From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: stable@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
patches@lists.linux.dev, syzbot <syzkaller@googlegroups.com>,
Eric Dumazet <edumazet@google.com>,
Bernard Pidoux <f6bvp@free.fr>,
"David S. Miller" <davem@davemloft.net>,
Sasha Levin <sashal@kernel.org>
Subject: [PATCH 5.10 13/75] net/rose: fix races in rose_kill_by_device()
Date: Wed, 3 Jan 2024 17:54:54 +0100 [thread overview]
Message-ID: <20240103164845.203898184@linuxfoundation.org> (raw)
In-Reply-To: <20240103164842.953224409@linuxfoundation.org>
5.10-stable review patch. If anyone has any objections, please let me know.
------------------
From: Eric Dumazet <edumazet@google.com>
[ Upstream commit 64b8bc7d5f1434c636a40bdcfcd42b278d1714be ]
syzbot found an interesting netdev refcounting issue in
net/rose/af_rose.c, thanks to CONFIG_NET_DEV_REFCNT_TRACKER=y [1]
Problem is that rose_kill_by_device() can change rose->device
while other threads do not expect the pointer to be changed.
We have to first collect sockets in a temporary array,
then perform the changes while holding the socket
lock and rose_list_lock spinlock (in this order)
Change rose_release() to also acquire rose_list_lock
before releasing the netdev refcount.
[1]
[ 1185.055088][ T7889] ref_tracker: reference already released.
[ 1185.061476][ T7889] ref_tracker: allocated in:
[ 1185.066081][ T7889] rose_bind+0x4ab/0xd10
[ 1185.070446][ T7889] __sys_bind+0x1ec/0x220
[ 1185.074818][ T7889] __x64_sys_bind+0x72/0xb0
[ 1185.079356][ T7889] do_syscall_64+0x40/0x110
[ 1185.083897][ T7889] entry_SYSCALL_64_after_hwframe+0x63/0x6b
[ 1185.089835][ T7889] ref_tracker: freed in:
[ 1185.094088][ T7889] rose_release+0x2f5/0x570
[ 1185.098629][ T7889] __sock_release+0xae/0x260
[ 1185.103262][ T7889] sock_close+0x1c/0x20
[ 1185.107453][ T7889] __fput+0x270/0xbb0
[ 1185.111467][ T7889] task_work_run+0x14d/0x240
[ 1185.116085][ T7889] get_signal+0x106f/0x2790
[ 1185.120622][ T7889] arch_do_signal_or_restart+0x90/0x7f0
[ 1185.126205][ T7889] exit_to_user_mode_prepare+0x121/0x240
[ 1185.131846][ T7889] syscall_exit_to_user_mode+0x1e/0x60
[ 1185.137293][ T7889] do_syscall_64+0x4d/0x110
[ 1185.141783][ T7889] entry_SYSCALL_64_after_hwframe+0x63/0x6b
[ 1185.148085][ T7889] ------------[ cut here ]------------
WARNING: CPU: 1 PID: 7889 at lib/ref_tracker.c:255 ref_tracker_free+0x61a/0x810 lib/ref_tracker.c:255
Modules linked in:
CPU: 1 PID: 7889 Comm: syz-executor.2 Not tainted 6.7.0-rc4-syzkaller-00162-g65c95f78917e #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 11/10/2023
RIP: 0010:ref_tracker_free+0x61a/0x810 lib/ref_tracker.c:255
Code: 00 44 8b 6b 18 31 ff 44 89 ee e8 21 62 f5 fc 45 85 ed 0f 85 a6 00 00 00 e8 a3 66 f5 fc 48 8b 34 24 48 89 ef e8 27 5f f1 05 90 <0f> 0b 90 bb ea ff ff ff e9 52 fd ff ff e8 84 66 f5 fc 4c 8d 6d 44
RSP: 0018:ffffc90004917850 EFLAGS: 00010202
RAX: 0000000000000201 RBX: ffff88802618f4c0 RCX: 0000000000000000
RDX: 0000000000000202 RSI: ffffffff8accb920 RDI: 0000000000000001
RBP: ffff8880269ea5b8 R08: 0000000000000001 R09: fffffbfff23e35f6
R10: ffffffff91f1afb7 R11: 0000000000000001 R12: 1ffff92000922f0c
R13: 0000000005a2039b R14: ffff88802618f4d8 R15: 00000000ffffffff
FS: 00007f0a720ef6c0(0000) GS:ffff8880b9900000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f43a819d988 CR3: 0000000076c64000 CR4: 00000000003506f0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
<TASK>
netdev_tracker_free include/linux/netdevice.h:4127 [inline]
netdev_put include/linux/netdevice.h:4144 [inline]
netdev_put include/linux/netdevice.h:4140 [inline]
rose_kill_by_device net/rose/af_rose.c:195 [inline]
rose_device_event+0x25d/0x330 net/rose/af_rose.c:218
notifier_call_chain+0xb6/0x3b0 kernel/notifier.c:93
call_netdevice_notifiers_info+0xbe/0x130 net/core/dev.c:1967
call_netdevice_notifiers_extack net/core/dev.c:2005 [inline]
call_netdevice_notifiers net/core/dev.c:2019 [inline]
__dev_notify_flags+0x1f5/0x2e0 net/core/dev.c:8646
dev_change_flags+0x122/0x170 net/core/dev.c:8682
dev_ifsioc+0x9ad/0x1090 net/core/dev_ioctl.c:529
dev_ioctl+0x224/0x1090 net/core/dev_ioctl.c:786
