diff options
Diffstat (limited to 'drivers/infiniband/sw')
60 files changed, 734 insertions, 807 deletions
diff --git a/drivers/infiniband/sw/rdmavt/ah.c b/drivers/infiniband/sw/rdmavt/ah.c index 63999239ed9e..56926617b064 100644 --- a/drivers/infiniband/sw/rdmavt/ah.c +++ b/drivers/infiniband/sw/rdmavt/ah.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 - 2019 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/ah.h b/drivers/infiniband/sw/rdmavt/ah.h index c11fdf637d64..50ddf802bdcc 100644 --- a/drivers/infiniband/sw/rdmavt/ah.h +++ b/drivers/infiniband/sw/rdmavt/ah.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/cq.c b/drivers/infiniband/sw/rdmavt/cq.c index 9fe4dcaa049a..82c3f5932249 100644 --- a/drivers/infiniband/sw/rdmavt/cq.c +++ b/drivers/infiniband/sw/rdmavt/cq.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 - 2018 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/cq.h b/drivers/infiniband/sw/rdmavt/cq.h index b0a948ec760b..d49b6d1a26cb 100644 --- a/drivers/infiniband/sw/rdmavt/cq.h +++ b/drivers/infiniband/sw/rdmavt/cq.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 - 2018 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/mad.c b/drivers/infiniband/sw/rdmavt/mad.c index 98a8fe3b04ef..846e014ecc55 100644 --- a/drivers/infiniband/sw/rdmavt/mad.c +++ b/drivers/infiniband/sw/rdmavt/mad.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/mad.h b/drivers/infiniband/sw/rdmavt/mad.h index 368be29eab37..705a94537b55 100644 --- a/drivers/infiniband/sw/rdmavt/mad.h +++ b/drivers/infiniband/sw/rdmavt/mad.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/mcast.c b/drivers/infiniband/sw/rdmavt/mcast.c index a123874e1ca7..59045bdce2a9 100644 --- a/drivers/infiniband/sw/rdmavt/mcast.c +++ b/drivers/infiniband/sw/rdmavt/mcast.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/mcast.h b/drivers/infiniband/sw/rdmavt/mcast.h index b96d86f9625b..7627e0d49d09 100644 --- a/drivers/infiniband/sw/rdmavt/mcast.h +++ b/drivers/infiniband/sw/rdmavt/mcast.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/mmap.c b/drivers/infiniband/sw/rdmavt/mmap.c index 4d2238f3f3c8..46e3b3e0643a 100644 --- a/drivers/infiniband/sw/rdmavt/mmap.c +++ b/drivers/infiniband/sw/rdmavt/mmap.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/mmap.h b/drivers/infiniband/sw/rdmavt/mmap.h index 7e92cf28e071..29aaca3e8b83 100644 --- a/drivers/infiniband/sw/rdmavt/mmap.h +++ b/drivers/infiniband/sw/rdmavt/mmap.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/mr.c b/drivers/infiniband/sw/rdmavt/mr.c index 8a1f2e285180..7a9afd5231d5 100644 --- a/drivers/infiniband/sw/rdmavt/mr.c +++ b/drivers/infiniband/sw/rdmavt/mr.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/mr.h b/drivers/infiniband/sw/rdmavt/mr.h index d17f1400b5f6..44afe2731741 100644 --- a/drivers/infiniband/sw/rdmavt/mr.h +++ b/drivers/infiniband/sw/rdmavt/mr.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/pd.c b/drivers/infiniband/sw/rdmavt/pd.c index ae62071969fa..3af8081dc6c7 100644 --- a/drivers/infiniband/sw/rdmavt/pd.c +++ b/drivers/infiniband/sw/rdmavt/pd.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/pd.h b/drivers/infiniband/sw/rdmavt/pd.h index 42a0ef3b7da3..552adaeb371f 100644 --- a/drivers/infiniband/sw/rdmavt/pd.h +++ b/drivers/infiniband/sw/rdmavt/pd.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index dc83d0ac6a38..e6203e26cc06 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 - 2020 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/qp.h b/drivers/infiniband/sw/rdmavt/qp.h index bd04be80723c..1a201d2bedd6 100644 --- a/drivers/infiniband/sw/rdmavt/qp.h +++ b/drivers/infiniband/sw/rdmavt/qp.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/rc.c b/drivers/infiniband/sw/rdmavt/rc.c index 4e5d4a27633c..7cd473302576 100644 --- a/drivers/infiniband/sw/rdmavt/rc.c +++ b/drivers/infiniband/sw/rdmavt/rc.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/srq.c b/drivers/infiniband/sw/rdmavt/srq.c index 14d196bde2a1..fe125bf85b27 100644 --- a/drivers/infiniband/sw/rdmavt/srq.c +++ b/drivers/infiniband/sw/rdmavt/srq.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/srq.h b/drivers/infiniband/sw/rdmavt/srq.h index 7d17372cd269..e654a9fa2989 100644 --- a/drivers/infiniband/sw/rdmavt/srq.h +++ b/drivers/infiniband/sw/rdmavt/srq.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/trace.c b/drivers/infiniband/sw/rdmavt/trace.c index 01704b8dd683..e31b9f3e752d 100644 --- a/drivers/infiniband/sw/rdmavt/trace.c +++ b/drivers/infiniband/sw/rdmavt/trace.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/trace.h b/drivers/infiniband/sw/rdmavt/trace.h index 30eb4a72ea7d..bdb6b9326b64 100644 --- a/drivers/infiniband/sw/rdmavt/trace.h +++ b/drivers/infiniband/sw/rdmavt/trace.h @@ -1,10 +1,10 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016, 2017 Intel Corporation. */ #define RDI_DEV_ENTRY(rdi) __string(dev, rvt_get_ibdev_name(rdi)) -#define RDI_DEV_ASSIGN(rdi) __assign_str(dev, rvt_get_ibdev_name(rdi)) +#define RDI_DEV_ASSIGN(rdi) __assign_str(dev) #include "trace_rvt.h" #include "trace_qp.h" diff --git a/drivers/infiniband/sw/rdmavt/trace_cq.h b/drivers/infiniband/sw/rdmavt/trace_cq.h index 30dd1d9bae26..54ce06e10b7f 100644 --- a/drivers/infiniband/sw/rdmavt/trace_cq.h +++ b/drivers/infiniband/sw/rdmavt/trace_cq.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 - 2018 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/trace_mr.h b/drivers/infiniband/sw/rdmavt/trace_mr.h index 1de7012000cb..0cb8e0a0565e 100644 --- a/drivers/infiniband/sw/rdmavt/trace_mr.h +++ b/drivers/infiniband/sw/rdmavt/trace_mr.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/trace_qp.h b/drivers/infiniband/sw/rdmavt/trace_qp.h index c28c81fcb32a..fa128f16ca3f 100644 --- a/drivers/infiniband/sw/rdmavt/trace_qp.h +++ b/drivers/infiniband/sw/rdmavt/trace_qp.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/trace_rc.h b/drivers/infiniband/sw/rdmavt/trace_rc.h index 833bf778b05d..9919d66c17c3 100644 --- a/drivers/infiniband/sw/rdmavt/trace_rc.h +++ b/drivers/infiniband/sw/rdmavt/trace_rc.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2017 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/trace_rvt.h b/drivers/infiniband/sw/rdmavt/trace_rvt.h index 9df6b0b8263b..a00489e66ddf 100644 --- a/drivers/infiniband/sw/rdmavt/trace_rvt.h +++ b/drivers/infiniband/sw/rdmavt/trace_rvt.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ @@ -24,7 +24,7 @@ TRACE_EVENT(rvt_dbg, ), TP_fast_assign( RDI_DEV_ASSIGN(rdi); - __assign_str(msg, msg); + __assign_str(msg); ), TP_printk("[%s]: %s", __get_str(dev), __get_str(msg)) ); diff --git a/drivers/infiniband/sw/rdmavt/trace_tx.h b/drivers/infiniband/sw/rdmavt/trace_tx.h index ff7d39a30768..dff18baa2765 100644 --- a/drivers/infiniband/sw/rdmavt/trace_tx.h +++ b/drivers/infiniband/sw/rdmavt/trace_tx.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/vt.c b/drivers/infiniband/sw/rdmavt/vt.c index d61f8de7f21c..5499025e8a0a 100644 --- a/drivers/infiniband/sw/rdmavt/vt.c +++ b/drivers/infiniband/sw/rdmavt/vt.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* * Copyright(c) 2016 - 2018 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rdmavt/vt.h b/drivers/infiniband/sw/rdmavt/vt.h index 461574e3f6a5..4d17333fa90e 100644 --- a/drivers/infiniband/sw/rdmavt/vt.h +++ b/drivers/infiniband/sw/rdmavt/vt.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* * Copyright(c) 2016 Intel Corporation. */ diff --git a/drivers/infiniband/sw/rxe/rxe.c b/drivers/infiniband/sw/rxe/rxe.c index 54c723a6edda..255677bc12b2 100644 --- a/drivers/infiniband/sw/rxe/rxe.c +++ b/drivers/infiniband/sw/rxe/rxe.c @@ -33,6 +33,8 @@ void rxe_dealloc(struct ib_device *ib_dev) if (rxe->tfm) crypto_free_shash(rxe->tfm); + + mutex_destroy(&rxe->usdev_lock); } /* initialize rxe device parameters */ @@ -160,8 +162,6 @@ void rxe_set_mtu(struct rxe_dev *rxe, unsigned int ndev_mtu) port->attr.active_mtu = mtu; port->mtu_cap = ib_mtu_enum_to_int(mtu); - - rxe_info_dev(rxe, "Set mtu to %d", port->mtu_cap); } /* called by ifc layer to create new rxe device. @@ -181,7 +181,7 @@ static int rxe_newlink(const char *ibdev_name, struct net_device *ndev) int err = 0; if (is_vlan_dev(ndev)) { - rxe_err("rxe creation allowed on top of a real device only"); + rxe_err("rxe creation allowed on top of a real device only\n"); err = -EPERM; goto err; } @@ -189,7 +189,7 @@ static int rxe_newlink(const char *ibdev_name, struct net_device *ndev) rxe = rxe_get_dev_from_net(ndev); if (rxe) { ib_device_put(&rxe->ib_dev); - rxe_err_dev(rxe, "already configured on %s", ndev->name); + rxe_err_dev(rxe, "already configured on %s\n", ndev->name); err = -EEXIST; goto err; } diff --git a/drivers/infiniband/sw/rxe/rxe.h b/drivers/infiniband/sw/rxe/rxe.h index d33dd6cf83d3..d8fb2c7af30a 100644 --- a/drivers/infiniband/sw/rxe/rxe.h +++ b/drivers/infiniband/sw/rxe/rxe.h @@ -38,7 +38,7 @@ #define RXE_ROCE_V2_SPORT (0xc000) -#define rxe_dbg(fmt, ...) pr_debug("%s: " fmt "\n", __func__, ##__VA_ARGS__) +#define rxe_dbg(fmt, ...) pr_debug("%s: " fmt, __func__, ##__VA_ARGS__) #define rxe_dbg_dev(rxe, fmt, ...) ibdev_dbg(&(rxe)->ib_dev, \ "%s: " fmt, __func__, ##__VA_ARGS__) #define rxe_dbg_uc(uc, fmt, ...) ibdev_dbg((uc)->ibuc.device, \ @@ -58,7 +58,7 @@ #define rxe_dbg_mw(mw, fmt, ...) ibdev_dbg((mw)->ibmw.device, \ "mw#%d %s: " fmt, (mw)->elem.index, __func__, ##__VA_ARGS__) -#define rxe_err(fmt, ...) pr_err_ratelimited("%s: " fmt "\n", __func__, \ +#define rxe_err(fmt, ...) pr_err_ratelimited("%s: " fmt, __func__, \ ##__VA_ARGS__) #define rxe_err_dev(rxe, fmt, ...) ibdev_err_ratelimited(&(rxe)->ib_dev, \ "%s: " fmt, __func__, ##__VA_ARGS__) @@ -79,7 +79,7 @@ #define rxe_err_mw(mw, fmt, ...) ibdev_err_ratelimited((mw)->ibmw.device, \ "mw#%d %s: " fmt, (mw)->elem.index, __func__, ##__VA_ARGS__) -#define rxe_info(fmt, ...) pr_info_ratelimited("%s: " fmt "\n", __func__, \ +#define rxe_info(fmt, ...) pr_info_ratelimited("%s: " fmt, __func__, \ ##__VA_ARGS__) #define rxe_info_dev(rxe, fmt, ...) ibdev_info_ratelimited(&(rxe)->ib_dev, \ "%s: " fmt, __func__, ##__VA_ARGS__) diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c index 5111735aafae..d48af2180745 100644 --- a/drivers/infiniband/sw/rxe/rxe_comp.c +++ b/drivers/infiniband/sw/rxe/rxe_comp.c @@ -122,25 +122,16 @@ void retransmit_timer(struct timer_list *t) spin_lock_irqsave(&qp->state_lock, flags); if (qp->valid) { qp->comp.timeout = 1; - rxe_sched_task(&qp->comp.task); + rxe_sched_task(&qp->send_task); } spin_unlock_irqrestore(&qp->state_lock, flags); } void rxe_comp_queue_pkt(struct rxe_qp *qp, struct sk_buff *skb) { - int must_sched; - + rxe_counter_inc(SKB_TO_PKT(skb)->rxe, RXE_CNT_SENDER_SCHED); skb_queue_tail(&qp->resp_pkts, skb); - - must_sched = skb_queue_len(&qp->resp_pkts) > 1; - if (must_sched != 0) - rxe_counter_inc(SKB_TO_PKT(skb)->rxe, RXE_CNT_COMPLETER_SCHED); - - if (must_sched) - rxe_sched_task(&qp->comp.task); - else - rxe_run_task(&qp->comp.task); + rxe_sched_task(&qp->send_task); } static inline enum comp_state get_wqe(struct rxe_qp *qp, @@ -325,7 +316,7 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, qp->comp.psn = pkt->psn; if (qp->req.wait_psn) { qp->req.wait_psn = 0; - rxe_sched_task(&qp->req.task); + qp->req.again = 1; } } return COMPST_ERROR_RETRY; @@ -433,7 +424,7 @@ static void make_send_cqe(struct rxe_qp *qp, struct rxe_send_wqe *wqe, } } else { if (wqe->status != IB_WC_WR_FLUSH_ERR) - rxe_err_qp(qp, "non-flush error status = %d", + rxe_err_qp(qp, "non-flush error status = %d\n", wqe->status); } } @@ -476,7 +467,7 @@ static void do_complete(struct rxe_qp *qp, struct rxe_send_wqe *wqe) */ if (qp->req.wait_fence) { qp->req.wait_fence = 0; - rxe_sched_task(&qp->req.task); + qp->req.again = 1; } } @@ -515,7 +506,7 @@ static inline enum comp_state complete_ack(struct rxe_qp *qp, if (qp->req.need_rd_atomic) { qp->comp.timeout_retry = 0; qp->req.need_rd_atomic = 0; - rxe_sched_task(&qp->req.task); + qp->req.again = 1; } } @@ -541,7 +532,7 @@ static inline enum comp_state complete_wqe(struct rxe_qp *qp, if (qp->req.wait_psn) { qp->req.wait_psn = 0; - rxe_sched_task(&qp->req.task); + qp->req.again = 1; } } @@ -582,7 +573,7 @@ static int flush_send_wqe(struct rxe_qp *qp, struct rxe_send_wqe *wqe) err = rxe_cq_post(qp->scq, &cqe, 0); if (err) - rxe_dbg_cq(qp->scq, "post cq failed, err = %d", err); + rxe_dbg_cq(qp->scq, "post cq failed, err = %d\n", err); return err; } @@ -597,6 +588,10 @@ static void flush_send_queue(struct rxe_qp *qp, bool notify) struct rxe_queue *q = qp->sq.queue; int err; + /* send queue never got created. nothing to do. */ + if (!qp->sq.queue) + return; + while ((wqe = queue_head(q, q->type))) { if (notify) { err = flush_send_wqe(qp, wqe); @@ -650,6 +645,8 @@ int rxe_completer(struct rxe_qp *qp) int ret; unsigned long flags; + qp->req.again = 0; + spin_lock_irqsave(&qp->state_lock, flags); if (!qp->valid || qp_state(qp) == IB_QPS_ERR || qp_state(qp) == IB_QPS_RESET) { @@ -733,7 +730,7 @@ int rxe_completer(struct rxe_qp *qp) if (qp->req.wait_psn) { qp->req.wait_psn = 0; - rxe_sched_task(&qp->req.task); + qp->req.again = 1; } state = COMPST_DONE; @@ -788,7 +785,7 @@ int rxe_completer(struct rxe_qp *qp) RXE_CNT_COMP_RETRY); qp->req.need_retry = 1; qp->comp.started_retry = 1; - rxe_sched_task(&qp->req.task); + qp->req.again = 1; } goto done; @@ -839,8 +836,9 @@ done: ret = 0; goto out; exit: - ret = -EAGAIN; + ret = (qp->req.again) ? 