From cf6f54e3f133229f02a90c04fe0ff9dd9d3264b4 Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Wed, 8 Jul 2015 11:46:36 +0200 Subject: [PATCH] UBIFS: Kill unneeded locking in ubifs_init_security Fixes the following lockdep splat: [ 1.244527] ============================================= [ 1.245193] [ INFO: possible recursive locking detected ] [ 1.245193] 4.2.0-rc1+ #37 Not tainted [ 1.245193] --------------------------------------------- [ 1.245193] cp/742 is trying to acquire lock: [ 1.245193] (&sb->s_type->i_mutex_key#9){+.+.+.}, at: [] ubifs_init_security+0x29/0xb0 [ 1.245193] [ 1.245193] but task is already holding lock: [ 1.245193] (&sb->s_type->i_mutex_key#9){+.+.+.}, at: [] path_openat+0x3af/0x1280 [ 1.245193] [ 1.245193] other info that might help us debug this: [ 1.245193] Possible unsafe locking scenario: [ 1.245193] [ 1.245193] CPU0 [ 1.245193] ---- [ 1.245193] lock(&sb->s_type->i_mutex_key#9); [ 1.245193] lock(&sb->s_type->i_mutex_key#9); [ 1.245193] [ 1.245193] *** DEADLOCK *** [ 1.245193] [ 1.245193] May be due to missing lock nesting notation [ 1.245193] [ 1.245193] 2 locks held by cp/742: [ 1.245193] #0: (sb_writers#5){.+.+.+}, at: [] mnt_want_write+0x1f/0x50 [ 1.245193] #1: (&sb->s_type->i_mutex_key#9){+.+.+.}, at: [] path_openat+0x3af/0x1280 [ 1.245193] [ 1.245193] stack backtrace: [ 1.245193] CPU: 2 PID: 742 Comm: cp Not tainted 4.2.0-rc1+ #37 [ 1.245193] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.7.5-0-ge51488c-20140816_022509-build35 04/01/2014 [ 1.245193] ffffffff8252d530 ffff88007b023a38 ffffffff814f6f49 ffffffff810b56c5 [ 1.245193] ffff88007c30cc80 ffff88007b023af8 ffffffff810a150d ffff88007b023a68 [ 1.245193] 000000008101302a ffff880000000000 00000008f447e23f ffffffff8252d500 [ 1.245193] Call Trace: [ 1.245193] [] dump_stack+0x4c/0x65 [ 1.245193] [] ? console_unlock+0x1c5/0x510 [ 1.245193] [] __lock_acquire+0x1a6d/0x1ea0 [ 1.245193] [] ? __lock_is_held+0x58/0x80 [ 1.245193] [] lock_acquire+0xd3/0x270 [ 1.245193] [] ? ubifs_init_security+0x29/0xb0 [ 1.245193] [] mutex_lock_nested+0x6b/0x3a0 [ 1.245193] [] ? ubifs_init_security+0x29/0xb0 [ 1.245193] [] ? ubifs_init_security+0x29/0xb0 [ 1.245193] [] ubifs_init_security+0x29/0xb0 [ 1.245193] [] ubifs_create+0xa6/0x1f0 [ 1.245193] [] ? path_openat+0x3af/0x1280 [ 1.245193] [] vfs_create+0x95/0xc0 [ 1.245193] [] path_openat+0x7cc/0x1280 [ 1.245193] [] ? __lock_acquire+0x543/0x1ea0 [ 1.245193] [] ? sched_clock_cpu+0x90/0xc0 [ 1.245193] [] ? calc_global_load_tick+0x60/0x90 [ 1.245193] [] ? sched_clock_cpu+0x90/0xc0 [ 1.245193] [] ? __alloc_fd+0xaf/0x180 [ 1.245193] [] do_filp_open+0x75/0xd0 [ 1.245193] [] ? _raw_spin_unlock+0x26/0x40 [ 1.245193] [] ? __alloc_fd+0xaf/0x180 [ 1.245193] [] do_sys_open+0x129/0x200 [ 1.245193] [] SyS_open+0x19/0x20 [ 1.245193] [] entry_SYSCALL_64_fastpath+0x12/0x6f While the lockdep splat is a false positive, becuase path_openat holds i_mutex of the parent directory and ubifs_init_security() tries to acquire i_mutex of a new inode, it reveals that taking i_mutex in ubifs_init_security() is in vain because it is only being called in the inode allocation path and therefore nobody else can see the inode yet. Cc: stable@vger.kernel.org # 3.20- Reported-and-tested-by: Boris Brezillon Reviewed-and-tested-by: Dongsheng Yang Signed-off-by: Richard Weinberger Signed-off-by: dedekind1@gmail.com --- fs/ubifs/xattr.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/fs/ubifs/xattr.c b/fs/ubifs/xattr.c index 96f3448b6eb40c..fd65b3f1923ccd 100644 --- a/fs/ubifs/xattr.c +++ b/fs/ubifs/xattr.c @@ -652,11 +652,8 @@ int ubifs_init_security(struct inode *dentry, struct inode *inode, { int err; - mutex_lock(&inode->i_mutex); err = security_inode_init_security(inode, dentry, qstr, &init_xattrs, 0); - mutex_unlock(&inode->i_mutex); - if (err) { struct ubifs_info *c = dentry->i_sb->s_fs_info; ubifs_err(c, "cannot initialize security for inode %lu, error %d",