^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0-or-later
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /* -*- mode: c; c-basic-offset: 8; -*-
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) * vim: noexpandtab sw=8 ts=8 sts=0:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) * Copyright (C) 2004 Oracle. All rights reserved.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8) * ----
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) * Callers for this were originally written against a very simple synchronus
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) * API. This implementation reflects those simple callers. Some day I'm sure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) * we'll need to move to a more robust posting/callback mechanism.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) * Transmit calls pass in kernel virtual addresses and block copying this into
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) * the socket's tx buffers via a usual blocking sendmsg. They'll block waiting
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) * for a failed socket to timeout. TX callers can also pass in a poniter to an
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) * 'int' which gets filled with an errno off the wire in response to the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) * message they send.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) * Handlers for unsolicited messages are registered. Each socket has a page
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) * that incoming data is copied into. First the header, then the data.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) * Handlers are called from only one thread with a reference to this per-socket
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) * page. This page is destroyed after the handler call, so it can't be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) * referenced beyond the call. Handlers may block but are discouraged from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) * doing so.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) * Any framing errors (bad magic, large payload lengths) close a connection.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) * Our sock_container holds the state we associate with a socket. It's current
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) * framing state is held there as well as the refcounting we do around when it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) * is safe to tear down the socket. The socket is only finally torn down from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) * the container when the container loses all of its references -- so as long
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) * as you hold a ref on the container you can trust that the socket is valid
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) * for use with kernel socket APIs.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) * Connections are initiated between a pair of nodes when the node with the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) * higher node number gets a heartbeat callback which indicates that the lower
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) * numbered node has started heartbeating. The lower numbered node is passive
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) * and only accepts the connection if the higher numbered node is heartbeating.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42) #include <linux/kernel.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) #include <linux/sched/mm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) #include <linux/jiffies.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) #include <linux/slab.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) #include <linux/idr.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) #include <linux/kref.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) #include <linux/net.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) #include <linux/export.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) #include <net/tcp.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) #include <linux/uaccess.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) #include "heartbeat.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) #include "tcp.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) #include "nodemanager.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) #define MLOG_MASK_PREFIX ML_TCP
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) #include "masklog.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) #include "quorum.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) #include "tcp_internal.h"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) #define SC_NODEF_FMT "node %s (num %u) at %pI4:%u"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) #define SC_NODEF_ARGS(sc) sc->sc_node->nd_name, sc->sc_node->nd_num, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) &sc->sc_node->nd_ipv4_address, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) ntohs(sc->sc_node->nd_ipv4_port)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) * In the following two log macros, the whitespace after the ',' just
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) * before ##args is intentional. Otherwise, gcc 2.95 will eat the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) * previous token if args expands to nothing.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) #define msglog(hdr, fmt, args...) do { \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) typeof(hdr) __hdr = (hdr); \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) mlog(ML_MSG, "[mag %u len %u typ %u stat %d sys_stat %d " \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) "key %08x num %u] " fmt, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) be16_to_cpu(__hdr->magic), be16_to_cpu(__hdr->data_len), \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) be16_to_cpu(__hdr->msg_type), be32_to_cpu(__hdr->status), \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) be32_to_cpu(__hdr->sys_status), be32_to_cpu(__hdr->key), \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80) be32_to_cpu(__hdr->msg_num) , ##args); \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) } while (0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) #define sclog(sc, fmt, args...) do { \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) typeof(sc) __sc = (sc); \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) mlog(ML_SOCKET, "[sc %p refs %d sock %p node %u page %p " \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) "pg_off %zu] " fmt, __sc, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) kref_read(&__sc->sc_kref), __sc->sc_sock, \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) __sc->sc_node->nd_num, __sc->sc_page, __sc->sc_page_off , \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) ##args); \
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) } while (0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) static DEFINE_RWLOCK(o2net_handler_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) static struct rb_root o2net_handler_tree = RB_ROOT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) static struct o2net_node o2net_nodes[O2NM_MAX_NODES];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) /* XXX someday we'll need better accounting */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) static struct socket *o2net_listen_sock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) * listen work is only queued by the listening socket callbacks on the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) * o2net_wq. teardown detaches the callbacks before destroying the workqueue.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) * quorum work is queued as sock containers are shutdown.. stop_listening
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) * tears down all the node's sock containers, preventing future shutdowns
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) * and queued quroum work, before canceling delayed quorum work and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) * destroying the work queue.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) static struct workqueue_struct *o2net_wq;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) static struct work_struct o2net_listen_work;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) static struct o2hb_callback_func o2net_hb_up, o2net_hb_down;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) #define O2NET_HB_PRI 0x1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) static struct o2net_handshake *o2net_hand;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) static struct o2net_msg *o2net_keep_req, *o2net_keep_resp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) static int o2net_sys_err_translations[O2NET_ERR_MAX] =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) {[O2NET_ERR_NONE] = 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) [O2NET_ERR_NO_HNDLR] = -ENOPROTOOPT,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) [O2NET_ERR_OVERFLOW] = -EOVERFLOW,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) [O2NET_ERR_DIED] = -EHOSTDOWN,};
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) /* can't quite avoid *all* internal declarations :/ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) static void o2net_sc_connect_completed(struct work_struct *work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) static void o2net_rx_until_empty(struct work_struct *work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) static void o2net_shutdown_sc(struct work_struct *work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) static void o2net_listen_data_ready(struct sock *sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) static void o2net_sc_send_keep_req(struct work_struct *work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) static void o2net_idle_timer(struct timer_list *t);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) static void o2net_sc_postpone_idle(struct o2net_sock_container *sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) static void o2net_sc_reset_idle_timer(struct o2net_sock_container *sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) #ifdef CONFIG_DEBUG_FS
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) static void o2net_init_nst(struct o2net_send_tracking *nst, u32 msgtype,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) u32 msgkey, struct task_struct *task, u8 node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) INIT_LIST_HEAD(&nst->st_net_debug_item);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) nst->st_task = task;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) nst->st_msg_type = msgtype;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) nst->st_msg_key = msgkey;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) nst->st_node = node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) static inline void o2net_set_nst_sock_time(struct o2net_send_tracking *nst)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) nst->st_sock_time = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) static inline void o2net_set_nst_send_time(struct o2net_send_tracking *nst)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) nst->st_send_time = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) static inline void o2net_set_nst_status_time(struct o2net_send_tracking *nst)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156) nst->st_status_time = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) static inline void o2net_set_nst_sock_container(struct o2net_send_tracking *nst,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) nst->st_sc = sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) static inline void o2net_set_nst_msg_id(struct o2net_send_tracking *nst,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) u32 msg_id)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) nst->st_id = msg_id;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) static inline void o2net_set_sock_timer(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) sc->sc_tv_timer = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) static inline void o2net_set_data_ready_time(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) sc->sc_tv_data_ready = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181) static inline void o2net_set_advance_start_time(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) sc->sc_tv_advance_start = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) static inline void o2net_set_advance_stop_time(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) sc->sc_tv_advance_stop = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) static inline void o2net_set_func_start_time(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) sc->sc_tv_func_start = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) static inline void o2net_set_func_stop_time(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) sc->sc_tv_func_stop = ktime_get();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) #else /* CONFIG_DEBUG_FS */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) # define o2net_init_nst(a, b, c, d, e)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) # define o2net_set_nst_sock_time(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) # define o2net_set_nst_send_time(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) # define o2net_set_nst_status_time(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) # define o2net_set_nst_sock_container(a, b)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207) # define o2net_set_nst_msg_id(a, b)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) # define o2net_set_sock_timer(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) # define o2net_set_data_ready_time(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) # define o2net_set_advance_start_time(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) # define o2net_set_advance_stop_time(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) # define o2net_set_func_start_time(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) # define o2net_set_func_stop_time(a)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) #endif /* CONFIG_DEBUG_FS */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) #ifdef CONFIG_OCFS2_FS_STATS