0 : -EAGAIN; out: + qp->req.again = 0; if (pkt) free_pkt(pkt); return ret; diff --git a/drivers/infiniband/sw/rxe/rxe_cq.c b/drivers/infiniband/sw/rxe/rxe_cq.c index d5486cbb3f10..fec87c9030ab 100644 --- a/drivers/infiniband/sw/rxe/rxe_cq.c +++ b/drivers/infiniband/sw/rxe/rxe_cq.c @@ -27,7 +27,7 @@ int rxe_cq_chk_attr(struct rxe_dev *rxe, struct rxe_cq *cq, if (cq) { count = queue_count(cq->queue, QUEUE_TYPE_TO_CLIENT); if (cqe < count) { - rxe_dbg_cq(cq, "cqe(%d) < current # elements in queue (%d)", + rxe_dbg_cq(cq, "cqe(%d) < current # elements in queue (%d)\n", cqe, count); goto err1; } @@ -96,7 +96,7 @@ int rxe_cq_post(struct rxe_cq *cq, struct rxe_cqe *cqe, int solicited) full = queue_full(cq->queue, QUEUE_TYPE_TO_CLIENT); if (unlikely(full)) { - rxe_err_cq(cq, "queue full"); + rxe_err_cq(cq, "queue full\n"); spin_unlock_irqrestore(&cq->cq_lock, flags); if (cq->ibcq.event_handler) { ev.device = cq->ibcq.device; diff --git a/drivers/infiniband/sw/rxe/rxe_hw_counters.c b/drivers/infiniband/sw/rxe/rxe_hw_counters.c index a012522b577a..437917a7d8f2 100644 --- a/drivers/infiniband/sw/rxe/rxe_hw_counters.c +++ b/drivers/infiniband/sw/rxe/rxe_hw_counters.c @@ -14,7 +14,7 @@ static const struct rdma_stat_desc rxe_counter_descs[] = { [RXE_CNT_RCV_RNR].name = "rcvd_rnr_err", [RXE_CNT_SND_RNR].name = "send_rnr_err", [RXE_CNT_RCV_SEQ_ERR].name = "rcvd_seq_err", - [RXE_CNT_COMPLETER_SCHED].name = "ack_deferred", + [RXE_CNT_SENDER_SCHED].name = "ack_deferred", [RXE_CNT_RETRY_EXCEEDED].name = "retry_exceeded_err", [RXE_CNT_RNR_RETRY_EXCEEDED].name = "retry_rnr_exceeded_err", [RXE_CNT_COMP_RETRY].name = "completer_retry_err", diff --git a/drivers/infiniband/sw/rxe/rxe_hw_counters.h b/drivers/infiniband/sw/rxe/rxe_hw_counters.h index 71f4d4fa9dc8..051f9e1c3852 100644 --- a/drivers/infiniband/sw/rxe/rxe_hw_counters.h +++ b/drivers/infiniband/sw/rxe/rxe_hw_counters.h @@ -18,7 +18,7 @@ enum rxe_counters { RXE_CNT_RCV_RNR, RXE_CNT_SND_RNR, RXE_CNT_RCV_SEQ_ERR, - RXE_CNT_COMPLETER_SCHED, + RXE_CNT_SENDER_SCHED, RXE_CNT_RETRY_EXCEEDED, RXE_CNT_RNR_RETRY_EXCEEDED, RXE_CNT_COMP_RETRY, diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h index 666e06a82bc9..ded46119151b 100644 --- a/drivers/infiniband/sw/rxe/rxe_loc.h +++ b/drivers/infiniband/sw/rxe/rxe_loc.h @@ -59,7 +59,7 @@ int rxe_mmap(struct ib_ucontext *context, struct vm_area_struct *vma); /* rxe_mr.c */ u8 rxe_get_next_key(u32 last_key); void rxe_mr_init_dma(int access, struct rxe_mr *mr); -int rxe_mr_init_user(struct rxe_dev *rxe, u64 start, u64 length, u64 iova, +int rxe_mr_init_user(struct rxe_dev *rxe, u64 start, u64 length, int access, struct rxe_mr *mr); int rxe_mr_init_fast(int max_pages, struct rxe_mr *mr); int rxe_flush_pmem_iova(struct rxe_mr *mr, u64 iova, unsigned int length); @@ -136,12 +136,6 @@ static inline int qp_mtu(struct rxe_qp *qp) return IB_MTU_4096; } -static inline int rcv_wqe_size(int max_sge) -{ - return sizeof(struct rxe_recv_wqe) + - max_sge * sizeof(struct ib_sge); -} - void free_rd_atomic_resource(struct resp_res *res); static inline void rxe_advance_resp_resource(struct rxe_qp *qp) @@ -170,7 +164,8 @@ void rxe_dealloc(struct ib_device *ib_dev); int rxe_completer(struct rxe_qp *qp); int rxe_requester(struct rxe_qp *qp); -int rxe_responder(struct rxe_qp *qp); +int rxe_sender(struct rxe_qp *qp); +int rxe_receiver(struct rxe_qp *qp); /* rxe_icrc.c */ int rxe_icrc_init(struct rxe_dev *rxe); diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c index f54042e9aeb2..da3dee520876 100644 --- a/drivers/infiniband/sw/rxe/rxe_mr.c +++ b/drivers/infiniband/sw/rxe/rxe_mr.c @@ -34,7 +34,7 @@ int mr_check_range(struct rxe_mr *mr, u64 iova, size_t length) case IB_MR_TYPE_MEM_REG: if (iova < mr->ibmr.iova || iova + length > mr->ibmr.iova + mr->ibmr.length) { - rxe_dbg_mr(mr, "iova/length out of range"); + rxe_dbg_mr(mr, "iova/length out of range\n"); return -EINVAL; } return 0; @@ -126,7 +126,7 @@ static int rxe_mr_fill_pages_from_sgt(struct rxe_mr *mr, struct sg_table *sgt) return xas_error(&xas); } -int rxe_mr_init_user(struct rxe_dev *rxe, u64 start, u64 length, u64 iova, +int rxe_mr_init_user(struct rxe_dev *rxe, u64 start, u64 length, int access, struct rxe_mr *mr) { struct ib_umem *umem; @@ -319,7 +319,7 @@ int rxe_mr_copy(struct rxe_mr *mr, u64 iova, void *addr, err = mr_check_range(mr, iova, length); if (unlikely(err)) { - rxe_dbg_mr(mr, "iova out of range"); + rxe_dbg_mr(mr, "iova out of range\n"); return err; } @@ -477,7 +477,7 @@ int rxe_mr_do_atomic_op(struct rxe_mr *mr, u64 iova, int opcode, u64 *va; if (unlikely(mr->state != RXE_MR_STATE_VALID)) { - rxe_dbg_mr(mr, "mr not in valid state"); + rxe_dbg_mr(mr, "mr not in valid state\n"); return RESPST_ERR_RKEY_VIOLATION; } @@ -490,7 +490,7 @@ int rxe_mr_do_atomic_op(struct rxe_mr *mr, u64 iova, int opcode, err = mr_check_range(mr, iova, sizeof(value)); if (err) { - rxe_dbg_mr(mr, "iova out of range"); + rxe_dbg_mr(mr, "iova out of range\n"); return RESPST_ERR_RKEY_VIOLATION; } page_offset = rxe_mr_iova_to_page_offset(mr, iova); @@ -501,7 +501,7 @@ int rxe_mr_do_atomic_op(struct rxe_mr *mr, u64 iova, int opcode, } if (unlikely(page_offset & 0x7)) { - rxe_dbg_mr(mr, "iova not aligned"); + rxe_dbg_mr(mr, "iova not aligned\n"); return RESPST_ERR_MISALIGNED_ATOMIC; } @@ -534,7 +534,7 @@ int rxe_mr_do_atomic_write(struct rxe_mr *mr, u64 iova, u64 value) /* See IBA oA19-28 */ if (unlikely(mr->state != RXE_MR_STATE_VALID)) { - rxe_dbg_mr(mr, "mr not in valid state"); + rxe_dbg_mr(mr, "mr not in valid state\n"); return RESPST_ERR_RKEY_VIOLATION; } @@ -548,7 +548,7 @@ int rxe_mr_do_atomic_write(struct rxe_mr *mr, u64 iova, u64 value) /* See IBA oA19-28 */ err = mr_check_range(mr, iova, sizeof(value)); if (unlikely(err)) { - rxe_dbg_mr(mr, "iova out of range"); + rxe_dbg_mr(mr, "iova out of range\n"); return RESPST_ERR_RKEY_VIOLATION; } page_offset = rxe_mr_iova_to_page_offset(mr, iova); @@ -560,7 +560,7 @@ int rxe_mr_do_atomic_write(struct rxe_mr *mr, u64 iova, u64 value) /* See IBA A19.4.2 */ if (unlikely(page_offset & 0x7)) { - rxe_dbg_mr(mr, "misaligned address"); + rxe_dbg_mr(mr, "misaligned address\n"); return RESPST_ERR_MISALIGNED_ATOMIC; } diff --git a/drivers/infiniband/sw/rxe/rxe_mw.c b/drivers/infiniband/sw/rxe/rxe_mw.c index d9312b5c9d20..379e65bfcd49 100644 --- a/drivers/infiniband/sw/rxe/rxe_mw.c +++ b/drivers/infiniband/sw/rxe/rxe_mw.c @@ -198,7 +198,7 @@ int rxe_bind_mw(struct rxe_qp *qp, struct rxe_send_wqe *wqe) } if (access & ~RXE_ACCESS_SUPPORTED_MW) { - rxe_err_mw(mw, "access %#x not supported", access); + rxe_err_mw(mw, "access %#x not supported\n", access); ret = -EOPNOTSUPP; goto err_drop_mr; } diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c index cd59666158b1..ca9a82e1c4c7 100644 --- a/drivers/infiniband/sw/rxe/rxe_net.c +++ b/drivers/infiniband/sw/rxe/rxe_net.c @@ -345,46 +345,52 @@ int rxe_prepare(struct rxe_av *av, struct rxe_pkt_info *pkt, static void rxe_skb_tx_dtor(struct sk_buff *skb) { - struct sock *sk = skb->sk; - struct rxe_qp *qp = sk->sk_user_data; - int skb_out = atomic_dec_return(&qp->skb_out); + struct net_device *ndev = skb->dev; + struct rxe_dev *rxe; + unsigned int qp_index; + struct rxe_qp *qp; + int skb_out; + + rxe = rxe_get_dev_from_net(ndev); + if (!rxe && is_vlan_dev(ndev)) + rxe = rxe_get_dev_from_net(vlan_dev_real_dev(ndev)); + if (WARN_ON(!rxe)) + return; + + qp_index = (int)(uintptr_t)skb->sk->sk_user_data; + if (!qp_index) + return; - if (unlikely(qp->need_req_skb && - skb_out < RXE_INFLIGHT_SKBS_PER_QP_LOW)) - rxe_sched_task(&qp->req.task); + qp = rxe_pool_get_index(&rxe->qp_pool, qp_index); + if (!qp) + goto put_dev; + + skb_out = atomic_dec_return(&qp->skb_out); + if (qp->need_req_skb && skb_out < RXE_INFLIGHT_SKBS_PER_QP_LOW) + rxe_sched_task(&qp->send_task); rxe_put(qp); +put_dev: + ib_device_put(&rxe->ib_dev); + sock_put(skb->sk); } static int rxe_send(struct sk_buff *skb, struct rxe_pkt_info *pkt) { int err; + struct sock *sk = pkt->qp->sk->sk; + sock_hold(sk); + skb->sk = sk; skb->destructor = rxe_skb_tx_dtor; - skb->sk = pkt->qp->sk->sk; - - rxe_get(pkt->qp); atomic_inc(&pkt->qp->skb_out); - if (skb->protocol == htons(ETH_P_IP)) { + if (skb->protocol == htons(ETH_P_IP)) err = ip_local_out(dev_net(skb_dst(skb)->dev), skb->sk, skb); - } else if (skb->protocol == htons(ETH_P_IPV6)) { + else err = ip6_local_out(dev_net(skb_dst(skb)->dev), skb->sk, skb); - } else { - rxe_dbg_qp(pkt->qp, "Unknown layer 3 protocol: %d\n", - skb->protocol); - atomic_dec(&pkt->qp->skb_out); - rxe_put(pkt->qp); - kfree_skb(skb); - return -EINVAL; - } - - if (unlikely(net_xmit_eval(err))) { - rxe_dbg_qp(pkt->qp, "error sending packet: %d\n", err); - return -EAGAIN; - } - return 0; + return err; } /* fix up a send packet to match the packets @@ -392,8 +398,15 @@ static int rxe_send(struct sk_buff *skb, struct rxe_pkt_info *pkt) */ static int rxe_loopback(struct sk_buff *skb, struct rxe_pkt_info *pkt) { + struct sock *sk = pkt->qp->sk->sk; + memcpy(SKB_TO_PKT(skb), pkt, sizeof(*pkt)); + sock_hold(sk); + skb->sk = sk; + skb->destructor = rxe_skb_tx_dtor; + atomic_inc(&pkt->qp->skb_out); + if (skb->protocol == htons(ETH_P_IP)) skb_pull(skb, sizeof(struct iphdr)); else @@ -440,12 +453,6 @@ int rxe_xmit_packet(struct rxe_qp *qp, struct rxe_pkt_info *pkt, return err; } - if ((qp_type(qp) != IB_QPT_RC) && - (pkt->mask & RXE_END_MASK)) { - pkt->wqe->state = wqe_state_done; - rxe_sched_task(&qp->comp.task); - } - rxe_counter_inc(rxe, RXE_CNT_SENT_PKTS); goto done; diff --git a/drivers/infiniband/sw/rxe/rxe_pool.c b/drivers/infiniband/sw/rxe/rxe_pool.c index 6215c6de3a84..67567d62195e 100644 --- a/drivers/infiniband/sw/rxe/rxe_pool.c +++ b/drivers/infiniband/sw/rxe/rxe_pool.c @@ -119,7 +119,7 @@ void rxe_pool_cleanup(struct rxe_pool *pool) int __rxe_add_to_pool(struct rxe_pool *pool, struct rxe_pool_elem *elem, bool sleepable) { - int err; + int err = -EINVAL; gfp_t gfp_flags; if (atomic_inc_return(&pool->num_elem) > pool->max_elem) @@ -147,7 +147,7 @@ int __rxe_add_to_pool(struct rxe_pool *pool, struct rxe_pool_elem *elem, err_cnt: atomic_dec(&pool->num_elem); - return -EINVAL; + return err; } void *rxe_pool_get_index(struct rxe_pool *pool, u32 index) diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c index a569b111a9d2..d2f7b5195c19 100644 --- a/drivers/infiniband/sw/rxe/rxe_qp.c +++ b/drivers/infiniband/sw/rxe/rxe_qp.c @@ -183,18 +183,68 @@ static void rxe_qp_init_misc(struct rxe_dev *rxe, struct rxe_qp *qp, atomic_set(&qp->skb_out, 0); } +static int rxe_init_sq(struct rxe_qp *qp, struct ib_qp_init_attr *init, + struct ib_udata *udata, + struct rxe_create_qp_resp __user *uresp) +{ + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); + int wqe_size; + int err; + + qp->sq.max_wr = init->cap.max_send_wr; + wqe_size = max_t(int, init->cap.max_send_sge * sizeof(struct ib_sge), + init->cap.max_inline_data); + qp->sq.max_sge = wqe_size / sizeof(struct ib_sge); + qp->sq.max_inline = wqe_size; + wqe_size += sizeof(struct rxe_send_wqe); + + qp->sq.queue = rxe_queue_init(rxe, &qp->sq.max_wr, wqe_size, + QUEUE_TYPE_FROM_CLIENT); + if (!qp->sq.queue) { + rxe_err_qp(qp, "Unable to allocate send queue\n"); + err = -ENOMEM; + goto err_out; + } + + /* prepare info for caller to mmap send queue if user space qp */ + err = do_mmap_info(rxe, uresp ? &uresp->sq_mi : NULL, udata, + qp->sq.queue->buf, qp->sq.queue->buf_size, + &qp->sq.queue->ip); + if (err) { + rxe_err_qp(qp, "do_mmap_info failed, err = %d\n", err); + goto err_free; + } + + /* return actual capabilities to caller which may be larger + * than requested + */ + init->cap.max_send_wr = qp->sq.max_wr; + init->cap.max_send_sge = qp->sq.max_sge; + init->cap.max_inline_data = qp->sq.max_inline; + + return 0; + +err_free: + vfree(qp->sq.queue->buf); + kfree(qp->sq.queue); + qp->sq.queue = NULL; +err_out: + return err; +} + static int rxe_qp_init_req(struct rxe_dev *rxe, struct rxe_qp *qp, struct ib_qp_init_attr *init, struct ib_udata *udata, struct rxe_create_qp_resp __user *uresp) { int err; - int wqe_size; - enum queue_type type; + + /* if we don't finish qp create make sure queue is valid */ + skb_queue_head_init(&qp->req_pkts); err = sock_create_kern(&init_net, AF_INET, SOCK_DGRAM, 0, &qp->sk); if (err < 0) return err; - qp->sk->sk->sk_user_data = qp; + qp->sk->sk->sk_user_data = (void *)(uintptr_t)qp->elem.index; /* pick a source UDP port number for this QP based on * the source QPN. this spreads traffic for different QPs @@ -204,32 +254,10 @@ static int rxe_qp_init_req(struct rxe_dev *rxe, struct rxe_qp *qp, * (0xc000 - 0xffff). */ qp->src_port = RXE_ROCE_V2_SPORT + (hash_32(qp_num(qp), 14) & 0x3fff); - qp->sq.max_wr = init->cap.max_send_wr; - - /* These caps are limited by rxe_qp_chk_cap() done by the caller */ - wqe_size = max_t(int, init->cap.max_send_sge * sizeof(struct ib_sge), - init->cap.max_inline_data); - qp->sq.max_sge = init->cap.max_send_sge = - wqe_size / sizeof(struct ib_sge); - qp->sq.max_inline = init->cap.max_inline_data = wqe_size; - wqe_size += sizeof(struct rxe_send_wqe); - type = QUEUE_TYPE_FROM_CLIENT; - qp->sq.queue = rxe_queue_init(rxe, &qp->sq.max_wr, - wqe_size, type); - if (!qp->sq.queue) - return -ENOMEM; - - err = do_mmap_info(rxe, uresp ? &uresp->sq_mi : NULL, udata, - qp->sq.queue->buf, qp->sq.queue->buf_size, - &qp->sq.queue->ip); - - if (err) { - vfree(qp->sq.queue->buf); - kfree(qp->sq.queue); - qp->sq.queue = NULL; + err = rxe_init_sq(qp, init, udata, uresp); + if (err) return err; - } qp->req.wqe_index = queue_get_producer(qp->sq.queue, QUEUE_TYPE_FROM_CLIENT); @@ -237,8 +265,7 @@ static int rxe_qp_init_req(struct rxe_dev *rxe, struct rxe_qp *qp, qp->req.opcode = -1; qp->comp.opcode = -1; - rxe_init_task(&qp->req.task, qp, rxe_requester); - rxe_init_task(&qp->comp.task, qp, rxe_completer); + rxe_init_task(&qp->send_task, qp, rxe_sender); qp->qp_timeout_jiffies = 0; /* Can't be set for UD/UC in modify_qp */ if (init->qp_type == IB_QPT_RC) { @@ -248,39 +275,68 @@ static int rxe_qp_init_req(struct rxe_dev *rxe, struct rxe_qp *qp, return 0; } +static int rxe_init_rq(struct rxe_qp *qp, struct ib_qp_init_attr *init, + struct ib_udata *udata, + struct rxe_create_qp_resp __user *uresp) +{ + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); + int wqe_size; + int err; + + qp->rq.max_wr = init->cap.max_recv_wr; + qp->rq.max_sge = init->cap.max_recv_sge; + wqe_size = sizeof(struct rxe_recv_wqe) + + qp->rq.max_sge*sizeof(struct ib_sge); + + qp->rq.queue = rxe_queue_init(rxe, &qp->rq.max_wr, wqe_size, + QUEUE_TYPE_FROM_CLIENT); + if (!qp->rq.queue) { + rxe_err_qp(qp, "Unable to allocate recv queue\n"); + err = -ENOMEM; + goto err_out; + } + + /* prepare info for caller to mmap recv queue if user space qp */ + err = do_mmap_info(rxe, uresp ? &uresp->rq_mi : NULL, udata, + qp->rq.queue->buf, qp->rq.queue->buf_size, + &qp->rq.queue->ip); + if (err) { + rxe_err_qp(qp, "do_mmap_info failed, err = %d\n", err); + goto err_free; + } + + /* return actual capabilities to caller which may be larger + * than requested + */ + init->cap.max_recv_wr = qp->rq.max_wr; + + return 0; + +err_free: + vfree(qp->rq.queue->buf); + kfree(qp->rq.queue); + qp->rq.queue = NULL; +err_out: + return err; +} + static int rxe_qp_init_resp(struct rxe_dev *rxe, struct rxe_qp *qp, struct ib_qp_init_attr *init, struct ib_udata *udata, struct rxe_create_qp_resp __user *uresp) { int err; - int wqe_size; - enum queue_type type; + + /* if we don't finish qp create make sure queue is valid */ + skb_queue_head_init(&qp->resp_pkts); if (!qp->srq) { - qp->rq.max_wr = init->cap.max_recv_wr; - qp->rq.max_sge = init->cap.max_recv_sge; - - wqe_size = rcv_wqe_size(qp->rq.max_sge); - - type = QUEUE_TYPE_FROM_CLIENT; - qp->rq.queue = rxe_queue_init(rxe, &qp->rq.max_wr, - wqe_size, type); - if (!qp->rq.queue) - return -ENOMEM; - - err = do_mmap_info(rxe, uresp ? &uresp->rq_mi : NULL, udata, - qp->rq.queue->buf, qp->rq.queue->buf_size, - &qp->rq.queue->ip); - if (err) { - vfree(qp->rq.queue->buf); - kfree(qp->rq.queue); - qp->rq.queue = NULL; + err = rxe_init_rq(qp, init, udata, uresp); + if (err) return err; - } } - rxe_init_task(&qp->resp.task, qp, rxe_responder); + rxe_init_task(&qp->recv_task, qp, rxe_receiver); qp->resp.opcode = OPCODE_NONE; qp->resp.msn = 0; @@ -307,10 +363,10 @@ int rxe_qp_from_init(struct rxe_dev *rxe, struct rxe_qp *qp, struct rxe_pd *pd, if (srq) rxe_get(srq); - qp->pd = pd; - qp->rcq = rcq; - qp->scq = scq; - qp->srq = srq; + qp->pd = pd; + qp->rcq = rcq; + qp->scq = scq; + qp->srq = srq; atomic_inc(&rcq->num_wq); atomic_inc(&scq->num_wq); @@ -457,14 +513,12 @@ err1: static void rxe_qp_reset(struct rxe_qp *qp) { /* stop tasks from running */ - rxe_disable_task(&qp->resp.task); - rxe_disable_task(&qp->comp.task); - rxe_disable_task(&qp->req.task); + rxe_disable_task(&qp->recv_task); + rxe_disable_task(&qp->send_task); /* drain work and packet queuesc */ - rxe_requester(qp); - rxe_completer(qp); - rxe_responder(qp); + rxe_sender(qp); + rxe_receiver(qp); if (qp->rq.queue) rxe_queue_reset(qp->rq.queue); @@ -491,9 +545,8 @@ static void rxe_qp_reset(struct rxe_qp *qp) cleanup_rd_atomic_resources(qp); /* reenable tasks */ - rxe_enable_task(&qp->resp.task); - rxe_enable_task(&qp->comp.task); - rxe_enable_task(&qp->req.task); + rxe_enable_task(&qp->recv_task); + rxe_enable_task(&qp->send_task); } /* move the qp to the error state */ @@ -505,9 +558,8 @@ void rxe_qp_error(struct rxe_qp *qp) qp->attr.qp_state = IB_QPS_ERR; /* drain work and packet queues */ - rxe_sched_task(&qp->resp.task); - rxe_sched_task(&qp->comp.task); - rxe_sched_task(&qp->req.task); + rxe_sched_task(&qp->recv_task); + rxe_sched_task(&qp->send_task); spin_unlock_irqrestore(&qp->state_lock, flags); } @@ -518,8 +570,7 @@ static void rxe_qp_sqd(struct rxe_qp *qp, struct ib_qp_attr *attr, spin_lock_irqsave(&qp->state_lock, flags); qp->attr.sq_draining = 1; - rxe_sched_task(&qp->comp.task); - rxe_sched_task(&qp->req.task); + rxe_sched_task(&qp->send_task); spin_unlock_irqrestore(&qp->state_lock, flags); } @@ -764,19 +815,15 @@ static void rxe_qp_do_cleanup(struct work_struct *work) del_timer_sync(&qp->rnr_nak_timer); } - if (qp->resp.task.func) - rxe_cleanup_task(&qp->resp.task); - - if (qp->req.task.func) - rxe_cleanup_task(&qp->req.task); + if (qp->recv_task.func) + rxe_cleanup_task(&qp->recv_task); - if (qp->comp.task.func) - rxe_cleanup_task(&qp->comp.task); + if (qp->send_task.func) + rxe_cleanup_task(&qp->send_task); /* flush out any receive wr's or pending requests */ - rxe_requester(qp); - rxe_completer(qp); - rxe_responder(qp); + rxe_sender(qp); + rxe_receiver(qp); if (qp->sq.queue) rxe_queue_cleanup(qp->sq.queue); diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c index 2171f19494bc..cd14c4c2dff9 100644 --- a/drivers/infiniband/sw/rxe/rxe_req.c +++ b/drivers/infiniband/sw/rxe/rxe_req.c @@ -108,7 +108,7 @@ void rnr_nak_timer(struct timer_list *t) /* request a send queue retry */ qp->req.need_retry = 1; qp->req.wait_for_rnr_timer = 0; - rxe_sched_task(&qp->req.task); + rxe_sched_task(&qp->send_task); } spin_unlock_irqrestore(&qp->state_lock, flags); } @@ -545,6 +545,8 @@ static void update_wqe_state(struct rxe_qp *qp, if (pkt->mask & RXE_END_MASK) { if (qp_type(qp) == IB_QPT_RC) wqe->state = wqe_state_pending; + else + wqe->state = wqe_state_done; } else { wqe->state = wqe_state_processing; } @@ -573,28 +575,6 @@ static void update_wqe_psn(struct rxe_qp *qp, qp->req.psn = (qp->req.psn + 1) & BTH_PSN_MASK; } -static void save_state(struct rxe_send_wqe *wqe, - struct rxe_qp *qp, - struct rxe_send_wqe *rollback_wqe, - u32 *rollback_psn) -{ - rollback_wqe->state = wqe->state; - rollback_wqe->first_psn = wqe->first_psn; - rollback_wqe->last_psn = wqe->last_psn; - *rollback_psn = qp->req.psn; -} - -static void rollback_state(struct rxe_send_wqe *wqe, - struct rxe_qp *qp, - struct rxe_send_wqe *rollback_wqe, - u32 rollback_psn) -{ - wqe->state = rollback_wqe->state; - wqe->first_psn = rollback_wqe->first_psn; - wqe->last_psn = rollback_wqe->last_psn; - qp->req.psn = rollback_psn; -} - static void update_state(struct rxe_qp *qp, struct rxe_pkt_info *pkt) { qp->req.opcode = pkt->opcode; @@ -653,12 +633,6 @@ static int rxe_do_local_ops(struct rxe_qp *qp, struct rxe_send_wqe *wqe) wqe->status = IB_WC_SUCCESS; qp->req.wqe_index = queue_next_index(qp->sq.queue, qp->req.wqe_index); - /* There is no ack coming for local work requests - * which can lead to a deadlock. So go ahead and complete - * it now. - */ - rxe_sched_task(&qp->comp.task); - return 0; } @@ -674,8 +648,6 @@ int rxe_requester(struct rxe_qp *qp) int opcode; int err; int ret; - struct rxe_send_wqe rollback_wqe; - u32 rollback_psn; struct rxe_queue *q = qp->sq.queue; struct rxe_ah *ah; struct rxe_av *av; @@ -784,7 +756,6 @@ int rxe_requester(struct rxe_qp *qp) qp->req.wqe_index); wqe->state = wqe_state_done; wqe->status = IB_WC_SUCCESS; - rxe_sched_task(&qp->comp.task); goto done; } payload = mtu; @@ -829,31 +800,14 @@ int rxe_requester(struct rxe_qp *qp) if (ah) rxe_put(ah); - /* - * To prevent a race on wqe access between requester and completer, - * wqe members state and psn need to be set before calling - * rxe_xmit_packet(). - * Otherwise, completer might initiate an unjustified retry flow. - */ - save_state(wqe, qp, &rollback_wqe, &rollback_psn); - update_wqe_state(qp, wqe, &pkt); - update_wqe_psn(qp, wqe, &pkt, payload); - err = rxe_xmit_packet(qp, &pkt, skb); if (err) { - qp->need_req_skb = 1; - - rollback_state(wqe, qp, &rollback_wqe, rollback_psn); - - if (err == -EAGAIN) { - rxe_sched_task(&qp->req.task); - goto exit; - } - wqe->status = IB_WC_LOC_QP_OP_ERR; goto err; } + update_wqe_state(qp, wqe, &pkt); + update_wqe_psn(qp, wqe, &pkt, payload); update_state(qp, &pkt); /* A non-zero return value will cause rxe_do_task to @@ -873,3 +827,20 @@ exit: out: return ret; } + +int rxe_sender(struct rxe_qp *qp) +{ + int req_ret; + int comp_ret; + + /* process the send queue */ + req_ret = rxe_requester(qp); + + /* process the response queue */ + comp_ret = rxe_completer(qp); + + /* exit the task loop if both requester and completer + * are ready + */ + return (req_ret && comp_ret) ? -EAGAIN : 0; +} diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 64c64f5f36a8..6596a85723c9 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -49,18 +49,8 @@ static char *resp_state_name[] = { /* rxe_recv calls here to add a request packet to the input queue */ void rxe_resp_queue_pkt(struct rxe_qp *qp, struct sk_buff *skb) { - int must_sched; - struct rxe_pkt_info *pkt = SKB_TO_PKT(skb); - skb_queue_tail(&qp->req_pkts, skb); - - must_sched = (pkt->opcode == IB_OPCODE_RC_RDMA_READ_REQUEST) || - (skb_queue_len(&qp->req_pkts) > 1); - - if (must_sched) - rxe_sched_task(&qp->resp.task); - else - rxe_run_task(&qp->resp.task); + rxe_sched_task(&qp->recv_task); } static inline enum resp_states get_req(struct rxe_qp *qp, @@ -354,6 +344,19 @@ static enum resp_states rxe_resp_check_length(struct rxe_qp *qp, * receive buffer later. For rmda operations additional * length checks are performed in check_rkey. */ + if ((qp_type(qp) == IB_QPT_GSI) || (qp_type(qp) == IB_QPT_UD)) { + unsigned int payload = payload_size(pkt); + unsigned int recv_buffer_len = 0; + int i; + + for (i = 0; i < qp->resp.wqe->dma.num_sge; i++) + recv_buffer_len += qp->resp.wqe->dma.sge[i].length; + if (payload + 40 > recv_buffer_len) { + rxe_dbg_qp(qp, "The receive buffer is too small for this UD packet.\n"); + return RESPST_ERR_LENGTH; + } + } + if (pkt->mask & RXE_PAYLOAD_MASK && ((qp_type(qp) == IB_QPT_RC) || (qp_type(qp) == IB_QPT_UC))) { unsigned int mtu = qp->mtu; @@ -362,18 +365,18 @@ static enum resp_states rxe_resp_check_length(struct rxe_qp *qp, if ((pkt->mask & RXE_START_MASK) && (pkt->mask & RXE_END_MASK)) { if (unlikely(payload > mtu)) { - rxe_dbg_qp(qp, "only packet too long"); + rxe_dbg_qp(qp, "only packet too long\n"); return RESPST_ERR_LENGTH; } } else if ((pkt->mask & RXE_START_MASK) || (pkt->mask & RXE_MIDDLE_MASK)) { if (unlikely(payload != mtu)) { - rxe_dbg_qp(qp, "first or middle packet not mtu"); + rxe_dbg_qp(qp, "first or middle packet not mtu\n"); return RESPST_ERR_LENGTH; } } else if (pkt->mask & RXE_END_MASK) { if (unlikely((payload == 0) || (payload > mtu))) { - rxe_dbg_qp(qp, "last packet zero or too long"); + rxe_dbg_qp(qp, "last packet zero or too long\n"); return RESPST_ERR_LENGTH; } } @@ -382,7 +385,7 @@ static enum resp_states rxe_resp_check_length(struct rxe_qp *qp, /* See IBA C9-94 */ if (pkt->mask & RXE_RETH_MASK) { if (reth_len(pkt) > (1U << 31)) { - rxe_dbg_qp(qp, "dma length too long"); + rxe_dbg_qp(qp, "dma length too long\n"); return RESPST_ERR_LENGTH; } } @@ -1133,7 +1136,7 @@ static enum resp_states do_complete(struct rxe_qp *qp, } } else { if (wc->status != IB_WC_WR_FLUSH_ERR) - rxe_err_qp(qp, "non-flush error status = %d", + rxe_err_qp(qp, "non-flush error status = %d\n", wc->status); } @@ -1442,7 +1445,7 @@ static int flush_recv_wqe(struct rxe_qp *qp, struct rxe_recv_wqe *wqe) err = rxe_cq_post(qp->rcq, &cqe, 0); if (err) - rxe_dbg_cq(qp->rcq, "post cq failed err = %d", err); + rxe_dbg_cq(qp->rcq, "post cq failed err = %d\n", err); return err; } @@ -1469,6 +1472,10 @@ static void flush_recv_queue(struct rxe_qp *qp, bool notify) return; } + /* recv queue not created. nothing to do. */ + if (!qp->rq.queue) + return; + while ((wqe = queue_head(q, q->type))) { if (notify) { err = flush_recv_wqe(qp, wqe); @@ -1481,7 +1488,7 @@ static void flush_recv_queue(struct rxe_qp *qp, bool notify) qp->resp.wqe = NULL; } -int rxe_responder(struct rxe_qp *qp) +int rxe_receiver(struct rxe_qp *qp) { struct rxe_dev *rxe = to_rdev(qp->ibqp.device); enum resp_states state; diff --git a/drivers/infiniband/sw/rxe/rxe_srq.c b/drivers/infiniband/sw/rxe/rxe_srq.c index 27ca82ec0826..3661cb627d28 100644 --- a/drivers/infiniband/sw/rxe/rxe_srq.c +++ b/drivers/infiniband/sw/rxe/rxe_srq.c @@ -45,40 +45,41 @@ int rxe_srq_from_init(struct rxe_dev *rxe, struct rxe_srq *srq, struct ib_srq_init_attr *init, struct ib_udata *udata, struct rxe_create_srq_resp __user *uresp) { - int err; - int srq_wqe_size; struct rxe_queue *q; - enum queue_type type; + int wqe_size; + int err; - srq->ibsrq.event_handler = init->event_handler; - srq->ibsrq.srq_context = init->srq_context; - srq->limit = init->attr.srq_limit; - srq->srq_num = srq->elem.index; - srq->rq.max_wr = init->attr.max_wr; - srq->rq.max_sge = init->attr.max_sge; + srq->ibsrq.event_handler = init->event_handler; + srq->ibsrq.srq_context = init->srq_context; + srq->limit = init->attr.srq_limit; + srq->srq_num = srq->elem.index; + srq->rq.max_wr = init->attr.max_wr; + srq->rq.max_sge = init->attr.max_sge; - srq_wqe_size = rcv_wqe_size(srq->rq.max_sge); + wqe_size = sizeof(struct rxe_recv_wqe) + + srq->rq.max_sge*sizeof(struct ib_sge); spin_lock_init(&srq->rq.producer_lock); spin_lock_init(&srq->rq.consumer_lock); - type = QUEUE_TYPE_FROM_CLIENT; - q = rxe_queue_init(rxe, &srq->rq.max_wr, srq_wqe_size, type); + q = rxe_queue_init(rxe, &srq->rq.max_wr, wqe_size, + QUEUE_TYPE_FROM_CLIENT); if (!q) { rxe_dbg_srq(srq, "Unable to allocate queue\n"); - return -ENOMEM; + err = -ENOMEM; + goto err_out; } - srq->rq.queue = q; - err = do_mmap_info(rxe, uresp ? &uresp->mi : NULL, udata, q->buf, q->buf_size, &q->ip); if (err) { - vfree(q->buf); - kfree(q); - return err; + rxe_dbg_srq(srq, "Unable to init mmap info for caller\n"); + goto err_free; } + srq->rq.queue = q; + init->attr.max_wr = srq->rq.max_wr; + if (uresp) { if (copy_to_user(&uresp->srq_num, &srq->srq_num, sizeof(uresp->srq_num))) { @@ -88,6 +89,12 @@ int rxe_srq_from_init(struct rxe_dev *rxe, struct rxe_srq *srq, } return 0; + +err_free: + vfree(q->buf); + kfree(q); +err_out: + return err; } int rxe_srq_chk_attr(struct rxe_dev *rxe, struct rxe_srq *srq, @@ -145,9 +152,10 @@ int rxe_srq_from_attr(struct rxe_dev *rxe, struct rxe_srq *srq, struct ib_srq_attr *attr, enum ib_srq_attr_mask mask, struct rxe_modify_srq_cmd *ucmd, struct ib_udata *udata) { - int err; struct rxe_queue *q = srq->rq.queue; struct mminfo __user *mi = NULL; + int wqe_size; + int err; if (mask & IB_SRQ_MAX_WR) { /* @@ -156,12 +164,16 @@ int rxe_srq_from_attr(struct rxe_dev *rxe, struct rxe_srq *srq, */ mi = u64_to_user_ptr(ucmd->mmap_info_addr); - err = rxe_queue_resize(q, &attr->max_wr, - rcv_wqe_size(srq->rq.max_sge), udata, mi, - &srq->rq.producer_lock, + wqe_size = sizeof(struct rxe_recv_wqe) + + srq->rq.max_sge*sizeof(struct ib_sge); + + err = rxe_queue_resize(q, &attr->max_wr, wqe_size, + udata, mi, &srq->rq.producer_lock, &srq->rq.consumer_lock); if (err) - goto err2; + goto err_free; + + srq->rq.max_wr = attr->max_wr; } if (mask & IB_SRQ_LIMIT) @@ -169,7 +181,7 @@ int rxe_srq_from_attr(struct rxe_dev *rxe, struct rxe_srq *srq, return 0; -err2: +err_free: rxe_queue_cleanup(q); srq->rq.queue = NULL; return err; diff --git a/drivers/infiniband/sw/rxe/rxe_task.c b/drivers/infiniband/sw/rxe/rxe_task.c index 1501120d4f52..80332638d9e3 100644 --- a/drivers/infiniband/sw/rxe/rxe_task.c +++ b/drivers/infiniband/sw/rxe/rxe_task.c @@ -156,7 +156,7 @@ static void do_task(struct rxe_task *task) default: WARN_ON(1); - rxe_dbg_qp(task->qp, "unexpected task state = %d", + rxe_dbg_qp(task->qp, "unexpected task state = %d\n", task->state); task->state = TASK_STATE_IDLE; } @@ -167,7 +167,7 @@ exit: if (WARN_ON(task->num_done != task->num_sched)) rxe_dbg_qp( task->qp, - "%ld tasks scheduled, %ld tasks done", + "%ld tasks scheduled, %ld tasks done\n", task->num_sched, task->num_done); } spin_unlock_irqrestore(&task->lock, flags); diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c index 903f0b71447e..de6238ee4379 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.c +++ b/drivers/infiniband/sw/rxe/rxe_verbs.c @@ -23,7 +23,7 @@ static int rxe_query_device(struct ib_device *ibdev, int err; if (udata->inlen || udata->outlen) { - rxe_dbg_dev(rxe, "malformed udata"); + rxe_dbg_dev(rxe, "malformed udata\n"); err = -EINVAL; goto err_out; } @@ -33,7 +33,7 @@ static int rxe_query_device(struct ib_device *ibdev, return 0; err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -45,7 +45,7 @@ static int rxe_query_port(struct ib_device *ibdev, if (port_num != 1) { err = -EINVAL; - rxe_dbg_dev(rxe, "bad port_num = %d", port_num); + rxe_dbg_dev(rxe, "bad port_num = %d\n", port_num); goto err_out; } @@ -67,7 +67,7 @@ static int rxe_query_port(struct ib_device *ibdev, return ret; err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -79,7 +79,7 @@ static int rxe_query_pkey(struct ib_device *ibdev, if (index != 0) { err = -EINVAL; - rxe_dbg_dev(rxe, "bad pkey index = %d", index); + rxe_dbg_dev(rxe, "bad pkey index = %d\n", index); goto err_out; } @@ -87,7 +87,7 @@ static int rxe_query_pkey(struct ib_device *ibdev, return 0; err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -100,7 +100,7 @@ static int rxe_modify_device(struct ib_device *ibdev, if (mask & ~(IB_DEVICE_MODIFY_SYS_IMAGE_GUID | IB_DEVICE_MODIFY_NODE_DESC)) { err = -EOPNOTSUPP; - rxe_dbg_dev(rxe, "unsupported mask = 0x%x", mask); + rxe_dbg_dev(rxe, "unsupported mask = 0x%x\n", mask); goto err_out; } @@ -115,7 +115,7 @@ static int rxe_modify_device(struct ib_device *ibdev, return 0; err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -128,14 +128,14 @@ static int rxe_modify_port(struct ib_device *ibdev, u32 port_num, if (port_num != 1) { err = -EINVAL; - rxe_dbg_dev(rxe, "bad port_num = %d", port_num); + rxe_dbg_dev(rxe, "bad port_num = %d\n", port_num); goto err_out; } //TODO is shutdown useful if (mask & ~(IB_PORT_RESET_QKEY_CNTR)) { err = -EOPNOTSUPP; - rxe_dbg_dev(rxe, "unsupported mask = 0x%x", mask); + rxe_dbg_dev(rxe, "unsupported mask = 0x%x\n", mask); goto err_out; } @@ -149,7 +149,7 @@ static int rxe_modify_port(struct ib_device *ibdev, u32 port_num, return 0; err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -161,14 +161,14 @@ static enum rdma_link_layer rxe_get_link_layer(struct ib_device *ibdev, if (port_num != 1) { err = -EINVAL; - rxe_dbg_dev(rxe, "bad port_num = %d", port_num); + rxe_dbg_dev(rxe, "bad port_num = %d\n", port_num); goto err_out; } return IB_LINK_LAYER_ETHERNET; err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -181,7 +181,7 @@ static int rxe_port_immutable(struct ib_device *ibdev, u32 port_num, if (port_num != 1) { err = -EINVAL; - rxe_dbg_dev(rxe, "bad port_num = %d", port_num); + rxe_dbg_dev(rxe, "bad port_num = %d\n", port_num); goto err_out; } @@ -197,7 +197,7 @@ static int rxe_port_immutable(struct ib_device *ibdev, u32 port_num, return 0; err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -210,7 +210,7 @@ static int rxe_alloc_ucontext(struct ib_ucontext *ibuc, struct ib_udata *udata) err = rxe_add_to_pool(&rxe->uc_pool, uc); if (err) - rxe_err_dev(rxe, "unable to create uc"); + rxe_err_dev(rxe, "unable to create uc\n"); return err; } @@ -222,7 +222,7 @@ static void rxe_dealloc_ucontext(struct ib_ucontext *ibuc) err = rxe_cleanup(uc); if (err) - rxe_err_uc(uc, "cleanup failed, err = %d", err); + rxe_err_uc(uc, "cleanup failed, err = %d\n", err); } /* pd */ @@ -234,14 +234,14 @@ static int rxe_alloc_pd(struct ib_pd *ibpd, struct ib_udata *udata) err = rxe_add_to_pool(&rxe->pd_pool, pd); if (err) { - rxe_dbg_dev(rxe, "unable to alloc pd"); + rxe_dbg_dev(rxe, "unable to alloc pd\n"); goto err_out; } return 0; err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -252,7 +252,7 @@ static int rxe_dealloc_pd(struct ib_pd *ibpd, struct ib_udata *udata) err = rxe_cleanup(pd); if (err) - rxe_err_pd(pd, "cleanup failed, err = %d", err); + rxe_err_pd(pd, "cleanup failed, err = %d\n", err); return 0; } @@ -279,7 +279,7 @@ static int rxe_create_ah(struct ib_ah *ibah, err = rxe_add_to_pool_ah(&rxe->ah_pool, ah, init_attr->flags & RDMA_CREATE_AH_SLEEPABLE); if (err) { - rxe_dbg_dev(rxe, "unable to create ah"); + rxe_dbg_dev(rxe, "unable to create ah\n"); goto err_out; } @@ -288,7 +288,7 @@ static int rxe_create_ah(struct ib_ah *ibah, err = rxe_ah_chk_attr(ah, init_attr->ah_attr); if (err) { - rxe_dbg_ah(ah, "bad attr"); + rxe_dbg_ah(ah, "bad attr\n"); goto err_cleanup; } @@ -298,7 +298,7 @@ static int rxe_create_ah(struct ib_ah *ibah, sizeof(uresp->ah_num)); if (err) { err = -EFAULT; - rxe_dbg_ah(ah, "unable to copy to user"); + rxe_dbg_ah(ah, "unable to copy to user\n"); goto err_cleanup; } } else if (ah->is_user) { @@ -314,9 +314,9 @@ static int rxe_create_ah(struct ib_ah *ibah, err_cleanup: cleanup_err = rxe_cleanup(ah); if (cleanup_err) - rxe_err_ah(ah, "cleanup failed, err = %d", cleanup_err); + rxe_err_ah(ah, "cleanup failed, err = %d\n", cleanup_err); err_out: - rxe_err_ah(ah, "returned err = %d", err); + rxe_err_ah(ah, "returned err = %d\n", err); return err; } @@ -327,7 +327,7 @@ static int rxe_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *attr) err = rxe_ah_chk_attr(ah, attr); if (err) { - rxe_dbg_ah(ah, "bad attr"); + rxe_dbg_ah(ah, "bad attr\n"); goto err_out; } @@ -336,7 +336,7 @@ static int rxe_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *attr) return 0; err_out: - rxe_err_ah(ah, "returned err = %d", err); + rxe_err_ah(ah, "returned err = %d\n", err); return err; } @@ -358,7 +358,7 @@ static int rxe_destroy_ah(struct ib_ah *ibah, u32 flags) err = rxe_cleanup_ah(ah, flags & RDMA_DESTROY_AH_SLEEPABLE); if (err) - rxe_err_ah(ah, "cleanup failed, err = %d", err); + rxe_err_ah(ah, "cleanup failed, err = %d\n", err); return 0; } @@ -376,7 +376,7 @@ static int rxe_create_srq(struct ib_srq *ibsrq, struct ib_srq_init_attr *init, if (udata) { if (udata->outlen < sizeof(*uresp)) { err = -EINVAL; - rxe_err_dev(rxe, "malformed udata"); + rxe_err_dev(rxe, "malformed udata\n"); goto err_out; } uresp = udata->outbuf; @@ -384,20 +384,20 @@ static int rxe_create_srq(struct ib_srq *ibsrq, struct ib_srq_init_attr *init, if (init->srq_type != IB_SRQT_BASIC) { err = -EOPNOTSUPP; - rxe_dbg_dev(rxe, "srq type = %d, not supported", + rxe_dbg_dev(rxe, "srq type = %d, not supported\n", init->srq_type); goto err_out; } err = rxe_srq_chk_init(rxe, init); if (err) { - rxe_dbg_dev(rxe, "invalid init attributes"); + rxe_dbg_dev(rxe, "invalid init attributes\n"); goto err_out; } err = rxe_add_to_pool(&rxe->srq_pool, srq); if (err) { - rxe_dbg_dev(rxe, "unable to create srq, err = %d", err); + rxe_dbg_dev(rxe, "unable to create srq, err = %d\n", err); goto err_out; } @@ -406,7 +406,7 @@ static int rxe_create_srq(struct ib_srq *ibsrq, struct ib_srq_init_attr *init, err = rxe_srq_from_init(rxe, srq, init, udata, uresp); if (err) { - rxe_dbg_srq(srq, "create srq failed, err = %d", err); + rxe_dbg_srq(srq, "create srq failed, err = %d\n", err); goto err_cleanup; } @@ -415,9 +415,9 @@ static int rxe_create_srq(struct ib_srq *ibsrq, struct ib_srq_init_attr *init, err_cleanup: cleanup_err = rxe_cleanup(srq); if (cleanup_err) - rxe_err_srq(srq, "cleanup failed, err = %d", cleanup_err); + rxe_err_srq(srq, "cleanup failed, err = %d\n", cleanup_err); err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -433,34 +433,34 @@ static int rxe_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr, if (udata) { if (udata->inlen < sizeof(cmd)) { err = -EINVAL; - rxe_dbg_srq(srq, "malformed udata"); + rxe_dbg_srq(srq, "malformed udata\n"); goto err_out; } err = ib_copy_from_udata(&cmd, udata, sizeof(cmd)); if (err) { err = -EFAULT; - rxe_dbg_srq(srq, "unable to read udata"); + rxe_dbg_srq(srq, "unable to read udata\n"); goto err_out; } } err = rxe_srq_chk_attr(rxe, srq, attr, mask); if (err) { - rxe_dbg_srq(srq, "bad init attributes"); + rxe_dbg_srq(srq, "bad init attributes\n"); goto err_out; } err = rxe_srq_from_attr(rxe, srq, attr, mask, &cmd, udata); if (err) { - rxe_dbg_srq(srq, "bad attr"); + rxe_dbg_srq(srq, "bad attr\n"); goto err_out; } return 0; err_out: - rxe_err_srq(srq, "returned err = %d", err); + rxe_err_srq(srq, "returned err = %d\n", err); return err; } @@ -471,7 +471,7 @@ static int rxe_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr) if (srq->error) { err = -EINVAL; - rxe_dbg_srq(srq, "srq in error state"); + rxe_dbg_srq(srq, "srq in error state\n"); goto err_out; } @@ -481,7 +481,7 @@ static int rxe_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr) return 0; err_out: - rxe_err_srq(srq, "returned err = %d", err); + rxe_err_srq(srq, "returned err = %d\n", err); return err; } @@ -505,7 +505,7 @@ static int rxe_post_srq_recv(struct ib_srq *ibsrq, const struct ib_recv_wr *wr, if (err) { *bad_wr = wr; - rxe_err_srq(srq, "returned err = %d", err); + rxe_err_srq(srq, "returned err = %d\n", err); } return err; @@ -518,7 +518,7 @@ static int rxe_destroy_srq(struct ib_srq *ibsrq, struct ib_udata *udata) err = rxe_cleanup(srq); if (err) - rxe_err_srq(srq, "cleanup failed, err = %d", err); + rxe_err_srq(srq, "cleanup failed, err = %d\n", err); return 0; } @@ -536,13 +536,13 @@ static int rxe_create_qp(struct ib_qp *ibqp, struct ib_qp_init_attr *init, if (udata) { if (udata->inlen) { err = -EINVAL; - rxe_dbg_dev(rxe, "malformed udata, err = %d", err); + rxe_dbg_dev(rxe, "malformed udata, err = %d\n", err); goto err_out; } if (udata->outlen < sizeof(*uresp)) { err = -EINVAL; - rxe_dbg_dev(rxe, "malformed udata, err = %d", err); + rxe_dbg_dev(rxe, "malformed udata, err = %d\n", err); goto err_out; } @@ -554,25 +554,25 @@ static int rxe_create_qp(struct ib_qp *ibqp, struct ib_qp_init_attr *init, if (init->create_flags) { err = -EOPNOTSUPP; - rxe_dbg_dev(rxe, "unsupported create_flags, err = %d", err); + rxe_dbg_dev(rxe, "unsupported create_flags, err = %d\n", err); goto err_out; } err = rxe_qp_chk_init(rxe, init); if (err) { - rxe_dbg_dev(rxe, "bad init attr, err = %d", err); + rxe_dbg_dev(rxe, "bad init attr, err = %d\n", err); goto err_out; } err = rxe_add_to_pool(&rxe->qp_pool, qp); if (err) { - rxe_dbg_dev(rxe, "unable to create qp, err = %d", err); + rxe_dbg_dev(rxe, "unable to create qp, err = %d\n", err); goto err_out; } err = rxe_qp_from_init(rxe, qp, pd, init, uresp, ibqp->pd, udata); if (err) { - rxe_dbg_qp(qp, "create qp failed, err = %d", err); + rxe_dbg_qp(qp, "create qp failed, err = %d\n", err); goto err_cleanup; } @@ -582,9 +582,9 @@ static int rxe_create_qp(struct ib_qp *ibqp, struct ib_qp_init_attr *init, err_cleanup: cleanup_err = rxe_cleanup(qp); if (cleanup_err) - rxe_err_qp(qp, "cleanup failed, err = %d", cleanup_err); + rxe_err_qp(qp, "cleanup failed, err = %d\n", cleanup_err); err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -597,20 +597,20 @@ static int rxe_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, if (mask & ~IB_QP_ATTR_STANDARD_BITS) { err = -EOPNOTSUPP; - rxe_dbg_qp(qp, "unsupported mask = 0x%x, err = %d", + rxe_dbg_qp(qp, "unsupported mask = 0x%x, err = %d\n", mask, err); goto err_out; } err = rxe_qp_chk_attr(rxe, qp, attr, mask); if (err) { - rxe_dbg_qp(qp, "bad mask/attr, err = %d", err); + rxe_dbg_qp(qp, "bad mask/attr, err = %d\n", err); goto err_out; } err = rxe_qp_from_attr(qp, attr, mask, udata); if (err) { - rxe_dbg_qp(qp, "modify qp failed, err = %d", err); + rxe_dbg_qp(qp, "modify qp failed, err = %d\n", err); goto err_out; } @@ -622,7 +622,7 @@ static int rxe_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, return 0; err_out: - rxe_err_qp(qp, "returned err = %d", err); + rxe_err_qp(qp, "returned err = %d\n", err); return err; } @@ -644,18 +644,18 @@ static int rxe_destroy_qp(struct ib_qp *ibqp, struct ib_udata *udata) err = rxe_qp_chk_destroy(qp); if (err) { - rxe_dbg_qp(qp, "unable to destroy qp, err = %d", err); + rxe_dbg_qp(qp, "unable to destroy qp, err = %d\n", err); goto err_out; } err = rxe_cleanup(qp); if (err) - rxe_err_qp(qp, "cleanup failed, err = %d", err); + rxe_err_qp(qp, "cleanup failed, err = %d\n", err); return 0; err_out: - rxe_err_qp(qp, "returned err = %d", err); + rxe_err_qp(qp, "returned err = %d\n", err); return err; } @@ -675,12 +675,12 @@ static int validate_send_wr(struct rxe_qp *qp, const struct ib_send_wr *ibwr, do { mask = wr_opcode_mask(ibwr->opcode, qp); if (!mask) { - rxe_err_qp(qp, "bad wr opcode for qp type"); + rxe_err_qp(qp, "bad wr opcode for qp type\n"); break; } if (num_sge > sq->max_sge) { - rxe_err_qp(qp, "num_sge > max_sge"); + rxe_err_qp(qp, "num_sge > max_sge\n"); break; } @@ -689,27 +689,27 @@ static int validate_send_wr(struct rxe_qp *qp, const struct ib_send_wr *ibwr, length += ibwr->sg_list[i].length; if (length > (1UL << 31)) { - rxe_err_qp(qp, "message length too long"); + rxe_err_qp(qp, "message length too long\n"); break; } if (mask & WR_ATOMIC_MASK) { if (length != 8) { - rxe_err_qp(qp, "atomic length != 8"); + rxe_err_qp(qp, "atomic length != 8\n"); break; } if (atomic_wr(ibwr)->remote_addr & 0x7) { - rxe_err_qp(qp, "misaligned atomic address"); + rxe_err_qp(qp, "misaligned atomic address\n"); break; } } if (ibwr->send_flags & IB_SEND_INLINE) { if (!