Hi Nicolas,

the different runs seem to be completely independent, so the only reason of the problem that I can guess is that generateNewMesh() generates a correct mesh in the first iteration, and an incorrect mesh in the second. Could you send the mesh that works, and the one that does not? Also, more lines of output of the error run would help, so that we could guess where in the code this happens.

r.

On 08/08/2016 02:11 PM, Nicolas DELAN wrote:

Hi,

Actually i am launching the attached case with a loop that generates a new mesh at each iteration then i launch sfepy with a subprocess.

for example :

for i in range(12) generateNewMesh() process = subprocess.Popen(["python", "simple.py", "thermalParamFile"])

When i do that, the first iteration works fine but the second crashes with an error.

Le mercredi 3 août 2016 15:06:00 UTC+2, Robert Cimrman a écrit :

...

Hi Nicolas,

could you send a minimal example that does that? I have no idea what is wrong without an additional information.

r.

On 08/03/2016 09:45 AM, Nicolas DELAN wrote:

...

Hi,

I try to launch several times sfepy using a loop but at the second iteration i have this error:

glibc detected *** python: malloc(): memory corruption: 0x0000000002e2d810

---

Is there any solution ?