summaryrefslogtreecommitdiff
path: root/drivers/infiniband
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2012-12-21 16:40:26 -0800
committerLinus Torvalds <torvalds@linux-foundation.org>2012-12-21 16:40:26 -0800
commit184e2516614f7055d4c3a2e63fd8a3eb95fff6d6 (patch)
tree9822dd3cc97f8cfed3cbda6167818b60355cc7ec /drivers/infiniband
parent0264405b84505f60ae00625f261e75a32c7ddf56 (diff)
parentd72623b665d84b1e07fe43854e83387fce8dd134 (diff)
Merge tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband
Pull more infiniband changes from Roland Dreier: "Second batch of InfiniBand/RDMA changes for 3.8: - cxgb4 changes to fix lookup engine hash collisions - mlx4 changes to make flow steering usable - fix to IPoIB to avoid pinning dst reference for too long" * tag 'rdma-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/roland/infiniband: RDMA/cxgb4: Fix bug for active and passive LE hash collision path RDMA/cxgb4: Fix LE hash collision bug for passive open connection RDMA/cxgb4: Fix LE hash collision bug for active open connection mlx4_core: Allow choosing flow steering mode mlx4_core: Adjustments to Flow Steering activation logic for SR-IOV mlx4_core: Fix error flow in the flow steering wrapper mlx4_core: Add QPN enforcement for flow steering rules set by VFs cxgb4: Add LE hash collision bug fix path in LLD driver cxgb4: Add T4 filter support IPoIB: Call skb_dst_drop() once skb is enqueued for sending
Diffstat (limited to 'drivers/infiniband')
-rw-r--r--drivers/infiniband/hw/cxgb4/cm.c791
-rw-r--r--drivers/infiniband/hw/cxgb4/device.c210
-rw-r--r--drivers/infiniband/hw/cxgb4/iw_cxgb4.h33
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_cm.c3
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_ib.c3
5 files changed, 924 insertions, 116 deletions
diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c
index 5de86968379d..c13745cde7fa 100644
--- a/drivers/infiniband/hw/cxgb4/cm.c
+++ b/drivers/infiniband/hw/cxgb4/cm.c
@@ -38,10 +38,12 @@
#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 "iw_cxgb4.h"
@@ -61,6 +63,14 @@ static char *states[] = {
NULL,
};
+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=1)");
@@ -265,6 +275,7 @@ void _c4iw_free_ep(struct kref *kref)
cxgb4_remove_tid(ep->com.dev->rdev.lldi.tids, 0, ep->hwtid);
dst_release(ep->dst);
cxgb4_l2t_release(ep->l2t);
+ remove_handle(ep->com.dev, &ep->com.dev->hwtid_idr, ep->hwtid);
}
kfree(ep);
}
@@ -441,6 +452,50 @@ 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);
}
+#define VLAN_NONE 0xfff
+#define FILTER_SEL_VLAN_NONE 0xffff
+#define FILTER_SEL_WIDTH_P_FC (3+1) /* port uses 3 bits, FCoE one bit */
+#define FILTER_SEL_WIDTH_VIN_P_FC \
+ (6 + 7 + FILTER_SEL_WIDTH_P_FC) /* 6 bits are unused, VF uses 7 bits*/
+#define FILTER_SEL_WIDTH_TAG_P_FC \
+ (3 + FILTER_SEL_WIDTH_VIN_P_FC) /* PF uses 3 bits */
+#define FILTER_SEL_WIDTH_VLD_TAG_P_FC (1 + FILTER_SEL_WIDTH_TAG_P_FC)
+
+static unsigned int select_ntuple(struct c4iw_dev *dev, struct dst_entry *dst,
+ struct l2t_entry *l2t)
+{
+ unsigned int ntuple = 0;
+ u32 viid;
+
+ switch (dev->rdev.lldi.filt_mode) {
+
+ /* default filter mode */
+ case HW_TPL_FR_MT_PR_IV_P_FC:
+ if (l2t->vlan == VLAN_NONE)
+ ntuple |= FILTER_SEL_VLAN_NONE << FILTER_SEL_WIDTH_P_FC;
+ else {
