From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-8.2 required=3.0 tests=DKIMWL_WL_HIGH,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH, MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_PASS autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id CD8C1C282C0 for ; Fri, 25 Jan 2019 03:40:25 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id B94D221855 for ; Fri, 25 Jan 2019 03:40:24 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (2048-bit key) header.d=oracle.com header.i=@oracle.com header.b="AXr9fk9g" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728032AbfAYDkX (ORCPT ); Thu, 24 Jan 2019 22:40:23 -0500 Received: from userp2130.oracle.com ([156.151.31.86]:35576 "EHLO userp2130.oracle.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726304AbfAYDkX (ORCPT ); Thu, 24 Jan 2019 22:40:23 -0500 Received: from pps.filterd (userp2130.oracle.com [127.0.0.1]) by userp2130.oracle.com (8.16.0.22/8.16.0.22) with SMTP id x0P3doAC080653; Fri, 25 Jan 2019 03:40:18 GMT DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=oracle.com; h=subject : to : references : from : message-id : date : mime-version : in-reply-to : content-type : content-transfer-encoding; s=corp-2018-07-02; bh=W7eTkN6YvPpRIkWpGbmFilXufPFJ2UYuOWc4sIfKIwU=; b=AXr9fk9g9Dj9DZPGehLP2NZXLAwnYD+oys69mz7eijc2j0XVr4uZ126nrGs4s461fckY XhMaSqQToMNCDMy2huh7MnsSwc+JHNIx9491tIZjz6W04MqdVGxs7NK0b/GcU66nBaQl lS5J8r9QGjBg3eQagmYUmlQQgkldgFBfYEcF2No+vp5k/+xuAYwIhQ+DdRbbBSEHs3kp FxkLkUiiJ9J+2stHpqoX9awhSbjHYmHnN8PvNyJfjat6EIdgblczxFpZ5+GXh/nEYogv GALiJ2tXvp+f51imh98tcWil55YQ3Wkop37h4tKgdxMIyPTqKeWHz/7cOerXtxkhITuC Tg== Received: from aserv0022.oracle.com (aserv0022.oracle.com [141.146.126.234]) by userp2130.oracle.com with ESMTP id 2q3uav3g3w-1 (version=TLSv1.2 cipher=ECDHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Fri, 25 Jan 2019 03:40:18 +0000 Received: from aserv0121.oracle.com (aserv0121.oracle.com [141.146.126.235]) by aserv0022.oracle.com (8.14.4/8.14.4) with ESMTP id x0P3eHwc013451 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Fri, 25 Jan 2019 03:40:17 GMT Received: from abhmp0019.oracle.com (abhmp0019.oracle.com [141.146.116.25]) by aserv0121.oracle.com (8.14.4/8.13.8) with ESMTP id x0P3eHfV000460; Fri, 25 Jan 2019 03:40:17 GMT Received: from [10.190.142.77] (/192.188.170.109) by default (Oracle Beehive Gateway v4.0) with ESMTP ; Thu, 24 Jan 2019 19:40:17 -0800 Subject: Re: [PATCH v2] Btrfs: avoid deadlock with memory reclaim due to allocation of devices To: fdmanana@kernel.org, linux-btrfs@vger.kernel.org, dsterba@suse.cz References: <20181213211725.14832-1-fdmanana@kernel.org> <20190111171759.19920-1-fdmanana@kernel.org> From: Anand Jain Message-ID: Date: Fri, 25 Jan 2019 11:40:13 +0800 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:52.0) Gecko/20100101 Thunderbird/52.9.1 MIME-Version: 1.0 In-Reply-To: <20190111171759.19920-1-fdmanana@kernel.org> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 7bit X-Proofpoint-Virus-Version: vendor=nai engine=5900 definitions=9146 signatures=668682 X-Proofpoint-Spam-Details: rule=notspam policy=default score=0 suspectscore=2 malwarescore=0 phishscore=0 bulkscore=0 spamscore=0 mlxscore=0 mlxlogscore=999 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.0.1-1810050000 definitions=main-1901250026 Sender: linux-btrfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-btrfs@vger.kernel.org On 01/12/2019 01:17 AM, fdmanana@kernel.org wrote: > From: Filipe Manana > > In a few places we are allocating a device using the GFP_KERNEL flag when > it is not safe to do so, because if reclaim is triggered it can cause a > transaction commit while we are holding the device list mutex. This mutex > is required in the transaction commit path (at write_all_supers() and > btrfs_update_commit_device_size()). > > So fix this by setting up a nofs memory allocation context in those cases. > > Fixes: 78f2c9e6dbb14 ("btrfs: device add and remove: use GFP_KERNEL") > Fixes: e0ae999414238 ("btrfs: preallocate device flush bio") > Signed-off-by: Filipe Manana > --- > > V2: Change the approach to fix the problem by setting up nofs contextes > where needed. So remaining functions which still does GFP_KERNEL are.. btrfs_init_dev_replace_tgtdev() btrfs_init_new_device() add_missing_dev() Normally most of the device allocations are already done at device_list_add(), except for the special three functions above, so can we simplify the fix to [1] as there