From patchwork Thu Jun 11 23:11:59 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: "Marciniszyn, Mike" X-Patchwork-Id: 6593291 Return-Path: X-Original-To: patchwork-linux-rdma@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork2.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.136]) by patchwork2.web.kernel.org (Postfix) with ESMTP id 3D1D2C0020 for ; Thu, 11 Jun 2015 23:12:07 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id E57A720623 for ; Thu, 11 Jun 2015 23:12:05 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 5EF842054A for ; Thu, 11 Jun 2015 23:12:04 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751850AbbFKXMB (ORCPT ); Thu, 11 Jun 2015 19:12:01 -0400 Received: from mga02.intel.com ([134.134.136.20]:39288 "EHLO mga02.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752361AbbFKXMB (ORCPT ); Thu, 11 Jun 2015 19:12:01 -0400 Received: from fmsmga003.fm.intel.com ([10.253.24.29]) by orsmga101.jf.intel.com with ESMTP; 11 Jun 2015 16:12:00 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.13,598,1427785200"; d="scan'208";a="506937849" Received: from sedona.ch.intel.com ([143.182.228.65]) by FMSMGA003.fm.intel.com with ESMTP; 11 Jun 2015 16:12:00 -0700 Received: from phlsvsles11.ph.intel.com (phlsvsles11.ph.intel.com [10.228.195.43]) by sedona.ch.intel.com (8.13.6/8.14.3/Standard MailSET/Hub) with ESMTP id t5BNBx5E003959; Thu, 11 Jun 2015 16:11:59 -0700 Received: from phlsvslse11.ph.intel.com (localhost [127.0.0.1]) by phlsvsles11.ph.intel.com with ESMTP id t5BNBxZd017334; Thu, 11 Jun 2015 19:11:59 -0400 Subject: [PATCH 39/41] IB/hfi1: add multicast routines To: dledford@redhat.com From: Mike Marciniszyn Cc: linux-rdma@vger.kernel.org Date: Thu, 11 Jun 2015 19:11:59 -0400 Message-ID: <20150611231159.16479.16513.stgit@phlsvslse11.ph.intel.com> In-Reply-To: <20150611230710.16479.62955.stgit@phlsvslse11.ph.intel.com> References: <20150611230710.16479.62955.stgit@phlsvslse11.ph.intel.com> User-Agent: StGit/0.16 MIME-Version: 1.0 Sender: linux-rdma-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, T_RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP Signed-off-by: Andrew Friedley Signed-off-by: Arthur Kepner Signed-off-by: Brendan Cunningham Signed-off-by: Brian Welty Signed-off-by: Caz Yokoyama Signed-off-by: Dean Luick Signed-off-by: Dennis Dalessandro Signed-off-by: Easwar Hariharan Signed-off-by: Harish Chegondi Signed-off-by: Ira Weiny Signed-off-by: Jim Snow Signed-off-by: John Gregor Signed-off-by: Jubin John Signed-off-by: Kaike Wan Signed-off-by: Kevin Pine Signed-off-by: Kyle Liddell Signed-off-by: Mike Marciniszyn Signed-off-by: Mitko Haralanov Signed-off-by: Ravi Krishnaswamy Signed-off-by: Sadanand Warrier Signed-off-by: Sanath Kumar Signed-off-by: Sudeep Dutt Signed-off-by: Vlad Danushevsky --- drivers/infiniband/hw/hfi1/verbs_mcast.c | 385 ++++++++++++++++++++++++++++++ 1 file changed, 385 insertions(+) create mode 100644 drivers/infiniband/hw/hfi1/verbs_mcast.c -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html diff --git a/drivers/infiniband/hw/hfi1/verbs_mcast.c b/drivers/infiniband/hw/hfi1/verbs_mcast.c new file mode 100644 index 0000000..afc6b4c --- /dev/null +++ b/drivers/infiniband/hw/hfi1/verbs_mcast.c @@ -0,0 +1,385 @@ +/* + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * Copyright(c) 2015 Intel Corporation. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Copyright(c) 2015 Intel Corporation. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in + * the documentation and/or other materials provided with the + * distribution. