On 5 June 2015 at 15:01, Balaji, Pavan <[email protected]> wrote:

Huh. I wonder why our testing is not catching this. Can you send us the exact test you used?

This is the minimal Python code able to reproduce the leak, translation to C is obvious. I would not be surprised if none of your tests have such a code path. from mpi4py import MPI win = MPI.Win.Create(MPI.BOTTOM, 1, MPI.INFO_NULL, MPI.COMM_SELF) info = win.Get_info() win.Set_info(info) info.Free() win.Free() -- Lisandro Dalcin ============ Research Scientist Computer, Electrical and Mathematical Sciences & Engineering (CEMSE) Numerical Porous Media Center (NumPor) King Abdullah University of Science and Technology (KAUST) http://numpor.kaust.edu.sa/ 4700 King Abdullah University of Science and Technology al-Khawarizmi Bldg (Bldg 1), Office # 4332 Thuwal 23955-6900, Kingdom of Saudi Arabia http://www.kaust.edu.sa Office Phone: +966 12 808-0459