| more problem_checkpoint() stuff -- successfully pass through the checkpoint dir name |
|
More...
|
about 12 years ago
|
| add a problem_checkpoint() and problem_restart() routine that is called from Castro_io.cpp at checkpoint and restart. This allows problems to store problem-specific information needed for restart. Note: checkpoint is called by the IO Processor only but restart is called by all processors. |
|
More...
|
about 12 years ago
|
| move the checkpoint, restart, and plotfile stuff into Castro_io.cpp -- starting to clean-up |
|
More...
|
about 12 years ago
|
| add the elapsed CPU time since the start of the simulation to the job_info file -- this is like in Maestro. The current elapsed time is stored in chkXXXXX/CPUtime. It is calculated in the code as numCores * (time since start of current run) + (time from restart file). where numCores is the number of MPI * OMP, and the time from restart file is what is read in from the chkpoint CPUtime file if we restarted. This keeps the time cumulative between restarts. |
|
More...
|
about 12 years ago
|
| comment |
|
More...
|
about 12 years ago
|
| more job_info stuff -- now species info |
|
More...
|
about 12 years ago
|
| more job_info syncing with Maestro -- now output the grid info |
|
More...
|
about 12 years ago
|
| sync up with the 3-d tracing -- max the pressure edge state with small_pres |
|
More...
|
about 12 years ago
|
| add the ppm_reference stuff to 2-d |
|
More...
|
about 12 years ago
|
| remove unneeded modules |
|
More...
|
about 12 years ago
|
| the colglaz logic was wrong -- there was a hardcoded parmeter that overrode the value of use_colglaz_in |
|
More...
|
about 12 years ago
|
| fix prototype |
|
More...
|
about 12 years ago
|
| make rot_freq a meth_param. Remove rot_freq from the argument list in the Rotate_?d.f90 routines. |
|
More...
|
about 12 years ago
|
| Made some subroutines public for radiation codes |
|
More...
|
about 12 years ago
|
| Nothing has really changed except that all Fortran 90 procedures that are not called from C++ live in modules now. |
|
More...
|
about 12 years ago
|
| fix the check on use_colglaz. Also put in a sanity check on ppm_reference |
|
More...
|
about 12 years ago
|
| Merge branch 'master' of gamera.lbl.gov:/usr/local/gitroot/Castro |
|
More...
|
about 12 years ago
|
| comments + remove the jumps that are not actually used (like the transverse velocities carried by the u-c and u+c waves). Also make ppm_reference = 2 -> 1. |
|
More...
|
about 12 years ago
|
| Fixed more mistaching argument bugs |
|
More...
|
about 12 years ago
|
| Fixed a bug. An actual argument was missing |
|
More...
|
about 12 years ago
|
| Fixed the intent of an argment. It should be inout instead of out. |
|
More...
|
about 12 years ago
|
| Fixed a bug |
|
More...
|
about 12 years ago
|
| Calls to sponge() weren't matching the declaration. domlo & domhi are integers. Changed the definition of sponge() to match. |
|
More...
|
about 12 years ago
|
| ppm_reference = 2 -- this is like 1, but it only sets the reference state to the jump carried by the fastest moving wave IF that wave is actually moving toward the interface. Doesn't seem to make a difference, but this should be more correct. |
|
More...
|
about 12 years ago
|
| Add one more OMP line. |
|
More...
|
about 12 years ago
|
| make grav_source_type a runtime parameter. Also fix type = 2 so that in the prediction, it does the same as type 1. |
|
More...
|
about 12 years ago
|
| OMP fix |
|
More...
|
about 12 years ago
|
| mainlined |
|
More...
|
about 12 years ago
|
| if use_colglaz = 1, we now use the Colella & Glaz Riemann solver. |
|
More...
|
about 12 years ago
|
| use_colglaz now in 3-d! |
|
More...
|
about 12 years ago
|
