From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, Eric Dumazet <edumazet@google.com>,
syzbot <syzkaller@googlegroups.com>,
"David S. Miller" <davem@davemloft.net>
Subject: [PATCH 5.4 20/88] bonding: make tx_rebalance_counter an atomic
Date: Mon, 13 Dec 2021 10:29:50 +0100 [thread overview]
Message-ID: <20211213092933.910826596@linuxfoundation.org> (raw)
In-Reply-To: <20211213092933.250314515@linuxfoundation.org>
From: Eric Dumazet <edumazet@google.com>
commit dac8e00fb640e9569cdeefd3ce8a75639e5d0711 upstream.
KCSAN reported a data-race [1] around tx_rebalance_counter
which can be accessed from different contexts, without
the protection of a lock/mutex.
[1]
BUG: KCSAN: data-race in bond_alb_init_slave / bond_alb_monitor
write to 0xffff888157e8ca24 of 4 bytes by task 7075 on cpu 0:
bond_alb_init_slave+0x713/0x860 drivers/net/bonding/bond_alb.c:1613
bond_enslave+0xd94/0x3010 drivers/net/bonding/bond_main.c:1949
do_set_master net/core/rtnetlink.c:2521 [inline]
__rtnl_newlink net/core/rtnetlink.c:3475 [inline]
rtnl_newlink+0x1298/0x13b0 net/core/rtnetlink.c:3506
rtnetlink_rcv_msg+0x745/0x7e0 net/core/rtnetlink.c:5571
netlink_rcv_skb+0x14e/0x250 net/netlink/af_netlink.c:2491
rtnetlink_rcv+0x18/0x20 net/core/rtnetlink.c:5589
netlink_unicast_kernel net/netlink/af_netlink.c:1319 [inline]
netlink_unicast+0x5fc/0x6c0 net/netlink/af_netlink.c:1345
netlink_sendmsg+0x6e1/0x7d0 net/netlink/af_netlink.c:1916
sock_sendmsg_nosec net/socket.c:704 [inline]
sock_sendmsg net/socket.c:724 [inline]
____sys_sendmsg+0x39a/0x510 net/socket.c:2409
___sys_sendmsg net/socket.c:2463 [inline]
__sys_sendmsg+0x195/0x230 net/socket.c:2492
__do_sys_sendmsg net/socket.c:2501 [inline]
__se_sys_sendmsg net/socket.c:2499 [inline]
__x64_sys_sendmsg+0x42/0x50 net/socket.c:2499
do_syscall_x64 arch/x86/entry/common.c:50 [inline]
do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80
entry_SYSCALL_64_after_hwframe+0x44/0xae
read to 0xffff888157e8ca24 of 4 bytes by task 1082 on cpu 1:
bond_alb_monitor+0x8f/0xc00 drivers/net/bonding/bond_alb.c:1511
process_one_work+0x3fc/0x980 kernel/workqueue.c:2298
worker_thread+0x616/0xa70 kernel/workqueue.c:2445
kthread+0x2c7/0x2e0 kernel/kthread.c:327
ret_from_fork+0x1f/0x30
value changed: 0x00000001 -> 0x00000064
Reported by Kernel Concurrency Sanitizer on:
CPU: 1 PID: 1082 Comm: kworker/u4:3 Not tainted 5.16.0-rc3-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Workqueue: bond1 bond_alb_monitor
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
drivers/net/bonding/bond_alb.c | 14 ++++++++------
include/net/bond_alb.h | 2 +-
2 files changed, 9 insertions(+), 7 deletions(-)
--- a/drivers/net/bonding/bond_alb.c
+++ b/drivers/net/bonding/bond_alb.c
@@ -1514,14 +1514,14 @@ void bond_alb_monitor(struct work_struct
struct slave *slave;
if (!bond_has_slaves(bond)) {
- bond_info->tx_rebalance_counter = 0;
+ atomic_set(&bond_info->tx_rebalance_counter, 0);
bond_info->lp_counter = 0;
goto re_arm;
}
rcu_read_lock();
- bond_info->tx_rebalance_counter++;
+ atomic_inc(&bond_info->tx_rebalance_counter);
bond_info->lp_counter++;
/* send learning packets */
@@ -1543,7 +1543,7 @@ void bond_alb_monitor(struct work_struct
}
/* rebalance tx traffic */
- if (bond_info->tx_rebalance_counter >= BOND_TLB_REBALANCE_TICKS) {
+ if (atomic_read(&bond_info->tx_rebalance_counter) >= BOND_TLB_REBALANCE_TICKS) {
bond_for_each_slave_rcu(bond, slave, iter) {
tlb_clear_slave(bond, slave, 1);
if (slave == rcu_access_pointer(bond->curr_active_slave)) {
@@ -1553,7 +1553,7 @@ void bond_alb_monitor(struct work_struct
bond_info->unbalanced_load = 0;
}
}
- bond_info->tx_rebalance_counter = 0;
+ atomic_set(&bond_info->tx_rebalance_counter, 0);
}
if (bond_info->rlb_enabled) {
@@ -1623,7 +1623,8 @@ int bond_alb_init_slave(struct bonding *
tlb_init_slave(slave);
/* order a rebalance ASAP */
- bond->alb_info.tx_rebalance_counter = BOND_TLB_REBALANCE_TICKS;
+ atomic_set(&bond->alb_info.tx_rebalance_counter,
+ BOND_TLB_REBALANCE_TICKS);
if (bond->alb_info.rlb_enabled)
bond->alb_info.rlb_rebalance = 1;
@@ -1660,7 +1661,8 @@ void bond_alb_handle_link_change(struct
rlb_clear_slave(bond, slave);
} else if (link == BOND_LINK_UP) {
/* order a rebalance ASAP */
- bond_info->tx_rebalance_counter = BOND_TLB_REBALANCE_TICKS;
+ atomic_set(&bond_info->tx_rebalance_counter,
+ BOND_TLB_REBALANCE_TICKS);
if (bond->alb_info.rlb_enabled) {
bond->alb_info.rlb_rebalance = 1;
/* If the updelay module parameter is smaller than the
--- a/include/net/bond_alb.h
+++ b/include/net/bond_alb.h
@@ -126,7 +126,7 @@ struct tlb_slave_info {
struct alb_bond_info {
struct tlb_client_info *tx_hashtbl; /* Dynamically allocated */
u32 unbalanced_load;
- int tx_rebalance_counter;
+ atomic_t tx_rebalance_counter;
int lp_counter;
/* -------- rlb parameters -------- */
int rlb_enabled;
next prev parent reply other threads:[~2021-12-13 9:55 UTC|newest]
