Charm++ parallel programming system

[Phil Miller <mille121 AT illinois.edu>]

On Fri, Jan 8, 2010 at 12:10, sefer baday 
<sefer.baday AT unibas.ch>
 wrote:
> Hello Phil,
>
> Thanks again, I will make full test.

Thank you. Could you do us the small favor of running the test in
tests/util under the build that didn't work (32 bit mpi-linux)? If
that crashes, then it would be very easy for us to catch this
explicitly in the future.

> Probably I had difficulties in compiling because of lack of my  knowledge in
> Linux.

Don't sell yourself short. Your knowledge was quite sufficient to
answer debugging questions without difficulty.

> The documentation is quite good.

Nevertheless, any improvements we can make are helpful to us, too.

> Have a nice weekend,

You too!

Phil

>
> sefer
>
> On 8 Jan 2010, at 17:28, Phil Miller wrote:
>
>> On Fri, Jan 8, 2010 at 11:20, sefer baday 
>> <sefer.baday AT unibas.ch>
>>  wrote:
>>>
>>> Hi Phil,
>>>
>>> I checked my system. it is 64 bit.
>>> First I tried the non optimized option and it didn't work.
>>> I also checked the path for mpi compiler and it was ok.
>>>
>>> Then I decided to build with the mpi-linux-am64 architecture and it
>>> successfully compiled.
>>>
>>> and I run megatest and got this output :
>>>
>>> ./charmrun ./pgm 12 6 +p2
>>>
>>> Running on 2 processors:  ./pgm 12 6
>>> Megatest is running on 2 processors.
>>> test 0: initiated [bitvector (jbooth)]
>>> test 0: completed (0.00 sec)
>>> test 1: initiated [immediatering (gengbin)]
>>> test 1: completed (0.01 sec)
>>> test 2: initiated [callback (olawlor)]
>>> test 2: completed (0.00 sec)
>>> test 3: initiated [reduction (olawlor)]
>>> test 3: completed (0.00 sec)
>>> test 4: initiated [inherit (olawlor)]
>>> test 4: completed (0.00 sec)
>>> test 5: initiated [templates (milind)]
>>> test 5: completed (0.00 sec)
>>> test 6: initiated [statistics (olawlor)]
>>> test 6: completed (0.00 sec)
>>> test 7: initiated [rotest (milind)]
>>> test 7: completed (0.00 sec)
>>> test 8: initiated [priotest (mlind)]
>>> test 8: completed (0.00 sec)
>>> test 9: initiated [priomsg (fang)]
>>> test 9: completed (0.00 sec)
>>> test 10: initiated [marshall (olawlor)]
>>> test 10: completed (0.01 sec)
>>> test 11: initiated [migration (jackie)]
>>> test 11: completed (0.00 sec)
>>> test 12: initiated [queens (jackie)]
>>> test 12: completed (0.00 sec)
>>> test 13: initiated [packtest (fang)]
>>> test 13: completed (0.00 sec)
>>> test 14: initiated [tempotest (fang)]
>>> test 14: completed (0.00 sec)
>>> test 15: initiated [arrayring (fang)]
>>> test 15: completed (0.00 sec)
>>> test 16: initiated [fib (jackie)]
>>> test 16: completed (0.00 sec)
>>> test 17: initiated [synctest (mjlang)]
>>> test 17: completed (0.00 sec)
>>> test 18: initiated [nodecast (milind)]
>>> test 18: completed (0.00 sec)
>>> test 19: initiated [groupcast (mjlang)]
>>> test 19: completed (0.00 sec)
>>> test 20: initiated [varraystest (milind)]
>>> test 20: completed (0.00 sec)
>>> test 21: initiated [varsizetest (mjlang)]
>>> test 21: completed (0.00 sec)
>>> test 22: initiated [nodering (milind)]
>>> test 22: completed (0.00 sec)
>>> test 23: initiated [groupring (milind)]
>>> test 23: completed (0.00 sec)
>>> test 24: initiated [multi immediatering (gengbin)]