sock_do_ioctl+0x198/0x270 net/socket.c:1234
sock_ioctl+0x22e/0x6b0 net/socket.c:1339
vfs_ioctl fs/ioctl.c:51 [inline]
__do_sys_ioctl fs/ioctl.c:871 [inline]
__se_sys_ioctl fs/ioctl.c:857 [inline]
__x64_sys_ioctl+0x18f/0x210 fs/ioctl.c:857
do_syscall_x64 arch/x86/entry/common.c:52 [inline]
do_syscall_64+0x40/0x110 arch/x86/entry/common.c:83
entry_SYSCALL_64_after_hwframe+0x63/0x6b
RIP: 0033:0x7f0a7147cba9
Code: 28 00 00 00 75 05 48 83 c4 28 c3 e8 e1 20 00 00 90 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b0 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f0a720ef0c8 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
RAX: ffffffffffffffda RBX: 00007f0a7159bf80 RCX: 00007f0a7147cba9
RDX: 0000000020000040 RSI: 0000000000008914 RDI: 0000000000000004
RBP: 00007f0a714c847a R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 000000000000000b R14: 00007f0a7159bf80 R15: 00007ffc8bb3a5f8
</TASK>
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Bernard Pidoux <f6bvp@free.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
net/rose/af_rose.c | 39 ++++++++++++++++++++++++++++++++++-----
1 file changed, 34 insertions(+), 5 deletions(-)
diff --git a/net/rose/af_rose.c b/net/rose/af_rose.c
index b3e7a92f1ec19..1d95ff34b13c9 100644
--- a/net/rose/af_rose.c
+++ b/net/rose/af_rose.c
@@ -181,21 +181,47 @@ void rose_kill_by_neigh(struct rose_neigh *neigh)
*/
static void rose_kill_by_device(struct net_device *dev)
{
- struct sock *s;
+ struct sock *sk, *array[16];
+ struct rose_sock *rose;
+ bool rescan;
+ int i, cnt;
+start:
+ rescan = false;
+ cnt = 0;
spin_lock_bh(&rose_list_lock);
- sk_for_each(s, &rose_list) {
- struct rose_sock *rose = rose_sk(s);
+ sk_for_each(sk, &rose_list) {
+ rose = rose_sk(sk);
+ if (rose->device == dev) {
+ if (cnt == ARRAY_SIZE(array)) {
+ rescan = true;
+ break;
+ }
+ sock_hold(sk);
+ array[cnt++] = sk;
+ }
+ }
+ spin_unlock_bh(&rose_list_lock);
+ for (i = 0; i < cnt; i++) {
+ sk = array[cnt];
+ rose = rose_sk(sk);
+ lock_sock(sk);
+ spin_lock_bh(&rose_list_lock);
if (rose->device == dev) {
- rose_disconnect(s, ENETUNREACH, ROSE_OUT_OF_ORDER, 0);
+ rose_disconnect(sk, ENETUNREACH, ROSE_OUT_OF_ORDER, 0);
if (rose->neighbour)
rose->neighbour->use--;
dev_put(rose->device);
rose->device = NULL;
}
+ spin_unlock_bh(&rose_list_lock);
+ release_sock(sk);
+ sock_put(sk);
+ cond_resched();
}
- spin_unlock_bh(&rose_list_lock);
+ if (rescan)
+ goto start;
}
/*
@@ -655,7 +681,10 @@ static int rose_release(struct socket *sock)
break;
}
+ spin_lock_bh(&rose_list_lock);
dev_put(rose->device);
+ rose->device = NULL;
+ spin_unlock_bh(&rose_list_lock);
sock->sk = NULL;
release_sock(sk);
sock_put(sk);
--
2.43.0
next prev parent reply other threads:[~2024-01-03 17:08 UTC|newest]
Thread overview: 86+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-01-03 16:54 [PATCH 5.10 00/75] 5.10.206-rc1 review Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 01/75] ksmbd: fix wrong name of SMB2_CREATE_ALLOCATION_SIZE Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 02/75] smb: client: fix OOB in smb2_query_reparse_point() Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 03/75] ARM: OMAP2+: Fix null pointer dereference and memory leak in omap_soc_device_init Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 04/75] reset: Fix crash when freeing non-existent optional resets Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 05/75] s390/vx: fix save/restore of fpu kernel context Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 06/75] wifi: mac80211: mesh_plink: fix matches_local logic Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 07/75] Revert "net/mlx5e: fix double free of encap_header" Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 08/75] net/mlx5e: Fix slab-out-of-bounds in mlx5_query_nic_vport_mac_list() Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 09/75] net/mlx5: Fix fw tracer first block check Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 10/75] net/mlx5e: Correct snprintf truncation handling for fw_version buffer used by representors Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 11/75] net: sched: ife: fix potential use-after-free Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 12/75] ethernet: atheros: fix a memleak in atl1e_setup_ring_resources Greg Kroah-Hartman
