On Thu, Jan 23, 2020 at 03:01:29PM -0800, Guenter Roeck wrote:
> On Wed, Jan 22, 2020 at 10:29:05AM +0100, Greg Kroah-Hartman wrote:
> > From: Christian Brauner <christian.brauner@xxxxxxxxxx>
> >
> > commit 6b3ad6649a4c75504edeba242d3fd36b3096a57f upstream.
> >
> > Commit 69f594a38967 ("ptrace: do not audit capability check when outputing /proc/pid/stat")
> > introduced the ability to opt out of audit messages for accesses to various
> > proc files since they are not violations of policy. While doing so it
> > somehow switched the check from ns_capable() to
> > has_ns_capability{_noaudit}(). That means it switched from checking the
> > subjective credentials of the task to using the objective credentials. This
> > is wrong since. ptrace_has_cap() is currently only used in
> > ptrace_may_access() And is used to check whether the calling task (subject)
> > has the CAP_SYS_PTRACE capability in the provided user namespace to operate
> > on the target task (object). According to the cred.h comments this would
> > mean the subjective credentials of the calling task need to be used.
> > This switches ptrace_has_cap() to use security_capable(). Because we only
> > call ptrace_has_cap() in ptrace_may_access() and in there we already have a
> > stable reference to the calling task's creds under rcu_read_lock() there's
> > no need to go through another series of dereferences and rcu locking done
> > in ns_capable{_noaudit}().
> >
> > As one example where this might be particularly problematic, Jann pointed
> > out that in combination with the upcoming IORING_OP_OPENAT feature, this
> > bug might allow unprivileged users to bypass the capability checks while
> > asynchronously opening files like /proc/*/mem, because the capability
> > checks for this would be performed against kernel credentials.
> >
> > To illustrate on the former point about this being exploitable: When
> > io_uring creates a new context it records the subjective credentials of the
> > caller. Later on, when it starts to do work it creates a kernel thread and
> > registers a callback. The callback runs with kernel creds for
> > ktask->real_cred and ktask->cred. To prevent this from becoming a
> > full-blown 0-day io_uring will call override_cred() and override
> > ktask->cred with the subjective credentials of the creator of the io_uring
> > instance. With ptrace_has_cap() currently looking at ktask->real_cred this
> > override will be ineffective and the caller will be able to open arbitray
> > proc files as mentioned above.
> > Luckily, this is currently not exploitable but will turn into a 0-day once
> > IORING_OP_OPENAT{2} land in v5.6. Fix it now!
> >
> > Cc: Oleg Nesterov <oleg@xxxxxxxxxx>
> > Cc: Eric Paris <eparis@xxxxxxxxxx>
> > Cc: stable@xxxxxxxxxxxxxxx
> > Reviewed-by: Kees Cook <keescook@xxxxxxxxxxxx>
> > Reviewed-by: Serge Hallyn <serge@xxxxxxxxxx>
> > Reviewed-by: Jann Horn <jannh@xxxxxxxxxx>
> > Fixes: 69f594a38967 ("ptrace: do not audit capability check when outputing /proc/pid/stat")
> > Signed-off-by: Christian Brauner <christian.brauner@xxxxxxxxxx>
> > Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
> >
> > ---
> > kernel/ptrace.c | 15 ++++++++++-----
> > 1 file changed, 10 insertions(+), 5 deletions(-)
> >
> > --- a/kernel/ptrace.c
> > +++ b/kernel/ptrace.c
> > @@ -258,12 +258,17 @@ static int ptrace_check_attach(struct ta
> > return ret;
> > }
> >
> > -static int ptrace_has_cap(struct user_namespace *ns, unsigned int mode)
> > +static bool ptrace_has_cap(const struct cred *cred, struct user_namespace *ns,
> > + unsigned int mode)
> > {
> > + int ret;
> > +
> > if (mode & PTRACE_MODE_NOAUDIT)
> > - return has_ns_capability_noaudit(current, ns, CAP_SYS_PTRACE);
> > + ret = security_capable(cred, ns, CAP_SYS_PTRACE);
> > else
> > - return has_ns_capability(current, ns, CAP_SYS_PTRACE);
> > + ret = security_capable(cred, ns, CAP_SYS_PTRACE);
> > +
> > + return ret == 0;
>
> This results in
> if (condition)
> do_something;
> else
> do_the_same;
>
> Is that really correct ? The upstream patch calls security_capable()
> with additional CAP_OPT_NOAUDIT vs. CAP_OPT_NONE parameter, which does
> make sense. But I don't really see the benefit of the change above.
Yeah, this is odd, and differs from the original version I applied to
the staging queue.
Sasha, you made this change to the patch, I'm guessing to make it build
properly in 4.14? Should I just have dropped it from there instead?
thanks,
greg k-h