>>> test 24: completed (0.01 sec)
>>> test 25: initiated [multi callback (olawlor)]
>>> test 25: completed (0.00 sec)
>>> test 26: initiated [multi reduction (olawlor)]
>>> test 26: completed (0.00 sec)
>>> test 27: initiated [multi statistics (olawlor)]
>>> test 27: completed (0.00 sec)
>>> test 28: initiated [multi priotest (mlind)]
>>> test 28: completed (0.00 sec)
>>> test 29: initiated [multi priomsg (fang)]
>>> test 29: completed (0.00 sec)
>>> test 30: initiated [multi marshall (olawlor)]
>>> test 30: completed (0.03 sec)
>>> test 31: initiated [multi migration (jackie)]
>>> test 31: completed (0.00 sec)
>>> test 32: initiated [multi packtest (fang)]
>>> test 32: completed (0.00 sec)
>>> test 33: initiated [multi tempotest (fang)]
>>> test 33: completed (0.00 sec)
>>> test 34: initiated [multi arrayring (fang)]
>>> test 34: completed (0.00 sec)
>>> test 35: initiated [multi fib (jackie)]
>>> test 35: completed (0.01 sec)
>>> test 36: initiated [multi synctest (mjlang)]
>>> test 36: completed (0.01 sec)
>>> test 37: initiated [multi nodecast (milind)]
>>> test 37: completed (0.00 sec)
>>> test 38: initiated [multi groupcast (mjlang)]
>>> test 38: completed (0.00 sec)
>>> test 39: initiated [multi varraystest (milind)]
>>> test 39: completed (0.00 sec)
>>> test 40: initiated [multi varsizetest (mjlang)]
>>> test 40: completed (0.00 sec)
>>> test 41: initiated [multi nodering (milind)]
>>> test 41: completed (0.00 sec)
>>> test 42: initiated [multi groupring (milind)]
>>> test 42: completed (0.00 sec)
>>> test 43: initiated [all-at-once]
>>> test 43: completed (0.02 sec)
>>> All tests completed, exiting
>>> End of program
>>>
>>>
>>> So , does this mean that it is ok ?
>>
>> Yes, it seems to. If you want a bit more assurance, you can run the
>> full test suite with "make test" inside the built charm directory.
>>
>> Phil
>>
>>>
>>>
>>> thanks
>>>
>>> sefer
>>>
>>>
>>>
>>>
>>> On 8 Jan 2010, at 16:39, Phil Miller wrote:
>>>
>>>> On Fri, Jan 8, 2010 at 10:19, sefer baday 
>>>> <sefer.baday AT unibas.ch>
>>>>  wrote:
>>>>>
>>>>> thank you for your reply. I really had hard time to compile charm++.
>>>>
>>>> I'm sorry to hear that. If you can tell us what you ran into, it would
>>>> help us improve the documentation.
>>>>
>>>>>> What build command did you run?
>>>>>
>>>>> ./build charm++  mpi-linux   -O  -DCMK_OPTIMIZE=1
>>>>
>>>> Just to check, is your system 32-bit or 64-bit? The output of 'uname
>>>> -a' would be helpful.
>>>>
>>>> Also, if the things suggested below don't lead to success, it might be
>>>> worthwhile testing a non-optimized build, leaving off the -O and
>>>> -DCMK_OPTIMIZE=1 options.
>>>>
>>>>> and I also changed  the c++ compiler as mpiicpc and fortran compiler as
>>>>> ifort  in the conv-mach.sh file
>>>>>
>>>>> CMK_CPP_CHARM='/lib/cpp -P'
>>>>> CMK_CPP_C='mpicc -E'
>>>>> CMK_CC='mpicc '
>>>>
>>>> Is this the Intel C compiler corresponding to the Intel C++ and
>>>> Fortran compilers configured below? It's the right binary name, but
>>>> there might be another MPI compiler by that name earlier on your path.
>>>>
>>>>> CMK_CXX='mpiicpc '
>>>>> CMK_CXXPP='mpiicpc -E '
>>>>> CMK_CF77='f77'
>>>>> CMK_CF90='ifort'
>>>>> CMK_RANLIB='ranlib'
>>>>> CMK_LIBS='-lckqt -lmpich -pthread'