Thread overview: 95+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-12-13 9:29 [PATCH 5.4 00/88] 5.4.165-rc1 review Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 01/88] serial: tegra: Change lower tolerance baud rate limit for tegra20 and tegra30 Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 02/88] ntfs: fix ntfs_test_inode and ntfs_init_locked_inode function type Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 03/88] HID: quirks: Add quirk for the Microsoft Surface 3 type-cover Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 04/88] HID: google: add eel USB id Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 05/88] HID: add hid_is_usb() function to make it simpler for USB detection Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 06/88] HID: add USB_HID dependancy to hid-prodikeys Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 07/88] HID: add USB_HID dependancy to hid-chicony Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 08/88] HID: add USB_HID dependancy on some USB HID drivers Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 09/88] HID: bigbenff: prevent null pointer dereference Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 10/88] HID: wacom: fix problems when device is not a valid USB device Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 11/88] HID: check for valid USB device for many HID drivers Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 12/88] can: kvaser_usb: get CAN clock frequency from device Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 13/88] can: kvaser_pciefd: kvaser_pciefd_rx_error_frame(): increase correct stats->{rx,tx}_errors counter Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 14/88] can: sja1000: fix use after free in ems_pcmcia_add_card() Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 15/88] nfc: fix potential NULL pointer deref in nfc_genl_dump_ses_done Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 16/88] selftests: netfilter: add a vrf+conntrack testcase Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 17/88] vrf: dont run conntrack on vrf with !dflt qdisc Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 18/88] bpf: Fix the off-by-two error in range markings Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 19/88] ice: ignore dropped packets during init Greg Kroah-Hartman
2021-12-13 9:29 ` Greg Kroah-Hartman [this message]
2021-12-13 9:29 ` [PATCH 5.4 21/88] nfp: Fix memory leak in nfp_cpp_area_cache_add() Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 22/88] seg6: fix the iif in the IPv6 socket control block Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 23/88] udp: using datalen to cap max gso segments Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 24/88] iavf: restore MSI state on reset Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 25/88] iavf: Fix reporting when setting descriptor count Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 26/88] IB/hfi1: Correct guard on eager buffer deallocation Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 27/88] mm: bdi: initialize bdi_min_ratio when bdi is unregistered Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 28/88] ALSA: ctl: Fix copy of updated id with element read/write Greg Kroah-Hartman
2021-12-13 9:29 ` [PATCH 5.4 29/88] ALSA: hda/realtek - Add headset Mic support for Lenovo ALC897 platform Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 30/88] ALSA: pcm: oss: Fix negative period/buffer sizes Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 31/88] ALSA: pcm: oss: Limit the period size to 16MB Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 32/88] ALSA: pcm: oss: Handle missing errors in snd_pcm_oss_change_params*() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 33/88] btrfs: clear extent buffer uptodate when we fail to write it Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 34/88] btrfs: replace the BUG_ON in btrfs_del_root_ref with proper error handling Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 35/88] nfsd: Fix nsfd startup race (again) Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 36/88] tracefs: Have new files inherit the ownership of their parent Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 37/88] clk: qcom: regmap-mux: fix parent clock lookup Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 38/88] drm/syncobj: Deal with signalled fences in drm_syncobj_find_fence Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 39/88] can: pch_can: pch_can_rx_normal: fix use after free Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 40/88] can: m_can: Disable and ignore ELO interrupt Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 41/88] x86/sme: Explicitly map new EFI memmap table as encrypted Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 42/88] libata: add horkage for ASMedia 1092 Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 43/88] wait: add wake_up_pollfree() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 44/88] binder: use wake_up_pollfree() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 45/88] signalfd: " Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 46/88] aio: keep poll requests on waitqueue until completed Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 47/88] aio: fix use-after-free due to missing POLLFREE handling Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 