Re: [PATCH v3 10/10] kfence: add test suite

[Marco Elver <elver@google.com>]

On Mon, 21 Sep 2020 at 19:13, Paul E. McKenney <paulmck@kernel.org> wrote:
>
> On Mon, Sep 21, 2020 at 03:26:11PM +0200, Marco Elver wrote:
> > Add KFENCE test suite, testing various error detection scenarios. Makes
> > use of KUnit for test organization. Since KFENCE's interface to obtain
> > error reports is via the console, the test verifies that KFENCE outputs
> > expected reports to the console.
> >
> > Reviewed-by: Dmitry Vyukov <dvyukov@google.com>
> > Co-developed-by: Alexander Potapenko <glider@google.com>
> > Signed-off-by: Alexander Potapenko <glider@google.com>
> > Signed-off-by: Marco Elver <elver@google.com>
>
> [ . . . ]
>
> > +/* Test SLAB_TYPESAFE_BY_RCU works. */
> > +static void test_memcache_typesafe_by_rcu(struct kunit *test)
> > +{
> > +     const size_t size = 32;
> > +     struct expect_report expect = {
> > +             .type = KFENCE_ERROR_UAF,
> > +             .fn = test_memcache_typesafe_by_rcu,
> > +     };
> > +
> > +     setup_test_cache(test, size, SLAB_TYPESAFE_BY_RCU, NULL);
> > +     KUNIT_EXPECT_TRUE(test, test_cache); /* Want memcache. */
> > +
> > +     expect.addr = test_alloc(test, size, GFP_KERNEL, ALLOCATE_ANY);
> > +     *expect.addr = 42;
> > +
> > +     rcu_read_lock();
> > +     test_free(expect.addr);
> > +     KUNIT_EXPECT_EQ(test, *expect.addr, (char)42);
> > +     rcu_read_unlock();
>
> It won't happen very often, but memory really could be freed at this point,
> especially in CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels ...

Ah, thanks for pointing it out.

> > +     /* No reports yet, memory should not have been freed on access. */
> > +     KUNIT_EXPECT_FALSE(test, report_available());
>
> ... so the above statement needs to go before the rcu_read_unlock().

You mean the comment (and not the KUNIT_EXPECT_FALSE that no reports
were generated), correct?

Admittedly, the whole comment is a bit imprecise, so I'll reword.

> > +     rcu_barrier(); /* Wait for free to happen. */
>
> But you are quite right that the memory is not -guaranteed- to be freed
> until we get here.

Right, I'll update the comment.

Thanks,
-- Marco