From patchwork Mon Mar 27 03:28:17 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Wen Gu X-Patchwork-Id: 13188560 X-Patchwork-Delegate: kuba@kernel.org Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 1BDFCC7619A for ; Mon, 27 Mar 2023 03:29:36 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S232215AbjC0D3c (ORCPT ); Sun, 26 Mar 2023 23:29:32 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:33544 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S232186AbjC0D2r (ORCPT ); Sun, 26 Mar 2023 23:28:47 -0400 Received: from out30-113.freemail.mail.aliyun.com (out30-113.freemail.mail.aliyun.com [115.124.30.113]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 373F9527C; Sun, 26 Mar 2023 20:28:37 -0700 (PDT) X-Alimail-AntiSpam: AC=PASS;BC=-1|-1;BR=01201311R131e4;CH=green;DM=||false|;DS=||;FP=0|-1|-1|-1|0|-1|-1|-1;HT=ay29a033018045168;MF=guwen@linux.alibaba.com;NM=1;PH=DS;RN=11;SR=0;TI=SMTPD_---0Vef4Ar5_1679887713; Received: from localhost(mailfrom:guwen@linux.alibaba.com fp:SMTPD_---0Vef4Ar5_1679887713) by smtp.aliyun-inc.com; Mon, 27 Mar 2023 11:28:35 +0800 From: Wen Gu To: kgraul@linux.ibm.com, wenjia@linux.ibm.com, jaka@linux.ibm.com, wintera@linux.ibm.com, davem@davemloft.net, edumazet@google.com, kuba@kernel.org, pabeni@redhat.com Cc: linux-s390@vger.kernel.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org Subject: [RFC PATCH net-next v4 7/9] net/smc: Avoid data copy from sndbuf to peer RMB in SMC-D Date: Mon, 27 Mar 2023 11:28:17 +0800 Message-Id: <1679887699-54797-8-git-send-email-guwen@linux.alibaba.com> X-Mailer: git-send-email 1.8.3.1 In-Reply-To: <1679887699-54797-1-git-send-email-guwen@linux.alibaba.com> References: <1679887699-54797-1-git-send-email-guwen@linux.alibaba.com> Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: kuba@kernel.org X-Patchwork-State: RFC This patch aims to avoid data copy from local sndbuf to peer RMB by attaching local sndbuf to peer RMB when DMBs have ISM_DMB_MAPPABLE attribute. After this, local sndbuf and peer RMB share the same physical memory. +----------+ +----------+ | socket A | | socket B | +----------+ +----------+ | ^ | +---------+ | regard as | | ----------| local sndbuf | B's | regard as | | RMB | local RMB |-------> | | +---------+ 1. Actions on local RMB. a. Create or reuse RMB when connection is created; b. Unuse RMB when connection is freed; c. Free RMB when link group is freed; 2. Actions on local sndbuf. a. Attach local sndbuf to peer RMB by the rtoken exchanged through CLC message. Since then, accessing local sndbuf is equivalent to accessing peer RMB b. sndbuf_desc is exclusive to specific connection and won't be added to lgr buffer pool for reuse. c. Local sndbuf is detached from peer RMB and freed when connection is freed. Therefore, the data written to local sndbuf will directly reach peer RMB. Signed-off-by: Wen Gu --- net/smc/af_smc.c | 14 +++++++++++ net/smc/smc_core.c | 70 +++++++++++++++++++++++++++++++++++++++++++++++++++++- net/smc/smc_core.h | 1 + 3 files changed, 84 insertions(+), 1 deletion(-) diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index c91600a..212d1b1 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -1378,6 +1378,12 @@ static int smc_connect_ism(struct smc_sock *smc, } smc_conn_save_peer_info(smc, aclc); + + if (smc_ism_dmb_mappable(smc->conn.lgr->smcd)) { + rc = smcd_buf_attach(smc); + if (rc) + goto connect_abort; + } smc_close_init(smc); smc_rx_init(smc); smc_tx_init(smc); @@ -2436,6 +2442,14 @@ static void smc_listen_work(struct work_struct *work) mutex_unlock(&smc_server_lgr_pending); } smc_conn_save_peer_info(new_smc, cclc); + + if (ini->is_smcd && + smc_ism_dmb_mappable(new_smc->conn.lgr->smcd)) { + rc = smcd_buf_attach(new_smc); + if (rc) + goto out_decl; + } + smc_listen_out_connected(new_smc); SMC_STAT_SERV_SUCC_INC(sock_net(newclcsock->sk), ini); goto out_free; diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c index 4543567..0fa26cc 100644 --- a/net/smc/smc_core.c +++ b/net/smc/smc_core.c @@ -1130,6 +1130,20 @@ static void smcr_buf_unuse(struct smc_buf_desc *buf_desc, bool is_rmb, } } +static void smcd_buf_detach(struct smc_connection *conn) +{ + struct smcd_dev *smcd = conn->lgr->smcd; + u64 peer_token = conn->peer_token; + + if (!conn->sndbuf_desc) + return; + + smc_ism_detach_dmb(smcd, peer_token); + + kfree(conn->sndbuf_desc); + conn->sndbuf_desc = NULL; +} + static void smc_buf_unuse(struct smc_connection *conn, struct smc_link_group *lgr) { @@ -1174,6 +1188,10 @@ void smc_conn_free(struct smc_connection *conn) if (!list_empty(&lgr->list)) smc_ism_unset_conn(conn); tasklet_kill(&conn->rx_tsklet); + + /* detach sndbuf from peer RMB */ + if (smc_ism_dmb_mappable(lgr->smcd)) + smcd_buf_detach(conn); } else { smc_cdc_wait_pend_tx_wr(conn); if (current_work() != &conn->abort_work) @@ -2425,15 +2443,23 @@ void smc_rmb_sync_sg_for_cpu(struct smc_connection *conn) */ int smc_buf_create(struct smc_sock *smc, bool is_smcd) { + bool sndbuf_created = false; int rc; + if (is_smcd && + smc_ism_dmb_mappable(smc->conn.lgr->smcd)) + goto create_rmb; + /* create send buffer */ rc = __smc_buf_create(smc, is_smcd, false); if (rc) return rc; + sndbuf_created = true; + +create_rmb: /* create rmb */ rc = __smc_buf_create(smc, is_smcd, true); - if (rc) { + if (rc && sndbuf_created) { down_write(&smc->conn.lgr->sndbufs_lock); list_del(&smc->conn.sndbuf_desc->list); up_write(&smc->conn.lgr->sndbufs_lock); @@ -2443,6 +2469,48 @@ int smc_buf_create(struct smc_sock *smc, bool is_smcd) return rc; } +int smcd_buf_attach(struct smc_sock *smc) +{ + struct smc_connection *conn = &smc->conn; + struct smcd_dev *smcd = conn->lgr->smcd; + u64 peer_token = conn->peer_token; + struct smc_buf_desc *buf_desc; + int rc; + + buf_desc = kzalloc(sizeof(*buf_desc), GFP_KERNEL); + if (!buf_desc) + return -ENOMEM; + + /* map local sndbuf desc to peer RMB, so operations on local + * sndbuf are equivalent to operations on peer RMB. + */ + rc = smc_ism_attach_dmb(smcd, peer_token, buf_desc); + if (rc) { + rc = SMC_CLC_DECL_MEM; + goto free; + } + + smc->sk.sk_sndbuf = buf_desc->len; + buf_desc->cpu_addr = (u8 *)buf_desc->cpu_addr + sizeof(struct smcd_cdc_msg); + buf_desc->len -= sizeof(struct smcd_cdc_msg); + conn->sndbuf_desc = buf_desc; + conn->sndbuf_desc->used = 1; + atomic_set(&conn->sndbuf_space, conn->sndbuf_desc->len); + return 0; + +free: + if (conn->rmb_desc) { + /* free local RMB as well */ + down_write(&conn->lgr->rmbs_lock); + list_del(&conn->rmb_desc->list); + up_write(&conn->lgr->rmbs_lock); + smc_buf_free(conn->lgr, true, conn->rmb_desc); + conn->rmb_desc = NULL; + } + kfree(buf_desc); + return rc; +} + static inline int smc_rmb_reserve_rtoken_idx(struct smc_link_group *lgr) { int i; diff --git a/net/smc/smc_core.h b/net/smc/smc_core.h index 1645fba..e52cf70 100644 --- a/net/smc/smc_core.h +++ b/net/smc/smc_core.h @@ -524,6 +524,7 @@ void smc_smcd_terminate(struct smcd_dev *dev, u64 peer_gid, void smc_smcd_terminate_all(struct smcd_dev *dev); void smc_smcr_terminate_all(struct smc_ib_device *smcibdev); int smc_buf_create(struct smc_sock *smc, bool is_smcd); +int smcd_buf_attach(struct smc_sock *smc); int smc_uncompress_bufsize(u8 compressed); int smc_rmb_rtoken_handling(struct smc_connection *conn, struct smc_link *link, struct smc_clc_msg_accept_confirm *clc);