summaryrefslogtreecommitdiff
path: root/drivers/infiniband
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/infiniband')
-rw-r--r--drivers/infiniband/core/cma.c9
-rw-r--r--drivers/infiniband/hw/amso1100/c2.c7
-rw-r--r--drivers/infiniband/hw/amso1100/c2.h8
-rw-r--r--drivers/infiniband/hw/amso1100/c2_ae.c1
-rw-r--r--drivers/infiniband/hw/amso1100/c2_pd.c4
-rw-r--r--drivers/infiniband/hw/amso1100/c2_qp.c4
-rw-r--r--drivers/infiniband/hw/amso1100/c2_rnic.c4
-rw-r--r--drivers/infiniband/hw/cxgb3/iwch_cm.c6
-rw-r--r--drivers/infiniband/hw/cxgb4/cm.c797
-rw-r--r--drivers/infiniband/hw/cxgb4/device.c210
-rw-r--r--drivers/infiniband/hw/cxgb4/iw_cxgb4.h33
-rw-r--r--drivers/infiniband/hw/ehca/ehca_main.c6
-rw-r--r--drivers/infiniband/hw/ehca/hcp_if.c20
-rw-r--r--drivers/infiniband/hw/ipath/ipath_driver.c12
-rw-r--r--drivers/infiniband/hw/ipath/ipath_init_chip.c10
-rw-r--r--drivers/infiniband/hw/mlx4/cm.c4
-rw-r--r--drivers/infiniband/hw/mlx4/cq.c34
-rw-r--r--drivers/infiniband/hw/mlx4/main.c27
-rw-r--r--drivers/infiniband/hw/mlx4/mlx4_ib.h1
-rw-r--r--drivers/infiniband/hw/mlx4/user.h12
-rw-r--r--drivers/infiniband/hw/mthca/mthca_main.c9
-rw-r--r--drivers/infiniband/hw/nes/nes.c6
-rw-r--r--drivers/infiniband/hw/nes/nes.h1
-rw-r--r--drivers/infiniband/hw/nes/nes_cm.c32
-rw-r--r--drivers/infiniband/hw/nes/nes_hw.c9
-rw-r--r--drivers/infiniband/hw/nes/nes_mgt.c42
-rw-r--r--drivers/infiniband/hw/nes/nes_nic.c13
-rw-r--r--drivers/infiniband/hw/nes/nes_verbs.c9
-rw-r--r--drivers/infiniband/hw/qib/qib_init.c12
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_cm.c3
-rw-r--r--drivers/infiniband/ulp/ipoib/ipoib_ib.c3
-rw-r--r--drivers/infiniband/ulp/srp/ib_srp.c314
-rw-r--r--drivers/infiniband/ulp/srp/ib_srp.h11
-rw-r--r--drivers/infiniband/ulp/srpt/ib_srpt.c178
-rw-r--r--drivers/infiniband/ulp/srpt/ib_srpt.h1
35 files changed, 1342 insertions, 510 deletions
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index a7568c34a1aa..d789eea32168 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -345,17 +345,17 @@ static int find_gid_port(struct ib_device *device, union ib_gid *gid, u8 port_nu
err = ib_query_port(device, port_num, &props);
if (err)
- return 1;
+ return err;
for (i = 0; i < props.gid_tbl_len; ++i) {
err = ib_query_gid(device, port_num, i, &tmp);
if (err)
- return 1;
+ return err;
if (!memcmp(&tmp, gid, sizeof tmp))
return 0;
}
- return -EAGAIN;
+ return -EADDRNOTAVAIL;
}
static int cma_acquire_dev(struct rdma_id_private *id_priv)
@@ -388,8 +388,7 @@ static int cma_acquire_dev(struct rdma_id_private *id_priv)
if (!ret) {
id_priv->id.port_num = port;
goto out;
- } else if (ret == 1)
- break;
+ }
}
}
}
diff --git a/drivers/infiniband/hw/amso1100/c2.c b/drivers/infiniband/hw/amso1100/c2.c
index 5ce7b9e8bff6..7275e727e0f5 100644
--- a/drivers/infiniband/hw/amso1100/c2.c
+++ b/drivers/infiniband/hw/amso1100/c2.c
@@ -920,8 +920,7 @@ static struct net_device *c2_devinit(struct c2_dev *c2dev,
return netdev;
}
-static int __devinit c2_probe(struct pci_dev *pcidev,
- const struct pci_device_id *ent)
+static int c2_probe(struct pci_dev *pcidev, const struct pci_device_id *ent)
{
int ret = 0, i;
unsigned long reg0_start, reg0_flags, reg0_len;
@@ -1191,7 +1190,7 @@ static int __devinit c2_probe(struct pci_dev *pcidev,
return ret;
}
-static void __devexit c2_remove(struct pci_dev *pcidev)
+static void c2_remove(struct pci_dev *pcidev)
{
struct c2_dev *c2dev = pci_get_drvdata(pcidev);
struct net_device *netdev = c2dev->netdev;
@@ -1236,7 +1235,7 @@ static struct pci_driver c2_pci_driver = {
.name = DRV_NAME,
.id_table = c2_pci_table,
.probe = c2_probe,
- .remove = __devexit_p(c2_remove),
+ .remove = c2_remove,
};
static int __init c2_init_module(void)
diff --git a/drivers/infiniband/hw/amso1100/c2.h b/drivers/infiniband/hw/amso1100/c2.h
index 6ae698e68775..ba7a1208ff9e 100644
--- a/drivers/infiniband/hw/amso1100/c2.h
+++ b/drivers/infiniband/hw/amso1100/c2.h
@@ -498,16 +498,16 @@ extern int c2_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr,
struct ib_send_wr **bad_wr);
extern int c2_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *ib_wr,
struct ib_recv_wr **bad_wr);
-extern void __devinit c2_init_qp_table(struct c2_dev *c2dev);
-extern void __devexit c2_cleanup_qp_table(struct c2_dev *c2dev);
+extern void c2_init_qp_table(struct c2_dev *c2dev);
+extern void c2_cleanup_qp_table(struct c2_dev *c2dev);
extern void c2_set_qp_state(struct c2_qp *, int);
extern struct c2_qp *c2_find_qpn(struct c2_dev *c2dev, int qpn);
/* PDs */
extern int c2_pd_alloc(struct c2_dev *c2dev, int privileged, struct c2_pd *pd);
extern void c2_pd_free(struct c2_dev *c2dev, struct c2_pd *pd);
-extern int __devinit c2_init_pd_table(struct c2_dev *c2dev);
-extern void __devexit c2_cleanup_pd_table(struct c2_dev *c2dev);
+extern int c2_init_pd_table(struct c2_dev *c2dev);
+extern void c2_cleanup_pd_table(struct c2_dev *c2dev);
/* CQs */
extern int c2_init_cq(struct c2_dev *c2dev, int entries,
diff --git a/drivers/infiniband/hw/amso1100/c2_ae.c b/drivers/infiniband/hw/amso1100/c2_ae.c
index 32d34e88d5cf..706cf97cbe8f 100644
--- a/drivers/infiniband/hw/amso1100/c2_ae.c
+++ b/drivers/infiniband/hw/amso1100/c2_ae.c
@@ -311,6 +311,7 @@ void c2_ae_event(struct c2_dev *c2dev, u32 mq_index)
if (cq->ibcq.event_handler)
cq->ibcq.event_handler(&ib_event,
cq->ibcq.cq_context);
+ break;
}
default:
diff --git a/drivers/infiniband/hw/amso1100/c2_pd.c b/drivers/infiniband/hw/amso1100/c2_pd.c
index 161f2a285351..f3e81dc357bb 100644
--- a/drivers/infiniband/hw/amso1100/c2_pd.c
+++ b/drivers/infiniband/hw/amso1100/c2_pd.c
@@ -70,7 +70,7 @@ void c2_pd_free(struct c2_dev *c2dev, struct c2_pd *pd)
spin_unlock(&c2dev->pd_table.lock);
}
-int __devinit c2_init_pd_table(struct c2_dev *c2dev)
+int c2_init_pd_table(struct c2_dev *c2dev)
{
c2dev->pd_table.last = 0;
@@ -84,7 +84,7 @@ int __devinit c2_init_pd_table(struct c2_dev *c2dev)
return 0;
}
-void __devexit c2_cleanup_pd_table(struct c2_dev *c2dev)
+void c2_cleanup_pd_table(struct c2_dev *c2dev)
{
kfree(c2dev->pd_table.table);
}
diff --git a/drivers/infiniband/hw/amso1100/c2_qp.c b/drivers/infiniband/hw/amso1100/c2_qp.c
index 0d7b6f23caff..28cd5cb51859 100644
--- a/drivers/infiniband/hw/amso1100/c2_qp.c
+++ b/drivers/infiniband/hw/amso1100/c2_qp.c
@@ -1010,13 +1010,13 @@ out:
return err;
}
-void __devinit c2_init_qp_table(struct c2_dev *c2dev)
+void c2_init_qp_table(struct c2_dev *c2dev)
{
spin_lock_init(&c2dev->qp_table.lock);
idr_init(&c2dev->qp_table.idr);
}
-void __devexit c2_cleanup_qp_table(struct c2_dev *c2dev)
+void c2_cleanup_qp_table(struct c2_dev *c2dev)
{
idr_destroy(&c2dev->qp_table.idr);
}
diff --git a/drivers/infiniband/hw/amso1100/c2_rnic.c b/drivers/infiniband/hw/amso1100/c2_rnic.c
index e4a73158fc7f..b7c986990053 100644
--- a/drivers/infiniband/hw/amso1100/c2_rnic.c
+++ b/drivers/infiniband/hw/amso1100/c2_rnic.c
@@ -442,7 +442,7 @@ static int c2_rnic_close(struct c2_dev *c2dev)
* involves initializing the various limits and resource pools that
* comprise the RNIC instance.
*/
-int __devinit c2_rnic_init(struct c2_dev *c2dev)
+int c2_rnic_init(struct c2_dev *c2dev)
{
int err;
u32 qsize, msgsize;
@@ -611,7 +611,7 @@ int __devinit c2_rnic_init(struct c2_dev *c2dev)
/*
* Called by c2_remove to cleanup the RNIC resources.