(mask & WR_INLINE_MASK)) { - rxe_err_qp(qp, "opcode doesn't support inline data"); + rxe_err_qp(qp, "opcode doesn't support inline data\n"); break; } if (length > sq->max_inline) { - rxe_err_qp(qp, "inline length too big"); + rxe_err_qp(qp, "inline length too big\n"); break; } } @@ -747,7 +747,7 @@ static int init_send_wr(struct rxe_qp *qp, struct rxe_send_wr *wr, case IB_WR_SEND: break; default: - rxe_err_qp(qp, "bad wr opcode %d for UD/GSI QP", + rxe_err_qp(qp, "bad wr opcode %d for UD/GSI QP\n", wr->opcode); return -EINVAL; } @@ -795,10 +795,9 @@ static int init_send_wr(struct rxe_qp *qp, struct rxe_send_wr *wr, case IB_WR_ATOMIC_WRITE: break; default: - rxe_err_qp(qp, "unsupported wr opcode %d", + rxe_err_qp(qp, "unsupported wr opcode %d\n", wr->opcode); return -EINVAL; - break; } } @@ -813,7 +812,7 @@ static void copy_inline_data_to_wqe(struct rxe_send_wqe *wqe, int i; for (i = 0; i < ibwr->num_sge; i++, sge++) { - memcpy(p, ib_virt_dma_to_page(sge->addr), sge->length); + memcpy(p, ib_virt_dma_to_ptr(sge->addr), sge->length); p += sge->length; } } @@ -871,7 +870,7 @@ static int post_one_send(struct rxe_qp *qp, const struct ib_send_wr *ibwr) full = queue_full(sq->queue, QUEUE_TYPE_FROM_ULP); if (unlikely(full)) { - rxe_err_qp(qp, "send queue full"); + rxe_err_qp(qp, "send queue full\n"); return -ENOMEM; } @@ -889,6 +888,7 @@ static int rxe_post_send_kernel(struct rxe_qp *qp, { int err = 0; unsigned long flags; + int good = 0; spin_lock_irqsave(&qp->sq.sq_lock, flags); while (ibwr) { @@ -896,18 +896,16 @@ static int rxe_post_send_kernel(struct rxe_qp *qp, if (err) { *bad_wr = ibwr; break; + } else { + good++; } ibwr = ibwr->next; } spin_unlock_irqrestore(&qp->sq.sq_lock, flags); - if (!err) - rxe_sched_task(&qp->req.task); - - spin_lock_irqsave(&qp->state_lock, flags); - if (qp_state(qp) == IB_QPS_ERR) - rxe_sched_task(&qp->comp.task); - spin_unlock_irqrestore(&qp->state_lock, flags); + /* kickoff processing of any posted wqes */ + if (good) + rxe_sched_task(&qp->send_task); return err; } @@ -923,21 +921,21 @@ static int rxe_post_send(struct ib_qp *ibqp, const struct ib_send_wr *wr, /* caller has already called destroy_qp */ if (WARN_ON_ONCE(!qp->valid)) { spin_unlock_irqrestore(&qp->state_lock, flags); - rxe_err_qp(qp, "qp has been destroyed"); + rxe_err_qp(qp, "qp has been destroyed\n"); return -EINVAL; } if (unlikely(qp_state(qp) < IB_QPS_RTS)) { spin_unlock_irqrestore(&qp->state_lock, flags); *bad_wr = wr; - rxe_err_qp(qp, "qp not ready to send"); + rxe_err_qp(qp, "qp not ready to send\n"); return -EINVAL; } spin_unlock_irqrestore(&qp->state_lock, flags); if (qp->is_user) { /* Utilize process context to do protocol processing */ - rxe_run_task(&qp->req.task); + rxe_sched_task(&qp->send_task); } else { err = rxe_post_send_kernel(qp, wr, bad_wr); if (err) @@ -960,13 +958,13 @@ static int post_one_recv(struct rxe_rq *rq, const struct ib_recv_wr *ibwr) full = queue_full(rq->queue, QUEUE_TYPE_FROM_ULP); if (unlikely(full)) { err = -ENOMEM; - rxe_dbg("queue full"); + rxe_dbg("queue full\n"); goto err_out; } if (unlikely(num_sge > rq->max_sge)) { err = -EINVAL; - rxe_dbg("bad num_sge > max_sge"); + rxe_dbg("bad num_sge > max_sge\n"); goto err_out; } @@ -977,7 +975,7 @@ static int post_one_recv(struct rxe_rq *rq, const struct ib_recv_wr *ibwr) /* IBA max message size is 2^31 */ if (length >= (1UL<<31)) { err = -EINVAL; - rxe_dbg("message length too long"); + rxe_dbg("message length too long\n"); goto err_out; } @@ -997,7 +995,7 @@ static int post_one_recv(struct rxe_rq *rq, const struct ib_recv_wr *ibwr) return 0; err_out: - rxe_dbg("returned err = %d", err); + rxe_dbg("returned err = %d\n", err); return err; } @@ -1013,7 +1011,7 @@ static int rxe_post_recv(struct ib_qp *ibqp, const struct ib_recv_wr *wr, /* caller has already called destroy_qp */ if (WARN_ON_ONCE(!qp->valid)) { spin_unlock_irqrestore(&qp->state_lock, flags); - rxe_err_qp(qp, "qp has been destroyed"); + rxe_err_qp(qp, "qp has been destroyed\n"); return -EINVAL; } @@ -1021,14 +1019,14 @@ static int rxe_post_recv(struct ib_qp *ibqp, const struct ib_recv_wr *wr, if (unlikely((qp_state(qp) < IB_QPS_INIT))) { spin_unlock_irqrestore(&qp->state_lock, flags); *bad_wr = wr; - rxe_dbg_qp(qp, "qp not ready to post recv"); + rxe_dbg_qp(qp, "qp not ready to post recv\n"); return -EINVAL; } spin_unlock_irqrestore(&qp->state_lock, flags); if (unlikely(qp->srq)) { *bad_wr = wr; - rxe_dbg_qp(qp, "qp has srq, use post_srq_recv instead"); + rxe_dbg_qp(qp, "qp has srq, use post_srq_recv instead\n"); return -EINVAL; } @@ -1047,7 +1045,7 @@ static int rxe_post_recv(struct ib_qp *ibqp, const struct ib_recv_wr *wr, spin_lock_irqsave(&qp->state_lock, flags); if (qp_state(qp) == IB_QPS_ERR) - rxe_sched_task(&qp->resp.task); + rxe_sched_task(&qp->recv_task); spin_unlock_irqrestore(&qp->state_lock, flags); return err; @@ -1066,7 +1064,7 @@ static int rxe_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, if (udata) { if (udata->outlen < sizeof(*uresp)) { err = -EINVAL; - rxe_dbg_dev(rxe, "malformed udata, err = %d", err); + rxe_dbg_dev(rxe, "malformed udata, err = %d\n", err); goto err_out; } uresp = udata->outbuf; @@ -1074,26 +1072,26 @@ static int rxe_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, if (attr->flags) { err = -EOPNOTSUPP; - rxe_dbg_dev(rxe, "bad attr->flags, err = %d", err); + rxe_dbg_dev(rxe, "bad attr->flags, err = %d\n", err); goto err_out; } err = rxe_cq_chk_attr(rxe, NULL, attr->cqe, attr->comp_vector); if (err) { - rxe_dbg_dev(rxe, "bad init attributes, err = %d", err); + rxe_dbg_dev(rxe, "bad init attributes, err = %d\n", err); goto err_out; } err = rxe_add_to_pool(&rxe->cq_pool, cq); if (err) { - rxe_dbg_dev(rxe, "unable to create cq, err = %d", err); + rxe_dbg_dev(rxe, "unable to create cq, err = %d\n", err); goto err_out; } err = rxe_cq_from_init(rxe, cq, attr->cqe, attr->comp_vector, udata, uresp); if (err) { - rxe_dbg_cq(cq, "create cq failed, err = %d", err); + rxe_dbg_cq(cq, "create cq failed, err = %d\n", err); goto err_cleanup; } @@ -1102,9 +1100,9 @@ static int rxe_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, err_cleanup: cleanup_err = rxe_cleanup(cq); if (cleanup_err) - rxe_err_cq(cq, "cleanup failed, err = %d", cleanup_err); + rxe_err_cq(cq, "cleanup failed, err = %d\n", cleanup_err); err_out: - rxe_err_dev(rxe, "returned err = %d", err); + rxe_err_dev(rxe, "returned err = %d\n", err); return err; } @@ -1118,7 +1116,7 @@ static int rxe_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata) if (udata) { if (udata->outlen < sizeof(*uresp)) { err = -EINVAL; - rxe_dbg_cq(cq, "malformed udata"); + rxe_dbg_cq(cq, "malformed udata\n"); goto err_out; } uresp = udata->outbuf; @@ -1126,20 +1124,20 @@ static int rxe_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata) err = rxe_cq_chk_attr(rxe, cq, cqe, 0); if (err) { - rxe_dbg_cq(cq, "bad attr, err = %d", err); + rxe_dbg_cq(cq, "bad attr, err = %d\n", err); goto err_out; } err = rxe_cq_resize_queue(cq, cqe, uresp, udata); if (err) { - rxe_dbg_cq(cq, "resize cq failed, err = %d", err); + rxe_dbg_cq(cq, "resize cq failed, err = %d\n", err); goto err_out; } return 0; err_out: - rxe_err_cq(cq, "returned err = %d", err); + rxe_err_cq(cq, "returned err = %d\n", err); return err; } @@ -1203,18 +1201,18 @@ static int rxe_destroy_cq(struct ib_cq *ibcq, struct ib_udata *udata) */ if (atomic_read(&cq->num_wq)) { err = -EINVAL; - rxe_dbg_cq(cq, "still in use"); + rxe_dbg_cq(cq, "still in use\n"); goto err_out; } err = rxe_cleanup(cq); if (err) - rxe_err_cq(cq, "cleanup failed, err = %d", err); + rxe_err_cq(cq, "cleanup failed, err = %d\n", err); return 0; err_out: - rxe_err_cq(cq, "returned err = %d", err); + rxe_err_cq(cq, "returned err = %d\n", err); return err; } @@ -1232,7 +1230,7 @@ static struct ib_mr *rxe_get_dma_mr(struct ib_pd *ibpd, int access) err = rxe_add_to_pool(&rxe->mr_pool, mr); if (err) { - rxe_dbg_dev(rxe, "unable to create mr"); + rxe_dbg_dev(rxe, "unable to create mr\n"); goto err_free; } @@ -1246,7 +1244,7 @@ static struct ib_mr *rxe_get_dma_mr(struct ib_pd *ibpd, int access) err_free: kfree(mr); - rxe_err_pd(pd, "returned err = %d", err); + rxe_err_pd(pd, "returned err = %d\n", err); return ERR_PTR(err); } @@ -1260,7 +1258,7 @@ static struct ib_mr *rxe_reg_user_mr(struct ib_pd *ibpd, u64 start, int err, cleanup_err; if (access & ~RXE_ACCESS_SUPPORTED_MR) { - rxe_err_pd(pd, "access = %#x not supported (%#x)", access, + rxe_err_pd(pd, "access = %#x not supported (%#x)\n", access, RXE_ACCESS_SUPPORTED_MR); return ERR_PTR(-EOPNOTSUPP); } @@ -1271,7 +1269,7 @@ static struct ib_mr *rxe_reg_user_mr(struct ib_pd *ibpd, u64 start, err = rxe_add_to_pool(&rxe->mr_pool, mr); if (err) { - rxe_dbg_pd(pd, "unable to create mr"); + rxe_dbg_pd(pd, "unable to create mr\n"); goto err_free; } @@ -1279,9 +1277,9 @@ static struct ib_mr *rxe_reg_user_mr(struct ib_pd *ibpd, u64 start, mr->ibmr.pd = ibpd; mr->ibmr.device = ibpd->device; - err = rxe_mr_init_user(rxe, start, length, iova, access, mr); + err = rxe_mr_init_user(rxe, start, length, access, mr); if (err) { - rxe_dbg_mr(mr, "reg_user_mr failed, err = %d", err); + rxe_dbg_mr(mr, "reg_user_mr failed, err = %d\n", err); goto err_cleanup; } @@ -1291,10 +1289,10 @@ static struct ib_mr *rxe_reg_user_mr(struct ib_pd *ibpd, u64 start, err_cleanup: cleanup_err = rxe_cleanup(mr); if (cleanup_err) - rxe_err_mr(mr, "cleanup failed, err = %d", cleanup_err); + rxe_err_mr(mr, "cleanup failed, err = %d\n", cleanup_err); err_free: kfree(mr); - rxe_err_pd(pd, "returned err = %d", err); + rxe_err_pd(pd, "returned err = %d\n", err); return ERR_PTR(err); } @@ -1311,7 +1309,7 @@ static struct ib_mr *rxe_rereg_user_mr(struct ib_mr *ibmr, int flags, * rereg_pd and rereg_access */ if (flags & ~RXE_MR_REREG_SUPPORTED) { - rxe_err_mr(mr, "flags = %#x not supported", flags); + rxe_err_mr(mr, "flags = %#x not supported\n", flags); return ERR_PTR(-EOPNOTSUPP); } @@ -1323,7 +1321,7 @@ static struct ib_mr *rxe_rereg_user_mr(struct ib_mr *ibmr, int flags, if (flags & IB_MR_REREG_ACCESS) { if (access & ~RXE_ACCESS_SUPPORTED_MR) { - rxe_err_mr(mr, "access = %#x not supported", access); + rxe_err_mr(mr, "access = %#x not supported\n", access); return ERR_PTR(-EOPNOTSUPP); } mr->access = access; @@ -1342,7 +1340,7 @@ static struct ib_mr *rxe_alloc_mr(struct ib_pd *ibpd, enum ib_mr_type mr_type, if (mr_type != IB_MR_TYPE_MEM_REG) { err = -EINVAL; - rxe_dbg_pd(pd, "mr type %d not supported, err = %d", + rxe_dbg_pd(pd, "mr type %d not supported, err = %d\n", mr_type, err); goto err_out; } @@ -1361,7 +1359,7 @@ static struct ib_mr *rxe_alloc_mr(struct ib_pd *ibpd, enum ib_mr_type mr_type, err = rxe_mr_init_fast(max_num_sg, mr); if (err) { - rxe_dbg_mr(mr, "alloc_mr failed, err = %d", err); + rxe_dbg_mr(mr, "alloc_mr failed, err = %d\n", err); goto err_cleanup; } @@ -1371,11 +1369,11 @@ static struct ib_mr *rxe_alloc_mr(struct ib_pd *ibpd, enum ib_mr_type mr_type, err_cleanup: cleanup_err = rxe_cleanup(mr); if (cleanup_err) - rxe_err_mr(mr, "cleanup failed, err = %d", err); + rxe_err_mr(mr, "cleanup failed, err = %d\n", err); err_free: kfree(mr); err_out: - rxe_err_pd(pd, "returned err = %d", err); + rxe_err_pd(pd, "returned err = %d\n", err); return ERR_PTR(err); } @@ -1387,19 +1385,19 @@ static int rxe_dereg_mr(struct ib_mr *ibmr, struct ib_udata *udata) /* See IBA 10.6.7.2.6 */ if (atomic_read(&mr->num_mw) > 0) { err = -EINVAL; - rxe_dbg_mr(mr, "mr has mw's bound"); + rxe_dbg_mr(mr, "mr has mw's bound\n"); goto err_out; } cleanup_err = rxe_cleanup(mr); if (cleanup_err) - rxe_err_mr(mr, "cleanup failed, err = %d", cleanup_err); + rxe_err_mr(mr, "cleanup failed, err = %d\n", cleanup_err); kfree_rcu_mightsleep(mr); return 0; err_out: - rxe_err_mr(mr, "returned err = %d", err); + rxe_err_mr(mr, "returned err = %d\n", err); return err; } diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.h b/drivers/infiniband/sw/rxe/rxe_verbs.h index ccb9d19ffe8a..3c1354f82283 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.h +++ b/drivers/infiniband/sw/rxe/rxe_verbs.h @@ -113,7 +113,7 @@ struct rxe_req_info { int need_retry; int wait_for_rnr_timer; int noack_pkts; - struct rxe_task task; + int again; }; struct rxe_comp_info { @@ -124,7 +124,6 @@ struct rxe_comp_info { int started_retry; u32 retry_cnt; u32 rnr_retry; - struct rxe_task task; }; enum rdatm_res_state { @@ -196,7 +195,6 @@ struct rxe_resp_info { unsigned int res_head; unsigned int res_tail; struct resp_res *res; - struct rxe_task task; }; struct rxe_qp { @@ -229,6 +227,9 @@ struct rxe_qp { struct sk_buff_head req_pkts; struct sk_buff_head resp_pkts; + struct rxe_task send_task; + struct rxe_task recv_task; + struct rxe_req_info req; struct rxe_comp_info comp; struct rxe_resp_info resp; diff --git a/drivers/infiniband/sw/siw/iwarp.h b/drivers/infiniband/sw/siw/iwarp.h index 3f1dedb50a0d..8cf69309827d 100644 --- a/drivers/infiniband/sw/siw/iwarp.h +++ b/drivers/infiniband/sw/siw/iwarp.