response, basically i followed the instruction
>>> at
>>> > nvidia website:
>>> >
>>> > http://www.nvidia.com/object/gromacs-installation.html
>>> >
>>> > I instaled GPU driver versoin 369, as well as CUDA, gcc, cmake. After
>>> that
>>> > it tells to run following:
>>> >
>>> > CC=gcc CXX=g++ cmake  -DGMX_OPENMP=ON -DGMX_GPU=ON -
>>> > DGPU_DEPLOYMENT_KIT_ROOT_DIR= -DGMX_BUILD_OWN_FFTW=ON -
>>> > DGMX_PREFER_STATIC_LIBS=ON -DCMAKE_BUILD_TYPE=Release -
>>> > DCMAKE_INSTALL_PREFIX=
>>> >
>>> > I tweaked as follows to fit my environment I setup (basically just
>>> > installation source and run directory:
>>> > CC=gcc CXX=g++ cmake /root/gromacs/gromacs-2016 -DGMX_OPENMP=ON
>>> > -DGMX_GPU=ON -DGPU_DEPLOYMENT_KIT_ROOT_DIR=/usr/local/cuda
>>> > -DGMX_BUILD_OWN_FFTW=ON -DGMX_PREFER_STATIC_LIBS=ON
>>> > -DCMAKE_BUILD_TYPE=Release
>>> > -DCMAKE_INSTALL_PREFIX=/root/gromacs/gromacs-2016-run inside build.sh
>>> shell
>>> > script.
>>> >
>>> > This is the full log, I highlighted toward the end section of code
>>> where it
>>> > started to fail. It appears something related to AVX. Thanks.,
>>> >
>>> > [root@localhost gromacs-2016]# ./build.sh
>>> > CMake Error: The source directory "/root/gromacs/gromacs-2016/-" does
>>> not
>>> > exist.
>>> > Specify --help for usage, or press the help button on the CMake GUI.
>>> > ./build.sh: line 2: -DGMX_BUILD_OWN_FFTW=ON: command not found
>>> > ./build.sh: line 3: -DCMAKE_BUILD_TYPE=Release: command not found
>>> >
>>>
>>> All the above is because your script needed editing to handle newlines
>>> from
>>> your copy-paste.
>>>
>>>
>>> > [root@localhost gromacs-2016]# nano -w build.sh
>>> > [root@localhost gromacs-2016]# pwd
>>> > /root/gromacs/gromacs-2016
>>> > [root@localhost gromacs-2016]# mkdir ~/gromacs/gromacs-run
>>> > [root@localhost gromacs-2016]# ./build.sh
>>> > -- The C compiler identification is GNU 4.8.5
>>> > -- The CXX compiler identification is GNU 4.8.5
>>> > -- Check for working C compiler: /usr/bin/gcc
>>> > -- Check for working C compiler: /usr/bin/gcc -- works
>>> > -- Detecting C compiler ABI info
>>> > -- Detecting C compiler ABI info - done
>>> > -- Check for working CXX compiler: /usr/bin/g++
>>> > -- Check for working CXX compiler: /usr/bin/g++ -- works
>>> > -- Detecting CXX compiler ABI info
>>> > -- Detecting CXX compiler ABI info - done
>>> > -- Performing Test CXXFLAG_STD_CXX0X
>>> > -- Performing Test CXXFLAG_STD_CXX0X - Success
>>> > -- Performing Test CXX11_SUPPORTED
>>> > -

Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-07-14 Thread Guyen Gn
Any ideas on this one? Thanks!

On Mon, Jul 3, 2017 at 9:33 PM, Guyen Gn  wrote:

