From patchwork Mon Dec 5 06:57:52 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Santosh Shilimkar X-Patchwork-Id: 702584 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming@ozlabs.org Delivered-To: patchwork-incoming@ozlabs.org Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 3tXFxj4XcDz9t1h for ; Mon, 5 Dec 2016 17:59:53 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752217AbcLEG7j (ORCPT ); Mon, 5 Dec 2016 01:59:39 -0500 Received: from userp1040.oracle.com ([156.151.31.81]:29317 "EHLO userp1040.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751847AbcLEG6T (ORCPT ); Mon, 5 Dec 2016 01:58:19 -0500 Received: from aserv0021.oracle.com (aserv0021.oracle.com [141.146.126.233]) by userp1040.oracle.com (Sentrion-MTA-4.3.2/Sentrion-MTA-4.3.2) with ESMTP id uB56wCPv022957 (version=TLSv1 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Mon, 5 Dec 2016 06:58:13 GMT Received: from aserv0122.oracle.com (aserv0122.oracle.com [141.146.126.236]) by aserv0021.oracle.com (8.13.8/8.14.4) with ESMTP id uB56wCdK007372 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=OK); Mon, 5 Dec 2016 06:58:12 GMT Received: from abhmp0015.oracle.com (abhmp0015.oracle.com [141.146.116.21]) by aserv0122.oracle.com (8.14.4/8.14.4) with ESMTP id uB56wCeA027191; Mon, 5 Dec 2016 06:58:12 GMT Received: from localhost.localdomain (/10.159.248.90) by default (Oracle Beehive Gateway v4.0) with ESMTP ; Sun, 04 Dec 2016 22:58:12 -0800 From: Santosh Shilimkar To: netdev@vger.kernel.org, davem@davemloft.net Cc: linux-kernel@vger.kernel.org, Santosh Shilimkar Subject: [net-next][PATCH 17/18] RDS: add receive message trace used by application Date: Sun, 4 Dec 2016 22:57:52 -0800 Message-Id: <1480921073-9140-18-git-send-email-santosh.shilimkar@oracle.com> X-Mailer: git-send-email 1.9.1 In-Reply-To: <1480921073-9140-1-git-send-email-santosh.shilimkar@oracle.com> References: <1480921073-9140-1-git-send-email-santosh.shilimkar@oracle.com> X-Source-IP: aserv0021.oracle.com [141.146.126.233] Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Socket option to tap receive path latency in various stages in nano seconds. It can be enabled on selective sockets using using SO_RDS_MSG_RXPATH_LATENCY socket option. RDS will return the data to application with RDS_CMSG_RXPATH_LATENCY in defined format. Scope is left to add more trace points for future without need of change in the interface. Reviewed-by: Sowmini Varadhan Signed-off-by: Santosh Shilimkar --- include/uapi/linux/rds.h | 33 +++++++++++++++++++++++++++++++++ net/rds/af_rds.c | 28 ++++++++++++++++++++++++++++ net/rds/ib_recv.c | 4 ++++ net/rds/rds.h | 10 ++++++++++ net/rds/recv.c | 32 +++++++++++++++++++++++++++++--- net/rds/tcp_recv.c | 5 +++++ 6 files changed, 109 insertions(+), 3 deletions(-) diff --git a/include/uapi/linux/rds.h b/include/uapi/linux/rds.h index 0f9265c..3833113 100644 --- a/include/uapi/linux/rds.h +++ b/include/uapi/linux/rds.h @@ -52,6 +52,13 @@ #define RDS_GET_MR_FOR_DEST 7 #define SO_RDS_TRANSPORT 8 +/* Socket option to tap receive path latency + * SO_RDS: SO_RDS_MSG_RXPATH_LATENCY + * Format used struct rds_rx_trace_so + */ +#define SO_RDS_MSG_RXPATH_LATENCY 10 + + /* supported values for SO_RDS_TRANSPORT */ #define RDS_TRANS_IB 0 #define RDS_TRANS_IWARP 1 @@ -77,6 +84,12 @@ * the same as for the GET_MR setsockopt. * RDS_CMSG_RDMA_STATUS (recvmsg) * Returns the status of a completed RDMA