From e60efa84252c059bde5f65fccc6af94478d39e3b Mon Sep 17 00:00:00 2001 From: Wang Shilong Date: Wed, 23 Apr 2014 19:33:34 +0800 Subject: Btrfs: avoid triggering bug_on() when we fail to start inode caching task When running stress test(including snapshots,balance,fstress), we trigger the following BUG_ON() which is because we fail to start inode caching task. [ 181.131945] kernel BUG at fs/btrfs/inode-map.c:179! [ 181.137963] invalid opcode: 0000 [#1] SMP [ 181.217096] CPU: 11 PID: 2532 Comm: btrfs Not tainted 3.14.0 #1 [ 181.240521] task: ffff88013b621b30 ti: ffff8800b6ada000 task.ti: ffff8800b6ada000 [ 181.367506] Call Trace: [ 181.371107] [] btrfs_return_ino+0x9e/0x110 [btrfs] [ 181.379191] [] btrfs_evict_inode+0x46b/0x4c0 [btrfs] [ 181.387464] [] ? autoremove_wake_function+0x40/0x40 [ 181.395642] [] evict+0x9e/0x190 [ 181.401882] [] iput+0xf3/0x180 [ 181.408025] [] btrfs_orphan_cleanup+0x1ee/0x430 [btrfs] [ 181.416614] [] btrfs_mksubvol.isra.29+0x3bd/0x450 [btrfs] [ 181.425399] [] btrfs_ioctl_snap_create_transid+0x186/0x190 [btrfs] [ 181.435059] [] btrfs_ioctl_snap_create_v2+0xeb/0x130 [btrfs] [ 181.444148] [] btrfs_ioctl+0xf76/0x2b90 [btrfs] [ 181.451971] [] ? handle_mm_fault+0x475/0xe80 [ 181.459509] [] ? __do_page_fault+0x1ec/0x520 [ 181.467046] [] ? do_mmap_pgoff+0x2f5/0x3c0 [ 181.474393] [] do_vfs_ioctl+0x2d8/0x4b0 [ 181.481450] [] SyS_ioctl+0x81/0xa0 [ 181.488021] [] system_call_fastpath+0x16/0x1b We should avoid triggering BUG_ON() here, instead, we output warning messages and clear inode_cache option. Signed-off-by: Wang Shilong Signed-off-by: Chris Mason --- fs/btrfs/inode-map.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) (limited to 'fs/btrfs/inode-map.c') diff --git a/fs/btrfs/inode-map.c b/fs/btrfs/inode-map.c index cc8ca193d83..8ad529e3e67 100644 --- a/fs/btrfs/inode-map.c +++ b/fs/btrfs/inode-map.c @@ -176,7 +176,11 @@ static void start_caching(struct btrfs_root *root) tsk = kthread_run(caching_kthread, root, "btrfs-ino-cache-%llu\n", root->root_key.objectid); - BUG_ON(IS_ERR(tsk)); /* -ENOMEM */ + if (IS_ERR(tsk)) { + btrfs_warn(root->fs_info, "failed to start inode caching task"); + btrfs_clear_and_info(root, CHANGE_INODE_CACHE, + "disabling inode map caching"); + } } int btrfs_find_free_ino(struct btrfs_root *root, u64 *objectid) -- cgit v1.2.3-70-g09d2 From 1c70d8fb4dfa95bee491816b2a6767b5ca1080e7 Mon Sep 17 00:00:00 2001 From: Miao Xie Date: Wed, 23 Apr 2014 19:33:36 +0800 Subject: Btrfs: fix inode caching vs tree log Currently, with inode cache enabled, we will reuse its inode id immediately after unlinking file, we may hit something like following: |->iput inode |->return inode id into inode cache |->create dir,fsync |->power off An easy way to reproduce this problem is: mkfs.btrfs -f /dev/sdb mount /dev/sdb /mnt -o inode_cache,commit=100 dd if=/dev/zero of=/mnt/data bs=1M count=10 oflag=sync inode_id=`ls -i /mnt/data | awk '{print $1}'` rm -f /mnt/data i=1 while [ 1 ] do mkdir /mnt/dir_$i test1=`stat /mnt/dir_$i | grep Inode: | awk '{print $4}'` if [ $test1 -eq $inode_id ] then dd if=/dev/zero of=/mnt/dir_$i/data bs=1M count=1 oflag=sync echo b > /proc/sysrq-trigger fi sleep 1 i=$(($i+1)) done mount /dev/sdb /mnt umount /dev/sdb btrfs check /dev/sdb We fix this problem by adding unlinked inode's id into pinned tree, and we can not reuse them until committing transaction. Cc: stable@vger.kernel.org Signed-off-by: Miao Xie Signed-off-by: Wang Shilong Signed-off-by: Chris Mason --- fs/btrfs/inode-map.c | 18 ++---------------- 1 file changed, 2 insertions(+), 16 deletions(-) (limited to 'fs/btrfs/inode-map.c') diff --git a/fs/btrfs/inode-map.c b/fs/btrfs/inode-map.c index 8ad529e3e67..86935f5ae29 100644 --- a/fs/btrfs/inode-map.c +++ b/fs/btrfs/inode-map.c @@ -209,24 +209,14 @@ again: void btrfs_return_ino(struct btrfs_root *root, u64 objectid) { - struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; struct btrfs_free_space_ctl *pinned = root->free_ino_pinned; if (!btrfs_test_opt(root, INODE_MAP_CACHE)) return; - again: if (root->cached == BTRFS_CACHE_FINISHED) { - __btrfs_add_free_space(ctl, objectid, 1); + __btrfs_add_free_space(pinned, objectid, 1); } else { - /* - * If we are in the process of caching free ino chunks, - * to avoid adding the same inode number to the free_ino - * tree twice due to cross transaction, we'll leave it - * in the pinned tree until a transaction is committed - * or the caching work is done. - */ - down_write(&root->fs_info->commit_root_sem); spin_lock(&root->cache_lock); if (root->cached == BTRFS_CACHE_FINISHED) { @@ -238,11 +228,7 @@ again: start_caching(root); - if (objectid <= root->cache_progress || - objectid >= root->highest_objectid) - __btrfs_add_free_space(ctl, objectid, 1); - else - __btrfs_add_free_space(pinned, objectid, 1); + __btrfs_add_free_space(pinned, objectid, 1); up_write(&root->fs_info->commit_root_sem); } -- cgit v1.2.3-70-g09d2