From 18513091af9483ba84328d42092bd4d42a3c958f Mon Sep 17 00:00:00 2001 From: Wang Xiaoguang Date: Mon, 25 Jul 2016 15:51:40 +0800 Subject: btrfs: update btrfs_space_info's bytes_may_use timely This patch can fix some false ENOSPC errors, below test script can reproduce one false ENOSPC error: #!/bin/bash dd if=/dev/zero of=fs.img bs=$((1024*1024)) count=128 dev=$(losetup --show -f fs.img) mkfs.btrfs -f -M $dev mkdir /tmp/mntpoint mount $dev /tmp/mntpoint cd /tmp/mntpoint xfs_io -f -c "falloc 0 $((64*1024*1024))" testfile Above script will fail for ENOSPC reason, but indeed fs still has free space to satisfy this request. Please see call graph: btrfs_fallocate() |-> btrfs_alloc_data_chunk_ondemand() | bytes_may_use += 64M |-> btrfs_prealloc_file_range() |-> btrfs_reserve_extent() |-> btrfs_add_reserved_bytes() | alloc_type is RESERVE_ALLOC_NO_ACCOUNT, so it does not | change bytes_may_use, and bytes_reserved += 64M. Now | bytes_may_use + bytes_reserved == 128M, which is greater | than btrfs_space_info's total_bytes, false enospc occurs. | Note, the bytes_may_use decrease operation will be done in | end of btrfs_fallocate(), which is too late. Here is another simple case for buffered write: CPU 1 | CPU 2 | |-> cow_file_range() |-> __btrfs_buffered_write() |-> btrfs_reserve_extent() | | | | | | | | | ..... | |-> btrfs_check_data_free_space() | | | | |-> extent_clear_unlock_delalloc() | In CPU 1, btrfs_reserve_extent()->find_free_extent()-> btrfs_add_reserved_bytes() do not decrease bytes_may_use, the decrease operation will be delayed to be done in extent_clear_unlock_delalloc(). Assume in this case, btrfs_reserve_extent() reserved 128MB data, CPU2's btrfs_check_data_free_space() tries to reserve 100MB data space. If 100MB > data_sinfo->total_bytes - data_sinfo->bytes_used - data_sinfo->bytes_reserved - data_sinfo->bytes_pinned - data_sinfo->bytes_readonly - data_sinfo->bytes_may_use btrfs_check_data_free_space() will try to allcate new data chunk or call btrfs_start_delalloc_roots(), or commit current transaction in order to reserve some free space, obviously a lot of work. But indeed it's not necessary as long as decreasing bytes_may_use timely, we still have free space, decreasing 128M from bytes_may_use. To fix this issue, this patch chooses to update bytes_may_use for both data and metadata in btrfs_add_reserved_bytes(). For compress path, real extent length may not be equal to file content length, so introduce a ram_bytes argument for btrfs_reserve_extent(), find_free_extent() and btrfs_add_reserved_bytes(), it's becasue bytes_may_use is increased by file content length. Then compress path can update bytes_may_use correctly. Also now we can discard RESERVE_ALLOC_NO_ACCOUNT, RESERVE_ALLOC and RESERVE_FREE. As we know, usually EXTENT_DO_ACCOUNTING is used for error path. In run_delalloc_nocow(), for inode marked as NODATACOW or extent marked as PREALLOC, we also need to update bytes_may_use, but can not pass EXTENT_DO_ACCOUNTING, because it also clears metadata reservation, so here we introduce EXTENT_CLEAR_DATA_RESV flag to indicate btrfs_clear_bit_hook() to update btrfs_space_info's bytes_may_use. Meanwhile __btrfs_prealloc_file_range() will call btrfs_free_reserved_data_space() internally for both sucessful and failed path, btrfs_prealloc_file_range()'s callers does not need to call btrfs_free_reserved_data_space() any more. Signed-off-by: Wang Xiaoguang Reviewed-by: Josef Bacik Signed-off-by: David Sterba Signed-off-by: Chris Mason --- fs/btrfs/file.c | 26 +++++++++++++++----------- 1 file changed, 15 insertions(+), 11 deletions(-) (limited to 'fs/btrfs/file.c') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 5842423f8f47..3391f2adf0c8 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -2675,6 +2675,7 @@ static long btrfs_fallocate(struct file *file, int mode, alloc_start = round_down(offset, blocksize); alloc_end = round_up(offset + len, blocksize); + cur_offset = alloc_start; /* Make sure we aren't being