>>>>> CMK_LD_LIBRARY_PATH="-Wl,-rpath,$CHARMLIBSO/"
>>>>> CMK_NATIVE_LIBS=''
>>>>> CMK_NATIVE_CC='gcc '
>>>>> CMK_NATIVE_LD='gcc'
>>>>> CMK_NATIVE_CXX='g++ '
>>>>> CMK_NATIVE_LDXX='g++'
>>>>> CMK_NATIVE_CC='gcc '
>>>>> CMK_NATIVE_CXX='g++ '
>>>>> CMK_F90LIBS='-L/usr/lib -L/opt/absoft/lib -lf90math -lfio -lU77
>>>>> -lf77math
>>>>> '
>>>>> CMK_MOD_NAME_ALLCAPS=1
>>>>> CMK_MOD_EXT="mod"
>>>>> CMK_F90_USE_MODDIR=1
>>>>> CMK_F90_MODINC="-p"
>>>>>
>>>>>> What version of Linux are you running this on (distribution and
>>>>>> release)?
>>>>>
>>>>> CentOS release 5.4
>>>>> linux kernel version : 2.6.18-92.el5
>>>>>
>>>>>> What MPI libraries and compilers are you using?
>>>>>
>>>>>
>>>>> mpicc and mpiicpc
>>>>> version : intel/ictce/p_3.2.020/Linux/impi/3.2.1.009
>>>>
>>>> This isn't the latest, but should be perfectly fine.
>>>>
>>>>>> Is this test failure repeatable?
>>>>>
>>>>> yes it always gives the same error
>>>>>
>>>>>> Does it happen with only 1 processor?
>>>>>
>>>>> simple tests run correctly.
>>>>>
>>>>> ./charmrun ./hello
>>>>
>>>> How about the megatest program on one processor?
>>>>
>>>>> Running on 1 processors:  ./hello
>>>>> Running Hello on 1 processors for 5 elements
>>>>> Hello 0 created
>>>>> Hello 1 created
>>>>> Hello 2 created
>>>>> Hello 3 created
>>>>> Hello 4 created
>>>>> Hi[17] from element 0
>>>>> Hi[18] from element 1
>>>>> Hi[19] from element 2
>>>>> Hi[20] from element 3
>>>>> Hi[21] from element 4
>>>>> All done
>>>>> End of program
>>>>>
>>>>>
>>>>>
>>>>>> Does it happen with more than 2 processors?
>>>>>
>>>>> yes, it gives the same error for 8 processors too.
>>>>>
>>>>>
>>>>>
>>>>> thanks a lot
>>>>>
>>>>> sefer
>>>>>
>>>>>
>>>>>
>>>>> On 8 Jan 2010, at 15:10, Phil Miller wrote:
>>>>>
>>>>>> Sefer,
>>>>>>
>>>>>> More information would be very helpful in diagnosing your issue.
>>>>>>
>>>>>> What build command did you run?
>>>>>>
>>>>>> What version of Linux are you running this on (distribution and
>>>>>> release)?
>>>>>>
>>>>>> What MPI libraries and compilers are you using?
>>>>>>
>>>>>> Is this test failure repeatable?
>>>>>>
>>>>>> Does it happen with only 1 processor?
>>>>>>
>>>>>> Does it happen with more than 2 processors?
>>>>>>
>>>>>> Phil
>>>>>>
>>>>>> On Fri, Jan 8, 2010 at 08:33, sefer baday 
>>>>>> <sefer.baday AT unibas.ch>
>>>>>> wrote:
>>>>>>>
>>>>>>> Hello,
>>>>>>>
>>>>>>> I am trying to compile charm++  on the linux cluster with mpi
>>>>>>> compiler.
>>>>>>>
>>>>>>> I successfully compiled charm++.  However when do megatest  it fails.
>>>>>>>
>>>>>>> the following is the error I got in the test run :
>>>>>>>
>>>>>>> ./charmrun ./pgm 12 6 +p2
>>>>>>>
>>>>>>> Running on 2 processors:  ./pgm 12 6
>>>>>>> Megatest is running on 2 processors.
>>>>>>> test 0: initiated [bitvector (jbooth)]
>>>>>>> test 0: completed (0.00 sec)
>>>>>>> test 1: initiated [immediatering (gengbin)]
>>>>>>> test 1: completed (0.01 sec)
>>>>>>> test 2: initiated [callback (olawlor)]
>>>>>>> rank 0 in job 1  bc2-login01_51184   caused collective abort of all
>>>>>>> ranks
>>>>>>>  exit status of rank 0: killed by signal 11
>>>>>>>
>>>>>>>
>>>>>>> Could you help me ?
>>>>>>>
>>>>>>> What could be the my mistake ?
>>>>>>>
>>>>>>> thanks
>>>>>>>
>>>>>>> sefer
>>>>>>>
>>>>>
>>>>>
>>>
>>>
>
>