From patchwork Thu Nov 16 02:20:41 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Li,Rongqing" X-Patchwork-Id: 13457526 X-Patchwork-Delegate: kuba@kernel.org Authentication-Results: smtp.subspace.kernel.org; dkim=none Received: from njjs-sys-mailin01.njjs.baidu.com (mx314.baidu.com [180.101.52.172]) by lindbergh.monkeyblade.net (Postfix) with ESMTP id CA9E08E; Wed, 15 Nov 2023 18:20:42 -0800 (PST) Received: from localhost (bjhw-sys-rpm015653cc5.bjhw.baidu.com [10.227.53.39]) by njjs-sys-mailin01.njjs.baidu.com (Postfix) with ESMTP id 38C9E7F00049; Thu, 16 Nov 2023 10:20:41 +0800 (CST) From: Li RongQing To: wenjia@linux.ibm.co, netdev@vger.kernel.org, linux-s390@vger.kernel.org, dust.li@linux.alibaba.com Subject: [PATCH][net-next] net/smc: avoid atomic_set and smp_wmb in the tx path when possible Date: Thu, 16 Nov 2023 10:20:41 +0800 Message-Id: <20231116022041.51959-1-lirongqing@baidu.com> X-Mailer: git-send-email 2.9.4 X-Spam-Level: * Precedence: bulk X-Mailing-List: netdev@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: X-Patchwork-Delegate: kuba@kernel.org there is rare possibility that conn->tx_pushing is not 1, since tx_pushing is just checked with 1, so move the setting tx_pushing to 1 after atomic_dec_and_test() return false, to avoid atomic_set and smp_wmb in tx path Signed-off-by: Li RongQing Reviewed-by: Dust Li --- net/smc/smc_tx.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/net/smc/smc_tx.c b/net/smc/smc_tx.c index 3b0ff3b..72dbdee 100644 --- a/net/smc/smc_tx.c +++ b/net/smc/smc_tx.c @@ -667,8 +667,6 @@ int smc_tx_sndbuf_nonempty(struct smc_connection *conn) return 0; again: - atomic_set(&conn->tx_pushing, 1); - smp_wmb(); /* Make sure tx_pushing is 1 before real send */ rc = __smc_tx_sndbuf_nonempty(conn); /* We need to check whether someone else have added some data into @@ -677,8 +675,11 @@ int smc_tx_sndbuf_nonempty(struct smc_connection *conn) * If so, we need to push again to prevent those data hang in the send * queue. */ - if (unlikely(!atomic_dec_and_test(&conn->tx_pushing))) + if (unlikely(!atomic_dec_and_test(&conn->tx_pushing))) { + atomic_set(&conn->tx_pushing, 1); + smp_wmb(); /* Make sure tx_pushing is 1 before real send */ goto again; + } return rc; }