give some crap mode */ if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) @@ -2767,7 +2768,6 @@ static long btrfs_fallocate(struct file *file, int mode, /* First, check if we exceed the qgroup limit */ INIT_LIST_HEAD(&reserve_list); - cur_offset = alloc_start; while (1) { em = btrfs_get_extent(inode, NULL, 0, cur_offset, alloc_end - cur_offset, 0); @@ -2794,6 +2794,14 @@ static long btrfs_fallocate(struct file *file, int mode, last_byte - cur_offset); if (ret < 0) break; + } else { + /* + * Do not need to reserve unwritten extent for this + * range, free reserved data space first, otherwise + * it'll result in false ENOSPC error. + */ + btrfs_free_reserved_data_space(inode, cur_offset, + last_byte - cur_offset); } free_extent_map(em); cur_offset = last_byte; @@ -2811,6 +2819,9 @@ static long btrfs_fallocate(struct file *file, int mode, range->start, range->len, 1 << inode->i_blkbits, offset + len, &alloc_hint); + else + btrfs_free_reserved_data_space(inode, range->start, + range->len); list_del(&range->list); kfree(range); } @@ -2845,18 +2856,11 @@ out_unlock: unlock_extent_cached(&BTRFS_I(inode)->io_tree, alloc_start, locked_end, &cached_state, GFP_KERNEL); out: - /* - * As we waited the extent range, the data_rsv_map must be empty - * in the range, as written data range will be released from it. - * And for prealloacted extent, it will also be released when - * its metadata is written. - * So this is completely used as cleanup. - */ - btrfs_qgroup_free_data(inode, alloc_start, alloc_end - alloc_start); inode_unlock(inode); /* Let go of our reservation. */ - btrfs_free_reserved_data_space(inode, alloc_start, - alloc_end - alloc_start); + if (ret != 0) + btrfs_free_reserved_data_space(inode, alloc_start, + alloc_end - cur_offset); return ret; } -- cgit v1.2.3 From 28a235931b56d4e7bdd51f6733daf95f2b269da8 Mon Sep 17 00:00:00 2001 From: Filipe Manana Date: Tue, 23 Aug 2016 21:13:51 +0100 Subject: Btrfs: fix lockdep warning on deadlock against an inode's log mutex Commit 44f714dae50a ("Btrfs: improve performance on fsync against new inode after rename/unlink"), which landed in 4.8-rc2, introduced a possibility for a deadlock due to double locking of an inode's log mutex by the same task, which lockdep reports with: [23045.433975] ============================================= [23045.434748] [ INFO: possible recursive locking detected ] [23045.435426] 4.7.0-rc6-btrfs-next-34+ #1 Not tainted [23045.436044] --------------------------------------------- [23045.436044] xfs_io/3688 is trying to acquire lock: [23045.436044] (&ei->log_mutex){+.+...}, at: [] btrfs_log_inode+0x13a/0xc95 [btrfs] [23045.436044] but task is already holding lock: [23045.436044] (&ei->log_mutex){+.+...}, at: [] btrfs_log_inode+0x13a/0xc95 [btrfs] [23045.436044] other info that might help us debug this: [23045.436044] Possible unsafe locking scenario: [23045.436044] CPU0 [23045.436044] ---- [23045.436044] lock(&ei->log_mutex); [23045.436044] lock(&ei->log_mutex); [23045.436044] *** DEADLOCK *** [23045.436044] May be due to missing lock nesting notation [23045.436044] 3 locks held by xfs_io/3688: [23045.436044] #0: (&sb->s_type->i_mutex_key#15){+.+...}, at: [] btrfs_sync_file+0x14e/0x425 [btrfs] [23045.436044] #1: (sb_internal#2){.+.+.+}, at: [] __sb_start_write+0x5f/0xb0 [23045.436044] #2: (&ei->log_mutex){+.+...}, at: [] btrfs_log_inode+0x13a/0xc95 [btrfs] [23045.436044] stack backtrace: [23045.436044] CPU: 4 PID: 3688 Comm: xfs_io Not tainted 4.7.0-rc6-btrfs-next-34+ #1 [23045.436044] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.9.1-0-gb3ef39f-prebuilt.qemu-project.org 04/01/2014 [23045.436044] 0000000000000000 ffff88022f5f7860 ffffffff8127074d ffffffff82a54b70 [23045.436044] ffffffff82a54b70 ffff88022f5f7920 ffffffff81092897 ffff880228015d68 [23045.436044] 0000000000000000 ffffffff82a54b70 ffffffff829c3f00 ffff880228015d68 [23045.436044] Call Trace: [23045.436044] [] dump_stack+0x67/0x90 [23045.436044] [] __lock_acquire+0xcbb/0xe4e [23045.436044] [] ? mark_lock+0x24/0x201 [23045.436044] [] ? mark_held_locks+0x5e/0x74 [23045.436044] [] lock_acquire+0x12f/0x1c3 [23045.436044] [] ? lock_acquire+0x12f/0x1c3 [23045.436044] [] ? btrfs_log_inode+0x13a/0xc95 [btrfs] [23045.436044] [] ? btrfs_log_inode+0x13a/0xc95 [btrfs] [23045.436044] [] mutex_lock_nested+0x77/0x3a7 [23045.436044] [] ? btrfs_log_inode+0x13a/0xc95 [btrfs] [23045.436044] [] ? btrfs_release_delayed_node+0xb/0xd [btrfs] [23045.436044] [] btrfs_log_inode+0x13a/0xc95 [btrfs] [23045.436044] [] ? btrfs_log_inode+0x13a/0xc95 [btrfs] [23045.436044] [] ? vprintk_emit+0x453/0x465 [23045.436044] [] btrfs_log_inode+0x66e/0xc95 [btrfs] [23045.436044] [] log_new_dir_dentries+0x26c/0x359 [btrfs] [23045.436044] [] btrfs_log_inode_parent+0x4a6/0x628 [btrfs] [23045.436044] [] btrfs_log_dentry_safe+0x5a/0x75 [btrfs] [23045.436044] [] btrfs_sync_file+0x304/0x425 [btrfs] [23045.436044] [] vfs_fsync_range+0x8c/0x9e [23045.436044] [] vfs_fsync+0x1c/0x1e [23045.436044] [] do_fsync+0x31/0x4a [23045.436044] [] SyS_fsync+0x10/0x14 [23045.436044] [] entry_SYSCALL_64_fastpath+0x18/0xa8 [23045.436044] [] ? trace_hardirqs_off_caller+0x3f/0xaa An example reproducer for this is: $ mkfs.btrfs -f /dev/sdb $ mount /dev/sdb /mnt $ mkdir /mnt/dir $ touch /mnt/dir/foo $ sync $ mv /mnt/dir/foo /mnt/dir/bar $ touch /mnt/dir/foo $ xfs_io -c "fsync" /mnt/dir/bar This is because while logging the inode of file bar we end up logging its parent directory (since its inode has an unlink_trans field matching the current transaction id due to the rename operation), which in turn logs the inodes for all its new dentries, so that the new inode for the new file named foo gets logged which in turn triggered another logging attempt for the inode we are fsync'ing, since that inode had an old name that corresponds to the name of the new inode. So fix this by ensuring that when logging the inode for a new dentry that has a name matching an old name of some other inode, we don't log again the original inode that we are fsync'ing. Fixes: 44f714dae50a ("Btrfs: improve performance on fsync against new inode after rename/unlink") Signed-off-by: Filipe Manana Signed-off-by: David Sterba Signed-off-by: Chris Mason --- fs/btrfs/file.c | 2 +- fs/btrfs/tree-log.c | 5 +++-- fs/btrfs/tree-log.h | 5 ++++- 3 files changed, 8 insertions(+), 4 deletions(-) (limited to 'fs/btrfs/file.c') diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 3391f2adf0c8..fea31a4a6e36 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -2070,7 +2070,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync) } trans->sync = true; - btrfs_init_log_ctx(&ctx); + btrfs_init_log_ctx(&ctx, inode); ret = btrfs_log_dentry_safe(trans, root, dentry, start, end, &ctx); if (ret < 0) { diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index ffe92da81b8a..e935035ac034 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -2823,7 +2823,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, */ mutex_unlock(&root->log_mutex); - btrfs_init_log_ctx(&root_log_ctx); + btrfs_init_log_ctx(&root_log_ctx, NULL); mutex_lock(&log_root_tree->log_mutex); atomic_inc(&log_root_tree->log_batch); @@ -4757,7 +4757,8 @@ again: if (ret < 0) { err = ret; goto out_unlock; - } else if (ret > 0) { + } else if (ret > 0 && ctx && + other_ino != btrfs_ino(ctx->inode)) { struct btrfs_key inode_key; struct inode *other_inode; diff --git a/fs/btrfs/tree-log.h b/fs/btrfs/tree-log.h index a9f1b75d080d..ab858e31ccbc 100644 --- a/fs/btrfs/tree-log.h +++ b/fs/btrfs/tree-log.h @@ -30,15 +30,18 @@ struct btrfs_log_ctx { int log_transid; int io_err; bool log_new_dentries; + struct inode *inode; struct list_head list; }; -static inline void btrfs_init_log_ctx(struct btrfs_log_ctx *ctx) +static inline void btrfs_init_log_ctx(struct btrfs_log_ctx *ctx, + struct inode *inode) { ctx->log_ret = 0; ctx->log_transid = 0; ctx->io_err = 0; ctx->log_new_dentries = false; + ctx->inode = inode; INIT_LIST_HEAD(&ctx->list); } -- cgit v1.2.3