*/
-void __devexit c2_rnic_term(struct c2_dev *c2dev)
+void c2_rnic_term(struct c2_dev *c2dev)
{
/* Close the open adapter instance */
diff --git a/drivers/infiniband/hw/cxgb3/iwch_cm.c b/drivers/infiniband/hw/cxgb3/iwch_cm.c
index aaf88ef9409c..3e094cd6a0e3 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_cm.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_cm.c
@@ -128,9 +128,8 @@ static void stop_ep_timer(struct iwch_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",
+ WARN(1, "%s timer stopped when its not running! ep %p state %u\n",
__func__, ep, ep->com.state);
- WARN_ON(1);
return;
}
del_timer_sync(&ep->timer);
@@ -1756,9 +1755,8 @@ static void ep_timeout(unsigned long arg)
__state_set(&ep->com, ABORTING);
break;
default:
- printk(KERN_ERR "%s unexpected state ep %p state %u\n",
+ WARN(1, "%s unexpected state ep %p state %u\n",
__func__, ep, ep->com.state);
- WARN_ON(1);
abort = 0;
}
spin_unlock_irqrestore(&ep->com.lock, flags);
diff --git a/drivers/infiniband/hw/cxgb4/cm.c b/drivers/infiniband/hw/cxgb4/cm.c
index 6cfd4d8fd0bd..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)");
@@ -151,9 +161,8 @@ static void 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! "
+ WARN(1, "%s timer stopped when its not running! "
"ep %p state %u\n", __func__, ep, ep->com.state);
- WARN_ON(1);
return;
}
del_timer_sync(&ep->timer);
@@ -266,6 +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);
}
@@ -442,6 +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;
@@ -464,7 +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) |
@@ -475,6 +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);
@@ -493,8 +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);
}
@@ -771,6 +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);
@@ -778,7 +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);
@@ -804,6 +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);
}
}
@@ -812,6 +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);
}
@@ -826,6 +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);
}
}
@@ -844,6 +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);
}
}
@@ -876,6 +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) {
@@ -916,6 +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;
}
@@ -932,6 +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);
}
}
@@ -1317,6 +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);
@@ -1339,9 +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.
@@ -1384,6 +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
*/
@@ -1393,6 +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;
@@ -1413,6 +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.
*/
@@ -1420,6 +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",
@@ -1437,6 +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);
@@ -1453,13 +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;
}
@@ -1511,14 +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) |
@@ -1530,6 +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);
@@ -1646,22 +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__);
@@ -1695,6 +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;
@@ -1716,6 +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);
@@ -1735,12 +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;
}
@@ -1760,6 +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:
@@ -1839,74 +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);
@@ -1927,6 +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().
@@ -2141,6 +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);
@@ -2170,6 +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);
@@ -2293,6 +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),
@@ -2338,6 +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);
@@ -2352,7 +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);
@@ -2371,30 +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);
@@ -2413,12 +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);
@@ -2482,10 +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;
}
@@ -2495,10 +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;
}
@@ -2521,7 +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)
@@ -2532,6 +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);
@@ -2551,9 +3121,8 @@ static void process_timeout(struct c4iw_ep *ep)
__state_set(&ep->com, ABORTING);
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);
@@ -2653,7 +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);
@@ -2661,7 +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:
@@ -2724,7 +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/hw/ehca/ehca_main.c b/drivers/infiniband/hw/ehca/ehca_main.c
index 832e7a7d0aee..f8a62918a88d 100644
--- a/drivers/infiniband/hw/ehca/ehca_main.c
+++ b/drivers/infiniband/hw/ehca/ehca_main.c
@@ -713,8 +713,8 @@ static struct attribute_group ehca_dev_attr_grp = {
.attrs = ehca_dev_attrs
};
-static int __devinit ehca_probe(struct platform_device *dev,
- const struct of_device_id *id)
+static int ehca_probe(struct platform_device *dev,
+ const struct of_device_id *id)
{
struct ehca_shca *shca;
const u64 *handle;
@@ -879,7 +879,7 @@ probe1:
return -EINVAL;
}
-static int __devexit ehca_remove(struct platform_device *dev)
+static int ehca_remove(struct platform_device *dev)
{
struct ehca_shca *shca = dev_get_drvdata(&dev->dev);
unsigned long flags;
diff --git a/drivers/infiniband/hw/ehca/hcp_if.c b/drivers/infiniband/hw/ehca/hcp_if.c
index 2d41d04fd959..89517ffb4389 100644
--- a/drivers/infiniband/hw/ehca/hcp_if.c
+++ b/drivers/infiniband/hw/ehca/hcp_if.c
@@ -90,26 +90,6 @@
static DEFINE_SPINLOCK(hcall_lock);
-static u32 get_longbusy_msecs(int longbusy_rc)
-{
- switch (longbusy_rc) {
- case H_LONG_BUSY_ORDER_1_MSEC:
- return 1;
- case H_LONG_BUSY_ORDER_10_MSEC:
- return 10;
- case H_LONG_BUSY_ORDER_100_MSEC:
- return 100;
- case H_LONG_BUSY_ORDER_1_SEC:
- return 1000;
- case H_LONG_BUSY_ORDER_10_SEC:
- return 10000;
- case H_LONG_BUSY_ORDER_100_SEC:
- return 100000;
- default:
- return 1;
- }
-}
-
static long ehca_plpar_hcall_norets(unsigned long opcode,
unsigned long arg1,
unsigned long arg2,
diff --git a/drivers/infiniband/hw/ipath/ipath_driver.c b/drivers/infiniband/hw/ipath/ipath_driver.c
index bfca37b2432f..7b371f545ece 100644
--- a/drivers/infiniband/hw/ipath/ipath_driver.c
+++ b/drivers/infiniband/hw/ipath/ipath_driver.c
@@ -127,9 +127,8 @@ const char *ipath_ibcstatus_str[] = {
"LTState1C", "LTState1D", "LTState1E", "LTState1F"
};
-static void __devexit ipath_remove_one(struct pci_dev *);
-static int __devinit ipath_init_one(struct pci_dev *,
- const struct pci_device_id *);
+static void ipath_remove_one(struct pci_dev *);
+static int ipath_init_one(struct pci_dev *, const struct pci_device_id *);
/* Only needed for registration, nothing else needs this info */
#define PCI_VENDOR_ID_PATHSCALE 0x1fc1
@@ -148,7 +147,7 @@ MODULE_DEVICE_TABLE(pci, ipath_pci_tbl);
static struct pci_driver ipath_driver = {
.name = IPATH_DRV_NAME,
.probe = ipath_init_one,
- .remove = __devexit_p(ipath_remove_one),
+ .remove = ipath_remove_one,
.id_table = ipath_pci_tbl,
.driver = {
.groups = ipath_driver_attr_groups,
@@ -392,8 +391,7 @@ done:
static void cleanup_device(struct ipath_devdata *dd);
-static int __devinit ipath_init_one(struct pci_dev *pdev,
- const struct pci_device_id *ent)
+static int ipath_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
{
int ret, len, j;
struct ipath_devdata *dd;
@@ -737,7 +735,7 @@ static void cleanup_device(struct ipath_devdata *dd)
kfree(tmp);
}
-static void __devexit ipath_remove_one(struct pci_dev *pdev)
+static void ipath_remove_one(struct pci_dev *pdev)
{
struct ipath_devdata *dd = pci_get_drvdata(pdev);
diff --git a/drivers/infiniband/hw/ipath/ipath_init_chip.c b/drivers/infiniband/hw/ipath/ipath_init_chip.c
index 49b09c697c7c..be2a60e142b0 100644
--- a/drivers/infiniband/hw/ipath/ipath_init_chip.c
+++ b/drivers/infiniband/hw/ipath/ipath_init_chip.c
@@ -719,16 +719,6 @@ int ipath_init_chip(struct ipath_devdata *dd, int reinit)
goto done;
/*
- * we ignore most issues after reporting them, but have to specially
- * handle hardware-disabled chips.
- */
- if (ret == 2) {
- /* unique error, known to ipath_init_one */
- ret = -EPERM;
- goto done;
- }
-
- /*
* We could bump this to allow for full rcvegrcnt + rcvtidcnt,