> Still did not work. change xMIC to xCore and log is below (if it was other
> way around, then originally it must have not worked).
> Script file is all oneline.
>
> -- Performing Test C_xCORE_AVX512_FLAG_WORKS
> -- Performing Test C_xCORE_AVX512_FLAG_WORKS - Failed
> -- Performing Test C_mavx512f_mfma_FLAG_WORKS
> -- Performing Test C_mavx512f_mfma_FLAG_WORKS - Failed
> -- Performing Test C_mavx512f_FLAG_WORKS
> -- Performing Test C_mavx512f_FLAG_WORKS - Failed
> -- Performing Test C_arch_AVX_FLAG_WORKS
> -- Performing Test C_arch_AVX_FLAG_WORKS - Failed
> -- Performing Test C_hgnu_FLAG_WORKS
> -- Performing Test C_hgnu_FLAG_WORKS - Failed
> -- Performing Test C_FLAG_WORKS
> -- Performing Test C_FLAG_WORKS - Success
> -- Performing Test C_COMPILE_WORKS
> -- Performing Test C_COMPILE_WORKS - Failed
> -- Compiler flag was valid, but executable did not build - perhaps update
> the binutils package
> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS
> -- Performing Test CXX_xCORE_AVX512_FLAG_WORKS - Failed
> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS
> -- Performing Test CXX_mavx512f_mfma_FLAG_WORKS - Failed
> -- Performing Test CXX_mavx512f_FLAG_WORKS
> -- Performing Test CXX_mavx512f_FLAG_WORKS - Failed
> -- Performing Test CXX_arch_AVX_FLAG_WORKS
> -- Performing Test CXX_arch_AVX_FLAG_WORKS - Failed
> -- Performing Test CXX_hgnu_FLAG_WORKS
> -- Performing Test CXX_hgnu_FLAG_WORKS - Failed
> -- Performing Test CXX_FLAG_WORKS
> -- Performing Test CXX_FLAG_WORKS - Success
> -- Performing Test CXX_COMPILE_WORKS
> -- Performing Test CXX_COMPILE_WORKS - Failed
> -- Compiler flag was valid, but executable did not build - perhaps update
> the binutils package
> CMake Error at cmake/gmxManageSimd.cmake:92 (message):
>   Found a compiler flag for AVX 512F support, but some other problem
> exists.
>   Update your assembler and/or linker, e.g.  in the binutils package of
> your
>   distribution.
> Call Stack (most recent call first):
>   cmake/gmxManageSimd.cmake:314 (gmx_give_fatal_error_when_
> simd_support_not_found)
>   CMakeLists.txt:666 (gmx_manage_simd)
>
>
> -- Configuring incomplete, errors occurred!
>
>
> On Mon, Jul 3, 2017 at 2:44 PM, Mark Abraham 
> wrote:
>
>> Hi,
>>
>> On Mon, Jul 3, 2017 at 10:50 PM Guyen Gn  wrote:
>>
>> > Sorry for long wait on response, basically i followed the instruction at
>> > nvidia website:
>> >
>> > http://www.nvidia.com/object/gromacs-installation.html
>> >
>> > I instaled GPU driver versoin 369, as well as CUDA, gcc, cmake. After
>> that
>> > it tells to run following:
>> >
>> > CC=gcc CXX=g++ cmake  -DGMX_OPENMP=ON -DGMX_GPU=ON -
>> > DGPU_DEPLOYMENT_KIT_ROOT_DIR= -DGMX_BUILD_OWN_FFTW=ON -
>> > DGMX_PREFER_STATIC_LIBS=ON -DCMAKE_BUILD_TYPE=Release -
>> > DCMAKE_INSTALL_PREFIX=
>> >
>> > I tweaked as follows to fit my environment I setup (basically just
>> > installation source and run directory:
>> > CC=gcc CXX=g++ cmake /root/gromacs/gromacs-2016 -DGMX_OPENMP=ON
>> > -DGMX_GPU=ON -DGPU_DEPLOYMENT_KIT_ROOT_DIR=/usr/local/cuda
>> > -DGMX_BUILD_OWN_FFTW=ON -DGMX_PREFER_STATIC_LIBS=ON
>> > -DCMAKE_BUILD_TYPE=Release
>> > -DCMAKE_INSTALL_PREFIX=/root/gromacs/gromacs-2016-run inside build.sh
>> shell
>> > script.
>> >
>> > This is the full log, I highlighted toward the end section of code
>> where it
>> > started to fail. It appears something related to AVX. Thanks.,
>> >
>> > [root@localhost gromacs-2016]# ./build.sh
>> > CMake Error: The source directory "/root/gromacs/gromacs-2016/-" does
>> not
>> > exist.
>> > Specify --help for usage, or press the help button on the CMake GUI.
>> > ./build.sh: line 2: -DGMX_BUILD_OWN_FFTW=ON: command not found
>> > ./build.sh: line 3: -DCMAKE_BUILD_TYPE=Release: command not found
>> >
>>
>> All the above is because your script needed editing to handle newlines
>> from
>> your copy-paste.
>>
>>
>> > [root@localhost gromacs-2016]# nano -w build.sh
>> > [root@localhost gromacs-2016]# pwd
>> > /root/gromacs/gromacs-2016
>> > [root@localhost gromacs-2016]# mkdir ~/gromacs/gromacs-run
>> > [root@localhost gromacs-2016]# ./build.sh
>> > -- The C compiler identification is GNU 4.8.5
>> > -- The CXX compiler identification is GNU 4.8.5
>> > -- Check for working C compiler: /usr/bin/gcc
>> > -- Check for working C compiler: /usr/bin/gcc -- works
>> > -- Detecting C compiler ABI info
>> > -- Detecting C compiler ABI info - done
>> > -- Check for working CXX compiler: /usr/bin/g++
>> > -- Check for working CXX compiler: /usr/bin/g++ -- works
>> > -- Detecting CXX compiler ABI info
>> > -- Detecting CXX compiler ABI info - done
>> > -- Performing Test CXXFLAG_STD_CXX0X
>> > -- Performing Test CXXFLAG_STD_CXX0X - Success
>> > -- Performing Test CXX11_SUPPORTED
>> > -- Performing Test CXX11_SUPPORTED - Success
>> > -- Performing Test CXX11_STDLIB_PRESENT
>> > -- Performing Test CXX11_STDLIB_PRESENT - Success
>> > -- Looking for NVIDIA GPUs present in the system
>> > -- 

Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-07-03 Thread Guyen Gn
Still did not work. change xMIC to xCore and log is below (if it was other
way around, then originally it must have not worked).
Script file is all oneline.

-- Performing Test C_xCORE_AVX512_FLAG_WORKS
-- Performing Test C_xCORE_AVX512_FLAG_WORKS - Failed
-- Performing Test C_mavx512f_mfma_FLAG_WORKS
-- Performing Test C_mavx512f_mfma_FLAG_WORKS - Failed
-- Performing Test C_mavx512f_FLAG_WORKS
-- Performing Test C_mavx512f_FLAG_WORKS - Failed
-- Performing Test C_arch_AVX_FLAG_WORKS
-- Performing Test C_arch_AVX_FLAG_WORKS - Failed
-- Performing Test C_hgnu_FLAG_WORKS
-- Performing Test C_hgnu_FLAG_WORKS - Failed
-- Performing Test C_FLAG_WORKS
-- Performing Test C_FLAG_WORKS - Success
-- Performing Test C_COMPILE_WORKS
-- Performing Test C_COMPILE_WORKS - Failed
-- Compiler flag was valid, but executable did not build - perhaps update
the binutils package
-- Performing Test CXX_xCORE_AVX512_FLAG_WORKS
-- Performing Test CXX_xCORE_AVX512_FLAG_WORKS - Failed
-- Performing Test CXX_mavx512f_mfma_FLAG_WORKS
-- Performing Test CXX_mavx512f_mfma_FLAG_WORKS - Failed
-- Performing Test CXX_mavx512f_FLAG_WORKS
-- Performing Test CXX_mavx512f_FLAG_WORKS - Failed
-- Performing Test CXX_arch_AVX_FLAG_WORKS
-- Performing Test CXX_arch_AVX_FLAG_WORKS - Failed
-- Performing Test CXX_hgnu_FLAG_WORKS
-- Performing Test CXX_hgnu_FLAG_WORKS - Failed
-- Performing Test CXX_FLAG_WORKS
-- Performing Test CXX_FLAG_WORKS - Success
-- Performing Test CXX_COMPILE_WORKS
-- Performing Test CXX_COMPILE_WORKS - Failed
-- Compiler flag was valid, but executable did not build - perhaps update
the binutils package
CMake Error at cmake/gmxManageSimd.cmake:92 (message):
  Found a compiler flag for AVX 512F support, but some other problem exists.
  Update your assembler and/or linker, e.g.  in the binutils package of your
  distribution.
Call Stack (most recent call first):
  cmake/gmxManageSimd.cmake:314
(gmx_give_fatal_error_when_simd_support_not_found)
  CMakeLists.txt:666 (gmx_manage_simd)


-- Configuring incomplete, errors occurred!