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) static ktime_t o2net_get_func_run_time(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) return ktime_sub(sc->sc_tv_func_stop, sc->sc_tv_func_start);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) static void o2net_update_send_stats(struct o2net_send_tracking *nst,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) sc->sc_tv_status_total = ktime_add(sc->sc_tv_status_total,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) ktime_sub(ktime_get(),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) nst->st_status_time));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) sc->sc_tv_send_total = ktime_add(sc->sc_tv_send_total,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) ktime_sub(nst->st_status_time,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) nst->st_send_time));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) sc->sc_tv_acquiry_total = ktime_add(sc->sc_tv_acquiry_total,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) ktime_sub(nst->st_send_time,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) nst->st_sock_time));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) sc->sc_send_count++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) static void o2net_update_recv_stats(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) sc->sc_tv_process_total = ktime_add(sc->sc_tv_process_total,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) o2net_get_func_run_time(sc));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) sc->sc_recv_count++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) # define o2net_update_send_stats(a, b)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) # define o2net_update_recv_stats(sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) #endif /* CONFIG_OCFS2_FS_STATS */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) static inline unsigned int o2net_reconnect_delay(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) return o2nm_single_cluster->cl_reconnect_delay_ms;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) static inline unsigned int o2net_keepalive_delay(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) return o2nm_single_cluster->cl_keepalive_delay_ms;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) static inline unsigned int o2net_idle_timeout(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) return o2nm_single_cluster->cl_idle_timeout_ms;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) static inline int o2net_sys_err_to_errno(enum o2net_system_error err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) int trans;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) BUG_ON(err >= O2NET_ERR_MAX);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) trans = o2net_sys_err_translations[err];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) /* Just in case we mess up the translation table above */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) BUG_ON(err != O2NET_ERR_NONE && trans == 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) return trans;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) static struct o2net_node * o2net_nn_from_num(u8 node_num)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) BUG_ON(node_num >= ARRAY_SIZE(o2net_nodes));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) return &o2net_nodes[node_num];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) static u8 o2net_num_from_nn(struct o2net_node *nn)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) BUG_ON(nn == NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) return nn - o2net_nodes;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) static int o2net_prep_nsw(struct o2net_node *nn, struct o2net_status_wait *nsw)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) ret = idr_alloc(&nn->nn_status_idr, nsw, 0, 0, GFP_ATOMIC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298) if (ret >= 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) nsw->ns_id = ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) list_add_tail(&nsw->ns_node_item, &nn->nn_status_list);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) init_waitqueue_head(&nsw->ns_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) nsw->ns_sys_status = O2NET_ERR_NONE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) nsw->ns_status = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) static void o2net_complete_nsw_locked(struct o2net_node *nn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) struct o2net_status_wait *nsw,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) enum o2net_system_error sys_status,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) s32 status)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) assert_spin_locked(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) if (!list_empty(&nsw->ns_node_item)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) list_del_init(&nsw->ns_node_item);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) nsw->ns_sys_status = sys_status;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) nsw->ns_status = status;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) idr_remove(&nn->nn_status_idr, nsw->ns_id);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) wake_up(&nsw->ns_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) static void o2net_complete_nsw(struct o2net_node *nn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) struct o2net_status_wait *nsw,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) u64 id, enum o2net_system_error sys_status,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) s32 status)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) if (nsw == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) if (id > INT_MAX)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) nsw = idr_find(&nn->nn_status_idr, id);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) if (nsw == NULL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) o2net_complete_nsw_locked(nn, nsw, sys_status, status);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) static void o2net_complete_nodes_nsw(struct o2net_node *nn)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) struct o2net_status_wait *nsw, *tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) unsigned int num_kills = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) assert_spin_locked(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357) list_for_each_entry_safe(nsw, tmp, &nn->nn_status_list, ns_node_item) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) o2net_complete_nsw_locked(nn, nsw, O2NET_ERR_DIED, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) num_kills++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) mlog(0, "completed %d messages for node %u\n", num_kills,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) o2net_num_from_nn(nn));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) static int o2net_nsw_completed(struct o2net_node *nn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) struct o2net_status_wait *nsw)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) int completed;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) completed = list_empty(&nsw->ns_node_item);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) return completed;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) static void sc_kref_release(struct kref *kref)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) struct o2net_sock_container *sc = container_of(kref,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) struct o2net_sock_container, sc_kref);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382) BUG_ON(timer_pending(&sc->sc_idle_timeout));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384) sclog(sc, "releasing\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) if (sc->sc_sock) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387) sock_release(sc->sc_sock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388) sc->sc_sock = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) o2nm_undepend_item(&sc->sc_node->nd_item);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392) o2nm_node_put(sc->sc_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) sc->sc_node = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395) o2net_debug_del_sc(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) if (sc->sc_page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398) __free_page(sc->sc_page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) kfree(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402) static void sc_put(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404) sclog(sc, "put\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) kref_put(&sc->sc_kref, sc_kref_release);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407) static void sc_get(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) sclog(sc, "get\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410) kref_get(&sc->sc_kref);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412) static struct o2net_sock_container *sc_alloc(struct o2nm_node *node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414) struct o2net_sock_container *sc, *ret = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415) struct page *page = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416) int status = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418) page = alloc_page(GFP_NOFS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) sc = kzalloc(sizeof(*sc), GFP_NOFS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) if (sc == NULL || page == NULL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423) kref_init(&sc->sc_kref);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424) o2nm_node_get(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425) sc->sc_node = node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427) /* pin the node item of the remote node */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428) status = o2nm_depend_item(&node->nd_item);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) if (status) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) mlog_errno(status);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431) o2nm_node_put(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) INIT_WORK(&sc->sc_connect_work, o2net_sc_connect_completed);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) INIT_WORK(&sc->sc_rx_work, o2net_rx_until_empty);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436) INIT_WORK(&sc->sc_shutdown_work, o2net_shutdown_sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) INIT_DELAYED_WORK(&sc->sc_keepalive_work, o2net_sc_send_keep_req);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439) timer_setup(&sc->sc_idle_timeout, o2net_idle_timer, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) sclog(sc, "alloced\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) ret = sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444) sc->sc_page = page;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445) o2net_debug_add_sc(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446) sc = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) page = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450) if (page)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451) __free_page(page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) kfree(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459) static void o2net_sc_queue_work(struct o2net_sock_container *sc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 461) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 462) sc_get(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 463) if (!queue_work(o2net_wq, work))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466) static void o2net_sc_queue_delayed_work(struct o2net_sock_container *sc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) struct delayed_work *work,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468) int delay)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470) sc_get(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) if (!queue_delayed_work(o2net_wq, work, delay))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474) static void o2net_sc_cancel_delayed_work(struct o2net_sock_container *sc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475) struct delayed_work *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477) if (cancel_delayed_work(work))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) static atomic_t o2net_connected_peers = ATOMIC_INIT(0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 482)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 483) int o2net_num_connected_peers(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 484) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 485) return atomic_read(&o2net_connected_peers);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 486) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 487)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 488) static void o2net_set_nn_state(struct o2net_node *nn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 489) struct o2net_sock_container *sc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 490) unsigned valid, int err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 491) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 492) int was_valid = nn->nn_sc_valid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 493) int was_err = nn->nn_persistent_error;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 494) struct o2net_sock_container *old_sc = nn->nn_sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 495)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 496) assert_spin_locked(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 497)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 498) if (old_sc && !sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 499) atomic_dec(&o2net_connected_peers);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 500) else if (!old_sc && sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 501) atomic_inc(&o2net_connected_peers);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 502)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 503) /* the node num comparison and single connect/accept path should stop
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 504) * an non-null sc from being overwritten with another */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 505) BUG_ON(sc && nn->nn_sc && nn->nn_sc != sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 506) mlog_bug_on_msg(err && valid, "err %d valid %u\n", err, valid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 507) mlog_bug_on_msg(valid && !sc, "valid %u sc %p\n", valid, sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 508)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 509) if (was_valid && !valid && err == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 510) err = -ENOTCONN;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 511)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 512) mlog(ML_CONN, "node %u sc: %p -> %p, valid %u -> %u, err %d -> %d\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 513) o2net_num_from_nn(nn), nn->nn_sc, sc, nn->nn_sc_valid, valid,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 514) nn->nn_persistent_error, err);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 515)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 516) nn->nn_sc = sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 517) nn->nn_sc_valid = valid ? 1 : 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 518) nn->nn_persistent_error = err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 519)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 520) /* mirrors o2net_tx_can_proceed() */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 521) if (nn->nn_persistent_error || nn->nn_sc_valid)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 522) wake_up(&nn->nn_sc_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 523)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 524) if (was_valid && !was_err && nn->nn_persistent_error) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 525) o2quo_conn_err(o2net_num_from_nn(nn));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 526) queue_delayed_work(o2net_wq, &nn->nn_still_up,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 527) msecs_to_jiffies(O2NET_QUORUM_DELAY_MS));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 528) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 529)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 530) if (was_valid && !valid) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 531) if (old_sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 532) printk(KERN_NOTICE "o2net: No longer connected to "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 533) SC_NODEF_FMT "\n", SC_NODEF_ARGS(old_sc));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 534) o2net_complete_nodes_nsw(nn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 535) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 536)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 537) if (!was_valid && valid) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 538) o2quo_conn_up(o2net_num_from_nn(nn));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 539) cancel_delayed_work(&nn->nn_connect_expired);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 540) printk(KERN_NOTICE "o2net: %s " SC_NODEF_FMT "\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 541) o2nm_this_node() > sc->sc_node->nd_num ?
