[gmx-users] Error while installing GROMACS-2018-rc1

Mark Abraham mark.j.abraham at gmail.com
Tue Jan 9 11:45:51 CET 2018


Hi,

Please do try again when we make the formal release shortly. If there's
still an issue, we'd really appreciate hearing from you, e.g. open a
redmine issue and attach a full output of cmake + make check. But so far,
we think the issue is caused by something outside of anything we can
attempt handle.

Mark

On Tue, Jan 9, 2018 at 10:44 AM Mark Abraham <mark.j.abraham at gmail.com>
wrote:

> Hi,
>
> Thanks again for the testing and feedback.
>
> On Wed, Jan 3, 2018 at 11:52 AM Yongchul Chung <chung at northwestern.edu>
> wrote:
>
>> Well, celebrated too soon. I received the following error message during
>> "make check" step.
>>
>>
>>
>> NOTE: The GPU has >25% less load than the CPU. This imbalance causes
>>
>>       performance loss.
>>
>>
>>                Core t (s)   Wall t (s)        (%)
>>
>>        Time:        0.034        0.034      100.0
>>
>>                  (ns/day)    (hour/ns)
>>
>> Performance:       53.584        0.448
>>
>
> Clearly this test ran on a GPU...
>
>
>> Opened
>>
>> /Users/ygchung/Desktop/codes/gromacs/build/src/programs/mdrun/tests/Testing/Temporary/PmeTest_ReproducesEnergies_spc-and-methanol_PmeAuto.edr
>> as single precision energy file
>>
>> Reading energy frame     20 time    0.020         Reading file
>>
>> /Users/ygchung/Desktop/codes/gromacs/build/src/programs/mdrun/tests/Testing/Temporary/PmeTest_ReproducesEnergies.tpr,
>> VERSION 2018-rc1-dev-20180103-871f2f460-unknown (single precision)
>>
>> Can not increase nstlist because an NVE ensemble is used
>>
>> Using 1 MPI thread
>>
>> 1 GPU auto-selected for this run.
>>
>> Mapping of GPU IDs to the 1 GPU task in the 1 rank on this node:
>>
>>   PP:0
>>
>> Compilation of source file
>> /Users/ygchung/Desktop/codes/gromacs/src/gromacs/mdlib/nbnxn_ocl/
>> nbnxn_ocl_kernels.cl failed!
>>
>> -- Used build options: -DWARP_SIZE_TEST=64 -D_AMD_SOURCE_
>> -DGMX_OCL_FASTGEN_ADD_TWINCUT -DEL_EWALD_ANA -DEELNAME=_ElecEw
>> -DVDWNAME=_VdwLJ -DCENTRAL=22 -DNBNXN_GPU_NCLUSTER_PER_SUPERCLUSTER=8
>> -DNBNXN_GPU_CLUSTER_SIZE=8 -DNBNXN_GPU_JGROUP_SIZE=4
>> -DGMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY=4 -DNBNXN_MIN_RSQ=3.82e-07f
>> -DIATYPE_SHMEM -cl-fast-relaxed-math
>> -I/Users/ygchung/Desktop/codes/gromacs/src/gromacs/mdlib/nbnxn_ocl
>>
> >
>
>> --------------LOG START---------------
>>
>> <program source>:80:10: fatal error: 'nbnxn_ocl_kernel_pruneonly.clh' file
>> not found
>>
>> #include "nbnxn_ocl_kernel_pruneonly.clh"
>>
>>          ^
>>
>
> Yet somehow this file was not found shortly afterwards. It's clearly being
> told to look in the right place (from the -I flag in the build options).
> That looks to me like an issue in the OpenCL driver / runtime. I see you're
> using the latest Mac version, but perhaps trying to update the OpenCL
> things will help with a more stable setup.
>
> Mark
>
> ---------------LOG END----------------
>>
>>
>> -------------------------------------------------------
>>
>> Program:     mdrun-test, version 2018-rc1-dev-20180103-871f2f460-unknown
>>
>> Source file: src/gromacs/gpu_utils/ocl_compiler.cpp (line 502)
>>
>> Function:    cl_program gmx::ocl::compileProgram(FILE *, const std::string
>> &, const std::string &, cl_context, cl_device_id, ocl_vendor_id_t)
>>
>>
>> Internal error (bug):
>>
>> Failed to compile NBNXN kernels for GPU #AMD Radeon Pro 560 Compute Engine
>>
>>   Could not build OpenCL program, error was CL_BUILD_PROGRAM_FAILURE
>>
>>
>> For more information and tips for troubleshooting, please check the
>> GROMACS
>>
>> website at http://www.gromacs.org/Documentation/Errors
>>
>> -------------------------------------------------------
>>
>>
>>       Start 33: MdrunMpiTests
>>
>> 33/39 Test #33: MdrunMpiTests ....................   Passed    2.69 sec
>>
>>       Start 34: regressiontests/simple
>>
>> 34/39 Test #34: regressiontests/simple ...........   Passed    2.33 sec
>>
>>       Start 35: regressiontests/complex
>>
>> 35/39 Test #35: regressiontests/complex ..........   Passed   53.42 sec
>>
>>       Start 36: regressiontests/kernel
