diff options
Diffstat (limited to 'drivers/infiniband/hw/cxgb4/cm.c')
| -rw-r--r-- | drivers/infiniband/hw/cxgb4/cm.c | 2410 |
1 files changed, 1986 insertions, 424 deletions
diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c index 8b00e6c46f0..768a0fb67dd 100644 --- a/drivers/infiniband/hw/cxgb4/cm.c +++ b/drivers/infiniband/hw/cxgb4/cm.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved. + * Copyright (c) 2009-2014 Chelsio, Inc. All rights reserved. * * This software is available to you under a choice of one of two * licenses. You may choose to be licensed under the terms of the GNU @@ -38,10 +38,16 @@ #include <linux/inetdevice.h> #include <linux/ip.h> #include <linux/tcp.h> +#include <linux/if_vlan.h> #include <net/neighbour.h> #include <net/netevent.h> #include <net/route.h> +#include <net/tcp.h> +#include <net/ip6_route.h> +#include <net/addrconf.h> + +#include <rdma/ib_addr.h> #include "iw_cxgb4.h" @@ -61,9 +67,17 @@ static char *states[] = { NULL, }; -static int dack_mode; +static int nocong; +module_param(nocong, int, 0644); +MODULE_PARM_DESC(nocong, "Turn of congestion control (default=0)"); + +static int enable_ecn; +module_param(enable_ecn, int, 0644); +MODULE_PARM_DESC(enable_ecn, "Enable ECN (default=0/disabled)"); + +static int dack_mode = 1; module_param(dack_mode, int, 0644); -MODULE_PARM_DESC(dack_mode, "Delayed ack mode (default=0)"); +MODULE_PARM_DESC(dack_mode, "Delayed ack mode (default=1)"); int c4iw_max_read_depth = 8; module_param(c4iw_max_read_depth, int, 0644); @@ -86,9 +100,9 @@ int c4iw_debug; module_param(c4iw_debug, int, 0644); MODULE_PARM_DESC(c4iw_debug, "Enable debug logging (default=0)"); -static int peer2peer; +static int peer2peer = 1; module_param(peer2peer, int, 0644); -MODULE_PARM_DESC(peer2peer, "Support peer2peer ULPs (default=0)"); +MODULE_PARM_DESC(peer2peer, "Support peer2peer ULPs (default=1)"); static int p2p_type = FW_RI_INIT_P2PTYPE_READ_REQ; module_param(p2p_type, int, 0644); @@ -103,7 +117,8 @@ MODULE_PARM_DESC(ep_timeout_secs, "CM Endpoint operation timeout " static int mpa_rev = 1; module_param(mpa_rev, int, 0644); MODULE_PARM_DESC(mpa_rev, "MPA Revision, 0 supports amso1100, " - "1 is spec compliant. (default=1)"); + "1 is RFC0544 spec compliant, 2 is IETF MPA Peer Connect Draft" + " compliant (default=1)"); static int markers_enabled; module_param(markers_enabled, int, 0644); @@ -132,31 +147,43 @@ static void connect_reply_upcall(struct c4iw_ep *ep, int status); static LIST_HEAD(timeout_list); static spinlock_t timeout_lock; +static void deref_qp(struct c4iw_ep *ep) +{ + c4iw_qp_rem_ref(&ep->com.qp->ibqp); + clear_bit(QP_REFERENCED, &ep->com.flags); +} + +static void ref_qp(struct c4iw_ep *ep) +{ + set_bit(QP_REFERENCED, &ep->com.flags); + c4iw_qp_add_ref(&ep->com.qp->ibqp); +} + static void start_ep_timer(struct c4iw_ep *ep) { PDBG("%s ep %p\n", __func__, ep); if (timer_pending(&ep->timer)) { - PDBG("%s stopped / restarted timer ep %p\n", __func__, ep); - del_timer_sync(&ep->timer); - } else - c4iw_get_ep(&ep->com); + pr_err("%s timer already started! ep %p\n", + __func__, ep); + return; + } + clear_bit(TIMEOUT, &ep->com.flags); + c4iw_get_ep(&ep->com); ep->timer.expires = jiffies + ep_timeout_secs * HZ; ep->timer.data = (unsigned long)ep; ep->timer.function = ep_timeout; add_timer(&ep->timer); } -static void stop_ep_timer(struct c4iw_ep *ep) +static int stop_ep_timer(struct c4iw_ep *ep) { - PDBG("%s ep %p\n", __func__, ep); - if (!timer_pending(&ep->timer)) { - printk(KERN_ERR "%s timer stopped when its not running! " - "ep %p state %u\n", __func__, ep, ep->com.state); - WARN_ON(1); - return; - } + PDBG("%s ep %p stopping\n", __func__, ep); del_timer_sync(&ep->timer); - c4iw_put_ep(&ep->com); + if (!test_and_set_bit(TIMEOUT, &ep->com.flags)) { + c4iw_put_ep(&ep->com); + return 0; + } + return 1; } static int c4iw_l2t_send(struct c4iw_rdev *rdev, struct sk_buff *skb, @@ -207,12 +234,16 @@ static void release_tid(struct c4iw_rdev *rdev, u32 hwtid, struct sk_buff *skb) static void set_emss(struct c4iw_ep *ep, u16 opt) { - ep->emss = ep->com.dev->rdev.lldi.mtus[GET_TCPOPT_MSS(opt)] - 40; + ep->emss = ep->com.dev->rdev.lldi.mtus[GET_TCPOPT_MSS(opt)] - + sizeof(struct iphdr) - sizeof(struct tcphdr); ep->mss = ep->emss; if (GET_TCPOPT_TSTAMP(opt)) ep->emss -= 12; if (ep->emss < 128) ep->emss = 128; + if (ep->emss & 7) + PDBG("Warning: misaligned mtu idx %u mss %u emss=%u\n", + GET_TCPOPT_MSS(opt), ep->mss, ep->emss); PDBG("%s mss_idx %u mss %u emss=%u\n", __func__, GET_TCPOPT_MSS(opt), ep->mss, ep->emss); } @@ -261,11 +292,20 @@ void _c4iw_free_ep(struct kref *kref) ep = container_of(kref, struct c4iw_ep, com.kref); PDBG("%s ep %p state %s\n", __func__, ep, states[state_read(&ep->com)]); + if (test_bit(QP_REFERENCED, &ep->com.flags)) + deref_qp(ep); if (test_bit(RELEASE_RESOURCES, &ep->com.flags)) { + remove_handle(ep->com.dev, &ep->com.dev->hwtid_idr, ep->hwtid); cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, ep->hwtid); dst_release(ep->dst); cxgb4_l2t_release(ep->l2t); } + if (test_bit(RELEASE_MAPINFO, &ep->com.flags)) { + print_addr(&ep->com, __func__, "remove_mapinfo/mapping"); + iwpm_remove_mapinfo(&ep->com.local_addr, + &ep->com.mapped_local_addr); + iwpm_remove_mapping(&ep->com.local_addr, RDMA_NL_C4IW); + } kfree(ep); } @@ -307,33 +347,78 @@ static struct sk_buff *get_skb(struct sk_buff *skb, int len, gfp_t gfp) } else { skb = alloc_skb(len, gfp); } + t4_set_arp_err_handler(skb, NULL, NULL); return skb; } -static struct rtable *find_route(struct c4iw_dev *dev, __be32 local_ip, +static struct net_device *get_real_dev(struct net_device *egress_dev) +{ + return rdma_vlan_dev_real_dev(egress_dev) ? : egress_dev; +} + +static int our_interface(struct c4iw_dev *dev, struct net_device *egress_dev) +{ + int i; + + egress_dev = get_real_dev(egress_dev); + for (i = 0; i < dev->rdev.lldi.nports; i++) + if (dev->rdev.lldi.ports[i] == egress_dev) + return 1; + return 0; +} + +static struct dst_entry *find_route6(struct c4iw_dev *dev, __u8 *local_ip, + __u8 *peer_ip, __be16 local_port, + __be16 peer_port, u8 tos, + __u32 sin6_scope_id) +{ + struct dst_entry *dst = NULL; + + if (IS_ENABLED(CONFIG_IPV6)) { + struct flowi6 fl6; + + memset(&fl6, 0, sizeof(fl6)); + memcpy(&fl6.daddr, peer_ip, 16); + memcpy(&fl6.saddr, local_ip, 16); + if (ipv6_addr_type(&fl6.daddr) & IPV6_ADDR_LINKLOCAL) + fl6.flowi6_oif = sin6_scope_id; + dst = ip6_route_output(&init_net, NULL, &fl6); + if (!dst) + goto out; + if (!our_interface(dev, ip6_dst_idev(dst)->dev) && + !(ip6_dst_idev(dst)->dev->flags & IFF_LOOPBACK)) { + dst_release(dst); + dst = NULL; + } + } + +out: + return dst; +} + +static struct dst_entry *find_route(struct c4iw_dev *dev, __be32 local_ip, __be32 peer_ip, __be16 local_port, __be16 peer_port, u8 tos) { struct rtable *rt; - struct flowi fl = { - .oif = 0, - .nl_u = { - .ip4_u = { - .daddr = peer_ip, - .saddr = local_ip, - .tos = tos} - }, - .proto = IPPROTO_TCP, - .uli_u = { - .ports = { - .sport = local_port, - .dport = peer_port} - } - }; - - if (ip_route_output_flow(&init_net, &rt, &fl, NULL, 0)) + struct flowi4 fl4; + struct neighbour *n; + + rt = ip_route_output_ports(&init_net, &fl4, NULL, peer_ip, local_ip, + peer_port, local_port, IPPROTO_TCP, + tos, 0); + if (IS_ERR(rt)) + return NULL; + n = dst_neigh_lookup(&rt->dst, &peer_ip); + if (!n) + return NULL; + if (!our_interface(dev, n->dev) && + !(n->dev->flags & IFF_LOOPBACK)) { + dst_release(&rt->dst); return NULL; - return rt; + } + neigh_release(n); + return &rt->dst; } static void arp_failure_discard(void *handle, struct sk_buff *skb) @@ -347,8 +432,17 @@ static void arp_failure_discard(void *handle, struct sk_buff *skb) */ static void act_open_req_arp_failure(void *handle, struct sk_buff *skb) { + struct c4iw_ep *ep = handle; + printk(KERN_ERR MOD "ARP failure duing connect\n"); kfree_skb(skb); + connect_reply_upcall(ep, -EHOSTUNREACH); + state_set(&ep->com, DEAD); + remove_handle(ep->com.dev, &ep->com.dev->atid_idr, ep->atid); + cxgb4_free_atid(ep->com.dev->rdev.lldi.tids, ep->atid); + dst_release(ep->dst); + cxgb4_l2t_release(ep->l2t); + c4iw_put_ep(&ep->com); } /* @@ -392,7 +486,7 @@ static void send_flowc(struct c4iw_ep *ep, struct sk_buff *skb) flowc->mnemval[5].mnemonic = FW_FLOWC_MNEM_RCVNXT; flowc->mnemval[5].val = cpu_to_be32(ep->rcv_seq); flowc->mnemval[6].mnemonic = FW_FLOWC_MNEM_SNDBUF; - flowc->mnemval[6].val = cpu_to_be32(snd_win); + flowc->mnemval[6].val = cpu_to_be32(ep->snd_win); flowc->mnemval[7].mnemonic = FW_FLOWC_MNEM_MSS; flowc->mnemval[7].val = cpu_to_be32(ep->emss); /* Pad WR to 16 byte boundary */ @@ -452,15 +546,80 @@ static int send_abort(struct c4iw_ep *ep, struct sk_buff *skb, gfp_t gfp) return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); } +/* + * c4iw_form_pm_msg - Form a port mapper message with mapping info + */ +static void c4iw_form_pm_msg(struct c4iw_ep *ep, + struct iwpm_sa_data *pm_msg) +{ + memcpy(&pm_msg->loc_addr, &ep->com.local_addr, + sizeof(ep->com.local_addr)); + memcpy(&pm_msg->rem_addr, &ep->com.remote_addr, + sizeof(ep->com.remote_addr)); +} + +/* + * c4iw_form_reg_msg - Form a port mapper message with dev info + */ +static void c4iw_form_reg_msg(struct c4iw_dev *dev, + struct iwpm_dev_data *pm_msg) +{ + memcpy(pm_msg->dev_name, dev->ibdev.name, IWPM_DEVNAME_SIZE); + memcpy(pm_msg->if_name, dev->rdev.lldi.ports[0]->name, + IWPM_IFNAME_SIZE); +} + +static void c4iw_record_pm_msg(struct c4iw_ep *ep, + struct iwpm_sa_data *pm_msg) +{ + memcpy(&ep->com.mapped_local_addr, &pm_msg->mapped_loc_addr, + sizeof(ep->com.mapped_local_addr)); + memcpy(&ep->com.mapped_remote_addr, &pm_msg->mapped_rem_addr, + sizeof(ep->com.mapped_remote_addr)); +} + +static void best_mtu(const unsigned short *mtus, unsigned short mtu, + unsigned int *idx, int use_ts) +{ + unsigned short hdr_size = sizeof(struct iphdr) + + sizeof(struct tcphdr) + + (use_ts ? 12 : 0); + unsigned short data_size = mtu - hdr_size; + + cxgb4_best_aligned_mtu(mtus, hdr_size, data_size, 8, idx); +} + static int send_connect(struct c4iw_ep *ep) { struct cpl_act_open_req *req; + struct cpl_t5_act_open_req *t5_req; + struct cpl_act_open_req6 *req6; + struct cpl_t5_act_open_req6 *t5_req6; struct sk_buff *skb; u64 opt0; u32 opt2; unsigned int mtu_idx; int wscale; - int wrlen = roundup(sizeof *req, 16); + int wrlen; + int sizev4 = is_t4(ep->com.dev->rdev.lldi.adapter_type) ? + sizeof(struct cpl_act_open_req) : + sizeof(struct cpl_t5_act_open_req); + int sizev6 = is_t4(ep->com.dev->rdev.lldi.adapter_type) ? + sizeof(struct cpl_act_open_req6) : + sizeof(struct cpl_t5_act_open_req6); + struct sockaddr_in *la = (struct sockaddr_in *) + &ep->com.mapped_local_addr; + struct sockaddr_in *ra = (struct sockaddr_in *) + &ep->com.mapped_remote_addr; + struct sockaddr_in6 *la6 = (struct sockaddr_in6 *) + &ep->com.mapped_local_addr; + struct sockaddr_in6 *ra6 = (struct sockaddr_in6 *) + &ep->com.mapped_remote_addr; + int win; + + wrlen = (ep->com.remote_addr.ss_family == AF_INET) ? + roundup(sizev4, 16) : + roundup(sizev6, 16); PDBG("%s ep %p atid %u\n", __func__, ep, ep->atid); @@ -472,9 +631,20 @@ static int send_connect(struct c4iw_ep *ep) } set_wr_txq(skb, CPL_PRIORITY_SETUP, ep->ctrlq_idx); - cxgb4_best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx); + best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx, + enable_tcp_timestamps); wscale = compute_wscale(rcv_win); - opt0 = KEEP_ALIVE(1) | + + /* + * Specify the largest window that will fit in opt0. The + * remainder will be specified in the rx_data_ack. + */ + win = ep->rcv_win >> 10; + if (win > RCV_BUFSIZ_MASK) + win = RCV_BUFSIZ_MASK; + + opt0 = (nocong ? NO_CONG(1) : 0) | + KEEP_ALIVE(1) | DELACK(1) | WND_SCALE(wscale) | MSS_IDX(mtu_idx) | @@ -482,8 +652,10 @@ static int send_connect(struct c4iw_ep *ep) TX_CHAN(ep->tx_chan) | SMAC_SEL(ep->smac_idx) | DSCP(ep->tos) | - RCV_BUFSIZ(rcv_win>>10); + ULP_MODE(ULP_MODE_TCPDDP) | + RCV_BUFSIZ(win); opt2 = RX_CHANNEL(0) | + CCTRL_ECN(enable_ecn) | RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid); if (enable_tcp_timestamps) opt2 |= TSTAMPS_EN(1); @@ -491,33 +663,127 @@ static int send_connect(struct c4iw_ep *ep) opt2 |= SACK_EN(1); if (wscale && enable_tcp_window_scaling) opt2 |= WND_SCALE_EN(1); - t4_set_arp_err_handler(skb, NULL, act_open_req_arp_failure); - - req = (struct cpl_act_open_req *) skb_put(skb, wrlen); - INIT_TP_WR(req, 0); - OPCODE_TID(req) = cpu_to_be32( - MK_OPCODE_TID(CPL_ACT_OPEN_REQ, ((ep->rss_qid<<14)|ep->atid))); - req->local_port = ep->com.local_addr.sin_port; - req->peer_port = ep->com.remote_addr.sin_port; - req->local_ip = ep->com.local_addr.sin_addr.s_addr; - req->peer_ip = ep->com.remote_addr.sin_addr.s_addr; - req->opt0 = cpu_to_be64(opt0); - req->params = 0; - req->opt2 = cpu_to_be32(opt2); + if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) { + opt2 |= T5_OPT_2_VALID; + opt2 |= V_CONG_CNTRL(CONG_ALG_TAHOE); + } + t4_set_arp_err_handler(skb, ep, act_open_req_arp_failure); + + if (is_t4(ep->com.dev->rdev.lldi.adapter_type)) { + if (ep->com.remote_addr.ss_family == AF_INET) { + req = (struct cpl_act_open_req *) skb_put(skb, wrlen); + INIT_TP_WR(req, 0); + OPCODE_TID(req) = cpu_to_be32( + MK_OPCODE_TID(CPL_ACT_OPEN_REQ, + ((ep->rss_qid << 14) | ep->atid))); + req->local_port = la->sin_port; + req->peer_port = ra->sin_port; + req->local_ip = la->sin_addr.s_addr; + req->peer_ip = ra->sin_addr.s_addr; + req->opt0 = cpu_to_be64(opt0); + req->params = cpu_to_be32(cxgb4_select_ntuple( + ep->com.dev->rdev.lldi.ports[0], + ep->l2t)); + req->opt2 = cpu_to_be32(opt2); + } else { + req6 = (struct cpl_act_open_req6 *)skb_put(skb, wrlen); + + INIT_TP_WR(req6, 0); + OPCODE_TID(req6) = cpu_to_be32( + MK_OPCODE_TID(CPL_ACT_OPEN_REQ6, + ((ep->rss_qid<<14)|ep->atid))); + req6->local_port = la6->sin6_port; + req6->peer_port = ra6->sin6_port; + req6->local_ip_hi = *((__be64 *) + (la6->sin6_addr.s6_addr)); + req6->local_ip_lo = *((__be64 *) + (la6->sin6_addr.s6_addr + 8)); + req6->peer_ip_hi = *((__be64 *) + (ra6->sin6_addr.s6_addr)); + req6->peer_ip_lo = *((__be64 *) + (ra6->sin6_addr.s6_addr + 8)); + req6->opt0 = cpu_to_be64(opt0); + req6->params = cpu_to_be32(cxgb4_select_ntuple( + ep->com.dev->rdev.lldi.ports[0], + ep->l2t)); + req6->opt2 = cpu_to_be32(opt2); + } + } else { + u32 isn = (prandom_u32() & ~7UL) - 1; + + opt2 |= T5_OPT_2_VALID; + opt2 |= CONG_CNTRL_VALID; /* OPT_2_ISS for T5 */ + if (peer2peer) + isn += 4; + + if (ep->com.remote_addr.ss_family == AF_INET) { + t5_req = (struct cpl_t5_act_open_req *) + skb_put(skb, wrlen); + INIT_TP_WR(t5_req, 0); + OPCODE_TID(t5_req) = cpu_to_be32( + MK_OPCODE_TID(CPL_ACT_OPEN_REQ, + ((ep->rss_qid << 14) | ep->atid))); + t5_req->local_port = la->sin_port; + t5_req->peer_port = ra->sin_port; + t5_req->local_ip = la->sin_addr.s_addr; + t5_req->peer_ip = ra->sin_addr.s_addr; + t5_req->opt0 = cpu_to_be64(opt0); + t5_req->params = cpu_to_be64(V_FILTER_TUPLE( + cxgb4_select_ntuple( + ep->com.dev->rdev.lldi.ports[0], + ep->l2t))); + t5_req->rsvd = cpu_to_be32(isn); + PDBG("%s snd_isn %u\n", __func__, + be32_to_cpu(t5_req->rsvd)); + t5_req->opt2 = cpu_to_be32(opt2); + } else { + t5_req6 = (struct cpl_t5_act_open_req6 *) + skb_put(skb, wrlen); + INIT_TP_WR(t5_req6, 0); + OPCODE_TID(t5_req6) = cpu_to_be32( + MK_OPCODE_TID(CPL_ACT_OPEN_REQ6, + ((ep->rss_qid<<14)|ep->atid))); + t5_req6->local_port = la6->sin6_port; + t5_req6->peer_port = ra6->sin6_port; + t5_req6->local_ip_hi = *((__be64 *) + (la6->sin6_addr.s6_addr)); + t5_req6->local_ip_lo = *((__be64 *) + (la6->sin6_addr.s6_addr + 8)); + t5_req6->peer_ip_hi = *((__be64 *) + (ra6->sin6_addr.s6_addr)); + t5_req6->peer_ip_lo = *((__be64 *) + (ra6->sin6_addr.s6_addr + 8)); + t5_req6->opt0 = cpu_to_be64(opt0); + t5_req6->params = (__force __be64)cpu_to_be32( + cxgb4_select_ntuple( + ep->com.dev->rdev.lldi.ports[0], + ep->l2t)); + t5_req6->rsvd = cpu_to_be32(isn); + PDBG("%s snd_isn %u\n", __func__, + be32_to_cpu(t5_req6->rsvd)); + t5_req6->opt2 = cpu_to_be32(opt2); + } + } + + set_bit(ACT_OPEN_REQ, &ep->com.history); return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); } -static void send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb) +static void send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb, + u8 mpa_rev_to_use) { int mpalen, wrlen; struct fw_ofld_tx_data_wr *req; struct mpa_message *mpa; + struct mpa_v2_conn_params mpa_v2_params; PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen); BUG_ON(skb_cloned(skb)); mpalen = sizeof(*mpa) + ep->plen; + if (mpa_rev_to_use == 2) + mpalen += sizeof(struct mpa_v2_conn_params); wrlen = roundup(mpalen + sizeof *req, 16); skb = get_skb(skb, wrlen, GFP_KERNEL); if (!skb) { @@ -543,12 +809,41 @@ static void send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb) mpa = (struct mpa_message *)(req + 1); memcpy(mpa->key, MPA_KEY_REQ, sizeof(mpa->key)); mpa->flags = (crc_enabled ? MPA_CRC : 0) | - (markers_enabled ? MPA_MARKERS : 0); + (markers_enabled ? MPA_MARKERS : 0) | + (mpa_rev_to_use == 2 ? MPA_ENHANCED_RDMA_CONN : 0); mpa->private_data_size = htons(ep->plen); - mpa->revision = mpa_rev; + mpa->revision = mpa_rev_to_use; + if (mpa_rev_to_use == 1) { + ep->tried_with_mpa_v1 = 1; + ep->retry_with_mpa_v1 = 0; + } + + if (mpa_rev_to_use == 2) { + mpa->private_data_size = htons(ntohs(mpa->private_data_size) + + sizeof (struct mpa_v2_conn_params)); + mpa_v2_params.ird = htons((u16)ep->ird); + mpa_v2_params.ord = htons((u16)ep->ord); + + if (peer2peer) { + mpa_v2_params.ird |= htons(MPA_V2_PEER2PEER_MODEL); + if (p2p_type == FW_RI_INIT_P2PTYPE_RDMA_WRITE) + mpa_v2_params.ord |= + htons(MPA_V2_RDMA_WRITE_RTR); + else if (p2p_type == FW_RI_INIT_P2PTYPE_READ_REQ) + mpa_v2_params.ord |= + htons(MPA_V2_RDMA_READ_RTR); + } + memcpy(mpa->private_data, &mpa_v2_params, + sizeof(struct mpa_v2_conn_params)); - if (ep->plen) - memcpy(mpa->private_data, ep->mpa_pkt + sizeof(*mpa), ep->plen); + if (ep->plen) + memcpy(mpa->private_data + + sizeof(struct mpa_v2_conn_params), + ep->mpa_pkt + sizeof(*mpa), ep->plen); + } else + if (ep->plen) + memcpy(mpa->private_data, + ep->mpa_pkt + sizeof(*mpa), ep->plen); /* * Reference the mpa skb. This ensures the data area @@ -561,8 +856,9 @@ static void send_mpa_req(struct c4iw_ep *ep, struct sk_buff *skb) ep->mpa_skb = skb; c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); start_ep_timer(ep); - state_set(&ep->com, MPA_REQ_SENT); + __state_set(&ep->com, MPA_REQ_SENT); ep->mpa_attr.initiator = 1; + ep->snd_seq += mpalen; return; } @@ -572,10 +868,13 @@ static int send_mpa_reject(struct c4iw_ep *ep, const void *pdata, u8 plen) struct fw_ofld_tx_data_wr *req; struct mpa_message *mpa; struct sk_buff *skb; + struct mpa_v2_conn_params mpa_v2_params; PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen); mpalen = sizeof(*mpa) + plen; + if (ep->mpa_attr.version == 2 && ep->mpa_attr.enhanced_rdma_conn) + mpalen += sizeof(struct mpa_v2_conn_params); wrlen = roundup(mpalen + sizeof *req, 16); skb = get_skb(NULL, wrlen, GFP_KERNEL); @@ -603,10 +902,31 @@ static int send_mpa_reject(struct c4iw_ep *ep, const void *pdata, u8 plen) memset(mpa, 0, sizeof(*mpa)); memcpy(mpa->key, MPA_KEY_REP, sizeof(mpa->key)); mpa->flags = MPA_REJECT; - mpa->revision = mpa_rev; + mpa->revision = ep->mpa_attr.version; mpa->private_data_size = htons(plen); - if (plen) - memcpy(mpa->private_data, pdata, plen); + + if (ep->mpa_attr.version == 2 && ep->mpa_attr.enhanced_rdma_conn) { + mpa->flags |= MPA_ENHANCED_RDMA_CONN; + mpa->private_data_size = htons(ntohs(mpa->private_data_size) + + sizeof (struct mpa_v2_conn_params)); + mpa_v2_params.ird = htons(((u16)ep->ird) | + (peer2peer ? MPA_V2_PEER2PEER_MODEL : + 0)); + mpa_v2_params.ord = htons(((u16)ep->ord) | (peer2peer ? + (p2p_type == + FW_RI_INIT_P2PTYPE_RDMA_WRITE ? + MPA_V2_RDMA_WRITE_RTR : p2p_type == + FW_RI_INIT_P2PTYPE_READ_REQ ? + MPA_V2_RDMA_READ_RTR : 0) : 0)); + memcpy(mpa->private_data, &mpa_v2_params, + sizeof(struct mpa_v2_conn_params)); + + if (ep->plen) + memcpy(mpa->private_data + + sizeof(struct mpa_v2_conn_params), pdata, plen); + } else + if (plen) + memcpy(mpa->private_data, pdata, plen); /* * Reference the mpa skb again. This ensures the data area @@ -618,6 +938,7 @@ static int send_mpa_reject(struct c4iw_ep *ep, const void *pdata, u8 plen) t4_set_arp_err_handler(skb, NULL, arp_failure_discard); BUG_ON(ep->mpa_skb); ep->mpa_skb = skb; + ep->snd_seq += mpalen; return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); } @@ -627,10 +948,13 @@ static int send_mpa_reply(struct c4iw_ep *ep, const void *pdata, u8 plen) struct fw_ofld_tx_data_wr *req; struct mpa_message *mpa; struct sk_buff *skb; + struct mpa_v2_conn_params mpa_v2_params; PDBG("%s ep %p tid %u pd_len %d\n", __func__, ep, ep->hwtid, ep->plen); mpalen = sizeof(*mpa) + plen; + if (ep->mpa_attr.version == 2 && ep->mpa_attr.enhanced_rdma_conn) + mpalen += sizeof(struct mpa_v2_conn_params); wrlen = roundup(mpalen + sizeof *req, 16); skb = get_skb(NULL, wrlen, GFP_KERNEL); @@ -659,10 +983,36 @@ static int send_mpa_reply(struct c4iw_ep *ep, const void *pdata, u8 plen) memcpy(mpa->key, MPA_KEY_REP, sizeof(mpa->key)); mpa->flags = (ep->mpa_attr.crc_enabled ? MPA_CRC : 0) | (markers_enabled ? MPA_MARKERS : 0); - mpa->revision = mpa_rev; + mpa->revision = ep->mpa_attr.version; mpa->private_data_size = htons(plen); - if (plen) - memcpy(mpa->private_data, pdata, plen); + + if (ep->mpa_attr.version == 2 && ep->mpa_attr.enhanced_rdma_conn) { + mpa->flags |= MPA_ENHANCED_RDMA_CONN; + mpa->private_data_size = htons(ntohs(mpa->private_data_size) + + sizeof (struct mpa_v2_conn_params)); + mpa_v2_params.ird = htons((u16)ep->ird); + mpa_v2_params.ord = htons((u16)ep->ord); + if (peer2peer && (ep->mpa_attr.p2p_type != + FW_RI_INIT_P2PTYPE_DISABLED)) { + mpa_v2_params.ird |= htons(MPA_V2_PEER2PEER_MODEL); + + if (p2p_type == FW_RI_INIT_P2PTYPE_RDMA_WRITE) + mpa_v2_params.ord |= + htons(MPA_V2_RDMA_WRITE_RTR); + else if (p2p_type == FW_RI_INIT_P2PTYPE_READ_REQ) + mpa_v2_params.ord |= + htons(MPA_V2_RDMA_READ_RTR); + } + + memcpy(mpa->private_data, &mpa_v2_params, + sizeof(struct mpa_v2_conn_params)); + + if (ep->plen) + memcpy(mpa->private_data + + sizeof(struct mpa_v2_conn_params), pdata, plen); + } else + if (plen) + memcpy(mpa->private_data, pdata, plen); /* * Reference the mpa skb. This ensures the data area @@ -672,7 +1022,8 @@ static int send_mpa_reply(struct c4iw_ep *ep, const void *pdata, u8 plen) skb_get(skb); t4_set_arp_err_handler(skb, NULL, arp_failure_discard); ep->mpa_skb = skb; - state_set(&ep->com, MPA_REP_SENT); + __state_set(&ep->com, MPA_REP_SENT); + ep->snd_seq += mpalen; return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); } @@ -689,11 +1040,13 @@ static int act_establish(struct c4iw_dev *dev, struct sk_buff *skb) PDBG("%s ep %p tid %u snd_isn %u rcv_isn %u\n", __func__, ep, tid, be32_to_cpu(req->snd_isn), be32_to_cpu(req->rcv_isn)); + mutex_lock(&ep->com.mutex); dst_confirm(ep->dst); /* setup the hwtid for this connection */ ep->hwtid = tid; cxgb4_insert_tid(t, ep, tid); + insert_handle(dev, &dev->hwtid_idr, ep, ep->hwtid); ep->snd_seq = be32_to_cpu(req->snd_isn); ep->rcv_seq = be32_to_cpu(req->rcv_isn); @@ -701,37 +1054,43 @@ static int act_establish(struct c4iw_dev *dev, struct sk_buff *skb) set_emss(ep, ntohs(req->tcp_opt)); /* dealloc the atid */ + remove_handle(ep->com.dev, &ep->com.dev->atid_idr, atid); cxgb4_free_atid(t, atid); + set_bit(ACT_ESTAB, &ep->com.history); /* start MPA negotiation */ send_flowc(ep, NULL); - send_mpa_req(ep, skb); - + if (ep->retry_with_mpa_v1) + send_mpa_req(ep, skb, 1); + else + send_mpa_req(ep, skb, mpa_rev); + mutex_unlock(&ep->com.mutex); return 0; } -static void close_complete_upcall(struct c4iw_ep *ep) +static void close_complete_upcall(struct c4iw_ep *ep, int status) { struct iw_cm_event event; PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CLOSE; + event.status = status; if (ep->com.cm_id) { PDBG("close complete delivered ep %p cm_id %p tid %u\n", ep, ep->com.cm_id, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); ep->com.cm_id->rem_ref(ep->com.cm_id); ep->com.cm_id = NULL; - ep->com.qp = NULL; + set_bit(CLOSE_UPCALL, &ep->com.history); } } static int abort_connection(struct c4iw_ep *ep, struct sk_buff *skb, gfp_t gfp) { PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); - close_complete_upcall(ep); - state_set(&ep->com, ABORTING); + __state_set(&ep->com, ABORTING); + set_bit(ABORT_CONN, &ep->com.history); return send_abort(ep, skb, gfp); } @@ -746,6 +1105,7 @@ static void peer_close_upcall(struct c4iw_ep *ep) PDBG("peer close delivered ep %p cm_id %p tid %u\n", ep, ep->com.cm_id, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); + set_bit(DISCONN_UPCALL, &ep->com.history); } } @@ -763,7 +1123,7 @@ static void peer_abort_upcall(struct c4iw_ep *ep) ep->com.cm_id->event_handler(ep->com.cm_id, &event); ep->com.cm_id->rem_ref(ep->com.cm_id); ep->com.cm_id = NULL; - ep->com.qp = NULL; + set_bit(ABORT_UPCALL, &ep->com.history); } } @@ -775,45 +1135,74 @@ static void connect_reply_upcall(struct c4iw_ep *ep, int status) memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CONNECT_REPLY; event.status = status; - event.local_addr = ep->com.local_addr; - event.remote_addr = ep->com.remote_addr; + memcpy(&event.local_addr, &ep->com.local_addr, + sizeof(ep->com.local_addr)); + memcpy(&event.remote_addr, &ep->com.remote_addr, + sizeof(ep->com.remote_addr)); if ((status == 0) || (status == -ECONNREFUSED)) { - event.private_data_len = ep->plen; - event.private_data = ep->mpa_pkt + sizeof(struct mpa_message); + if (!ep->tried_with_mpa_v1) { + /* this means MPA_v2 is used */ + event.private_data_len = ep->plen - + sizeof(struct mpa_v2_conn_params); + event.private_data = ep->mpa_pkt + + sizeof(struct mpa_message) + + sizeof(struct mpa_v2_conn_params); + } else { + /* this means MPA_v1 is used */ + event.private_data_len = ep->plen; + event.private_data = ep->mpa_pkt + + sizeof(struct mpa_message); + } } PDBG("%s ep %p tid %u status %d\n", __func__, ep, ep->hwtid, status); + set_bit(CONN_RPL_UPCALL, &ep->com.history); ep->com.cm_id->event_handler(ep->com.cm_id, &event); if (status < 0) { ep->com.cm_id->rem_ref(ep->com.cm_id); ep->com.cm_id = NULL; - ep->com.qp = NULL; } } -static void connect_request_upcall(struct c4iw_ep *ep) +static int connect_request_upcall(struct c4iw_ep *ep) { struct iw_cm_event event; + int ret; PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_CONNECT_REQUEST; - event.local_addr = ep->com.local_addr; - event.remote_addr = ep->com.remote_addr; - event.private_data_len = ep->plen; - event.private_data = ep->mpa_pkt + sizeof(struct mpa_message); + memcpy(&event.local_addr, &ep->com.local_addr, + sizeof(ep->com.local_addr)); + memcpy(&event.remote_addr, &ep->com.remote_addr, + sizeof(ep->com.remote_addr)); event.provider_data = ep; - if (state_read(&ep->parent_ep->com) != DEAD) { - c4iw_get_ep(&ep->com); - ep->parent_ep->com.cm_id->event_handler( - ep->parent_ep->com.cm_id, - &event); + if (!ep->tried_with_mpa_v1) { + /* this means MPA_v2 is used */ + event.ord = ep->ord; + event.ird = ep->ird; + event.private_data_len = ep->plen - + sizeof(struct mpa_v2_conn_params); + event.private_data = ep->mpa_pkt + sizeof(struct mpa_message) + + sizeof(struct mpa_v2_conn_params); + } else { + /* this means MPA_v1 is used. Send max supported */ + event.ord = c4iw_max_read_depth; + event.ird = c4iw_max_read_depth; + event.private_data_len = ep->plen; + event.private_data = ep->mpa_pkt + sizeof(struct mpa_message); } + c4iw_get_ep(&ep->com); + ret = ep->parent_ep->com.cm_id->event_handler(ep->parent_ep->com.cm_id, + &event); + if (ret) + c4iw_put_ep(&ep->com); + set_bit(CONNREQ_UPCALL, &ep->com.history); c4iw_put_ep(&ep->parent_ep->com); - ep->parent_ep = NULL; + return ret; } static void established_upcall(struct c4iw_ep *ep) @@ -823,9 +1212,12 @@ static void established_upcall(struct c4iw_ep *ep) PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); memset(&event, 0, sizeof(event)); event.event = IW_CM_EVENT_ESTABLISHED; + event.ird = ep->ird; + event.ord = ep->ord; if (ep->com.cm_id) { PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); ep->com.cm_id->event_handler(ep->com.cm_id, &event); + set_bit(ESTAB_UPCALL, &ep->com.history); } } @@ -842,6 +1234,14 @@ static int update_rx_credits(struct c4iw_ep *ep, u32 credits) return 0; } + /* + * If we couldn't specify the entire rcv window at connection setup + * due to the limit in the number of bits in the RCV_BUFSIZ field, + * then add the overage in to the credits returned. + */ + if (ep->rcv_win > RCV_BUFSIZ_MASK * 1024) + credits += ep->rcv_win - RCV_BUFSIZ_MASK * 1024; + req = (struct cpl_rx_data_ack *) skb_put(skb, wrlen); memset(req, 0, wrlen); INIT_TP_WR(req, ep->hwtid); @@ -855,24 +1255,27 @@ static int update_rx_credits(struct c4iw_ep *ep, u32 credits) return credits; } -static void process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) +static int process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) { struct mpa_message *mpa; + struct mpa_v2_conn_params *mpa_v2_params; u16 plen; + u16 resp_ird, resp_ord; + u8 rtr_mismatch = 0, insuff_ird = 0; struct c4iw_qp_attributes attrs; enum c4iw_qp_attr_mask mask; int err; + int disconnect = 0; PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); /* - * Stop mpa timer. If it expired, then the state has - * changed and we bail since ep_timeout already aborted - * the connection. + * Stop mpa timer. If it expired, then + * we ignore the MPA reply. process_timeout() + * will abort the connection. */ - stop_ep_timer(ep); - if (state_read(&ep->com) != MPA_REQ_SENT) - return; + if (stop_ep_timer(ep)) + return 0; /* * If we get more than the supported amount of private data @@ -894,11 +1297,13 @@ static void process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) * if we don't even have the mpa message, then bail. */ if (ep->mpa_pkt_len < sizeof(*mpa)) - return; + return 0; mpa = (struct mpa_message *) ep->mpa_pkt; /* Validate MPA header. */ - if (mpa->revision != mpa_rev) { + if (mpa->revision > mpa_rev) { + printk(KERN_ERR MOD "%s MPA version mismatch. Local = %d," + " Received = %d\n", __func__, mpa_rev, mpa->revision); err = -EPROTO; goto err; } @@ -932,7 +1337,7 @@ static void process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) * We'll continue process when more data arrives. */ if (ep->mpa_pkt_len < (sizeof(*mpa) + plen)) - return; + return 0; if (mpa->flags & MPA_REJECT) { err = -ECONNREFUSED; @@ -944,17 +1349,70 @@ static void process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) * start reply message including private data. And * the MPA header is valid. */ - state_set(&ep->com, FPDU_MODE); + __state_set(&ep->com, FPDU_MODE); ep->mpa_attr.crc_enabled = (mpa->flags & MPA_CRC) | crc_enabled ? 1 : 0; ep->mpa_attr.recv_marker_enabled = markers_enabled; ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0; - ep->mpa_attr.version = mpa_rev; - ep->mpa_attr.p2p_type = peer2peer ? p2p_type : - FW_RI_INIT_P2PTYPE_DISABLED; + ep->mpa_attr.version = mpa->revision; + ep->mpa_attr.p2p_type = FW_RI_INIT_P2PTYPE_DISABLED; + + if (mpa->revision == 2) { + ep->mpa_attr.enhanced_rdma_conn = + mpa->flags & MPA_ENHANCED_RDMA_CONN ? 1 : 0; + if (ep->mpa_attr.enhanced_rdma_conn) { + mpa_v2_params = (struct mpa_v2_conn_params *) + (ep->mpa_pkt + sizeof(*mpa)); + resp_ird = ntohs(mpa_v2_params->ird) & + MPA_V2_IRD_ORD_MASK; + resp_ord = ntohs(mpa_v2_params->ord) & + MPA_V2_IRD_ORD_MASK; + + /* + * This is a double-check. Ideally, below checks are + * not required since ird/ord stuff has been taken + * care of in c4iw_accept_cr + */ + if ((ep->ird < resp_ord) || (ep->ord > resp_ird)) { + err = -ENOMEM; + ep->ird = resp_ord; + ep->ord = resp_ird; + insuff_ird = 1; + } + + if (ntohs(mpa_v2_params->ird) & + MPA_V2_PEER2PEER_MODEL) { + if (ntohs(mpa_v2_params->ord) & + MPA_V2_RDMA_WRITE_RTR) + ep->mpa_attr.p2p_type = + FW_RI_INIT_P2PTYPE_RDMA_WRITE; + else if (ntohs(mpa_v2_params->ord) & + MPA_V2_RDMA_READ_RTR) + ep->mpa_attr.p2p_type = + FW_RI_INIT_P2PTYPE_READ_REQ; + } + } + } else if (mpa->revision == 1) + if (peer2peer) + ep->mpa_attr.p2p_type = p2p_type; + PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, " - "xmit_marker_enabled=%d, version=%d\n", __func__, - ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, - ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version); + "xmit_marker_enabled=%d, version=%d p2p_type=%d local-p2p_type = " + "%d\n", __func__, ep->mpa_attr.crc_enabled, + ep->mpa_attr.recv_marker_enabled, + ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version, + ep->mpa_attr.p2p_type, p2p_type); + + /* + * If responder's RTR does not match with that of initiator, assign + * FW_RI_INIT_P2PTYPE_DISABLED in mpa attributes so that RTR is not + * generated when moving QP to RTS state. + * A TERM message will be sent after QP has moved to RTS state + */ + if ((ep->mpa_attr.version == 2) && peer2peer && + (ep->mpa_attr.p2p_type != p2p_type)) { + ep->mpa_attr.p2p_type = FW_RI_INIT_P2PTYPE_DISABLED; + rtr_mismatch = 1; + } attrs.mpa_attr = ep->mpa_attr; attrs.max_ird = ep->ird; @@ -971,31 +1429,66 @@ static void process_mpa_reply(struct c4iw_ep *ep, struct sk_buff *skb) ep->com.qp, mask, &attrs, 1); if (err) goto err; + + /* + * If responder's RTR requirement did not match with what initiator + * supports, generate TERM message + */ + if (rtr_mismatch) { + printk(KERN_ERR "%s: RTR mismatch, sending TERM\n", __func__); + attrs.layer_etype = LAYER_MPA | DDP_LLP; + attrs.ecode = MPA_NOMATCH_RTR; + attrs.next_state = C4IW_QP_STATE_TERMINATE; + attrs.send_term = 1; + err = c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp, + C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); + err = -ENOMEM; + disconnect = 1; + goto out; + } + + /* + * Generate TERM if initiator IRD is not sufficient for responder + * provided ORD. Currently, we do the same behaviour even when + * responder provided IRD is also not sufficient as regards to + * initiator ORD. + */ + if (insuff_ird) { + printk(KERN_ERR "%s: Insufficient IRD, sending TERM\n", + __func__); + attrs.layer_etype = LAYER_MPA | DDP_LLP; + attrs.ecode = MPA_INSUFF_IRD; + attrs.next_state = C4IW_QP_STATE_TERMINATE; + attrs.send_term = 1; + err = c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp, + C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); + err = -ENOMEM; + disconnect = 1; + goto out; + } goto out; err: - state_set(&ep->com, ABORTING); + __state_set(&ep->com, ABORTING); send_abort(ep, skb, GFP_KERNEL); out: connect_reply_upcall(ep, err); - return; + return disconnect; } static void process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) { struct mpa_message *mpa; + struct mpa_v2_conn_params *mpa_v2_params; u16 plen; PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); - if (state_read(&ep->com) != MPA_REQ_WAIT) - return; - /* * If we get more than the supported amount of private data * then we must fail this connection. */ if (ep->mpa_pkt_len + skb->len > sizeof(ep->mpa_pkt)) { - stop_ep_timer(ep); + (void)stop_ep_timer(ep); abort_connection(ep, skb, GFP_KERNEL); return; } @@ -1017,18 +1510,21 @@ static void process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) return; PDBG("%s enter (%s line %u)\n", __func__, __FILE__, __LINE__); - stop_ep_timer(ep); mpa = (struct mpa_message *) ep->mpa_pkt; /* * Validate MPA Header. */ - if (mpa->revision != mpa_rev) { + if (mpa->revision > mpa_rev) { + printk(KERN_ERR MOD "%s MPA version mismatch. Local = %d," + " Received = %d\n", __func__, mpa_rev, mpa->revision); + (void)stop_ep_timer(ep); abort_connection(ep, skb, GFP_KERNEL); return; } if (memcmp(mpa->key, MPA_KEY_REQ, sizeof(mpa->key))) { + (void)stop_ep_timer(ep); abort_connection(ep, skb, GFP_KERNEL); return; } @@ -1039,6 +1535,7 @@ static void process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) * Fail if there's too much private data. */ if (plen > MPA_MAX_PRIVATE_DATA) { + (void)stop_ep_timer(ep); abort_connection(ep, skb, GFP_KERNEL); return; } @@ -1047,6 +1544,7 @@ static void process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) * If plen does not account for pkt size */ if (ep->mpa_pkt_len > (sizeof(*mpa) + plen)) { + (void)stop_ep_timer(ep); abort_connection(ep, skb, GFP_KERNEL); return; } @@ -1066,19 +1564,61 @@ static void process_mpa_request(struct c4iw_ep *ep, struct sk_buff *skb) ep->mpa_attr.crc_enabled = (mpa->flags & MPA_CRC) | crc_enabled ? 1 : 0; ep->mpa_attr.recv_marker_enabled = markers_enabled; ep->mpa_attr.xmit_marker_enabled = mpa->flags & MPA_MARKERS ? 1 : 0; - ep->mpa_attr.version = mpa_rev; - ep->mpa_attr.p2p_type = peer2peer ? p2p_type : - FW_RI_INIT_P2PTYPE_DISABLED; + ep->mpa_attr.version = mpa->revision; + if (mpa->revision == 1) + ep->tried_with_mpa_v1 = 1; + ep->mpa_attr.p2p_type = FW_RI_INIT_P2PTYPE_DISABLED; + + if (mpa->revision == 2) { + ep->mpa_attr.enhanced_rdma_conn = + mpa->flags & MPA_ENHANCED_RDMA_CONN ? 