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 542) "Connected to" : "Accepted connection from",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 543) SC_NODEF_ARGS(sc));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 544) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 545)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 546) /* trigger the connecting worker func as long as we're not valid,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 547) * it will back off if it shouldn't connect. This can be called
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 548) * from node config teardown and so needs to be careful about
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 549) * the work queue actually being up. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 550) if (!valid && o2net_wq) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 551) unsigned long delay;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 552) /* delay if we're within a RECONNECT_DELAY of the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 553) * last attempt */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 554) delay = (nn->nn_last_connect_attempt +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 555) msecs_to_jiffies(o2net_reconnect_delay()))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 556) - jiffies;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 557) if (delay > msecs_to_jiffies(o2net_reconnect_delay()))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 558) delay = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 559) mlog(ML_CONN, "queueing conn attempt in %lu jiffies\n", delay);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 560) queue_delayed_work(o2net_wq, &nn->nn_connect_work, delay);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 561)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 562) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 563) * Delay the expired work after idle timeout.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 564) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 565) * We might have lots of failed connection attempts that run
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 566) * through here but we only cancel the connect_expired work when
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 567) * a connection attempt succeeds. So only the first enqueue of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 568) * the connect_expired work will do anything. The rest will see
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 569) * that it's already queued and do nothing.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 570) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 571) delay += msecs_to_jiffies(o2net_idle_timeout());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 572) queue_delayed_work(o2net_wq, &nn->nn_connect_expired, delay);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 573) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 574)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 575) /* keep track of the nn's sc ref for the caller */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 576) if ((old_sc == NULL) && sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 577) sc_get(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 578) if (old_sc && (old_sc != sc)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 579) o2net_sc_queue_work(old_sc, &old_sc->sc_shutdown_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 580) sc_put(old_sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 581) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 582) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 583)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 584) /* see o2net_register_callbacks() */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 585) static void o2net_data_ready(struct sock *sk)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 586) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 587) void (*ready)(struct sock *sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 588) struct o2net_sock_container *sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 589)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 590) read_lock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 591) sc = sk->sk_user_data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 592) if (sc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 593) sclog(sc, "data_ready hit\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 594) o2net_set_data_ready_time(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 595) o2net_sc_queue_work(sc, &sc->sc_rx_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 596) ready = sc->sc_data_ready;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 597) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 598) ready = sk->sk_data_ready;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 599) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 600) read_unlock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 601)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 602) ready(sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 603) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 604)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 605) /* see o2net_register_callbacks() */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 606) static void o2net_state_change(struct sock *sk)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 607) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 608) void (*state_change)(struct sock *sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 609) struct o2net_sock_container *sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 610)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 611) read_lock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 612) sc = sk->sk_user_data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 613) if (sc == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 614) state_change = sk->sk_state_change;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 615) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 616) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 617)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 618) sclog(sc, "state_change to %d\n", sk->sk_state);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 619)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 620) state_change = sc->sc_state_change;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 621)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 622) switch(sk->sk_state) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 623) /* ignore connecting sockets as they make progress */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 624) case TCP_SYN_SENT:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 625) case TCP_SYN_RECV:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 626) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 627) case TCP_ESTABLISHED:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 628) o2net_sc_queue_work(sc, &sc->sc_connect_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 629) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 630) default:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 631) printk(KERN_INFO "o2net: Connection to " SC_NODEF_FMT
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 632) " shutdown, state %d\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 633) SC_NODEF_ARGS(sc), sk->sk_state);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 634) o2net_sc_queue_work(sc, &sc->sc_shutdown_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 635) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 636) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 637) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 638) read_unlock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 639) state_change(sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 640) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 641)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 642) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 643) * we register callbacks so we can queue work on events before calling
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 644) * the original callbacks. our callbacks our careful to test user_data
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 645) * to discover when they've reaced with o2net_unregister_callbacks().
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 646) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 647) static void o2net_register_callbacks(struct sock *sk,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 648) struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 649) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 650) write_lock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 651)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 652) /* accepted sockets inherit the old listen socket data ready */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 653) if (sk->sk_data_ready == o2net_listen_data_ready) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 654) sk->sk_data_ready = sk->sk_user_data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 655) sk->sk_user_data = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 656) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 657)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 658) BUG_ON(sk->sk_user_data != NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 659) sk->sk_user_data = sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 660) sc_get(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 661)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 662) sc->sc_data_ready = sk->sk_data_ready;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 663) sc->sc_state_change = sk->sk_state_change;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 664) sk->sk_data_ready = o2net_data_ready;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 665) sk->sk_state_change = o2net_state_change;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 666)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 667) mutex_init(&sc->sc_send_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 668)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 669) write_unlock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 670) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 671)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 672) static int o2net_unregister_callbacks(struct sock *sk,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 673) struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 674) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 675) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 676)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 677) write_lock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 678) if (sk->sk_user_data == sc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 679) ret = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 680) sk->sk_user_data = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 681) sk->sk_data_ready = sc->sc_data_ready;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 682) sk->sk_state_change = sc->sc_state_change;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 683) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 684) write_unlock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 685)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 686) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 687) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 688)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 689) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 690) * this is a little helper that is called by callers who have seen a problem
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 691) * with an sc and want to detach it from the nn if someone already hasn't beat
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 692) * them to it. if an error is given then the shutdown will be persistent
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 693) * and pending transmits will be canceled.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 694) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 695) static void o2net_ensure_shutdown(struct o2net_node *nn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 696) struct o2net_sock_container *sc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 697) int err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 698) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 699) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 700) if (nn->nn_sc == sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 701) o2net_set_nn_state(nn, NULL, 0, err);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 702) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 703) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 704)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 705) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 706) * This work queue function performs the blocking parts of socket shutdown. A
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 707) * few paths lead here. set_nn_state will trigger this callback if it sees an
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 708) * sc detached from the nn. state_change will also trigger this callback
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 709) * directly when it sees errors. In that case we need to call set_nn_state
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 710) * ourselves as state_change couldn't get the nn_lock and call set_nn_state
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 711) * itself.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 712) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 713) static void o2net_shutdown_sc(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 714) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 715) struct o2net_sock_container *sc =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 716) container_of(work, struct o2net_sock_container,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 717) sc_shutdown_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 718) struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 719)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 720) sclog(sc, "shutting down\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 721)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 722) /* drop the callbacks ref and call shutdown only once */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 723) if (o2net_unregister_callbacks(sc->sc_sock->sk, sc)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 724) /* we shouldn't flush as we're in the thread, the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 725) * races with pending sc work structs are harmless */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 726) del_timer_sync(&sc->sc_idle_timeout);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 727) o2net_sc_cancel_delayed_work(sc, &sc->sc_keepalive_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 728) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 729) kernel_sock_shutdown(sc->sc_sock, SHUT_RDWR);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 730) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 731)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 732) /* not fatal so failed connects before the other guy has our
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 733) * heartbeat can be retried */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 734) o2net_ensure_shutdown(nn, sc, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 735) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 736) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 737)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 738) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 739)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 740) static int o2net_handler_cmp(struct o2net_msg_handler *nmh, u32 msg_type,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 741) u32 key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 742) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 743) int ret = memcmp(&nmh->nh_key, &key, sizeof(key));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 744)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 745) if (ret == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 746) ret = memcmp(&nmh->nh_msg_type, &msg_type, sizeof(msg_type));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 747)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 748) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 749) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 750)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 751) static struct o2net_msg_handler *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 752) o2net_handler_tree_lookup(u32 msg_type, u32 key, struct rb_node ***ret_p,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 753) struct rb_node **ret_parent)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 754) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 755) struct rb_node **p = &o2net_handler_tree.rb_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 756) struct rb_node *parent = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 757) struct o2net_msg_handler *nmh, *ret = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 758) int cmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 759)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 760) while (*p) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 761) parent = *p;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 762) nmh = rb_entry(parent, struct o2net_msg_handler, nh_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 763) cmp = o2net_handler_cmp(nmh, msg_type, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 764)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 765) if (cmp < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 766) p = &(*p)->rb_left;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 767) else if (cmp > 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 768) p = &(*p)->rb_right;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 769) else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 770) ret = nmh;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 771) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 772) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 773) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 774)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 775) if (ret_p != NULL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 776) *ret_p = p;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 777) if (ret_parent != NULL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 778) *ret_parent = parent;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 779)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 780) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 781) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 782)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 783) static void o2net_handler_kref_release(struct kref *kref)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 784) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 785) struct o2net_msg_handler *nmh;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 786) nmh = container_of(kref, struct o2net_msg_handler, nh_kref);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 787)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 788) kfree(nmh);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 789) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 790)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 791) static void o2net_handler_put(struct o2net_msg_handler *nmh)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 792) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 793) kref_put(&nmh->nh_kref, o2net_handler_kref_release);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 794) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 795)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 796) /* max_len is protection for the handler func. incoming messages won't
