[Trilinos-Users] runtime error with 36.4M non zeros, about ritz values

Mike Atambo mikeat4999 at gmail.com
Wed Feb 11 05:51:40 MST 2015


Hi,
i was attempting to create a  large sparse matrix, (about 36M non zeros),
and  it seems when fillcomplete is called, i meet this:

terminate called after throwing an instance of 'std::runtime_error'
terminate called after throwing an instance of 'std::runtime_error'
  what():
 /home/matambo/install/apps/trilinos-gnu/include/Tpetra_CrsMatrix_def.hpp:2651:

Throw number = 1

Throw test that evaluated to true: (gblerr)

Tpetra::CrsMatrix<std::complex<double>, int, int,
KokkosClassic::SerialNode>::globalAssemble: non-local entries correspond to
invalid rows.
  what():
 /home/matambo/install/apps/trilinos-gnu/include/Tpetra_CrsMatrix_def.hpp:2651:

Throw number = 1

Throw test that evaluated to true: (gblerr)

Tpetra::CrsMatrix<std::complex<double>, int, int,
KokkosClassic::SerialNode>::globalAssemble: non-local entries correspond to
invalid rows.
[gpu01:28251] *** Process received signal ***
[gpu01:28250] *** Process received signal ***
[gpu01:28251] Signal: Aborted (6)
[gpu01:28251] Signal code:  (-6)
[gpu01:28250] Signal: Aborted (6)
[gpu01:28250] Signal code:  (-6)
[gpu01:28250] [ 0] /lib64/libpthread.so.0() [0x335fc0f500]
[gpu01:28250] [ 1] /lib64/libc.so.6(gsignal+0x35) [0x335f8328a5]
[gpu01:28250] [ 2] /lib64/libc.so.6(abort+0x175) [0x335f834085]
[gpu01:28250] [ 3]
/usr/lib64/libstdc++.so.6(_ZN9__gnu_cxx27__verbose_terminate_handlerEv+0x12d)
[0x367a6bea5d]
[gpu01:28250] [ 4] /usr/lib64/libstdc++.so.6() [0x367a6bcbe6]
[gpu01:28250] [ 5] /usr/lib64/libstdc++.so.6() [0x367a6bcc13]
[gpu01:28250] [ 6] /usr/lib64/libstdc++.so.6() [0x367a6bcd0e]
[gpu01:28250] [ 7]
./kryanasazi.x(_ZN6Tpetra9CrsMatrixISt7complexIdEiiN13KokkosClassic10SerialNodeEE14globalAssembleEv+0x701)
[0x46d051]
[gpu01:28250] [ 8]
./kryanasazi.x(_ZN6Tpetra9CrsMatrixISt7complexIdEiiN13KokkosClassic10SerialNodeEE12fillCompleteERKN7Teuchos3RCPIKNS_3MapIiiS4_EEEESD_RKNS7_INS6_13ParameterListEEE+0x4e0)
[0x4670c0]
[gpu01:28250] [ 9]
./kryanasazi.x(_ZN6Tpetra9CrsMatrixISt7complexIdEiiN13KokkosClassic10SerialNodeEE12fillCompleteERKN7Teuchos3RCPINS6_13ParameterListEEE+0x47d)
[0x466a9d]
[gpu01:28250] [10]
./kryanasazi.x(_Z9GenStatesIN6Tpetra9CrsMatrixISt7complexIdEiiN13KokkosClassic10SerialNodeEEEEN7Teuchos3RCPIKT_EERKNS8_IKNS7_4CommIiEEEERSoim+0x14b6)
[0x461f16]
[gpu01:28250] [11] ./kryanasazi.x(main+0x17b) [0x45eccb]
[gpu01:28250] [12] /lib64/libc.so.6(__libc_start_main+0xfd) [0x335f81ecdd]
[gpu01:28250] [13] ./kryanasazi.x() [0x45ea89]
[gpu01:28250] *** End of error message ***
[gpu01:28251] [ 0] /lib64/libpthread.so.0() [0x335fc0f500]
[gpu01:28251] [ 1] /lib64/libc.so.6(gsignal+0x35) [0x335f8328a5]
[gpu01:28251] [ 2] /lib64/libc.so.6(abort+0x175) [0x335f834085]
[gpu01:28251] [ 3]
/usr/lib64/libstdc++.so.6(_ZN9__gnu_cxx27__verbose_terminate_handlerEv+0x12d)
[0x367a6bea5d]
[gpu01:28251] [ 4] /usr/lib64/libstdc++.so.6() [0x367a6bcbe6]
[gpu01:28251] [ 5] /usr/lib64/libstdc++.so.6() [0x367a6bcc13]
[gpu01:28251] [ 6] /usr/lib64/libstdc++.so.6() [0x367a6bcd0e]
[gpu01:28251] [ 7]
./kryanasazi.x(_ZN6Tpetra9CrsMatrixISt7complexIdEiiN13KokkosClassic10SerialNodeEE14globalAssembleEv+0x701)
[0x46d051]
[gpu01:28251] [ 8]
./kryanasazi.x(_ZN6Tpetra9CrsMatrixISt7complexIdEiiN13KokkosClassic10SerialNodeEE12fillCompleteERKN7Teuchos3RCPIKNS_3MapIiiS4_EEEESD_RKNS7_INS6_13ParameterListEEE+0x4e0)
[0x4670c0]
[gpu01:28251] [ 9]
./kryanasazi.x(_ZN6Tpetra9CrsMatrixISt7complexIdEiiN13KokkosClassic10SerialNodeEE12fillCompleteERKN7Teuchos3RCPINS6_13ParameterListEEE+0x47d)
[0x466a9d]
[gpu01:28251] [10]
./kryanasazi.x(_Z9GenStatesIN6Tpetra9CrsMatrixISt7complexIdEiiN13KokkosClassic10SerialNodeEEEEN7Teuchos3RCPIKT_EERKNS8_IKNS7_4CommIiEEEERSoim+0x14b6)
[0x461f16]
[gpu01:28251] [11] ./kryanasazi.x(main+0x17b) [0x45eccb]
[gpu01:28251] [12] /lib64/libc.so.6(__libc_start_main+0xfd) [0x335f81ecdd]
[gpu01:28251] [13] ./kryanasazi.x() [0x45ea89]
[gpu01:28251] *** End of error message ***
--------------------------------------------------------------------------
mpirun noticed that process rank 1 with PID 28251 on node gpu01 exited on
signal 6 (Aborted).
--------------------------------------------------------------------------


2. Question:
For smaller matrices,  where i meet some convergence issue, the krylov
solver prints out
a rizt value, that  from other computation is  close to the   eigenvalue
that im looking for,  is there any way to access this ritz value
programaticaly even  when convergence is not achieved?

-- 
M. O. Atambo
mikeat4999 at gmail.com
matambo at ictp.it
Ext .139
Room 209.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://software.sandia.gov/pipermail/trilinos-users/attachments/20150211/03990978/attachment.html>


More information about the Trilinos-Users mailing list