+ ntuple |= l2t->vlan << FILTER_SEL_WIDTH_P_FC;
+ ntuple |= 1 << FILTER_SEL_WIDTH_VLD_TAG_P_FC;
+ }
+ ntuple |= l2t->lport << S_PORT | IPPROTO_TCP <<
+ FILTER_SEL_WIDTH_VLD_TAG_P_FC;
+ break;
+ case HW_TPL_FR_MT_PR_OV_P_FC: {
+ viid = cxgb4_port_viid(l2t->neigh->dev);
+
+ ntuple |= FW_VIID_VIN_GET(viid) << FILTER_SEL_WIDTH_P_FC;
+ ntuple |= FW_VIID_PFN_GET(viid) << FILTER_SEL_WIDTH_VIN_P_FC;
+ ntuple |= FW_VIID_VIVLD_GET(viid) << FILTER_SEL_WIDTH_TAG_P_FC;
+ ntuple |= l2t->lport << S_PORT | IPPROTO_TCP <<
+ FILTER_SEL_WIDTH_VLD_TAG_P_FC;
+ break;
+ }
+ default:
+ break;
+ }
+ return ntuple;
+}
+
static int send_connect(struct c4iw_ep *ep)
{
struct cpl_act_open_req *req;
@@ -463,7 +518,8 @@ static int send_connect(struct c4iw_ep *ep)
cxgb4_best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx);
wscale = compute_wscale(rcv_win);
- opt0 = KEEP_ALIVE(1) |
+ opt0 = (nocong ? NO_CONG(1) : 0) |
+ KEEP_ALIVE(1) |
DELACK(1) |
WND_SCALE(wscale) |
MSS_IDX(mtu_idx) |
@@ -474,6 +530,7 @@ static int send_connect(struct c4iw_ep *ep)
ULP_MODE(ULP_MODE_TCPDDP) |
RCV_BUFSIZ(rcv_win>>10);
opt2 = RX_CHANNEL(0) |
+ CCTRL_ECN(enable_ecn) |
RSS_QUEUE_VALID | RSS_QUEUE(ep->rss_qid);
if (enable_tcp_timestamps)
opt2 |= TSTAMPS_EN(1);
@@ -492,8 +549,9 @@ static int send_connect(struct c4iw_ep *ep)
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->params = cpu_to_be32(select_ntuple(ep->com.dev, ep->dst, ep->l2t));
req->opt2 = cpu_to_be32(opt2);
+ set_bit(ACT_OPEN_REQ, &ep->com.history);
return c4iw_l2t_send(&ep->com.dev->rdev, skb, ep->l2t);
}
@@ -770,6 +828,7 @@ static int act_establish(struct c4iw_dev *dev, struct sk_buff *skb)
/* 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);
@@ -777,7 +836,9 @@ 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);
@@ -803,6 +864,7 @@ static void close_complete_upcall(struct c4iw_ep *ep)
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);
}
}
@@ -811,6 +873,7 @@ 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);
+ set_bit(ABORT_CONN, &ep->com.history);
return send_abort(ep, skb, gfp);
}
@@ -825,6 +888,7 @@ 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);
}
}
@@ -843,6 +907,7 @@ static void peer_abort_upcall(struct c4iw_ep *ep)
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);
}
}
@@ -875,6 +940,7 @@ static void connect_reply_upcall(struct c4iw_ep *ep, int status)
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) {
@@ -915,6 +981,7 @@ static void connect_request_upcall(struct c4iw_ep *ep)
ep->parent_ep->com.cm_id,
&event);
}
+ set_bit(CONNREQ_UPCALL, &ep->com.history);
c4iw_put_ep(&ep->parent_ep->com);
ep->parent_ep = NULL;
}
@@ -931,6 +998,7 @@ static void established_upcall(struct c4iw_ep *ep)
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);
}
}
@@ -1316,6 +1384,7 @@ 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;
ep = lookup_tid(t, tid);
PDBG("%s ep %p tid %u dlen %u\n", __func__, ep, ep->hwtid, dlen);
@@ -1338,9 +1407,9 @@ static int rx_data(struct c4iw_dev *dev, struct sk_buff *skb)
case MPA_REP_SENT:
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);
+ pr_err("%s Unexpected streaming data." \
+ " ep %p state %d tid %u status %d\n",
+ __func__, ep, state_read(&ep->com), ep->hwtid, status);
/*
* The ep will timeout and inform the ULP of the failure.