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 797) * be given to the handler if their payload is longer than the max. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 798) int o2net_register_handler(u32 msg_type, u32 key, u32 max_len,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 799) o2net_msg_handler_func *func, void *data,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 800) o2net_post_msg_handler_func *post_func,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 801) struct list_head *unreg_list)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 802) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 803) struct o2net_msg_handler *nmh = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 804) struct rb_node **p, *parent;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 805) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 806)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 807) if (max_len > O2NET_MAX_PAYLOAD_BYTES) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 808) mlog(0, "max_len for message handler out of range: %u\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 809) max_len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 810) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 811) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 812) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 813)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 814) if (!msg_type) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 815) mlog(0, "no message type provided: %u, %p\n", msg_type, func);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 816) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 817) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 818)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 819) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 820) if (!func) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 821) mlog(0, "no message handler provided: %u, %p\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 822) msg_type, func);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 823) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 824) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 825) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 826)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 827) nmh = kzalloc(sizeof(struct o2net_msg_handler), GFP_NOFS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 828) if (nmh == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 829) ret = -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 830) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 831) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 832)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 833) nmh->nh_func = func;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 834) nmh->nh_func_data = data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 835) nmh->nh_post_func = post_func;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 836) nmh->nh_msg_type = msg_type;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 837) nmh->nh_max_len = max_len;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 838) nmh->nh_key = key;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 839) /* the tree and list get this ref.. they're both removed in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 840) * unregister when this ref is dropped */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 841) kref_init(&nmh->nh_kref);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 842) INIT_LIST_HEAD(&nmh->nh_unregister_item);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 843)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 844) write_lock(&o2net_handler_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 845) if (o2net_handler_tree_lookup(msg_type, key, &p, &parent))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 846) ret = -EEXIST;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 847) else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 848) rb_link_node(&nmh->nh_node, parent, p);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 849) rb_insert_color(&nmh->nh_node, &o2net_handler_tree);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 850) list_add_tail(&nmh->nh_unregister_item, unreg_list);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 851)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 852) mlog(ML_TCP, "registered handler func %p type %u key %08x\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 853) func, msg_type, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 854) /* we've had some trouble with handlers seemingly vanishing. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 855) mlog_bug_on_msg(o2net_handler_tree_lookup(msg_type, key, &p,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 856) &parent) == NULL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 857) "couldn't find handler we *just* registered "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 858) "for type %u key %08x\n", msg_type, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 859) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 860) write_unlock(&o2net_handler_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 861)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 862) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 863) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 864) kfree(nmh);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 865)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 866) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 867) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 868) EXPORT_SYMBOL_GPL(o2net_register_handler);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 869)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 870) void o2net_unregister_handler_list(struct list_head *list)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 871) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 872) struct o2net_msg_handler *nmh, *n;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 873)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 874) write_lock(&o2net_handler_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 875) list_for_each_entry_safe(nmh, n, list, nh_unregister_item) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 876) mlog(ML_TCP, "unregistering handler func %p type %u key %08x\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 877) nmh->nh_func, nmh->nh_msg_type, nmh->nh_key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 878) rb_erase(&nmh->nh_node, &o2net_handler_tree);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 879) list_del_init(&nmh->nh_unregister_item);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 880) kref_put(&nmh->nh_kref, o2net_handler_kref_release);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 881) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 882) write_unlock(&o2net_handler_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 883) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 884) EXPORT_SYMBOL_GPL(o2net_unregister_handler_list);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 885)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 886) static struct o2net_msg_handler *o2net_handler_get(u32 msg_type, u32 key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 887) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 888) struct o2net_msg_handler *nmh;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 889)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 890) read_lock(&o2net_handler_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 891) nmh = o2net_handler_tree_lookup(msg_type, key, NULL, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 892) if (nmh)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 893) kref_get(&nmh->nh_kref);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 894) read_unlock(&o2net_handler_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 895)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 896) return nmh;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 897) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 898)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 899) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 900)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 901) static int o2net_recv_tcp_msg(struct socket *sock, void *data, size_t len)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 902) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 903) struct kvec vec = { .iov_len = len, .iov_base = data, };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 904) struct msghdr msg = { .msg_flags = MSG_DONTWAIT, };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 905) iov_iter_kvec(&msg.msg_iter, READ, &vec, 1, len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 906) return sock_recvmsg(sock, &msg, MSG_DONTWAIT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 907) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 908)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 909) static int o2net_send_tcp_msg(struct socket *sock, struct kvec *vec,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 910) size_t veclen, size_t total)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 911) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 912) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 913) struct msghdr msg = {.msg_flags = 0,};
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 914)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 915) if (sock == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 916) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 917) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 918) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 919)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 920) ret = kernel_sendmsg(sock, &msg, vec, veclen, total);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 921) if (likely(ret == total))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 922) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 923) mlog(ML_ERROR, "sendmsg returned %d instead of %zu\n", ret, total);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 924) if (ret >= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 925) ret = -EPIPE; /* should be smarter, I bet */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 926) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 927) mlog(0, "returning error: %d\n", ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 928) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 929) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 930)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 931) static void o2net_sendpage(struct o2net_sock_container *sc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 932) void *kmalloced_virt,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 933) size_t size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 934) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 935) struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 936) ssize_t ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 937)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 938) while (1) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 939) mutex_lock(&sc->sc_send_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 940) ret = sc->sc_sock->ops->sendpage(sc->sc_sock,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 941) virt_to_page(kmalloced_virt),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 942) offset_in_page(kmalloced_virt),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 943) size, MSG_DONTWAIT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 944) mutex_unlock(&sc->sc_send_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 945) if (ret == size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 946) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 947) if (ret == (ssize_t)-EAGAIN) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 948) mlog(0, "sendpage of size %zu to " SC_NODEF_FMT
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 949) " returned EAGAIN\n", size, SC_NODEF_ARGS(sc));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 950) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 951) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 952) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 953) mlog(ML_ERROR, "sendpage of size %zu to " SC_NODEF_FMT
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 954) " failed with %zd\n", size, SC_NODEF_ARGS(sc), ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 955) o2net_ensure_shutdown(nn, sc, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 956) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 957) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 958) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 959)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 960) static void o2net_init_msg(struct o2net_msg *msg, u16 data_len, u16 msg_type, u32 key)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 961) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 962) memset(msg, 0, sizeof(struct o2net_msg));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 963) msg->magic = cpu_to_be16(O2NET_MSG_MAGIC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 964) msg->data_len = cpu_to_be16(data_len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 965) msg->msg_type = cpu_to_be16(msg_type);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 966) msg->sys_status = cpu_to_be32(O2NET_ERR_NONE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 967) msg->status = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 968) msg->key = cpu_to_be32(key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 969) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 970)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 971) static int o2net_tx_can_proceed(struct o2net_node *nn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 972) struct o2net_sock_container **sc_ret,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 973) int *error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 974) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 975) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 976)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 977) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 978) if (nn->nn_persistent_error) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 979) ret = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 980) *sc_ret = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 981) *error = nn->nn_persistent_error;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 982) } else if (nn->nn_sc_valid) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 983) kref_get(&nn->nn_sc->sc_kref);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 984)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 985) ret = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 986) *sc_ret = nn->nn_sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 987) *error = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 988) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 989) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 990)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 991) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 992) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 993)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 994) /* Get a map of all nodes to which this node is currently connected to */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 995) void o2net_fill_node_map(unsigned long *map, unsigned bytes)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 996) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 997) struct o2net_sock_container *sc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 998) int node, ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 999)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1000) BUG_ON(bytes < (BITS_TO_LONGS(O2NM_MAX_NODES) * sizeof(unsigned long)));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1001)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1002) memset(map, 0, bytes);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1003) for (node = 0; node < O2NM_MAX_NODES; ++node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1004) if (!o2net_tx_can_proceed(o2net_nn_from_num(node), &sc, &ret))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1005) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1006) if (!ret) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1007) set_bit(node, map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1008) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1009) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1010) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1011) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1012) EXPORT_SYMBOL_GPL(o2net_fill_node_map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1013)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1014) int o2net_send_message_vec(u32 msg_type, u32 key, struct kvec *caller_vec,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1015) size_t caller_veclen, u8 target_node, int *status)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1016) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1017) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1018) struct o2net_msg *msg = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1019) size_t veclen, caller_bytes = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1020) struct kvec *vec = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1021) struct o2net_sock_container *sc = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1022) struct o2net_node *nn = o2net_nn_from_num(target_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1023) struct o2net_status_wait nsw = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1024) .ns_node_item = LIST_HEAD_INIT(nsw.ns_node_item),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1025) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1026) struct o2net_send_tracking nst;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1027)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1028) o2net_init_nst(&nst, msg_type, key, current, target_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1029)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1030) if (o2net_wq == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1031) mlog(0, "attempt to tx without o2netd running\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1032) ret = -ESRCH;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1033) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1034) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1035)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1036) if (caller_veclen == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1037) mlog(0, "bad kvec array length\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1038) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1039) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1040) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1041)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1042) caller_bytes = iov_length((struct iovec *)caller_vec, caller_veclen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1043) if (caller_bytes > O2NET_MAX_PAYLOAD_BYTES) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1044) mlog(0, "total payload len %zu too large\n", caller_bytes);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1045) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1046) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1047) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1048)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1049) if (target_node == o2nm_this_node()) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1050) ret = -ELOOP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1051) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1052) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1053)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1054) o2net_debug_add_nst(&nst);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1055)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1056) o2net_set_nst_sock_time(&nst);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1057)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1058) wait_event(nn->nn_sc_wq, o2net_tx_can_proceed(nn, &sc, &ret));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1059) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1060) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1061)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1062) o2net_set_nst_sock_container(&nst, sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1063)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1064) veclen = caller_veclen + 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1065) vec = kmalloc_array(veclen, sizeof(struct kvec), GFP_ATOMIC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1066) if (vec == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1067) mlog(0, "failed to %zu element kvec!\n", veclen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1068) ret = -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1069) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1070) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1071)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1072) msg = kmalloc(sizeof(struct o2net_msg), GFP_ATOMIC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1073) if (!msg) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1074) mlog(0, "failed to allocate a o2net_msg!\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1075) ret = -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1076) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1077) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1078)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1079) o2net_init_msg(msg, caller_bytes, msg_type, key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1080)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1081) vec[0].iov_len = sizeof(struct o2net_msg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1082) vec[0].iov_base = msg;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1083) memcpy(&vec[1], caller_vec, caller_veclen * sizeof(struct kvec));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1084)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1085) ret = o2net_prep_nsw(nn, &nsw);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1086) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1087) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1088)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1089) msg->msg_num = cpu_to_be32(nsw.ns_id);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1090) o2net_set_nst_msg_id(&nst, nsw.ns_id);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1091)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1092) o2net_set_nst_send_time(&nst);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1093)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1094) /* finally, convert the message header to network byte-order