1 : 0; + if (ep->mpa_attr.enhanced_rdma_conn) { + mpa_v2_params = (struct mpa_v2_conn_params *) + (ep->mpa_pkt + sizeof(*mpa)); + ep->ird = ntohs(mpa_v2_params->ird) & + MPA_V2_IRD_ORD_MASK; + ep->ord = ntohs(mpa_v2_params->ord) & + MPA_V2_IRD_ORD_MASK; + if (ntohs(mpa_v2_params->ird) & MPA_V2_PEER2PEER_MODEL) + if (peer2peer) { + if (ntohs(mpa_v2_params->ord) & + MPA_V2_RDMA_WRITE_RTR) + ep->mpa_attr.p2p_type = + FW_RI_INIT_P2PTYPE_RDMA_WRITE; + else if (ntohs(mpa_v2_params->ord) & + MPA_V2_RDMA_READ_RTR) + ep->mpa_attr.p2p_type = + FW_RI_INIT_P2PTYPE_READ_REQ; + } + } + } else if (mpa->revision == 1) + if (peer2peer) + ep->mpa_attr.p2p_type = p2p_type; + PDBG("%s - crc_enabled=%d, recv_marker_enabled=%d, " "xmit_marker_enabled=%d, version=%d p2p_type=%d\n", __func__, ep->mpa_attr.crc_enabled, ep->mpa_attr.recv_marker_enabled, ep->mpa_attr.xmit_marker_enabled, ep->mpa_attr.version, ep->mpa_attr.p2p_type); - state_set(&ep->com, MPA_REQ_RCVD); - - /* drive upcall */ - connect_request_upcall(ep); + /* + * If the endpoint timer already expired, then we ignore + * the start request. process_timeout() will abort + * the connection. + */ + if (!stop_ep_timer(ep)) { + __state_set(&ep->com, MPA_REQ_RCVD); + + /* drive upcall */ + mutex_lock(&ep->parent_ep->com.mutex); + if (ep->parent_ep->com.state != DEAD) { + if (connect_request_upcall(ep)) + abort_connection(ep, skb, GFP_KERNEL); + } else { + abort_connection(ep, skb, GFP_KERNEL); + } + mutex_unlock(&ep->parent_ep->com.mutex); + } return; } @@ -1089,38 +1629,49 @@ static int rx_data(struct c4iw_dev *dev, struct sk_buff *skb) unsigned int dlen = ntohs(hdr->len); unsigned int tid = GET_TID(hdr); struct tid_info *t = dev->rdev.lldi.tids; + __u8 status = hdr->status; + int disconnect = 0; ep = lookup_tid(t, tid); + if (!ep) + return 0; PDBG("%s ep %p tid %u dlen %u\n", __func__, ep, ep->hwtid, dlen); skb_pull(skb, sizeof(*hdr)); skb_trim(skb, dlen); - - ep->rcv_seq += dlen; - BUG_ON(ep->rcv_seq != (ntohl(hdr->seq) + dlen)); + mutex_lock(&ep->com.mutex); /* update RX credits */ update_rx_credits(ep, dlen); - switch (state_read(&ep->com)) { + switch (ep->com.state) { case MPA_REQ_SENT: - process_mpa_reply(ep, skb); + ep->rcv_seq += dlen; + disconnect = process_mpa_reply(ep, skb); break; case MPA_REQ_WAIT: + ep->rcv_seq += dlen; process_mpa_request(ep, skb); break; - case MPA_REP_SENT: + case FPDU_MODE: { + struct c4iw_qp_attributes attrs; + BUG_ON(!ep->com.qp); + if (status) + pr_err("%s Unexpected streaming data." \ + " qpid %u ep %p state %d tid %u status %d\n", + __func__, ep->com.qp->wq.sq.qid, ep, + ep->com.state, ep->hwtid, status); + attrs.next_state = C4IW_QP_STATE_TERMINATE; + c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp, + C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); + disconnect = 1; break; + } default: - printk(KERN_ERR MOD "%s Unexpected streaming data." - " ep %p state %d tid %u\n", - __func__, ep, state_read(&ep->com), ep->hwtid); - - /* - * The ep will timeout and inform the ULP of the failure. - * See ep_timeout(). - */ break; } + mutex_unlock(&ep->com.mutex); + if (disconnect) + c4iw_ep_disconnect(ep, 0, GFP_KERNEL); return 0; } @@ -1133,11 +1684,15 @@ static int abort_rpl(struct c4iw_dev *dev, struct sk_buff *skb) struct tid_info *t = dev->rdev.lldi.tids; ep = lookup_tid(t, tid); + if (!ep) { + printk(KERN_WARNING MOD "Abort rpl to freed endpoint\n"); + return 0; + } PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); - BUG_ON(!ep); mutex_lock(&ep->com.mutex); switch (ep->com.state) { case ABORTING: + c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); __state_set(&ep->com, DEAD); release = 1; break; @@ -1153,6 +1708,78 @@ static int abort_rpl(struct c4iw_dev *dev, struct sk_buff *skb) return 0; } +static void send_fw_act_open_req(struct c4iw_ep *ep, unsigned int atid) +{ + struct sk_buff *skb; + struct fw_ofld_connection_wr *req; + unsigned int mtu_idx; + int wscale; + struct sockaddr_in *sin; + int win; + + skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); + req = (struct fw_ofld_connection_wr *)__skb_put(skb, sizeof(*req)); + memset(req, 0, sizeof(*req)); + req->op_compl = htonl(V_WR_OP(FW_OFLD_CONNECTION_WR)); + req->len16_pkd = htonl(FW_WR_LEN16(DIV_ROUND_UP(sizeof(*req), 16))); + req->le.filter = cpu_to_be32(cxgb4_select_ntuple( + ep->com.dev->rdev.lldi.ports[0], + ep->l2t)); + sin = (struct sockaddr_in *)&ep->com.mapped_local_addr; + req->le.lport = sin->sin_port; + req->le.u.ipv4.lip = sin->sin_addr.s_addr; + sin = (struct sockaddr_in *)&ep->com.mapped_remote_addr; + req->le.pport = sin->sin_port; + req->le.u.ipv4.pip = sin->sin_addr.s_addr; + req->tcb.t_state_to_astid = + htonl(V_FW_OFLD_CONNECTION_WR_T_STATE(TCP_SYN_SENT) | + V_FW_OFLD_CONNECTION_WR_ASTID(atid)); + req->tcb.cplrxdataack_cplpassacceptrpl = + htons(F_FW_OFLD_CONNECTION_WR_CPLRXDATAACK); + req->tcb.tx_max = (__force __be32) jiffies; + req->tcb.rcv_adv = htons(1); + best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx, + enable_tcp_timestamps); + wscale = compute_wscale(rcv_win); + + /* + * Specify the largest window that will fit in opt0. The + * remainder will be specified in the rx_data_ack. + */ + win = ep->rcv_win >> 10; + if (win > RCV_BUFSIZ_MASK) + win = RCV_BUFSIZ_MASK; + + req->tcb.opt0 = (__force __be64) (TCAM_BYPASS(1) | + (nocong ? NO_CONG(1) : 0) | + KEEP_ALIVE(1) | + DELACK(1) | + WND_SCALE(wscale) | + MSS_IDX(mtu_idx) | + L2T_IDX(ep->l2t->idx) | + TX_CHAN(ep->tx_chan) | + SMAC_SEL(ep->smac_idx) | + DSCP(ep->tos) | + ULP_MODE(ULP_MODE_TCPDDP) | + RCV_BUFSIZ(win)); + req->tcb.opt2 = (__force __be32) (PACE(1) | + TX_QUEUE(ep->com.dev->rdev.lldi.tx_modq[ep->tx_chan]) | + RX_CHANNEL(0) | + CCTRL_ECN(enable_ecn) | + RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid)); + if (enable_tcp_timestamps) + req->tcb.opt2 |= (__force __be32) TSTAMPS_EN(1); + if (enable_tcp_sack) + req->tcb.opt2 |= (__force __be32) SACK_EN(1); + if (wscale && enable_tcp_window_scaling) + req->tcb.opt2 |= (__force __be32) WND_SCALE_EN(1); + req->tcb.opt0 = cpu_to_be64((__force u64) req->tcb.opt0); + req->tcb.opt2 = cpu_to_be32((__force u32) req->tcb.opt2); + set_wr_txq(skb, CPL_PRIORITY_CONTROL, ep->ctrlq_idx); + set_bit(ACT_OFLD_CONN, &ep->com.history); + c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); +} + /* * Return whether a failed active open has allocated a TID */ @@ -1162,6 +1789,190 @@ static inline int act_open_has_tid(int status) status != CPL_ERR_ARP_MISS; } +/* Returns whether a CPL status conveys negative advice. + */ +static int is_neg_adv(unsigned int status) +{ + return status == CPL_ERR_RTX_NEG_ADVICE || + status == CPL_ERR_PERSIST_NEG_ADVICE || + status == CPL_ERR_KEEPALV_NEG_ADVICE; +} + +static void set_tcp_window(struct c4iw_ep *ep, struct port_info *pi) +{ + ep->snd_win = snd_win; + ep->rcv_win = rcv_win; + PDBG("%s snd_win %d rcv_win %d\n", __func__, ep->snd_win, ep->rcv_win); +} + +#define ACT_OPEN_RETRY_COUNT 2 + +static int import_ep(struct c4iw_ep *ep, int iptype, __u8 *peer_ip, + struct dst_entry *dst, struct c4iw_dev *cdev, + bool clear_mpa_v1) +{ + struct neighbour *n; + int err, step; + struct net_device *pdev; + + n = dst_neigh_lookup(dst, peer_ip); + if (!n) + return -ENODEV; + + rcu_read_lock(); + err = -ENOMEM; + if (n->dev->flags & IFF_LOOPBACK) { + if (iptype == 4) + pdev = ip_dev_find(&init_net, *(__be32 *)peer_ip); + else if (IS_ENABLED(CONFIG_IPV6)) + for_each_netdev(&init_net, pdev) { + if (ipv6_chk_addr(&init_net, + (struct in6_addr *)peer_ip, + pdev, 1)) + break; + } + else + pdev = NULL; + + if (!pdev) { + err = -ENODEV; + goto out; + } + ep->l2t = cxgb4_l2t_get(cdev->rdev.lldi.l2t, + n, pdev, 0); + if (!ep->l2t) + goto out; + ep->mtu = pdev->mtu; + ep->tx_chan = cxgb4_port_chan(pdev); + ep->smac_idx = (cxgb4_port_viid(pdev) & 0x7F) << 1; + step = cdev->rdev.lldi.ntxq / + cdev->rdev.lldi.nchan; + ep->txq_idx = cxgb4_port_idx(pdev) * step; + step = cdev->rdev.lldi.nrxq / + cdev->rdev.lldi.nchan; + ep->ctrlq_idx = cxgb4_port_idx(pdev); + ep->rss_qid = cdev->rdev.lldi.rxq_ids[ + cxgb4_port_idx(pdev) * step]; + set_tcp_window(ep, (struct port_info *)netdev_priv(pdev)); + dev_put(pdev); + } else { + pdev = get_real_dev(n->dev); + ep->l2t = cxgb4_l2t_get(cdev->rdev.lldi.l2t, + n, pdev, 0); + if (!ep->l2t) + goto out; + ep->mtu = dst_mtu(dst); + ep->tx_chan = cxgb4_port_chan(pdev); + ep->smac_idx = (cxgb4_port_viid(pdev) & 0x7F) << 1; + step = cdev->rdev.lldi.ntxq / + cdev->rdev.lldi.nchan; + ep->txq_idx = cxgb4_port_idx(pdev) * step; + ep->ctrlq_idx = cxgb4_port_idx(pdev); + step = cdev->rdev.lldi.nrxq / + cdev->rdev.lldi.nchan; + ep->rss_qid = cdev->rdev.lldi.rxq_ids[ + cxgb4_port_idx(pdev) * step]; + set_tcp_window(ep, (struct port_info *)netdev_priv(pdev)); + + if (clear_mpa_v1) { + ep->retry_with_mpa_v1 = 0; + ep->tried_with_mpa_v1 = 0; + } + } + err = 0; +out: + rcu_read_unlock(); + + neigh_release(n); + + return err; +} + +static int c4iw_reconnect(struct c4iw_ep *ep) +{ + int err = 0; + struct sockaddr_in *laddr = (struct sockaddr_in *) + &ep->com.cm_id->local_addr; + struct sockaddr_in *raddr = (struct sockaddr_in *) + &ep->com.cm_id->remote_addr; + struct sockaddr_in6 *laddr6 = (struct sockaddr_in6 *) + &ep->com.cm_id->local_addr; + struct sockaddr_in6 *raddr6 = (struct sockaddr_in6 *) + &ep->com.cm_id->remote_addr; + int iptype; + __u8 *ra; + + PDBG("%s qp %p cm_id %p\n", __func__, ep->com.qp, ep->com.cm_id); + init_timer(&ep->timer); + + /* + * Allocate an active TID to initiate a TCP connection. + */ + ep->atid = cxgb4_alloc_atid(ep->com.dev->rdev.lldi.tids, ep); + if (ep->atid == -1) { + pr_err("%s - cannot alloc atid.\n", __func__); + err = -ENOMEM; + goto fail2; + } + insert_handle(ep->com.dev, &ep->com.dev->atid_idr, ep, ep->atid); + + /* find a route */ + if (ep->com.cm_id->local_addr.ss_family == AF_INET) { + ep->dst = find_route(ep->com.dev, laddr->sin_addr.s_addr, + raddr->sin_addr.s_addr, laddr->sin_port, + raddr->sin_port, 0); + iptype = 4; + ra = (__u8 *)&raddr->sin_addr; + } else { + ep->dst = find_route6(ep->com.dev, laddr6->sin6_addr.s6_addr, + raddr6->sin6_addr.s6_addr, + laddr6->sin6_port, raddr6->sin6_port, 0, + raddr6->sin6_scope_id); + iptype = 6; + ra = (__u8 *)&raddr6->sin6_addr; + } + if (!ep->dst) { + pr_err("%s - cannot find route.\n", __func__); + err = -EHOSTUNREACH; + goto fail3; + } + err = import_ep(ep, iptype, ra, ep->dst, ep->com.dev, false); + if (err) { + pr_err("%s - cannot alloc l2e.\n", __func__); + goto fail4; + } + + PDBG("%s txq_idx %u tx_chan %u smac_idx %u rss_qid %u l2t_idx %u\n", + __func__, ep->txq_idx, ep->tx_chan, ep->smac_idx, ep->rss_qid, + ep->l2t->idx); + + state_set(&ep->com, CONNECTING); + ep->tos = 0; + + /* send connect request to rnic */ + err = send_connect(ep); + if (!err) + goto out; + + cxgb4_l2t_release(ep->l2t); +fail4: + dst_release(ep->dst); +fail3: + remove_handle(ep->com.dev, &ep->com.dev->atid_idr, ep->atid); + cxgb4_free_atid(ep->com.dev->rdev.lldi.tids, ep->atid); +fail2: + /* + * remember to send notification to upper layer. + * We are in here so the upper layer is not aware that this is + * re-connect attempt and so, upper layer is still waiting for + * response of 1st connect request. + */ + connect_reply_upcall(ep, -ECONNRESET); + c4iw_put_ep(&ep->com); +out: + return err; +} + static int act_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb) { struct c4iw_ep *ep; @@ -1170,24 +1981,81 @@ static int act_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb) ntohl(rpl->atid_status))); struct tid_info *t = dev->rdev.lldi.tids; int status = GET_AOPEN_STATUS(ntohl(rpl->atid_status)); + struct sockaddr_in *la; + struct sockaddr_in *ra; + struct sockaddr_in6 *la6; + struct sockaddr_in6 *ra6; ep = lookup_atid(t, atid); + la = (struct sockaddr_in *)&ep->com.mapped_local_addr; + ra = (struct sockaddr_in *)&ep->com.mapped_remote_addr; + la6 = (struct sockaddr_in6 *)&ep->com.mapped_local_addr; + ra6 = (struct sockaddr_in6 *)&ep->com.mapped_remote_addr; PDBG("%s ep %p atid %u status %u errno %d\n", __func__, ep, atid, status, status2errno(status)); - if (status == CPL_ERR_RTX_NEG_ADVICE) { + if (is_neg_adv(status)) { printk(KERN_WARNING MOD "Connection problems for atid %u\n", atid); return 0; } + set_bit(ACT_OPEN_RPL, &ep->com.history); + + /* + * Log interesting failures. + */ + switch (status) { + case CPL_ERR_CONN_RESET: + case CPL_ERR_CONN_TIMEDOUT: + break; + case CPL_ERR_TCAM_FULL: + mutex_lock(&dev->rdev.stats.lock); + dev->rdev.stats.tcam_full++; + mutex_unlock(&dev->rdev.stats.lock); + if (ep->com.local_addr.ss_family == AF_INET && + dev->rdev.lldi.enable_fw_ofld_conn) { + send_fw_act_open_req(ep, + GET_TID_TID(GET_AOPEN_ATID( + ntohl(rpl->atid_status)))); + return 0; + } + break; + case CPL_ERR_CONN_EXIST: + if (ep->retry_count++ < ACT_OPEN_RETRY_COUNT) { + set_bit(ACT_RETRY_INUSE, &ep->com.history); + remove_handle(ep->com.dev, &ep->com.dev->atid_idr, + atid); + cxgb4_free_atid(t, atid); + dst_release(ep->dst); + cxgb4_l2t_release(ep->l2t); + c4iw_reconnect(ep); + return 0; + } + break; + default: + if (ep->com.local_addr.ss_family == AF_INET) { + pr_info("Active open failure - atid %u status %u errno %d %pI4:%u->%pI4:%u\n", + atid, status, status2errno(status), + &la->sin_addr.s_addr, ntohs(la->sin_port), + &ra->sin_addr.s_addr, ntohs(ra->sin_port)); + } else { + pr_info("Active open failure - atid %u status %u errno %d %pI6:%u->%pI6:%u\n", + atid, status, status2errno(status), + la6->sin6_addr.s6_addr, ntohs(la6->sin6_port), + ra6->sin6_addr.s6_addr, ntohs(ra6->sin6_port)); + } + break; + } + connect_reply_upcall(ep, status2errno(status)); state_set(&ep->com, DEAD); if (status && act_open_has_tid(status)) cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, GET_TID(rpl)); + remove_handle(ep->com.dev, &ep->com.dev->atid_idr, atid); cxgb4_free_atid(t, atid); dst_release(ep->dst); cxgb4_l2t_release(ep->l2t); @@ -1204,39 +2072,17 @@ static int pass_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb) struct c4iw_listen_ep *ep = lookup_stid(t, stid); if (!ep) { - printk(KERN_ERR MOD "stid %d lookup failure!\n", stid); - return 0; + PDBG("%s stid %d lookup failure!