On Mon, Jul 3, 2017 at 2:44 PM, Mark Abraham 
wrote:

> Hi,
>
> On Mon, Jul 3, 2017 at 10:50 PM Guyen Gn  wrote:
>
> > Sorry for long wait on response, basically i followed the instruction at
> > nvidia website:
> >
> > http://www.nvidia.com/object/gromacs-installation.html
> >
> > I instaled GPU driver versoin 369, as well as CUDA, gcc, cmake. After
> that
> > it tells to run following:
> >
> > CC=gcc CXX=g++ cmake  -DGMX_OPENMP=ON -DGMX_GPU=ON -
> > DGPU_DEPLOYMENT_KIT_ROOT_DIR= -DGMX_BUILD_OWN_FFTW=ON -
> > DGMX_PREFER_STATIC_LIBS=ON -DCMAKE_BUILD_TYPE=Release -
> > DCMAKE_INSTALL_PREFIX=
> >
> > I tweaked as follows to fit my environment I setup (basically just
> > installation source and run directory:
> > CC=gcc CXX=g++ cmake /root/gromacs/gromacs-2016 -DGMX_OPENMP=ON
> > -DGMX_GPU=ON -DGPU_DEPLOYMENT_KIT_ROOT_DIR=/usr/local/cuda
> > -DGMX_BUILD_OWN_FFTW=ON -DGMX_PREFER_STATIC_LIBS=ON
> > -DCMAKE_BUILD_TYPE=Release
> > -DCMAKE_INSTALL_PREFIX=/root/gromacs/gromacs-2016-run inside build.sh
> shell
> > script.
> >
> > This is the full log, I highlighted toward the end section of code where
> it
> > started to fail. It appears something related to AVX. Thanks.,
> >
> > [root@localhost gromacs-2016]# ./build.sh
> > CMake Error: The source directory "/root/gromacs/gromacs-2016/-" does not
> > exist.
> > Specify --help for usage, or press the help button on the CMake GUI.
> > ./build.sh: line 2: -DGMX_BUILD_OWN_FFTW=ON: command not found
> > ./build.sh: line 3: -DCMAKE_BUILD_TYPE=Release: command not found
> >
>
> All the above is because your script needed editing to handle newlines from
> your copy-paste.
>
>
> > [root@localhost gromacs-2016]# nano -w build.sh
> > [root@localhost gromacs-2016]# pwd
> > /root/gromacs/gromacs-2016
> > [root@localhost gromacs-2016]# mkdir ~/gromacs/gromacs-run
> > [root@localhost gromacs-2016]# ./build.sh
> > -- The C compiler identification is GNU 4.8.5
> > -- The CXX compiler identification is GNU 4.8.5
> > -- Check for working C compiler: /usr/bin/gcc
> > -- Check for working C compiler: /usr/bin/gcc -- works
> > -- Detecting C compiler ABI info
> > -- Detecting C compiler ABI info - done
> > -- Check for working CXX compiler: /usr/bin/g++
> > -- Check for working CXX compiler: /usr/bin/g++ -- works
> > -- Detecting CXX compiler ABI info
> > -- Detecting CXX compiler ABI info - done
> > -- Performing Test CXXFLAG_STD_CXX0X
> > -- Performing Test CXXFLAG_STD_CXX0X - Success
> > -- Performing Test CXX11_SUPPORTED
> > -- Performing Test CXX11_SUPPORTED - Success
> > -- Performing Test CXX11_STDLIB_PRESENT
> > -- Performing Test CXX11_STDLIB_PRESENT - Success
> > -- Looking for NVIDIA GPUs present in the system
> > -- Number of NVIDIA GPUs detected: 2
> > -- Found CUDA: /usr/local/cuda (found suitable version "8.0", minimum
> > required is "5.0")
> > -- Could NOT find NVML (missing:  NVML_INCLUDE_DIR)
> > -- Checking for GCC x86 inline asm
> > -- Checking for GCC x86 inli

Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-07-03 Thread Mark Abraham
Hi,

On Mon, Jul 3, 2017 at 10:50 PM Guyen Gn  wrote:

