Re: [PATCH bpf-next v6 2/2] selftests/bpf: Add test to verify freeing the special fields when update [lru_,]percpu_hash maps

[Leon Hwang]

On 2025/11/11 21:38, Leon Hwang wrote:
> 
> 
> On 2025/11/7 10:00, Yonghong Song wrote:
>>
>>
>> On 11/5/25 7:14 AM, Leon Hwang wrote:
>>> Add test to verify that updating [lru_,]percpu_hash maps decrements
>>> refcount when BPF_KPTR_REF objects are involved.
>>>
>>> The tests perform the following steps:
>>>
>>> 1. Call update_elem() to insert an initial value.
>>> 2. Use bpf_refcount_acquire() to increment the refcount.
>>> 3. Store the node pointer in the map value.
>>> 4. Add the node to a linked list.
>>> 5. Probe-read the refcount and verify it is *2*.
>>> 6. Call update_elem() again to trigger refcount decrement.
>>> 7. Probe-read the refcount and verify it is *1*.
>>>
>>> Signed-off-by: Leon Hwang <leon.hwang@xxxxxxxxx>
>>
>> LGTM with a few nits below.
>>
>> Acked-by: Yonghong Song <yonghong.song@xxxxxxxxx>
>>
> 
> Hi Yonghong,
> 
> Thanks for your review and ack.
> 
>>> ---
>>>   .../bpf/prog_tests/refcounted_kptr.c          | 57 ++++++++++++++++++
>>>   .../selftests/bpf/progs/refcounted_kptr.c     | 60 +++++++++++++++++++
>>>   2 files changed, 117 insertions(+)
>>>
>>> diff --git a/tools/testing/selftests/bpf/prog_tests/refcounted_kptr.c
>>> b/tools/testing/selftests/bpf/prog_tests/refcounted_kptr.c
>>> index d6bd5e16e6372..086f679fa3f61 100644
>>> --- a/tools/testing/selftests/bpf/prog_tests/refcounted_kptr.c
>>> +++ b/tools/testing/selftests/bpf/prog_tests/refcounted_kptr.c
>>> @@ -44,3 +44,60 @@ void test_refcounted_kptr_wrong_owner(void)
>>>       ASSERT_OK(opts.retval, "rbtree_wrong_owner_remove_fail_a2 retval");
>>>       refcounted_kptr__destroy(skel);
>>>   }
>>> +
>>> +void test_percpu_hash_refcounted_kptr_refcount_leak(void)
>>> +{
>>> +    struct refcounted_kptr *skel;
>>> +    int cpu_nr, fd, err, key = 0;
>>> +    struct bpf_map *map;
>>> +    size_t values_sz;
>>> +    u64 *values;
>>> +    LIBBPF_OPTS(bpf_test_run_opts, opts,
>>> +            .data_in = &pkt_v4,
>>> +            .data_size_in = sizeof(pkt_v4),
>>> +            .repeat = 1,
>>> +    );
>>> +
>>> +    cpu_nr = libbpf_num_possible_cpus();
>>> +    if (!ASSERT_GT(cpu_nr, 0, "libbpf_num_possible_cpus"))
>>> +        return;
>>> +
>>> +    values = calloc(cpu_nr, sizeof(u64));
>>> +    if (!ASSERT_OK_PTR(values, "calloc values"))
>>> +        return;
>>> +
>>> +    skel = refcounted_kptr__open_and_load();
>>> +    if (!ASSERT_OK_PTR(skel, "refcounted_kptr__open_and_load")) {
>>> +        free(values);
>>> +        return;
>>> +    }
>>> +
>>> +    values_sz = cpu_nr * sizeof(u64);
>>> +    memset(values, 0, values_sz);
>>> +
>>> +    map = skel->maps.percpu_hash;
>>> +    err = bpf_map__update_elem(map, &key, sizeof(key), values,
>>> values_sz, 0);
>>> +    if (!ASSERT_OK(err, "bpf_map__update_elem"))
>>> +        goto out;
>>> +
