Re: [PATCH] selftests/bpf: test accounting of tail calls when prog is NULL

[Hari Bathini <hbathini@linux.ibm.com>]

On 17/02/26 6:23 pm, Ilya Leoshkevich wrote:
> 
> On 2/16/26 10:08, Hari Bathini wrote:
>> Test whether tail call count is incorrectly accounted for, when the
>> tail call fails due to a missing BPF program.
>>
>> Signed-off-by: Hari Bathini <hbathini@linux.ibm.com>
>> ---
>>
>> - powerpc64 BPF JIT has been incorrectly accounting for tailcall count
>>    even when BPF program to tailcall into is missing. A simple change
>>    to one of the tailcall selftests could have flagged it earlier.
>>
>>      https://lore.kernel.org/all/20260216065639.1750181-2- 
>> hbathini@linux.ibm.com/
>>
>>   tools/testing/selftests/bpf/progs/tailcall3.c | 5 ++++-
>>   1 file changed, 4 insertions(+), 1 deletion(-)
>>
>> diff --git a/tools/testing/selftests/bpf/progs/tailcall3.c b/tools/ 
>> testing/selftests/bpf/progs/tailcall3.c
>> index f60bcd7b8d4b..204f19c30a3e 100644
>> --- a/tools/testing/selftests/bpf/progs/tailcall3.c
>> +++ b/tools/testing/selftests/bpf/progs/tailcall3.c
>> @@ -5,7 +5,7 @@
>>   struct {
>>       __uint(type, BPF_MAP_TYPE_PROG_ARRAY);
>> -    __uint(max_entries, 1);
>> +    __uint(max_entries, 2);
>>       __uint(key_size, sizeof(__u32));
>>       __uint(value_size, sizeof(__u32));
>>   } jmp_table SEC(".maps");
>> @@ -23,6 +23,9 @@ int classifier_0(struct __sk_buff *skb)
>>   SEC("tc")
>>   int entry(struct __sk_buff *skb)
>>   {
>> +    /* prog == NULL case */
>> +    bpf_tail_call_static(skb, &jmp_table, 1);
>> +
>>       bpf_tail_call_static(skb, &jmp_table, 0);
>>       return 0;
>>   }
> 
> Seems like the x86_64 JIT passes this already, but the interpreter still 
> has:
> 
>          if (unlikely(tail_call_cnt >= MAX_TAIL_CALL_CNT))
>              goto out;
> 
>          tail_call_cnt++;
> 
>          prog = READ_ONCE(array->ptrs[index]);
>          if (!prog)
>              goto out;
> 
> Should it be changed too? Regardless:

Yeah, I guess. Posted 
https://lore.kernel.org/bpf/20260220062959.195101-1-hbathini@linux.ibm.com/

> 
> Reviewed-by: Ilya Leoshkevich <iii@linux.ibm.com>

Thanks, Ilya.

- Hari