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ diff --git a/drivers/infiniband/sw/siw/siw.h b/drivers/infiniband/sw/siw/siw.h index 2f3a9cda3850..75253f2b3e3d 100644 --- a/drivers/infiniband/sw/siw/siw.h +++ b/drivers/infiniband/sw/siw/siw.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ @@ -74,6 +74,7 @@ struct siw_device { u32 vendor_part_id; int numa_node; + char raw_gid[ETH_ALEN]; /* physical port state (only one port per device) */ enum ib_port_state state; @@ -120,11 +121,10 @@ struct siw_page_chunk { }; struct siw_umem { + struct ib_umem *base_mem; struct siw_page_chunk *page_chunk; int num_pages; - bool writable; u64 fp_addr; /* First page base address */ - struct mm_struct *owning_mm; }; struct siw_pble { @@ -136,7 +136,7 @@ struct siw_pble { struct siw_pbl { unsigned int num_buf; unsigned int max_buf; - struct siw_pble pbe[]; + struct siw_pble pbe[] __counted_by(max_buf); }; /* @@ -288,10 +288,11 @@ struct siw_rx_stream { int skb_offset; /* offset in skb */ int skb_copied; /* processed bytes in skb */ + enum siw_rx_state state; + union iwarp_hdr hdr; struct mpa_trailer trailer; - - enum siw_rx_state state; + struct shash_desc *mpa_crc_hd; /* * For each FPDU, main RX loop runs through 3 stages: @@ -313,7 +314,6 @@ struct siw_rx_stream { u64 ddp_to; u32 inval_stag; /* Stag to be invalidated */ - struct shash_desc *mpa_crc_hd; u8 rx_suspend : 1; u8 pad : 2; /* # of pad bytes expected */ u8 rdmap_op : 4; /* opcode of current frame */ @@ -417,10 +417,10 @@ struct siw_iwarp_tx { struct siw_qp { struct ib_qp base_qp; struct siw_device *sdev; + int tx_cpu; struct kref ref; struct completion qp_free; struct list_head devq; - int tx_cpu; struct siw_qp_attrs attrs; struct siw_cep *cep; @@ -465,7 +465,6 @@ struct siw_qp { } term_info; struct rdma_user_mmap_entry *sq_entry; /* mmap info for SQE array */ struct rdma_user_mmap_entry *rq_entry; /* mmap info for RQE array */ - struct rcu_head rcu; }; /* helper macros */ @@ -530,11 +529,12 @@ void siw_qp_llp_data_ready(struct sock *sk); void siw_qp_llp_write_space(struct sock *sk); /* QP TX path functions */ +int siw_create_tx_threads(void); +void siw_stop_tx_threads(void); int siw_run_sq(void *arg); int siw_qp_sq_process(struct siw_qp *qp); int siw_sq_start(struct siw_qp *qp); int siw_activate_tx(struct siw_qp *qp); -void siw_stop_tx_thread(int nr_cpu); int siw_get_tx_cpu(struct siw_device *sdev); void siw_put_tx_cpu(int cpu); @@ -657,7 +657,7 @@ static inline struct siw_sqe *orq_get_free(struct siw_qp *qp) static inline int siw_orq_empty(struct siw_qp *qp) { - return qp->orq[qp->orq_get % qp->attrs.orq_size].flags == 0 ? 1 : 0; + return orq_get_current(qp)->flags == 0 ? 1 : 0; } static inline struct siw_sqe *irq_alloc_free(struct siw_qp *qp) diff --git a/drivers/infiniband/sw/siw/siw_cm.c b/drivers/infiniband/sw/siw/siw_cm.c index da530c0404da..86323918a570 100644 --- a/drivers/infiniband/sw/siw/siw_cm.c +++ b/drivers/infiniband/sw/siw/siw_cm.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Fredy Neeser */ @@ -41,16 +41,6 @@ static int siw_cm_upcall(struct siw_cep *cep, enum iw_cm_event_type reason, static void siw_sk_assign_cm_upcalls(struct sock *sk) { - write_lock_bh(&sk->sk_callback_lock); - sk->sk_state_change = siw_cm_llp_state_change; - sk->sk_data_ready = siw_cm_llp_data_ready; - sk->sk_write_space = siw_cm_llp_write_space; - sk->sk_error_report = siw_cm_llp_error_report; - write_unlock_bh(&sk->sk_callback_lock); -} - -static void siw_sk_save_upcalls(struct sock *sk) -{ struct siw_cep *cep = sk_to_cep(sk); write_lock_bh(&sk->sk_callback_lock); @@ -58,6 +48,11 @@ static void siw_sk_save_upcalls(struct sock *sk) cep->sk_data_ready = sk->sk_data_ready; cep->sk_write_space = sk->sk_write_space; cep->sk_error_report = sk->sk_error_report; + + sk->sk_state_change = siw_cm_llp_state_change; + sk->sk_data_ready = siw_cm_llp_data_ready; + sk->sk_write_space = siw_cm_llp_write_space; + sk->sk_error_report = siw_cm_llp_error_report; write_unlock_bh(&sk->sk_callback_lock); } @@ -156,7 +151,6 @@ static void siw_cep_socket_assoc(struct siw_cep *cep, struct socket *s) siw_cep_get(cep); s->sk->sk_user_data = cep; - siw_sk_save_upcalls(s->sk); siw_sk_assign_cm_upcalls(s->sk); } @@ -364,6 +358,24 @@ static int siw_cm_upcall(struct siw_cep *cep, enum iw_cm_event_type reason, return id->event_handler(id, &event); } +static void siw_free_cm_id(struct siw_cep *cep) +{ + if (!cep->cm_id) + return; + + cep->cm_id->rem_ref(cep->cm_id); + cep->cm_id = NULL; +} + +static void siw_destroy_cep_sock(struct siw_cep *cep) +{ + if (cep->sock) { + siw_socket_disassoc(cep->sock); + sock_release(cep->sock); + cep->sock = NULL; + } +} + /* * siw_qp_cm_drop() * @@ -393,8 +405,7 @@ void siw_qp_cm_drop(struct siw_qp *qp, int schedule) } siw_dbg_cep(cep, "immediate close, state %d\n", cep->state); - if (qp->term_info.valid) - siw_send_terminate(qp); + siw_send_terminate(qp); if (cep->cm_id) { switch (cep->state) { @@ -416,20 +427,12 @@ void siw_qp_cm_drop(struct siw_qp *qp, int schedule) default: break; } - cep->cm_id->rem_ref(cep->cm_id); - cep->cm_id = NULL; + siw_free_cm_id(cep); siw_cep_put(cep); } cep->state = SIW_EPSTATE_CLOSED; - if (cep->sock) { - siw_socket_disassoc(cep->sock); - /* - * Immediately close socket - */ - sock_release(cep->sock); - cep->sock = NULL; - } + siw_destroy_cep_sock(cep); if (cep->qp) { cep->qp = NULL; siw_qp_put(qp); @@ -445,6 +448,12 @@ void siw_cep_put(struct siw_cep *cep) kref_put(&cep->ref, __siw_cep_dealloc); } +static void siw_cep_set_free_and_put(struct siw_cep *cep) +{ + siw_cep_set_free(cep); + siw_cep_put(cep); +} + void siw_cep_get(struct siw_cep *cep) { kref_get(&cep->ref); @@ -976,6 +985,7 @@ static void siw_accept_newconn(struct siw_cep *cep) siw_cep_put(cep); new_cep->listen_cep = NULL; if (rv) { + siw_cancel_mpatimer(new_cep); siw_cep_set_free(new_cep); goto error; } @@ -1060,7 +1070,7 @@ static void siw_cm_work_handler(struct work_struct *w) /* * QP scheduled LLP close */ - if (cep->qp && cep->qp->term_info.valid) + if (cep->qp) siw_send_terminate(cep->qp); if (cep->cm_id) @@ -1100,9 +1110,12 @@ static void siw_cm_work_handler(struct work_struct *w) /* * Socket close before MPA request received. */ - siw_dbg_cep(cep, "no mpareq: drop listener\n"); - siw_cep_put(cep->listen_cep); - cep->listen_cep = NULL; + if (cep->listen_cep) { + siw_dbg_cep(cep, + "no mpareq: drop listener\n"); + siw_cep_put(cep->listen_cep); + cep->listen_cep = NULL; + } } } release_cep = 1; @@ -1171,8 +1184,7 @@ static void siw_cm_work_handler(struct work_struct *w) cep->sock = NULL; } if (cep->cm_id) { - cep->cm_id->rem_ref(cep->cm_id); - cep->cm_id = NULL; + siw_free_cm_id(cep); siw_cep_put(cep); } } @@ -1227,7 +1239,11 @@ static void siw_cm_llp_data_ready(struct sock *sk) if (!cep) goto out; - siw_dbg_cep(cep, "state: %d\n", cep->state); + siw_dbg_cep(cep, "cep state: %d, socket state %d\n", + cep->state, sk->sk_state); + + if (sk->sk_state != TCP_ESTABLISHED) + goto out; switch (cep->state) { case SIW_EPSTATE_RDMA_MODE: @@ -1501,16 +1517,13 @@ error: cep->cm_id = NULL; id->rem_ref(id); - siw_cep_put(cep); qp->cep = NULL; siw_cep_put(cep); cep->state = SIW_EPSTATE_CLOSED; - siw_cep_set_free(cep); - - siw_cep_put(cep); + siw_cep_set_free_and_put(cep); } else if (s) { sock_release(s); @@ -1541,7 +1554,7 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) struct siw_cep *cep = (struct siw_cep *)id->provider_data; struct siw_qp *qp; struct siw_qp_attrs qp_attrs; - int rv, max_priv_data = MPA_MAX_PRIVDATA; + int rv = -EINVAL, max_priv_data = MPA_MAX_PRIVDATA; bool wait_for_peer_rts = false; siw_cep_set_inuse(cep); @@ -1557,26 +1570,17 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) if (cep->state != SIW_EPSTATE_RECVD_MPAREQ) { siw_dbg_cep(cep, "out of state\n"); - - siw_cep_set_free(cep); - siw_cep_put(cep); - - return -ECONNRESET; + rv = -ECONNRESET; + goto free_cep; } qp = siw_qp_id2obj(sdev, params->qpn); if (!qp) { WARN(1, "[QP %d] does not exist\n", params->qpn); - siw_cep_set_free(cep); - siw_cep_put(cep); - - return -EINVAL; + goto free_cep; } down_write(&qp->state_lock); - if (qp->attrs.state > SIW_QP_STATE_RTR) { - rv = -EINVAL; - up_write(&qp->state_lock); - goto error; - } + if (qp->attrs.state > SIW_QP_STATE_RTR) + goto error_unlock; siw_dbg_cep(cep, "[QP %d]\n", params->qpn); if (try_gso && cep->mpa.hdr.params.bits & MPA_RR_FLAG_GSO_EXP) { @@ -1590,9 +1594,7 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) "[QP %u]: ord %d (max %d), ird %d (max %d)\n", qp_id(qp), params->ord, sdev->attrs.max_ord, params->ird, sdev->attrs.max_ird); - rv = -EINVAL; - up_write(&qp->state_lock); - goto error; + goto error_unlock; } if (cep->enhanced_rdma_conn_est) max_priv_data -= sizeof(struct mpa_v2_data); @@ -1602,9 +1604,7 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) cep, "[QP %u]: private data length: %d (max %d)\n", qp_id(qp), params->private_data_len, max_priv_data); - rv = -EINVAL; - up_write(&qp->state_lock); - goto error; + goto error_unlock; } if (cep->enhanced_rdma_conn_est) { if (params->ord > cep->ord) { @@ -1613,9 +1613,7 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) } else { cep->ird = params->ird; cep->ord = params->ord; - rv = -EINVAL; - up_write(&qp->state_lock); - goto error; + goto error_unlock; } } if (params->ird < cep->ird) { @@ -1624,8 +1622,7 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) params->ird = cep->ird; else { rv = -ENOMEM; - up_write(&qp->state_lock); - goto error; + goto error_unlock; } } if (cep->mpa.v2_ctrl.ord & @@ -1672,7 +1669,6 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) SIW_QP_ATTR_ORD | SIW_QP_ATTR_IRD | SIW_QP_ATTR_MPA); up_write(&qp->state_lock); - if (rv) goto error; @@ -1695,27 +1691,23 @@ int siw_accept(struct iw_cm_id *id, struct iw_cm_conn_param *params) siw_cep_set_free(cep); return 0; + +error_unlock: + up_write(&qp->state_lock); error: - siw_socket_disassoc(cep->sock); - sock_release(cep->sock); - cep->sock = NULL; + siw_destroy_cep_sock(cep); cep->state = SIW_EPSTATE_CLOSED; - if (cep->cm_id) { - cep->cm_id->rem_ref(id); - cep->cm_id = NULL; - } + siw_free_cm_id(cep); if (qp->cep) { siw_cep_put(cep); qp->cep = NULL; } cep->qp = NULL; siw_qp_put(qp); - - siw_cep_set_free(cep); - siw_cep_put(cep); - +free_cep: + siw_cep_set_free_and_put(cep); return rv; } @@ -1737,8 +1729,7 @@ int siw_reject(struct iw_cm_id *id, const void *pdata, u8 pd_len) if (cep->state != SIW_EPSTATE_RECVD_MPAREQ) { siw_dbg_cep(cep, "out of state\n"); - siw_cep_set_free(cep); - siw_cep_put(cep); /* put last reference */ + siw_cep_set_free_and_put(cep); /* put last reference */ return -ECONNRESET; } @@ -1749,14 +1740,11 @@ int siw_reject(struct iw_cm_id *id, const void *pdata, u8 pd_len) cep->mpa.hdr.params.bits |= MPA_RR_FLAG_REJECT; /* reject */ siw_send_mpareqrep(cep, pdata, pd_len); } - siw_socket_disassoc(cep->sock); - sock_release(cep->sock); - cep->sock = NULL; + siw_destroy_cep_sock(cep); cep->state = SIW_EPSTATE_CLOSED; - siw_cep_set_free(cep); - siw_cep_put(cep); + siw_cep_set_free_and_put(cep); return 0; } @@ -1883,16 +1871,12 @@ error: if (cep) { siw_cep_set_inuse(cep); - if (cep->cm_id) { - cep->cm_id->rem_ref(cep->cm_id); - cep->cm_id = NULL; - } + siw_free_cm_id(cep); cep->sock = NULL; siw_socket_disassoc(s); cep->state = SIW_EPSTATE_CLOSED; - siw_cep_set_free(cep); - siw_cep_put(cep); + siw_cep_set_free_and_put(cep); } sock_release(s); @@ -1916,18 +1900,14 @@ static void siw_drop_listeners(struct iw_cm_id *id) siw_cep_set_inuse(cep); - if (cep->cm_id) { - cep->cm_id->rem_ref(cep->cm_id); - cep->cm_id = NULL; - } + siw_free_cm_id(cep); if (cep->sock) { siw_socket_disassoc(cep->sock); sock_release(cep->sock); cep->sock = NULL; } cep->state = SIW_EPSTATE_CLOSED; - siw_cep_set_free(cep); - siw_cep_put(cep); + siw_cep_set_free_and_put(cep); } } diff --git a/drivers/infiniband/sw/siw/siw_cm.h b/drivers/infiniband/sw/siw/siw_cm.h index 8c59cb3e2868..7011c8a8ee7b 100644 --- a/drivers/infiniband/sw/siw/siw_cm.h +++ b/drivers/infiniband/sw/siw/siw_cm.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Greg Joyce <greg@opengridcomputing.com> */ diff --git a/drivers/infiniband/sw/siw/siw_cq.c b/drivers/infiniband/sw/siw/siw_cq.c index 403029de6b92..f3c2226aff94 100644 --- a/drivers/infiniband/sw/siw/siw_cq.c +++ b/drivers/infiniband/sw/siw/siw_cq.