Problem with boundary functions in 1D hydro shock tube about athena

[tomidakn]

Sorry, this is my fault. The boundary routine is being largely modified during implementation of MeshBlocks and MPI. This had been already fixed in my version but I have not committed it yet. I will commit as soon as possible once I confirm MPI is working correctly.

[c-white]

After implementing the fix I suggested in my local copy, a new error shows up. This might be a sign of a separate bug, so I'll record it here.

python configure.py --prob shock_tube --coord cartesian -debug
make clean
make
cd bin
./athena -i ../inputs/hydro/athinput.sod

results in

### FATAL ERROR in function [ParameterInput::GetReal]
Parameter name 'problem_id' not found in block 'job'

Some debug results:

gdb athena
(gdb) b parameter_input.cpp:ParameterInput::GetString
(gdb) run -i ../inputs/hydro/athinput.sod
(gdb) c
(gdb) b InputBlock::GetPtrToLine
(gdb) c
(gdb) n
(gdb) p name.c_str()
$1 = 0x100204318 "problem_id"
(gdb) p name.length()
$2 = 10
(gdb) p pl->param_name.c_str()
$3 = 0x100200de8 "problem_id"
(gdb) p pl->param_name.length()
$4 = 4607182418800017408

It seems that when searching for a line with "problem_id" it finds the line, but the C++ std::string is corrupted with a bad length. Note that std::string::compare() will not return a match if this parameter is not the same, even if the underlying C character arrays are the same up to a terminating \0.