@@ -1383,6 +1452,63 @@ 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;
+
+ 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(select_ntuple(ep->com.dev, ep->dst,
+ ep->l2t));
+ req->le.lport = ep->com.local_addr.sin_port;
+ req->le.pport = ep->com.remote_addr.sin_port;
+ req->le.u.ipv4.lip = ep->com.local_addr.sin_addr.s_addr;
+ req->le.u.ipv4.pip = ep->com.remote_addr.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 = jiffies;
+ req->tcb.rcv_adv = htons(1);
+ cxgb4_best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx);
+ wscale = compute_wscale(rcv_win);
+ req->tcb.opt0 = 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(rcv_win >> 10);
+ req->tcb.opt2 = 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 |= TSTAMPS_EN(1);
+ if (enable_tcp_sack)
+ req->tcb.opt2 |= SACK_EN(1);
+ if (wscale && enable_tcp_window_scaling)
+ req->tcb.opt2 |= WND_SCALE_EN(1);
+ req->tcb.opt0 = cpu_to_be64(req->tcb.opt0);
+ req->tcb.opt2 = cpu_to_be32(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
*/
@@ -1392,6 +1518,111 @@ static inline int act_open_has_tid(int status)
status != CPL_ERR_ARP_MISS;
}
+#define ACT_OPEN_RETRY_COUNT 2
+
+static int c4iw_reconnect(struct c4iw_ep *ep)
+{
+ int err = 0;
+ struct rtable *rt;
+ struct port_info *pi;
+ struct net_device *pdev;
+ int step;
+ struct neighbour *neigh;
+
+ 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 */
+ rt = find_route(ep->com.dev,
+ ep->com.cm_id->local_addr.sin_addr.s_addr,
+ ep->com.cm_id->remote_addr.sin_addr.s_addr,
+ ep->com.cm_id->local_addr.sin_port,
+ ep->com.cm_id->remote_addr.sin_port, 0);
+ if (!rt) {
+ pr_err("%s - cannot find route.\n", __func__);
+ err = -EHOSTUNREACH;
+ goto fail3;
+ }
+ ep->dst = &rt->dst;
+
+ neigh = dst_neigh_lookup(ep->dst,
+ &ep->com.cm_id->remote_addr.sin_addr.s_addr);
+ /* get a l2t entry */
+ if (neigh->dev->flags & IFF_LOOPBACK) {
+ PDBG("%s LOOPBACK\n", __func__);
+ pdev = ip_dev_find(&init_net,
+ ep->com.cm_id->remote_addr.sin_addr.s_addr);
+ ep->l2t = cxgb4_l2t_get(ep->com.dev->rdev.lldi.l2t,
+ neigh, pdev, 0);
+ pi = (struct port_info *)netdev_priv(pdev);
+ ep->mtu = pdev->mtu;
+ ep->tx_chan = cxgb4_port_chan(pdev);
+ ep->smac_idx = (cxgb4_port_viid(pdev) & 0x7F) << 1;
+ dev_put(pdev);
+ } else {
+ ep->l2t = cxgb4_l2t_get(ep->com.dev->rdev.lldi.l2t,
+ neigh, neigh->dev, 0);
+ pi = (struct port_info *)netdev_priv(neigh->dev);
+ ep->mtu = dst_mtu(ep->dst);
+ ep->tx_chan = cxgb4_port_chan(neigh->dev);
+ ep->smac_idx = (cxgb4_port_viid(neigh->dev) &
+ 0x7F) << 1;
+ }
+
+ step = ep->com.dev->rdev.lldi.ntxq / ep->com.dev->rdev.lldi.nchan;
+ ep->txq_idx = pi->port_id * step;
+ ep->ctrlq_idx = pi->port_id;
+ step = ep->com.dev->rdev.lldi.nrxq / ep->com.dev->rdev.lldi.nchan;
+ ep->rss_qid = ep->com.dev->rdev.lldi.rxq_ids[pi->port_id * step];
+
+ if (!ep->l2t) {
+ pr_err("%s - cannot alloc l2e.\n", __func__);
+ err = -ENOMEM;
+ 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;
@@ -1412,6 +1643,8 @@ static int act_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
return 0;
}
+ set_bit(ACT_OPEN_RPL, &ep->com.history);
+
/*
* Log interesting failures.
*/
@@ -1419,6 +1652,29 @@ static int act_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
case CPL_ERR_CONN_RESET:
case CPL_ERR_CONN_TIMEDOUT:
break;
+ case CPL_ERR_TCAM_FULL:
+ if (dev->rdev.lldi.enable_fw_ofld_conn) {
+ mutex_lock(&dev->rdev.stats.lock);
+ dev->rdev.stats.tcam_full++;
+ mutex_unlock(&dev->rdev.stats.lock);
+ 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:
printk(KERN_INFO MOD "Active open failure - "
"atid %u status %u errno %d %pI4:%u->%pI4:%u\n",
@@ -1436,6 +1692,7 @@ static int act_open_rpl(struct c4iw_dev *dev, struct sk_buff *skb)
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);
@@ -1452,13 +1709,14 @@ 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));
c4iw_wake_up(&ep->com.wr_wait, status2errno(rpl->status));
+out:
return 0;
}