operation. + * RDS_CMSG_RXPATH_LATENCY(recvmsg) + * Returns rds message latencies in various stages of receive + * path in nS. Its set per socket using SO_RDS_MSG_RXPATH_LATENCY + * socket option. Legitimate points are defined in + * enum rds_message_rxpath_latency. More points can be added in + * future. CSMG format is struct rds_cmsg_rx_trace. */ #define RDS_CMSG_RDMA_ARGS 1 #define RDS_CMSG_RDMA_DEST 2 @@ -87,6 +100,7 @@ #define RDS_CMSG_ATOMIC_CSWP 7 #define RDS_CMSG_MASKED_ATOMIC_FADD 8 #define RDS_CMSG_MASKED_ATOMIC_CSWP 9 +#define RDS_CMSG_RXPATH_LATENCY 11 #define RDS_INFO_FIRST 10000 #define RDS_INFO_COUNTERS 10000 @@ -171,6 +185,25 @@ struct rds_info_rdma_connection { uint32_t rdma_mr_size; }; +/* RDS message Receive Path Latency points */ +enum rds_message_rxpath_latency { + RDS_MSG_RX_HDR_TO_DGRAM_START = 0, + RDS_MSG_RX_DGRAM_REASSEMBLE, + RDS_MSG_RX_DGRAM_DELIVERED, + RDS_MSG_RX_DGRAM_TRACE_MAX +}; + +struct rds_rx_trace_so { + u8 rx_traces; + u8 rx_trace_pos[RDS_MSG_RX_DGRAM_TRACE_MAX]; +}; + +struct rds_cmsg_rx_trace { + u8 rx_traces; + u8 rx_trace_pos[RDS_MSG_RX_DGRAM_TRACE_MAX]; + u64 rx_trace[RDS_MSG_RX_DGRAM_TRACE_MAX]; +}; + /* * Congestion monitoring. * Congestion control in RDS happens at the host connection diff --git a/net/rds/af_rds.c b/net/rds/af_rds.c index 2ac1e61..fd821740 100644 --- a/net/rds/af_rds.c +++ b/net/rds/af_rds.c @@ -298,6 +298,30 @@ static int rds_enable_recvtstamp(struct sock *sk, char __user *optval, return 0; } +static int rds_recv_track_latency(struct rds_sock *rs, char __user *optval, + int optlen) +{ + struct rds_rx_trace_so trace; + int i; + + if (optlen != sizeof(struct rds_rx_trace_so)) + return -EFAULT; + + if (copy_from_user(&trace, optval, sizeof(trace))) + return -EFAULT; + + rs->rs_rx_traces = trace.rx_traces; + for (i = 0; i < rs->rs_rx_traces; i++) { + if (trace.rx_trace_pos[i] > RDS_MSG_RX_DGRAM_TRACE_MAX) { + rs->rs_rx_traces = 0; + return -EFAULT; + } + rs->rs_rx_trace[i] = trace.rx_trace_pos[i]; + } + + return 0; +} + static int rds_setsockopt(struct socket *sock, int level, int optname, char __user *optval, unsigned int optlen) { @@ -338,6 +362,9 @@ static int rds_setsockopt(struct socket *sock, int level, int optname, ret = rds_enable_recvtstamp(sock->sk, optval, optlen); release_sock(sock->sk); break; + case SO_RDS_MSG_RXPATH_LATENCY: + ret = rds_recv_track_latency(rs, optval, optlen); + break; default: ret = -ENOPROTOOPT; } @@ -484,6 +511,7 @@ static int __rds_create(struct socket *sock, struct sock *sk, int protocol) INIT_LIST_HEAD(&rs->rs_cong_list); spin_lock_init(&rs->rs_rdma_lock); rs->rs_rdma_keys = RB_ROOT; + rs->rs_rx_traces = 0; spin_lock_bh(&rds_sock_lock); list_add_tail(&rs->rs_item, &rds_sock_list); diff --git a/net/rds/ib_recv.c b/net/rds/ib_recv.c index 4b0f126..e10624a 100644 --- a/net/rds/ib_recv.c +++ b/net/rds/ib_recv.c @@ -911,8 +911,12 @@ static void rds_ib_process_recv(struct rds_connection *conn, ic->i_ibinc = ibinc; hdr = &ibinc->ii_inc.i_hdr; + ibinc->ii_inc.i_rx_lat_trace[RDS_MSG_RX_HDR] = + local_clock(); memcpy(hdr, ihdr, sizeof(*hdr)); ic->i_recv_data_rem = be32_to_cpu(hdr->h_len); + ibinc->ii_inc.i_rx_lat_trace[RDS_MSG_RX_START] = + local_clock(); rdsdebug("ic %p ibinc %p rem %u flag 0x%x\n", ic, ibinc, ic->i_recv_data_rem, hdr->h_flags); diff --git a/net/rds/rds.h b/net/rds/rds.h index f713194..07fff73 100644 --- a/net/rds/rds.h +++ b/net/rds/rds.h @@ -253,6 +253,11 @@ struct