>>> +    fd = bpf_program__fd(skel->progs.percpu_hash_refcount_leak);
>>> +    err = bpf_prog_test_run_opts(fd, &opts);
>>> +    if (!ASSERT_OK(err, "bpf_prog_test_run_opts"))
>>> +        goto out;
>>> +    if (!ASSERT_EQ(opts.retval, 2, "opts.retval"))
>>> +        goto out;
>>> +
>>> +    err = bpf_map__update_elem(map, &key, sizeof(key), values,
>>> values_sz, 0);
>>> +    if (!ASSERT_OK(err, "bpf_map__update_elem"))
>>> +        goto out;
>>> +
>>> +    fd = bpf_program__fd(skel->progs.check_percpu_hash_refcount);
>>> +    err = bpf_prog_test_run_opts(fd, &opts);
>>> +    ASSERT_OK(err, "bpf_prog_test_run_opts");
>>> +    ASSERT_EQ(opts.retval, 1, "opts.retval");
>>> +
>>> +out:
>>> +    refcounted_kptr__destroy(skel);
>>> +    free(values);
>>> +}
>>> +
>>
>> Empty line here.
>>
>>> diff --git a/tools/testing/selftests/bpf/progs/refcounted_kptr.c b/
>>> tools/testing/selftests/bpf/progs/refcounted_kptr.c
>>> index 893a4fdb4b6e9..1aca85d86aebc 100644
>>> --- a/tools/testing/selftests/bpf/progs/refcounted_kptr.c
>>> +++ b/tools/testing/selftests/bpf/progs/refcounted_kptr.c
>>> @@ -568,4 +568,64 @@ int
>>> BPF_PROG(rbtree_sleepable_rcu_no_explicit_rcu_lock,
>>>       return 0;
>>>   }
>>>   +private(kptr_ref) u64 ref;
>>> +
>>> +static int probe_read_refcount(void)
>>> +{
>>> +    u32 refcount;
>>> +
>>> +    bpf_probe_read_kernel(&refcount, sizeof(refcount), (void *) ref);
>>> +    return refcount;
>>> +}
>>> +
>>> +static int __insert_in_list(struct bpf_list_head *head, struct
>>> bpf_spin_lock *lock,
>>> +                struct node_data __kptr **node)
>>> +{
>>> +    struct node_data *node_new, *node_ref, *node_old;
>>> +
>>> +    node_new = bpf_obj_new(typeof(*node_new));
>>> +    if (!node_new)
>>> +        return -1;
>>> +
>>> +    node_ref = bpf_refcount_acquire(node_new);
>>> +    node_old = bpf_kptr_xchg(node, node_new);
>>
>> Change the above to node_old = bpf_kptr_xchg(node, node_node_ref); might
>> be better for reasoning although node_ref/node_new are the same.
>>
> 
> Nope — node_ref and node_new are different for the verifier.

They are the same in theory.

The verifier failure was likely caused by something else, but I'm not
sure of the exact reason.

> 
> When trying node_old = bpf_kptr_xchg(node, node_ref), the verifier reported:
> 
> [verifier log snipped for brevity...]
> ; bpf_obj_drop(node_ref); @ refcounted_kptr.c:594
> 26: (bf) r1 = r6                      ; R1=scalar(id=7) R6=scalar(id=7)
> refs=3
> 27: (b7) r2 = 0                       ; R2=0 refs=3
> 28: (85) call bpf_obj_drop_impl#54490
> R1 must be referenced or trusted
> processed 27 insns (limit 1000000) max_states_per_insn 0 total_states 2
> peak_states 2 mark_read 0
> 
> So the verifier rejected it because R6 became scalar(id=7) from
> ptr_node_data(ref_obj_id=4).
> 
> ---
> 
> Hi Alexei, could you please drop the extra empty line when applying this
> patch?
> 
> Then I don't need to send another revision.
> 
> Thanks,
> Leon
> 
> [...]
>