@@ -1510,14 +1768,15 @@ static void accept_cr(struct c4iw_ep *ep, __be32 peer_ip, struct sk_buff *skb,
skb_get(skb);
cxgb4_best_mtu(ep->com.dev->rdev.lldi.mtus, ep->mtu, &mtu_idx);
wscale = compute_wscale(rcv_win);
- opt0 = KEEP_ALIVE(1) |
+ 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) |
+ DSCP(ep->tos >> 2) |
ULP_MODE(ULP_MODE_TCPDDP) |
RCV_BUFSIZ(rcv_win>>10);
opt2 = RX_CHANNEL(0) |
@@ -1529,6 +1788,15 @@ 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);
+ }
rpl = cplhdr(skb);
INIT_TP_WR(rpl, ep->hwtid);
@@ -1645,22 +1913,30 @@ out:
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 rtable *rt;
- __be32 local_ip, peer_ip;
+ __be32 local_ip, peer_ip = 0;
__be16 local_port, peer_port;
int err;
+ u16 peer_mss = ntohs(req->tcpopt.mss);
parent_ep = lookup_stid(t, stid);
- PDBG("%s parent ep %p tid %u\n", __func__, parent_ep, hwtid);
-
+ if (!parent_ep) {
+ PDBG("%s connect request on invalid stid %d\n", __func__, stid);
+ goto reject;
+ }
get_4tuple(req, &local_ip, &peer_ip, &local_port, &peer_port);
+ PDBG("%s parent ep %p hwtid %u laddr 0x%x raddr 0x%x lport %d " \
+ "rport %d peer_mss %d\n", __func__, parent_ep, hwtid,
+ ntohl(local_ip), ntohl(peer_ip), ntohs(local_port),
+ ntohs(peer_port), peer_mss);
+
if (state_read(&parent_ep->com) != LISTEN) {
printk(KERN_ERR "%s - listening ep not in LISTEN\n",
__func__);
@@ -1694,6 +1970,9 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb)
goto reject;
}
+ if (peer_mss && child_ep->mtu > (peer_mss + 40))
+ child_ep->mtu = peer_mss + 40;
+
state_set(&child_ep->com, CONNECTING);
child_ep->com.dev = dev;
child_ep->com.cm_id = NULL;
@@ -1715,6 +1994,7 @@ static int pass_accept_req(struct c4iw_dev *dev, struct sk_buff *skb)
init_timer(&child_ep->timer);
cxgb4_insert_tid(t, child_ep, hwtid);
accept_cr(child_ep, peer_ip, skb, req);
+ set_bit(PASS_ACCEPT_REQ, &child_ep->com.history);
goto out;
reject:
reject_cr(dev, hwtid, peer_ip, skb);
@@ -1734,12 +2014,17 @@ 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));
+ insert_handle(dev, &dev->hwtid_idr, ep, ep->hwtid);
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;
}
@@ -1759,6 +2044,7 @@ static int peer_close(struct c4iw_dev *dev, struct sk_buff *skb)
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:
@@ -1838,74 +2124,6 @@ static int is_neg_adv_abort(unsigned int status)
status == CPL_ERR_PERSIST_NEG_ADVICE;
}
-static int c4iw_reconnect(struct c4iw_ep *ep)
-{
- struct rtable *rt;
- int err = 0;
-
- 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) {
- printk(KERN_ERR MOD "%s - cannot alloc atid.\n", __func__);
- err = -ENOMEM;
- goto fail2;
- }
-
- /* find a route */
- rt = find_route(ep->com.dev,
- ep->com.cm_id->local_addr.sin_addr.s_addr,
- ep->com.cm_id->remote_addr.sin_addr.s_addr,
- ep->com.cm_id->local_addr.sin_port,
- ep->com.cm_id->remote_addr.sin_port, 0);
- if (!rt) {
- printk(KERN_ERR MOD "%s - cannot find route.\n", __func__);
- err = -EHOSTUNREACH;
- goto fail3;
- }
- ep->dst = &rt->dst;
-
- err = import_ep(ep, ep->com.cm_id->remote_addr.sin_addr.s_addr,
- ep->dst, ep->com.dev, false);
- if (err) {
- printk(KERN_ERR MOD "%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:
- 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 peer_abort(struct c4iw_dev *dev, struct sk_buff *skb)
{
struct cpl_abort_req_rss *req = cplhdr(skb);
@@ -1926,6 +2144,7 @@ static int peer_abort(struct c4iw_dev *dev, struct sk_buff *skb)
}
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().
@@ -2140,6 +2359,7 @@ int c4iw_reject_cr(struct iw_cm_id *cm_id, const void *pdata, u8 pdata_len)
c4iw_put_ep(&ep->com);
return -ECONNRESET;
}
+ set_bit(ULP_REJECT, &ep->com.history);
BUG_ON(state_read(&ep->com) != MPA_REQ_RCVD);
if (mpa_rev == 0)
abort_connection(ep, NULL, GFP_KERNEL);
@@ -2169,6 +2389,7 @@ int c4iw_accept_cr(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
BUG_ON(state_read(&ep->com) != 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);