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1095) * and send */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1096) mutex_lock(&sc->sc_send_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1097) ret = o2net_send_tcp_msg(sc->sc_sock, vec, veclen,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1098) sizeof(struct o2net_msg) + caller_bytes);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1099) mutex_unlock(&sc->sc_send_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1100) msglog(msg, "sending returned %d\n", ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1101) if (ret < 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1102) mlog(0, "error returned from o2net_send_tcp_msg=%d\n", ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1103) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1104) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1105)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1106) /* wait on other node's handler */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1107) o2net_set_nst_status_time(&nst);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1108) wait_event(nsw.ns_wq, o2net_nsw_completed(nn, &nsw));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1109)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1110) o2net_update_send_stats(&nst, sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1111)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1112) /* Note that we avoid overwriting the callers status return
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1113) * variable if a system error was reported on the other
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1114) * side. Callers beware. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1115) ret = o2net_sys_err_to_errno(nsw.ns_sys_status);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1116) if (status && !ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1117) *status = nsw.ns_status;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1118)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1119) mlog(0, "woken, returning system status %d, user status %d\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1120) ret, nsw.ns_status);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1121) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1122) o2net_debug_del_nst(&nst); /* must be before dropping sc and node */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1123) if (sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1124) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1125) kfree(vec);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1126) kfree(msg);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1127) o2net_complete_nsw(nn, &nsw, 0, 0, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1128) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1129) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1130) EXPORT_SYMBOL_GPL(o2net_send_message_vec);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1131)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1132) int o2net_send_message(u32 msg_type, u32 key, void *data, u32 len,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1133) u8 target_node, int *status)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1134) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1135) struct kvec vec = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1136) .iov_base = data,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1137) .iov_len = len,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1138) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1139) return o2net_send_message_vec(msg_type, key, &vec, 1,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1140) target_node, status);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1141) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1142) EXPORT_SYMBOL_GPL(o2net_send_message);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1143)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1144) static int o2net_send_status_magic(struct socket *sock, struct o2net_msg *hdr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1145) enum o2net_system_error syserr, int err)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1146) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1147) struct kvec vec = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1148) .iov_base = hdr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1149) .iov_len = sizeof(struct o2net_msg),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1150) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1151)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1152) BUG_ON(syserr >= O2NET_ERR_MAX);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1153)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1154) /* leave other fields intact from the incoming message, msg_num
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1155) * in particular */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1156) hdr->sys_status = cpu_to_be32(syserr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1157) hdr->status = cpu_to_be32(err);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1158) hdr->magic = cpu_to_be16(O2NET_MSG_STATUS_MAGIC); // twiddle the magic
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1159) hdr->data_len = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1160)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1161) msglog(hdr, "about to send status magic %d\n", err);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1162) /* hdr has been in host byteorder this whole time */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1163) return o2net_send_tcp_msg(sock, &vec, 1, sizeof(struct o2net_msg));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1164) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1165)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1166) /* this returns -errno if the header was unknown or too large, etc.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1167) * after this is called the buffer us reused for the next message */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1168) static int o2net_process_message(struct o2net_sock_container *sc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1169) struct o2net_msg *hdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1170) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1171) struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1172) int ret = 0, handler_status;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1173) enum o2net_system_error syserr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1174) struct o2net_msg_handler *nmh = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1175) void *ret_data = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1176)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1177) msglog(hdr, "processing message\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1178)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1179) o2net_sc_postpone_idle(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1180)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1181) switch(be16_to_cpu(hdr->magic)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1182) case O2NET_MSG_STATUS_MAGIC:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1183) /* special type for returning message status */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1184) o2net_complete_nsw(nn, NULL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1185) be32_to_cpu(hdr->msg_num),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1186) be32_to_cpu(hdr->sys_status),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1187) be32_to_cpu(hdr->status));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1188) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1189) case O2NET_MSG_KEEP_REQ_MAGIC:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1190) o2net_sendpage(sc, o2net_keep_resp,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1191) sizeof(*o2net_keep_resp));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1192) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1193) case O2NET_MSG_KEEP_RESP_MAGIC:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1194) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1195) case O2NET_MSG_MAGIC:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1196) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1197) default:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1198) msglog(hdr, "bad magic\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1199) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1200) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1201) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1202) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1203)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1204) /* find a handler for it */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1205) handler_status = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1206) nmh = o2net_handler_get(be16_to_cpu(hdr->msg_type),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1207) be32_to_cpu(hdr->key));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1208) if (!nmh) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1209) mlog(ML_TCP, "couldn't find handler for type %u key %08x\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1210) be16_to_cpu(hdr->msg_type), be32_to_cpu(hdr->key));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1211) syserr = O2NET_ERR_NO_HNDLR;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1212) goto out_respond;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1213) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1214)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1215) syserr = O2NET_ERR_NONE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1216)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1217) if (be16_to_cpu(hdr->data_len) > nmh->nh_max_len)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1218) syserr = O2NET_ERR_OVERFLOW;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1219)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1220) if (syserr != O2NET_ERR_NONE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1221) goto out_respond;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1222)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1223) o2net_set_func_start_time(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1224) sc->sc_msg_key = be32_to_cpu(hdr->key);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1225) sc->sc_msg_type = be16_to_cpu(hdr->msg_type);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1226) handler_status = (nmh->nh_func)(hdr, sizeof(struct o2net_msg) +
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1227) be16_to_cpu(hdr->data_len),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1228) nmh->nh_func_data, &ret_data);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1229) o2net_set_func_stop_time(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1230)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1231) o2net_update_recv_stats(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1232)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1233) out_respond:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1234) /* this destroys the hdr, so don't use it after this */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1235) mutex_lock(&sc->sc_send_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1236) ret = o2net_send_status_magic(sc->sc_sock, hdr, syserr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1237) handler_status);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1238) mutex_unlock(&sc->sc_send_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1239) hdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1240) mlog(0, "sending handler status %d, syserr %d returned %d\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1241) handler_status, syserr, ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1242)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1243) if (nmh) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1244) BUG_ON(ret_data != NULL && nmh->nh_post_func == NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1245) if (nmh->nh_post_func)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1246) (nmh->nh_post_func)(handler_status, nmh->nh_func_data,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1247) ret_data);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1248) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1249)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1250) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1251) if (nmh)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1252) o2net_handler_put(nmh);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1253) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1254) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1255)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1256) static int o2net_check_handshake(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1257) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1258) struct o2net_handshake *hand = page_address(sc->sc_page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1259) struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1260)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1261) if (hand->protocol_version != cpu_to_be64(O2NET_PROTOCOL_VERSION)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1262) printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " Advertised net "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1263) "protocol version %llu but %llu is required. "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1264) "Disconnecting.\n", SC_NODEF_ARGS(sc),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1265) (unsigned long long)be64_to_cpu(hand->protocol_version),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1266) O2NET_PROTOCOL_VERSION);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1267)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1268) /* don't bother reconnecting if its the wrong version. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1269) o2net_ensure_shutdown(nn, sc, -ENOTCONN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1270) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1271) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1272)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1273) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1274) * Ensure timeouts are consistent with other nodes, otherwise
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1275) * we can end up with one node thinking that the other must be down,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1276) * but isn't. This can ultimately cause corruption.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1277) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1278) if (be32_to_cpu(hand->o2net_idle_timeout_ms) !=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1279) o2net_idle_timeout()) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1280) printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a network "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1281) "idle timeout of %u ms, but we use %u ms locally. "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1282) "Disconnecting.\n", SC_NODEF_ARGS(sc),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1283) be32_to_cpu(hand->o2net_idle_timeout_ms),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1284) o2net_idle_timeout());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1285) o2net_ensure_shutdown(nn, sc, -ENOTCONN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1286) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1287) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1288)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1289) if (be32_to_cpu(hand->o2net_keepalive_delay_ms) !=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1290) o2net_keepalive_delay()) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1291) printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a keepalive "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1292) "delay of %u ms, but we use %u ms locally. "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1293) "Disconnecting.\n", SC_NODEF_ARGS(sc),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1294) be32_to_cpu(hand->o2net_keepalive_delay_ms),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1295) o2net_keepalive_delay());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1296) o2net_ensure_shutdown(nn, sc, -ENOTCONN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1297) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1298) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1299)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1300) if (be32_to_cpu(hand->o2hb_heartbeat_timeout_ms) !=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1301) O2HB_MAX_WRITE_TIMEOUT_MS) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1302) printk(KERN_NOTICE "o2net: " SC_NODEF_FMT " uses a heartbeat "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1303) "timeout of %u ms, but we use %u ms locally. "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1304) "Disconnecting.\n", SC_NODEF_ARGS(sc),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1305) be32_to_cpu(hand->o2hb_heartbeat_timeout_ms),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1306) O2HB_MAX_WRITE_TIMEOUT_MS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1307) o2net_ensure_shutdown(nn, sc, -ENOTCONN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1308) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1309) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1310)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1311) sc->sc_handshake_ok = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1312)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1313) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1314) /* set valid and queue the idle timers only if it hasn't been
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1315) * shut down already */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1316) if (nn->nn_sc == sc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1317) o2net_sc_reset_idle_timer(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1318) atomic_set(&nn->nn_timeout, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1319) o2net_set_nn_state(nn, sc, 1, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1320) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1321) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1322)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1323) /* shift everything up as though it wasn't there */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1324) sc->sc_page_off -= sizeof(struct o2net_handshake);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1325) if (sc->sc_page_off)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1326) memmove(hand, hand + 1, sc->sc_page_off);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1327)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1328) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1329) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1330)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1331) /* this demuxes the queued rx bytes into header or payload bits and calls