isn't much toll if the above three were also under GFP_NOFS. [1] -------------------------------------------- diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index b2c1d26f577e..3c350ab6535b 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -387,7 +387,7 @@ static struct btrfs_device *__alloc_device(void) { struct btrfs_device *dev; - dev = kzalloc(sizeof(*dev), GFP_KERNEL); + dev = kzalloc(sizeof(*dev), GFP_NOFS); if (!dev) return ERR_PTR(-ENOMEM); @@ -395,7 +395,7 @@ static struct btrfs_device *__alloc_device(void) * Preallocate a bio that's always going to be used for flushing device * barriers and matches the device lifespan */ - dev->flush_bio = bio_alloc_bioset(GFP_KERNEL, 0, NULL); + dev->flush_bio = bio_alloc_bioset(GFP_NOFS, 0, NULL); if (!dev->flush_bio) { kfree(dev); return ERR_PTR(-ENOMEM); -------------------------------------------- Thanks. > fs/btrfs/volumes.c | 33 ++++++++++++++++++++++++++++++--- > 1 file changed, 30 insertions(+), 3 deletions(-) > > diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c > index 2576b1a379c9..663566baae78 100644 > --- a/fs/btrfs/volumes.c > +++ b/fs/btrfs/volumes.c > @@ -14,6 +14,7 @@ > #include > #include > #include > +#include > #include "ctree.h" > #include "extent_map.h" > #include "disk-io.h" > @@ -988,20 +989,29 @@ static noinline struct btrfs_device *device_list_add(const char *path, > } > > if (!device) { > + unsigned int nofs_flag; > + > if (fs_devices->opened) { > mutex_unlock(&fs_devices->device_list_mutex); > return ERR_PTR(-EBUSY); > } > > + /* > + * Setup nofs context because we are holding the device list > + * mutex, which is required for a transaction commit. > + */ > + nofs_flag = memalloc_nofs_save(); > device = btrfs_alloc_device(NULL, &devid, > disk_super->dev_item.uuid); > if (IS_ERR(device)) { > + memalloc_nofs_restore(nofs_flag); > mutex_unlock(&fs_devices->device_list_mutex); > /* we can safely leave the fs_devices entry around */ > return device; > } > > - name = rcu_string_strdup(path, GFP_NOFS); > + name = rcu_string_strdup(path, GFP_KERNEL); > + memalloc_nofs_restore(nofs_flag); > if (!name) { > btrfs_free_device(device); > mutex_unlock(&fs_devices->device_list_mutex); > @@ -1137,11 +1147,19 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig) > /* We have held the volume lock, it is safe to get the devices. */ > list_for_each_entry(orig_dev, &orig->devices, dev_list) { > struct rcu_string *name; > + unsigned int nofs_flag; > > + /* > + * Setup nofs context because we are holding the device list > + * mutex, which is required for a transaction commit. > + */ > + nofs_flag = memalloc_nofs_save(); > device = btrfs_alloc_device(NULL, &orig_dev->devid, > orig_dev->uuid); > - if (IS_ERR(device)) > + if (IS_ERR(device)) { > + memalloc_nofs_restore(nofs_flag); > goto error; > + } > > /* > * This is ok to do without rcu read locked because we hold the > @@ -1151,12 +1169,14 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig) > name = rcu_string_strdup(orig_dev->name->str, > GFP_KERNEL); > if (!name) { > + memalloc_nofs_restore(nofs_flag); > btrfs_free_device(device); > goto error; > } > rcu_assign_pointer(device->name, name); > } > > + memalloc_nofs_restore(nofs_flag); > list_add(&device->dev_list, &fs_devices->devices); > device->fs_devices = fs_devices; > fs_devices->num_devices++; > @@ -1262,6 +1282,7 @@ static void btrfs_close_one_device(struct btrfs_device *device) > struct btrfs_fs_devices *fs_devices = device->fs_devices; > struct btrfs_device *new_device; > struct rcu_string *name; > + unsigned int nofs_flag; > > if (device->bdev) > fs_devices->open_devices--; > @@ -1277,17 +1298,23 @@ static void btrfs_close_one_device(struct btrfs_device *device) > > btrfs_close_bdev(device); > > + /* > + * Setup nofs context because we are holding the device list > + * mutex, which is required for a transaction commit. > + */ > + nofs_flag = memalloc_nofs_save(); > new_device = btrfs_alloc_device(NULL, &device->devid, > device->uuid); > BUG_ON(IS_ERR(new_device)); /* -ENOMEM */ > > /* Safe because we are under uuid_mutex */ > if (device->name) { > - name = rcu_string_strdup(device->name->str, GFP_NOFS); > + name = rcu_string_strdup(device->name->str, GFP_KERNEL); > BUG_ON(!name); /* -ENOMEM */ > rcu_assign_pointer(new_device->name, name); > } > > + memalloc_nofs_restore(nofs_flag); > list_replace_rcu(&device->dev_list, &new_device->dev_list); > new_device->fs_devices = device->fs_devices; > >