From 977940b85e3c3b6b5b8ffbc1235b23e35284382f Mon Sep 17 00:00:00 2001 From: Arthur Kepner Date: Wed, 4 Nov 2015 21:10:10 -0500 Subject: staging/rdma/hfi1: optionally prescan rx queue for {B, F}ECNs - UC, RC To more rapidly respond to Explicit Congestion Notifications, prescan the receive queue, and process FECNs, and BECNs first. When a UC, or RC packet containing a FECN, or BECN is found, immediately react to the ECN (either by returning a CNP, or adjusting the injection rate). Afterward, the packet will be processed normally. Reviewed-by: Mike Marciniszyn Reviewed-by: Dennis Dalessandro Signed-off-by: Arthur Kepner Signed-off-by: Ira Weiny Signed-off-by: Greg Kroah-Hartman diff --git a/drivers/staging/rdma/hfi1/driver.c b/drivers/staging/rdma/hfi1/driver.c index cfcbe41..9a4ec09 100644 --- a/drivers/staging/rdma/hfi1/driver.c +++ b/drivers/staging/rdma/hfi1/driver.c @@ -436,59 +436,58 @@ static inline void init_packet(struct hfi1_ctxtdata *rcd, #ifndef CONFIG_PRESCAN_RXQ static void prescan_rxq(struct hfi1_packet *packet) {} -#else /* CONFIG_PRESCAN_RXQ */ +#else /* !CONFIG_PRESCAN_RXQ */ static int prescan_receive_queue; static void process_ecn(struct hfi1_qp *qp, struct hfi1_ib_header *hdr, struct hfi1_other_headers *ohdr, - u64 rhf, struct ib_grh *grh) + u64 rhf, u32 bth1, struct ib_grh *grh) { struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); - u32 bth1; + u32 rqpn = 0; + u16 rlid; u8 sc5, svc_type; - int is_fecn, is_becn; switch (qp->ibqp.qp_type) { + case IB_QPT_SMI: + case IB_QPT_GSI: case IB_QPT_UD: + rlid = be16_to_cpu(hdr->lrh[3]); + rqpn = be32_to_cpu(ohdr->u.ud.deth[1]) & HFI1_QPN_MASK; svc_type = IB_CC_SVCTYPE_UD; break; - case IB_QPT_UC: /* LATER */ - case IB_QPT_RC: /* LATER */ + case IB_QPT_UC: + rlid = qp->remote_ah_attr.dlid; + rqpn = qp->remote_qpn; + svc_type = IB_CC_SVCTYPE_UC; + break; + case IB_QPT_RC: + rlid = qp->remote_ah_attr.dlid; + rqpn = qp->remote_qpn; + svc_type = IB_CC_SVCTYPE_RC; + break; default: return; } - is_fecn = (be32_to_cpu(ohdr->bth[1]) >> HFI1_FECN_SHIFT) & - HFI1_FECN_MASK; - is_becn = (be32_to_cpu(ohdr->bth[1]) >> HFI1_BECN_SHIFT) & - HFI1_BECN_MASK; - sc5 = (be16_to_cpu(hdr->lrh[0]) >> 12) & 0xf; if (rhf_dc_info(rhf)) sc5 |= 0x10; - if (is_fecn) { - u32 src_qpn = be32_to_cpu(ohdr->u.ud.deth[1]) & HFI1_QPN_MASK; + if (bth1 & HFI1_FECN_SMASK) { u16 pkey = (u16)be32_to_cpu(ohdr->bth[0]); u16 dlid = be16_to_cpu(hdr->lrh[1]); - u16 slid = be16_to_cpu(hdr->lrh[3]); - return_cnp(ibp, qp, src_qpn, pkey, dlid, slid, sc5, grh); + return_cnp(ibp, qp, rqpn, pkey, dlid, rlid, sc5, grh); } - if (is_becn) { + if (bth1 & HFI1_BECN_SMASK) { struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); - u32 lqpn = be32_to_cpu(ohdr->bth[1]) & HFI1_QPN_MASK; + u32 lqpn = bth1 & HFI1_QPN_MASK; u8 sl = ibp->sc_to_sl[sc5]; - process_becn(ppd, sl, 0, lqpn, 0, svc_type); + process_becn(ppd, sl, rlid, lqpn, rqpn, svc_type); } - - /* turn off BECN, or FECN */ - bth1 = be32_to_cpu(ohdr->bth[1]); - bth1 &= ~(HFI1_FECN_MASK << HFI1_FECN_SHIFT); - bth1 &= ~(HFI1_BECN_MASK << HFI1_BECN_SHIFT); - ohdr->bth[1] = cpu_to_be32(bth1); } struct ps_mdata { @@ -508,7 +507,6 @@ static inline void init_ps_mdata(struct ps_mdata *mdata, mdata->rcd = rcd; mdata->rsize = packet->rsize; mdata->maxcnt = packet->maxcnt; - mdata->ps_head = packet->rhqoff; if (HFI1_CAP_IS_KSET(DMA_RTAIL)) { @@ -564,7 +562,7 @@ static void prescan_rxq(struct hfi1_packet *packet) struct hfi1_other_headers *ohdr; struct ib_grh *grh = NULL; u64 rhf = rhf_to_cpu(rhf_addr); - u32 etype = rhf_rcv_type(rhf), qpn; + u32 etype = rhf_rcv_type(rhf), qpn, bth1; int is_ecn = 0; u8 lnh; @@ -586,15 +584,13 @@ static void prescan_rxq(struct hfi1_packet *packet) } else goto next; /* just in case */ - is_ecn |= be32_to_cpu(ohdr->bth[1]) & - (HFI1_FECN_MASK << HFI1_FECN_SHIFT); - is_ecn |= be32_to_cpu(ohdr->bth[1]) & - (HFI1_BECN_MASK << HFI1_BECN_SHIFT); + bth1 = be32_to_cpu(ohdr->bth[1]); + is_ecn = !!(bth1 & (HFI1_FECN_SMASK | HFI1_BECN_SMASK)); if (!is_ecn) goto next; - qpn = be32_to_cpu(ohdr->bth[1]) & HFI1_QPN_MASK; + qpn = bth1 & HFI1_QPN_MASK; rcu_read_lock(); qp = hfi1_lookup_qpn(ibp, qpn); @@ -603,8 +599,12 @@ static void prescan_rxq(struct hfi1_packet *packet) goto next; } - process_ecn(qp, hdr, ohdr, rhf, grh); + process_ecn(qp, hdr, ohdr, rhf, bth1, grh); rcu_read_unlock(); + + /* turn off BECN, FECN */ + bth1 &= ~(HFI1_FECN_SMASK | HFI1_BECN_SMASK); + ohdr->bth[1] = cpu_to_be32(bth1); next: update_ps_mdata(&mdata); } diff --git a/drivers/staging/rdma/hfi1/rc.c b/drivers/staging/rdma/hfi1/rc.c index 5fc93bb..fd23907 100644 --- a/drivers/staging/rdma/hfi1/rc.c +++ b/drivers/staging/rdma/hfi1/rc.c @@ -1978,7 +1978,7 @@ void hfi1_rc_rcv(struct hfi1_packet *packet) } psn = be32_to_cpu(ohdr->bth[2]); - opcode = bth0 >> 24; + opcode = (bth0 >> 24) & 0xff; /* * Process responses (ACKs) before anything else. Note that the @@ -2391,19 +2391,19 @@ void hfi1_rc_hdrerr( struct hfi1_ibport *ibp = to_iport(qp->ibqp.device, qp->port_num); int diff; u32 opcode; - u32 psn; + u32 psn, bth0; /* Check for GRH */ ohdr = &hdr->u.oth; if (has_grh) ohdr = &hdr->u.l.oth; - opcode = be32_to_cpu(ohdr->bth[0]); - if (hfi1_ruc_check_hdr(ibp, hdr, has_grh, qp, opcode)) + bth0 = be32_to_cpu(ohdr->bth[0]); + if (hfi1_ruc_check_hdr(ibp, hdr, has_grh, qp, bth0)) return; psn = be32_to_cpu(ohdr->bth[2]); - opcode >>= 24; + opcode = (bth0 >> 24) & 0xff; /* Only deal with RDMA Writes for now */ if (opcode < IB_OPCODE_RC_RDMA_READ_RESPONSE_FIRST) { diff --git a/drivers/staging/rdma/hfi1/uc.c b/drivers/staging/rdma/hfi1/uc.c index 6095039..4f2a788 