Re: [PATCH -next] audit: let the caller of audit_get_sk() ensure that net is valid

[cuigaosheng <cuigaosheng1@huawei.com>]

> I'd prefer to keep this check in place, and I think it may be required
> to ensure proper behavior in kauditd_thread()/kauditd_send_queue().

ok，thanks for taking time to review this patch.

On 2023/3/23 23:40, Paul Moore wrote:
> On Thu, Mar 16, 2023 at 11:38 PM Gaosheng Cui <cuigaosheng1@huawei.com> wrote:
>> The caller of audit_get_sk() must ensure that net is valid, otherwise
>> there will be null-prt-defer, for example, in the netlink_unicast().
>>
>> Signed-off-by: Gaosheng Cui <cuigaosheng1@huawei.com>
>> ---
>>   kernel/audit.c | 7 ++-----
>>   1 file changed, 2 insertions(+), 5 deletions(-)
>>
>> diff --git a/kernel/audit.c b/kernel/audit.c
>> index 9bc0b0301198..6a42a3801d01 100644
>> --- a/kernel/audit.c
>> +++ b/kernel/audit.c
>> @@ -285,16 +285,13 @@ static pid_t auditd_pid_vnr(void)
>>    * @net: the destination network namespace
>>    *
>>    * Description:
>> - * Returns the sock pointer if valid, NULL otherwise.  The caller must ensure
>> - * that a reference is held for the network namespace while the sock is in use.
>> + * Returns the sock pointer. The caller must ensure than net is valid and
>> + * a reference is held for the network namespace while the sock is in use.
>>    */
>>   static struct sock *audit_get_sk(const struct net *net)
>>   {
>>          struct audit_net *aunet;
>>
>> -       if (!net)
>> -               return NULL;
> I'd prefer to keep this check in place, and I think it may be required
> to ensure proper behavior in kauditd_thread()/kauditd_send_queue().
>
>>          aunet = net_generic(net, audit_net_id);
>>          return aunet->sk;
>>   }
>> --
>> 2.25.1