* but then it no longer nicely fits power of two, and since
* we now use routines that backend onto __get_free_pages, the
diff --git a/drivers/infiniband/hw/mlx4/cm.c b/drivers/infiniband/hw/mlx4/cm.c
index 80079e5a2e30..dbc99d41605c 100644
--- a/drivers/infiniband/hw/mlx4/cm.c
+++ b/drivers/infiniband/hw/mlx4/cm.c
@@ -268,15 +268,15 @@ static void schedule_delayed(struct ib_device *ibdev, struct id_map_entry *id)
struct mlx4_ib_sriov *sriov = &to_mdev(ibdev)->sriov;
unsigned long flags;
- spin_lock_irqsave(&sriov->going_down_lock, flags);
spin_lock(&sriov->id_map_lock);
+ spin_lock_irqsave(&sriov->going_down_lock, flags);
/*make sure that there is no schedule inside the scheduled work.*/
if (!sriov->is_going_down) {
id->scheduled_delete = 1;
schedule_delayed_work(&id->timeout, CM_CLEANUP_CACHE_TIMEOUT);
}
- spin_unlock(&sriov->id_map_lock);
spin_unlock_irqrestore(&sriov->going_down_lock, flags);
+ spin_unlock(&sriov->id_map_lock);
}
int mlx4_ib_multiplex_cm_handler(struct ib_device *ibdev, int port, int slave_id,
diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index c9eb6a6815ce..ae67df35dd4d 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -66,7 +66,7 @@ static void mlx4_ib_cq_event(struct mlx4_cq *cq, enum mlx4_event type)
static void *get_cqe_from_buf(struct mlx4_ib_cq_buf *buf, int n)
{
- return mlx4_buf_offset(&buf->buf, n * sizeof (struct mlx4_cqe));
+ return mlx4_buf_offset(&buf->buf, n * buf->entry_size);
}
static void *get_cqe(struct mlx4_ib_cq *cq, int n)
@@ -77,8 +77,9 @@ static void *get_cqe(struct mlx4_ib_cq *cq, int n)
static void *get_sw_cqe(struct mlx4_ib_cq *cq, int n)
{
struct mlx4_cqe *cqe = get_cqe(cq, n & cq->ibcq.cqe);
+ struct mlx4_cqe *tcqe = ((cq->buf.entry_size == 64) ? (cqe + 1) : cqe);
- return (!!(cqe->owner_sr_opcode & MLX4_CQE_OWNER_MASK) ^
+ return (!!(tcqe->owner_sr_opcode & MLX4_CQE_OWNER_MASK) ^
!!(n & (cq->ibcq.cqe + 1))) ? NULL : cqe;
}
@@ -99,12 +100,13 @@ static int mlx4_ib_alloc_cq_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq_buf *
{
int err;
- err = mlx4_buf_alloc(dev->dev, nent * sizeof(struct mlx4_cqe),
+ err = mlx4_buf_alloc(dev->dev, nent * dev->dev->caps.cqe_size,
PAGE_SIZE * 2, &buf->buf);
if (err)
goto out;
+ buf->entry_size = dev->dev->caps.cqe_size;
err = mlx4_mtt_init(dev->dev, buf->buf.npages, buf->buf.page_shift,
&buf->mtt);
if (err)
@@ -120,8 +122,7 @@ err_mtt:
mlx4_mtt_cleanup(dev->dev, &buf->mtt);
err_buf:
- mlx4_buf_free(dev->dev, nent * sizeof(struct mlx4_cqe),
- &buf->buf);
+ mlx4_buf_free(dev->dev, nent * buf->entry_size, &buf->buf);
out:
return err;
@@ -129,7 +130,7 @@ out:
static void mlx4_ib_free_cq_buf(struct mlx4_ib_dev *dev, struct mlx4_ib_cq_buf *buf, int cqe)
{
- mlx4_buf_free(dev->dev, (cqe + 1) * sizeof(struct mlx4_cqe), &buf->buf);
+ mlx4_buf_free(dev->dev, (cqe + 1) * buf->entry_size, &buf->buf);
}
static int mlx4_ib_get_cq_umem(struct mlx4_ib_dev *dev, struct ib_ucontext *context,
@@ -137,8 +138,9 @@ static int mlx4_ib_get_cq_umem(struct mlx4_ib_dev *dev, struct ib_ucontext *cont
u64 buf_addr, int cqe)
{
int err;
+ int cqe_size = dev->dev->caps.cqe_size;
- *umem = ib_umem_get(context, buf_addr, cqe * sizeof (struct mlx4_cqe),
+ *umem = ib_umem_get(context, buf_addr, cqe * cqe_size,
IB_ACCESS_LOCAL_WRITE, 1);
if (IS_ERR(*umem))
return PTR_ERR(*umem);
@@ -331,16 +333,23 @@ static void mlx4_ib_cq_resize_copy_cqes(struct mlx4_ib_cq *cq)
{
struct mlx4_cqe *cqe, *new_cqe;
int i;
+ int cqe_size = cq->buf.entry_size;
+ int cqe_inc = cqe_size == 64 ? 1 : 0;
i = cq->mcq.cons_index;
cqe = get_cqe(cq, i & cq->ibcq.cqe);
+ cqe += cqe_inc;
+
while ((cqe->owner_sr_opcode & MLX4_CQE_OPCODE_MASK) != MLX4_CQE_OPCODE_RESIZE) {
new_cqe = get_cqe_from_buf(&cq->resize_buf->buf,
(i + 1) & cq->resize_buf->cqe);
- memcpy(new_cqe, get_cqe(cq, i & cq->ibcq.cqe), sizeof(struct mlx4_cqe));
+ memcpy(new_cqe, get_cqe(cq, i & cq->ibcq.cqe), cqe_size);
+ new_cqe += cqe_inc;
+
new_cqe->owner_sr_opcode = (cqe->owner_sr_opcode & ~MLX4_CQE_OWNER_MASK) |
(((i + 1) & (cq->resize_buf->cqe + 1)) ? MLX4_CQE_OWNER_MASK : 0);
cqe = get_cqe(cq, ++i & cq->ibcq.cqe);
+ cqe += cqe_inc;
}
++cq->mcq.cons_index;
}
@@ -438,6 +447,7 @@ err_buf:
out:
mutex_unlock(&cq->resize_mutex);
+
return err;
}
@@ -586,6 +596,9 @@ repoll:
if (!cqe)
return -EAGAIN;
+ if (cq->buf.entry_size == 64)
+ cqe++;
+
++cq->mcq.cons_index;
/*
@@ -807,6 +820,7 @@ void __mlx4_ib_cq_clean(struct mlx4_ib_cq *cq, u32 qpn, struct mlx4_ib_srq *srq)
int nfreed = 0;
struct mlx4_cqe *cqe, *dest;
u8 owner_bit;
+ int cqe_inc = cq->buf.entry_size == 64 ? 1 : 0;
/*
* First we need to find the current producer index, so we
@@ -825,12 +839,16 @@ void __mlx4_ib_cq_clean(struct mlx4_ib_cq *cq, u32 qpn, struct mlx4_ib_srq *srq)
*/
while ((int) --prod_index - (int) cq->mcq.cons_index >= 0) {
cqe = get_cqe(cq, prod_index & cq->ibcq.cqe);
+ cqe += cqe_inc;
+
if ((be32_to_cpu(cqe->vlan_my_qpn) & MLX4_CQE_QPN_MASK) == qpn) {
if (srq && !(cqe->owner_sr_opcode & MLX4_CQE_IS_SEND_MASK))
mlx4_ib_free_srq_wqe(srq, be16_to_cpu(cqe->wqe_index));
++nfreed;
} else if (nfreed) {
dest = get_cqe(cq, (prod_index + nfreed) & cq->ibcq.cqe);
+ dest += cqe_inc;
+
owner_bit = dest->owner_sr_opcode & MLX4_CQE_OWNER_MASK;
memcpy(dest, cqe, sizeof *cqe);
dest->owner_sr_opcode = owner_bit |
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 718ec6b2bad2..e7d81c0d1ac5 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -563,15 +563,24 @@ static struct ib_ucontext *mlx4_ib_alloc_ucontext(struct ib_device *ibdev,
{
struct mlx4_ib_dev *dev = to_mdev(ibdev);
struct mlx4_ib_ucontext *context;
+ struct mlx4_ib_alloc_ucontext_resp_v3 resp_v3;
struct mlx4_ib_alloc_ucontext_resp resp;
int err;
if (!dev->ib_active)
return ERR_PTR(-EAGAIN);
- resp.qp_tab_size = dev->dev->caps.num_qps;
- resp.bf_reg_size = dev->dev->caps.bf_reg_size;
- resp.bf_regs_per_page = dev->dev->caps.bf_regs_per_page;
+ if (ibdev->uverbs_abi_ver == MLX4_IB_UVERBS_NO_DEV_CAPS_ABI_VERSION) {
+ resp_v3.qp_tab_size = dev->dev->caps.num_qps;
+ resp_v3.bf_reg_size = dev->dev->caps.bf_reg_size;
+ resp_v3.bf_regs_per_page = dev->dev->caps.bf_regs_per_page;
+ } else {
+ resp.dev_caps = dev->dev->caps.userspace_caps;
+ resp.qp_tab_size = dev->dev->caps.num_qps;
+ resp.bf_reg_size = dev->dev->caps.bf_reg_size;
+ resp.bf_regs_per_page = dev->dev->caps.bf_regs_per_page;
+ resp.cqe_size = dev->dev->caps.cqe_size;
+ }
context = kmalloc(sizeof *context, GFP_KERNEL);
if (!context)
@@ -586,7 +595,11 @@ static struct ib_ucontext *mlx4_ib_alloc_ucontext(struct ib_device *ibdev,
INIT_LIST_HEAD(&context->db_page_list);
mutex_init(&context->db_page_mutex);
- err = ib_copy_to_udata(udata, &resp, sizeof resp);
+ if (ibdev->uverbs_abi_ver == MLX4_IB_UVERBS_NO_DEV_CAPS_ABI_VERSION)
+ err = ib_copy_to_udata(udata, &resp_v3, sizeof(resp_v3));
+ else
+ err = ib_copy_to_udata(udata, &resp, sizeof(resp));
+
if (err) {
mlx4_uar_free(to_mdev(ibdev)->dev, &context->uar);
kfree(context);
@@ -1342,7 +1355,11 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
ibdev->ib_dev.num_comp_vectors = dev->caps.num_comp_vectors;
ibdev->ib_dev.dma_device = &dev->pdev->dev;
- ibdev->ib_dev.uverbs_abi_ver = MLX4_IB_UVERBS_ABI_VERSION;
+ if (dev->caps.userspace_caps)
+ ibdev->ib_dev.uverbs_abi_ver = MLX4_IB_UVERBS_ABI_VERSION;
+ else
+ ibdev->ib_dev.uverbs_abi_ver = MLX4_IB_UVERBS_NO_DEV_CAPS_ABI_VERSION;
+
ibdev->ib_dev.uverbs_cmd_mask =
(1ull << IB_USER_VERBS_CMD_GET_CONTEXT) |
(1ull << IB_USER_VERBS_CMD_QUERY_DEVICE) |
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index e04cbc9a54a5..dcd845bc30f0 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -90,6 +90,7 @@ struct mlx4_ib_xrcd {
struct mlx4_ib_cq_buf {
struct mlx4_buf buf;
struct mlx4_mtt mtt;
+ int entry_size;
};
struct mlx4_ib_cq_resize {
diff --git a/drivers/infiniband/hw/mlx4/user.h b/drivers/infiniband/hw/mlx4/user.h
index 13beedeeef9f..07e6769ef43b 100644
--- a/drivers/infiniband/hw/mlx4/user.h
+++ b/drivers/infiniband/hw/mlx4/user.h
@@ -40,7 +40,9 @@
* Increment this value if any changes that break userspace ABI
* compatibility are made.
*/
-#define MLX4_IB_UVERBS_ABI_VERSION 3
+
+#define MLX4_IB_UVERBS_NO_DEV_CAPS_ABI_VERSION 3
+#define MLX4_IB_UVERBS_ABI_VERSION 4
/*
* Make sure that all structs defined in this file remain laid out so
@@ -50,10 +52,18 @@
* instead.
