Error in master version c892a263
In the above version I have two errors:
- on CPU build : the code runs correctly up to Physical_time ~ 527: see below figure in which semi major axis is plotted versus time for c892a263 and compared to a non crashing case for 3daf035a.
Also the disk80.h5 has correct results, while in disk90.h5 the disk has collapsed -> density values close to zero.
nan appear in the orbitJupiter.dat and planetJupiterdat file. I do not know what is going on,
For example the orbitJupiter.dat file:
5.275200000000e+02 9.571928895717e-01 5.580508236465e-03 9.856890
274223e-01 7.627263948596e-02 1.120976196628e+00 -2.901589791249e
+00
5.338000000000e+02 -nan 0.000000000000e+00 -nan -nan -nan
-nan
the test case is on Jean-Zay: in /gpfsscratch/rech/oth/roth002/MigrationAccretion/M35Sigma01CPUk6 for c892a263 and in /gpfsscratch/rech/oth/roth002/MigrationAccretion/M35Sigma01CPUk63da for 3daf035a
The problem is related to what I have reported in #679 (closed) after compiling on GPU on the same test I have troubles in orbit files from the beginning of the simulation.