[for-next,v6,06/12] Add memory access through MWs

Message ID	20200921200356.8627-7-rpearson@hpe.com (mailing list archive)
State	Superseded
Headers	show Return-Path: <SRS0=5Yt0=C6=vger.kernel.org=linux-rdma-owner@kernel.org> From: Bob Pearson <rpearsonhpe@gmail.com> To: jgg@nvidia.com, zyjzyj2000@gmail.com, linux-rdma@vger.kernel.org Cc: Bob Pearson <rpearson@hpe.com> Subject: [PATCH for-next v6 06/12] Add memory access through MWs Date: Mon, 21 Sep 2020 15:03:50 -0500 Message-Id: <20200921200356.8627-7-rpearson@hpe.com> In-Reply-To: <20200921200356.8627-1-rpearson@hpe.com> References: <20200921200356.8627-1-rpearson@hpe.com> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Precedence: bulk
Series	rdma_rxe: API extensions \| expand [for-next,v6,00/12] rdma_rxe: API extensions [for-next,v6,01/12] rdma_rxe: Separat MEM into MR and MW objects. [for-next,v6,02/12] rdma_rxe: Enable MW objects [for-next,v6,03/12] rdma_rxe: Let pools support both keys and indices [for-next,v6,04/12] rdma_rxe: Add alloc_mw and dealloc_mw verbs [for-next,v6,05/12] rdma_rxe: Add bind_mw and invalidate_mw verbs [for-next,v6,06/12] Add memory access through MWs [for-next,v6,07/12] rdma_rxe: Add support for ibv_query_device_ex [for-next,v6,08/12] rdma_rxe: Add support for extended CQ operations [for-next,v6,09/12] rdma_rxe: Add support for extended QP operations [for-next,v6,10/12] rdma_rxe: Fix pool related bugs [for-next,v6,11/12] rdma_rxe: Fix mcast group allocation bug [for-next,v6,12/12] rdma_rxe: Fix bugs in the multicast receive path

diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h index 1ee6270d3f2a..b824aeeb8b6b 100644 --- a/drivers/infiniband/sw/rxe/rxe_loc.h +++ b/drivers/infiniband/sw/rxe/rxe_loc.h @@ -100,29 +100,30 @@ enum lookup_type { lookup_remote, }; -struct rxe_mr *lookup_mr(struct rxe_pd *pd, int access, u32 key, - enum lookup_type type); - -int mr_check_range(struct rxe_mr *mr, u64 iova, size_t length); - -void rxe_mr_cleanup(struct rxe_pool_entry *arg); - int advance_dma_data(struct rxe_dma_info *dma, unsigned int length); int rxe_invalidate_mr(struct rxe_qp *qp, struct rxe_mr *mr); +int rxe_mr_check_access(struct rxe_qp *qp, struct rxe_mr *mr, + int access, u64 va, u32 resid); + +void rxe_mr_cleanup(struct rxe_pool_entry *arg); + /* rxe_mw.c */ struct ib_mw *rxe_alloc_mw(struct ib_pd *ibpd, enum ib_mw_type type, struct ib_udata *udata); int rxe_dealloc_mw(struct ib_mw *ibmw); -void rxe_mw_cleanup(struct rxe_pool_entry *arg); - int rxe_bind_mw(struct rxe_qp *qp, struct rxe_send_wqe *wqe); int rxe_invalidate_mw(struct rxe_qp *qp, struct rxe_mw *mw); +int rxe_mw_check_access(struct rxe_qp *qp, struct rxe_mw *mw, + int access, u64 va, u32 resid); + +void rxe_mw_cleanup(struct rxe_pool_entry *arg); + /* rxe_net.c */ void rxe_loopback(struct sk_buff *skb); int rxe_send(struct rxe_pkt_info *pkt, struct sk_buff *skb); diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c index 3f7c9b84f99b..3a7f814f4b81 100644 --- a/drivers/infiniband/sw/rxe/rxe_mr.c +++ b/drivers/infiniband/sw/rxe/rxe_mr.c @@ -21,7 +21,7 @@ static void rxe_set_mr_lkey(struct rxe_mr *mr) goto again; } -int mr_check_range(struct rxe_mr *mr, u64 iova, size_t length) +static int mr_check_range(struct rxe_mr *mr, u64 iova, size_t length) { switch (mr->type) { case RXE_MR_TYPE_DMA: @@ -380,6 +380,25 @@ int rxe_mr_copy(struct rxe_mr *mr, u64 iova, void *addr, int length, return err; } +static struct rxe_mr *lookup_mr(struct rxe_pd *pd, int access, u32 lkey) +{ + struct rxe_mr *mr; + struct rxe_dev *rxe = to_rdev(pd->ibpd.device); + + mr = rxe_pool_get_key(&rxe->mr_pool, &lkey); + if (!mr) + return NULL; + + if (unlikely((mr->ibmr.lkey != lkey) || (mr->pd != pd) || + (access && !(access & mr->access)) || + (mr->state != RXE_MEM_STATE_VALID))) { + rxe_drop_ref(mr); + return NULL; + } + + return mr; +} + /* copy data in or out of a wqe, i.e. sg list * under the control of a dma descriptor */ @@ -409,7 +428,7 @@ int copy_data( } if (sge->length && (offset < sge->length)) { - mr = lookup_mr(pd, access, sge->lkey, lookup_local); + mr = lookup_mr(pd, access, sge->lkey); if (!mr) { err = -EINVAL; goto err1; @@ -434,8 +453,7 @@ int copy_data( } if (sge->length) { - mr = lookup_mr(pd, access, sge->lkey, - lookup_local); + mr = lookup_mr(pd, access, sge->lkey); if (!mr) { err = -EINVAL; goto err1; @@ -510,34 +528,6 @@ int advance_dma_data(struct rxe_dma_info *dma, unsigned int length) return 0; } -/* (1) find the mr corresponding to lkey/rkey - * depending on lookup_type - * (2) verify that the (qp) pd matches the mr pd - * (3) verify that the mr can support the requested access - * (4) verify that mr state is valid - */ -struct rxe_mr *lookup_mr(struct rxe_pd *pd, int access, u32 key, - enum lookup_type type) -{ - struct rxe_mr *mr; - struct rxe_dev *rxe = to_rdev(pd->ibpd.device); - - mr = rxe_pool_get_key(&rxe->mr_pool, &key); - if (!mr) - return NULL; - - if (unlikely((type == lookup_local && mr->lkey != key) || - (type == lookup_remote && mr->rkey != key) || - mr->pd != pd || - (access && !(access & mr->access)) || - mr->state != RXE_MEM_STATE_VALID)) { - rxe_drop_ref(mr); - mr = NULL; - } - - return mr; -} - int rxe_invalidate_mr(struct rxe_qp *qp, struct rxe_mr *mr) { /* TODO there are API rules being ignored here @@ -548,6 +538,34 @@ int rxe_invalidate_mr(struct rxe_qp *qp, struct rxe_mr *mr) return 0; } +int rxe_mr_check_access(struct rxe_qp *qp, struct rxe_mr *mr, + int access, u64 va, u32 resid) +{ + int ret; + struct rxe_pd *pd = to_rpd(mr->ibmr.pd); + + if (unlikely(mr->state != RXE_MEM_STATE_VALID)) { + pr_err("attempt to access a MR that is not in the valid state\n"); + return -EINVAL; + } + + /* C10-56 */ + if (unlikely(pd != qp->pd)) { + pr_err("attempt to access a MR with a different PD than the QP\n"); + return -EINVAL; + } + + /* C10-57 */ + if (unlikely(access && !