If the CLR runs out of memory during the virtual fence app, it appears that either the native stack gets corrupted, or processPhase receives an argument that is a CLR buffer but the getBuffer() call returns a null pointer. Has anybody else seen something like that?
If the CLR runs out of memory during the virtual fence app, it appears that either the native stack gets corrupted, or processPhase receives an argument that is a CLR buffer but the getBuffer() call returns a null pointer. Has anybody else seen something like that?