[MPTCP] Re: [bug report] mptcp: Implement MPTCP receive path

[MPTCP] Re: [bug report] mptcp: Implement MPTCP receive path

[Florian Westphal]

[-- Attachment #1: Type: text/plain, Size: 1306 bytes --]

Dan Carpenter <dan.carpenter(a)oracle.com> wrote:
> Hello Mat Martineau,
> 
> The patch 648ef4b88673: "mptcp: Implement MPTCP receive path" from
> Jan 21, 2020, leads to the following static checker warning:
> 
> 	net/mptcp/subflow.c:558 subflow_data_ready()
> 	warn: test_bit() takes a bit number

[..]

> net/mptcp/subflow.c
>    544  static void subflow_data_ready(struct sock *sk)
>    545  {
>    546          struct mptcp_subflow_context *subflow = mptcp_subflow_ctx(sk);
>    547          struct sock *parent = subflow->conn;
>    548  
>    549          if (!parent || !subflow->mp_capable) {
>    550                  subflow->tcp_data_ready(sk);
>    551  
>    552                  if (parent)
>    553                          parent->sk_data_ready(parent);
>    554                  return;
>    555          }
>    556  
>    557          if (mptcp_subflow_data_available(sk)) {
>    558                  set_bit(MPTCP_DATA_READY, &mptcp_sk(parent)->flags);
>                                 ^^^^^^^^^^^^^^^^
> This is "BIT(0)" but it should just be "0".  The double shift bug will
> not cause a problem so long as it is done consistently and it doesn't go
> higher than BIT(6).

Indeed, thanks for reporting.  I've sent a simple patch to toss the
BIT() part.