From patchwork Wed Nov 8 10:06:15 2017 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Sagi Grimberg X-Patchwork-Id: 10048205 Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork.web.codeaurora.org (Postfix) with ESMTP id 0DD9D6032D for ; Wed, 8 Nov 2017 10:06:53 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id E9B6029514 for ; Wed, 8 Nov 2017 10:06:52 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id DED9629584; Wed, 8 Nov 2017 10:06:52 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-5.1 required=2.0 tests=BAYES_00,DKIM_SIGNED, RCVD_IN_DNSWL_HI, T_DKIM_INVALID, URIBL_BLACK autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 65BCB29514 for ; Wed, 8 Nov 2017 10:06:52 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751448AbdKHKGu (ORCPT ); Wed, 8 Nov 2017 05:06:50 -0500 Received: from bombadil.infradead.org ([65.50.211.133]:57685 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751141AbdKHKGs (ORCPT ); Wed, 8 Nov 2017 05:06:48 -0500 DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=infradead.org; s=bombadil.20170209; h=References:In-Reply-To:Message-Id: Date:Subject:Cc:To:From:Sender:Reply-To:MIME-Version:Content-Type: Content-Transfer-Encoding:Content-ID:Content-Description:Resent-Date: Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID:List-Id: List-Help:List-Unsubscribe:List-Subscribe:List-Post:List-Owner:List-Archive; bh=r6jghvPVq8Sw7+204Y8hbLaRmzefxHawO78kWySPsAw=; b=b5ZM1OWg+F/OqMt68OANhim5o U81MrskR4Hw2VgZ3PnfMZ678To5/Rg9CCrj07TLoyL7NmFITvLBTEutHYrKs8FBZs/57Cjuu5gERJ mBbA2QNFFRrtmWG0oLQX56iUHTS8HqLLHR3QSA6CAuvrWtigfGldYlHE69lb9YkOZh9PsxCF+wiAq JNYYNOroQ+We4mIT19ZG4roT97skw4qV9lqoWnvo9kvVo89HCXUdPO4cAtoMhzxbcgyuo7NfS99f+ wau0XKPpvQbMxhfEx5hu2cvJxeEpHQDlb7DnJneer/fuVrjSn0SsVEaGQ//wokd9e4qQikkyeYpoI 7EH5lcx4g==; Received: from [31.154.58.122] (helo=bombadil.infradead.org) by bombadil.infradead.org with esmtpsa (Exim 4.87 #1 (Red Hat Linux)) id 1eCNFt-0001fb-V1; Wed, 08 Nov 2017 10:06:46 +0000 From: Sagi Grimberg To: linux-rdma@vger.kernel.org, linux-nvme@lists.infradead.org Cc: Christoph Hellwig , Max Gurtuvoy Subject: [PATCH v2 2/3] nvme-rdma: don't complete requests before a send work request has completed Date: Wed, 8 Nov 2017 12:06:15 +0200 Message-Id: <20171108100616.26605-3-sagi@grimberg.me> X-Mailer: git-send-email 2.14.1 In-Reply-To: <20171108100616.26605-1-sagi@grimberg.me> References: <20171108100616.26605-1-sagi@grimberg.me> Sender: linux-rdma-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP In order to guarantee that the HCA will never get an access violation (either from invalidated rkey or from iommu) when retrying a send operation we must complete a request only when both send completion and the nvme cqe has arrived. We need to set the send/recv completions flags atomically because we might have more than a single context accessing the request concurrently (one is cq irq-poll context and the other is user-polling used in IOCB_HIPRI). Only then we are safe to invalidate the rkey (if needed), unmap the host buffers, and complete the IO. Signed-off-by: Sagi Grimberg --- drivers/nvme/host/rdma.c | 52 ++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 44 insertions(+), 8 deletions(-) diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c index c765f1d20141..998f7cb445d9 100644 --- a/drivers/nvme/host/rdma.c +++ b/drivers/nvme/host/rdma.c @@ -67,6 +67,10 @@ struct nvme_rdma_request { struct nvme_request req; struct ib_mr *mr; struct nvme_rdma_qe sqe; + struct nvme_completion