@@ -2292,6 +2513,7 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
err = -ENOMEM;
goto fail2;
}
+ insert_handle(dev, &dev->atid_idr, ep, ep->atid);
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),
@@ -2337,6 +2559,7 @@ int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param)
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:
cm_id->rem_ref(cm_id);
@@ -2351,7 +2574,6 @@ int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog)
struct c4iw_dev *dev = to_c4iw_dev(cm_id->device);
struct c4iw_listen_ep *ep;
-
might_sleep();
ep = alloc_ep(sizeof(*ep), GFP_KERNEL);
@@ -2370,30 +2592,54 @@ int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog)
/*
* 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->stid = cxgb4_alloc_sftid(dev->rdev.lldi.tids, PF_INET, ep);
+ else
+ ep->stid = cxgb4_alloc_stid(dev->rdev.lldi.tids, PF_INET, ep);
+
if (ep->stid == -1) {
printk(KERN_ERR MOD "%s - cannot alloc stid.\n", __func__);
err = -ENOMEM;
goto fail2;
}
-
+ insert_handle(dev, &dev->stid_idr, ep, ep->stid);
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)
- goto fail3;
-
- /* wait for pass_open_rpl */
- err = c4iw_wait_for_reply(&ep->com.dev->rdev, &ep->com.wr_wait, 0, 0,
- __func__);
+ if (dev->rdev.lldi.enable_fw_ofld_conn) {
+ do {
+ err = cxgb4_create_server_filter(
+ ep->com.dev->rdev.lldi.ports[0], ep->stid,
+ ep->com.local_addr.sin_addr.s_addr,
+ ep->com.local_addr.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, ep->com.local_addr.sin_addr.s_addr,
+ ep->com.local_addr.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) {
cm_id->provider_data = ep;
goto out;
}
-fail3:
+ pr_err("%s cxgb4_create_server/filter failed err %d " \
+ "stid %d laddr %08x lport %d\n", \
+ __func__, err, ep->stid,
+ ntohl(ep->com.local_addr.sin_addr.s_addr),
+ ntohs(ep->com.local_addr.sin_port));
cxgb4_free_stid(ep->com.dev->rdev.lldi.tids, ep->stid, PF_INET);
fail2:
cm_id->rem_ref(cm_id);
@@ -2412,12 +2658,19 @@ 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__);
+ if (ep->com.dev->rdev.lldi.enable_fw_ofld_conn) {
+ 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 = listen_stop(ep);
+ 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, PF_INET);
done:
cm_id->rem_ref(cm_id);
@@ -2481,10 +2734,13 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp)
if (close) {
if (abrupt) {
+ set_bit(EP_DISC_ABORT, &ep->com.history);
close_complete_upcall(ep);
ret = send_abort(ep, NULL, gfp);
- } else
+ } else {
+ set_bit(EP_DISC_CLOSE, &ep->com.history);
ret = send_halfclose(ep, gfp);
+ }
if (ret)
fatal = 1;
}
@@ -2494,10 +2750,323 @@ int c4iw_ep_disconnect(struct c4iw_ep *ep, int abrupt, gfp_t gfp)
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, 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 *)cpu_to_be64(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,
+ htonl(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;
+ u8 intf;
+ struct cpl_rx_pkt *cpl = cplhdr(skb);
+ struct cpl_pass_accept_req *req;
+ struct tcp_options_received tmp_opt;
+
+ /* Store values from cpl_rx_pkt in temporary location. */
+ vlantag = cpl->vlan;
+ len = cpl->len;
+ l2info = cpl->l2info;
+ hdr_len = 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, 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(htonl(l2info))) |
+ F_SYN_XACT_MATCH);
+ req->hdr_len = cpu_to_be32(V_SYN_RX_CHAN(G_RX_CHAN(htonl(l2info))) |
+ V_TCP_HDR_LEN(G_RX_TCPHDR_LEN(htons(hdr_len))) |
+ V_IP_HDR_LEN(G_RX_IPHDR_LEN(htons(hdr_len))) |
+ V_ETH_HDR_LEN(G_RX_ETHHDR_LEN(htonl(l2info))));
+ req->vlan = vlantag;
+ req->len = 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);
+
+ 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 = 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 = cpu_to_be64((u64)skb);
+
+ set_wr_txq(req_skb, CPL_PRIORITY_CONTROL, port_id);
+ cxgb4_ofld_send(dev->rdev.lldi.ports[0], 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 rtable *rt;
+ struct c4iw_ep *lep;
+ u16 window;
+ struct port_info *pi;