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS + * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT + * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR + * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT + * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, + * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT + * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, + * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY + * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT + * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE + * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + * + */ + +#include + +#include "hfi.h" + +/** + * mcast_qp_alloc - alloc a struct to link a QP to mcast GID struct + * @qp: the QP to link + */ +static struct hfi1_mcast_qp *mcast_qp_alloc(struct hfi1_qp *qp) +{ + struct hfi1_mcast_qp *mqp; + + mqp = kmalloc(sizeof(*mqp), GFP_KERNEL); + if (!mqp) + goto bail; + + mqp->qp = qp; + atomic_inc(&qp->refcount); + +bail: + return mqp; +} + +static void mcast_qp_free(struct hfi1_mcast_qp *mqp) +{ + struct hfi1_qp *qp = mqp->qp; + + /* Notify hfi1_destroy_qp() if it is waiting. */ + if (atomic_dec_and_test(&qp->refcount)) + wake_up(&qp->wait); + + kfree(mqp); +} + +/** + * mcast_alloc - allocate the multicast GID structure + * @mgid: the multicast GID + * + * A list of QPs will be attached to this structure. + */ +static struct hfi1_mcast *mcast_alloc(union ib_gid *mgid) +{ + struct hfi1_mcast *mcast; + + mcast = kmalloc(sizeof(*mcast), GFP_KERNEL); + if (!mcast) + goto bail; + + mcast->mgid = *mgid; + INIT_LIST_HEAD(&mcast->qp_list); + init_waitqueue_head(&mcast->wait); + atomic_set(&mcast->refcount, 0); + mcast->n_attached = 0; + +bail: + return mcast; +} + +static void mcast_free(struct hfi1_mcast *mcast) +{ + struct hfi1_mcast_qp *p, *tmp; + + list_for_each_entry_safe(p, tmp, &mcast->qp_list, list) + mcast_qp_free(p); + + kfree(mcast); +} + +/** + * hfi1_mcast_find - search the global table for the given multicast GID + * @ibp: the IB port structure + * @mgid: the multicast GID to search for + * + * Returns NULL if not found. + * + * The caller is responsible for decrementing the reference count if found. + */ +struct hfi1_mcast *hfi1_mcast_find(struct hfi1_ibport *ibp, union ib_gid *mgid) +{ + struct rb_node *n; + unsigned long flags; + struct hfi1_mcast *mcast; + + spin_lock_irqsave(&ibp->lock, flags); + n = ibp->mcast_tree.rb_node; + while (n) { + int ret; + + mcast = rb_entry(n, struct hfi1_mcast, rb_node); + + ret = memcmp(mgid->raw, mcast->mgid.raw, + sizeof(union ib_gid)); + if (ret < 0) + n = n->rb_left; + else if (ret > 0) + n = n->rb_right; + else { + atomic_inc(&mcast->refcount); + spin_unlock_irqrestore(&ibp->lock, flags); + goto bail; + } + } + spin_unlock_irqrestore(&ibp->lock, flags); + + mcast = NULL; + +bail: + return mcast; +} + +/** + * mcast_add - insert mcast GID into table and attach QP struct + * @mcast: the mcast GID table + * @mqp: the QP to attach + * + * Return zero if both were added. Return EEXIST if the GID was already in + * the table but the QP was added. Return ESRCH if the QP was already + * attached and neither structure was added. + */ +static int mcast_add(struct hfi1_ibdev *dev, struct hfi1_ibport *ibp, + struct hfi1_mcast *mcast, struct hfi1_mcast_qp *mqp) +{ + struct rb_node **n = &ibp->mcast_tree.rb_node; + struct rb_node *pn = NULL; + int ret; + + spin_lock_irq(&ibp->lock); + + while (*n) { + struct hfi1_mcast *tmcast; + struct hfi1_mcast_qp *p; + + pn = *n; + tmcast = rb_entry(pn, struct hfi1_mcast, rb_node); + + ret = memcmp(mcast->mgid.raw, tmcast->mgid.raw, + sizeof(union ib_gid)); + if (ret < 0) { + n = &pn->rb_left; + continue; + } + if (ret > 0) { + n = &pn->rb_right; + continue; + } + + /* Search the QP list to see if this is already there. */ + list_for_each_entry_rcu(p, &tmcast->qp_list, list) { + if (p->qp == mqp->qp) { + ret = ESRCH; + goto bail; + } + } + if (tmcast->n_attached == hfi1_max_mcast_qp_attached) { + ret = ENOMEM; + goto bail; + } + + tmcast->n_attached++; + + list_add_tail_rcu(&mqp->list, &tmcast->qp_list); + ret = EEXIST; + goto bail; + } + + spin_lock(&dev->n_mcast_grps_lock); + if (dev->n_mcast_grps_allocated == hfi1_max_mcast_grps) { + spin_unlock(&dev->n_mcast_grps_lock); + ret = ENOMEM; + goto bail; + } + + dev->n_mcast_grps_allocated++; + spin_unlock(&dev->n_mcast_grps_lock); + + mcast->n_attached++; + + list_add_tail_rcu(&mqp->list, &mcast->qp_list); + + atomic_inc(&mcast->refcount); + rb_link_node(&mcast->rb_node, pn, n); + rb_insert_color(&mcast->rb_node, &ibp->mcast_tree); + + ret = 0; + +bail: + spin_unlock_irq(&ibp->lock); + + return ret; +} + +int hfi1_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) +{ + struct hfi1_qp *qp = to_iqp(ibqp); + struct hfi1_ibdev *dev = to_idev(ibqp->device); + struct hfi1_ibport *ibp; + struct hfi1_mcast *mcast; + struct hfi1_mcast_qp *mqp; + int ret; + + if (ibqp->qp_num <= 1 || qp->state == IB_QPS_RESET) { + ret = -EINVAL; + goto bail; + } + + /* + * Allocate data structures since its better to do this outside of + * spin locks and it will most likely be needed. + */ + mcast = mcast_alloc(gid); + if (mcast == NULL) { + ret = -ENOMEM; + goto bail; + } + mqp = mcast_qp_alloc(qp); + if (mqp == NULL) { + mcast_free(mcast); + ret = -ENOMEM; + goto bail; + } + ibp = to_iport(ibqp->device, qp->port_num); + switch (mcast_add(dev, ibp, mcast, mqp)) { + case ESRCH: + /* Neither was used: OK to attach the same QP twice. */ + mcast_qp_free(mqp); + mcast_free(mcast); + break; + + case EEXIST: /* The mcast wasn't used */ + mcast_free(mcast); + break; + + case ENOMEM: + /* Exceeded the maximum number of mcast groups. */ + mcast_qp_free(mqp); + mcast_free(mcast); + ret = -ENOMEM; + goto bail; + + default: + break; + } + + ret = 0; + +bail: + return ret; +} + +int hfi1_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) +{ + struct hfi1_qp *qp = to_iqp(ibqp); + struct hfi1_ibdev *dev = to_idev(ibqp->device); + struct hfi1_ibport *ibp = to_iport(ibqp->device, qp->port_num); + struct hfi1_mcast *mcast = NULL; + struct hfi1_mcast_qp *p, *tmp; + struct rb_node *n; + int last = 0; + int ret; + + if (ibqp->qp_num <= 1 || qp->state == IB_QPS_RESET) { + ret = -EINVAL; + goto bail; + } + + spin_lock_irq(&ibp->lock); + + /* Find the GID in the mcast table. */ + n = ibp->mcast_tree.rb_node; + while (1) { + if (n == NULL) { + spin_unlock_irq(&ibp->lock); + ret = -EINVAL; + goto bail; + } + + mcast = rb_entry(n, struct hfi1_mcast, rb_node); + ret = memcmp(gid->raw, mcast->mgid.raw, + sizeof(union ib_gid)); + if (ret < 0) + n = n->rb_left; + else if (ret > 0) + n = n->rb_right; + else + break; + } + + /* Search the QP list. */ + list_for_each_entry_safe(p, tmp, &mcast->qp_list, list) { + if (p->qp != qp) + continue; + /* + * We found it, so remove it, but don't poison the forward + * link until we are sure there are no list walkers. + */ + list_del_rcu(&p->list); + mcast->n_attached--; + + /* If this was the last attached QP, remove the GID too. */ + if (list_empty(&mcast->qp_list)) { + rb_erase(&mcast->rb_node, &ibp->mcast_tree); + last = 1; + } + break; + } + + spin_unlock_irq(&ibp->lock); + + if (p) { + /* + * Wait for any list walkers to finish before freeing the + * list element. + */ + wait_event(mcast->wait, atomic_read(&mcast->refcount) <= 1); + mcast_qp_free(p); + } + if (last) { + atomic_dec(&mcast->refcount); + wait_event(mcast->wait, !atomic_read(&mcast->refcount)); + mcast_free(mcast); + spin_lock_irq(&dev->n_mcast_grps_lock); + dev->n_mcast_grps_allocated--; + spin_unlock_irq(&dev->n_mcast_grps_lock); + } + + ret = 0; + +bail: + return ret; +} + +int hfi1_mcast_tree_empty(struct hfi1_ibport *ibp) +{ + return ibp->mcast_tree.rb_node == NULL; +}