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1332) * handlers as each full message is read off the socket. it returns -error,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1333) * == 0 eof, or > 0 for progress made.*/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1334) static int o2net_advance_rx(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1335) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1336) struct o2net_msg *hdr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1337) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1338) void *data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1339) size_t datalen;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1340)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1341) sclog(sc, "receiving\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1342) o2net_set_advance_start_time(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1343)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1344) if (unlikely(sc->sc_handshake_ok == 0)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1345) if(sc->sc_page_off < sizeof(struct o2net_handshake)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1346) data = page_address(sc->sc_page) + sc->sc_page_off;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1347) datalen = sizeof(struct o2net_handshake) - sc->sc_page_off;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1348) ret = o2net_recv_tcp_msg(sc->sc_sock, data, datalen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1349) if (ret > 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1350) sc->sc_page_off += ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1351) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1352)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1353) if (sc->sc_page_off == sizeof(struct o2net_handshake)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1354) o2net_check_handshake(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1355) if (unlikely(sc->sc_handshake_ok == 0))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1356) ret = -EPROTO;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1357) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1358) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1359) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1360)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1361) /* do we need more header? */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1362) if (sc->sc_page_off < sizeof(struct o2net_msg)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1363) data = page_address(sc->sc_page) + sc->sc_page_off;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1364) datalen = sizeof(struct o2net_msg) - sc->sc_page_off;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1365) ret = o2net_recv_tcp_msg(sc->sc_sock, data, datalen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1366) if (ret > 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1367) sc->sc_page_off += ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1368) /* only swab incoming here.. we can
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1369) * only get here once as we cross from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1370) * being under to over */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1371) if (sc->sc_page_off == sizeof(struct o2net_msg)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1372) hdr = page_address(sc->sc_page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1373) if (be16_to_cpu(hdr->data_len) >
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1374) O2NET_MAX_PAYLOAD_BYTES)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1375) ret = -EOVERFLOW;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1376) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1377) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1378) if (ret <= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1379) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1380) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1381)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1382) if (sc->sc_page_off < sizeof(struct o2net_msg)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1383) /* oof, still don't have a header */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1384) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1385) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1386)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1387) /* this was swabbed above when we first read it */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1388) hdr = page_address(sc->sc_page);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1389)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1390) msglog(hdr, "at page_off %zu\n", sc->sc_page_off);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1391)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1392) /* do we need more payload? */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1393) if (sc->sc_page_off - sizeof(struct o2net_msg) < be16_to_cpu(hdr->data_len)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1394) /* need more payload */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1395) data = page_address(sc->sc_page) + sc->sc_page_off;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1396) datalen = (sizeof(struct o2net_msg) + be16_to_cpu(hdr->data_len)) -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1397) sc->sc_page_off;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1398) ret = o2net_recv_tcp_msg(sc->sc_sock, data, datalen);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1399) if (ret > 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1400) sc->sc_page_off += ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1401) if (ret <= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1402) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1403) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1404)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1405) if (sc->sc_page_off - sizeof(struct o2net_msg) == be16_to_cpu(hdr->data_len)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1406) /* we can only get here once, the first time we read
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1407) * the payload.. so set ret to progress if the handler
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1408) * works out. after calling this the message is toast */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1409) ret = o2net_process_message(sc, hdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1410) if (ret == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1411) ret = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1412) sc->sc_page_off = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1413) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1414)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1415) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1416) sclog(sc, "ret = %d\n", ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1417) o2net_set_advance_stop_time(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1418) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1419) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1420)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1421) /* this work func is triggerd by data ready. it reads until it can read no
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1422) * more. it interprets 0, eof, as fatal. if data_ready hits while we're doing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1423) * our work the work struct will be marked and we'll be called again. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1424) static void o2net_rx_until_empty(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1425) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1426) struct o2net_sock_container *sc =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1427) container_of(work, struct o2net_sock_container, sc_rx_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1428) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1429)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1430) do {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1431) ret = o2net_advance_rx(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1432) } while (ret > 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1433)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1434) if (ret <= 0 && ret != -EAGAIN) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1435) struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1436) sclog(sc, "saw error %d, closing\n", ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1437) /* not permanent so read failed handshake can retry */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1438) o2net_ensure_shutdown(nn, sc, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1439) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1440)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1441) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1442) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1443)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1444) static void o2net_initialize_handshake(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1445) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1446) o2net_hand->o2hb_heartbeat_timeout_ms = cpu_to_be32(
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1447) O2HB_MAX_WRITE_TIMEOUT_MS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1448) o2net_hand->o2net_idle_timeout_ms = cpu_to_be32(o2net_idle_timeout());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1449) o2net_hand->o2net_keepalive_delay_ms = cpu_to_be32(
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1450) o2net_keepalive_delay());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1451) o2net_hand->o2net_reconnect_delay_ms = cpu_to_be32(
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1452) o2net_reconnect_delay());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1453) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1454)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1455) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1456)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1457) /* called when a connect completes and after a sock is accepted. the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1458) * rx path will see the response and mark the sc valid */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1459) static void o2net_sc_connect_completed(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1460) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1461) struct o2net_sock_container *sc =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1462) container_of(work, struct o2net_sock_container,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1463) sc_connect_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1464)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1465) mlog(ML_MSG, "sc sending handshake with ver %llu id %llx\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1466) (unsigned long long)O2NET_PROTOCOL_VERSION,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1467) (unsigned long long)be64_to_cpu(o2net_hand->connector_id));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1468)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1469) o2net_initialize_handshake();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1470) o2net_sendpage(sc, o2net_hand, sizeof(*o2net_hand));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1471) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1472) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1473)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1474) /* this is called as a work_struct func. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1475) static void o2net_sc_send_keep_req(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1476) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1477) struct o2net_sock_container *sc =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1478) container_of(work, struct o2net_sock_container,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1479) sc_keepalive_work.work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1480)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1481) o2net_sendpage(sc, o2net_keep_req, sizeof(*o2net_keep_req));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1482) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1483) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1484)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1485) /* socket shutdown does a del_timer_sync against this as it tears down.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1486) * we can't start this timer until we've got to the point in sc buildup
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1487) * where shutdown is going to be involved */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1488) static void o2net_idle_timer(struct timer_list *t)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1489) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1490) struct o2net_sock_container *sc = from_timer(sc, t, sc_idle_timeout);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1491) struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1492) #ifdef CONFIG_DEBUG_FS
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1493) unsigned long msecs = ktime_to_ms(ktime_get()) -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1494) ktime_to_ms(sc->sc_tv_timer);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1495) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1496) unsigned long msecs = o2net_idle_timeout();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1497) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1498)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1499) printk(KERN_NOTICE "o2net: Connection to " SC_NODEF_FMT " has been "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1500) "idle for %lu.%lu secs.\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1501) SC_NODEF_ARGS(sc), msecs / 1000, msecs % 1000);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1502)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1503) /* idle timerout happen, don't shutdown the connection, but
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1504) * make fence decision. Maybe the connection can recover before
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1505) * the decision is made.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1506) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1507) atomic_set(&nn->nn_timeout, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1508) o2quo_conn_err(o2net_num_from_nn(nn));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1509) queue_delayed_work(o2net_wq, &nn->nn_still_up,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1510) msecs_to_jiffies(O2NET_QUORUM_DELAY_MS));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1511)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1512) o2net_sc_reset_idle_timer(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1513)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1514) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1515)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1516) static void o2net_sc_reset_idle_timer(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1517) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1518) o2net_sc_cancel_delayed_work(sc, &sc->sc_keepalive_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1519) o2net_sc_queue_delayed_work(sc, &sc->sc_keepalive_work,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1520) msecs_to_jiffies(o2net_keepalive_delay()));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1521) o2net_set_sock_timer(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1522) mod_timer(&sc->sc_idle_timeout,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1523) jiffies + msecs_to_jiffies(o2net_idle_timeout()));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1524) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1525)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1526) static void o2net_sc_postpone_idle(struct o2net_sock_container *sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1527) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1528) struct o2net_node *nn = o2net_nn_from_num(sc->sc_node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1529)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1530) /* clear fence decision since the connection recover from timeout*/
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1531) if (atomic_read(&nn->nn_timeout)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1532) o2quo_conn_up(o2net_num_from_nn(nn));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1533) cancel_delayed_work(&nn->nn_still_up);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1534) atomic_set(&nn->nn_timeout, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1535) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1536)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1537) /* Only push out an existing timer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1538) if (timer_pending(&sc->sc_idle_timeout))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1539) o2net_sc_reset_idle_timer(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1540) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1541)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1542) /* this work func is kicked whenever a path sets the nn state which doesn't
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1543) * have valid set. This includes seeing hb come up, losing a connection,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1544) * having a connect attempt fail, etc. This centralizes the logic which decides
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1545) * if a connect attempt should be made or if we should give up and all future
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1546) * transmit attempts should fail */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1547) static void o2net_start_connect(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1548) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1549) struct o2net_node *nn =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1550) container_of(work, struct o2net_node, nn_connect_work.work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1551) struct o2net_sock_container *sc = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1552) struct o2nm_node *node = NULL, *mynode = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1553) struct socket *sock = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1554) struct sockaddr_in myaddr = {0, }, remoteaddr = {0, };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1555) int ret = 0, stop;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1556) unsigned int timeout;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1557) unsigned int nofs_flag;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1558)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1559) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1560) * sock_create allocates the sock with GFP_KERNEL. We must
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1561) * prevent the filesystem from being reentered by memory reclaim.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1562) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1563) nofs_flag = memalloc_nofs_save();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1564) /* if we're greater we initiate tx, otherwise we accept */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1565) if (o2nm_this_node() <= o2net_num_from_nn(nn))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1566) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1567)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1568) /* watch for racing with tearing a node down */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1569) node = o2nm_get_node_by_num(o2net_num_from_nn(nn));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1570) if (node == NULL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1571) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1572)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1573) mynode = o2nm_get_node_by_num(o2nm_this_node());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1574) if (mynode == NULL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1575) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1576)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1577) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1578) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1579) * see if we already have one pending or have given up.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1580) * For nn_timeout, it is set when we close the connection
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1581) * because of the idle time out. So it means that we have