>>
>> 36/39 Test #36: regressiontests/kernel ...........   Passed   32.33 sec
>>
>>       Start 37: regressiontests/freeenergy
>>
>> 37/39 Test #37: regressiontests/freeenergy .......   Passed    5.70 sec
>>
>>       Start 38: regressiontests/pdb2gmx
>>
>> 38/39 Test #38: regressiontests/pdb2gmx ..........   Passed   15.66 sec
>>
>>       Start 39: regressiontests/rotation
>>
>> 39/39 Test #39: regressiontests/rotation .........   Passed    3.10 sec
>>
>>
>> 97% tests passed, 1 tests failed out of 39
>>
>>
>> Label Time Summary:
>>
>> GTest              =  19.86 sec*proc (33 tests)
>>
>> IntegrationTest    =  18.38 sec*proc (3 tests)
>>
>> MpiTest            =   2.73 sec*proc (3 tests)
>>
>> UnitTest           =   1.48 sec*proc (30 tests)
>>
>>
>> Total Test time (real) = 132.48 sec
>>
>>
>> The following tests FAILED:
>>
>> 32 - MdrunTests (Failed)
>>
>> Errors while running CTest
>>
>> make[3]: *** [CMakeFiles/run-ctest-nophys] Error 8
>>
>> make[2]: *** [CMakeFiles/run-ctest-nophys.dir/all] Error 2
>>
>> make[1]: *** [CMakeFiles/check.dir/rule] Error 2
>>
>> make: *** [check] Error 2
>>
>>
>>
>>
>> On Wed, Jan 3, 2018 at 7:42 PM, Yongchul Chung <chung at northwestern.edu>
>> wrote:
>>
>> > Thanks, I have successfully installed the release-2018 version of
>> GROMACS
>> > on Macbook Pro 2017 HighSierra.
>> >
>> >
>> > On Tue, Jan 2, 2018 at 1:43 PM, Mark Abraham <mark.j.abraham at gmail.com>
>> > wrote:
>> >
>> >> Hi,
>> >>
>> >> Thanks, we have found and prepared a fix for an issue that looks very
>> >> similar (https://gerrit.gromacs.org/#/c/7400). Apparently the Mac
>> OpenCL
>> >> compilation cannot tolerate any extra spaces in the command used to
>> >> compile
>> >> the kernels used on the device.
>> >>
>> >> You could either compile GROMACS without OpenCL support for now, or if
>> you
>> >> would like to see if we have fixed the issue for you, that link enables
>> >> you
>> >> to download a tarball that contains the fix! We'd love to hear that it
>> >> works.
>> >>
>> >> Mark
>> >>
>> >> On Mon, Jan 1, 2018, 20:05 Yongchul Chung <chung at northwestern.edu>
>> wrote:
>> >>
>> >> > I am trying to install the latest GROMACS on Macbook Pro 2017 15-inch
>> >> (High
>> >> > Sierra) and got the following error message when issuing "make"
>> command
>> >> > during the installation steps.
>> >> >
>> >> > My command for cmake was: cmake .. -DGMX_BUILD_OWN_FFTW=ON
>> >> > -DREGRESSIONTEST_DOWNLOAD=ON -DGMX_GPU=ON -DGMX_USE_OPENCL=ON
>> >> >
>> >> > Thanks in advance for the help.
>> >> >
>> >> > [ 91%] Building CXX object
>> >> >
>> >> > src/gromacs/CMakeFiles/libgromacs.dir/mdlib/nbnxn_ocl/nbnxn_
>> >> ocl_data_mgmt.cpp.o
>> >> >
>> >> > In file included from
>> >> >
>> >> > /Users/ygchung/Desktop/codes/gromacs-2018-rc1/src/gromacs/md
>> >> lib/nbnxn_ocl/nbnxn_ocl_data_mgmt.cpp:54:
>> >> >
>> >> > In file included from
>> >> >
>> >> > /Users/ygchung/Desktop/codes/gromacs-2018-rc1/src/gromacs/gp
>> >> u_utils/oclutils.h:46:
>> >> >
>> >> > In file included from
>> >> >
>> >> > /Users/ygchung/Desktop/codes/gromacs-2018-rc1/src/gromacs/gp
>> >> u_utils/gmxopencl.h:61:
>> >> >
>> >> > In file included from
>> >> > /System/Library/Frameworks/OpenCL.framework/Headers/opencl.h:16:
>> >> >
>> >> > In file included from
>> >> > /System/Library/Frameworks/OpenCL.framework/Headers/gcl.h:23:
>> >> >
>> >> > */usr/include/dispatch/dispatch.h:38:45: **error: **expected value in
>> >> > expression*
>> >> >
>> >> > #if !defined(HAVE_UNISTD_H) || HAVE_UNISTD_H
>> >> >
>> >> > *                                            ^*
>> >> >
>> >> > 1 error generated.
>> >> >
>> >> > make[2]: ***
>> >> >
>> >> > [src/gromacs/CMakeFiles/libgromacs.dir/mdlib/nbnxn_ocl/
>> >> nbnxn_ocl_data_mgmt.cpp.o]
>> >> > Error 1
>> >> >
>> >> > make[1]: *** [src/gromacs/CMakeFiles/libgromacs.dir/all] Error 2
>> >> >
>> >> > make: *** [all] Error 2
>> >> > --
>> >> > 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-request at 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-request at 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-request at gromacs.org.
>>
>


More information about the gromacs.org_gmx-users mailing list