*/
+struct mlx4_ib_alloc_ucontext_resp_v3 {
+ __u32 qp_tab_size;
+ __u16 bf_reg_size;
+ __u16 bf_regs_per_page;
+};
+
struct mlx4_ib_alloc_ucontext_resp {
+ __u32 dev_caps;
__u32 qp_tab_size;
__u16 bf_reg_size;
__u16 bf_regs_per_page;
+ __u32 cqe_size;
};
struct mlx4_ib_alloc_pd_resp {
diff --git a/drivers/infiniband/hw/mthca/mthca_main.c b/drivers/infiniband/hw/mthca/mthca_main.c
index aa12a533ae9e..87897b95666d 100644
--- a/drivers/infiniband/hw/mthca/mthca_main.c
+++ b/drivers/infiniband/hw/mthca/mthca_main.c
@@ -130,7 +130,7 @@ static int log_mtts_per_seg = ilog2(MTHCA_MTT_SEG_SIZE / 8);
module_param_named(log_mtts_per_seg, log_mtts_per_seg, int, 0444);
MODULE_PARM_DESC(log_mtts_per_seg, "Log2 number of MTT entries per segment (1-5)");
-static char mthca_version[] __devinitdata =
+static char mthca_version[] =
DRV_NAME ": Mellanox InfiniBand HCA driver v"
DRV_VERSION " (" DRV_RELDATE ")\n";
@@ -1139,8 +1139,7 @@ int __mthca_restart_one(struct pci_dev *pdev)
return __mthca_init_one(pdev, hca_type);
}
-static int __devinit mthca_init_one(struct pci_dev *pdev,
- const struct pci_device_id *id)
+static int mthca_init_one(struct pci_dev *pdev, const struct pci_device_id *id)
{
int ret;
@@ -1162,7 +1161,7 @@ static int __devinit mthca_init_one(struct pci_dev *pdev,
return ret;
}
-static void __devexit mthca_remove_one(struct pci_dev *pdev)
+static void mthca_remove_one(struct pci_dev *pdev)
{
mutex_lock(&mthca_device_mutex);
__mthca_remove_one(pdev);
@@ -1199,7 +1198,7 @@ static struct pci_driver mthca_driver = {
.name = DRV_NAME,
.id_table = mthca_pci_table,
.probe = mthca_init_one,
- .remove = __devexit_p(mthca_remove_one)
+ .remove = mthca_remove_one,
};
static void __init __mthca_check_profile_val(const char *name, int *pval,
diff --git a/drivers/infiniband/hw/nes/nes.c b/drivers/infiniband/hw/nes/nes.c
index 748db2d3e465..5b152a366dff 100644
--- a/drivers/infiniband/hw/nes/nes.c
+++ b/drivers/infiniband/hw/nes/nes.c
@@ -444,7 +444,7 @@ static irqreturn_t nes_interrupt(int irq, void *dev_id)
/**
* nes_probe - Device initialization
*/
-static int __devinit nes_probe(struct pci_dev *pcidev, const struct pci_device_id *ent)
+static int nes_probe(struct pci_dev *pcidev, const struct pci_device_id *ent)
{
struct net_device *netdev = NULL;
struct nes_device *nesdev = NULL;
@@ -749,7 +749,7 @@ static int __devinit nes_probe(struct pci_dev *pcidev, const struct pci_device_i
/**
* nes_remove - unload from kernel
*/
-static void __devexit nes_remove(struct pci_dev *pcidev)
+static void nes_remove(struct pci_dev *pcidev)
{
struct nes_device *nesdev = pci_get_drvdata(pcidev);
struct net_device *netdev;
@@ -810,7 +810,7 @@ static struct pci_driver nes_pci_driver = {
.name = DRV_NAME,
.id_table = nes_pci_table,
.probe = nes_probe,
- .remove = __devexit_p(nes_remove),
+ .remove = nes_remove,
};
static ssize_t nes_show_adapter(struct device_driver *ddp, char *buf)
diff --git a/drivers/infiniband/hw/nes/nes.h b/drivers/infiniband/hw/nes/nes.h
index 5cac29e6bc1c..33cc58941a3e 100644
--- a/drivers/infiniband/hw/nes/nes.h
+++ b/drivers/infiniband/hw/nes/nes.h
@@ -532,6 +532,7 @@ void nes_iwarp_ce_handler(struct nes_device *, struct nes_hw_cq *);
int nes_destroy_cqp(struct nes_device *);
int nes_nic_cm_xmit(struct sk_buff *, struct net_device *);
void nes_recheck_link_status(struct work_struct *work);
+void nes_terminate_timeout(unsigned long context);
/* nes_nic.c */
struct net_device *nes_netdev_init(struct nes_device *, void __iomem *);
diff --git a/drivers/infiniband/hw/nes/nes_cm.c b/drivers/infiniband/hw/nes/nes_cm.c
index cfaacaf6bf5f..22ea67eea5dc 100644
--- a/drivers/infiniband/hw/nes/nes_cm.c
+++ b/drivers/infiniband/hw/nes/nes_cm.c
@@ -629,11 +629,9 @@ static void build_rdma0_msg(struct nes_cm_node *cm_node, struct nes_qp **nesqp_a
case SEND_RDMA_READ_ZERO:
default:
- if (cm_node->send_rdma0_op != SEND_RDMA_READ_ZERO) {
- printk(KERN_ERR "%s[%u]: Unsupported RDMA0 len operation=%u\n",
- __func__, __LINE__, cm_node->send_rdma0_op);
- WARN_ON(1);
- }
+ if (cm_node->send_rdma0_op != SEND_RDMA_READ_ZERO)
+ WARN(1, "Unsupported RDMA0 len operation=%u\n",
+ cm_node->send_rdma0_op);
nes_debug(NES_DBG_CM, "Sending first rdma operation.\n");
wqe->wqe_words[NES_IWARP_SQ_WQE_MISC_IDX] =
cpu_to_le32(NES_IWARP_SQ_OP_RDMAR);
@@ -671,7 +669,6 @@ int schedule_nes_timer(struct nes_cm_node *cm_node, struct sk_buff *skb,
struct nes_cm_core *cm_core = cm_node->cm_core;
struct nes_timer_entry *new_send;
int ret = 0;
- u32 was_timer_set;
new_send = kzalloc(sizeof(*new_send), GFP_ATOMIC);
if (!new_send)
@@ -723,12 +720,8 @@ int schedule_nes_timer(struct nes_cm_node *cm_node, struct sk_buff *skb,
}
}
- was_timer_set = timer_pending(&cm_core->tcp_timer);
-
- if (!was_timer_set) {
- cm_core->tcp_timer.expires = new_send->timetosend;
- add_timer(&cm_core->tcp_timer);
- }
+ if (!timer_pending(&cm_core->tcp_timer))
+ mod_timer(&cm_core->tcp_timer, new_send->timetosend);
return ret;
}
@@ -946,10 +939,8 @@ static void nes_cm_timer_tick(unsigned long pass)
}
if (settimer) {
- if (!timer_pending(&cm_core->tcp_timer)) {
- cm_core->tcp_timer.expires = nexttimeout;
- add_timer(&cm_core->tcp_timer);
- }
+ if (!timer_pending(&cm_core->tcp_timer))
+ mod_timer(&cm_core->tcp_timer, nexttimeout);
}
}
@@ -1314,8 +1305,6 @@ static int mini_cm_del_listen(struct nes_cm_core *cm_core,
static inline int mini_cm_accelerated(struct nes_cm_core *cm_core,
struct nes_cm_node *cm_node)
{
- u32 was_timer_set;
-
cm_node->accelerated = 1;
if (cm_node->accept_pend) {
@@ -1325,11 +1314,8 @@ static inline int mini_cm_accelerated(struct nes_cm_core *cm_core,
BUG_ON(atomic_read(&cm_node->listener->pend_accepts_cnt) < 0);
}
- was_timer_set = timer_pending(&cm_core->tcp_timer);
- if (!was_timer_set) {
- cm_core->tcp_timer.expires = jiffies + NES_SHORT_TIME;
- add_timer(&cm_core->tcp_timer);
- }
+ if (!timer_pending(&cm_core->tcp_timer))
+ mod_timer(&cm_core->tcp_timer, (jiffies + NES_SHORT_TIME));
return 0;
}
diff --git a/drivers/infiniband/hw/nes/nes_hw.c b/drivers/infiniband/hw/nes/nes_hw.c
index fe7965ee4096..67647e264611 100644
--- a/drivers/infiniband/hw/nes/nes_hw.c
+++ b/drivers/infiniband/hw/nes/nes_hw.c
@@ -75,7 +75,6 @@ static void nes_process_iwarp_aeqe(struct nes_device *nesdev,
static void process_critical_error(struct nes_device *nesdev);
static void nes_process_mac_intr(struct nes_device *nesdev, u32 mac_number);
static unsigned int nes_reset_adapter_ne020(struct nes_device *nesdev, u8 *OneG_Mode);
-static void nes_terminate_timeout(unsigned long context);
static void nes_terminate_start_timer(struct nes_qp *nesqp);
#ifdef CONFIG_INFINIBAND_NES_DEBUG
@@ -3520,7 +3519,7 @@ static void nes_terminate_received(struct nes_device *nesdev,
}
/* Timeout routine in case terminate fails to complete */
-static void nes_terminate_timeout(unsigned long context)
+void nes_terminate_timeout(unsigned long context)
{
struct nes_qp *nesqp = (struct nes_qp *)(unsigned long)context;
@@ -3530,11 +3529,7 @@ static void nes_terminate_timeout(unsigned long context)
/* Set a timer in case hw cannot complete the terminate sequence */
static void nes_terminate_start_timer(struct nes_qp *nesqp)
{
- init_timer(&nesqp->terminate_timer);
- nesqp->terminate_timer.function = nes_terminate_timeout;
- nesqp->terminate_timer.expires = jiffies + HZ;
- nesqp->terminate_timer.data = (unsigned long)nesqp;
- add_timer(&nesqp->terminate_timer);
+ mod_timer(&nesqp->terminate_timer, (jiffies + HZ));
}
/**
diff --git a/drivers/infiniband/hw/nes/nes_mgt.c b/drivers/infiniband/hw/nes/nes_mgt.c
index 3ba7be369452..416645259b0f 100644
--- a/drivers/infiniband/hw/nes/nes_mgt.c
+++ b/drivers/infiniband/hw/nes/nes_mgt.c
@@ -210,6 +210,9 @@ static struct sk_buff *nes_get_next_skb(struct nes_device *nesdev, struct nes_qp
}
while (1) {
+ if (skb_queue_empty(&nesqp->pau_list))
+ goto out;
+
seq = nes_get_seq(skb, ack, wnd, fin_rcvd, rst_rcvd);
if (seq == nextseq) {
if (skb->len || processacks)
@@ -218,14 +221,13 @@ static struct sk_buff *nes_get_next_skb(struct nes_device *nesdev, struct nes_qp
goto out;
}
- if (skb->next == (struct sk_buff *)&nesqp->pau_list)
- goto out;
-
old_skb = skb;
skb = skb->next;
skb_unlink(old_skb, &nesqp->pau_list);
nes_mgt_free_skb(nesdev, old_skb, PCI_DMA_TODEVICE);
nes_rem_ref_cm_node(nesqp->cm_node);
+ if (skb == (struct sk_buff *)&nesqp->pau_list)
+ goto out;
}
return skb;
@@ -245,7 +247,6 @@ static int get_fpdu_info(struct nes_device *nesdev, struct nes_qp *nesqp,
struct nes_rskb_cb *cb;
struct pau_fpdu_info *fpdu_info = NULL;
struct pau_fpdu_frag frags[MAX_FPDU_FRAGS];
- unsigned long flags;
u32 fpdu_len = 0;
u32 tmp_len;
int frag_cnt = 0;
@@ -260,12 +261,10 @@ static int get_fpdu_info(struct nes_device *nesdev, struct nes_qp *nesqp,
*pau_fpdu_info = NULL;
- spin_lock_irqsave(&nesqp->pau_lock, flags);
skb = nes_get_next_skb(nesdev, nesqp, NULL, nesqp->pau_rcv_nxt, &ack, &wnd, &fin_rcvd, &rst_rcvd);
- if (!skb) {
- spin_unlock_irqrestore(&nesqp->pau_lock, flags);
+ if (!skb)