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1582) * at least connected to that node successfully once,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1583) * now try to connect to it again.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1584) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1585) timeout = atomic_read(&nn->nn_timeout);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1586) stop = (nn->nn_sc ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1587) (nn->nn_persistent_error &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1588) (nn->nn_persistent_error != -ENOTCONN || timeout == 0)));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1589) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1590) if (stop)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1591) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1592)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1593) nn->nn_last_connect_attempt = jiffies;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1594)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1595) sc = sc_alloc(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1596) if (sc == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1597) mlog(0, "couldn't allocate sc\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1598) ret = -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1599) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1600) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1601)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1602) ret = sock_create(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1603) if (ret < 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1604) mlog(0, "can't create socket: %d\n", ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1605) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1606) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1607) sc->sc_sock = sock; /* freed by sc_kref_release */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1608)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1609) sock->sk->sk_allocation = GFP_ATOMIC;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1610)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1611) myaddr.sin_family = AF_INET;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1612) myaddr.sin_addr.s_addr = mynode->nd_ipv4_address;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1613) myaddr.sin_port = htons(0); /* any port */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1614)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1615) ret = sock->ops->bind(sock, (struct sockaddr *)&myaddr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1616) sizeof(myaddr));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1617) if (ret) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1618) mlog(ML_ERROR, "bind failed with %d at address %pI4\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1619) ret, &mynode->nd_ipv4_address);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1620) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1621) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1622)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1623) tcp_sock_set_nodelay(sc->sc_sock->sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1624) tcp_sock_set_user_timeout(sock->sk, O2NET_TCP_USER_TIMEOUT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1625)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1626) o2net_register_callbacks(sc->sc_sock->sk, sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1627)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1628) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1629) /* handshake completion will set nn->nn_sc_valid */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1630) o2net_set_nn_state(nn, sc, 0, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1631) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1632)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1633) remoteaddr.sin_family = AF_INET;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1634) remoteaddr.sin_addr.s_addr = node->nd_ipv4_address;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1635) remoteaddr.sin_port = node->nd_ipv4_port;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1636)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1637) ret = sc->sc_sock->ops->connect(sc->sc_sock,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1638) (struct sockaddr *)&remoteaddr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1639) sizeof(remoteaddr),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1640) O_NONBLOCK);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1641) if (ret == -EINPROGRESS)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1642) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1643)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1644) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1645) if (ret && sc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1646) printk(KERN_NOTICE "o2net: Connect attempt to " SC_NODEF_FMT
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1647) " failed with errno %d\n", SC_NODEF_ARGS(sc), ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1648) /* 0 err so that another will be queued and attempted
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1649) * from set_nn_state */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1650) o2net_ensure_shutdown(nn, sc, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1651) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1652) if (sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1653) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1654) if (node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1655) o2nm_node_put(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1656) if (mynode)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1657) o2nm_node_put(mynode);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1658)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1659) memalloc_nofs_restore(nofs_flag);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1660) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1661) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1662)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1663) static void o2net_connect_expired(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1664) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1665) struct o2net_node *nn =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1666) container_of(work, struct o2net_node, nn_connect_expired.work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1667)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1668) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1669) if (!nn->nn_sc_valid) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1670) printk(KERN_NOTICE "o2net: No connection established with "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1671) "node %u after %u.%u seconds, check network and"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1672) " cluster configuration.\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1673) o2net_num_from_nn(nn),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1674) o2net_idle_timeout() / 1000,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1675) o2net_idle_timeout() % 1000);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1676)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1677) o2net_set_nn_state(nn, NULL, 0, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1678) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1679) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1680) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1681)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1682) static void o2net_still_up(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1683) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1684) struct o2net_node *nn =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1685) container_of(work, struct o2net_node, nn_still_up.work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1686)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1687) o2quo_hb_still_up(o2net_num_from_nn(nn));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1688) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1689)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1690) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1691)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1692) void o2net_disconnect_node(struct o2nm_node *node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1693) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1694) struct o2net_node *nn = o2net_nn_from_num(node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1695)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1696) /* don't reconnect until it's heartbeating again */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1697) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1698) atomic_set(&nn->nn_timeout, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1699) o2net_set_nn_state(nn, NULL, 0, -ENOTCONN);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1700) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1701)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1702) if (o2net_wq) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1703) cancel_delayed_work(&nn->nn_connect_expired);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1704) cancel_delayed_work(&nn->nn_connect_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1705) cancel_delayed_work(&nn->nn_still_up);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1706) flush_workqueue(o2net_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1707) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1708) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1709)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1710) static void o2net_hb_node_down_cb(struct o2nm_node *node, int node_num,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1711) void *data)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1712) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1713) o2quo_hb_down(node_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1714)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1715) if (!node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1716) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1717)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1718) if (node_num != o2nm_this_node())
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1719) o2net_disconnect_node(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1720)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1721) BUG_ON(atomic_read(&o2net_connected_peers) < 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1722) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1723)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1724) static void o2net_hb_node_up_cb(struct o2nm_node *node, int node_num,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1725) void *data)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1726) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1727) struct o2net_node *nn = o2net_nn_from_num(node_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1728)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1729) o2quo_hb_up(node_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1730)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1731) BUG_ON(!node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1732)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1733) /* ensure an immediate connect attempt */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1734) nn->nn_last_connect_attempt = jiffies -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1735) (msecs_to_jiffies(o2net_reconnect_delay()) + 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1736)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1737) if (node_num != o2nm_this_node()) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1738) /* believe it or not, accept and node heartbeating testing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1739) * can succeed for this node before we got here.. so
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1740) * only use set_nn_state to clear the persistent error
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1741) * if that hasn't already happened */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1742) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1743) atomic_set(&nn->nn_timeout, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1744) if (nn->nn_persistent_error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1745) o2net_set_nn_state(nn, NULL, 0, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1746) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1747) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1748) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1749)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1750) void o2net_unregister_hb_callbacks(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1751) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1752) o2hb_unregister_callback(NULL, &o2net_hb_up);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1753) o2hb_unregister_callback(NULL, &o2net_hb_down);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1754) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1755)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1756) int o2net_register_hb_callbacks(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1757) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1758) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1759)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1760) o2hb_setup_callback(&o2net_hb_down, O2HB_NODE_DOWN_CB,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1761) o2net_hb_node_down_cb, NULL, O2NET_HB_PRI);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1762) o2hb_setup_callback(&o2net_hb_up, O2HB_NODE_UP_CB,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1763) o2net_hb_node_up_cb, NULL, O2NET_HB_PRI);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1764)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1765) ret = o2hb_register_callback(NULL, &o2net_hb_up);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1766) if (ret == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1767) ret = o2hb_register_callback(NULL, &o2net_hb_down);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1768)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1769) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1770) o2net_unregister_hb_callbacks();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1771)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1772) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1773) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1774)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1775) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1776)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1777) static int o2net_accept_one(struct socket *sock, int *more)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1778) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1779) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1780) struct sockaddr_in sin;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1781) struct socket *new_sock = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1782) struct o2nm_node *node = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1783) struct o2nm_node *local_node = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1784) struct o2net_sock_container *sc = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1785) struct o2net_node *nn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1786) unsigned int nofs_flag;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1787)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1788) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1789) * sock_create_lite allocates the sock with GFP_KERNEL. We must
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1790) * prevent the filesystem from being reentered by memory reclaim.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1791) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1792) nofs_flag = memalloc_nofs_save();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1793)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1794) BUG_ON(sock == NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1795) *more = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1796) ret = sock_create_lite(sock->sk->sk_family, sock->sk->sk_type,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1797) sock->sk->sk_protocol, &new_sock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1798) if (ret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1799) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1800)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1801) new_sock->type = sock->type;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1802) new_sock->ops = sock->ops;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1803) ret = sock->ops->accept(sock, new_sock, O_NONBLOCK, false);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1804) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1805) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1806)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1807) *more = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1808) new_sock->sk->sk_allocation = GFP_ATOMIC;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1809)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1810) tcp_sock_set_nodelay(new_sock->sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1811) tcp_sock_set_user_timeout(new_sock->sk, O2NET_TCP_USER_TIMEOUT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1812)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1813) ret = new_sock->ops->getname(new_sock, (struct sockaddr *) &sin, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1814) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1815) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1816)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1817) node = o2nm_get_node_by_ip(sin.sin_addr.s_addr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1818) if (node == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1819) printk(KERN_NOTICE "o2net: Attempt to connect from unknown "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1820) "node at %pI4:%d\n", &sin.sin_addr.s_addr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1821) ntohs(sin.sin_port));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1822) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1823) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1824) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1825)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1826) if (o2nm_this_node() >= node->nd_num) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1827) local_node = o2nm_get_node_by_num(o2nm_this_node());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1828) if (local_node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1829) printk(KERN_NOTICE "o2net: Unexpected connect attempt "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1830) "seen at node '%s' (%u, %pI4:%d) from "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1831) "node '%s' (%u, %pI4:%d)\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1832) local_node->nd_name, local_node->nd_num,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1833) &(local_node->nd_ipv4_address),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1834) ntohs(local_node->nd_ipv4_port),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1835) node->nd_name,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1836) node->nd_num, &sin.sin_addr.s_addr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1837) ntohs(sin.sin_port));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1838) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1839) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1840) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1841)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1842) /* this happens all the time when the other node sees our heartbeat
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1843) * and tries to connect before we see their heartbeat */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1844) if (!o2hb_check_node_heartbeating_from_callback(node->nd_num)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1845) mlog(ML_CONN, "attempt to connect from node '%s' at "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1846) "%pI4:%d but it isn't heartbeating\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1847) node->nd_name, &sin.sin_addr.s_addr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1848) ntohs(sin.sin_port));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1849) ret = -EINVAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1850) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1851) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1852)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1853) nn = o2net_nn_from_num(node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1854)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1855) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1856) if (nn->nn_sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1857) ret = -EBUSY;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1858) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1859) ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1860) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1861) if (ret) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1862) printk(KERN_NOTICE "o2net: Attempt to connect from node '%s' "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1863) "at %pI4:%d but it already has an open connection\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1864) node->nd_name, &sin.sin_addr.s_addr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1865) ntohs(sin.sin_port));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1866) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1867) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1868)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1869) sc = sc_alloc(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1870) if (sc == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1871) ret = -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1872) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1873) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1874)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1875) sc->sc_sock = new_sock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1876) new_sock = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1877)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1878) spin_lock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1879) atomic_set(&nn->nn_timeout, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1880) o2net_set_nn_state(nn, sc, 0, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1881) spin_unlock(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1882)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1883) o2net_register_callbacks(sc->sc_sock->sk, sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1884) o2net_sc_queue_work(sc, &sc->sc_rx_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1885)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1886) o2net_initialize_handshake();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1887) o2net_sendpage(sc, o2net_hand, sizeof(*o2net_hand));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1888)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1889) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1890) if (new_sock)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1891) sock_release(new_sock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1892) if (node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1893) o2nm_node_put(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1894) if (local_node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1895) o2nm_node_put(local_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1896) if (sc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1897) sc_put(sc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1898)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1899) memalloc_nofs_restore(nofs_flag);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1900) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1901) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1902)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1903) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1904) * This function is invoked in response to one or more