cqe; + spinlock_t lock; + bool send_completed; + bool resp_completed; struct ib_sge sge[1 + NVME_RDMA_MAX_INLINE_SEGMENTS]; u32 num_sge; int nents; @@ -332,6 +336,8 @@ static int nvme_rdma_init_request(struct blk_mq_tag_set *set, struct ib_device *ibdev = dev->dev; int ret; + spin_lock_init(&req->lock); + ret = nvme_rdma_alloc_qe(ibdev, &req->sqe, sizeof(struct nvme_command), DMA_TO_DEVICE); if (ret) @@ -1154,6 +1160,8 @@ static int nvme_rdma_map_data(struct nvme_rdma_queue *queue, req->num_sge = 1; req->inline_data = false; req->mr->need_inval = false; + req->send_completed = false; + req->resp_completed = false; c->common.flags |= NVME_CMD_SGL_METABUF; @@ -1190,13 +1198,30 @@ static int nvme_rdma_map_data(struct nvme_rdma_queue *queue, static void nvme_rdma_send_done(struct ib_cq *cq, struct ib_wc *wc) { - if (unlikely(wc->status != IB_WC_SUCCESS)) + struct nvme_rdma_qe *qe = + container_of(wc->wr_cqe, struct nvme_rdma_qe, cqe); + struct nvme_rdma_request *req = + container_of(qe, struct nvme_rdma_request, sqe); + struct request *rq = blk_mq_rq_from_pdu(req); + unsigned long flags; + bool end; + + if (unlikely(wc->status != IB_WC_SUCCESS)) { nvme_rdma_wr_error(cq, wc, "SEND"); + return; + } + + spin_lock_irqsave(&req->lock, flags); + req->send_completed = true; + end = req->resp_completed; + spin_unlock_irqrestore(&req->lock, flags); + if (end) + nvme_end_request(rq, req->cqe.status, req->cqe.result); } static int nvme_rdma_post_send(struct nvme_rdma_queue *queue, struct nvme_rdma_qe *qe, struct ib_sge *sge, u32 num_sge, - struct ib_send_wr *first) + struct ib_send_wr *first, bool signal) { struct ib_send_wr wr, *bad_wr; int ret; @@ -1212,7 +1237,7 @@ static int nvme_rdma_post_send(struct nvme_rdma_queue *queue, wr.sg_list = sge; wr.num_sge = num_sge; wr.opcode = IB_WR_SEND; - wr.send_flags = IB_SEND_SIGNALED; + wr.send_flags = signal ? IB_SEND_SIGNALED : 0; if (first) first->next = ≀ @@ -1286,7 +1311,7 @@ static void nvme_rdma_submit_async_event(struct nvme_ctrl *arg, int aer_idx) ib_dma_sync_single_for_device(dev, sqe->dma, sizeof(*cmd), DMA_TO_DEVICE); - ret = nvme_rdma_post_send(queue, sqe, &sge, 1, NULL); + ret = nvme_rdma_post_send(queue, sqe, &sge, 1, NULL, false); WARN_ON_ONCE(ret); } @@ -1296,6 +1321,8 @@ static int nvme_rdma_process_nvme_rsp(struct nvme_rdma_queue *queue, struct request *rq; struct nvme_rdma_request *req; int ret = 0; + unsigned long flags; + bool end; rq = blk_mq_tag_to_rq(nvme_rdma_tagset(queue), cqe->command_id); if (!rq) { @@ -1307,14 +1334,23 @@ static int nvme_rdma_process_nvme_rsp(struct nvme_rdma_queue *queue, } req = blk_mq_rq_to_pdu(rq); - if (rq->tag == tag) - ret = 1; + req->cqe.status = cqe->status; + req->cqe.result = cqe->result; if ((wc->wc_flags & IB_WC_WITH_INVALIDATE) && wc->ex.invalidate_rkey == req->mr->rkey) req->mr->need_inval = false; - nvme_end_request(rq, cqe->status, cqe->result); + spin_lock_irqsave(&req->lock, flags); + req->resp_completed = true; + end = req->send_completed; + spin_unlock_irqrestore(&req->lock, flags); + if (end) { + if (rq->tag == tag) + ret = 1; + nvme_end_request(rq, req->cqe.status, req->cqe.result); + } + return ret; } @@ -1603,7 +1639,7 @@ static blk_status_t nvme_rdma_queue_rq(struct blk_mq_hw_ctx *hctx, sizeof(struct nvme_command), DMA_TO_DEVICE); err = nvme_rdma_post_send(queue, sqe, req->sge, req->num_sge, - req->mr->need_inval ? &req->reg_wr.wr : NULL); + req->mr->need_inval ? &req->reg_wr.wr : NULL, true); if (unlikely(err)) { nvme_rdma_unmap_data(queue, rq); goto err;