goto out;
- }
+
cb = (struct nes_rskb_cb *)&skb->cb[0];
if (skb->len) {
fpdu_len = be16_to_cpu(*(__be16 *) skb->data) + MPA_FRAMING;
@@ -290,10 +289,9 @@ static int get_fpdu_info(struct nes_device *nesdev, struct nes_qp *nesqp,
skb = nes_get_next_skb(nesdev, nesqp, skb,
nesqp->pau_rcv_nxt + frag_tot, &ack, &wnd, &fin_rcvd, &rst_rcvd);
- if (!skb) {
- spin_unlock_irqrestore(&nesqp->pau_lock, flags);
+ if (!skb)
goto out;
- } else if (rst_rcvd) {
+ if (rst_rcvd) {
/* rst received in the middle of fpdu */
for (; i >= 0; i--) {
skb_unlink(frags[i].skb, &nesqp->pau_list);
@@ -320,8 +318,6 @@ static int get_fpdu_info(struct nes_device *nesdev, struct nes_qp *nesqp,
frag_cnt = 1;
}
- spin_unlock_irqrestore(&nesqp->pau_lock, flags);
-
/* Found one */
fpdu_info = kzalloc(sizeof(*fpdu_info), GFP_ATOMIC);
if (fpdu_info == NULL) {
@@ -383,9 +379,8 @@ static int get_fpdu_info(struct nes_device *nesdev, struct nes_qp *nesqp,
if (frags[i].skb->len == 0) {
/* Pull skb off the list - it will be freed in the callback */
- spin_lock_irqsave(&nesqp->pau_lock, flags);
- skb_unlink(frags[i].skb, &nesqp->pau_list);
- spin_unlock_irqrestore(&nesqp->pau_lock, flags);
+ if (!skb_queue_empty(&nesqp->pau_list))
+ skb_unlink(frags[i].skb, &nesqp->pau_list);
} else {
/* Last skb still has data so update the seq */
iph = (struct iphdr *)(cb->data_start + ETH_HLEN);
@@ -414,14 +409,18 @@ static int forward_fpdus(struct nes_vnic *nesvnic, struct nes_qp *nesqp)
struct pau_fpdu_info *fpdu_info;
struct nes_hw_cqp_wqe *cqp_wqe;
struct nes_cqp_request *cqp_request;
+ unsigned long flags;
u64 u64tmp;
u32 u32tmp;
int rc;
while (1) {
+ spin_lock_irqsave(&nesqp->pau_lock, flags);
rc = get_fpdu_info(nesdev, nesqp, &fpdu_info);
- if (fpdu_info == NULL)
+ if (rc || (fpdu_info == NULL)) {
+ spin_unlock_irqrestore(&nesqp->pau_lock, flags);
return rc;
+ }
cqp_request = fpdu_info->cqp_request;
cqp_wqe = &cqp_request->cqp_wqe;
@@ -447,7 +446,7 @@ static int forward_fpdus(struct nes_vnic *nesvnic, struct nes_qp *nesqp)
set_wqe_32bit_value(cqp_wqe->wqe_words, NES_NIC_SQ_WQE_FRAG0_LOW_IDX,
lower_32_bits(u64tmp));
set_wqe_32bit_value(cqp_wqe->wqe_words, NES_NIC_SQ_WQE_FRAG0_HIGH_IDX,
- upper_32_bits(u64tmp >> 32));
+ upper_32_bits(u64tmp));
set_wqe_32bit_value(cqp_wqe->wqe_words, NES_NIC_SQ_WQE_FRAG1_LOW_IDX,
lower_32_bits(fpdu_info->frags[0].physaddr));
@@ -475,6 +474,7 @@ static int forward_fpdus(struct nes_vnic *nesvnic, struct nes_qp *nesqp)
atomic_set(&cqp_request->refcount, 1);
nes_post_cqp_request(nesdev, cqp_request);
+ spin_unlock_irqrestore(&nesqp->pau_lock, flags);
}
return 0;
@@ -649,11 +649,9 @@ static void nes_chg_qh_handler(struct nes_device *nesdev, struct nes_cqp_request
nesqp = qh_chg->nesqp;
/* Should we handle the bad completion */
- if (cqp_request->major_code) {
- printk(KERN_ERR PFX "Invalid cqp_request major_code=0x%x\n",
+ if (cqp_request->major_code)
+ WARN(1, PFX "Invalid cqp_request major_code=0x%x\n",
cqp_request->major_code);
- WARN_ON(1);
- }
switch (nesqp->pau_state) {
case PAU_DEL_QH:
diff --git a/drivers/infiniband/hw/nes/nes_nic.c b/drivers/infiniband/hw/nes/nes_nic.c
index 0564be757d82..9542e1644a5c 100644
--- a/drivers/infiniband/hw/nes/nes_nic.c
+++ b/drivers/infiniband/hw/nes/nes_nic.c
@@ -944,12 +944,13 @@ static void nes_netdev_set_multicast_list(struct net_device *netdev)
addr,
perfect_filter_register_address+(mc_index * 8),
mc_nic_index);
- macaddr_high = ((u16) addr[0]) << 8;
- macaddr_high += (u16) addr[1];
- macaddr_low = ((u32) addr[2]) << 24;
- macaddr_low += ((u32) addr[3]) << 16;
- macaddr_low += ((u32) addr[4]) << 8;
- macaddr_low += (u32) addr[5];
+ macaddr_high = ((u8) addr[0]) << 8;
+ macaddr_high += (u8) addr[1];
+ macaddr_low = ((u8) addr[2]) << 24;
+ macaddr_low += ((u8) addr[3]) << 16;
+ macaddr_low += ((u8) addr[4]) << 8;
+ macaddr_low += (u8) addr[5];
+
nes_write_indexed(nesdev,
perfect_filter_register_address+(mc_index * 8),
macaddr_low);
diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c
index cd0ecb215cca..07e4fbad987a 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.c
+++ b/drivers/infiniband/hw/nes/nes_verbs.c
@@ -1404,6 +1404,9 @@ static struct ib_qp *nes_create_qp(struct ib_pd *ibpd,
}
nesqp->sig_all = (init_attr->sq_sig_type == IB_SIGNAL_ALL_WR);
+ init_timer(&nesqp->terminate_timer);
+ nesqp->terminate_timer.function = nes_terminate_timeout;
+ nesqp->terminate_timer.data = (unsigned long)nesqp;
/* update the QP table */
nesdev->nesadapter->qp_table[nesqp->hwqp.qp_id-NES_FIRST_QPN] = nesqp;
@@ -1413,7 +1416,6 @@ static struct ib_qp *nes_create_qp(struct ib_pd *ibpd,
return &nesqp->ibqp;
}
-
/**
* nes_clean_cq
*/
@@ -2559,6 +2561,11 @@ static struct ib_mr *nes_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
return ibmr;
case IWNES_MEMREG_TYPE_QP:
case IWNES_MEMREG_TYPE_CQ:
+ if (!region->length) {
+ nes_debug(NES_DBG_MR, "Unable to register zero length region for CQ\n");
+ ib_umem_release(region);
+ return ERR_PTR(-EINVAL);
+ }
nespbl = kzalloc(sizeof(*nespbl), GFP_KERNEL);
if (!nespbl) {
nes_debug(NES_DBG_MR, "Unable to allocate PBL\n");
diff --git a/drivers/infiniband/hw/qib/qib_init.c b/drivers/infiniband/hw/qib/qib_init.c
index 4443adfcd9ee..ddf066d9abb6 100644
--- a/drivers/infiniband/hw/qib/qib_init.c
+++ b/drivers/infiniband/hw/qib/qib_init.c
@@ -1134,9 +1134,8 @@ void qib_disable_after_error(struct qib_devdata *dd)
*dd->devstatusp |= QIB_STATUS_HWERROR;
}
-static void __devexit qib_remove_one(struct pci_dev *);
-static int __devinit qib_init_one(struct pci_dev *,
- const struct pci_device_id *);
+static void qib_remove_one(struct pci_dev *);
+static int qib_init_one(struct pci_dev *, const struct pci_device_id *);
#define DRIVER_LOAD_MSG "QLogic " QIB_DRV_NAME " loaded: "
#define PFX QIB_DRV_NAME ": "
@@ -1153,7 +1152,7 @@ MODULE_DEVICE_TABLE(pci, qib_pci_tbl);
struct pci_driver qib_driver = {
.name = QIB_DRV_NAME,
.probe = qib_init_one,
- .remove = __devexit_p(qib_remove_one),
+ .remove = qib_remove_one,
.id_table = qib_pci_tbl,
.err_handler = &qib_pci_err_handler,
};
@@ -1342,8 +1341,7 @@ static void qib_postinit_cleanup(struct qib_devdata *dd)
qib_free_devdata(dd);
}
-static int __devinit qib_init_one(struct pci_dev *pdev,
- const struct pci_device_id *ent)
+static int qib_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
{
int ret, j, pidx, initfail;
struct qib_devdata *dd = NULL;
@@ -1448,7 +1446,7 @@ bail:
return ret;
}
-static void __devexit qib_remove_one(struct pci_dev *pdev)
+static void qib_remove_one(struct pci_dev *pdev)
{
struct qib_devdata *dd = pci_get_drvdata(pdev);
int ret;
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))
diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c
index 922d845f76b0..d5088ce78290 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.c
+++ b/drivers/infiniband/ulp/srp/ib_srp.c
@@ -222,27 +222,29 @@ static int srp_new_cm_id(struct srp_target_port *target)
static int srp_create_target_ib(struct srp_target_port *target)
{
struct ib_qp_init_attr *init_attr;
+ struct ib_cq *recv_cq, *send_cq;
+ struct ib_qp *qp;
int ret;
init_attr = kzalloc(sizeof *init_attr, GFP_KERNEL);
if (!init_attr)
return -ENOMEM;
- target->recv_cq = ib_create_cq(target->srp_host->srp_dev->dev,
- srp_recv_completion, NULL, target, SRP_RQ_SIZE, 0);
- if (IS_ERR(target->recv_cq)) {
- ret = PTR_ERR(target->recv_cq);
+ recv_cq = ib_create_cq(target->srp_host->srp_dev->dev,
+ srp_recv_completion, NULL, target, SRP_RQ_SIZE, 0);
+ if (IS_ERR(recv_cq)) {
+ ret = PTR_ERR(recv_cq);
goto err;
}
- target->send_cq = ib_create_cq(target->srp_host->srp_dev->dev,
- srp_send_completion, NULL, target, SRP_SQ_SIZE, 0);
- if (IS_ERR(target->send_cq)) {
- ret = PTR_ERR(target->send_cq);
+ send_cq = ib_create_cq(target->srp_host->srp_dev->dev,
+ srp_send_completion, NULL, target, SRP_SQ_SIZE, 0);
+ if (IS_ERR(send_cq)) {
+ ret = PTR_ERR(send_cq);
goto err_recv_cq;
}
- ib_req_notify_cq(target->recv_cq, IB_CQ_NEXT_COMP);
+ ib_req_notify_cq(recv_cq, IB_CQ_NEXT_COMP);
init_attr->event_handler = srp_qp_event;
init_attr->cap.max_send_wr = SRP_SQ_SIZE;
@@ -251,30 +253,41 @@ static int srp_create_target_ib(struct srp_target_port *target)
init_attr->cap.max_send_sge = 1;
init_attr->sq_sig_type = IB_SIGNAL_ALL_WR;
init_attr->qp_type = IB_QPT_RC;
- init_attr->send_cq = target->send_cq;
- init_attr->recv_cq = target->recv_cq;
+ init_attr->send_cq = send_cq;
+ init_attr->recv_cq = recv_cq;
- target->qp = ib_create_qp(target->srp_host->srp_dev->pd, init_attr);
- if (IS_ERR(target->qp)) {
- ret = PTR_ERR(target->qp);
+ qp = ib_create_qp(target->srp_host->srp_dev->pd, init_attr);
+ if (IS_ERR(qp)) {
+ ret = PTR_ERR(qp);
goto err_send_cq;
}
- ret = srp_init_qp(target, target->qp);
+ ret = srp_init_qp(target, qp);
if (ret)
goto err_qp;
+ if (target->qp)
+ ib_destroy_qp(target->qp);
+ if (target->recv_cq)
+ ib_destroy_cq(target->recv_cq);
+ if (target->send_cq)
+ ib_destroy_cq(target->send_cq);
+
+ target->qp = qp;
+ target->recv_cq = recv_cq;
+ target->send_cq = send_cq;
+