+ struct net_device *pdev;
+ u16 rss_qid;
+ 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 = cpu_to_be32(rss->hash_val) - dev->rdev.lldi.tids->sftid_base
+ + dev->rdev.lldi.tids->nstids;
+
+ 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;
+ }
+
+ if (G_RX_ETHHDR_LEN(ntohl(cpl->l2info)) == 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);
+
+ rt = find_route(dev, iph->daddr, iph->saddr, tcph->dest, tcph->source,
+ iph->tos);
+ if (!rt) {
+ pr_err("%s - failed to find dst entry!\n",
+ __func__);
+ goto reject;
+ }
+ dst = &rt->dst;
+ neigh = dst_neigh_lookup_skb(dst, skb);
+
+ 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 {
+ e = cxgb4_l2t_get(dev->rdev.lldi.l2t, neigh,
+ neigh->dev, 0);
+ pi = (struct port_info *)netdev_priv(neigh->dev);
+ tx_chan = cxgb4_port_chan(neigh->dev);
+ }
+ 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 = htons(tcph->window);
+
+ /* Calcuate filter portion for LE region. */
+ filter = cpu_to_be32(select_ntuple(dev, dst, 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;
}
@@ -2520,7 +3089,8 @@ 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)
@@ -2531,6 +3101,7 @@ 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);
@@ -2651,7 +3222,7 @@ 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);
@@ -2659,7 +3230,8 @@ static int fw6_msg(struct c4iw_dev *dev, struct sk_buff *skb)
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:
@@ -2722,7 +3294,8 @@ c4iw_handler_func c4iw_handlers[NUM_CPL_CMDS] = {
[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)
diff --git a/drivers/infiniband/hw/cxgb4/device.c b/drivers/infiniband/hw/cxgb4/device.c
index cb4ecd783700..ba11c76c0b5a 100644
--- a/drivers/infiniband/hw/cxgb4/device.c
+++ b/drivers/infiniband/hw/cxgb4/device.c
@@ -279,6 +279,11 @@ static int stats_show(struct seq_file *seq, void *v)
seq_printf(seq, " DB State: %s Transitions %llu\n",
db_state_str[dev->db_state],
dev->rdev.stats.db_state_transitions);
+ seq_printf(seq, "TCAM_FULL: %10llu\n", dev->rdev.stats.tcam_full);
+ seq_printf(seq, "ACT_OFLD_CONN_FAILS: %10llu\n",
+ dev->rdev.stats.act_ofld_conn_fails);
+ seq_printf(seq, "PAS_OFLD_CONN_FAILS: %10llu\n",
+ dev->rdev.stats.pas_ofld_conn_fails);
return 0;
}
@@ -309,6 +314,9 @@ static ssize_t stats_clear(struct file *file, const char __user *buf,
dev->rdev.stats.db_empty = 0;
dev->rdev.stats.db_drop = 0;
dev->rdev.stats.db_state_transitions = 0;
+ dev->rdev.stats.tcam_full = 0;
+ dev->rdev.stats.act_ofld_conn_fails = 0;
+ dev->rdev.stats.pas_ofld_conn_fails = 0;
mutex_unlock(&dev->rdev.stats.lock);
return count;
}
@@ -322,6 +330,113 @@ static const struct file_operations stats_debugfs_fops = {
.write = stats_clear,
};
+static int dump_ep(int id, void *p, void *data)
+{
+ struct c4iw_ep *ep = p;
+ struct c4iw_debugfs_data *epd = data;
+ int space;
+ int cc;
+
+ space = epd->bufsize - epd->pos - 1;
+ if (space == 0)
+ return 1;
+
+ cc = snprintf(epd->buf + epd->pos, space,
+ "ep %p cm_id %p qp %p state %d flags 0x%lx history 0x%lx "
+ "hwtid %d atid %d %pI4:%d <-> %pI4:%d\n",
+ ep, ep->com.cm_id, ep->com.qp, (int)ep->com.state,
+ ep->com.flags, ep->com.history, ep->hwtid, ep->atid,
+ &ep->com.local_addr.sin_addr.s_addr,
+ ntohs(ep->com.local_addr.sin_port),
+ &ep->com.remote_addr.sin_addr.s_addr,
+ ntohs(ep->com.remote_addr.sin_port));
+ if (cc < space)
+ epd->pos += cc;
+ return 0;
+}
+
+static int dump_listen_ep(int id, void *p, void *data)
+{
+ struct c4iw_listen_ep *ep = p;
+ struct c4iw_debugfs_data *epd = data;
+ int space;
+ int cc;
+
+ space = epd->bufsize - epd->pos - 1;
+ if (space == 0)
+ return 1;
+
+ cc = snprintf(epd->buf + epd->pos, space,
+ "ep %p cm_id %p state %d flags 0x%lx stid %d backlog %d "
+ "%pI4:%d\n", ep, ep->com.cm_id, (int)ep->com.state,
+ ep->com.flags, ep->stid, ep->backlog,
+ &ep->com.local_addr.sin_addr.s_addr,
+ ntohs(ep->com.local_addr.sin_port));
+ if (cc < space)
+ epd->pos += cc;
+ return 0;
+}
+
