[CP2K:2378] Re: neither PGI nor Intel compiler working?

[Ondrej Marsalek]

Hi again,

disabling the optimizations did help. Now I get these failures:

$ grep FAIL regtest-03.log
                    QS/regtest-hfx/H2O-hfx-emd.inp         RUNTIME FAIL
            QS/regtest-hfx/H2O-hfx-emd-restart.inp         RUNTIME FAIL
                    QS/regtest-epr-1/NO2-EPR-3.inp         RUNTIME FAIL
                    QS/regtest-epr-2/NO2-EPR-1.inp         RUNTIME FAIL
                   FARMING/regtest-1/farming-4.inp         RUNTIME FAIL
        QS/regtest-hfx-periodic/H2O-trunc-auto.inp         RUNTIME FAIL
     QS/regtest-hfx-periodic/H2O-trunc-auto-md.inp         RUNTIME FAIL
        QS/regtest-hfx-periodic/CH3-trunc-auto.inp         RUNTIME FAIL
     QS/regtest-hfx-periodic/CH3-trunc-auto-md.inp         RUNTIME FAIL
number of FAILED  tests 9

I run the tests in parallel with 2 CPUs. Are there any more of these
that are expected to fail (except for farming, that's clear)? I'll try
to find some time to have a look at whatever should work and does not.

Best,
Ondrej


On Tue, Nov 3, 2009 at 12:12, flo <fsch... at pci.uzh.ch> wrote:
>
> Hi Ondrej,
>
> could you sent me the output files from RT propagation. I will have a
> look what's going wrong. Just in case these are only the EMD related
> tests, could you do me a favor and try to compile the
> rt_propagation_utils.F file with -O0 and see if the problems persist.
>
> Thanks
> Flo
>
> On Nov 3, 12:07 pm, Ondrej Marsalek <ondrej.... at gmail.com> wrote:
>> Hi Ilya,
>>
>> running the full test suite, I got some problem in RT propagation
>> (that printed some warning indicating that they are expected to fail)
>> and a few problems in HFX, which I have not investigated yet.
>> Otherwise, it looks fine.
>>
>> The arch file is attached. If you have any suggestions for further
>> tuning, please let me know.
>>
>> Best,
>> Ondrej
>>
>> On Tue, Nov 3, 2009 at 11:47, Ilya <ikupc... at gmail.com> wrote:
>>
>> > Hello Ondrej.
>>
>> > I have ifort 11.0, but I'm still getting Nan's in the output.
>>
>> > Could you post your arch file used for 11.0.084, please.
>>
>> > Thanks.
>>
>> > Ilya
>>
>> > On Oct 30, 12:33 am, Ondrej Marsalek <ondrej.... at gmail.com>
>> > wrote:
>> >> A quick update. Using the libint wrapper, I have compiled with
>> >> 11.0.084 and a quick test says that at least one of the previously
>> >> failing tests now does work.
>>
>> >> The question that now remains - is there any point in trying to get
>> >> 11.1.x working? Any experience?
>>
>> >> Ondrej
>>
>> >> On Thu, Oct 29, 2009 at 19:38, Ondrej Marsalek
>>
>> >> <ondrej.... at gmail.com> wrote:
>> >> > Hi,
>>
>> >> > this sounds promising. I would like to ask which version of the Intel
>> >> > compiler to use. The readme file for libint use lists 11.1.056 as
>> >> > having full ISO_C_BINDING support, so presumably it works. On the
>> >> > other hand, this page:
>>
>> >> >http://software.intel.com/en-us/articles/build-cp2k-using-intel-fortr...
>>
>> >> > says that 11.1 should not be used. I do realize that that information
>> >> > is from July, but it is still there.
>>
>> >> > Any suggestions appreciated,
>> >> > Ondrej
>>
>> >> > On Wed, Oct 28, 2009 at 16:21, Urban <bors... at pci.uzh.ch> wrote:
>>
>> >> >> There was a known issue that resulted in an error in
>> >> >> cp_fm_cholesky_decompose when run with certain processor numbers. That
>> >> >> issue, along with the Intel compiler issues, should have been fixed
>> >> >> with the commits of Oct. 27.
>>
>> >> >> On Oct 26, 4:46 pm, NUCP2K <chenh... at gmail.com> wrote:
>> >> >>> Dear cp2k community,
>>
>> >> >>>   I just downloaded the latest CVS snapshot and tried to compile the
>> >> >>> codes using my once working ARCH files for PGI and Intel compilers on
>> >> >>> Linux-x86-64 MPI machines. Unfortunately, none of the two compilers is
>> >> >>> working though both can get the codes compiled successfully. It seems
>> >> >>> to me that the "cp_fm_cholesky_decompose" is the cause of the runtime
>> >> >>> trouble since my testing job always exit abnormally when the
>> >> >>> subroutine is called.
>>
>> >> >>>   Anyone has the experience on successfully compiling and running the
>> >> >>> latest cp2k code on Linux PC MPI clusters? The sharing of your ARCH
>> >> >>> files is highly appreciated.
>>
>> >> >>>   Thanks.
>>
>> >> >>> Hanning
>>
>>
>>
>>  Linux-x86-64-intel.popt
>> 1KViewDownload
> >
>