kfree(init_attr);
return 0;
err_qp:
- ib_destroy_qp(target->qp);
+ ib_destroy_qp(qp);
err_send_cq:
- ib_destroy_cq(target->send_cq);
+ ib_destroy_cq(send_cq);
err_recv_cq:
- ib_destroy_cq(target->recv_cq);
+ ib_destroy_cq(recv_cq);
err:
kfree(init_attr);
@@ -289,6 +302,9 @@ static void srp_free_target_ib(struct srp_target_port *target)
ib_destroy_cq(target->send_cq);
ib_destroy_cq(target->recv_cq);
+ target->qp = NULL;
+ target->send_cq = target->recv_cq = NULL;
+
for (i = 0; i < SRP_RQ_SIZE; ++i)
srp_free_iu(target->srp_host, target->rx_ring[i]);
for (i = 0; i < SRP_SQ_SIZE; ++i)
@@ -428,34 +444,50 @@ static int srp_send_req(struct srp_target_port *target)
return status;
}
-static void srp_disconnect_target(struct srp_target_port *target)
+static bool srp_queue_remove_work(struct srp_target_port *target)
{
- /* XXX should send SRP_I_LOGOUT request */
+ bool changed = false;
- init_completion(&target->done);
- if (ib_send_cm_dreq(target->cm_id, NULL, 0)) {
- shost_printk(KERN_DEBUG, target->scsi_host,
- PFX "Sending CM DREQ failed\n");
- return;
+ spin_lock_irq(&target->lock);
+ if (target->state != SRP_TARGET_REMOVED) {
+ target->state = SRP_TARGET_REMOVED;
+ changed = true;
}
- wait_for_completion(&target->done);
+ spin_unlock_irq(&target->lock);
+
+ if (changed)
+ queue_work(system_long_wq, &target->remove_work);
+
+ return changed;
}
-static bool srp_change_state(struct srp_target_port *target,
- enum srp_target_state old,
- enum srp_target_state new)
+static bool srp_change_conn_state(struct srp_target_port *target,
+ bool connected)
{
bool changed = false;
spin_lock_irq(&target->lock);
- if (target->state == old) {
- target->state = new;
+ if (target->connected != connected) {
+ target->connected = connected;
changed = true;
}
spin_unlock_irq(&target->lock);
+
return changed;
}
+static void srp_disconnect_target(struct srp_target_port *target)
+{
+ if (srp_change_conn_state(target, false)) {
+ /* XXX should send SRP_I_LOGOUT request */
+
+ if (ib_send_cm_dreq(target->cm_id, NULL, 0)) {
+ shost_printk(KERN_DEBUG, target->scsi_host,
+ PFX "Sending CM DREQ failed\n");
+ }
+ }
+}
+
static void srp_free_req_data(struct srp_target_port *target)
{
struct ib_device *ibdev = target->srp_host->srp_dev->dev;
@@ -489,32 +521,50 @@ static void srp_del_scsi_host_attr(struct Scsi_Host *shost)
device_remove_file(&shost->shost_dev, *attr);
}
-static void srp_remove_work(struct work_struct *work)
+static void srp_remove_target(struct srp_target_port *target)
{
- struct srp_target_port *target =
- container_of(work, struct srp_target_port, work);
-
- if (!srp_change_state(target, SRP_TARGET_DEAD, SRP_TARGET_REMOVED))
- return;
-
- spin_lock(&target->srp_host->target_lock);
- list_del(&target->list);
- spin_unlock(&target->srp_host->target_lock);
+ WARN_ON_ONCE(target->state != SRP_TARGET_REMOVED);
srp_del_scsi_host_attr(target->scsi_host);
srp_remove_host(target->scsi_host);
scsi_remove_host(target->scsi_host);
+ srp_disconnect_target(target);
ib_destroy_cm_id(target->cm_id);
srp_free_target_ib(target);
srp_free_req_data(target);
scsi_host_put(target->scsi_host);
}
+static void srp_remove_work(struct work_struct *work)
+{
+ struct srp_target_port *target =
+ container_of(work, struct srp_target_port, remove_work);
+
+ WARN_ON_ONCE(target->state != SRP_TARGET_REMOVED);
+
+ spin_lock(&target->srp_host->target_lock);
+ list_del(&target->list);
+ spin_unlock(&target->srp_host->target_lock);
+
+ srp_remove_target(target);
+}
+
+static void srp_rport_delete(struct srp_rport *rport)
+{
+ struct srp_target_port *target = rport->lld_data;
+
+ srp_queue_remove_work(target);
+}
+
static int srp_connect_target(struct srp_target_port *target)
{
int retries = 3;
int ret;
+ WARN_ON_ONCE(target->connected);
+
+ target->qp_in_error = false;
+
ret = srp_lookup_path(target);
if (ret)
return ret;
@@ -534,6 +584,7 @@ static int srp_connect_target(struct srp_target_port *target)
*/
switch (target->status) {
case 0:
+ srp_change_conn_state(target, true);
return 0;
case SRP_PORT_REDIRECT:
@@ -646,13 +697,14 @@ static void srp_reset_req(struct srp_target_port *target, struct srp_request *re
static int srp_reconnect_target(struct srp_target_port *target)
{
- struct ib_qp_attr qp_attr;
- struct ib_wc wc;
+ struct Scsi_Host *shost = target->scsi_host;
int i, ret;
- if (!srp_change_state(target, SRP_TARGET_LIVE, SRP_TARGET_CONNECTING))
+ if (target->state != SRP_TARGET_LIVE)
return -EAGAIN;
+ scsi_target_block(&shost->shost_gendev);
+
srp_disconnect_target(target);
/*
* Now get a new local CM ID so that we avoid confusing the
@@ -660,21 +712,11 @@ static int srp_reconnect_target(struct srp_target_port *target)
*/
ret = srp_new_cm_id(target);
if (ret)
- goto err;
+ goto unblock;
- qp_attr.qp_state = IB_QPS_RESET;
- ret = ib_modify_qp(target->qp, &qp_attr, IB_QP_STATE);
- if (ret)
- goto err;
-
- ret = srp_init_qp(target, target->qp);
+ ret = srp_create_target_ib(target);
if (ret)
- goto err;
-
- while (ib_poll_cq(target->recv_cq, 1, &wc) > 0)
- ; /* nothing */
- while (ib_poll_cq(target->send_cq, 1, &wc) > 0)
- ; /* nothing */
+ goto unblock;
for (i = 0; i < SRP_CMD_SQ_SIZE; ++i) {
struct srp_request *req = &target->req_ring[i];
@@ -686,13 +728,16 @@ static int srp_reconnect_target(struct srp_target_port *target)
for (i = 0; i < SRP_SQ_SIZE; ++i)
list_add(&target->tx_ring[i]->list, &target->free_tx);
- target->qp_in_error = 0;
ret = srp_connect_target(target);
+
+unblock:
+ scsi_target_unblock(&shost->shost_gendev, ret == 0 ? SDEV_RUNNING :
+ SDEV_TRANSPORT_OFFLINE);
+
if (ret)
goto err;
- if (!srp_change_state(target, SRP_TARGET_CONNECTING, SRP_TARGET_LIVE))
- ret = -EAGAIN;
+ shost_printk(KERN_INFO, target->scsi_host, PFX "reconnect succeeded\n");
return ret;
@@ -705,17 +750,8 @@ err:
* However, we have to defer the real removal because we
* are in the context of the SCSI error handler now, which
* will deadlock if we call scsi_remove_host().
- *
- * Schedule our work inside the lock to avoid a race with
- * the flush_scheduled_work() in srp_remove_one().
*/
- spin_lock_irq(&target->lock);
- if (target->state == SRP_TARGET_CONNECTING) {
- target->state = SRP_TARGET_DEAD;
- INIT_WORK(&target->work, srp_remove_work);
- queue_work(ib_wq, &target->work);
- }
- spin_unlock_irq(&target->lock);
+ srp_queue_remove_work(target);
return ret;
}
@@ -1262,6 +1298,19 @@ static void srp_handle_recv(struct srp_target_port *target, struct ib_wc *wc)
PFX "Recv failed with error code %d\n", res);
}
+static void srp_handle_qp_err(enum ib_wc_status wc_status,
+ enum ib_wc_opcode wc_opcode,
+ struct srp_target_port *target)
+{
+ if (target->connected && !target->qp_in_error) {
+ shost_printk(KERN_ERR, target->scsi_host,
+ PFX "failed %s status %d\n",
+ wc_opcode & IB_WC_RECV ? "receive" : "send",
+ wc_status);
+ }
+ target->qp_in_error = true;
+}
+
static void srp_recv_completion(struct ib_cq *cq, void *target_ptr)
{
struct srp_target_port *target = target_ptr;
@@ -1269,15 +1318,11 @@ static void srp_recv_completion(struct ib_cq *cq, void *target_ptr)
ib_req_notify_cq(cq, IB_CQ_NEXT_COMP);
while (ib_poll_cq(cq, 1, &wc) > 0) {
- if (wc.status) {
- shost_printk(KERN_ERR, target->scsi_host,
- PFX "failed receive status %d\n",
- wc.status);
- target->qp_in_error = 1;
- break;
+ if (likely(wc.status == IB_WC_SUCCESS)) {
+ srp_handle_recv(target, &wc);
+ } else {
+ srp_handle_qp_err(wc.status, wc.opcode, target);
}
-
- srp_handle_recv(target, &wc);
}
}
@@ -1288,16 +1333,12 @@ static void srp_send_completion(struct ib_cq *cq, void *target_ptr)
struct srp_iu *iu;
while (ib_poll_cq(cq, 1, &wc) > 0) {
- if (wc.status) {
- shost_printk(KERN_ERR, target->scsi_host,
- PFX "failed send status %d\n",
- wc.status);
- target->qp_in_error = 1;
- break;
+ if (likely(wc.status == IB_WC_SUCCESS)) {
+ iu = (struct srp_iu *) (uintptr_t) wc.wr_id;
+ list_add(&iu->list, &target->free_tx);
+ } else {
+ srp_handle_qp_err(wc.status, wc.opcode, target);
}
-
- iu = (struct srp_iu *) (uintptr_t) wc.wr_id;
- list_add(&iu->list, &target->free_tx);
}
}
@@ -1311,16 +1352,6 @@ static int srp_queuecommand(struct Scsi_Host *shost, struct scsi_cmnd *scmnd)
unsigned long flags;
int len;
- if (target->state == SRP_TARGET_CONNECTING)
- goto err;
-
- if (target->state == SRP_TARGET_DEAD ||
- target->state == SRP_TARGET_REMOVED) {
- scmnd->result = DID_BAD_TARGET << 16;
- scmnd->scsi_done(scmnd);
- return 0;
- }
-
spin_lock_irqsave(&target->lock, flags);
iu = __srp_get_tx_iu(target, SRP_IU_CMD);
if (!iu)
@@ -1377,7 +1408,6 @@ err_iu:
err_unlock:
spin_unlock_irqrestore(&target->lock, flags);
-err:
return SCSI_MLQUEUE_HOST_BUSY;
}
@@ -1419,6 +1449,33 @@ err:
return -ENOMEM;
}
+static uint32_t srp_compute_rq_tmo(struct ib_qp_attr *qp_attr, int attr_mask)
+{
+ uint64_t T_tr_ns, max_compl_time_ms;
+ uint32_t rq_tmo_jiffies;
+
+ /*
+ * According to section 11.2.4.2 in the IBTA spec (Modify Queue Pair,
+ * table 91), both the QP timeout and the retry count have to be set
+ * for RC QP's during the RTR to RTS transition.