+static int ep_release(struct inode *inode, struct file *file)
+{
+ struct c4iw_debugfs_data *epd = file->private_data;
+ if (!epd) {
+ pr_info("%s null qpd?\n", __func__);
+ return 0;
+ }
+ vfree(epd->buf);
+ kfree(epd);
+ return 0;
+}
+
+static int ep_open(struct inode *inode, struct file *file)
+{
+ struct c4iw_debugfs_data *epd;
+ int ret = 0;
+ int count = 1;
+
+ epd = kmalloc(sizeof(*epd), GFP_KERNEL);
+ if (!epd) {
+ ret = -ENOMEM;
+ goto out;
+ }
+ epd->devp = inode->i_private;
+ epd->pos = 0;
+
+ spin_lock_irq(&epd->devp->lock);
+ idr_for_each(&epd->devp->hwtid_idr, count_idrs, &count);
+ idr_for_each(&epd->devp->atid_idr, count_idrs, &count);
+ idr_for_each(&epd->devp->stid_idr, count_idrs, &count);
+ spin_unlock_irq(&epd->devp->lock);
+
+ epd->bufsize = count * 160;
+ epd->buf = vmalloc(epd->bufsize);
+ if (!epd->buf) {
+ ret = -ENOMEM;
+ goto err1;
+ }
+
+ spin_lock_irq(&epd->devp->lock);
+ idr_for_each(&epd->devp->hwtid_idr, dump_ep, epd);
+ idr_for_each(&epd->devp->atid_idr, dump_ep, epd);
+ idr_for_each(&epd->devp->stid_idr, dump_listen_ep, epd);
+ spin_unlock_irq(&epd->devp->lock);
+
+ file->private_data = epd;
+ goto out;
+err1:
+ kfree(epd);
+out:
+ return ret;
+}
+
+static const struct file_operations ep_debugfs_fops = {
+ .owner = THIS_MODULE,
+ .open = ep_open,
+ .release = ep_release,
+ .read = debugfs_read,
+};
+
static int setup_debugfs(struct c4iw_dev *devp)
{
struct dentry *de;
@@ -344,6 +459,11 @@ static int setup_debugfs(struct c4iw_dev *devp)
if (de && de->d_inode)
de->d_inode->i_size = 4096;
+ de = debugfs_create_file("eps", S_IWUSR, devp->debugfs_root,
+ (void *)devp, &ep_debugfs_fops);
+ if (de && de->d_inode)
+ de->d_inode->i_size = 4096;
+
return 0;
}
@@ -475,6 +595,9 @@ static void c4iw_dealloc(struct uld_ctx *ctx)
idr_destroy(&ctx->dev->cqidr);
idr_destroy(&ctx->dev->qpidr);
idr_destroy(&ctx->dev->mmidr);
+ idr_destroy(&ctx->dev->hwtid_idr);
+ idr_destroy(&ctx->dev->stid_idr);
+ idr_destroy(&ctx->dev->atid_idr);
iounmap(ctx->dev->rdev.oc_mw_kva);
ib_dealloc_device(&ctx->dev->ibdev);
ctx->dev = NULL;
@@ -532,6 +655,9 @@ static struct c4iw_dev *c4iw_alloc(const struct cxgb4_lld_info *infop)
idr_init(&devp->cqidr);
idr_init(&devp->qpidr);
idr_init(&devp->mmidr);
+ idr_init(&devp->hwtid_idr);
+ idr_init(&devp->stid_idr);
+ idr_init(&devp->atid_idr);
spin_lock_init(&devp->lock);
mutex_init(&devp->rdev.stats.lock);
mutex_init(&devp->db_mutex);
@@ -577,14 +703,76 @@ out:
return ctx;
}
+static inline struct sk_buff *copy_gl_to_skb_pkt(const struct pkt_gl *gl,
+ const __be64 *rsp,
+ u32 pktshift)
+{
+ struct sk_buff *skb;
+
+ /*
+ * Allocate space for cpl_pass_accept_req which will be synthesized by
+ * driver. Once the driver synthesizes the request the skb will go
+ * through the regular cpl_pass_accept_req processing.
+ * The math here assumes sizeof cpl_pass_accept_req >= sizeof
+ * cpl_rx_pkt.
+ */
+ skb = alloc_skb(gl->tot_len + sizeof(struct cpl_pass_accept_req) +
+ sizeof(struct rss_header) - pktshift, GFP_ATOMIC);
+ if (unlikely(!skb))
+ return NULL;
+
+ __skb_put(skb, gl->tot_len + sizeof(struct cpl_pass_accept_req) +
+ sizeof(struct rss_header) - pktshift);
+
+ /*
+ * This skb will contain:
+ * rss_header from the rspq descriptor (1 flit)
+ * cpl_rx_pkt struct from the rspq descriptor (2 flits)
+ * space for the difference between the size of an
+ * rx_pkt and pass_accept_req cpl (1 flit)
+ * the packet data from the gl
+ */
+ skb_copy_to_linear_data(skb, rsp, sizeof(struct cpl_pass_accept_req) +
+ sizeof(struct rss_header));
+ skb_copy_to_linear_data_offset(skb, sizeof(struct rss_header) +
+ sizeof(struct cpl_pass_accept_req),
+ gl->va + pktshift,
+ gl->tot_len - pktshift);
+ return skb;
+}
+
+static inline int recv_rx_pkt(struct c4iw_dev *dev, const struct pkt_gl *gl,
+ const __be64 *rsp)
+{
+ unsigned int opcode = *(u8 *)rsp;
+ struct sk_buff *skb;
+
+ if (opcode != CPL_RX_PKT)
+ goto out;
+
+ skb = copy_gl_to_skb_pkt(gl , rsp, dev->rdev.lldi.sge_pktshift);