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ diff --git a/drivers/infiniband/sw/siw/siw_main.c b/drivers/infiniband/sw/siw/siw_main.c index 65b5cda5457b..61ad8ca3d1a2 100644 --- a/drivers/infiniband/sw/siw/siw_main.c +++ b/drivers/infiniband/sw/siw/siw_main.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ @@ -75,8 +75,7 @@ static int siw_device_register(struct siw_device *sdev, const char *name) return rv; } - siw_dbg(base_dev, "HWaddr=%pM\n", sdev->netdev->dev_addr); - + siw_dbg(base_dev, "HWaddr=%pM\n", sdev->raw_gid); return 0; } @@ -88,29 +87,6 @@ static void siw_device_cleanup(struct ib_device *base_dev) xa_destroy(&sdev->mem_xa); } -static int siw_create_tx_threads(void) -{ - int cpu, assigned = 0; - - for_each_online_cpu(cpu) { - /* Skip HT cores */ - if (cpu % cpumask_weight(topology_sibling_cpumask(cpu))) - continue; - - siw_tx_thread[cpu] = - kthread_run_on_cpu(siw_run_sq, - (unsigned long *)(long)cpu, - cpu, "siw_tx/%u"); - if (IS_ERR(siw_tx_thread[cpu])) { - siw_tx_thread[cpu] = NULL; - continue; - } - - assigned++; - } - return assigned; -} - static int siw_dev_qualified(struct net_device *netdev) { /* @@ -133,6 +109,17 @@ static struct { int num_nodes; } siw_cpu_info; +static void siw_destroy_cpulist(int number) +{ + int i = 0; + + while (i < number) + kfree(siw_cpu_info.tx_valid_cpus[i++]); + + kfree(siw_cpu_info.tx_valid_cpus); + siw_cpu_info.tx_valid_cpus = NULL; +} + static int siw_init_cpulist(void) { int i, num_nodes = nr_node_ids; @@ -162,24 +149,11 @@ static int siw_init_cpulist(void) out_err: siw_cpu_info.num_nodes = 0; - while (--i >= 0) - kfree(siw_cpu_info.tx_valid_cpus[i]); - kfree(siw_cpu_info.tx_valid_cpus); - siw_cpu_info.tx_valid_cpus = NULL; + siw_destroy_cpulist(i); return -ENOMEM; } -static void siw_destroy_cpulist(void) -{ - int i = 0; - - while (i < siw_cpu_info.num_nodes) - kfree(siw_cpu_info.tx_valid_cpus[i++]); - - kfree(siw_cpu_info.tx_valid_cpus); -} - /* * Choose CPU with least number of active QP's from NUMA node of * TX interface. @@ -313,24 +287,19 @@ static struct siw_device *siw_device_create(struct net_device *netdev) return NULL; base_dev = &sdev->base_dev; - sdev->netdev = netdev; - if (netdev->type != ARPHRD_LOOPBACK && netdev->type != ARPHRD_NONE) { - addrconf_addr_eui48((unsigned char *)&base_dev->node_guid, - netdev->dev_addr); + if (netdev->addr_len) { + memcpy(sdev->raw_gid, netdev->dev_addr, + min_t(unsigned int, netdev->addr_len, ETH_ALEN)); } else { /* - * This device does not have a HW address, - * but connection mangagement lib expects gid != 0 + * This device does not have a HW address, but + * connection mangagement requires a unique gid. */ - size_t len = min_t(size_t, strlen(base_dev->name), 6); - char addr[6] = { }; - - memcpy(addr, base_dev->name, len); - addrconf_addr_eui48((unsigned char *)&base_dev->node_guid, - addr); + eth_random_addr(sdev->raw_gid); } + addrconf_addr_eui48((u8 *)&base_dev->node_guid, sdev->raw_gid); base_dev->uverbs_cmd_mask |= BIT_ULL(IB_USER_VERBS_CMD_POST_SEND); @@ -535,7 +504,6 @@ static struct rdma_link_ops siw_link_ops = { static __init int siw_init_module(void) { int rv; - int nr_cpu; if (SENDPAGE_THRESH < SIW_MAX_INLINE) { pr_info("siw: sendpage threshold too small: %u\n", @@ -580,40 +548,30 @@ static __init int siw_init_module(void) return 0; out_error: - for (nr_cpu = 0; nr_cpu < nr_cpu_ids; nr_cpu++) { - if (siw_tx_thread[nr_cpu]) { - siw_stop_tx_thread(nr_cpu); - siw_tx_thread[nr_cpu] = NULL; - } - } + siw_stop_tx_threads(); + if (siw_crypto_shash) crypto_free_shash(siw_crypto_shash); pr_info("SoftIWARP attach failed. Error: %d\n", rv); siw_cm_exit(); - siw_destroy_cpulist(); + siw_destroy_cpulist(siw_cpu_info.num_nodes); return rv; } static void __exit siw_exit_module(void) { - int cpu; + siw_stop_tx_threads(); - for_each_possible_cpu(cpu) { - if (siw_tx_thread[cpu]) { - siw_stop_tx_thread(cpu); - siw_tx_thread[cpu] = NULL; - } - } unregister_netdevice_notifier(&siw_netdev_nb); rdma_link_unregister(&siw_link_ops); ib_unregister_driver(RDMA_DRIVER_SIW); siw_cm_exit(); - siw_destroy_cpulist(); + siw_destroy_cpulist(siw_cpu_info.num_nodes); if (siw_crypto_shash) crypto_free_shash(siw_crypto_shash); diff --git a/drivers/infiniband/sw/siw/siw_mem.c b/drivers/infiniband/sw/siw/siw_mem.c index e6e25f15567d..dcb963607c8b 100644 --- a/drivers/infiniband/sw/siw/siw_mem.c +++ b/drivers/infiniband/sw/siw/siw_mem.c @@ -1,10 +1,11 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ #include <linux/gfp.h> #include <rdma/ib_verbs.h> +#include <rdma/ib_umem.h> #include <linux/dma-mapping.h> #include <linux/slab.h> #include <linux/sched/mm.h> @@ -13,18 +14,20 @@ #include "siw.h" #include "siw_mem.h" +/* Stag lookup is based on its index part only (24 bits). */ +#define SIW_STAG_MAX_INDEX 0x00ffffff + /* - * Stag lookup is based on its index part only (24 bits). * The code avoids special Stag of zero and tries to randomize * STag values between 1 and SIW_STAG_MAX_INDEX. */ int siw_mem_add(struct siw_device *sdev, struct siw_mem *m) { - struct xa_limit limit = XA_LIMIT(1, 0x00ffffff); + struct xa_limit limit = XA_LIMIT(1, SIW_STAG_MAX_INDEX); u32 id, next; get_random_bytes(&next, 4); - next &= 0x00ffffff; + next &= SIW_STAG_MAX_INDEX; if (xa_alloc_cyclic(&sdev->mem_xa, &id, m, limit, &next, GFP_KERNEL) < 0) @@ -60,28 +63,17 @@ struct siw_mem *siw_mem_id2obj(struct siw_device *sdev, int stag_index) return NULL; } -static void siw_free_plist(struct siw_page_chunk *chunk, int num_pages, - bool dirty) -{ - unpin_user_pages_dirty_lock(chunk->plist, num_pages, dirty); -} - -void siw_umem_release(struct siw_umem *umem, bool dirty) +void siw_umem_release(struct siw_umem *umem) { - struct mm_struct *mm_s = umem->owning_mm; int i, num_pages = umem->num_pages; - for (i = 0; num_pages; i++) { - int to_free = min_t(int, PAGES_PER_CHUNK, num_pages); + if (umem->base_mem) + ib_umem_release(umem->base_mem); - siw_free_plist(&umem->page_chunk[i], to_free, - umem->writable && dirty); + for (i = 0; num_pages > 0; i++) { kfree(umem->page_chunk[i].plist); - num_pages -= to_free; + num_pages -= PAGES_PER_CHUNK; } - atomic64_sub(umem->num_pages, &mm_s->pinned_vm); - - mmdrop(mm_s); kfree(umem->page_chunk); kfree(umem); } @@ -91,7 +83,7 @@ int siw_mr_add_mem(struct siw_mr *mr, struct ib_pd *pd, void *mem_obj, { struct siw_device *sdev = to_siw_dev(pd->device); struct siw_mem *mem = kzalloc(sizeof(*mem), GFP_KERNEL); - struct xa_limit limit = XA_LIMIT(1, 0x00ffffff); + struct xa_limit limit = XA_LIMIT(1, SIW_STAG_MAX_INDEX); u32 id, next; if (!mem) @@ -107,7 +99,7 @@ int siw_mr_add_mem(struct siw_mr *mr, struct ib_pd *pd, void *mem_obj, kref_init(&mem->ref); get_random_bytes(&next, 4); - next &= 0x00ffffff; + next &= SIW_STAG_MAX_INDEX; if (xa_alloc_cyclic(&sdev->mem_xa, &id, mem, limit, &next, GFP_KERNEL) < 0) { @@ -145,7 +137,7 @@ void siw_free_mem(struct kref *ref) if (!mem->is_mw && mem->mem_obj) { if (mem->is_pbl == 0) - siw_umem_release(mem->umem, true); + siw_umem_release(mem->umem); else kfree(mem->pbl); } @@ -362,18 +354,16 @@ struct siw_pbl *siw_pbl_alloc(u32 num_buf) return pbl; } -struct siw_umem *siw_umem_get(u64 start, u64 len, bool writable) +struct siw_umem *siw_umem_get(struct ib_device *base_dev, u64 start, + u64 len, int rights) { struct siw_umem *umem; - struct mm_struct *mm_s; + struct ib_umem *base_mem; + struct sg_page_iter sg_iter; + struct sg_table *sgt; u64 first_page_va; - unsigned long mlock_limit; - unsigned int foll_flags = FOLL_LONGTERM; int num_pages, num_chunks, i, rv = 0; - if (!can_do_mlock()) - return ERR_PTR(-EPERM); - if (!len) return ERR_PTR(-EINVAL); @@ -385,65 +375,50 @@ struct siw_umem *siw_umem_get(u64 start, u64 len, bool writable) if (!umem) return ERR_PTR(-ENOMEM); - mm_s = current->mm; - umem->owning_mm = mm_s; - umem->writable = writable; - - mmgrab(mm_s); - - if (writable) - foll_flags |= FOLL_WRITE; - - mmap_read_lock(mm_s); - - mlock_limit = rlimit(RLIMIT_MEMLOCK) >> PAGE_SHIFT; - - if (atomic64_add_return(num_pages, &mm_s->pinned_vm) > mlock_limit) { - rv = -ENOMEM; - goto out_sem_up; - } - umem->fp_addr = first_page_va; - umem->page_chunk = kcalloc(num_chunks, sizeof(struct siw_page_chunk), GFP_KERNEL); if (!umem->page_chunk) { rv = -ENOMEM; - goto out_sem_up; + goto err_out; } - for (i = 0; num_pages; i++) { + base_mem = ib_umem_get(base_dev, start, len, rights); + if (IS_ERR(base_mem)) { + rv = PTR_ERR(base_mem); + siw_dbg(base_dev, "Cannot pin user memory: %d\n", rv); + goto err_out; + } + umem->fp_addr = first_page_va; + umem->base_mem = base_mem; + + sgt = &base_mem->sgt_append.sgt; + __sg_page_iter_start(&sg_iter, sgt->sgl, sgt->orig_nents, 0); + + if (!__sg_page_iter_next(&sg_iter)) { + rv = -EINVAL; + goto err_out; + } + for (i = 0; num_pages > 0; i++) { int nents = min_t(int, num_pages, PAGES_PER_CHUNK); struct page **plist = kcalloc(nents, sizeof(struct page *), GFP_KERNEL); if (!plist) { rv = -ENOMEM; - goto out_sem_up; + goto err_out; } umem->page_chunk[i].plist = plist; - while (nents) { - rv = pin_user_pages(first_page_va, nents, foll_flags, - plist); - if (rv < 0) - goto out_sem_up; - - umem->num_pages += rv; - first_page_va += rv * PAGE_SIZE; - plist += rv; - nents -= rv; - num_pages -= rv; + while (nents--) { + *plist = sg_page_iter_page(&sg_iter); + umem->num_pages++; + num_pages--; + plist++; + if (!__sg_page_iter_next(&sg_iter)) + break; } } -out_sem_up: - mmap_read_unlock(mm_s); - - if (rv > 0) - return umem; - - /* Adjust accounting for pages not pinned */ - if (num_pages) - atomic64_sub(num_pages, &mm_s->pinned_vm); - - siw_umem_release(umem, false); + return umem; +err_out: + siw_umem_release(umem); return ERR_PTR(rv); } diff --git a/drivers/infiniband/sw/siw/siw_mem.h b/drivers/infiniband/sw/siw/siw_mem.h index f911287576d1..e74cfcd6dbc1 100644 --- a/drivers/infiniband/sw/siw/siw_mem.h +++ b/drivers/infiniband/sw/siw/siw_mem.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ @@ -6,8 +6,9 @@ #ifndef _SIW_MEM_H #define _SIW_MEM_H -struct siw_umem *siw_umem_get(u64 start, u64 len, bool writable); -void siw_umem_release(struct siw_umem *umem, bool dirty); +struct siw_umem *siw_umem_get(struct ib_device *base_dave, u64 start, + u64 len, int rights); +void siw_umem_release(struct siw_umem *umem); struct siw_pbl *siw_pbl_alloc(u32 num_buf); dma_addr_t siw_pbl_get_buffer(struct siw_pbl *pbl, u64 off, int *len, int *idx); struct siw_mem *siw_mem_id2obj(struct siw_device *sdev, int stag_index); diff --git a/drivers/infiniband/sw/siw/siw_qp.c b/drivers/infiniband/sw/siw/siw_qp.c index 81e9bbd9ebda..da92cfa2073d 100644 --- a/drivers/infiniband/sw/siw/siw_qp.c +++ b/drivers/infiniband/sw/siw/siw_qp.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ @@ -204,7 +204,7 @@ static int siw_qp_readq_init(struct siw_qp *qp, int irq_size, int orq_size) { if (irq_size) { irq_size = roundup_pow_of_two(irq_size); - qp->irq = vzalloc(irq_size * sizeof(struct siw_sqe)); + qp->irq = vcalloc(irq_size, sizeof(struct siw_sqe)); if (!qp->irq) { qp->attrs.irq_size = 0; return -ENOMEM; @@ -212,7 +212,7 @@ static int siw_qp_readq_init(struct siw_qp *qp, int irq_size, int orq_size) } if (orq_size) { orq_size = roundup_pow_of_two(orq_size); - qp->orq = vzalloc(orq_size * sizeof(struct siw_sqe)); + qp->orq = vcalloc(orq_size, sizeof(struct siw_sqe)); if (!qp->orq) { qp->attrs.orq_size = 0; qp->attrs.irq_size = 0; @@ -1183,7 +1183,7 @@ int siw_rqe_complete(struct siw_qp *qp, struct siw_rqe *rqe, u32 bytes, /* * siw_sq_flush() * - * Flush SQ and ORRQ entries to CQ. + * Flush SQ and ORQ entries to CQ. * * Must be called with QP state write lock held. * Therefore, SQ and ORQ lock must not be taken. diff --git a/drivers/infiniband/sw/siw/siw_qp_rx.c b/drivers/infiniband/sw/siw/siw_qp_rx.c index 58bbf738e4e5..ed4fc39718b4 100644 --- a/drivers/infiniband/sw/siw/siw_qp_rx.c +++ b/drivers/infiniband/sw/siw/siw_qp_rx.