From patchwork Thu Nov 15 21:44:44 2018 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Juliet Kim X-Patchwork-Id: 998543 X-Patchwork-Delegate: davem@davemloft.net Return-Path: X-Original-To: patchwork-incoming-netdev@ozlabs.org Delivered-To: patchwork-incoming-netdev@ozlabs.org Authentication-Results: ozlabs.org; spf=none (mailfrom) smtp.mailfrom=vger.kernel.org (client-ip=209.132.180.67; helo=vger.kernel.org; envelope-from=netdev-owner@vger.kernel.org; receiver=) Authentication-Results: ozlabs.org; dmarc=none (p=none dis=none) header.from=linux.vnet.ibm.com Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by ozlabs.org (Postfix) with ESMTP id 42wvzK60vNz9sB5 for ; Fri, 16 Nov 2018 08:45:01 +1100 (AEDT) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1726258AbeKPHy1 (ORCPT ); Fri, 16 Nov 2018 02:54:27 -0500 Received: from mx0b-001b2d01.pphosted.com ([148.163.158.5]:53004 "EHLO mx0a-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1725860AbeKPHy1 (ORCPT ); Fri, 16 Nov 2018 02:54:27 -0500 Received: from pps.filterd (m0098420.ppops.net [127.0.0.1]) by mx0b-001b2d01.pphosted.com (8.16.0.22/8.16.0.22) with SMTP id wAFLibL8051689 for ; Thu, 15 Nov 2018 16:44:53 -0500 Received: from e32.co.us.ibm.com (e32.co.us.ibm.com [32.97.110.150]) by mx0b-001b2d01.pphosted.com with ESMTP id 2nsewyepq9-1 (version=TLSv1.2 cipher=AES256-GCM-SHA384 bits=256 verify=NOT) for ; Thu, 15 Nov 2018 16:44:52 -0500 Received: from localhost by e32.co.us.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Thu, 15 Nov 2018 21:44:52 -0000 Received: from b03cxnp08028.gho.boulder.ibm.com (9.17.130.20) by e32.co.us.ibm.com (192.168.1.132) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; (version=TLSv1/SSLv3 cipher=AES256-GCM-SHA384 bits=256/256) Thu, 15 Nov 2018 21:44:47 -0000 Received: from b03ledav001.gho.boulder.ibm.com (b03ledav001.gho.boulder.ibm.com [9.17.130.232]) by b03cxnp08028.gho.boulder.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id wAFLikwj19202070 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=FAIL); Thu, 15 Nov 2018 21:44:46 GMT Received: from b03ledav001.gho.boulder.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 89E7C6E052; Thu, 15 Nov 2018 21:44:46 +0000 (GMT) Received: from b03ledav001.gho.boulder.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id A71236E04C; Thu, 15 Nov 2018 21:44:45 +0000 (GMT) Received: from juliets-mbp.austin.ibm.com (unknown [9.41.179.214]) by b03ledav001.gho.boulder.ibm.com (Postfix) with ESMTPS; Thu, 15 Nov 2018 21:44:45 +0000 (GMT) To: linuxppc-dev@lists.ozlabs.org, netdev@vger.kernel.org, linux-kernel@vger.kernel.org Cc: Benjamin Herrenschmidt , Paul Mackerras , Michael Ellerman , Thomas Falcon , Michael Bringmann , "David S. Miller" , Tyrel Datwyler From: Juliet Kim Subject: [PATCH] net/ibmnvic: Fix deadlock problem in reset Date: Thu, 15 Nov 2018 15:44:44 -0600 User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.14; rv:60.0) Gecko/20100101 Thunderbird/60.3.0 MIME-Version: 1.0 Content-Language: en-US X-TM-AS-GCONF: 00 x-cbid: 18111521-0004-0000-0000-000014B33762 X-IBM-SpamModules-Scores: X-IBM-SpamModules-Versions: BY=3.00010057; HX=3.00000242; KW=3.00000007; PH=3.00000004; SC=3.00000270; SDB=6.01117969; UDB=6.00579908; IPR=6.00898078; MB=3.00024183; MTD=3.00000008; XFM=3.00000015; UTC=2018-11-15 21:44:50 X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 18111521-0005-0000-0000-000089861661 Message-Id: <25faa351-51d3-3aeb-a1c9-e8d3259604dd@linux.vnet.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:, , definitions=2018-11-15_16:, , signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 priorityscore=1501 malwarescore=0 suspectscore=2 phishscore=0 bulkscore=0 spamscore=0 clxscore=1011 lowpriorityscore=0 mlxscore=0 impostorscore=0 mlxlogscore=999 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.0.1-1810050000 definitions=main-1811150189 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Subject: [PATCH] net/ibmnvic: Fix deadlock problem in reset From: Juliet Kim This patch changes to use rtnl_lock only during a reset to avoid deadlock that could occur when a thread operating close is holding rtnl_lock and waiting for reset_lock acquired by another thread, which is waiting for rtnl_lock in order to set the number of tx/rx queues during a reset. Also, we now setting the number of tx/rx queues during a soft reset for failover or LPM events. Signed-off-by: Juliet Kim ---  drivers/net/ethernet/ibm/ibmvnic.c |   59 +++++++++++++-----------------------  drivers/net/ethernet/ibm/ibmvnic.h |    2 +  2 files changed, 22 insertions(+), 39 deletions(-) diff --git a/drivers/net/ethernet/ibm/ibmvnic.c b/drivers/net/ethernet/ibm/ibmvnic.c index 7893bef..4a5de59 100644 --- a/drivers/net/ethernet/ibm/ibmvnic.c +++ b/drivers/net/ethernet/ibm/ibmvnic.c @@ -1103,20 +1103,15 @@ static int ibmvnic_open(struct net_device *netdev)         return 0;     } -   