100644 --- a/drivers/staging/rdma/hfi1/uc.c +++ b/drivers/staging/rdma/hfi1/uc.c @@ -268,7 +268,7 @@ void hfi1_uc_rcv(struct hfi1_packet *packet) u32 tlen = packet->tlen; struct hfi1_qp *qp = packet->qp; struct hfi1_other_headers *ohdr = packet->ohdr; - u32 opcode; + u32 bth0, opcode; u32 hdrsize = packet->hlen; u32 psn; u32 pad; @@ -278,10 +278,9 @@ void hfi1_uc_rcv(struct hfi1_packet *packet) int has_grh = rcv_flags & HFI1_HAS_GRH; int ret; u32 bth1; - struct ib_grh *grh = NULL; - opcode = be32_to_cpu(ohdr->bth[0]); - if (hfi1_ruc_check_hdr(ibp, hdr, has_grh, qp, opcode)) + bth0 = be32_to_cpu(ohdr->bth[0]); + if (hfi1_ruc_check_hdr(ibp, hdr, has_grh, qp, bth0)) return; bth1 = be32_to_cpu(ohdr->bth[1]); @@ -303,6 +302,7 @@ void hfi1_uc_rcv(struct hfi1_packet *packet) } if (bth1 & HFI1_FECN_SMASK) { + struct ib_grh *grh = NULL; u16 pkey = (u16)be32_to_cpu(ohdr->bth[0]); u16 slid = be16_to_cpu(hdr->lrh[3]); u16 dlid = be16_to_cpu(hdr->lrh[1]); @@ -310,13 +310,16 @@ void hfi1_uc_rcv(struct hfi1_packet *packet) u8 sc5; sc5 = ibp->sl_to_sc[qp->remote_ah_attr.sl]; + if (has_grh) + grh = &hdr->u.l.grh; - return_cnp(ibp, qp, src_qp, pkey, dlid, slid, sc5, grh); + return_cnp(ibp, qp, src_qp, pkey, dlid, slid, sc5, + grh); } } psn = be32_to_cpu(ohdr->bth[2]); - opcode >>= 24; + opcode = (bth0 >> 24) & 0xff; /* Compare the PSN verses the expected PSN. */ if (unlikely(cmp_psn(psn, qp->r_psn) != 0)) { diff --git a/drivers/staging/rdma/hfi1/verbs.c b/drivers/staging/rdma/hfi1/verbs.c index 9beb0aa..ea1d9e6 100644 --- a/drivers/staging/rdma/hfi1/verbs.c +++ b/drivers/staging/rdma/hfi1/verbs.c @@ -2152,11 +2152,40 @@ void hfi1_schedule_send(struct hfi1_qp *qp) void hfi1_cnp_rcv(struct hfi1_packet *packet) { struct hfi1_ibport *ibp = &packet->rcd->ppd->ibport_data; - - if (packet->qp->ibqp.qp_type == IB_QPT_UC) - hfi1_uc_rcv(packet); - else if (packet->qp->ibqp.qp_type == IB_QPT_UD) - hfi1_ud_rcv(packet); - else + struct hfi1_pportdata *ppd = ppd_from_ibp(ibp); + struct hfi1_ib_header *hdr = packet->hdr; + struct hfi1_qp *qp = packet->qp; + u32 lqpn, rqpn = 0; + u16 rlid = 0; + u8 sl, sc5, sc4_bit, svc_type; + bool sc4_set = has_sc4_bit(packet); + + switch (packet->qp->ibqp.qp_type) { + case IB_QPT_UC: + rlid = qp->remote_ah_attr.dlid; + rqpn = qp->remote_qpn; + svc_type = IB_CC_SVCTYPE_UC; + break; + case IB_QPT_RC: + rlid = qp->remote_ah_attr.dlid; + rqpn = qp->remote_qpn; + svc_type = IB_CC_SVCTYPE_RC; + break; + case IB_QPT_SMI: + case IB_QPT_GSI: + case IB_QPT_UD: + svc_type = IB_CC_SVCTYPE_UD; + break; + default: ibp->n_pkt_drops++; + return; + } + + sc4_bit = sc4_set << 4; + sc5 = (be16_to_cpu(hdr->lrh[0]) >> 12) & 0xf; + sc5 |= sc4_bit; + sl = ibp->sc_to_sl[sc5]; + lqpn = qp->ibqp.qp_num; + + process_becn(ppd, sl, rlid, lqpn, rqpn, svc_type); } -- cgit v0.10.2