Re: [PATCH net-next v8 5/7] mptcp: support rstreason for passive reset

[Matthieu Baerts <matttbe@kernel.org>]

Hi Jason,

On 23/04/2024 09:21, Jason Xing wrote:
> From: Jason Xing <kernelxing@tencent.com>
> 
> It relys on what reset options in the skb are as rfc8684 says. Reusing

(if you have something else to fix, 'checkpatch.pl --codespell' reported
a warning here: s/relys/relies/)

> this logic can save us much energy. This patch replaces most of the prior
> NOT_SPECIFIED reasons.
> 
> Signed-off-by: Jason Xing <kernelxing@tencent.com>
> ---
>  net/mptcp/protocol.h | 28 ++++++++++++++++++++++++++++
>  net/mptcp/subflow.c  | 22 +++++++++++++++++-----
>  2 files changed, 45 insertions(+), 5 deletions(-)
> 
> diff --git a/net/mptcp/protocol.h b/net/mptcp/protocol.h
> index fdfa843e2d88..bbcb8c068aae 100644
> --- a/net/mptcp/protocol.h
> +++ b/net/mptcp/protocol.h
> @@ -581,6 +581,34 @@ mptcp_subflow_ctx_reset(struct mptcp_subflow_context *subflow)
>  	WRITE_ONCE(subflow->local_id, -1);
>  }
>  
> +/* Convert reset reasons in MPTCP to enum sk_rst_reason type */
> +static inline enum sk_rst_reason
> +sk_rst_convert_mptcp_reason(u32 reason)
> +{
> +	switch (reason) {
> +	case MPTCP_RST_EUNSPEC:
> +		return SK_RST_REASON_MPTCP_RST_EUNSPEC;
> +	case MPTCP_RST_EMPTCP:
> +		return SK_RST_REASON_MPTCP_RST_EMPTCP;
> +	case MPTCP_RST_ERESOURCE:
> +		return SK_RST_REASON_MPTCP_RST_ERESOURCE;
> +	case MPTCP_RST_EPROHIBIT:
> +		return SK_RST_REASON_MPTCP_RST_EPROHIBIT;
> +	case MPTCP_RST_EWQ2BIG:
> +		return SK_RST_REASON_MPTCP_RST_EWQ2BIG;
> +	case MPTCP_RST_EBADPERF:
> +		return SK_RST_REASON_MPTCP_RST_EBADPERF;
> +	case MPTCP_RST_EMIDDLEBOX:
> +		return SK_RST_REASON_MPTCP_RST_EMIDDLEBOX;
> +	default:
> +		/**

I guess here as well, it should be '/*' instead of '/**'. But I guess
that's fine, this file is probably not scanned. Anyway, if you have to
send a new version, please fix this as well.

(Also, this helper might require '#include <net/rstreason.h>', but our
CI is fine with it, it is also added in the next commit, and probably
already included via include/net/request_sock.h. So I guess that's fine.)


Other than that, it looks good to me:

Reviewed-by: Matthieu Baerts (NGI0) <matttbe@kernel.org>

Cheers,
Matt
-- 
Sponsored by the NGI0 Core fund.