From patchwork Wed Sep 3 13:35:38 2014 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Miao Xie X-Patchwork-Id: 4834641 Return-Path: X-Original-To: patchwork-linux-btrfs@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork2.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.19.201]) by patchwork2.web.kernel.org (Postfix) with ESMTP id 24C97C0338 for ; Wed, 3 Sep 2014 13:35:52 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 71C9420219 for ; Wed, 3 Sep 2014 13:35:50 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 892CD2020E for ; Wed, 3 Sep 2014 13:35:48 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932808AbaICNeh (ORCPT ); Wed, 3 Sep 2014 09:34:37 -0400 Received: from cn.fujitsu.com ([59.151.112.132]:13195 "EHLO heian.cn.fujitsu.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S932291AbaICNeN (ORCPT ); Wed, 3 Sep 2014 09:34:13 -0400 X-IronPort-AV: E=Sophos;i="5.04,457,1406563200"; d="scan'208";a="35461825" Received: from localhost (HELO edo.cn.fujitsu.com) ([10.167.33.5]) by heian.cn.fujitsu.com with ESMTP; 03 Sep 2014 21:31:16 +0800 Received: from G08CNEXCHPEKD01.g08.fujitsu.local (localhost.localdomain [127.0.0.1]) by edo.cn.fujitsu.com (8.14.3/8.13.1) with ESMTP id s83DY9qD006665 for ; Wed, 3 Sep 2014 21:34:09 +0800 Received: from miao.fnst.cn.fujitsu.com (10.167.226.169) by G08CNEXCHPEKD01.g08.fujitsu.local (10.167.33.89) with Microsoft SMTP Server (TLS) id 14.3.181.6; Wed, 3 Sep 2014 21:34:20 +0800 From: Miao Xie To: Subject: [PATCH 09/18] Btrfs: fix unprotected device's variants on 32bits machine Date: Wed, 3 Sep 2014 21:35:38 +0800 Message-ID: <1409751347-5296-9-git-send-email-miaox@cn.fujitsu.com> X-Mailer: git-send-email 1.9.3 In-Reply-To: <1409751347-5296-1-git-send-email-miaox@cn.fujitsu.com> References: <1409751347-5296-1-git-send-email-miaox@cn.fujitsu.com> MIME-Version: 1.0 X-Originating-IP: [10.167.226.169] Sender: linux-btrfs-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-btrfs@vger.kernel.org X-Spam-Status: No, score=-8.6 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=ham 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 ->total_bytes,->disk_total_bytes,->bytes_used is protected by chunk lock when we change them, but sometimes we read them without any lock, and we might get unexpected value. We fix this problem like inode's i_size. Signed-off-by: Miao Xie --- fs/btrfs/dev-replace.c | 15 +++++---- fs/btrfs/ioctl.c | 6 ++-- fs/btrfs/volumes.c | 48 +++++++++++++++++------------ fs/btrfs/volumes.h | 84 ++++++++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 124 insertions(+), 29 deletions(-) diff --git a/fs/btrfs/dev-replace.c b/fs/btrfs/dev-replace.c index 1be03d8..da7ac14 100644 --- a/fs/btrfs/dev-replace.c +++ b/fs/btrfs/dev-replace.c @@ -418,7 +418,7 @@ int btrfs_dev_replace_start(struct btrfs_root *root, /* the disk copy procedure reuses the scrub code */ ret = btrfs_scrub_dev(fs_info, src_device->devid, 0, - src_device->total_bytes, + btrfs_device_get_total_bytes(src_device), &dev_replace->scrub_progress, 0, 1); ret = btrfs_dev_replace_finishing(root->fs_info, ret); @@ -555,11 +555,12 @@ static int btrfs_dev_replace_finishing(struct btrfs_fs_info *fs_info, memcpy(uuid_tmp, tgt_device->uuid, sizeof(uuid_tmp)); memcpy(tgt_device->uuid, src_device->uuid, sizeof(tgt_device->uuid)); memcpy(src_device->uuid, uuid_tmp, sizeof(src_device->uuid)); - tgt_device->total_bytes = src_device->total_bytes; - tgt_device->disk_total_bytes = src_device->disk_total_bytes; + btrfs_device_set_total_bytes(tgt_device, src_device->total_bytes); + btrfs_device_set_disk_total_bytes(tgt_device, + src_device->disk_total_bytes); + btrfs_device_set_bytes_used(tgt_device, src_device->bytes_used); ASSERT(list_empty(&src_device->resized_list)); tgt_device->commit_total_bytes = src_device->commit_total_bytes; - tgt_device->bytes_used = src_device->bytes_used; tgt_device->commit_bytes_used = src_device->bytes_used; if (fs_info->sb->s_bdev == src_device->bdev) fs_info->sb->s_bdev = tgt_device->bdev; @@ -650,6 +651,7 @@ void btrfs_dev_replace_status(struct btrfs_fs_info *fs_info, struct btrfs_ioctl_dev_replace_args *args) { struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace; + struct btrfs_device *srcdev; btrfs_dev_replace_lock(dev_replace); /* even if !dev_replace_is_valid, the values are good enough for @@ -672,8 +674,9 @@ void btrfs_dev_replace_status(struct btrfs_fs_info *fs_info, break; case BTRFS_IOCTL_DEV_REPLACE_STATE_STARTED: case BTRFS_IOCTL_DEV_REPLACE_STATE_SUSPENDED: + srcdev = dev_replace->srcdev; args->status.progress_1000 = div64_u64(dev_replace->cursor_left, - div64_u64(dev_replace->srcdev->total_bytes, 1000)); + div64_u64(btrfs_device_get_total_bytes(srcdev), 1000)); break; } btrfs_dev_replace_unlock(dev_replace); @@ -832,7 +835,7 @@ static int btrfs_dev_replace_continue_on_mount(struct btrfs_fs_info *fs_info) ret = btrfs_scrub_dev(fs_info, dev_replace->srcdev->devid, dev_replace->committed_cursor_left, - dev_replace->srcdev->total_bytes, + btrfs_device_get_total_bytes(dev_replace->srcdev), &dev_replace->scrub_progress, 0, 1); ret = btrfs_dev_replace_finishing(fs_info, ret); WARN_ON(ret); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index c692c36..e78d9f9 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1550,7 +1550,7 @@ static noinline int btrfs_ioctl_resize(struct file *file, goto out_free; } - old_size = device->total_bytes; + old_size = btrfs_device_get_total_bytes(device); if (mod < 0) { if (new_size > old_size) { @@ -2732,8 +2732,8 @@ static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg) } di_args->devid = dev->devid; - di_args->bytes_used = dev->bytes_used; - di_args->total_bytes = dev->total_bytes; + di_args->bytes_used = btrfs_device_get_bytes_used(dev); + di_args->total_bytes = btrfs_device_get_total_bytes(dev); memcpy(di_args->uuid, dev->uuid, sizeof(di_args->uuid)); if (dev->name) { struct rcu_string *name; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index d8e4a3d..41da102 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -1308,7 +1308,7 @@ again: if (device->bytes_used > 0) { u64 len = btrfs_dev_extent_length(leaf, extent); - device->bytes_used -= len; + btrfs_device_set_bytes_used(device, device->bytes_used - len); spin_lock(&root->fs_info->free_chunk_lock); root->fs_info->free_chunk_space += len; spin_unlock(&root->fs_info->free_chunk_lock); @@ -1462,8 +1462,10 @@ static int btrfs_add_device(struct btrfs_trans_handle *trans, btrfs_set_device_io_align(leaf, dev_item, device->io_align); btrfs_set_device_io_width(leaf, dev_item, device->io_width); btrfs_set_device_sector_size(leaf, dev_item, device->sector_size); - btrfs_set_device_total_bytes(leaf, dev_item, device->disk_total_bytes); - btrfs_set_device_bytes_used(leaf, dev_item, device->bytes_used); + btrfs_set_device_total_bytes(leaf, dev_item, + btrfs_device_get_disk_total_bytes(device)); + btrfs_set_device_bytes_used(leaf, dev_item, + btrfs_device_get_bytes_used(device)); btrfs_set_device_group(leaf, dev_item, 0); btrfs_set_device_seek_speed(leaf, dev_item, 0); btrfs_set_device_bandwidth(leaf, dev_item, 0); @@ -2330,7 +2332,8 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path, } - if (i_size_read(bdev->bd_inode) < srcdev->total_bytes) { + if (i_size_read(bdev->bd_inode) < + btrfs_device_get_total_bytes(srcdev)) { btrfs_err(fs_info, "target device is smaller than source device!"); ret = -EINVAL; goto error; @@ -2360,11 +2363,11 @@ int btrfs_init_dev_replace_tgtdev(struct btrfs_root *root, char *device_path, device->io_width = root->sectorsize; device->io_align = root->sectorsize; device->sector_size = root->sectorsize; - device->total_bytes = srcdev->total_bytes; - device->disk_total_bytes = srcdev->disk_total_bytes; + device->total_bytes = btrfs_device_get_total_bytes(srcdev); + device->disk_total_bytes = btrfs_device_get_disk_total_bytes(srcdev); + device->bytes_used = btrfs_device_get_bytes_used(srcdev); ASSERT(list_empty(&srcdev->resized_list)); device->commit_total_bytes = srcdev->commit_total_bytes; - device->bytes_used = srcdev->bytes_used; device->commit_bytes_used = device->bytes_used; device->dev_root = fs_info->dev_root; device->bdev = bdev; @@ -2435,8 +2438,10 @@ static noinline int btrfs_update_device(struct btrfs_trans_handle *trans, btrfs_set_device_io_align(leaf, dev_item, device->io_align); btrfs_set_device_io_width(leaf, dev_item, device->io_width); btrfs_set_device_sector_size(leaf, dev_item, device->sector_size); - btrfs_set_device_total_bytes(leaf, dev_item, device->disk_total_bytes); - btrfs_set_device_bytes_used(leaf, dev_item, device->bytes_used); + btrfs_set_device_total_bytes(leaf, dev_item, + btrfs_device_get_disk_total_bytes(device)); + btrfs_set_device_bytes_used(leaf, dev_item, + btrfs_device_get_bytes_used(device)); btrfs_mark_buffer_dirty(leaf); out: @@ -2464,8 +2469,8 @@ static int __btrfs_grow_device(struct btrfs_trans_handle *trans, btrfs_set_super_total_bytes(super_copy, old_total + diff); device->fs_devices->total_rw_bytes += diff; - device->total_bytes = new_size; - device->disk_total_bytes = new_size; + btrfs_device_set_total_bytes(device, new_size); + btrfs_device_set_disk_total_bytes(device, new_size); btrfs_clear_space_info_full(device->dev_root->fs_info); if (list_empty(&device->resized_list)) list_add_tail(&device->resized_list, @@ -3110,11 +3115,12 @@ static int __btrfs_balance(struct btrfs_fs_info *fs_info) /* step one make some room on all the devices */ devices = &fs_info->fs_devices->devices; list_for_each_entry(device, devices, dev_list) { - old_size = device->total_bytes; + old_size = btrfs_device_get_total_bytes(device); size_to_free = div_factor(old_size, 1); size_to_free = min(size_to_free, (u64)1 * 1024 * 1024); if (!device->writeable || - device->total_bytes - device->bytes_used > size_to_free || + btrfs_device_get_total_bytes(device) - + btrfs_device_get_bytes_used(device) > size_to_free || device->is_tgtdev_for_dev_replace) continue; @@ -3920,8 +3926,8 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size) struct btrfs_key key; struct btrfs_super_block *super_copy = root->fs_info->super_copy; u64 old_total = btrfs_super_total_bytes(super_copy); - u64 old_size = device->total_bytes; - u64 diff = device->total_bytes - new_size; + u64 old_size = btrfs_device_get_total_bytes(device); + u64 diff = old_size - new_size; if (device->is_tgtdev_for_dev_replace) return -EINVAL; @@ -3934,7 +3940,7 @@ int btrfs_shrink_device(struct btrfs_device *device, u64 new_size) lock_chunks(root); - device->total_bytes = new_size; + btrfs_device_set_total_bytes(device, new_size); if (device->writeable) { device->fs_devices->total_rw_bytes -= diff; spin_lock(&root->fs_info->free_chunk_lock); @@ -4000,7 +4006,7 @@ again: ret = -ENOSPC; lock_chunks(root); - device->total_bytes = old_size; + btrfs_device_set_total_bytes(device, old_size); if (device->writeable) device->fs_devices->total_rw_bytes += diff; spin_lock(&root->fs_info->free_chunk_lock); @@ -4018,7 +4024,7 @@ again: } lock_chunks(root); - device->disk_total_bytes = new_size; + btrfs_device_set_disk_total_bytes(device, new_size); if (list_empty(&device->resized_list)) list_add_tail(&device->resized_list, &root->fs_info->fs_devices->resized_devices); @@ -4429,8 +4435,10 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, if (ret) goto error_del_extent; - for (i = 0; i < map->num_stripes; i++) - map->stripes[i].dev->bytes_used += stripe_size; + for (i = 0; i < map->num_stripes; i++) { + num_bytes = map->stripes[i].dev->bytes_used + stripe_size; + btrfs_device_set_bytes_used(map->stripes[i].dev, num_bytes); + } spin_lock(&extent_root->fs_info->free_chunk_lock); extent_root->fs_info->free_chunk_space -= (stripe_size * diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index f79d532..76600a3 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -32,6 +32,19 @@ struct btrfs_pending_bios { struct bio *tail; }; +/* + * Use sequence counter to get consistent device stat data on + * 32-bit processors. + */ +#if BITS_PER_LONG==32 && defined(CONFIG_SMP) +#include +#define __BTRFS_NEED_DEVICE_DATA_ORDERED +#define btrfs_device_data_ordered_init(device) \ + seqcount_init(&device->data_seqcount) +#else +#define btrfs_device_data_ordered_init(device) do { } while (0) +#endif + struct btrfs_device { struct list_head dev_list; struct list_head dev_alloc_list; @@ -61,6 +74,10 @@ struct btrfs_device { int can_discard; int is_tgtdev_for_dev_replace; +#ifdef __BTRFS_NEED_DEVICE_DATA_ORDERED + seqcount_t data_seqcount; +#endif + /* the internal btrfs device id */ u64 devid; @@ -133,6 +150,73 @@ struct btrfs_device { atomic_t dev_stat_values[BTRFS_DEV_STAT_VALUES_MAX]; }; +/* + * If we read those variants at the context of their own lock, we needn't + * use the following helpers, reading them directly is safe. + */ +#if BITS_PER_LONG==32 && defined(CONFIG_SMP) +#define BTRFS_DEVICE_GETSET_FUNCS(name) \ +static inline u64 \ +btrfs_device_get_##name(const struct btrfs_device *dev) \ +{ \ + u64 size; \ + unsigned int seq; \ + \ + do { \ + seq = read_seqcount_begin(&dev->data_seqcount); \ + size = dev->name; \ + } while (read_seqcount_retry(&dev->data_seqcount, seq)); \ + return size; \ +} \ + \ +static inline void \ +btrfs_device_set_##name(struct btrfs_device *dev, u64 size) \ +{ \ + preempt_disable(); \ + write_seqcount_begin(&dev->data_seqcount); \ + dev->name = size; \ + write_seqcount_end(&dev->data_seqcount); \ + preempt_enable(); \ +} +#elif BITS_PER_LONG==32 && defined(CONFIG_PREEMPT) +#define BTRFS_DEVICE_GETSET_FUNCS(name) \ +static inline u64 \ +btrfs_device_get_##name(const struct btrfs_device *dev) \ +{ \ + u64 size; \ + \ + preempt_disable(); \ + size = dev->name; \ + preempt_enable(); \ + return size; \ +} \ + \ +static inline void \ +btrfs_device_set_##name(struct btrfs_device *dev, u64 size) \ +{ \ + preempt_disable(); \ + dev->name = size; \ + preempt_enable(); \ +} +#else +#define BTRFS_DEVICE_GETSET_FUNCS(name) \ +static inline u64 \ +btrfs_device_get_##name(const struct btrfs_device *dev) \ +{ \ + return dev->name; \ +} \ + \ +static inline void \ +btrfs_device_set_##name(struct btrfs_device *dev, u64 size) \ +{ \ + dev->name = size; \ +} +#endif + +BTRFS_DEVICE_GETSET_FUNCS(total_bytes); +BTRFS_DEVICE_GETSET_FUNCS(disk_total_bytes); +BTRFS_DEVICE_GETSET_FUNCS(bytes_used); + struct btrfs_fs_devices { u8 fsid[BTRFS_FSID_SIZE]; /* FS specific uuid */