> Sorry for long wait on response, basically i followed the instruction at
> nvidia website:
>
> http://www.nvidia.com/object/gromacs-installation.html
>
> I instaled GPU driver versoin 369, as well as CUDA, gcc, cmake. After that
> it tells to run following:
>
> CC=gcc CXX=g++ cmake  -DGMX_OPENMP=ON -DGMX_GPU=ON -
> DGPU_DEPLOYMENT_KIT_ROOT_DIR= -DGMX_BUILD_OWN_FFTW=ON -
> DGMX_PREFER_STATIC_LIBS=ON -DCMAKE_BUILD_TYPE=Release -
> DCMAKE_INSTALL_PREFIX=
>
> I tweaked as follows to fit my environment I setup (basically just
> installation source and run directory:
> CC=gcc CXX=g++ cmake /root/gromacs/gromacs-2016 -DGMX_OPENMP=ON
> -DGMX_GPU=ON -DGPU_DEPLOYMENT_KIT_ROOT_DIR=/usr/local/cuda
> -DGMX_BUILD_OWN_FFTW=ON -DGMX_PREFER_STATIC_LIBS=ON
> -DCMAKE_BUILD_TYPE=Release
> -DCMAKE_INSTALL_PREFIX=/root/gromacs/gromacs-2016-run inside build.sh shell
> script.
>
> This is the full log, I highlighted toward the end section of code where it
> started to fail. It appears something related to AVX. Thanks.,
>
> [root@localhost gromacs-2016]# ./build.sh
> CMake Error: The source directory "/root/gromacs/gromacs-2016/-" does not
> exist.
> Specify --help for usage, or press the help button on the CMake GUI.
> ./build.sh: line 2: -DGMX_BUILD_OWN_FFTW=ON: command not found
> ./build.sh: line 3: -DCMAKE_BUILD_TYPE=Release: command not found
>

All the above is because your script needed editing to handle newlines from
your copy-paste.


> [root@localhost gromacs-2016]# nano -w build.sh
> [root@localhost gromacs-2016]# pwd
> /root/gromacs/gromacs-2016
> [root@localhost gromacs-2016]# mkdir ~/gromacs/gromacs-run
> [root@localhost gromacs-2016]# ./build.sh
> -- The C compiler identification is GNU 4.8.5
> -- The CXX compiler identification is GNU 4.8.5
> -- Check for working C compiler: /usr/bin/gcc
> -- Check for working C compiler: /usr/bin/gcc -- works
> -- Detecting C compiler ABI info
> -- Detecting C compiler ABI info - done
> -- Check for working CXX compiler: /usr/bin/g++
> -- Check for working CXX compiler: /usr/bin/g++ -- works
> -- Detecting CXX compiler ABI info
> -- Detecting CXX compiler ABI info - done
> -- Performing Test CXXFLAG_STD_CXX0X
> -- Performing Test CXXFLAG_STD_CXX0X - Success
> -- Performing Test CXX11_SUPPORTED
> -- Performing Test CXX11_SUPPORTED - Success
> -- Performing Test CXX11_STDLIB_PRESENT
> -- Performing Test CXX11_STDLIB_PRESENT - Success
> -- Looking for NVIDIA GPUs present in the system
> -- Number of NVIDIA GPUs detected: 2
> -- Found CUDA: /usr/local/cuda (found suitable version "8.0", minimum
> required is "5.0")
> -- Could NOT find NVML (missing:  NVML_INCLUDE_DIR)
> -- Checking for GCC x86 inline asm
> -- Checking for GCC x86 inline asm - supported
> -- Detecting best SIMD instructions for this CPU
> -- Detected best SIMD instructions for this CPU - AVX_512
> -- Try OpenMP C flag = [-fopenmp]
> -- Performing Test OpenMP_FLAG_DETECTED
> -- Performing Test OpenMP_FLAG_DETECTED - Success
> -- Try OpenMP CXX flag = [-fopenmp]
> -- Performing Test OpenMP_FLAG_DETECTED
> -- Performing Test OpenMP_FLAG_DETECTED - Success
> -- Found OpenMP: -fopenmp
> -- Performing Test CFLAGS_EXCESS_PREC
> -- Performing Test CFLAGS_EXCESS_PREC - Success
> -- Performing Test CFLAGS_COPT
> -- Performing Test CFLAGS_COPT - Success
> -- Performing Test CFLAGS_NOINLINE
> -- Performing Test CFLAGS_NOINLINE - Success
> -- Performing Test CXXFLAGS_EXCESS_PREC
> -- Performing Test CXXFLAGS_EXCESS_PREC - Success
> -- Performing Test CXXFLAGS_COPT
> -- Performing Test CXXFLAGS_COPT - Success
> -- Performing Test CXXFLAGS_NOINLINE
> -- Performing Test CXXFLAGS_NOINLINE - Success
> -- Looking for include file unistd.h
> -- Looking for include file unistd.h - found
> -- Looking for include file pwd.h
> -- Looking for include file pwd.h - found
> -- Looking for include file dirent.h
> -- Looking for include file dirent.h - found
> -- Looking for include file time.h
> -- Looking for include file time.h - found
> -- Looking for include file sys/time.h
> -- Looking for include file sys/time.h - found
> -- Looking for include file io.h
> -- Looking for include file io.h - not found
> -- Looking for include file sched.h
> -- Looking for include file sched.h - found
> -- Looking for include file regex.h
> -- Looking for include file regex.h - found
> -- Looking for gettimeofday
> -- Looking for gettimeofday - found
> -- Looking for sysconf
> -- Looking for sysconf - found
> -- Looking for nice
> -- Looking for nice - found
> -- Looking for fsync
> -- Looking for fsync - found
> -- Looking for _fileno
> -- Looking for _fileno - not found
> -- Looking for fileno
> -- Looking for fileno - found
> -- Looking for _commit
> -- Looking for _commit - not found
> -- Looking for sigaction
> -- Looking for sigaction - found
> -- Performing Test HAVE_BUILTIN_CLZ
> -- Performing Test HAVE_BUILTIN_CLZ - Success
> -- Performing Test HAVE_BUILTIN_CLZLL
> -- Performing Test HAVE_BUILT

Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-07-03 Thread Guyen Gn
Sorry for long wait on response, basically i followed the instruction at
nvidia website:

http://www.nvidia.com/object/gromacs-installation.html

I instaled GPU driver versoin 369, as well as CUDA, gcc, cmake. After that
it tells to run following:

CC=gcc CXX=g++ cmake  -DGMX_OPENMP=ON -DGMX_GPU=ON -
DGPU_DEPLOYMENT_KIT_ROOT_DIR= -DGMX_BUILD_OWN_FFTW=ON -
DGMX_PREFER_STATIC_LIBS=ON -DCMAKE_BUILD_TYPE=Release -
DCMAKE_INSTALL_PREFIX=

I tweaked as follows to fit my environment I setup (basically just
installation source and run directory:
CC=gcc CXX=g++ cmake /root/gromacs/gromacs-2016 -DGMX_OPENMP=ON
-DGMX_GPU=ON -DGPU_DEPLOYMENT_KIT_ROOT_DIR=/usr/local/cuda
-DGMX_BUILD_OWN_FFTW=ON -DGMX_PREFER_STATIC_LIBS=ON
-DCMAKE_BUILD_TYPE=Release
-DCMAKE_INSTALL_PREFIX=/root/gromacs/gromacs-2016-run inside build.sh shell
script.

This is the full log, I highlighted toward the end section of code where it
started to fail. It appears something related to AVX. Thanks.,