+ */
+ WARN_ON_ONCE((attr_mask & (IB_QP_TIMEOUT | IB_QP_RETRY_CNT)) !=
+ (IB_QP_TIMEOUT | IB_QP_RETRY_CNT));
+
+ /*
+ * Set target->rq_tmo_jiffies to one second more than the largest time
+ * it can take before an error completion is generated. See also
+ * C9-140..142 in the IBTA spec for more information about how to
+ * convert the QP Local ACK Timeout value to nanoseconds.
+ */
+ T_tr_ns = 4096 * (1ULL << qp_attr->timeout);
+ max_compl_time_ms = qp_attr->retry_cnt * 4 * T_tr_ns;
+ do_div(max_compl_time_ms, NSEC_PER_MSEC);
+ rq_tmo_jiffies = msecs_to_jiffies(max_compl_time_ms + 1000);
+
+ return rq_tmo_jiffies;
+}
+
static void srp_cm_rep_handler(struct ib_cm_id *cm_id,
struct srp_login_rsp *lrsp,
struct srp_target_port *target)
@@ -1478,6 +1535,8 @@ static void srp_cm_rep_handler(struct ib_cm_id *cm_id,
if (ret)
goto error_free;
+ target->rq_tmo_jiffies = srp_compute_rq_tmo(qp_attr, attr_mask);
+
ret = ib_modify_qp(target->qp, qp_attr, attr_mask);
if (ret)
goto error_free;
@@ -1599,6 +1658,7 @@ static int srp_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event)
case IB_CM_DREQ_RECEIVED:
shost_printk(KERN_WARNING, target->scsi_host,
PFX "DREQ received - connection closed\n");
+ srp_change_conn_state(target, false);
if (ib_send_cm_drep(cm_id, NULL, 0))
shost_printk(KERN_ERR, target->scsi_host,
PFX "Sending CM DREP failed\n");
@@ -1608,7 +1668,6 @@ static int srp_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event)
shost_printk(KERN_ERR, target->scsi_host,
PFX "connection closed\n");
- comp = 1;
target->status = 0;
break;
@@ -1636,10 +1695,6 @@ static int srp_send_tsk_mgmt(struct srp_target_port *target,
struct srp_iu *iu;
struct srp_tsk_mgmt *tsk_mgmt;
- if (target->state == SRP_TARGET_DEAD ||
- target->state == SRP_TARGET_REMOVED)
- return -1;
-
init_completion(&target->tsk_mgmt_done);
spin_lock_irq(&target->lock);
@@ -1729,6 +1784,21 @@ static int srp_reset_host(struct scsi_cmnd *scmnd)
return ret;
}
+static int srp_slave_configure(struct scsi_device *sdev)
+{
+ struct Scsi_Host *shost = sdev->host;
+ struct srp_target_port *target = host_to_target(shost);
+ struct request_queue *q = sdev->request_queue;
+ unsigned long timeout;
+
+ if (sdev->type == TYPE_DISK) {
+ timeout = max_t(unsigned, 30 * HZ, target->rq_tmo_jiffies);
+ blk_queue_rq_timeout(q, timeout);
+ }
+
+ return 0;
+}
+
static ssize_t show_id_ext(struct device *dev, struct device_attribute *attr,
char *buf)
{
@@ -1861,6 +1931,7 @@ static struct scsi_host_template srp_template = {
.module = THIS_MODULE,
.name = "InfiniBand SRP initiator",
.proc_name = DRV_NAME,
+ .slave_configure = srp_slave_configure,
.info = srp_target_info,
.queuecommand = srp_queuecommand,
.eh_abort_handler = srp_abort,
@@ -1894,11 +1965,14 @@ static int srp_add_target(struct srp_host *host, struct srp_target_port *target)
return PTR_ERR(rport);
}
+ rport->lld_data = target;
+
spin_lock(&host->target_lock);
list_add_tail(&target->list, &host->target_list);
spin_unlock(&host->target_lock);
target->state = SRP_TARGET_LIVE;
+ target->connected = false;
scsi_scan_target(&target->scsi_host->shost_gendev,
0, target->scsi_id, SCAN_WILD_CARD, 0);
@@ -2188,6 +2262,7 @@ static ssize_t srp_create_target(struct device *dev,
sizeof (struct srp_indirect_buf) +
target->cmd_sg_cnt * sizeof (struct srp_direct_buf);
+ INIT_WORK(&target->remove_work, srp_remove_work);
spin_lock_init(&target->lock);
INIT_LIST_HEAD(&target->free_tx);
INIT_LIST_HEAD(&target->free_reqs);
@@ -2232,7 +2307,6 @@ static ssize_t srp_create_target(struct device *dev,
if (ret)
goto err_free_ib;
- target->qp_in_error = 0;
ret = srp_connect_target(target);
if (ret) {
shost_printk(KERN_ERR, target->scsi_host,
@@ -2422,8 +2496,7 @@ static void srp_remove_one(struct ib_device *device)
{
struct srp_device *srp_dev;
struct srp_host *host, *tmp_host;
- LIST_HEAD(target_list);
- struct srp_target_port *target, *tmp_target;
+ struct srp_target_port *target;
srp_dev = ib_get_client_data(device, &srp_client);
@@ -2436,35 +2509,17 @@ static void srp_remove_one(struct ib_device *device)
wait_for_completion(&host->released);
/*
- * Mark all target ports as removed, so we stop queueing
- * commands and don't try to reconnect.
+ * Remove all target ports.
*/
spin_lock(&host->target_lock);
- list_for_each_entry(target, &host->target_list, list) {
- spin_lock_irq(&target->lock);
- target->state = SRP_TARGET_REMOVED;
- spin_unlock_irq(&target->lock);
- }
+ list_for_each_entry(target, &host->target_list, list)
+ srp_queue_remove_work(target);
spin_unlock(&host->target_lock);
/*
- * Wait for any reconnection tasks that may have
- * started before we marked our target ports as
- * removed, and any target port removal tasks.
+ * Wait for target port removal tasks.
*/
- flush_workqueue(ib_wq);
-
- list_for_each_entry_safe(target, tmp_target,
- &host->target_list, list) {
- srp_del_scsi_host_attr(target->scsi_host);
- srp_remove_host(target->scsi_host);
- scsi_remove_host(target->scsi_host);
- srp_disconnect_target(target);
- ib_destroy_cm_id(target->cm_id);
- srp_free_target_ib(target);
- srp_free_req_data(target);
- scsi_host_put(target->scsi_host);
- }
+ flush_workqueue(system_long_wq);
kfree(host);
}
@@ -2478,6 +2533,7 @@ static void srp_remove_one(struct ib_device *device)
}
static struct srp_function_template ib_srp_transport_functions = {
+ .rport_delete = srp_rport_delete,
};
static int __init srp_init_module(void)
diff --git a/drivers/infiniband/ulp/srp/ib_srp.h b/drivers/infiniband/ulp/srp/ib_srp.h
index 020caf0c3789..de2d0b3c0bfe 100644
--- a/drivers/infiniband/ulp/srp/ib_srp.h
+++ b/drivers/infiniband/ulp/srp/ib_srp.h
@@ -80,9 +80,7 @@ enum {
enum srp_target_state {
SRP_TARGET_LIVE,
- SRP_TARGET_CONNECTING,
- SRP_TARGET_DEAD,
- SRP_TARGET_REMOVED
+ SRP_TARGET_REMOVED,
};
enum srp_iu_type {
@@ -163,6 +161,9 @@ struct srp_target_port {
struct ib_sa_query *path_query;
int path_query_id;
+ u32 rq_tmo_jiffies;
+ bool connected;
+
struct ib_cm_id *cm_id;
int max_ti_iu_len;
@@ -173,12 +174,12 @@ struct srp_target_port {
struct srp_iu *rx_ring[SRP_RQ_SIZE];
struct srp_request req_ring[SRP_CMD_SQ_SIZE];
- struct work_struct work;
+ struct work_struct remove_work;
struct list_head list;
struct completion done;
int status;
- int qp_in_error;
+ bool qp_in_error;
struct completion tsk_mgmt_done;
u8 tsk_mgmt_status;
diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c
index cf23c46185b2..c09d41b1a2ff 100644
--- a/drivers/infiniband/ulp/srpt/ib_srpt.c
+++ b/drivers/infiniband/ulp/srpt/ib_srpt.c
@@ -1269,7 +1269,6 @@ static struct srpt_send_ioctx *srpt_get_send_ioctx(struct srpt_rdma_ch *ch)
return ioctx;
BUG_ON(ioctx->ch != ch);
- kref_init(&ioctx->kref);
spin_lock_init(&ioctx->spinlock);
ioctx->state = SRPT_STATE_NEW;
ioctx->n_rbuf = 0;
@@ -1291,39 +1290,6 @@ static struct srpt_send_ioctx *srpt_get_send_ioctx(struct srpt_rdma_ch *ch)
}
/**
- * srpt_put_send_ioctx() - Free up resources.
- */
-static void srpt_put_send_ioctx(struct srpt_send_ioctx *ioctx)
-{
- struct srpt_rdma_ch *ch;
- unsigned long flags;
-
- BUG_ON(!ioctx);
- ch = ioctx->ch;
- BUG_ON(!ch);
-
- WARN_ON(srpt_get_cmd_state(ioctx) != SRPT_STATE_DONE);
-
- srpt_unmap_sg_to_ib_sge(ioctx->ch, ioctx);
- transport_generic_free_cmd(&ioctx->cmd, 0);
-
- if (ioctx->n_rbuf > 1) {
- kfree(ioctx->rbufs);
- ioctx->rbufs = NULL;
- ioctx->n_rbuf = 0;
- }
-
- spin_lock_irqsave(&ch->spinlock, flags);
- list_add(&ioctx->free_list, &ch->free_list);
- spin_unlock_irqrestore(&ch->spinlock, flags);
-}
-
-static void srpt_put_send_ioctx_kref(struct kref *kref)
-{
- srpt_put_send_ioctx(container_of(kref, struct srpt_send_ioctx, kref));
-}
-
-/**
* srpt_abort_cmd() - Abort a SCSI command.
* @ioctx: I/O context associated with the SCSI command.
* @context: Preferred execution context.