\n", __func__, stid); + goto out; } PDBG("%s ep %p status %d error %d\n", __func__, ep, rpl->status, status2errno(rpl->status)); - ep->com.wr_wait.ret = status2errno(rpl->status); - ep->com.wr_wait.done = 1; - wake_up(&ep->com.wr_wait.wait); + c4iw_wake_up(&ep->com.wr_wait, status2errno(rpl->status)); +out: return 0; } -static int listen_stop(struct c4iw_listen_ep *ep) -{ - struct sk_buff *skb; - struct cpl_close_listsvr_req *req; - - PDBG("%s ep %p\n", __func__, ep); - skb = get_skb(NULL, sizeof(*req), GFP_KERNEL); - if (!skb) { - printk(KERN_ERR MOD "%s - failed to alloc skb\n", __func__); - return -ENOMEM; - } - req = (struct cpl_close_listsvr_req *) skb_put(skb, sizeof(*req)); - INIT_TP_WR(req, 0); - OPCODE_TID(req) = cpu_to_be32(MK_OPCODE_TID(CPL_CLOSE_LISTSRV_REQ, - ep->stid)); - req->reply_ctrl = cpu_to_be16( - QUEUENO(ep->com.dev->rdev.lldi.rxq_ids[0])); - set_wr_txq(skb, CPL_PRIORITY_SETUP, 0); - return c4iw_ofld_send(&ep->com.dev->rdev, skb); -} - static int close_listsrv_rpl(struct c4iw_dev *dev, struct sk_buff *skb) { struct cpl_close_listsvr_rpl *rpl = cplhdr(skb); @@ -1245,13 +2091,11 @@ static int close_listsrv_rpl(struct c4iw_dev *dev, struct sk_buff *skb) struct c4iw_listen_ep *ep = lookup_stid(t, stid); PDBG("%s ep %p\n", __func__, ep); - ep->com.wr_wait.ret = status2errno(rpl->status); - ep->com.wr_wait.done = 1; - wake_up(&ep->com.wr_wait.wait); + c4iw_wake_up(&ep->com.wr_wait, status2errno(rpl->status)); return 0; } -static void accept_cr(struct c4iw_ep *ep, __be32 peer_ip, struct sk_buff *skb, +static void accept_cr(struct c4iw_ep *ep, struct sk_buff *skb, struct cpl_pass_accept_req *req) { struct cpl_pass_accept_rpl *rpl; @@ -1259,22 +2103,47 @@ static void accept_cr(struct c4iw_ep *ep, __be32 peer_ip, struct sk_buff *skb, u64 opt0; u32 opt2; int wscale; + struct cpl_t5_pass_accept_rpl *rpl5 = NULL; + int win; PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); BUG_ON(skb_cloned(skb)); - skb_trim(skb, sizeof(*rpl)); + skb_get(skb); - cxgb4_best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx); + rpl = cplhdr(skb); + if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) { + skb_trim(skb, roundup(sizeof(*rpl5), 16)); + rpl5 = (void *)rpl; + INIT_TP_WR(rpl5, ep->hwtid); + } else { + skb_trim(skb, sizeof(*rpl)); + INIT_TP_WR(rpl, ep->hwtid); + } + OPCODE_TID(rpl) = cpu_to_be32(MK_OPCODE_TID(CPL_PASS_ACCEPT_RPL, + ep->hwtid)); + + best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx, + enable_tcp_timestamps && req->tcpopt.tstamp); wscale = compute_wscale(rcv_win); - opt0 = KEEP_ALIVE(1) | + + /* + * Specify the largest window that will fit in opt0. The + * remainder will be specified in the rx_data_ack. + */ + win = ep->rcv_win >> 10; + if (win > RCV_BUFSIZ_MASK) + win = RCV_BUFSIZ_MASK; + opt0 = (nocong ? NO_CONG(1) : 0) | + KEEP_ALIVE(1) | DELACK(1) | WND_SCALE(wscale) | MSS_IDX(mtu_idx) | L2T_IDX(ep->l2t->idx) | TX_CHAN(ep->tx_chan) | SMAC_SEL(ep->smac_idx) | - DSCP(ep->tos) | - RCV_BUFSIZ(rcv_win>>10); + DSCP(ep->tos >> 2) | + ULP_MODE(ULP_MODE_TCPDDP) | + RCV_BUFSIZ(win); opt2 = RX_CHANNEL(0) | RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid); @@ -1284,47 +2153,72 @@ static void accept_cr(struct c4iw_ep *ep, __be32 peer_ip, struct sk_buff *skb, opt2 |= SACK_EN(1); if (wscale && enable_tcp_window_scaling) opt2 |= WND_SCALE_EN(1); + if (enable_ecn) { + const struct tcphdr *tcph; + u32 hlen = ntohl(req->hdr_len); + + tcph = (const void *)(req + 1) + G_ETH_HDR_LEN(hlen) + + G_IP_HDR_LEN(hlen); + if (tcph->ece && tcph->cwr) + opt2 |= CCTRL_ECN(1); + } + if (is_t5(ep->com.dev->rdev.lldi.adapter_type)) { + u32 isn = (prandom_u32() & ~7UL) - 1; + opt2 |= T5_OPT_2_VALID; + opt2 |= V_CONG_CNTRL(CONG_ALG_TAHOE); + opt2 |= CONG_CNTRL_VALID; /* OPT_2_ISS for T5 */ + rpl5 = (void *)rpl; + memset(&rpl5->iss, 0, roundup(sizeof(*rpl5)-sizeof(*rpl), 16)); + if (peer2peer) + isn += 4; + rpl5->iss = cpu_to_be32(isn); + PDBG("%s iss %u\n", __func__, be32_to_cpu(rpl5->iss)); + } - rpl = cplhdr(skb); - INIT_TP_WR(rpl, ep->hwtid); - OPCODE_TID(rpl) = cpu_to_be32(MK_OPCODE_TID(CPL_PASS_ACCEPT_RPL, - ep->hwtid)); rpl->opt0 = cpu_to_be64(opt0); rpl->opt2 = cpu_to_be32(opt2); set_wr_txq(skb, CPL_PRIORITY_SETUP, ep->ctrlq_idx); + t4_set_arp_err_handler(skb, NULL, arp_failure_discard); c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t); return; } -static void reject_cr(struct c4iw_dev *dev, u32 hwtid, __be32 peer_ip, - struct sk_buff *skb) +static void reject_cr(struct c4iw_dev *dev, u32 hwtid, struct sk_buff *skb) { - PDBG("%s c4iw_dev %p tid %u peer_ip %x\n", __func__, dev, hwtid, - peer_ip); + PDBG("%s c4iw_dev %p tid %u\n", __func__, dev, hwtid); BUG_ON(skb_cloned(skb)); skb_trim(skb, sizeof(struct cpl_tid_release)); - skb_get(skb); release_tid(&dev->rdev, hwtid, skb); return; } -static void get_4tuple(struct cpl_pass_accept_req *req, - __be32 *local_ip, __be32 *peer_ip, +static void get_4tuple(struct cpl_pass_accept_req *req, int *iptype, + __u8 *local_ip, __u8 *peer_ip, __be16 *local_port, __be16 *peer_port) { int eth_len = G_ETH_HDR_LEN(be32_to_cpu(req->hdr_len)); int ip_len = G_IP_HDR_LEN(be32_to_cpu(req->hdr_len)); struct iphdr *ip = (struct iphdr *)((u8 *)(req + 1) + eth_len); + struct ipv6hdr *ip6 = (struct ipv6hdr *)((u8 *)(req + 1) + eth_len); struct tcphdr *tcp = (struct tcphdr *) ((u8 *)(req + 1) + eth_len + ip_len); - PDBG("%s saddr 0x%x daddr 0x%x sport %u dport %u\n", __func__, - ntohl(ip->saddr), ntohl(ip->daddr), ntohs(tcp->source), - ntohs(tcp->dest)); - - *peer_ip = ip->saddr; - *local_ip = ip->daddr; + if (ip->version == 4) { + PDBG("%s saddr 0x%x daddr 0x%x sport %u dport %u\n", __func__, + ntohl(ip->saddr), ntohl(ip->daddr), ntohs(tcp->source), + ntohs(tcp->dest)); + *iptype = 4; + memcpy(peer_ip, &ip->saddr, 4); + memcpy(local_ip, &ip->daddr, 4); + } else { + PDBG("%s saddr %pI6 daddr %pI6 sport %u dport %u\n", __func__, + ip6->saddr.s6_addr, ip6->daddr.s6_addr, ntohs(tcp->source), + ntohs(tcp->dest)); + *iptype = 6; + memcpy(peer_ip, ip6->saddr.s6_addr, 16); + memcpy(local_ip, ip6->daddr.s6_addr, 16); + } *peer_port = tcp->source; *local_port = tcp->dest; @@ -1333,27 +2227,24 @@ static void get_4tuple(struct cpl_pass_accept_req *req, static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) { - struct c4iw_ep *child_ep, *parent_ep; + struct c4iw_ep *child_ep = NULL, *parent_ep; struct cpl_pass_accept_req *req = cplhdr(skb); unsigned int stid = GET_POPEN_TID(ntohl(req->tos_stid)); struct tid_info *t = dev->rdev.lldi.tids; unsigned int hwtid = GET_TID(req); struct dst_entry *dst; - struct l2t_entry *l2t; - struct rtable *rt; - __be32 local_ip, peer_ip; + __u8 local_ip[16], peer_ip[16]; __be16 local_port, peer_port; - struct net_device *pdev; - u32 tx_chan, smac_idx; - u16 rss_qid; - u32 mtu; - int step; - int txq_idx, ctrlq_idx; + int err; + u16 peer_mss = ntohs(req->tcpopt.mss); + int iptype; + unsigned short hdrs; parent_ep = lookup_stid(t, stid); - PDBG("%s parent ep %p tid %u\n", __func__, parent_ep, hwtid); - - get_4tuple(req, &local_ip, &peer_ip, &local_port, &peer_port); + if (!parent_ep) { + PDBG("%s connect request on invalid stid %d\n", __func__, stid); + goto reject; + } if (state_read(&parent_ep->com) != LISTEN) { printk(KERN_ERR "%s - listening ep not in LISTEN\n", @@ -1361,88 +2252,96 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb) goto reject; } + get_4tuple(req, &iptype, local_ip, peer_ip, &local_port, &peer_port); + /* Find output route */ - rt = find_route(dev, local_ip, peer_ip, local_port, peer_port, - GET_POPEN_TOS(ntohl(req->tos_stid))); - if (!rt) { + if (iptype == 4) { + PDBG("%s parent ep %p hwtid %u laddr %pI4 raddr %pI4 lport %d rport %d peer_mss %d\n" + , __func__, parent_ep, hwtid, + local_ip, peer_ip, ntohs(local_port), + ntohs(peer_port), peer_mss); + dst = find_route(dev, *(__be32 *)local_ip, *(__be32 *)peer_ip, + local_port, peer_port, + GET_POPEN_TOS(ntohl(req->tos_stid))); + } else { + PDBG("%s parent ep %p hwtid %u laddr %pI6 raddr %pI6 lport %d rport %d peer_mss %d\n" + , __func__, parent_ep, hwtid, + local_ip, peer_ip, ntohs(local_port), + ntohs(peer_port), peer_mss); + dst = find_route6(dev, local_ip, peer_ip, local_port, peer_port, + PASS_OPEN_TOS(ntohl(req->tos_stid)), + ((struct sockaddr_in6 *) + &parent_ep->com.local_addr)->sin6_scope_id); + } + if (!dst) { printk(KERN_ERR MOD "%s - failed to find dst entry!\n", __func__); goto reject; } - dst = &rt->dst; - if (dst->neighbour->dev->flags & IFF_LOOPBACK) { - pdev = ip_dev_find(&init_net, peer_ip); - BUG_ON(!pdev); - l2t = cxgb4_l2t_get(dev->rdev.lldi.l2t, dst->neighbour, - pdev, 0); - mtu = pdev->mtu; - tx_chan = cxgb4_port_chan(pdev); - smac_idx = (cxgb4_port_viid(pdev) & 0x7F) << 1; - step = dev->rdev.lldi.ntxq / dev->rdev.lldi.nchan; - txq_idx = cxgb4_port_idx(pdev) * step; - ctrlq_idx = cxgb4_port_idx(pdev); - step = dev->rdev.lldi.nrxq / dev->rdev.lldi.nchan; - rss_qid = dev->rdev.lldi.rxq_ids[cxgb4_port_idx(pdev) * step]; - dev_put(pdev); - } else { - l2t = cxgb4_l2t_get(dev->rdev.lldi.l2t, dst->neighbour, - dst->neighbour->dev, 0); - mtu = dst_mtu(dst); - tx_chan = cxgb4_port_chan(dst->neighbour->dev); - smac_idx = (cxgb4_port_viid(dst->neighbour->dev) & 0x7F) << 1; - step = dev->rdev.lldi.ntxq / dev->rdev.lldi.nchan; - txq_idx = cxgb4_port_idx(dst->neighbour->dev) * step; - ctrlq_idx = cxgb4_port_idx(dst->neighbour->dev); - step = dev->rdev.lldi.nrxq / dev->rdev.lldi.nchan; - rss_qid = dev->rdev.lldi.rxq_ids[ - cxgb4_port_idx(dst->neighbour->dev) * step]; - } - if (!l2t) { - printk(KERN_ERR MOD "%s - failed to allocate l2t entry!\n", + + child_ep = alloc_ep(sizeof(*child_ep), GFP_KERNEL); + if (!child_ep) { + printk(KERN_ERR MOD "%s - failed to allocate ep entry!\n", __func__); dst_release(dst); goto reject; } - child_ep = alloc_ep(sizeof(*child_ep), GFP_KERNEL); - if (!child_ep) { - printk(KERN_ERR MOD "%s - failed to allocate ep entry!\n", + err = import_ep(child_ep, iptype, peer_ip, dst, dev, false); + if (err) { + printk(KERN_ERR MOD "%s - failed to allocate l2t entry!\n", __func__); - cxgb4_l2t_release(l2t); dst_release(dst); + kfree(child_ep); goto reject; } + + hdrs = sizeof(struct iphdr) + sizeof(struct tcphdr) + + ((enable_tcp_timestamps && req->tcpopt.tstamp) ? 12 : 0); + if (peer_mss && child_ep->mtu > (peer_mss + hdrs)) + child_ep->mtu = peer_mss + hdrs; + state_set(&child_ep->com, CONNECTING); child_ep->com.dev = dev; child_ep->com.cm_id = NULL; - child_ep->com.local_addr.sin_family = PF_INET; - child_ep->com.local_addr.sin_port = local_port; - child_ep->com.local_addr.sin_addr.s_addr = local_ip; - child_ep->com.remote_addr.sin_family = PF_INET; - child_ep->com.remote_addr.sin_port = peer_port; - child_ep->com.remote_addr.sin_addr.s_addr = peer_ip; + if (iptype == 4) { + struct sockaddr_in *sin = (struct sockaddr_in *) + &child_ep->com.local_addr; + sin->sin_family = PF_INET; + sin->sin_port = local_port; + sin->sin_addr.s_addr = *(__be32 *)local_ip; + sin = (struct sockaddr_in *)&child_ep->com.remote_addr; + sin->sin_family = PF_INET; + sin->sin_port = peer_port; + sin->sin_addr.s_addr = *(__be32 *)peer_ip; + } else { + struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *) + &child_ep->com.local_addr; + sin6->sin6_family = PF_INET6; + sin6->sin6_port = local_port; + memcpy(sin6->sin6_addr.s6_addr, local_ip, 16); + sin6 = (struct sockaddr_in6 *)&child_ep->com.remote_addr; + sin6->sin6_family = PF_INET6; + sin6->sin6_port = peer_port; + memcpy(sin6->sin6_addr.s6_addr, peer_ip, 16); + } c4iw_get_ep(&parent_ep->com); child_ep->parent_ep = parent_ep; child_ep->tos = GET_POPEN_TOS(ntohl(req->tos_stid)); - child_ep->l2t = l2t; child_ep->dst = dst; child_ep->hwtid = hwtid; - child_ep->tx_chan = tx_chan; - child_ep->smac_idx = smac_idx; - child_ep->rss_qid = rss_qid; - child_ep->mtu = mtu; - child_ep->txq_idx = txq_idx; - child_ep->ctrlq_idx = ctrlq_idx; PDBG("%s tx_chan %u smac_idx %u rss_qid %u\n", __func__, - tx_chan, smac_idx, rss_qid); + child_ep->tx_chan, child_ep->smac_idx, child_ep->rss_qid); init_timer(&child_ep->timer); cxgb4_insert_tid(t, child_ep, hwtid); - accept_cr(child_ep, peer_ip, skb, req); + insert_handle(dev, &dev->hwtid_idr, child_ep, child_ep->hwtid); + accept_cr(child_ep, skb, req); + set_bit(PASS_ACCEPT_REQ, &child_ep->com.history); goto out; reject: - reject_cr(dev, hwtid, peer_ip, skb); + reject_cr(dev, hwtid, skb); out: return 0; } @@ -1459,12 +2358,16 @@ static int pass_establish(struct c4iw_dev *dev, struct sk_buff *skb) ep->snd_seq = be32_to_cpu(req->snd_isn); ep->rcv_seq = be32_to_cpu(req->rcv_isn); + PDBG("%s ep %p hwtid %u tcp_opt 0x%02x\n", __func__, ep, tid, + ntohs(req->tcp_opt)); + set_emss(ep, ntohs(req->tcp_opt)); dst_confirm(ep->dst); state_set(&ep->com, MPA_REQ_WAIT); start_ep_timer(ep); send_flowc(ep, skb); + set_bit(PASS_ESTAB, &ep->com.history); return 0; } @@ -1476,14 +2379,15 @@ static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb) struct c4iw_qp_attributes attrs; int disconnect = 1; int release = 0; - int closing = 0; struct tid_info *t = dev->rdev.lldi.tids; unsigned int tid = GET_TID(hdr); + int ret; ep = lookup_tid(t, tid); PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); dst_confirm(ep->dst); + set_bit(PEER_CLOSE, &ep->com.history); mutex_lock(&ep->com.mutex); switch (ep->com.state) { case MPA_REQ_WAIT: @@ -1502,23 +2406,24 @@ static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb) * in rdma connection migration (see c4iw_accept_cr()). */ __state_set(&ep->com, CLOSING); - ep->com.wr_wait.done = 1; - ep->com.wr_wait.ret = -ECONNRESET; PDBG("waking