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1905) * pending accepts at softIRQ level. We must drain the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1906) * entire que before returning.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1907) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1908)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1909) static void o2net_accept_many(struct work_struct *work)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1910) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1911) struct socket *sock = o2net_listen_sock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1912) int more;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1913)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1914) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1915) * It is critical to note that due to interrupt moderation
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1916) * at the network driver level, we can't assume to get a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1917) * softIRQ for every single conn since tcp SYN packets
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1918) * can arrive back-to-back, and therefore many pending
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1919) * accepts may result in just 1 softIRQ. If we terminate
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1920) * the o2net_accept_one() loop upon seeing an err, what happens
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1921) * to the rest of the conns in the queue? If no new SYN
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1922) * arrives for hours, no softIRQ will be delivered,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1923) * and the connections will just sit in the queue.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1924) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1925)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1926) for (;;) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1927) o2net_accept_one(sock, &more);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1928) if (!more)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1929) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1930) cond_resched();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1931) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1932) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1933)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1934) static void o2net_listen_data_ready(struct sock *sk)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1935) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1936) void (*ready)(struct sock *sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1937)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1938) read_lock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1939) ready = sk->sk_user_data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1940) if (ready == NULL) { /* check for teardown race */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1941) ready = sk->sk_data_ready;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1942) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1943) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1944)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1945) /* This callback may called twice when a new connection
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1946) * is being established as a child socket inherits everything
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1947) * from a parent LISTEN socket, including the data_ready cb of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1948) * the parent. This leads to a hazard. In o2net_accept_one()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1949) * we are still initializing the child socket but have not
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1950) * changed the inherited data_ready callback yet when
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1951) * data starts arriving.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1952) * We avoid this hazard by checking the state.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1953) * For the listening socket, the state will be TCP_LISTEN; for the new
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1954) * socket, will be TCP_ESTABLISHED. Also, in this case,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1955) * sk->sk_user_data is not a valid function pointer.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1956) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1957)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1958) if (sk->sk_state == TCP_LISTEN) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1959) queue_work(o2net_wq, &o2net_listen_work);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1960) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1961) ready = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1962) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1963)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1964) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1965) read_unlock_bh(&sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1966) if (ready != NULL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1967) ready(sk);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1968) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1969)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1970) static int o2net_open_listening_sock(__be32 addr, __be16 port)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1971) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1972) struct socket *sock = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1973) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1974) struct sockaddr_in sin = {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1975) .sin_family = PF_INET,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1976) .sin_addr = { .s_addr = addr },
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1977) .sin_port = port,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1978) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1979)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1980) ret = sock_create(PF_INET, SOCK_STREAM, IPPROTO_TCP, &sock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1981) if (ret < 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1982) printk(KERN_ERR "o2net: Error %d while creating socket\n", ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1983) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1984) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1985)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1986) sock->sk->sk_allocation = GFP_ATOMIC;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1987)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1988) write_lock_bh(&sock->sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1989) sock->sk->sk_user_data = sock->sk->sk_data_ready;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1990) sock->sk->sk_data_ready = o2net_listen_data_ready;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1991) write_unlock_bh(&sock->sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1992)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1993) o2net_listen_sock = sock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1994) INIT_WORK(&o2net_listen_work, o2net_accept_many);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1995)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1996) sock->sk->sk_reuse = SK_CAN_REUSE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1997) ret = sock->ops->bind(sock, (struct sockaddr *)&sin, sizeof(sin));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1998) if (ret < 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1999) printk(KERN_ERR "o2net: Error %d while binding socket at "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2000) "%pI4:%u\n", ret, &addr, ntohs(port));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2001) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2002) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2003)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2004) ret = sock->ops->listen(sock, 64);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2005) if (ret < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2006) printk(KERN_ERR "o2net: Error %d while listening on %pI4:%u\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2007) ret, &addr, ntohs(port));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2008)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2009) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2010) if (ret) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2011) o2net_listen_sock = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2012) if (sock)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2013) sock_release(sock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2014) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2015) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2016) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2017)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2018) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2019) * called from node manager when we should bring up our network listening
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2020) * socket. node manager handles all the serialization to only call this
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2021) * once and to match it with o2net_stop_listening(). note,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2022) * o2nm_this_node() doesn't work yet as we're being called while it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2023) * is being set up.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2024) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2025) int o2net_start_listening(struct o2nm_node *node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2026) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2027) int ret = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2028)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2029) BUG_ON(o2net_wq != NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2030) BUG_ON(o2net_listen_sock != NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2031)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2032) mlog(ML_KTHREAD, "starting o2net thread...\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2033) o2net_wq = alloc_ordered_workqueue("o2net", WQ_MEM_RECLAIM);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2034) if (o2net_wq == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2035) mlog(ML_ERROR, "unable to launch o2net thread\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2036) return -ENOMEM; /* ? */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2037) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2038)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2039) ret = o2net_open_listening_sock(node->nd_ipv4_address,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2040) node->nd_ipv4_port);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2041) if (ret) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2042) destroy_workqueue(o2net_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2043) o2net_wq = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2044) } else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2045) o2quo_conn_up(node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2046)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2047) return ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2048) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2049)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2050) /* again, o2nm_this_node() doesn't work here as we're involved in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2051) * tearing it down */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2052) void o2net_stop_listening(struct o2nm_node *node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2053) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2054) struct socket *sock = o2net_listen_sock;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2055) size_t i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2056)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2057) BUG_ON(o2net_wq == NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2058) BUG_ON(o2net_listen_sock == NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2059)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2060) /* stop the listening socket from generating work */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2061) write_lock_bh(&sock->sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2062) sock->sk->sk_data_ready = sock->sk->sk_user_data;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2063) sock->sk->sk_user_data = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2064) write_unlock_bh(&sock->sk->sk_callback_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2065)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2066) for (i = 0; i < ARRAY_SIZE(o2net_nodes); i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2067) struct o2nm_node *node = o2nm_get_node_by_num(i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2068) if (node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2069) o2net_disconnect_node(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2070) o2nm_node_put(node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2071) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2072) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2073)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2074) /* finish all work and tear down the work queue */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2075) mlog(ML_KTHREAD, "waiting for o2net thread to exit....\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2076) destroy_workqueue(o2net_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2077) o2net_wq = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2078)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2079) sock_release(o2net_listen_sock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2080) o2net_listen_sock = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2081)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2082) o2quo_conn_err(node->nd_num);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2083) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2084)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2085) /* ------------------------------------------------------------ */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2086)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2087) int o2net_init(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2088) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2089) unsigned long i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2090)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2091) o2quo_init();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2092)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2093) o2net_debugfs_init();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2094)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2095) o2net_hand = kzalloc(sizeof(struct o2net_handshake), GFP_KERNEL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2096) o2net_keep_req = kzalloc(sizeof(struct o2net_msg), GFP_KERNEL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2097) o2net_keep_resp = kzalloc(sizeof(struct o2net_msg), GFP_KERNEL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2098) if (!o2net_hand || !o2net_keep_req || !o2net_keep_resp)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2099) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2100)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2101) o2net_hand->protocol_version = cpu_to_be64(O2NET_PROTOCOL_VERSION);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2102) o2net_hand->connector_id = cpu_to_be64(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2103)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2104) o2net_keep_req->magic = cpu_to_be16(O2NET_MSG_KEEP_REQ_MAGIC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2105) o2net_keep_resp->magic = cpu_to_be16(O2NET_MSG_KEEP_RESP_MAGIC);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2106)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2107) for (i = 0; i < ARRAY_SIZE(o2net_nodes); i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2108) struct o2net_node *nn = o2net_nn_from_num(i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2109)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2110) atomic_set(&nn->nn_timeout, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2111) spin_lock_init(&nn->nn_lock);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2112) INIT_DELAYED_WORK(&nn->nn_connect_work, o2net_start_connect);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2113) INIT_DELAYED_WORK(&nn->nn_connect_expired,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2114) o2net_connect_expired);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2115) INIT_DELAYED_WORK(&nn->nn_still_up, o2net_still_up);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2116) /* until we see hb from a node we'll return einval */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2117) nn->nn_persistent_error = -ENOTCONN;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2118) init_waitqueue_head(&nn->nn_sc_wq);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2119) idr_init(&nn->nn_status_idr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2120) INIT_LIST_HEAD(&nn->nn_status_list);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2121) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2122)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2123) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2124)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2125) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2126) kfree(o2net_hand);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2127) kfree(o2net_keep_req);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2128) kfree(o2net_keep_resp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2129) o2net_debugfs_exit();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2130) o2quo_exit();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2131) return -ENOMEM;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2132) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2133)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2134) void o2net_exit(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2135) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2136) o2quo_exit();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2137) kfree(o2net_hand);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2138) kfree(o2net_keep_req);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2139) kfree(o2net_keep_resp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2140) o2net_debugfs_exit();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2141) }