(access & mr->access))) { + pr_err("attempt to access a MR without required access rights\n"); + return -EINVAL; + } + + ret = mr_check_range(mr, va, resid); + + return ret; +} + void rxe_mr_cleanup(struct rxe_pool_entry *arg) { struct rxe_mr *mr = container_of(arg, typeof(*mr), pelem); diff --git a/drivers/infiniband/sw/rxe/rxe_mw.c b/drivers/infiniband/sw/rxe/rxe_mw.c index 51bc71c98654..50f3152d3b57 100644 --- a/drivers/infiniband/sw/rxe/rxe_mw.c +++ b/drivers/infiniband/sw/rxe/rxe_mw.c @@ -318,11 +318,6 @@ int rxe_bind_mw(struct rxe_qp *qp, struct rxe_send_wqe *wqe) static int check_invalidate_mw(struct rxe_qp *qp, struct rxe_mw *mw) { - if (unlikely(mw->state != RXE_MEM_STATE_VALID)) { - pr_err_once("attempt to invalidate a MW that is not valid\n"); - return -EINVAL; - } - /* o10-37.2.26 */ if (unlikely(mw->ibmw.type == IB_MW_TYPE_1)) { pr_err_once("attempt to invalidate a type 1 MW\n"); @@ -336,9 +331,11 @@ static void do_invalidate_mw(struct rxe_mw *mw) { mw->qp = NULL; - rxe_drop_ref(mw->mr); - atomic_dec(&mw->mr->num_mw); - mw->mr = NULL; + if (mw->mr) { + atomic_dec(&mw->mr->num_mw); + mw->mr = NULL; + rxe_drop_ref(mw->mr); + } mw->access = 0; mw->addr = 0; @@ -364,6 +361,50 @@ int rxe_invalidate_mw(struct rxe_qp *qp, struct rxe_mw *mw) return ret; } +int rxe_mw_check_access(struct rxe_qp *qp, struct rxe_mw *mw, + int access, u64 va, u32 resid) +{ + struct rxe_pd *pd = to_rpd(mw->ibmw.pd); + + if (unlikely(mw->state != RXE_MEM_STATE_VALID)) { + pr_err_once("attempt to access a MW that is not valid\n"); + return -EINVAL; + } + + /* C10-76.2.1 */ + if (unlikely((mw->ibmw.type == IB_MW_TYPE_1) && (pd != qp->pd))) { + pr_err_once("attempt to access a type 1 MW with a different PD than the QP\n"); + return -EINVAL; + } + + /* o10-37.2.43 */ + if (unlikely((mw->ibmw.type == IB_MW_TYPE_2) && (mw->qp != qp))) { + pr_err_once("attempt to access a type 2 MW that is associated with a different QP\n"); + return -EINVAL; + } + + /* C10-77 */ + if (unlikely(access && !(access & mw->access))) { + pr_err_once("attempt to access a MW without sufficient access\n"); + return -EINVAL; + } + + if (mw->access & IB_ZERO_BASED) { + if (unlikely((va + resid) > mw->length)) { + pr_err_once("attempt to access a ZB MW out of bounds\n"); + return -EINVAL; + } + } else { + if (unlikely((va < mw->addr) || + ((va + resid) > (mw->addr + mw->length)))) { + pr_err_once("attempt to access a VA MW out of bounds\n"); + return -EINVAL; + } + } + + return 0; +} + void rxe_mw_cleanup(struct rxe_pool_entry *arg) { struct rxe_mw *mw = container_of(arg, typeof(*mw), pelem); diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c index 1944d2bbc43b..e4de0f4a3d69 100644 --- a/drivers/infiniband/sw/rxe/rxe_req.c +++ b/drivers/infiniband/sw/rxe/rxe_req.c @@ -645,7 +645,6 @@ int rxe_requester(void *arg) case IB_WR_BIND_MW: ret = rxe_bind_mw(qp, wqe); if (ret) { - wqe->state = wqe_state_done; wqe->status = IB_WC_MW_BIND_ERR; goto err; } @@ -653,6 +652,7 @@ int rxe_requester(void *arg) default: pr_err_once("unexpected LOCAL WR opcode = %d\n", wqe->wr.opcode); + wqe->status = IB_WC_LOC_QP_OP_ERR; goto err; } @@ -698,13 +698,7 @@ int rxe_requester(void *arg) payload = (mask & RXE_WRITE_OR_SEND) ? wqe->dma.resid : 0; if (payload > mtu) { if (qp_type(qp) == IB_QPT_UD) { - /* C10-93.1.1: If the total sum of all the buffer lengths