Re: [PATCH v3] proc: Ensure we see the exit of each process tid exactly

[Eric W. Biederman]

ebiederm@xxxxxxxxxxxx (Eric W. Biederman) writes:
> Take a good look over the code base and see if I can see any more
> issues like was found in next_tgid and repost the core patches.

Oleg,

I am reading through kernel/ptrace.c and I am seeing a lot of:

	spin_lock(&child->sighand->siglock);

In places where I don't see anything guaranteeing the child is stopped
such as ptrace_freeze_traced.  Are all of those places safe or
do some of the need to be transformed into lock_task_sighand in
case the process is current running?

I might just be reading the code to quickly and missing what keeps the
code from executing exec and changing sighand.

Eric