Re: [Libmesh-users] how to install PETSC with libmesh

2008-02-29 Thread yzhu
Thanks very much. I will have a try. yunfei Quoting [EMAIL PROTECTED]: > I had problems building libmesh with PETSc 2.3.3 but with 2.3.2-p10 > everything worked fine (under ubuntu 7.10). These were the steps: > > cvs -d:pserver:[EMAIL PROTECTED]:/cvsroot/libmesh co > libmesh > cd libmesh > cd con

Re: [Libmesh-users] how to install PETSC with libmesh

2008-02-29 Thread yzhu
sorry, do you mean I had better using --with-mpi=0 by "libMesh doesn't really support using an MPI configuration that differs from the configuration in PETSc" ? I wonder what commands and options you have used when installed PETSc? Could you give some of these ? Thanks. yunfei Quoting Roy Stogne

Re: [Libmesh-users] how to install PETSC with libmesh

2008-02-29 Thread yzhu
Thanks for your reply. Now I think the make errors are : "macro "MPI_ALLgather" requires 7 arguments, but only 6 given. MPI_ALLgather was not declared in this scope " Actually, in the process of install PETSC, it has prompted that MPI compiler wrappers are invalid. so I use: --with-mpi=0 Actuall

Re: [Libmesh-users] how to install PETSC with libmesh

2008-02-29 Thread yzhu
Thanks for email. yes, I have noticed configure script can not find a fortran compiler. IT staff told me that there is no gfortran in that computer, I know the problem is in PETSC, so I am just trying to install it. I am totally new to linux, so it's very hard for me to deal with these commands

[Libmesh-users] how to install PETSC with libmesh

2008-02-29 Thread yzhu
< Configuring library with real number support >>> <<< Configuring library with reference counting support >>> <<< Configuring library with Netcdf support >>> <<< Configuring library with Exodus API support >>>

[Libmesh-users] (no subject)

2008-02-29 Thread yzhu
< Configuring library with real number support >>> <<< Configuring library with reference counting support >>> <<< Configuring library with Netcdf support >>> <<< Configuring library with Exodus API support >>>