Re: [PATCH] btrfs: fix warning when creating a directory with smack enabled
From: Casey Schaufler <casey@schaufler-ca.com>
Date: 2021-02-26 20:08:05
On 2/26/2021 9:51 AM, fdmanana@kernel.org wrote:
From: Filipe Manana <redacted>
When we have smack enabled, during the creation of a directory smack may
attempt to add a "smack transmute" xattr on the inode, which results in
the following warning and trace:
[ 220.732359] ------------[ cut here ]------------
[ 220.732398] WARNING: CPU: 3 PID: 2548 at fs/btrfs/transaction.c:537 start_transaction+0x489/0x4f0
[ 220.732400] Modules linked in: nft_objref nf_conntrack_netbios_ns (...)
[ 220.732439] CPU: 3 PID: 2548 Comm: mkdir Not tainted 5.9.0-rc2smack+ #81
[ 220.732441] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014
[ 220.732444] RIP: 0010:start_transaction+0x489/0x4f0
[ 220.732447] Code: e9 be fc ff ff (...)
[ 220.732449] RSP: 0018:ffffc90001887d10 EFLAGS: 00010202
[ 220.732452] RAX: ffff88816f1e0000 RBX: 0000000000000201 RCX: 0000000000000003
[ 220.732454] RDX: 0000000000000201 RSI: 0000000000000002 RDI: ffff888177849000
[ 220.732456] RBP: ffff888177849000 R08: 0000000000000001 R09: 0000000000000004
[ 220.732458] R10: ffffffff825e8f7a R11: 0000000000000003 R12: ffffffffffffffe2
[ 220.732460] R13: 0000000000000000 R14: ffff88803d884270 R15: ffff8881680d8000
[ 220.732463] FS: 00007f67317b8440(0000) GS:ffff88817bcc0000(0000) knlGS:0000000000000000
[ 220.732465] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 220.732467] CR2: 00007f67247a22a8 CR3: 000000004bfbc002 CR4: 0000000000370ee0
[ 220.732472] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 220.732474] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 220.732475] Call Trace:
[ 220.732480] ? slab_free_freelist_hook+0xea/0x1b0
[ 220.732483] ? trace_hardirqs_on+0x1c/0xe0
[ 220.732490] btrfs_setxattr_trans+0x3c/0xf0
[ 220.732496] __vfs_setxattr+0x63/0x80
[ 220.732502] smack_d_instantiate+0x2d3/0x360
[ 220.732507] security_d_instantiate+0x29/0x40
[ 220.732511] d_instantiate_new+0x38/0x90
[ 220.732515] btrfs_mkdir+0x1cf/0x1e0
[ 220.732521] vfs_mkdir+0x14f/0x200
[ 220.732525] do_mkdirat+0x6d/0x110
[ 220.732531] do_syscall_64+0x2d/0x40
[ 220.732534] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 220.732537] RIP: 0033:0x7f673196ae6b
[ 220.732540] Code: 8b 05 11 (...)
[ 220.732542] RSP: 002b:00007ffc3c679b18 EFLAGS: 00000246 ORIG_RAX: 0000000000000053
[ 220.732545] RAX: ffffffffffffffda RBX: 00000000000001ff RCX: 00007f673196ae6b
[ 220.732547] RDX: 0000000000000000 RSI: 00000000000001ff RDI: 00007ffc3c67a30d
[ 220.732549] RBP: 00007ffc3c67a30d R08: 00000000000001ff R09: 0000000000000000
[ 220.732551] R10: 000055d3e39fe930 R11: 0000000000000246 R12: 0000000000000000
[ 220.732553] R13: 00007ffc3c679cd8 R14: 00007ffc3c67a30d R15: 00007ffc3c679ce0
[ 220.732563] irq event stamp: 11029
[ 220.732566] hardirqs last enabled at (11037): [<ffffffff81153fe6>] console_unlock+0x486/0x670
[ 220.732569] hardirqs last disabled at (11044): [<ffffffff81153c01>] console_unlock+0xa1/0x670
[ 220.732572] softirqs last enabled at (8864): [<ffffffff81e0102f>] asm_call_on_stack+0xf/0x20
[ 220.732575] softirqs last disabled at (8851): [<ffffffff81e0102f>] asm_call_on_stack+0xf/0x20
[ 220.732577] ---[ end trace 8f958916039daced ]---
This happens because at btrfs_mkdir() we call d_instantiate_new() while
holding a transaction handle, which results in the following call chain:
btrfs_mkdir()
trans = btrfs_start_transaction(root, 5);
d_instantiate_new()
smack_d_instantiate()
__vfs_setxattr()
btrfs_setxattr_trans()
btrfs_start_transaction()
start_transaction()
WARN_ON()
--> a tansaction start has TRANS_EXTWRITERS
set in its type
h->orig_rsv = h->block_rsv
h->block_rsv = NULL
btrfs_end_transaction(trans)
Besides the warning triggered at start_transaction.c, we set the handle's
block_rsv to NULL which may cause some surprises later on.
So fix this by making btrfs_setxattr_trans() not start a transaction when
we already have a handle on one, stored in current->journal_info, and use
that handle. We are good to use the handle because at btrfs_mkdir() we did
reserve space for the xattr and the inode item.
Reported-by: Casey Schaufler <casey@schaufler-ca.com>
Link: https://lore.kernel.org/linux-btrfs/434d856f-bd7b-4889-a6ec-e81aaebfa735@schaufler-ca.com/ (local)
Signed-off-by: Filipe Manana <redacted>This passes the Smack tests. Acked-by: Casey Schaufler <casey@schaufler-ca.com> Tested-by: Casey Schaufler <casey@schaufler-ca.com>
quoted hunk ↗ jump to hunk
--- fs/btrfs/xattr.c | 31 +++++++++++++++++++++++++++---- 1 file changed, 27 insertions(+), 4 deletions(-)diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c index af6246f36a9e..03135dbb318a 100644 --- a/fs/btrfs/xattr.c +++ b/fs/btrfs/xattr.c@@ -229,11 +229,33 @@ int btrfs_setxattr_trans(struct inode *inode, const char *name, { struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_trans_handle *trans; + const bool start_trans = (current->journal_info == NULL); int ret; - trans = btrfs_start_transaction(root, 2); - if (IS_ERR(trans)) - return PTR_ERR(trans); + if (start_trans) { + /* + * 1 unit for inserting/updating/deleting the xattr + * 1 unit for the inode item update + */ + trans = btrfs_start_transaction(root, 2); + if (IS_ERR(trans)) + return PTR_ERR(trans); + } else { + /* + * This can happen when smack is enabled and a directory is being + * created. It happens through d_instantiate_new(), which calls + * smack_d_instantiate(), which in turn calls __vfs_setxattr() to + * set the transmute xattr (XATTR_NAME_SMACKTRANSMUTE) on the + * inode. We have already reserved space for the xattr and inode + * update at btrfs_mkdir(), so just use the transaction handle. + * We don't join or start a transaction, as that will reset the + * block_rsv of the handle and trigger a warning for the start + * case. + */ + ASSERT(strncmp(name, XATTR_SECURITY_PREFIX, + XATTR_SECURITY_PREFIX_LEN) == 0); + trans = current->journal_info; + } ret = btrfs_setxattr(trans, inode, name, value, size, flags); if (ret)@@ -244,7 +266,8 @@ int btrfs_setxattr_trans(struct inode *inode, const char *name, ret = btrfs_update_inode(trans, root, BTRFS_I(inode)); BUG_ON(ret); out: - btrfs_end_transaction(trans); + if (start_trans) + btrfs_end_transaction(trans); return ret; }