The reed solomon decoder throws a seg-fault if it is run repeatedly in a loop, even if a new Reed_Solomon object is created every time in the loop iteration.
Attached code and Makefile shows when it fails. The observed behavior is that when m=3, the encoder and decoder work fine inside the loop. When m>3, unexpected results ensue.
A "little" of time has passed, but anyway.
I've experienced no such a behaviour for m = 2, 3, ... 8 (inclusive).
It was tested against the freshly built ITPP 4.3.1 (master branch, commit: bb5c7e95f40e8fdb5c3f3d01a84bcbaf76f3676d, Ubuntu).
What version and OS was it?