Messages in this thread | | | Date | Wed, 5 Sep 2018 20:29:07 +0200 (CEST) | From | Jiri Kosina <> | Subject | Re: [PATCH v3 1/3] ptrace: Provide ___ptrace_may_access() that can be applied on arbitrary tasks |
| |
On Wed, 5 Sep 2018, Andrea Arcangeli wrote:
> ptrace_has_cap(tcred->user_ns, mode) is supposed to eventually lockup > hard if called from scheduler as it does some locking, and we fixed > that already half a year ago. > > Not sure how it's still unfixed in Jiri's codebase after so long, or > if it's an issue specific to 3.10 and upstream gets away without this.
We haven't got any lockup reports in our kernels (and we do carry a variant of this patch), so it might be somehow specific to 3.10.
> diff --git a/kernel/ptrace.c b/kernel/ptrace.c > index eb7862f185ff..4a8d0dd73c93 100644 > --- a/kernel/ptrace.c > +++ b/kernel/ptrace.c > @@ -285,7 +285,8 @@ int ___ptrace_may_access(struct task_struct *tracer, > gid_eq(caller_gid, tcred->sgid) && > gid_eq(caller_gid, tcred->gid)) > goto ok; > - if (ptrace_has_cap(tcred->user_ns, mode)) > + if (!(mode & PTRACE_MODE_NOACCESS_CHK) && > + ptrace_has_cap(tcred->user_ns, mode)) > goto ok; > rcu_read_unlock(); > return -EPERM; > @@ -296,7 +297,8 @@ ok: > dumpable = get_dumpable(task->mm); > rcu_read_lock(); > if (dumpable != SUID_DUMP_USER && > - !ptrace_has_cap(__task_cred(task)->user_ns, mode)) { > + ((mode & PTRACE_MODE_NOACCESS_CHK) || > + !ptrace_has_cap(__task_cred(task)->user_ns, mode))) { > rcu_read_unlock(); > return -EPERM;
I will look into this whether it's still applicable or not, thanks a lot for the pointer.
(and no, my testing of the patch I sent on current tree didn't produce any hangs -- was there a reliable way to trigger it on 3.10?).
Thanks,
-- Jiri Kosina SUSE Labs
| |