Ruediger Pluem wrote:
Both shared memory segments have different shm ids but are using the same file.
I've always argued that it's not valid API test to attempt to remove a file backed SHM before the SHM is closed, so yes, this is a test bug. Joe disagrees, IIRC.