up ep %p tid %u\n", ep, ep->hwtid); - wake_up(&ep->com.wr_wait.wait); + c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); break; case MPA_REP_SENT: __state_set(&ep->com, CLOSING); - ep->com.wr_wait.done = 1; - ep->com.wr_wait.ret = -ECONNRESET; PDBG("waking up ep %p tid %u\n", ep, ep->hwtid); - wake_up(&ep->com.wr_wait.wait); + c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); break; case FPDU_MODE: start_ep_timer(ep); __state_set(&ep->com, CLOSING); - closing = 1; - peer_close_upcall(ep); + attrs.next_state = C4IW_QP_STATE_CLOSING; + ret = c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp, + C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); + if (ret != -ECONNRESET) { + peer_close_upcall(ep); + disconnect = 1; + } break; case ABORTING: disconnect = 0; @@ -1528,13 +2433,13 @@ static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb) disconnect = 0; break; case MORIBUND: - stop_ep_timer(ep); + (void)stop_ep_timer(ep); if (ep->com.cm_id && ep->com.qp) { attrs.next_state = C4IW_QP_STATE_IDLE; c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp, C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); } - close_complete_upcall(ep); + close_complete_upcall(ep, 0); __state_set(&ep->com, DEAD); release = 1; disconnect = 0; @@ -1546,11 +2451,6 @@ static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb) BUG_ON(1); } mutex_unlock(&ep->com.mutex); - if (closing) { - attrs.next_state = C4IW_QP_STATE_CLOSING; - c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp, - C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); - } if (disconnect) c4iw_ep_disconnect(ep, 0, GFP_KERNEL); if (release) @@ -1558,15 +2458,6 @@ static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb) return 0; } -/* - * Returns whether an ABORT_REQ_RSS message is a negative advice. - */ -static int is_neg_adv_abort(unsigned int status) -{ - return status == CPL_ERR_RTX_NEG_ADVICE || - status == CPL_ERR_PERSIST_NEG_ADVICE; -} - static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) { struct cpl_abort_req_rss *req = cplhdr(skb); @@ -1580,31 +2471,47 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) unsigned int tid = GET_TID(req); ep = lookup_tid(t, tid); - if (is_neg_adv_abort(req->status)) { + if (is_neg_adv(req->status)) { PDBG("%s neg_adv_abort ep %p tid %u\n", __func__, ep, ep->hwtid); return 0; } PDBG("%s ep %p tid %u state %u\n", __func__, ep, ep->hwtid, ep->com.state); + set_bit(PEER_ABORT, &ep->com.history); /* * Wake up any threads in rdma_init() or rdma_fini(). + * However, this is not needed if com state is just + * MPA_REQ_SENT */ - ep->com.wr_wait.done = 1; - ep->com.wr_wait.ret = -ECONNRESET; - wake_up(&ep->com.wr_wait.wait); + if (ep->com.state != MPA_REQ_SENT) + c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); mutex_lock(&ep->com.mutex); switch (ep->com.state) { case CONNECTING: break; case MPA_REQ_WAIT: - stop_ep_timer(ep); + (void)stop_ep_timer(ep); break; case MPA_REQ_SENT: - stop_ep_timer(ep); - connect_reply_upcall(ep, -ECONNRESET); + (void)stop_ep_timer(ep); + if (mpa_rev == 1 || (mpa_rev == 2 && ep->tried_with_mpa_v1)) + connect_reply_upcall(ep, -ECONNRESET); + else { + /* + * we just don't send notification upwards because we + * want to retry with mpa_v1 without upper layers even + * knowing it. + * + * do some housekeeping so as to re-initiate the + * connection + */ + PDBG("%s: mpa_rev=%d. Retrying with mpav1\n", __func__, + mpa_rev); + ep->retry_with_mpa_v1 = 1; + } break; case MPA_REP_SENT: break; @@ -1640,7 +2547,9 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) dst_confirm(ep->dst); if (ep->com.state != ABORTING) { __state_set(&ep->com, DEAD); - release = 1; + /* we don't release if we want to retry with mpa_v1 */ + if (!ep->retry_with_mpa_v1) + release = 1; } mutex_unlock(&ep->com.mutex); @@ -1660,6 +2569,14 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb) out: if (release) release_ep_resources(ep); + else if (ep->retry_with_mpa_v1) { + remove_handle(ep->com.dev, &ep->com.dev->hwtid_idr, ep->hwtid); + cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, ep->hwtid); + dst_release(ep->dst); + cxgb4_l2t_release(ep->l2t); + c4iw_reconnect(ep); + } + return 0; } @@ -1684,7 +2601,7 @@ static int close_con_rpl(struct c4iw_dev *dev, struct sk_buff *skb) __state_set(&ep->com, MORIBUND); break; case MORIBUND: - stop_ep_timer(ep); + (void)stop_ep_timer(ep); if ((ep->com.cm_id) && (ep->com.qp)) { attrs.next_state = C4IW_QP_STATE_IDLE; c4iw_modify_qp(ep->com.qp->rhp, @@ -1692,7 +2609,7 @@ static int close_con_rpl(struct c4iw_dev *dev, struct sk_buff *skb) C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); } - close_complete_upcall(ep); + close_complete_upcall(ep, 0); __state_set(&ep->com, DEAD); release = 1; break; @@ -1720,14 +2637,14 @@ static int terminate(struct c4iw_dev *dev, struct sk_buff *skb) ep = lookup_tid(t, tid); BUG_ON(!ep); - if (ep->com.qp) { + if (ep && ep->com.qp) { printk(KERN_WARNING MOD "TERM received tid %u qpid %u\n", tid, ep->com.qp->wq.sq.qid); attrs.next_state = C4IW_QP_STATE_TERMINATE; c4iw_modify_qp(ep->com.qp->rhp, ep->com.qp, C4IW_QP_ATTR_NEXT_STATE, &attrs, 1); } else - printk(KERN_WARNING MOD "TERM received tid %u no qp\n", tid); + printk(KERN_WARNING MOD "TERM received tid %u no ep/qp\n", tid); return 0; } @@ -1767,21 +2684,28 @@ static int fw4_ack(struct c4iw_dev *dev, struct sk_buff *skb) int c4iw_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len) { - int err; + int err = 0; + int disconnect = 0; struct c4iw_ep *ep = to_ep(cm_id); PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); - if (state_read(&ep->com) == DEAD) { + mutex_lock(&ep->com.mutex); + if (ep->com.state == DEAD) { + mutex_unlock(&ep->com.mutex); c4iw_put_ep(&ep->com); return -ECONNRESET; } - BUG_ON(state_read(&ep->com) != MPA_REQ_RCVD); + set_bit(ULP_REJECT, &ep->com.history); + BUG_ON(ep->com.state != MPA_REQ_RCVD); if (mpa_rev == 0) abort_connection(ep, NULL, GFP_KERNEL); else { err = send_mpa_reject(ep, pdata, pdata_len); - err = c4iw_ep_disconnect(ep, 0, GFP_KERNEL); + disconnect = 1; } + mutex_unlock(&ep->com.mutex); + if (disconnect) + err = c4iw_ep_disconnect(ep, 0, GFP_KERNEL); c4iw_put_ep(&ep->com); return 0; } @@ -1796,14 +2720,17 @@ int c4iw_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) struct c4iw_qp *qp = get_qhp(h, conn_param->qpn); PDBG("%s ep %p tid %u\n", __func__, ep, ep->hwtid); - if (state_read(&ep->com) == DEAD) { + + mutex_lock(&ep->com.mutex); + if (ep->com.state == DEAD) { err = -ECONNRESET; goto err; } - BUG_ON(state_read(&ep->com) != MPA_REQ_RCVD); + BUG_ON(ep->com.state != MPA_REQ_RCVD); BUG_ON(!qp); + set_bit(ULP_ACCEPT, &ep->com.history); if ((conn_param->ord > c4iw_max_read_depth) || (conn_param->ird > c4iw_max_read_depth)) { abort_connection(ep, NULL, GFP_KERNEL); @@ -1811,18 +2738,41 @@ int c4iw_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) goto err; } - cm_id->add_ref(cm_id); - ep->com.cm_id = cm_id; - ep->com.qp = qp; + if (ep->mpa_attr.version == 2 && ep->mpa_attr.enhanced_rdma_conn) { + if (conn_param->ord > ep->ird) { + ep->ird = conn_param->ird; + ep->ord = conn_param->ord; + send_mpa_reject(ep, conn_param->private_data, + conn_param->private_data_len); + abort_connection(ep, NULL, GFP_KERNEL); + err = -ENOMEM; + goto err; + } + if (conn_param->ird > ep->ord) { + if (!ep->ord) + conn_param->ird = 1; + else { + abort_connection(ep, NULL, GFP_KERNEL); + err = -ENOMEM; + goto err; + } + } + } ep->ird = conn_param->ird; ep->ord = conn_param->ord; - if (peer2peer && ep->ird == 0) - ep->ird = 1; + if (ep->mpa_attr.version != 2) + if (peer2peer && ep->ird == 0) + ep->ird = 1; PDBG("%s %d ird %d ord %d\n", __func__, __LINE__, ep->ird, ep->ord); + cm_id->add_ref(cm_id); + ep->com.cm_id = cm_id; + ep->com.qp = qp; + ref_qp(ep); + /* bind QP to EP and move to RTS */ attrs.mpa_attr = ep->mpa_attr; attrs.max_ird = ep->ird; @@ -1846,27 +2796,95 @@ int c4iw_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (err) goto err1; - state_set(&ep->com, FPDU_MODE); + __state_set(&ep->com, FPDU_MODE); established_upcall(ep); + mutex_unlock(&ep->com.mutex); c4iw_put_ep(&ep->com); return 0; err1: ep->com.cm_id = NULL; - ep->com.qp = NULL; cm_id->rem_ref(cm_id); err: + mutex_unlock(&ep->com.mutex); c4iw_put_ep(&ep->com); return err; } +static int pick_local_ipaddrs(struct c4iw_dev *dev, struct iw_cm_id *cm_id) +{ + struct in_device *ind; + int found = 0; + struct sockaddr_in *laddr = (struct sockaddr_in *)&cm_id->local_addr; + struct sockaddr_in *raddr = (struct sockaddr_in *)&cm_id->remote_addr; + + ind = in_dev_get(dev->rdev.lldi.ports[0]); + if (!ind) + return -EADDRNOTAVAIL; + for_primary_ifa(ind) { + laddr->sin_addr.s_addr = ifa->ifa_address; + raddr->sin_addr.s_addr = ifa->ifa_address; + found = 1; + break; + } + endfor_ifa(ind); + in_dev_put(ind); + return found ? 0 : -EADDRNOTAVAIL; +} + +static int get_lladdr(struct net_device *dev, struct in6_addr *addr, + unsigned char banned_flags) +{ + struct inet6_dev *idev; + int err = -EADDRNOTAVAIL; + + rcu_read_lock(); + idev = __in6_dev_get(dev); + if (idev != NULL) { + struct inet6_ifaddr *ifp; + + read_lock_bh(&idev->lock); + list_for_each_entry(ifp, &idev->addr_list, if_list) { + if (ifp->scope == IFA_LINK && + !(ifp->flags & banned_flags)) { + memcpy(addr, &ifp->addr, 16); + err = 0; + break; + } + } + read_unlock_bh(&idev->lock); + } + rcu_read_unlock(); + return err; +} + +static int pick_local_ip6addrs(struct c4iw_dev *dev, struct iw_cm_id *cm_id) +{ + struct in6_addr uninitialized_var(addr); + struct sockaddr_in6 *la6 = (struct sockaddr_in6 *)&cm_id->local_addr; + struct sockaddr_in6 *ra6 = (struct sockaddr_in6 *)&cm_id->remote_addr; + + if (get_lladdr(dev->rdev.lldi.ports[0], &addr, IFA_F_TENTATIVE)) { + memcpy(la6->sin6_addr.s6_addr, &addr, 16); + memcpy(ra6->sin6_addr.s6_addr, &addr, 16); + return 0; + } + return -EADDRNOTAVAIL; +} + int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) { - int err = 0; struct c4iw_dev *dev = to_c4iw_dev(cm_id->device); struct c4iw_ep *ep; - struct rtable *rt; - struct net_device *pdev; - int step; + int err = 0; + struct sockaddr_in *laddr; + struct sockaddr_in *raddr; + struct sockaddr_in6 *laddr6; + struct sockaddr_in6 *raddr6; + struct iwpm_dev_data pm_reg_msg; + struct iwpm_sa_data pm_msg; + __u8 *ra; + int iptype; + int iwpm_err = 0; if ((conn_param->ord > c4iw_max_read_depth) || (conn_param->ird > c4iw_max_read_depth)) { @@ -1894,7 +2912,12 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) ep->com.dev = dev; ep->com.cm_id = cm_id; ep->com.qp = get_qhp(dev, conn_param->qpn); - BUG_ON(!ep->com.qp); + if (!ep->com.qp) { + PDBG("%s qpn 0x%x not found!\n", __func__, conn_param->qpn); + err = -EINVAL; + goto fail1; + } + ref_qp(ep); PDBG("%s qpn 0x%x qp %p cm_id %p\n", __func__, conn_param->qpn, ep->com.qp, cm_id); @@ -1905,69 +2928,103 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) if (ep->atid == -1) { printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__); err = -ENOMEM; - goto fail2; + goto fail1; + } + insert_handle(dev, &dev->atid_idr, ep, ep->atid); + + memcpy(&ep->com.local_addr, &cm_id->local_addr, + sizeof(ep->com.local_addr)); + memcpy(&ep->com.remote_addr, &cm_id->remote_addr, + sizeof(ep->com.remote_addr)); + + /* No port mapper available, go with the specified peer information */ + memcpy(&ep->com.mapped_local_addr, &cm_id->local_addr, + sizeof(ep->com.mapped_local_addr)); + memcpy(&ep->com.mapped_remote_addr, &cm_id->remote_addr, + sizeof(ep->com.mapped_remote_addr)); + + c4iw_form_reg_msg(dev, &pm_reg_msg); + iwpm_err = iwpm_register_pid(&pm_reg_msg, RDMA_NL_C4IW); + if (iwpm_err) { + PDBG("%s: Port Mapper reg pid fail (err = %d).\n", + __func__, iwpm_err); + } + if (iwpm_valid_pid() && !iwpm_err) { + c4iw_form_pm_msg(ep, &pm_msg); + iwpm_err = iwpm_add_and_query_mapping(&pm_msg, RDMA_NL_C4IW); + if (iwpm_err) + PDBG("%s: Port Mapper query fail (err = %d).