[root@localhost gromacs-2016]# ./build.sh
CMake Error: The source directory "/root/gromacs/gromacs-2016/-" does not
exist.
Specify --help for usage, or press the help button on the CMake GUI.
./build.sh: line 2: -DGMX_BUILD_OWN_FFTW=ON: command not found
./build.sh: line 3: -DCMAKE_BUILD_TYPE=Release: command not found
[root@localhost gromacs-2016]# nano -w build.sh
[root@localhost gromacs-2016]# pwd
/root/gromacs/gromacs-2016
[root@localhost gromacs-2016]# mkdir ~/gromacs/gromacs-run
[root@localhost gromacs-2016]# ./build.sh
-- The C compiler identification is GNU 4.8.5
-- The CXX compiler identification is GNU 4.8.5
-- Check for working C compiler: /usr/bin/gcc
-- Check for working C compiler: /usr/bin/gcc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/g++
-- Check for working CXX compiler: /usr/bin/g++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Performing Test CXXFLAG_STD_CXX0X
-- Performing Test CXXFLAG_STD_CXX0X - Success
-- Performing Test CXX11_SUPPORTED
-- Performing Test CXX11_SUPPORTED - Success
-- Performing Test CXX11_STDLIB_PRESENT
-- Performing Test CXX11_STDLIB_PRESENT - Success
-- Looking for NVIDIA GPUs present in the system
-- Number of NVIDIA GPUs detected: 2
-- Found CUDA: /usr/local/cuda (found suitable version "8.0", minimum
required is "5.0")
-- Could NOT find NVML (missing:  NVML_INCLUDE_DIR)
-- Checking for GCC x86 inline asm
-- Checking for GCC x86 inline asm - supported
-- Detecting best SIMD instructions for this CPU
-- Detected best SIMD instructions for this CPU - AVX_512
-- Try OpenMP C flag = [-fopenmp]
-- Performing Test OpenMP_FLAG_DETECTED
-- Performing Test OpenMP_FLAG_DETECTED - Success
-- Try OpenMP CXX flag = [-fopenmp]
-- Performing Test OpenMP_FLAG_DETECTED
-- Performing Test OpenMP_FLAG_DETECTED - Success
-- Found OpenMP: -fopenmp
-- Performing Test CFLAGS_EXCESS_PREC
-- Performing Test CFLAGS_EXCESS_PREC - Success
-- Performing Test CFLAGS_COPT
-- Performing Test CFLAGS_COPT - Success
-- Performing Test CFLAGS_NOINLINE
-- Performing Test CFLAGS_NOINLINE - Success
-- Performing Test CXXFLAGS_EXCESS_PREC
-- Performing Test CXXFLAGS_EXCESS_PREC - Success
-- Performing Test CXXFLAGS_COPT
-- Performing Test CXXFLAGS_COPT - Success
-- Performing Test CXXFLAGS_NOINLINE
-- Performing Test CXXFLAGS_NOINLINE - Success
-- Looking for include file unistd.h
-- Looking for include file unistd.h - found
-- Looking for include file pwd.h
-- Looking for include file pwd.h - found
-- Looking for include file dirent.h
-- Looking for include file dirent.h - found
-- Looking for include file time.h
-- Looking for include file time.h - found
-- Looking for include file sys/time.h
-- Looking for include file sys/time.h - found
-- Looking for include file io.h
-- Looking for include file io.h - not found
-- Looking for include file sched.h
-- Looking for include file sched.h - found
-- Looking for include file regex.h
-- Looking for include file regex.h - found
-- Looking for gettimeofday
-- Looking for gettimeofday - found
-- Looking for sysconf
-- Looking for sysconf - found
-- Looking for nice
-- Looking for nice - found
-- Looking for fsync
-- Looking for fsync - found
-- Looking for _fileno
-- Looking for _fileno - not found
-- Looking for fileno
-- Looking for fileno - found
-- Looking for _commit
-- Looking for _commit - not found
-- Looking for sigaction
-- Looking for sigaction - found
-- Performing Test HAVE_BUILTIN_CLZ
-- Performing Test HAVE_BUILTIN_CLZ - Success
-- Performing Test HAVE_BUILTIN_CLZLL
-- Performing Test HAVE_BUILTIN_CLZLL - Success
-- Looking for clock_gettime in rt
-- Looking for clock_gettime in rt - found
-- Looking for feenableexcept in m
-- Looking for feenableexcept in m - found
-- Checking for sched.h GNU affinity API
-- Performing Test sched_affinity_compile
-- Performing Test sched_affinity_compile - Success
-- Looking for include file mm_malloc.h
-- Looking for include file m

Re: [gmx-users] AVX related compiler error during build with P100 RHEL7

2017-06-19 Thread Mark Abraham
Hi,

We can't support third party build guides, but please follow ours and ask
questions about it here. Provide full details.

Mark

On Mon, 19 Jun 2017 23:39 Guyen Gn  wrote:

> I have used the Nvidia's gromacs page to buid and run the gromacs.
> Few months back, I have used the Xeon CPU to build and run
> successfully on RHEL7.2. This also used the M6 GPU card.
> Now the same build is failing with AVX related compiler error if I use
> latest Skylake processor and P100 GPU. I have tried several
> permutations and all are same. May I get a support here? if so, I can
> send the details (build error log, system configuration, version of
> gromacs used etc.,)
> --
> Gromacs Users mailing list
>
> * Please search the archive at
> http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before
> posting!
>
> * Can't post? Read http://www.gromacs.org/Support/Mailing_Lists
>
> * For (un)subscribe requests visit
> https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or
> send a mail to gmx-users-requ...@gromacs.org.
>
-- 
Gromacs Users mailing list

* Please search the archive at 
http://www.gromacs.org/Support/Mailing_Lists/GMX-Users_List before posting!

* Can't post? Read http://www.gromacs.org/Support/Mailing_Lists

* For (un)subscribe requests visit
https://maillist.sys.kth.se/mailman/listinfo/gromacs.org_gmx-users or send a 
mail to gmx-users-requ...@gromacs.org.