mutex_lock(&adapter->reset_lock); -     if (adapter->state != VNIC_CLOSED) {         rc = ibmvnic_login(netdev); -       if (rc) { -           mutex_unlock(&adapter->reset_lock); +       if (rc)             return rc; -       }         rc = init_resources(adapter);         if (rc) {             netdev_err(netdev, "failed to initialize resources\n");             release_resources(adapter); -           mutex_unlock(&adapter->reset_lock);             return rc;         }     } @@ -1124,8 +1119,6 @@ static int ibmvnic_open(struct net_device *netdev)     rc = __ibmvnic_open(netdev);     netif_carrier_on(netdev); -   mutex_unlock(&adapter->reset_lock); -     return rc;  } @@ -1269,10 +1262,8 @@ static int ibmvnic_close(struct net_device *netdev)         return 0;     } -   mutex_lock(&adapter->reset_lock);     rc = __ibmvnic_close(netdev);     ibmvnic_cleanup(netdev); -   mutex_unlock(&adapter->reset_lock);     return rc;  } @@ -1820,20 +1811,15 @@ static int do_reset(struct ibmvnic_adapter *adapter,                 return rc;         } else if (adapter->req_rx_queues != old_num_rx_queues ||                adapter->req_tx_queues != old_num_tx_queues) { -           adapter->map_id = 1;             release_rx_pools(adapter);             release_tx_pools(adapter); -           rc = init_rx_pools(netdev); -           if (rc) -               return rc; -           rc = init_tx_pools(netdev); -           if (rc) -               return rc; -             release_napi(adapter); -           rc = init_napi(adapter); +           release_vpd_data(adapter); + +           rc = init_resources(adapter);             if (rc)                 return rc; +         } else {             rc = reset_tx_pools(adapter);             if (rc) @@ -1917,17 +1903,8 @@ static int do_hard_reset(struct ibmvnic_adapter *adapter,         adapter->state = VNIC_PROBED;         return 0;     } -   /* netif_set_real_num_xx_queues needs to take rtnl lock here -    * unless wait_for_reset is set, in which case the rtnl lock -    * has already been taken before initializing the reset -    */ -   if (!adapter->wait_for_reset) { -       rtnl_lock(); -       rc = init_resources(adapter); -       rtnl_unlock(); -   } else { -       rc = init_resources(adapter); -   } + +   rc = init_resources(adapter);     if (rc)         return rc; @@ -1986,13 +1963,21 @@ static void __ibmvnic_reset(struct work_struct *work)     struct ibmvnic_rwi *rwi;     struct ibmvnic_adapter *adapter;     struct net_device *netdev; +   bool we_lock_rtnl = false;     u32 reset_state;     int rc = 0;     adapter = container_of(work, struct ibmvnic_adapter, ibmvnic_reset);     netdev = adapter->netdev; -   mutex_lock(&adapter->reset_lock); +   /* netif_set_real_num_xx_queues needs to take rtnl lock here +    * unless wait_for_reset is set, in which case the rtnl lock +    * has already been taken before initializing the reset +    */ +   if (!adapter->wait_for_reset) { +       rtnl_lock(); +       we_lock_rtnl = true; +   }     reset_state = adapter->state;     rwi = get_next_rwi(adapter); @@ -2020,12 +2005,11 @@ static void __ibmvnic_reset(struct work_struct *work)     if (rc) {         netdev_dbg(adapter->netdev, "Reset failed\n");         free_all_rwi(adapter); -       mutex_unlock(&adapter->reset_lock); -       return;     }     adapter->resetting = false; -   mutex_unlock(&adapter->reset_lock); +   if (we_lock_rtnl) +       rtnl_unlock();  }  static int ibmvnic_reset(struct ibmvnic_adapter *adapter, @@ -4768,7 +4752,6 @@ static int ibmvnic_probe(struct vio_dev *dev, const struct vio_device_id *id)     INIT_WORK(&adapter->ibmvnic_reset, __ibmvnic_reset);     INIT_LIST_HEAD(&adapter->rwi_list); -   mutex_init(&adapter->reset_lock);     mutex_init(&adapter->rwi_lock);     adapter->resetting = false; @@ -4840,8 +4823,8 @@ static int ibmvnic_remove(struct vio_dev *dev)     struct ibmvnic_adapter *adapter = netdev_priv(netdev);     adapter->state = VNIC_REMOVING; -   unregister_netdev(netdev); -   mutex_lock(&adapter->reset_lock); +   rtnl_lock(); +   unregister_netdevice(netdev);     release_resources(adapter);     release_sub_crqs(adapter, 1); @@ -4852,7 +4835,7 @@ static int ibmvnic_remove(struct vio_dev *dev)     adapter->state = VNIC_REMOVED; -   mutex_unlock(&adapter->reset_lock); +   rtnl_unlock();     device_remove_file(&dev->dev, &dev_attr_failover);     free_netdev(netdev);     dev_set_drvdata(&dev->dev, NULL); diff --git a/drivers/net/ethernet/ibm/ibmvnic.h b/drivers/net/ethernet/ibm/ibmvnic.h index 18103b8..99c4f8d 100644 --- a/drivers/net/ethernet/ibm/ibmvnic.h +++ b/drivers/net/ethernet/ibm/ibmvnic.h @@ -1075,7 +1075,7 @@ struct ibmvnic_adapter {     struct tasklet_struct tasklet;     enum vnic_state state;     enum ibmvnic_reset_reason reset_reason; -   struct mutex reset_lock, rwi_lock; +   struct mutex rwi_lock;     struct list_head rwi_list;     struct work_struct ibmvnic_reset;     bool resetting;