48/88] tracefs: Set all files to the same group ownership as the mount option Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 49/88] block: fix ioprio_get(IOPRIO_WHO_PGRP) vs setuid(2) Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 50/88] qede: validate non LSO skb length Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 51/88] ASoC: qdsp6: q6routing: Fix return value from msm_routing_put_audio_mixer Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 52/88] i40e: Fix failed opcode appearing if handling messages from VF Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 53/88] i40e: Fix pre-set max number of queues for VF Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 54/88] mtd: rawnand: fsmc: Take instruction delay into account Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 55/88] mtd: rawnand: fsmc: Fix timing computation Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 56/88] dt-bindings: net: Reintroduce PHY no lane swap binding Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 57/88] tools build: Remove needless libpython-version feature check that breaks test-all fast path Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 58/88] net: cdc_ncm: Allow for dwNtbOutMaxSize to be unset or zero Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 59/88] net: altera: set a couple error code in probe() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 60/88] net: fec: only clear interrupt of handling queue in fec_enet_rx_queue() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 61/88] net, neigh: clear whole pneigh_entry at alloc time Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 62/88] net/qla3xxx: fix an error code in ql_adapter_up() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 63/88] selftests/fib_tests: Rework fib_rp_filter_test() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 64/88] USB: gadget: detect too-big endpoint 0 requests Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 65/88] USB: gadget: zero allocate endpoint 0 buffers Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 66/88] usb: core: config: fix validation of wMaxPacketValue entries Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 67/88] xhci: Remove CONFIG_USB_DEFAULT_PERSIST to prevent xHCI from runtime suspending Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 68/88] usb: core: config: using bit mask instead of individual bits Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 69/88] xhci: avoid race between disable slot command and host runtime suspend Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 70/88] iio: trigger: Fix reference counting Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 71/88] iio: trigger: stm32-timer: fix MODULE_ALIAS Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 72/88] iio: stk3310: Dont return error code in interrupt handler Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 73/88] iio: mma8452: Fix trigger reference couting Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 74/88] iio: ltr501: Dont return error code in trigger handler Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 75/88] iio: kxsd9: " Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 76/88] iio: itg3200: Call iio_trigger_notify_done() on error Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 77/88] iio: dln2-adc: Fix lockdep complaint Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 78/88] iio: dln2: Check return value of devm_iio_trigger_register() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 79/88] iio: at91-sama5d2: Fix incorrect sign extension Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 80/88] iio: adc: axp20x_adc: fix charging current reporting on AXP22x Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 81/88] iio: ad7768-1: Call iio_trigger_notify_done() on error Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 82/88] iio: accel: kxcjk-1013: Fix possible memory leak in probe and remove Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 83/88] irqchip/armada-370-xp: Fix return value of armada_370_xp_msi_alloc() Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 84/88] irqchip/armada-370-xp: Fix support for Multi-MSI interrupts Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 85/88] irqchip/irq-gic-v3-its.c: Force synchronisation when issuing INVALL Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 86/88] irqchip: nvic: Fix offset for Interrupt Priority Offsets Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 87/88] misc: fastrpc: fix improper packet size calculation Greg Kroah-Hartman
2021-12-13 9:30 ` [PATCH 5.4 88/88] bpf: Add selftests to cover packet access corner cases Greg Kroah-Hartman
2021-12-13 19:55 ` [PATCH 5.4 00/88] 5.4.165-rc1 review Guenter Roeck
2021-12-13 20:24 ` Florian Fainelli
2021-12-13 20:28 ` Shuah Khan
2021-12-14 12:29 ` Naresh Kamboju
2021-12-14 12:41 ` Sudip Mukherjee
2021-12-15 0:50 ` Samuel Zou
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=20211213092933.910826596@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=linux-kernel@vger.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