rds_ext_header_rdma_dest { #define RDS_EXTHDR_GEN_NUM 6 #define __RDS_EXTHDR_MAX 16 /* for now */ +#define RDS_RX_MAX_TRACES (RDS_MSG_RX_DGRAM_TRACE_MAX + 1) +#define RDS_MSG_RX_HDR 0 +#define RDS_MSG_RX_START 1 +#define RDS_MSG_RX_END 2 +#define RDS_MSG_RX_CMSG 3 struct rds_incoming { atomic_t i_refcount; @@ -265,6 +270,7 @@ struct rds_incoming { rds_rdma_cookie_t i_rdma_cookie; struct timeval i_rx_tstamp; + u64 i_rx_lat_trace[RDS_RX_MAX_TRACES]; }; struct rds_mr { @@ -575,6 +581,10 @@ struct rds_sock { unsigned char rs_recverr, rs_cong_monitor; u32 rs_hash_initval; + + /* Socket receive path trace points*/ + u8 rs_rx_traces; + u8 rs_rx_trace[RDS_MSG_RX_DGRAM_TRACE_MAX]; }; static inline struct rds_sock *rds_sk_to_rs(const struct sock *sk) diff --git a/net/rds/recv.c b/net/rds/recv.c index ba19eee..8b7e7b7 100644 --- a/net/rds/recv.c +++ b/net/rds/recv.c @@ -43,6 +43,8 @@ void rds_inc_init(struct rds_incoming *inc, struct rds_connection *conn, __be32 saddr) { + int i; + atomic_set(&inc->i_refcount, 1); INIT_LIST_HEAD(&inc->i_item); inc->i_conn = conn; @@ -50,6 +52,9 @@ void rds_inc_init(struct rds_incoming *inc, struct rds_connection *conn, inc->i_rdma_cookie = 0; inc->i_rx_tstamp.tv_sec = 0; inc->i_rx_tstamp.tv_usec = 0; + + for (i = 0; i < RDS_RX_MAX_TRACES; i++) + inc->i_rx_lat_trace[i] = 0; } EXPORT_SYMBOL_GPL(rds_inc_init); @@ -373,6 +378,7 @@ void rds_recv_incoming(struct rds_connection *conn, __be32 saddr, __be32 daddr, if (sock_flag(sk, SOCK_RCVTSTAMP)) do_gettimeofday(&inc->i_rx_tstamp); rds_inc_addref(inc); + inc->i_rx_lat_trace[RDS_MSG_RX_END] = local_clock(); list_add_tail(&inc->i_item, &rs->rs_recv_queue); __rds_wake_sk_sleep(sk); } else { @@ -534,7 +540,7 @@ static int rds_cmsg_recv(struct rds_incoming *inc, struct msghdr *msg, ret = put_cmsg(msg, SOL_RDS, RDS_CMSG_RDMA_DEST, sizeof(inc->i_rdma_cookie), &inc->i_rdma_cookie); if (ret) - return ret; + goto out; } if ((inc->i_rx_tstamp.tv_sec != 0) && @@ -543,10 +549,30 @@ static int rds_cmsg_recv(struct rds_incoming *inc, struct msghdr *msg, sizeof(struct timeval), &inc->i_rx_tstamp); if (ret) - return ret; + goto out; } - return 0; + if (rs->rs_rx_traces) { + struct rds_cmsg_rx_trace t; + int i, j; + + inc->i_rx_lat_trace[RDS_MSG_RX_CMSG] = local_clock(); + t.rx_traces = rs->rs_rx_traces; + for (i = 0; i < rs->rs_rx_traces; i++) { + j = rs->rs_rx_trace[i]; + t.rx_trace_pos[i] = j; + t.rx_trace[i] = inc->i_rx_lat_trace[j + 1] - + inc->i_rx_lat_trace[j]; + } + + ret = put_cmsg(msg, SOL_RDS, RDS_CMSG_RXPATH_LATENCY, + sizeof(t), &t); + if (ret) + goto out; + } + +out: + return ret; } int rds_recvmsg(struct socket *sock, struct msghdr *msg, size_t size, diff --git a/net/rds/tcp_recv.c b/net/rds/tcp_recv.c index ad4892e..e006ef8 100644 --- a/net/rds/tcp_recv.c +++ b/net/rds/tcp_recv.c @@ -180,6 +180,9 @@ static int rds_tcp_data_recv(read_descriptor_t *desc, struct sk_buff *skb, rdsdebug("alloced tinc %p\n", tinc); rds_inc_path_init(&tinc->ti_inc, cp, cp->cp_conn->c_faddr); + tinc->ti_inc.i_rx_lat_trace[RDS_MSG_RX_HDR] = + local_clock(); + /* * XXX * we might be able to use the __ variants when * we've already serialized at a higher level. @@ -204,6 +207,8 @@ static int rds_tcp_data_recv(read_descriptor_t *desc, struct sk_buff *skb, /* could be 0 for a 0 len message */ tc->t_tinc_data_rem = be32_to_cpu(tinc->ti_inc.i_hdr.h_len); + tinc->ti_inc.i_rx_lat_trace[RDS_MSG_RX_START] = + local_clock(); } }