+ if (skb == NULL)
+ goto out;
+
+ if (c4iw_handlers[opcode] == NULL) {
+ pr_info("%s no handler opcode 0x%x...\n", __func__,
+ opcode);
+ kfree_skb(skb);
+ goto out;
+ }
+ c4iw_handlers[opcode](dev, skb);
+ return 1;
+out:
+ return 0;
+}
+
static int c4iw_uld_rx_handler(void *handle, const __be64 *rsp,
const struct pkt_gl *gl)
{
struct uld_ctx *ctx = handle;
struct c4iw_dev *dev = ctx->dev;
struct sk_buff *skb;
- const struct cpl_act_establish *rpl;
- unsigned int opcode;
+ u8 opcode;
if (gl == NULL) {
/* omit RSS and rsp_ctrl at end of descriptor */
@@ -601,19 +789,29 @@ static int c4iw_uld_rx_handler(void *handle, const __be64 *rsp,
u32 qid = be32_to_cpu(rc->pldbuflen_qid);
c4iw_ev_handler(dev, qid);
return 0;
+ } else if (unlikely(*(u8 *)rsp != *(u8 *)gl->va)) {
+ if (recv_rx_pkt(dev, gl, rsp))
+ return 0;
+
+ pr_info("%s: unexpected FL contents at %p, " \
+ "RSS %#llx, FL %#llx, len %u\n",
+ pci_name(ctx->lldi.pdev), gl->va,
+ (unsigned long long)be64_to_cpu(*rsp),
+ (unsigned long long)be64_to_cpu(*(u64 *)gl->va),
+ gl->tot_len);
+
+ return 0;
} else {
skb = cxgb4_pktgl_to_skb(gl, 128, 128);
if (unlikely(!skb))
goto nomem;
}
- rpl = cplhdr(skb);
- opcode = rpl->ot.opcode;
-
+ opcode = *(u8 *)rsp;
if (c4iw_handlers[opcode])
c4iw_handlers[opcode](dev, skb);
else
- printk(KERN_INFO "%s no handler opcode 0x%x...\n", __func__,
+ pr_info("%s no handler opcode 0x%x...\n", __func__,
opcode);
return 0;
diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
index 9beb3a9f0336..9c1644fb0259 100644
--- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
+++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
@@ -130,6 +130,9 @@ struct c4iw_stats {
u64 db_empty;
u64 db_drop;
u64 db_state_transitions;
+ u64 tcam_full;
+ u64 act_ofld_conn_fails;
+ u64 pas_ofld_conn_fails;
};
struct c4iw_rdev {
@@ -223,6 +226,9 @@ struct c4iw_dev {
struct dentry *debugfs_root;
enum db_state db_state;
int qpcnt;
+ struct idr hwtid_idr;
+ struct idr atid_idr;
+ struct idr stid_idr;
};
static inline struct c4iw_dev *to_c4iw_dev(struct ib_device *ibdev)
@@ -712,6 +718,31 @@ enum c4iw_ep_flags {
CLOSE_SENT = 3,
};
+enum c4iw_ep_history {
+ ACT_OPEN_REQ = 0,
+ ACT_OFLD_CONN = 1,
+ ACT_OPEN_RPL = 2,
+ ACT_ESTAB = 3,
+ PASS_ACCEPT_REQ = 4,
+ PASS_ESTAB = 5,
+ ABORT_UPCALL = 6,
+ ESTAB_UPCALL = 7,
+ CLOSE_UPCALL = 8,
+ ULP_ACCEPT = 9,
+ ULP_REJECT = 10,
+ TIMEDOUT = 11,
+ PEER_ABORT = 12,
+ PEER_CLOSE = 13,
+ CONNREQ_UPCALL = 14,
+ ABORT_CONN = 15,
+ DISCONN_UPCALL = 16,
+ EP_DISC_CLOSE = 17,
+ EP_DISC_ABORT = 18,
+ CONN_RPL_UPCALL = 19,
+ ACT_RETRY_NOMEM = 20,
+ ACT_RETRY_INUSE = 21
+};
+
struct c4iw_ep_common {
struct iw_cm_id *cm_id;
struct c4iw_qp *qp;
@@ -723,6 +754,7 @@ struct c4iw_ep_common {
struct sockaddr_in remote_addr;
struct c4iw_wr_wait wr_wait;
unsigned long flags;
+ unsigned long history;
};
struct c4iw_listen_ep {
@@ -760,6 +792,7 @@ struct c4iw_ep {
u8 tos;
u8 retry_with_mpa_v1;
u8 tried_with_mpa_v1;
+ unsigned int retry_count;
};
static inline struct c4iw_ep *to_ep(struct iw_cm_id *cm_id)
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
index 72ae63f0072d..03103d2bd641 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c
@@ -752,6 +752,9 @@ void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_
dev->trans_start = jiffies;
++tx->tx_head;
+ skb_orphan(skb);
+ skb_dst_drop(skb);
+
if (++priv->tx_outstanding == ipoib_sendq_size) {
ipoib_dbg(priv, "TX ring 0x%x full, stopping kernel net queue\n",
tx->qp->qp_num);
diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
index f10221f40803..a1bca70e20aa 100644
--- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c
+++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c
@@ -615,8 +615,9 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb,
address->last_send = priv->tx_head;
++priv->tx_head;
- skb_orphan(skb);
+ skb_orphan(skb);
+ skb_dst_drop(skb);
}
if (unlikely(priv->tx_outstanding > MAX_SEND_CQE))