\n", + __func__, iwpm_err); + else + c4iw_record_pm_msg(ep, &pm_msg); } + if (iwpm_create_mapinfo(&ep->com.local_addr, + &ep->com.mapped_local_addr, RDMA_NL_C4IW)) { + iwpm_remove_mapping(&ep->com.local_addr, RDMA_NL_C4IW); + err = -ENOMEM; + goto fail1; + } + print_addr(&ep->com, __func__, "add_query/create_mapinfo"); + set_bit(RELEASE_MAPINFO, &ep->com.flags); - PDBG("%s saddr 0x%x sport 0x%x raddr 0x%x rport 0x%x\n", __func__, - ntohl(cm_id->local_addr.sin_addr.s_addr), - ntohs(cm_id->local_addr.sin_port), - ntohl(cm_id->remote_addr.sin_addr.s_addr), - ntohs(cm_id->remote_addr.sin_port)); + laddr = (struct sockaddr_in *)&ep->com.mapped_local_addr; + raddr = (struct sockaddr_in *)&ep->com.mapped_remote_addr; + laddr6 = (struct sockaddr_in6 *)&ep->com.mapped_local_addr; + raddr6 = (struct sockaddr_in6 *) &ep->com.mapped_remote_addr; - /* find a route */ - rt = find_route(dev, - cm_id->local_addr.sin_addr.s_addr, - cm_id->remote_addr.sin_addr.s_addr, - cm_id->local_addr.sin_port, - cm_id->remote_addr.sin_port, 0); - if (!rt) { + if (cm_id->remote_addr.ss_family == AF_INET) { + iptype = 4; + ra = (__u8 *)&raddr->sin_addr; + + /* + * Handle loopback requests to INADDR_ANY. + */ + if ((__force int)raddr->sin_addr.s_addr == INADDR_ANY) { + err = pick_local_ipaddrs(dev, cm_id); + if (err) + goto fail1; + } + + /* find a route */ + PDBG("%s saddr %pI4 sport 0x%x raddr %pI4 rport 0x%x\n", + __func__, &laddr->sin_addr, ntohs(laddr->sin_port), + ra, ntohs(raddr->sin_port)); + ep->dst = find_route(dev, laddr->sin_addr.s_addr, + raddr->sin_addr.s_addr, laddr->sin_port, + raddr->sin_port, 0); + } else { + iptype = 6; + ra = (__u8 *)&raddr6->sin6_addr; + + /* + * Handle loopback requests to INADDR_ANY. + */ + if (ipv6_addr_type(&raddr6->sin6_addr) == IPV6_ADDR_ANY) { + err = pick_local_ip6addrs(dev, cm_id); + if (err) + goto fail1; + } + + /* find a route */ + PDBG("%s saddr %pI6 sport 0x%x raddr %pI6 rport 0x%x\n", + __func__, laddr6->sin6_addr.s6_addr, + ntohs(laddr6->sin6_port), + raddr6->sin6_addr.s6_addr, ntohs(raddr6->sin6_port)); + ep->dst = find_route6(dev, laddr6->sin6_addr.s6_addr, + raddr6->sin6_addr.s6_addr, + laddr6->sin6_port, raddr6->sin6_port, 0, + raddr6->sin6_scope_id); + } + if (!ep->dst) { printk(KERN_ERR MOD "%s - cannot find route.\n", __func__); err = -EHOSTUNREACH; - goto fail3; + goto fail2; } - ep->dst = &rt->dst; - /* get a l2t entry */ - if (ep->dst->neighbour->dev->flags & IFF_LOOPBACK) { - PDBG("%s LOOPBACK\n", __func__); - pdev = ip_dev_find(&init_net, - cm_id->remote_addr.sin_addr.s_addr); - ep->l2t = cxgb4_l2t_get(ep->com.dev->rdev.lldi.l2t, - ep->dst->neighbour, - pdev, 0); - ep->mtu = pdev->mtu; - ep->tx_chan = cxgb4_port_chan(pdev); - ep->smac_idx = (cxgb4_port_viid(pdev) & 0x7F) << 1; - step = ep->com.dev->rdev.lldi.ntxq / - ep->com.dev->rdev.lldi.nchan; - ep->txq_idx = cxgb4_port_idx(pdev) * step; - step = ep->com.dev->rdev.lldi.nrxq / - ep->com.dev->rdev.lldi.nchan; - ep->ctrlq_idx = cxgb4_port_idx(pdev); - ep->rss_qid = ep->com.dev->rdev.lldi.rxq_ids[ - cxgb4_port_idx(pdev) * step]; - dev_put(pdev); - } else { - ep->l2t = cxgb4_l2t_get(ep->com.dev->rdev.lldi.l2t, - ep->dst->neighbour, - ep->dst->neighbour->dev, 0); - ep->mtu = dst_mtu(ep->dst); - ep->tx_chan = cxgb4_port_chan(ep->dst->neighbour->dev); - ep->smac_idx = (cxgb4_port_viid(ep->dst->neighbour->dev) & - 0x7F) << 1; - step = ep->com.dev->rdev.lldi.ntxq / - ep->com.dev->rdev.lldi.nchan; - ep->txq_idx = cxgb4_port_idx(ep->dst->neighbour->dev) * step; - ep->ctrlq_idx = cxgb4_port_idx(ep->dst->neighbour->dev); - step = ep->com.dev->rdev.lldi.nrxq / - ep->com.dev->rdev.lldi.nchan; - ep->rss_qid = ep->com.dev->rdev.lldi.rxq_ids[ - cxgb4_port_idx(ep->dst->neighbour->dev) * step]; - } - if (!ep->l2t) { + err = import_ep(ep, iptype, ra, ep->dst, ep->com.dev, true); + if (err) { printk(KERN_ERR MOD "%s - cannot alloc l2e.\n", __func__); - err = -ENOMEM; - goto fail4; + goto fail3; } PDBG("%s txq_idx %u tx_chan %u smac_idx %u rss_qid %u l2t_idx %u\n", @@ -1976,8 +3033,6 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) state_set(&ep->com, CONNECTING); ep->tos = 0; - ep->com.local_addr = cm_id->local_addr; - ep->com.remote_addr = cm_id->remote_addr; /* send connect request to rnic */ err = send_connect(ep); @@ -1985,23 +3040,82 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param) goto out; cxgb4_l2t_release(ep->l2t); -fail4: - dst_release(ep->dst); fail3: - cxgb4_free_atid(ep->com.dev->rdev.lldi.tids, ep->atid); + dst_release(ep->dst); fail2: + remove_handle(ep->com.dev, &ep->com.dev->atid_idr, ep->atid); + cxgb4_free_atid(ep->com.dev->rdev.lldi.tids, ep->atid); +fail1: cm_id->rem_ref(cm_id); c4iw_put_ep(&ep->com); out: return err; } +static int create_server6(struct c4iw_dev *dev, struct c4iw_listen_ep *ep) +{ + int err; + struct sockaddr_in6 *sin6 = (struct sockaddr_in6 *) + &ep->com.mapped_local_addr; + + c4iw_init_wr_wait(&ep->com.wr_wait); + err = cxgb4_create_server6(ep->com.dev->rdev.lldi.ports[0], + ep->stid, &sin6->sin6_addr, + sin6->sin6_port, + ep->com.dev->rdev.lldi.rxq_ids[0]); + if (!err) + err = c4iw_wait_for_reply(&ep->com.dev->rdev, + &ep->com.wr_wait, + 0, 0, __func__); + if (err) + pr_err("cxgb4_create_server6/filter failed err %d stid %d laddr %pI6 lport %d\n", + err, ep->stid, + sin6->sin6_addr.s6_addr, ntohs(sin6->sin6_port)); + return err; +} + +static int create_server4(struct c4iw_dev *dev, struct c4iw_listen_ep *ep) +{ + int err; + struct sockaddr_in *sin = (struct sockaddr_in *) + &ep->com.mapped_local_addr; + + if (dev->rdev.lldi.enable_fw_ofld_conn) { + do { + err = cxgb4_create_server_filter( + ep->com.dev->rdev.lldi.ports[0], ep->stid, + sin->sin_addr.s_addr, sin->sin_port, 0, + ep->com.dev->rdev.lldi.rxq_ids[0], 0, 0); + if (err == -EBUSY) { + set_current_state(TASK_UNINTERRUPTIBLE); + schedule_timeout(usecs_to_jiffies(100)); + } + } while (err == -EBUSY); + } else { + c4iw_init_wr_wait(&ep->com.wr_wait); + err = cxgb4_create_server(ep->com.dev->rdev.lldi.ports[0], + ep->stid, sin->sin_addr.s_addr, sin->sin_port, + 0, ep->com.dev->rdev.lldi.rxq_ids[0]); + if (!err) + err = c4iw_wait_for_reply(&ep->com.dev->rdev, + &ep->com.wr_wait, + 0, 0, __func__); + } + if (err) + pr_err("cxgb4_create_server/filter failed err %d stid %d laddr %pI4 lport %d\n" + , err, ep->stid, + &sin->sin_addr, ntohs(sin->sin_port)); + return err; +} + int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog) { int err = 0; struct c4iw_dev *dev = to_c4iw_dev(cm_id->device); struct c4iw_listen_ep *ep; - + struct iwpm_dev_data pm_reg_msg; + struct iwpm_sa_data pm_msg; + int iwpm_err = 0; might_sleep(); @@ -2016,36 +3130,70 @@ int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog) ep->com.cm_id = cm_id; ep->com.dev = dev; ep->backlog = backlog; - ep->com.local_addr = cm_id->local_addr; + memcpy(&ep->com.local_addr, &cm_id->local_addr, + sizeof(ep->com.local_addr)); /* * Allocate a server TID. */ - ep->stid = cxgb4_alloc_stid(dev->rdev.lldi.tids, PF_INET, ep); + if (dev->rdev.lldi.enable_fw_ofld_conn && + ep->com.local_addr.ss_family == AF_INET) + ep->stid = cxgb4_alloc_sftid(dev->rdev.lldi.tids, + cm_id->local_addr.ss_family, ep); + else + ep->stid = cxgb4_alloc_stid(dev->rdev.lldi.tids, + cm_id->local_addr.ss_family, ep); + if (ep->stid == -1) { printk(KERN_ERR MOD "%s - cannot alloc stid.\n", __func__); err = -ENOMEM; goto fail2; } - - state_set(&ep->com, LISTEN); - c4iw_init_wr_wait(&ep->com.wr_wait); - err = cxgb4_create_server(ep->com.dev->rdev.lldi.ports[0], ep->stid, - ep->com.local_addr.sin_addr.s_addr, - ep->com.local_addr.sin_port, - ep->com.dev->rdev.lldi.rxq_ids[0]); - if (err) + insert_handle(dev, &dev->stid_idr, ep, ep->stid); + + /* No port mapper available, go with the specified info */ + memcpy(&ep->com.mapped_local_addr, &cm_id->local_addr, + sizeof(ep->com.mapped_local_addr)); + + c4iw_form_reg_msg(dev, &pm_reg_msg); + iwpm_err = iwpm_register_pid(&pm_reg_msg, RDMA_NL_C4IW); + if (iwpm_err) { + PDBG("%s: Port Mapper reg pid fail (err = %d).\n", + __func__, iwpm_err); + } + if (iwpm_valid_pid() && !iwpm_err) { + memcpy(&pm_msg.loc_addr, &ep->com.local_addr, + sizeof(ep->com.local_addr)); + iwpm_err = iwpm_add_mapping(&pm_msg, RDMA_NL_C4IW); + if (iwpm_err) + PDBG("%s: Port Mapper query fail (err = %d).\n", + __func__, iwpm_err); + else + memcpy(&ep->com.mapped_local_addr, + &pm_msg.mapped_loc_addr, + sizeof(ep->com.mapped_local_addr)); + } + if (iwpm_create_mapinfo(&ep->com.local_addr, + &ep->com.mapped_local_addr, RDMA_NL_C4IW)) { + err = -ENOMEM; goto fail3; + } + print_addr(&ep->com, __func__, "add_mapping/create_mapinfo"); - /* wait for pass_open_rpl */ - err = c4iw_wait_for_reply(&ep->com.dev->rdev, &ep->com.wr_wait, 0, 0, - __func__); + set_bit(RELEASE_MAPINFO, &ep->com.flags); + state_set(&ep->com, LISTEN); + if (ep->com.local_addr.ss_family == AF_INET) + err = create_server4(dev, ep); + else + err = create_server6(dev, ep); if (!err) { cm_id->provider_data = ep; goto out; } + fail3: - cxgb4_free_stid(ep->com.dev->rdev.lldi.tids, ep->stid, PF_INET); + cxgb4_free_stid(ep->com.dev->rdev.lldi.tids, ep->stid, + ep->com.local_addr.ss_family); fail2: cm_id->rem_ref(cm_id); c4iw_put_ep(&ep->com); @@ -2063,13 +3211,24 @@ int c4iw_destroy_listen(struct iw_cm_id *cm_id) might_sleep(); state_set(&ep->com, DEAD); - c4iw_init_wr_wait(&ep->com.wr_wait); - err = listen_stop(ep); - if (err) - goto done; - err = c4iw_wait_for_reply(&ep->com.dev->rdev, &ep->com.wr_wait, 0, 0, - __func__); - cxgb4_free_stid(ep->com.dev->rdev.lldi.tids, ep->stid, PF_INET); + if (ep->com.dev->rdev.lldi.enable_fw_ofld_conn && + ep->com.local_addr.ss_family == AF_INET) { + err = cxgb4_remove_server_filter( + ep->com.dev->rdev.lldi.ports[0], ep->stid, + ep->com.dev->rdev.lldi.rxq_ids[0], 0); + } else { + c4iw_init_wr_wait(&ep->com.wr_wait); + err = cxgb4_remove_server( + ep->com.dev->rdev.lldi.ports[0], ep->stid, + ep->com.dev->rdev.lldi.rxq_ids[0], 0); + if (err) + goto done; + err = c4iw_wait_for_reply(&ep->com.dev->rdev, &ep->com.wr_wait, + 0, 0, __func__); + } + remove_handle(ep->com.dev, &ep->com.dev->stid_idr, ep->stid); + cxgb4_free_stid(ep->com.dev->rdev.lldi.tids, ep->stid, + ep->com.local_addr.ss_family); done: cm_id->rem_ref(cm_id); c4iw_put_ep(&ep->com); @@ -2091,7 +3250,7 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) rdev = &ep->com.dev->rdev; if (c4iw_fatal_error(rdev)) { fatal = 1; - close_complete_upcall(ep); + close_complete_upcall(ep, -EIO); ep->com.state = DEAD; } switch (ep->com.state) { @@ -2113,7 +3272,7 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) if (!test_and_set_bit(CLOSE_SENT, &ep->com.flags)) { close = 1; if (abrupt) { - stop_ep_timer(ep); + (void)stop_ep_timer(ep); ep->com.state = ABORTING; } else ep->com.state = MORIBUND; @@ -2130,24 +3289,369 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp) break; } - mutex_unlock(&ep->com.mutex); if (close) { - if (abrupt) - ret = abort_connection(ep, NULL, gfp); - else + if (abrupt) { + set_bit(EP_DISC_ABORT, &ep->com.history); + close_complete_upcall(ep, -ECONNRESET); + ret = send_abort(ep, NULL, gfp); + } else { + set_bit(EP_DISC_CLOSE, &ep->com.history); ret = send_halfclose(ep, gfp); + } if (ret) fatal = 1; } + mutex_unlock(&ep->com.mutex); if (fatal) release_ep_resources(ep); return ret; } -static int async_event(struct c4iw_dev *dev, struct sk_buff *skb) +static void active_ofld_conn_reply(struct c4iw_dev *dev, struct sk_buff *skb, + struct cpl_fw6_msg_ofld_connection_wr_rpl *req) +{ + struct c4iw_ep *ep; + int atid = be32_to_cpu(req->tid); + + ep = (struct c4iw_ep *)lookup_atid(dev->rdev.lldi.tids, + (__force u32) req->tid); + if (!ep) + return; + + switch (req->retval) { + case FW_ENOMEM: + set_bit(ACT_RETRY_NOMEM, &ep->com.history); + if (ep->retry_count++ < ACT_OPEN_RETRY_COUNT) { + send_fw_act_open_req(ep, atid); + return; + } + case FW_EADDRINUSE: + set_bit(ACT_RETRY_INUSE, &ep->com.history); + if (ep->retry_count++ < ACT_OPEN_RETRY_COUNT) { + send_fw_act_open_req(ep, atid); + return; + } + break; + default: + pr_info("%s unexpected ofld conn wr retval %d\n", + __func__, req->retval); + break; + } + pr_err("active ofld_connect_wr failure %d atid %d\n", + req->retval, atid); + mutex_lock(&dev->rdev.stats.lock); + dev->rdev.stats.act_ofld_conn_fails++; + mutex_unlock(&dev->rdev.stats.lock); + connect_reply_upcall(ep, status2errno(req->retval)); + state_set(&ep->com, DEAD); + remove_handle(dev, &dev->atid_idr, atid); + cxgb4_free_atid(dev->rdev.lldi.tids, atid); + dst_release(ep->dst); + cxgb4_l2t_release(ep->l2t); + c4iw_put_ep(&ep->com); +} + +static void passive_ofld_conn_reply(struct c4iw_dev *dev, struct sk_buff *skb, + struct cpl_fw6_msg_ofld_connection_wr_rpl *req) +{ + struct sk_buff *rpl_skb; + struct cpl_pass_accept_req *cpl; + int ret; + + rpl_skb = (struct sk_buff *)(unsigned long)req->cookie; + BUG_ON(!rpl_skb); + if (req->retval) { + PDBG("%s passive open failure %d\n", __func__, req->retval); + mutex_lock(&dev->rdev.stats.lock); + dev->rdev.stats.pas_ofld_conn_fails++; + mutex_unlock(&dev->rdev.stats.lock); + kfree_skb(rpl_skb); + } else { + cpl = (struct cpl_pass_accept_req *)cplhdr(rpl_skb); + OPCODE_TID(cpl) = htonl(MK_OPCODE_TID(CPL_PASS_ACCEPT_REQ, + (__force u32) htonl( + (__force u32) req->tid))); + ret = pass_accept_req(dev, rpl_skb); + if (!ret) + kfree_skb(rpl_skb); + } + return; +} + +static int deferred_fw6_msg(struct c4iw_dev *dev, struct sk_buff *skb) { struct cpl_fw6_msg *rpl = cplhdr(skb); - c4iw_ev_dispatch(dev, (struct t4_cqe *)&rpl->data[0]); + struct cpl_fw6_msg_ofld_connection_wr_rpl *req; + + switch (rpl->type) { + case FW6_TYPE_CQE: + c4iw_ev_dispatch(dev, (struct t4_cqe *)&rpl->data[0]); + break; + case FW6_TYPE_OFLD_CONNECTION_WR_RPL: + req = (struct cpl_fw6_msg_ofld_connection_wr_rpl *)rpl->data; + switch (req->t_state) { + case TCP_SYN_SENT: + active_ofld_conn_reply(dev, skb, req); + break; + case TCP_SYN_RECV: + passive_ofld_conn_reply(dev, skb, req); + break; + default: + pr_err("%s unexpected ofld conn wr state %d\n", + __func__, req->t_state); + break; + } + break; + } + return 0; +} + +static void build_cpl_pass_accept_req(struct sk_buff *skb, int stid , u8 tos) +{ + u32 l2info; + u16 vlantag, len, hdr_len, eth_hdr_len; + u8 intf; + struct cpl_rx_pkt *cpl = cplhdr(skb); + struct cpl_pass_accept_req *req; + struct tcp_options_received tmp_opt; + struct c4iw_dev *dev; + + dev = *((struct c4iw_dev **) (skb->cb + sizeof(void *))); + /* Store values from cpl_rx_pkt in temporary location. */ + vlantag = (__force u16) cpl->vlan; + len = (__force u16) cpl->len; + l2info = (__force u32) cpl->l2info; + hdr_len = (__force u16) cpl->hdr_len; + intf = cpl->iff; + + __skb_pull(skb, sizeof(*req) + sizeof(struct rss_header)); + + /* + * We need to parse the TCP options from SYN packet. + * to generate cpl_pass_accept_req. + */ + memset(&tmp_opt, 0, sizeof(tmp_opt)); + tcp_clear_options(&tmp_opt); + tcp_parse_options(skb, &tmp_opt, 0, NULL); + + req = (struct cpl_pass_accept_req *)__skb_push(skb, sizeof(*req)); + memset(req, 0, sizeof(*req)); + req->l2info = cpu_to_be16(V_SYN_INTF(intf) | + V_SYN_MAC_IDX(G_RX_MACIDX( + (__force int) htonl(l2info))) | + F_SYN_XACT_MATCH); + eth_hdr_len = is_t4(dev->rdev.lldi.adapter_type) ? + G_RX_ETHHDR_LEN((__force int) htonl(l2info)) : + G_RX_T5_ETHHDR_LEN((__force int) htonl(l2info)); + req->hdr_len = cpu_to_be32(V_SYN_RX_CHAN(G_RX_CHAN( + (__force int) htonl(l2info))) | + V_TCP_HDR_LEN(G_RX_TCPHDR_LEN( + (__force int) htons(hdr_len))) | + V_IP_HDR_LEN(G_RX_IPHDR_LEN( + (__force int) htons(hdr_len))) | + V_ETH_HDR_LEN(G_RX_ETHHDR_LEN(eth_hdr_len))); + req->vlan = (__force __be16) vlantag; + req->len = (__force __be16) len; + req->tos_stid = cpu_to_be32(PASS_OPEN_TID(stid) | + PASS_OPEN_TOS(tos)); + req->tcpopt.mss = htons(tmp_opt.mss_clamp); + if (tmp_opt.wscale_ok) + req->tcpopt.wsf = tmp_opt.snd_wscale; + req->tcpopt.tstamp = tmp_opt.saw_tstamp; + if (tmp_opt.sack_ok) + req->tcpopt.sack = 1; + OPCODE_TID(req) = htonl(MK_OPCODE_TID(CPL_PASS_ACCEPT_REQ, 0)); + return; +} + +static void send_fw_pass_open_req(struct c4iw_dev *dev, struct sk_buff *skb, + __be32 laddr, __be16 lport, + __be32 raddr, __be16 rport, + u32 rcv_isn, u32 filter, u16 window, + u32 rss_qid, u8 port_id) +{ + struct sk_buff *req_skb; + struct fw_ofld_connection_wr *req; + struct cpl_pass_accept_req *cpl = cplhdr(skb); + int ret; + + req_skb = alloc_skb(sizeof(struct fw_ofld_connection_wr), GFP_KERNEL); + req = (struct fw_ofld_connection_wr *)__skb_put(req_skb, sizeof(*req)); + memset(req, 0, sizeof(*req)); + req->op_compl = htonl(V_WR_OP(FW_OFLD_CONNECTION_WR) | FW_WR_COMPL(1)); + req->len16_pkd = htonl(FW_WR_LEN16(DIV_ROUND_UP(sizeof(*req), 16))); + req->le.version_cpl = htonl(F_FW_OFLD_CONNECTION_WR_CPL); + req->le.filter = (__force __be32) filter; + req->le.lport = lport; + req->le.pport = rport; + req->le.u.ipv4.lip = laddr; + req->le.u.ipv4.pip = raddr; + req->tcb.rcv_nxt = htonl(rcv_isn + 1); + req->tcb.rcv_adv = htons(window); + req->tcb.t_state_to_astid = + htonl(V_FW_OFLD_CONNECTION_WR_T_STATE(TCP_SYN_RECV) | + V_FW_OFLD_CONNECTION_WR_RCV_SCALE(cpl->tcpopt.wsf) | + V_FW_OFLD_CONNECTION_WR_ASTID( + GET_PASS_OPEN_TID(ntohl(cpl->tos_stid)))); + + /* + * We store the qid in opt2 which will be used by the firmware + * to send us the wr response. + */ + req->tcb.opt2 = htonl(V_RSS_QUEUE(rss_qid)); + + /* + * We initialize the MSS index in TCB to 0xF. + * So that when driver sends cpl_pass_accept_rpl + * TCB picks up the correct value. If this was 0 + * TP will ignore any value > 0 for MSS index. + */ + req->tcb.opt0 = cpu_to_be64(V_MSS_IDX(0xF)); + req->cookie = (unsigned long)skb; + + set_wr_txq(req_skb, CPL_PRIORITY_CONTROL, port_id); + ret = cxgb4_ofld_send(dev->rdev.lldi.ports[0], req_skb); + if (ret < 0) { + pr_err("%s - cxgb4_ofld_send error %d - dropping\n", __func__, + ret); + kfree_skb(skb); + kfree_skb(req_skb); + } +} + +/* + * Handler for CPL_RX_PKT message. Need to handle cpl_rx_pkt + * messages when a filter is being used instead of server to + * redirect a syn packet. When packets hit filter they are redirected + * to the offload queue and driver tries to establish the connection + * using firmware work request. + */ +static int rx_pkt(struct c4iw_dev *dev, struct sk_buff *skb) +{ + int stid; + unsigned int filter; + struct ethhdr *eh = NULL; + struct vlan_ethhdr *vlan_eh = NULL; + struct iphdr *iph; + struct tcphdr *tcph; + struct rss_header *rss = (void *)skb->data; + struct cpl_rx_pkt *cpl = (void *)skb->data; + struct cpl_pass_accept_req *req = (void *)(rss + 1); + struct l2t_entry *e; + struct dst_entry *dst; + struct c4iw_ep *lep; + u16 window; + struct port_info *pi; + struct net_device *pdev; + u16 rss_qid, eth_hdr_len; + int step; + u32 tx_chan; + struct neighbour *neigh; + + /* Drop all non-SYN packets */ + if (!(cpl->l2info & cpu_to_be32(F_RXF_SYN))) + goto reject; + + /* + * Drop all packets which did not hit the filter. + * Unlikely to happen. + */ + if (!(rss->filter_hit && rss->filter_tid)) + goto reject; + + /* + * Calculate the server tid from filter hit index from cpl_rx_pkt. + */ + stid = (__force int) cpu_to_be32((__force u32) rss->hash_val); + + lep = (struct c4iw_ep *)lookup_stid(dev->rdev.lldi.tids, stid); + if (!lep) { + PDBG("%s connect request on invalid stid %d\n", __func__, stid); + goto reject; + } + + eth_hdr_len = is_t4(dev->rdev.lldi.adapter_type) ? + G_RX_ETHHDR_LEN(htonl(cpl->l2info)) : + G_RX_T5_ETHHDR_LEN(htonl(cpl->l2info)); + if (eth_hdr_len == ETH_HLEN) { + eh = (struct ethhdr *)(req + 1); + iph = (struct iphdr *)(eh + 1); + } else { + vlan_eh = (struct vlan_ethhdr *)(req + 1); + iph = (struct iphdr *)(vlan_eh + 1); + skb->vlan_tci = ntohs(cpl->vlan); + } + + if (iph->version != 0x4) + goto reject; + + tcph = (struct tcphdr *)(iph + 1); + skb_set_network_header(skb, (void *)iph - (void *)rss); + skb_set_transport_header(skb, (void *)tcph - (void *)rss); + skb_get(skb); + + PDBG("%s lip 0x%x lport %u pip 0x%x pport %u tos %d\n", __func__, + ntohl(iph->daddr), ntohs(tcph->dest), ntohl(iph->saddr), + ntohs(tcph->source), iph->tos); + + dst = find_route(dev, iph->daddr, iph->saddr, tcph->dest, tcph->source, + iph->tos); + if (!dst) { + pr_err("%s - failed to find dst entry!\n", + __func__); + goto reject; + } + neigh = dst_neigh_lookup_skb(dst, skb); + + if (!neigh) { + pr_err("%s - failed to allocate neigh!\n", + __func__); + goto free_dst; + } + + if (neigh->dev->flags & IFF_LOOPBACK) { + pdev = ip_dev_find(&init_net, iph->daddr); + e = cxgb4_l2t_get(dev->rdev.lldi.l2t, neigh, + pdev, 0); + pi = (struct port_info *)netdev_priv(pdev); + tx_chan = cxgb4_port_chan(pdev); + dev_put(pdev); + } else { + pdev = get_real_dev(neigh->dev); + e = cxgb4_l2t_get(dev->rdev.lldi.l2t, neigh, + pdev, 0); + pi = (struct port_info *)netdev_priv(pdev); + tx_chan = cxgb4_port_chan(pdev); + } + neigh_release(neigh); + if (!e) { + pr_err("%s - failed to allocate l2t entry!\n", + __func__); + goto free_dst; + } + + step = dev->rdev.lldi.nrxq / dev->rdev.lldi.nchan; + rss_qid = dev->rdev.lldi.rxq_ids[pi->port_id * step]; + window = (__force u16) htons((__force u16)tcph->window); + + /* Calcuate filter portion for LE region. */ + filter = (__force unsigned int) cpu_to_be32(cxgb4_select_ntuple( + dev->rdev.lldi.ports[0], + e)); + + /* + * Synthesize the cpl_pass_accept_req. We have everything except the + * TID. Once firmware sends a reply with TID we update the TID field + * in cpl and pass it through the regular cpl_pass_accept_req path. + */ + build_cpl_pass_accept_req(skb, stid, iph->tos); + send_fw_pass_open_req(dev, skb, iph->daddr, tcph->dest, iph->saddr, + tcph->source, ntohl(tcph->seq), filter, window, + rss_qid, pi->port_id); + cxgb4_l2t_release(e); +free_dst: + dst_release(dst); +reject: return 0; } @@ -2170,7 +3674,8 @@ static c4iw_handler_func work_handlers[NUM_CPL_CMDS] = { [CPL_CLOSE_CON_RPL] = close_con_rpl, [CPL_RDMA_TERMINATE] = terminate, [CPL_FW4_ACK] = fw4_ack, - [CPL_FW6_MSG] = async_event + [CPL_FW6_MSG] = deferred_fw6_msg, + [CPL_RX_PKT] = rx_pkt }; static void process_timeout(struct c4iw_ep *ep) @@ -2181,6 +3686,7 @@ static void process_timeout(struct c4iw_ep *ep) mutex_lock(&ep->com.mutex); PDBG("%s ep %p tid %u state %d\n", __func__, ep, ep->hwtid, ep->com.state); + set_bit(TIMEDOUT, &ep->com.history); switch (ep->com.state) { case MPA_REQ_SENT: __state_set(&ep->com, ABORTING); @@ -2198,16 +3704,26 @@ static void process_timeout(struct c4iw_ep *ep) &attrs, 1); } __state_set(&ep->com, ABORTING); + close_complete_upcall(ep, -ETIMEDOUT); + break; + case ABORTING: + case DEAD: + + /* + * These states are expected if the ep timed out at the same + * time as another thread was calling stop_ep_timer(). + * So we silently do nothing for these states. + */ + abort = 0; break; default: - printk(KERN_ERR "%s unexpected state ep %p tid %u state %u\n", + WARN(1, "%s unexpected state ep %p tid %u state %u\n", __func__, ep, ep->hwtid, ep->com.state); - WARN_ON(1); abort = 0; } - mutex_unlock(&ep->com.mutex); if (abort) abort_connection(ep, NULL, GFP_KERNEL); + mutex_unlock(&ep->com.mutex); c4iw_put_ep(&ep->com); } @@ -2221,6 +3737,8 @@ static void process_timedout_eps(void) tmp = timeout_list.next; list_del(tmp); + tmp->next = NULL; + tmp->prev = NULL; spin_unlock_irq(&timeout_lock); ep = list_entry(tmp, struct c4iw_ep, entry); process_timeout(ep); @@ -2237,6 +3755,7 @@ static void process_work(struct work_struct *work) unsigned int opcode; int ret; + process_timedout_eps(); while ((skb = skb_dequeue(&rxq))) { rpl = cplhdr(skb); dev = *((struct c4iw_dev **) (skb->cb + sizeof(void *))); @@ -2246,8 +3765,8 @@ static void process_work(struct work_struct *work) ret = work_handlers[opcode](dev, skb); if (!ret) kfree_skb(skb); + process_timedout_eps(); } - process_timedout_eps(); } static DECLARE_WORK(skb_work, process_work); @@ -2255,11 +3774,21 @@ static DECLARE_WORK(skb_work, process_work); static void ep_timeout(unsigned long arg) { struct c4iw_ep *ep = (struct c4iw_ep *)arg; + int kickit = 0; spin_lock(&timeout_lock); - list_add_tail(&ep->entry, &timeout_list); + if (!test_and_set_bit(TIMEOUT, &ep->com.flags)) { + /* + * Only insert if it is not already on the list. + */ + if (!ep->entry.next) { + list_add_tail(&ep->entry, &timeout_list); + kickit = 1; + } + } spin_unlock(&timeout_lock); - queue_work(workq, &skb_work); + if (kickit) + queue_work(workq, &skb_work); } /* @@ -2302,21 +3831,16 @@ static int fw6_msg(struct c4iw_dev *dev, struct sk_buff *skb) PDBG("%s type %u\n", __func__, rpl->type); switch (rpl->type) { - case 1: + case FW6_TYPE_WR_RPL: ret = (int)((be64_to_cpu(rpl->data[0]) >> 8) & 0xff); wr_waitp = (struct c4iw_wr_wait *)(__force unsigned long) rpl->data[1]; PDBG("%s wr_waitp %p ret %u\n", __func__, wr_waitp, ret); - if (wr_waitp) { - if (ret) - wr_waitp->ret = -ret; - else - wr_waitp->ret = 0; - wr_waitp->done = 1; - wake_up(&wr_waitp->wait); - } + if (wr_waitp) + c4iw_wake_up(wr_waitp, ret ? -ret : 0); kfree_skb(skb); break; - case 2: + case FW6_TYPE_CQE: + case FW6_TYPE_OFLD_CONNECTION_WR_RPL: sched(dev, skb); break; default: @@ -2328,6 +3852,43 @@ static int fw6_msg(struct c4iw_dev *dev, struct sk_buff *skb) return 0; } +static int peer_abort_intr(struct c4iw_dev *dev, struct sk_buff *skb) +{ + struct cpl_abort_req_rss *req = cplhdr(skb); + struct c4iw_ep *ep; + struct tid_info *t = dev->rdev.lldi.tids; + unsigned int tid = GET_TID(req); + + ep = lookup_tid(t, tid); + if (!ep) { + printk(KERN_WARNING MOD + "Abort on non-existent endpoint, tid %d\n", tid); + kfree_skb(skb); + return 0; + } + if (is_neg_adv(req->status)) { + PDBG("%s neg_adv_abort ep %p tid %u\n", __func__, ep, + ep->hwtid); + kfree_skb(skb); + return 0; + } + PDBG("%s ep %p tid %u state %u\n", __func__, ep, ep->hwtid, + ep->com.state); + + /* + * Wake up any threads in rdma_init() or rdma_fini(). + * However, if we are on MPAv2 and want to retry with MPAv1 + * then, don't wake up yet. + */ + if (mpa_rev == 2 && !ep->tried_with_mpa_v1) { + if (ep->com.state != MPA_REQ_SENT) + c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); + } else + c4iw_wake_up(&ep->com.wr_wait, -ECONNRESET); + sched(dev, skb); + return 0; +} + /* * Most upcalls from the T4 Core go to sched() to * schedule the processing on a work queue. @@ -2344,11 +3905,12 @@ c4iw_handler_func c4iw_handlers[NUM_CPL_CMDS] = { [CPL_PASS_ESTABLISH] = sched, [CPL_PEER_CLOSE] = sched, [CPL_CLOSE_CON_RPL] = sched, - [CPL_ABORT_REQ_RSS] = sched, + [CPL_ABORT_REQ_RSS] = peer_abort_intr, [CPL_RDMA_TERMINATE] = sched, [CPL_FW4_ACK] = sched, [CPL_SET_TCB_RPL] = set_tcb_rpl, - [CPL_FW6_MSG] = fw6_msg + [CPL_FW6_MSG] = fw6_msg, + [CPL_RX_PKT] = sched }; int __init c4iw_cm_init(void) @@ -2363,7 +3925,7 @@ int __init c4iw_cm_init(void) return 0; } -void __exit c4iw_cm_term(void) +void c4iw_cm_term(void) { WARN_ON(!list_empty(&timeout_list)); flush_workqueue(workq); |