On Mon, Mar 27, 2017 at 12:21:15AM +0800, Xin Long wrote:
> David Laight noticed the support for MSG_MORE with datamsg->force_delay
> didn't really work as we expected, as the first msg with MSG_MORE set
> would always block the following chunks' dequeuing.
>
> This Patch is to rewrite it by saving the MSG_MORE flag into assoc as
> David Laight suggested.
>
> asoc->force_delay is used to save MSG_MORE flag before a msg is sent.
> All chunks in queue would not be sent out if asoc->force_delay is set
> by the msg with MSG_MORE flag, until a new msg without MSG_MORE flag
> clears asoc->force_delay.
>
> Note that this change would not affect the flush is generated by other
> triggers, like asoc->state != ESTABLISHED, queue size > pmtu etc.
>
> v1->v2:
> Not clear asoc->force_delay after sending the msg with MSG_MORE flag.
>
> Fixes: 4ea0c32f5f42 ("sctp: add support for MSG_MORE")
> Signed-off-by: Xin Long <lucien....@gmail.com>
> ---
> include/net/sctp/structs.h | 2 +-
> net/sctp/output.c | 2 +-
> net/sctp/socket.c | 2 +-
> 3 files changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/include/net/sctp/structs.h b/include/net/sctp/structs.h
> index 592dece..8caa5ee 100644
> --- a/include/net/sctp/structs.h
> +++ b/include/net/sctp/structs.h
> @@ -499,7 +499,6 @@ struct sctp_datamsg {
> /* Did the messenge fail to send? */
> int send_error;
> u8 send_failed:1,
> - force_delay:1,
> can_delay; /* should this message be Nagle delayed */
> };
>
> @@ -1878,6 +1877,7 @@ struct sctp_association {
>
> __u8 need_ecne:1, /* Need to send an ECNE Chunk? */
> temp:1, /* Is it a temporary association? */
> + force_delay:1,
> prsctp_enable:1,
> reconf_enable:1;
>
> diff --git a/net/sctp/output.c b/net/sctp/output.c
> index 1224421..73fd178 100644
> --- a/net/sctp/output.c
> +++ b/net/sctp/output.c
> @@ -704,7 +704,7 @@ static sctp_xmit_t sctp_packet_can_append_data(struct
> sctp_packet *packet,
> */
>
> if ((sctp_sk(asoc->base.sk)->nodelay || inflight == 0) &&
> - !chunk->msg->force_delay)
> + !asoc->force_delay)
How is this going to not block the flush on asoc->state != ESTABLISHED?
AFAICT b7018d0b6300 ("sctp: flush out queue once assoc state falls into
SHUTDOWN_PENDING") need to clear asoc->force_delay too.
Case I have in mind is the same old one:
- app send a msg with MSG_MORE
- close the asoc, without sending the final msg
> /* Nothing unacked */
> return SCTP_XMIT_OK;
>
> diff --git a/net/sctp/socket.c b/net/sctp/socket.c
> index 0f378ea..baa269a 100644
> --- a/net/sctp/socket.c
> +++ b/net/sctp/socket.c
> @@ -1965,7 +1965,7 @@ static int sctp_sendmsg(struct sock *sk, struct msghdr
> *msg, size_t msg_len)
> err = PTR_ERR(datamsg);
> goto out_free;
> }
> - datamsg->force_delay = !!(msg->msg_flags & MSG_MORE);
> + asoc->force_delay = !!(msg->msg_flags & MSG_MORE);
>
> /* Now send the (possibly) fragmented message. */
> list_for_each_entry(chunk, &datamsg->chunks, frag_list) {
> --
> 2.1.0
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-sctp" in
> the body of a message to majord...@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
