Messages in this thread | | | From | Arnd Bergmann <> | Subject | Re: Possible mem leak in copy_process() | Date | Thu, 17 Apr 2008 06:17:34 +0200 |
| |
On Wednesday 16 April 2008, Jesper Juhl wrote: > > Perhaps it can never happen the way Coverity thinks it can. You are a much > better judge of that than I, but it looks to me like it's at least > possible in theory - in which case we have a potential leak every time we > create a new process and that can't be good...
I think you're just looking at the common case here, which is correct: The pid is allocated for task creation and stays around as long as the task does. It gets freed when the last reference to it goes away, usually in the put_pid() called from exit().
Arnd <><
| |