@@ -1359,8 +1325,14 @@ static int srpt_abort_cmd(struct srpt_send_ioctx *ioctx)
}
spin_unlock_irqrestore(&ioctx->spinlock, flags);
- if (state == SRPT_STATE_DONE)
+ if (state == SRPT_STATE_DONE) {
+ struct srpt_rdma_ch *ch = ioctx->ch;
+
+ BUG_ON(ch->sess == NULL);
+
+ target_put_sess_cmd(ch->sess, &ioctx->cmd);
goto out;
+ }
pr_debug("Aborting cmd with state %d and tag %lld\n", state,
ioctx->tag);
@@ -1395,11 +1367,11 @@ static int srpt_abort_cmd(struct srpt_send_ioctx *ioctx)
spin_lock_irqsave(&ioctx->cmd.t_state_lock, flags);
ioctx->cmd.transport_state |= CMD_T_LUN_STOP;
spin_unlock_irqrestore(&ioctx->cmd.t_state_lock, flags);
- kref_put(&ioctx->kref, srpt_put_send_ioctx_kref);
+ target_put_sess_cmd(ioctx->ch->sess, &ioctx->cmd);
break;
case SRPT_STATE_MGMT_RSP_SENT:
srpt_set_cmd_state(ioctx, SRPT_STATE_DONE);
- kref_put(&ioctx->kref, srpt_put_send_ioctx_kref);
+ target_put_sess_cmd(ioctx->ch->sess, &ioctx->cmd);
break;
default:
WARN_ON("ERROR: unexpected command state");
@@ -1457,11 +1429,13 @@ static void srpt_handle_send_comp(struct srpt_rdma_ch *ch,
&& state != SRPT_STATE_DONE))
pr_debug("state = %d\n", state);
- if (state != SRPT_STATE_DONE)
- kref_put(&ioctx->kref, srpt_put_send_ioctx_kref);
- else
+ if (state != SRPT_STATE_DONE) {
+ srpt_unmap_sg_to_ib_sge(ch, ioctx);
+ transport_generic_free_cmd(&ioctx->cmd, 0);
+ } else {
printk(KERN_ERR "IB completion has been received too late for"
" wr_id = %u.\n", ioctx->ioctx.index);
+ }
}
/**
@@ -1712,10 +1686,10 @@ out_err:
static int srpt_check_stop_free(struct se_cmd *cmd)
{
- struct srpt_send_ioctx *ioctx;
+ struct srpt_send_ioctx *ioctx = container_of(cmd,
+ struct srpt_send_ioctx, cmd);
- ioctx = container_of(cmd, struct srpt_send_ioctx, cmd);
- return kref_put(&ioctx->kref, srpt_put_send_ioctx_kref);
+ return target_put_sess_cmd(ioctx->ch->sess, &ioctx->cmd);
}
/**
@@ -1730,12 +1704,12 @@ static int srpt_handle_cmd(struct srpt_rdma_ch *ch,
uint64_t unpacked_lun;
u64 data_len;
enum dma_data_direction dir;
- int ret;
+ sense_reason_t ret;
+ int rc;
BUG_ON(!send_ioctx);
srp_cmd = recv_ioctx->ioctx.buf;
- kref_get(&send_ioctx->kref);
cmd = &send_ioctx->cmd;
send_ioctx->tag = srp_cmd->tag;
@@ -1755,40 +1729,26 @@ static int srpt_handle_cmd(struct srpt_rdma_ch *ch,
break;
}
- ret = srpt_get_desc_tbl(send_ioctx, srp_cmd, &dir, &data_len);
- if (ret) {
+ if (srpt_get_desc_tbl(send_ioctx, srp_cmd, &dir, &data_len)) {
printk(KERN_ERR "0x%llx: parsing SRP descriptor table failed.\n",
srp_cmd->tag);
- cmd->se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- cmd->scsi_sense_reason = TCM_INVALID_CDB_FIELD;
- kref_put(&send_ioctx->kref, srpt_put_send_ioctx_kref);
+ ret = TCM_INVALID_CDB_FIELD;
goto send_sense;
}
- cmd->data_length = data_len;
- cmd->data_direction = dir;
unpacked_lun = srpt_unpack_lun((uint8_t *)&srp_cmd->lun,
sizeof(srp_cmd->lun));
- if (transport_lookup_cmd_lun(cmd, unpacked_lun) < 0) {
- kref_put(&send_ioctx->kref, srpt_put_send_ioctx_kref);
+ rc = target_submit_cmd(cmd, ch->sess, srp_cmd->cdb,
+ &send_ioctx->sense_data[0], unpacked_lun, data_len,
+ MSG_SIMPLE_TAG, dir, TARGET_SCF_ACK_KREF);
+ if (rc != 0) {
+ ret = TCM_LOGICAL_UNIT_COMMUNICATION_FAILURE;
goto send_sense;
}
- ret = target_setup_cmd_from_cdb(cmd, srp_cmd->cdb);
- if (ret < 0) {
- kref_put(&send_ioctx->kref, srpt_put_send_ioctx_kref);
- if (cmd->se_cmd_flags & SCF_SCSI_RESERVATION_CONFLICT) {
- srpt_queue_status(cmd);
- return 0;
- } else
- goto send_sense;
- }
-
- transport_handle_cdb_direct(cmd);
return 0;
send_sense:
- transport_send_check_condition_and_sense(cmd, cmd->scsi_sense_reason,
- 0);
+ transport_send_check_condition_and_sense(cmd, ret, 0);
return -1;
}
@@ -1865,9 +1825,11 @@ static void srpt_handle_tsk_mgmt(struct srpt_rdma_ch *ch,
{
struct srp_tsk_mgmt *srp_tsk;
struct se_cmd *cmd;
+ struct se_session *sess = ch->sess;
uint64_t unpacked_lun;
+ uint32_t tag = 0;
int tcm_tmr;
- int res;
+ int rc;
BUG_ON(!send_ioctx);
@@ -1882,39 +1844,32 @@ static void srpt_handle_tsk_mgmt(struct srpt_rdma_ch *ch,
send_ioctx->tag = srp_tsk->tag;
tcm_tmr = srp_tmr_to_tcm(srp_tsk->tsk_mgmt_func);
if (tcm_tmr < 0) {
- send_ioctx->cmd.se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
send_ioctx->cmd.se_tmr_req->response =
TMR_TASK_MGMT_FUNCTION_NOT_SUPPORTED;
- goto process_tmr;
- }
- res = core_tmr_alloc_req(cmd, NULL, tcm_tmr, GFP_KERNEL);
- if (res < 0) {
- send_ioctx->cmd.se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- send_ioctx->cmd.se_tmr_req->response = TMR_FUNCTION_REJECTED;
- goto process_tmr;
+ goto fail;
}
-
unpacked_lun = srpt_unpack_lun((uint8_t *)&srp_tsk->lun,
sizeof(srp_tsk->lun));
- res = transport_lookup_tmr_lun(&send_ioctx->cmd, unpacked_lun);
- if (res) {
- pr_debug("rejecting TMR for LUN %lld\n", unpacked_lun);
- send_ioctx->cmd.se_cmd_flags |= SCF_SCSI_CDB_EXCEPTION;
- send_ioctx->cmd.se_tmr_req->response = TMR_LUN_DOES_NOT_EXIST;
- goto process_tmr;
- }
-
- if (srp_tsk->tsk_mgmt_func == SRP_TSK_ABORT_TASK)
- srpt_rx_mgmt_fn_tag(send_ioctx, srp_tsk->task_tag);
-
-process_tmr:
- kref_get(&send_ioctx->kref);
- if (!(send_ioctx->cmd.se_cmd_flags & SCF_SCSI_CDB_EXCEPTION))
- transport_generic_handle_tmr(&send_ioctx->cmd);
- else
- transport_send_check_condition_and_sense(cmd,
- cmd->scsi_sense_reason, 0);
+ if (srp_tsk->tsk_mgmt_func == SRP_TSK_ABORT_TASK) {
+ rc = srpt_rx_mgmt_fn_tag(send_ioctx, srp_tsk->task_tag);
+ if (rc < 0) {
+ send_ioctx->cmd.se_tmr_req->response =
+ TMR_TASK_DOES_NOT_EXIST;
+ goto fail;
+ }
+ tag = srp_tsk->task_tag;
+ }
+ rc = target_submit_tmr(&send_ioctx->cmd, sess, NULL, unpacked_lun,
+ srp_tsk, tcm_tmr, GFP_KERNEL, tag,
+ TARGET_SCF_ACK_KREF);
+ if (rc != 0) {
+ send_ioctx->cmd.se_tmr_req->response = TMR_FUNCTION_REJECTED;
+ goto fail;
+ }
+ return;
+fail:
+ transport_send_check_condition_and_sense(cmd, 0, 0); // XXX:
}
/**
@@ -1956,10 +1911,6 @@ static void srpt_handle_new_iu(struct srpt_rdma_ch *ch,
}
}
- transport_init_se_cmd(&send_ioctx->cmd, &srpt_target->tf_ops, ch->sess,
- 0, DMA_NONE, MSG_SIMPLE_TAG,
- send_ioctx->sense_data);
-
switch (srp_cmd->opcode) {
case SRP_CMD:
srpt_handle_cmd(ch, recv_ioctx, send_ioctx);
@@ -2365,6 +2316,7 @@ static void srpt_release_channel_work(struct work_struct *w)
{
struct srpt_rdma_ch *ch;
struct srpt_device *sdev;
+ struct se_session *se_sess;
ch = container_of(w, struct srpt_rdma_ch, release_work);
pr_debug("ch = %p; ch->sess = %p; release_done = %p\n", ch, ch->sess,
@@ -2373,8 +2325,13 @@ static void srpt_release_channel_work(struct work_struct *w)
sdev = ch->sport->sdev;
BUG_ON(!sdev);
- transport_deregister_session_configfs(ch->sess);
- transport_deregister_session(ch->sess);
+ se_sess = ch->sess;
+ BUG_ON(!se_sess);
+
+ target_wait_for_sess_cmds(se_sess, 0);
+
+ transport_deregister_session_configfs(se_sess);
+ transport_deregister_session(se_sess);
ch->sess = NULL;
srpt_destroy_ch_ib(ch);
@@ -3099,7 +3056,7 @@ static int srpt_queue_response(struct se_cmd *cmd)
ioctx->tag);
srpt_unmap_sg_to_ib_sge(ch, ioctx);
srpt_set_cmd_state(ioctx, SRPT_STATE_DONE);
- kref_put(&ioctx->kref, srpt_put_send_ioctx_kref);
+ target_put_sess_cmd(ioctx->ch->sess, &ioctx->cmd);
}
out:
@@ -3490,6 +3447,23 @@ static u32 srpt_tpg_get_inst_index(struct se_portal_group *se_tpg)
static void srpt_release_cmd(struct se_cmd *se_cmd)
{
+ struct srpt_send_ioctx *ioctx = container_of(se_cmd,
+ struct srpt_send_ioctx, cmd);
+ struct srpt_rdma_ch *ch = ioctx->ch;
+ unsigned long flags;
+
+ WARN_ON(ioctx->state != SRPT_STATE_DONE);
+ WARN_ON(ioctx->mapped_sg_count != 0);
+
+ if (ioctx->n_rbuf > 1) {
+ kfree(ioctx->rbufs);
+ ioctx->rbufs = NULL;
+ ioctx->n_rbuf = 0;
+ }
+
+ spin_lock_irqsave(&ch->spinlock, flags);
+ list_add(&ioctx->free_list, &ch->free_list);
+ spin_unlock_irqrestore(&ch->spinlock, flags);
}
/**
diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.h b/drivers/infiniband/ulp/srpt/ib_srpt.h
index 61e52b830816..4caf55cda7b1 100644
--- a/drivers/infiniband/ulp/srpt/ib_srpt.h
+++ b/drivers/infiniband/ulp/srpt/ib_srpt.h
@@ -228,7 +228,6 @@ struct srpt_recv_ioctx {
struct srpt_send_ioctx {
struct srpt_ioctx ioctx;
struct srpt_rdma_ch *ch;
- struct kref kref;
struct rdma_iu *rdma_ius;
struct srp_direct_buf *rbufs;
struct srp_direct_buf single_rbuf;