specified for a - * UD message exceeds the MTU of the port as returned by QueryHCA, the CI - * shall not emit any packets for this message. Further, the CI shall not - * generate an error due to this condition. - */ - - /* fake a successful UD send */ + /* C10-93.1.1: fake a successful UD send */ wqe->first_psn = qp->req.psn; wqe->last_psn = qp->req.psn; qp->req.psn = (qp->req.psn + 1) & BTH_PSN_MASK; @@ -769,6 +763,8 @@ int rxe_requester(void *arg) * to be called again */ wqe->state = wqe_state_error; + qp->req.wqe_index = next_index(qp->sq.queue, + qp->req.wqe_index); __rxe_do_task(&qp->comp.task); ret = -EAGAIN; goto done; @@ -784,8 +780,7 @@ int rxe_requester(void *arg) again: /* we come here if we are done with the current wqe but want to - * get called again. Mostly we loop back to next wqe so should - * be all one way or the other + * get called again. */ ret = 0; goto done; diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index 4d688a50d301..91595c23bc16 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -393,6 +393,8 @@ static enum resp_states check_rkey(struct rxe_qp *qp, struct rxe_pkt_info *pkt) { struct rxe_mr *mr = NULL; + struct rxe_mw *mw = NULL; + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); u64 va; u32 rkey; u32 resid; @@ -400,6 +402,7 @@ static enum resp_states check_rkey(struct rxe_qp *qp, int mtu = qp->mtu; enum resp_states state; int access; + unsigned long flags; if (pkt->mask & (RXE_READ_MASK | RXE_WRITE_MASK)) { if (pkt->mask & RXE_RETH_MASK) { @@ -407,6 +410,7 @@ static enum resp_states check_rkey(struct rxe_qp *qp, qp->resp.rkey = reth_rkey(pkt); qp->resp.resid = reth_len(pkt); qp->resp.length = reth_len(pkt); + qp->resp.offset = 0; } access = (pkt->mask & RXE_READ_MASK) ? IB_ACCESS_REMOTE_READ : IB_ACCESS_REMOTE_WRITE; @@ -414,6 +418,7 @@ static enum resp_states check_rkey(struct rxe_qp *qp, qp->resp.va = atmeth_va(pkt); qp->resp.rkey = atmeth_rkey(pkt); qp->resp.resid = sizeof(u64); + qp->resp.offset = 0; access = IB_ACCESS_REMOTE_ATOMIC; } else { return RESPST_EXECUTE; @@ -431,20 +436,46 @@ static enum resp_states check_rkey(struct rxe_qp *qp, resid = qp->resp.resid; pktlen = payload_size(pkt); - mr = lookup_mr(qp->pd, access, rkey, lookup_remote); - if (!mr) { - state = RESPST_ERR_RKEY_VIOLATION; - goto err; - } + /* check rkey on each packet because someone could + * have invalidated, deallocated or unregistered it + * since the last packet + */ + if (rkey & IS_MW) { + mw = rxe_pool_get_key(&rxe->mw_pool, &rkey); + if (!mw) { + pr_err_once("no MW found with rkey = 0x%08x\n", rkey); + state = RESPST_ERR_RKEY_VIOLATION; + goto err; + } - if (unlikely(mr->state == RXE_MEM_STATE_FREE)) { - state = RESPST_ERR_RKEY_VIOLATION; - goto err; - } + spin_lock_irqsave(&mw->lock, flags); + if (rxe_mw_check_access(qp, mw, access, va, resid)) { + spin_unlock_irqrestore(&mw->lock, flags); + rxe_drop_ref(mw); + state = RESPST_ERR_RKEY_VIOLATION; + goto err; + } - if (mr_check_range(mr, va, resid)) { - state = RESPST_ERR_RKEY_VIOLATION; - goto err; + mr = mw->mr; + rxe_add_ref(mr); + + if (mw->access & IB_ZERO_BASED) + qp->resp.offset = mw->addr; + + spin_unlock_irqrestore(&mw->lock, flags); + rxe_drop_ref(mw); + } else { + mr = rxe_pool_get_key(&rxe->mr_pool, &rkey); + if (!mr || (mr->rkey != rkey)) { + pr_err_once("no MR found with rkey = 0x%08x\n", rkey); + state = RESPST_ERR_RKEY_VIOLATION; + goto err; + } + + if (rxe_mr_check_access(qp, mr, access, va, resid)) { + state = RESPST_ERR_RKEY_VIOLATION; + goto err; + } } if (pkt->mask & RXE_WRITE_MASK) { @@ -500,8 +531,8 @@ static enum resp_states write_data_in(struct rxe_qp *qp, int err; int data_len = payload_size(pkt); - err = rxe_mr_copy(qp->resp.mr, qp->resp.va, payload_addr(pkt), - data_len, to_mr_obj, NULL); + err = rxe_mr_copy(qp->resp.mr, qp->resp.va + qp->resp.offset, + payload_addr(pkt), data_len, to_mr_obj, NULL); if (err) { rc = RESPST_ERR_RKEY_VIOLATION; goto out; @@ -520,7 +551,6 @@ static DEFINE_SPINLOCK(atomic_ops_lock); static enum resp_states process_atomic(struct rxe_qp *qp, struct rxe_pkt_info *pkt) { - u64 iova = atmeth_va(pkt); u64 *vaddr; enum resp_states ret; struct rxe_mr *mr = qp->resp.mr; @@ -530,7 +560,7 @@ static enum resp_states process_atomic(struct rxe_qp *qp, goto out; } - vaddr = iova_to_vaddr(mr, iova, sizeof(u64)); + vaddr = iova_to_vaddr(mr, qp->resp.va + qp->resp.offset, sizeof(u64)); /* check vaddr is 8 bytes aligned. */ if (!vaddr || (uintptr_t)vaddr & 7) { @@ -655,8 +685,10 @@ static enum resp_states read_reply(struct rxe_qp *qp, res->type = RXE_READ_MASK; res->replay = 0; - res->read.va = qp->resp.va; - res->read.va_org = qp->resp.va; + res->read.va = qp->resp.va + + qp->resp.offset; + res->read.va_org = qp->resp.va + + qp->resp.offset; res->first_psn = req_pkt->psn; @@ -1336,7 +1368,10 @@ int rxe_responder(void *arg) /* Class C */ do_class_ac_error(qp, AETH_NAK_REM_ACC_ERR, IB_WC_REM_ACCESS_ERR); - state = RESPST_COMPLETE; + if (qp->resp.wqe) + state = RESPST_COMPLETE; + else + state = RESPST_ACKNOWLEDGE; } else { qp->resp.drop_msg = 1; if (qp->srq) { @@ -1364,7 +1399,10 @@ int rxe_responder(void *arg) /* Class C */ do_class_ac_error(qp, AETH_NAK_INVALID_REQ, IB_WC_REM_INV_REQ_ERR); - state = RESPST_COMPLETE; + if (qp->resp.wqe) + state = RESPST_COMPLETE; + else + state = RESPST_ACKNOWLEDGE; } else if (qp->srq) { /* UC/UD - class E */ qp->resp.status = IB_WC_REM_INV_REQ_ERR; @@ -1380,7 +1418,10 @@ int rxe_responder(void *arg) /* All, Class A. */ do_class_ac_error(qp, AETH_NAK_REM_OP_ERR, IB_WC_LOC_QP_OP_ERR); - state = RESPST_COMPLETE; + if (qp->resp.wqe) + state = RESPST_COMPLETE; + else + state = RESPST_ACKNOWLEDGE; break; case RESPST_ERR_CQ_OVERFLOW: diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.h b/drivers/infiniband/sw/rxe/rxe_verbs.h index 2fb5581edd8a..b24a9a0878c2 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.h +++ b/drivers/infiniband/sw/rxe/rxe_verbs.h @@ -183,6 +183,7 @@ struct rxe_resp_info { /* RDMA read / atomic only */ u64 va; + u64 offset; struct rxe_mr *mr; u32 resid; u32 rkey;

[for-next,v6,06/12] Add memory access through MWs

Commit Message

Patch