2024-01-03 16:54 ` Greg Kroah-Hartman [this message]
2024-01-03 16:54 ` [PATCH 5.10 14/75] net: check vlan filter feature in vlan_vids_add_by_dev() and vlan_vids_del_by_dev() Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 15/75] afs: Fix the dynamic roots d_delete to always delete unused dentries Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 16/75] afs: Fix dynamic root lookup DNS check Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 17/75] net: warn if gso_type isnt set for a GSO SKB Greg Kroah-Hartman
2024-01-03 16:54 ` [PATCH 5.10 18/75] net: check dev->gso_max_size in gso_features_check() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 19/75] keys, dns: Allow key types (eg. DNS) to be reclaimed immediately on expiry Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 20/75] afs: Fix overwriting of result of DNS query Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 21/75] i2c: aspeed: Handle the coalesced stop conditions with the start conditions Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 22/75] pinctrl: at91-pio4: use dedicated lock class for IRQ Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 23/75] ALSA: hda/hdmi: Add quirk to force pin connectivity on NUC10 Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 24/75] ALSA: hda/hdmi: add force-connect quirk for NUC5CPYB Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 25/75] smb: client: fix NULL deref in asn1_ber_decoder() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 26/75] btrfs: do not allow non subvolume root targets for snapshot Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 27/75] interconnect: Treat xlate() returning NULL node as an error Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 28/75] iio: imu: inv_mpu6050: fix an error code problem in inv_mpu6050_read_raw Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 29/75] interconnect: qcom: sm8250: Enable sync_state Greg Kroah-Hartman
2024-01-03 17:29 ` Konrad Dybcio
2024-01-03 16:55 ` [PATCH 5.10 30/75] Input: ipaq-micro-keys - add error handling for devm_kmemdup Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 31/75] scsi: bnx2fc: Fix skb double free in bnx2fc_rcv() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 32/75] iio: common: ms_sensors: ms_sensors_i2c: fix humidity conversion time table Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 33/75] iio: adc: ti_am335x_adc: Fix return value check of tiadc_request_dma() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 34/75] wifi: cfg80211: Add my certificate Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 35/75] wifi: cfg80211: fix certs build to not depend on file order Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 36/75] USB: serial: ftdi_sio: update Actisense PIDs constant names Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 37/75] USB: serial: option: add Quectel EG912Y module support Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 38/75] USB: serial: option: add Foxconn T99W265 with new baseline Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 39/75] USB: serial: option: add Quectel RM500Q R13 firmware support Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 40/75] Bluetooth: hci_event: Fix not checking if HCI_OP_INQUIRY has been sent Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 41/75] Bluetooth: L2CAP: Send reject on command corrupted request Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 42/75] Input: soc_button_array - add mapping for airplane mode button Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 43/75] net: 9p: avoid freeing uninit memory in p9pdu_vreadf Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 44/75] net: rfkill: gpio: set GPIO direction Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 45/75] net: ks8851: Fix TX stall caused by TX buffer overrun Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 