[petsc-users] valgrind error only with MUMPS

Dominik Szczerba dominik at itis.ethz.ch
Sat Nov 5 08:52:23 CDT 2011


>> there are no errors reported. Problem with MUMPS?
>
>   Yes. Looks like they are being sloppy. Whether this is an actual error condition or irrelevant only they would know. Report it to them.
>
>   Barry

Reported. They responded that the valgrind error is harmless and will be fixed.

Dominik

>> ==19601== Syscall param writev(vector[...]) points to uninitialised byte(s)
>> ==19601==    at 0x6791F59: writev (writev.c:56)
>> ==19601==    by 0x552CD8: MPIDU_Sock_writev (sock_immed.i:610)
>> ==19601==    by 0x5550DB: MPIDI_CH3_iSendv (ch3_isendv.c:84)
>> ==19601==    by 0x53B83A: MPIDI_CH3_EagerContigIsend (ch3u_eager.c:541)
>> ==19601==    by 0x53DA03: MPID_Isend (mpid_isend.c:130)
>> ==19601==    by 0x1440C71: PMPI_Isend (isend.c:124)
>> ==19601==    by 0x1453DED: MPI_ISEND (isendf.c:190)
>> ==19601==    by 0x124069F: __dmumps_comm_buffer_MOD_dmumps_62
>> (dmumps_comm_buffer.F:567)
>> ==19601==    by 0x12C1B4C: dmumps_242_ (dmumps_part2.F:739)
>> ==19601==    by 0x120C987: dmumps_249_ (dmumps_part8.F:6541)
>> ==19601==    by 0x120599F: dmumps_245_ (dmumps_part8.F:3885)
>> ==19601==    by 0x1229BCF: dmumps_301_ (dmumps_part8.F:2120)
>> ==19601==    by 0x128D272: dmumps_ (dmumps_part1.F:665)
>> ==19601==    by 0x1178753: dmumps_f77_ (dmumps_part3.F:6651)
>> ==19601==    by 0x1151A79: dmumps_c (mumps_c.c:422)
>> ==19601==    by 0x791075: MatSolve_MUMPS (mumps.c:547)
>> ==19601==    by 0x73ABA0: MatSolve (matrix.c:3108)
>> ==19601==    by 0x7FC5B9: PCApply_LU (lu.c:204)
>> ==19601==    by 0xD95AAB: PCApply (precon.c:383)
>> ==19601==    by 0xD98A70: PCApplyBAorAB (precon.c:609)
>> ==19601==  Address 0x86df168 is 8 bytes inside a block of size 144 alloc'd
>> ==19601==    at 0x4C28F9F: malloc (vg_replace_malloc.c:236)
>> ==19601==    by 0x124235E: __dmumps_comm_buffer_MOD_dmumps_2
>> (dmumps_comm_buffer.F:175)
>> ==19601==    by 0x12426E7: __dmumps_comm_buffer_MOD_dmumps_55
>> (dmumps_comm_buffer.F:123)
>> ==19601==    by 0x121F28B: dmumps_301_ (dmumps_part8.F:989)
>> ==19601==    by 0x128D272: dmumps_ (dmumps_part1.F:665)
>> ==19601==    by 0x1178753: dmumps_f77_ (dmumps_part3.F:6651)
>> ==19601==    by 0x1151A79: dmumps_c (mumps_c.c:422)
>> ==19601==    by 0x791075: MatSolve_MUMPS (mumps.c:547)
>> ==19601==    by 0x73ABA0: MatSolve (matrix.c:3108)
>> ==19601==    by 0x7FC5B9: PCApply_LU (lu.c:204)
>> ==19601==    by 0xD95AAB: PCApply (precon.c:383)
>> ==19601==    by 0xD98A70: PCApplyBAorAB (precon.c:609)
>> ==19601==    by 0x8D817F: KSPSolve_BCGS (bcgs.c:79)
>> ==19601==    by 0x86B189: KSPSolve (itfunc.c:423)
>> ==19601==    by 0x56013D: PetscLinearSystem::Solve() (PetscLinearSystem.cxx:221)
>> ==19601==    by 0x4E34F6: SolidSolver::Solve() (SolidSolver.cxx:1153)
>> ==19601==    by 0x4EFF56: main (SolidSolverMain.cxx:684)


More information about the petsc-users mailing list