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ @@ -405,6 +405,20 @@ out: return wqe; } +static int siw_rx_data(struct siw_mem *mem_p, struct siw_rx_stream *srx, + unsigned int *pbl_idx, u64 addr, int bytes) +{ + int rv; + + if (mem_p->mem_obj == NULL) + rv = siw_rx_kva(srx, ib_virt_dma_to_ptr(addr), bytes); + else if (!mem_p->is_pbl) + rv = siw_rx_umem(srx, mem_p->umem, addr, bytes); + else + rv = siw_rx_pbl(srx, pbl_idx, mem_p, addr, bytes); + return rv; +} + /* * siw_proc_send: * @@ -485,17 +499,8 @@ int siw_proc_send(struct siw_qp *qp) break; } mem_p = *mem; - if (mem_p->mem_obj == NULL) - rv = siw_rx_kva(srx, - ib_virt_dma_to_ptr(sge->laddr + frx->sge_off), - sge_bytes); - else if (!mem_p->is_pbl) - rv = siw_rx_umem(srx, mem_p->umem, - sge->laddr + frx->sge_off, sge_bytes); - else - rv = siw_rx_pbl(srx, &frx->pbl_idx, mem_p, - sge->laddr + frx->sge_off, sge_bytes); - + rv = siw_rx_data(mem_p, srx, &frx->pbl_idx, + sge->laddr + frx->sge_off, sge_bytes); if (unlikely(rv != sge_bytes)) { wqe->processed += rcvd_bytes; @@ -598,17 +603,8 @@ int siw_proc_write(struct siw_qp *qp) return -EINVAL; } - if (mem->mem_obj == NULL) - rv = siw_rx_kva(srx, - (void *)(uintptr_t)(srx->ddp_to + srx->fpdu_part_rcvd), - bytes); - else if (!mem->is_pbl) - rv = siw_rx_umem(srx, mem->umem, - srx->ddp_to + srx->fpdu_part_rcvd, bytes); - else - rv = siw_rx_pbl(srx, &frx->pbl_idx, mem, - srx->ddp_to + srx->fpdu_part_rcvd, bytes); - + rv = siw_rx_data(mem, srx, &frx->pbl_idx, + srx->ddp_to + srx->fpdu_part_rcvd, bytes); if (unlikely(rv != bytes)) { siw_init_terminate(qp, TERM_ERROR_LAYER_DDP, DDP_ETYPE_CATASTROPHIC, @@ -849,17 +845,8 @@ int siw_proc_rresp(struct siw_qp *qp) mem_p = *mem; bytes = min(srx->fpdu_part_rem, srx->skb_new); - - if (mem_p->mem_obj == NULL) - rv = siw_rx_kva(srx, - ib_virt_dma_to_ptr(sge->laddr + wqe->processed), - bytes); - else if (!mem_p->is_pbl) - rv = siw_rx_umem(srx, mem_p->umem, sge->laddr + wqe->processed, - bytes); - else - rv = siw_rx_pbl(srx, &frx->pbl_idx, mem_p, - sge->laddr + wqe->processed, bytes); + rv = siw_rx_data(mem_p, srx, &frx->pbl_idx, + sge->laddr + wqe->processed, bytes); if (rv != bytes) { wqe->wc_status = SIW_WC_GENERAL_ERR; rv = -EINVAL; @@ -881,6 +868,13 @@ error_term: return rv; } +static void siw_update_skb_rcvd(struct siw_rx_stream *srx, u16 length) +{ + srx->skb_offset += length; + srx->skb_new -= length; + srx->skb_copied += length; +} + int siw_proc_terminate(struct siw_qp *qp) { struct siw_rx_stream *srx = &qp->rx_stream; @@ -925,9 +919,7 @@ int siw_proc_terminate(struct siw_qp *qp) goto out; infop += to_copy; - srx->skb_offset += to_copy; - srx->skb_new -= to_copy; - srx->skb_copied += to_copy; + siw_update_skb_rcvd(srx, to_copy); srx->fpdu_part_rcvd += to_copy; srx->fpdu_part_rem -= to_copy; @@ -949,9 +941,7 @@ int siw_proc_terminate(struct siw_qp *qp) term->flag_m ? "valid" : "invalid"); } out: - srx->skb_new -= to_copy; - srx->skb_offset += to_copy; - srx->skb_copied += to_copy; + siw_update_skb_rcvd(srx, to_copy); srx->fpdu_part_rcvd += to_copy; srx->fpdu_part_rem -= to_copy; @@ -970,9 +960,7 @@ static int siw_get_trailer(struct siw_qp *qp, struct siw_rx_stream *srx) skb_copy_bits(skb, srx->skb_offset, tbuf, avail); - srx->skb_new -= avail; - srx->skb_offset += avail; - srx->skb_copied += avail; + siw_update_skb_rcvd(srx, avail); srx->fpdu_part_rem -= avail; if (srx->fpdu_part_rem) @@ -1023,12 +1011,8 @@ static int siw_get_hdr(struct siw_rx_stream *srx) skb_copy_bits(skb, srx->skb_offset, (char *)c_hdr + srx->fpdu_part_rcvd, bytes); + siw_update_skb_rcvd(srx, bytes); srx->fpdu_part_rcvd += bytes; - - srx->skb_new -= bytes; - srx->skb_offset += bytes; - srx->skb_copied += bytes; - if (srx->fpdu_part_rcvd < MIN_DDP_HDR) return -EAGAIN; @@ -1091,12 +1075,8 @@ static int siw_get_hdr(struct siw_rx_stream *srx) skb_copy_bits(skb, srx->skb_offset, (char *)c_hdr + srx->fpdu_part_rcvd, bytes); + siw_update_skb_rcvd(srx, bytes); srx->fpdu_part_rcvd += bytes; - - srx->skb_new -= bytes; - srx->skb_offset += bytes; - srx->skb_copied += bytes; - if (srx->fpdu_part_rcvd < hdrlen) return -EAGAIN; } diff --git a/drivers/infiniband/sw/siw/siw_qp_tx.c b/drivers/infiniband/sw/siw/siw_qp_tx.c index 7c7a51d36d0c..64ad9e0895bd 100644 --- a/drivers/infiniband/sw/siw/siw_qp_tx.c +++ b/drivers/infiniband/sw/siw/siw_qp_tx.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ @@ -34,6 +34,15 @@ static struct page *siw_get_pblpage(struct siw_mem *mem, u64 addr, int *idx) return NULL; } +static struct page *siw_get_page(struct siw_mem *mem, struct siw_sge *sge, + unsigned long offset, int *pbl_idx) +{ + if (!mem->is_pbl) + return siw_get_upage(mem->umem, sge->laddr + offset); + else + return siw_get_pblpage(mem, sge->laddr + offset, pbl_idx); +} + /* * Copy short payload at provided destination payload address */ @@ -67,11 +76,7 @@ static int siw_try_1seg(struct siw_iwarp_tx *c_tx, void *paddr) char *buffer; int pbl_idx = 0; - if (!mem->is_pbl) - p = siw_get_upage(mem->umem, sge->laddr); - else - p = siw_get_pblpage(mem, sge->laddr, &pbl_idx); - + p = siw_get_page(mem, sge, 0, &pbl_idx); if (unlikely(!p)) return -EFAULT; @@ -85,13 +90,7 @@ static int siw_try_1seg(struct siw_iwarp_tx *c_tx, void *paddr) memcpy(paddr, buffer + off, part); kunmap_local(buffer); - if (!mem->is_pbl) - p = siw_get_upage(mem->umem, - sge->laddr + part); - else - p = siw_get_pblpage(mem, - sge->laddr + part, - &pbl_idx); + p = siw_get_page(mem, sge, part, &pbl_idx); if (unlikely(!p)) return -EFAULT; @@ -249,14 +248,10 @@ static int siw_qp_prepare_tx(struct siw_iwarp_tx *c_tx) /* * Do complete CRC if enabled and short packet */ - if (c_tx->mpa_crc_hd) { - crypto_shash_init(c_tx->mpa_crc_hd); - if (crypto_shash_update(c_tx->mpa_crc_hd, - (u8 *)&c_tx->pkt, - c_tx->ctrl_len)) - return -EINVAL; - crypto_shash_final(c_tx->mpa_crc_hd, (u8 *)crc); - } + if (c_tx->mpa_crc_hd && + crypto_shash_digest(c_tx->mpa_crc_hd, (u8 *)&c_tx->pkt, + c_tx->ctrl_len, (u8 *)crc) != 0) + return -EINVAL; c_tx->ctrl_len += MPA_CRC_SIZE; return PKT_COMPLETE; @@ -297,8 +292,7 @@ static int siw_tx_ctrl(struct siw_iwarp_tx *c_tx, struct socket *s, (char *)&c_tx->pkt.ctrl + c_tx->ctrl_sent, .iov_len = c_tx->ctrl_len - c_tx->ctrl_sent }; - int rv = kernel_sendmsg(s, &msg, &iov, 1, - c_tx->ctrl_len - c_tx->ctrl_sent); + int rv = kernel_sendmsg(s, &msg, &iov, 1, iov.iov_len); if (rv >= 0) { c_tx->ctrl_sent += rv; @@ -502,13 +496,7 @@ static int siw_tx_hdt(struct siw_iwarp_tx *c_tx, struct socket *s) if (!is_kva) { struct page *p; - if (mem->is_pbl) - p = siw_get_pblpage( - mem, sge->laddr + sge_off, - &pbl_idx); - else - p = siw_get_upage(mem->umem, - sge->laddr + sge_off); + p = siw_get_page(mem, sge, sge_off, &pbl_idx); if (unlikely(!p)) { siw_unmap_pages(iov, kmap_mask, seg); wqe->processed -= c_tx->bytes_unsent; @@ -1009,13 +997,12 @@ static int siw_qp_sq_proc_local(struct siw_qp *qp, struct siw_wqe *wqe) * MPA FPDUs, each containing a DDP segment. * * SQ processing may occur in user context as a result of posting - * new WQE's or from siw_sq_work_handler() context. Processing in + * new WQE's or from siw_tx_thread context. Processing in * user context is limited to non-kernel verbs users. * * SQ processing may get paused anytime, possibly in the middle of a WR * or FPDU, if insufficient send space is available. SQ processing - * gets resumed from siw_sq_work_handler(), if send space becomes - * available again. + * gets resumed from siw_tx_thread, if send space becomes available again. * * Must be called with the QP state read-locked. * @@ -1208,10 +1195,45 @@ struct tx_task_t { static DEFINE_PER_CPU(struct tx_task_t, siw_tx_task_g); -void siw_stop_tx_thread(int nr_cpu) +int siw_create_tx_threads(void) +{ + int cpu, assigned = 0; + + for_each_online_cpu(cpu) { + struct tx_task_t *tx_task; + + /* Skip HT cores */ + if (cpu % cpumask_weight(topology_sibling_cpumask(cpu))) + continue; + + tx_task = &per_cpu(siw_tx_task_g, cpu); + init_llist_head(&tx_task->active); + init_waitqueue_head(&tx_task->waiting); + + siw_tx_thread[cpu] = + kthread_run_on_cpu(siw_run_sq, + (unsigned long *)(long)cpu, + cpu, "siw_tx/%u"); + if (IS_ERR(siw_tx_thread[cpu])) { + siw_tx_thread[cpu] = NULL; + continue; + } + assigned++; + } + return assigned; +} + +void siw_stop_tx_threads(void) { - kthread_stop(siw_tx_thread[nr_cpu]); - wake_up(&per_cpu(siw_tx_task_g, nr_cpu).waiting); + int cpu; + + for_each_possible_cpu(cpu) { + if (siw_tx_thread[cpu]) { + kthread_stop(siw_tx_thread[cpu]); + wake_up(&per_cpu(siw_tx_task_g, cpu).waiting); + siw_tx_thread[cpu] = NULL; + } + } } int siw_run_sq(void *data) @@ -1221,9 +1243,6 @@ int siw_run_sq(void *data) struct siw_qp *qp; struct tx_task_t *tx_task = &per_cpu(siw_tx_task_g, nr_cpu); - init_llist_head(&tx_task->active); - init_waitqueue_head(&tx_task->waiting); - while (1) { struct llist_node *fifo_list = NULL; @@ -1239,13 +1258,7 @@ int siw_run_sq(void *data) * llist_del_all returns a list with newest entry first. * Re-order list for fairness among QP's. */ - while (active) { - struct llist_node *tmp = active; - - active = llist_next(active); - tmp->next = fifo_list; - fifo_list = tmp; - } + fifo_list = llist_reverse_order(active); while (fifo_list) { qp = container_of(fifo_list, struct siw_qp, tx_list); fifo_list = llist_next(fifo_list); diff --git a/drivers/infiniband/sw/siw/siw_verbs.c b/drivers/infiniband/sw/siw/siw_verbs.c index 398ec13db624..ecf0444666b4 100644 --- a/drivers/infiniband/sw/siw/siw_verbs.c +++ b/drivers/infiniband/sw/siw/siw_verbs.c @@ -1,4 +1,4 @@ -// SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause +// SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ @@ -19,6 +19,15 @@ #include "siw_verbs.h" #include "siw_mem.h" +static int siw_qp_state_to_ib_qp_state[SIW_QP_STATE_COUNT] = { + [SIW_QP_STATE_IDLE] = IB_QPS_INIT, + [SIW_QP_STATE_RTR] = IB_QPS_RTR, + [SIW_QP_STATE_RTS] = IB_QPS_RTS, + [SIW_QP_STATE_CLOSING] = IB_QPS_SQD, + [SIW_QP_STATE_TERMINATE] = IB_QPS_SQE, + [SIW_QP_STATE_ERROR] = IB_QPS_ERR +}; + static int ib_qp_state_to_siw_qp_state[IB_QPS_ERR + 1] = { [IB_QPS_RESET] = SIW_QP_STATE_IDLE, [IB_QPS_INIT] = SIW_QP_STATE_IDLE, @@ -66,12 +75,9 @@ int siw_mmap(struct ib_ucontext *ctx, struct vm_area_struct *vma) entry = to_siw_mmap_entry(rdma_entry); rv = remap_vmalloc_range(vma, entry->address, 0); - if (rv) { + if (rv) pr_warn("remap_vmalloc_range failed: %lu, %zu\n", vma->vm_pgoff, size); - goto out; - } -out: rdma_user_mmap_entry_put(rdma_entry); return rv; @@ -157,7 +163,7 @@ int siw_query_device(struct ib_device *base_dev, struct ib_device_attr *attr, attr->vendor_part_id = sdev->vendor_part_id; addrconf_addr_eui48((u8 *)&attr->sys_image_guid, - sdev->netdev->dev_addr); + sdev->raw_gid); return 0; } @@ -218,7 +224,7 @@ int siw_query_gid(struct ib_device *base_dev, u32 port, int idx, /* subnet_prefix == interface_id == 0; */ memset(gid, 0, sizeof(*gid)); - memcpy(&gid->raw[0], sdev->netdev->dev_addr, 6); + memcpy(gid->raw, sdev->raw_gid, ETH_ALEN); return 0; } @@ -336,11 +342,10 @@ int siw_create_qp(struct ib_qp *ibqp, struct ib_qp_init_attr *attrs, goto err_atomic; } /* - * NOTE: we allow for zero element SQ and RQ WQE's SGL's - * but not for a QP unable to hold any WQE (SQ + RQ) + * NOTE: we don't allow for a QP unable to hold any SQ WQE */ - if (attrs->cap.max_send_wr + attrs->cap.max_recv_wr == 0) { - siw_dbg(base_dev, "QP must have send or receive queue\n"); + if (attrs->cap.max_send_wr == 0) { + siw_dbg(base_dev, "QP must have send queue\n"); rv = -EINVAL; goto err_atomic; } @@ -360,28 +365,21 @@ int siw_create_qp(struct ib_qp *ibqp, struct ib_qp_init_attr *attrs, if (rv) goto err_atomic; - num_sqe = attrs->cap.max_send_wr; - num_rqe = attrs->cap.max_recv_wr; /* All queue indices are derived from modulo operations * on a free running 'get' (consumer) and 'put' (producer) * unsigned counter. Having queue sizes at power of two * avoids handling counter wrap around. */ - if (num_sqe) - num_sqe = roundup_pow_of_two(num_sqe); - else { - /* Zero sized SQ is not supported */ - rv = -EINVAL; - goto err_out_xa; - } + num_sqe = roundup_pow_of_two(attrs->cap.max_send_wr); + num_rqe = attrs->cap.max_recv_wr; if (num_rqe) num_rqe = roundup_pow_of_two(num_rqe); if (udata) qp->sendq = vmalloc_user(num_sqe * sizeof(struct siw_sqe)); else - qp->sendq = vzalloc(num_sqe * sizeof(struct siw_sqe)); + qp->sendq = vcalloc(num_sqe, sizeof(struct siw_sqe)); if (qp->sendq == NULL) { rv = -ENOMEM; @@ -414,7 +412,7 @@ int siw_create_qp(struct ib_qp *ibqp, struct ib_qp_init_attr *attrs, qp->recvq = vmalloc_user(num_rqe * sizeof(struct siw_rqe)); else - qp->recvq = vzalloc(num_rqe * sizeof(struct siw_rqe)); + qp->recvq = vcalloc(num_rqe, sizeof(struct siw_rqe)); if (qp->recvq == NULL) { rv = -ENOMEM; @@ -515,6 +513,7 @@ int siw_query_qp(struct ib_qp *base_qp, struct ib_qp_attr *qp_attr, } else { return -EINVAL; } + qp_attr->qp_state = siw_qp_state_to_ib_qp_state[qp->attrs.state]; qp_attr->cap.max_inline_data = SIW_MAX_INLINE; qp_attr->cap.max_send_wr = qp->attrs.sq_size; qp_attr->cap.max_send_sge = qp->attrs.sq_max_sges; @@ -1321,8 +1320,6 @@ struct ib_mr *siw_reg_user_mr(struct ib_pd *pd, u64 start, u64 len, struct siw_umem *umem = NULL; struct siw_ureq_reg_mr ureq; struct siw_device *sdev = to_siw_dev(pd->device); - - unsigned long mem_limit = rlimit(RLIMIT_MEMLOCK); int rv; siw_dbg_pd(pd, "start: 0x%pK, va: 0x%pK, len: %llu\n", @@ -1338,20 +1335,7 @@ struct ib_mr *siw_reg_user_mr(struct ib_pd *pd, u64 start, u64 len, rv = -EINVAL; goto err_out; } - if (mem_limit != RLIM_INFINITY) { - unsigned long num_pages = - (PAGE_ALIGN(len + (start & ~PAGE_MASK))) >> PAGE_SHIFT; - mem_limit >>= PAGE_SHIFT; - - if (num_pages > mem_limit - current->mm->locked_vm) { - siw_dbg_pd(pd, "pages req %lu, max %lu, lock %lu\n", - num_pages, mem_limit, - current->mm->locked_vm); - rv = -ENOMEM; - goto err_out; - } - } - umem = siw_umem_get(start, len, ib_access_writable(rights)); + umem = siw_umem_get(pd->device, start, len, rights); if (IS_ERR(umem)) { rv = PTR_ERR(umem); siw_dbg_pd(pd, "getting user memory failed: %d\n", rv); @@ -1404,7 +1388,7 @@ err_out: kfree_rcu(mr, rcu); } else { if (umem) - siw_umem_release(umem, false); + siw_umem_release(umem); } return ERR_PTR(rv); } @@ -1494,7 +1478,7 @@ int siw_map_mr_sg(struct ib_mr *base_mr, struct scatterlist *sl, int num_sle, if (pbl->max_buf < num_sle) { siw_dbg_mem(mem, "too many SGE's: %d > %d\n", - mem->pbl->max_buf, num_sle); + num_sle, pbl->max_buf); return -ENOMEM; } for_each_sg(sl, slp, num_sle, i) { @@ -1624,7 +1608,7 @@ int siw_create_srq(struct ib_srq *base_srq, srq->recvq = vmalloc_user(srq->num_rqe * sizeof(struct siw_rqe)); else - srq->recvq = vzalloc(srq->num_rqe * sizeof(struct siw_rqe)); + srq->recvq = vcalloc(srq->num_rqe, sizeof(struct siw_rqe)); if (srq->recvq == NULL) { rv = -ENOMEM; diff --git a/drivers/infiniband/sw/siw/siw_verbs.h b/drivers/infiniband/sw/siw/siw_verbs.h index 09964234f8d3..4b57a4fb7237 100644 --- a/drivers/infiniband/sw/siw/siw_verbs.h +++ b/drivers/infiniband/sw/siw/siw_verbs.h @@ -1,4 +1,4 @@ -/* SPDX-License-Identifier: GPL-2.0 or BSD-3-Clause */ +/* SPDX-License-Identifier: GPL-2.0 OR BSD-3-Clause */ /* Authors: Bernard Metzler <bmt@zurich.ibm.com> */ /* Copyright (c) 2008-2019, IBM Corporation */ |