46/75] dt-bindings: nvmem: mxs-ocotp: Document fsl,ocotp Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 47/75] tracing / synthetic: Disable events after testing in synth_event_gen_test_init() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 48/75] bus: ti-sysc: Flush posted write only after srst_udelay Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 49/75] lib/vsprintf: Fix %pfwf when current node refcount == 0 Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 50/75] x86/alternatives: Sync core before enabling interrupts Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 51/75] 9p/net: fix possible memory leak in p9_check_errors() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 52/75] ARM: dts: Fix occasional boot hang for am3 usb Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 53/75] Bluetooth: SMP: Convert BT_ERR/BT_DBG to bt_dev_err/bt_dev_dbg Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 54/75] Bluetooth: use inclusive language in SMP Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 55/75] Bluetooth: MGMT/SMP: Fix address type when using SMP over BREDR/LE Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 56/75] usb: fotg210-hcd: delete an incorrect bounds test Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 57/75] smb: client: fix OOB in SMB2_query_info_init() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 58/75] smb: client: fix OOB in smbCalcSize() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 59/75] Bluetooth: af_bluetooth: Fix Use-After-Free in bt_sock_recvmsg Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 60/75] spi: atmel: Switch to transfer_one transfer method Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 61/75] spi: atmel: Fix CS and initialization bug Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 62/75] scsi: core: Add scsi_prot_ref_tag() helper Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 63/75] scsi: core: Introduce scsi_get_sector() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 64/75] scsi: core: Make scsi_get_lba() return the LBA Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 65/75] scsi: core: Use scsi_cmd_to_rq() instead of scsi_cmnd.request Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 66/75] scsi: core: Use a structure member to track the SCSI command submitter Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 67/75] scsi: core: Always send batch on reset or error handling command Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 68/75] ring-buffer: Fix wake ups when buffer_percent is set to 100 Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 69/75] tracing: Fix blocked reader of snapshot buffer Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 70/75] netfilter: nf_tables: skip set commit for deleted/destroyed sets Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 71/75] dm-integrity: dont modify bios immutable bio_vec in integrity_metadata() Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 72/75] tracing/kprobes: Return EADDRNOTAVAIL when func matches several symbols Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 73/75] Revert "MIPS: Loongson64: Enable DMA noncoherent support" Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 74/75] Bluetooth: SMP: Fix crash when receiving new connection when debug is enabled Greg Kroah-Hartman
2024-01-03 16:55 ` [PATCH 5.10 75/75] spi: atmel: Fix PDC transfer setup bug Greg Kroah-Hartman
2024-01-03 19:15 ` [PATCH 5.10 00/75] 5.10.206-rc1 review Florian Fainelli
2024-01-04 0:40 ` Dominique Martinet
2024-01-04 9:16 ` Daniel Díaz
2024-01-04 9:25 ` Greg Kroah-Hartman
2024-01-08 17:47 ` Francis Laniel
2024-01-04 12:11 ` Jon Hunter
2024-01-04 12:15 ` Pavel Machek
2024-01-05 1:03 ` Guenter Roeck
2024-01-05 11:12 ` Shreeya Patel
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20240103164845.203898184@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=f6bvp@free.fr \
--cc=patches@lists.linux.dev \
--cc=sashal@kernel.org \
--cc=stable@vger.kernel.org \
--cc=syzkaller@googlegroups.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox