* [PATCH v11 01/11] can: bcm: defer rx_op deallocation to workqueue to fix thrtimer UAF
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 18:10 ` sashiko-bot
2026-07-09 17:56 ` [PATCH v11 02/11] can: bcm: fix lockless bound/ifindex race and silent RX_SETUP failure Oliver Hartkopp via B4 Relay
` (9 subsequent siblings)
10 siblings, 1 reply; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde, Lee Jones, Feng Xue
From: Lee Jones <lee@kernel.org>
Commit f1b4e32aca08 ("can: bcm: use call_rcu() instead of costly
synchronize_rcu()") replaced synchronize_rcu() in bcm_delete_rx_op()
with call_rcu() and introduced the RX_NO_AUTOTIMER flag.
However, this flag check was omitted for thrtimer in the packet rx
fast-path. During BCM RX operation teardown, a concurrent RCU reader
(bcm_rx_handler) can race and re-arm thrtimer via
bcm_rx_update_and_send() after call_rcu() has been scheduled. Once
the RCU grace period elapses, bcm_op is freed. The subsequently
firing thrtimer then dereferences the deallocated op, causing a UAF.
Adding flag checks to the rx fast-path (bcm_rx_update_and_send) does not
fully close the TOCTOU race and introduces latency for every CAN frame.
Conversely, calling hrtimer_cancel() directly inside the RCU callback
(softirq context) is fatal as hrtimer_cancel() can sleep, triggering
a "scheduling while atomic" panic.
Resolve this by deferring the timer cancellation and memory free to a
dedicated unbound workqueue (bcm_wq). The RCU callback now queues a
work item to bcm_wq, which safely cancels both timers and deallocates
memory in sleepable process context. A dedicated workqueue is used to
prevent system-wide WQ saturation and is cleanly flushed/destroyed
on module unload to avoid rmmod page faults.
Since the deferred work can now outlive the calling context by an
unbounded amount, also take a reference on op->sk when it is assigned
and drop it only once the deferred work has cancelled both timers, so a
socket can no longer be freed out from under a still-armed timer whose
callback (bcm_send_to_user()) dereferences op->sk.
Fixes: f1b4e32aca08 ("can: bcm: use call_rcu() instead of costly synchronize_rcu()")
Tested-by: Feng Xue <feng.xue@outlook.com>
Tested-by: Oliver Hartkopp <socketcan@hartkopp.net>
Signed-off-by: Lee Jones <lee@kernel.org>
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 37 ++++++++++++++++++++++++++++++++++---
1 file changed, 34 insertions(+), 3 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index a4bef2c48a55..bdf53241bd7b 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -56,10 +56,11 @@
#include <linux/can.h>
#include <linux/can/core.h>
#include <linux/can/skb.h>
#include <linux/can/bcm.h>
#include <linux/slab.h>
+#include <linux/workqueue.h>
#include <linux/spinlock.h>
#include <net/can.h>
#include <net/sock.h>
#include <net/net_namespace.h>
@@ -90,10 +91,12 @@ MODULE_LICENSE("Dual BSD/GPL");
MODULE_AUTHOR("Oliver Hartkopp <oliver.hartkopp@volkswagen.de>");
MODULE_ALIAS("can-proto-2");
#define BCM_MIN_NAMELEN CAN_REQUIRED_SIZE(struct sockaddr_can, can_ifindex)
+static struct workqueue_struct *bcm_wq;
+
/*
* easy access to the first 64 bit of can(fd)_frame payload. cp->data is
* 64 bit aligned so the offset has to be multiples of 8 which is ensured
* by the only callers in bcm_rx_cmp_to_index() bcm_rx_handler().
*/
@@ -103,10 +106,11 @@ static inline u64 get_u64(const struct canfd_frame *cp, int offset)
}
struct bcm_op {
struct list_head list;
struct rcu_head rcu;
+ struct work_struct work;
int ifindex;
canid_t can_id;
u32 flags;
unsigned long frames_abs, frames_filtered;
struct bcm_timeval ival1, ival2;
@@ -791,23 +795,40 @@ static struct bcm_op *bcm_find_op(struct list_head *ops,
}
return NULL;
}
-static void bcm_free_op_rcu(struct rcu_head *rcu_head)
+static void bcm_free_op_work(struct work_struct *work)
{
- struct bcm_op *op = container_of(rcu_head, struct bcm_op, rcu);
+ struct bcm_op *op = container_of(work, struct bcm_op, work);
+
+ hrtimer_cancel(&op->timer);
+ hrtimer_cancel(&op->thrtimer);
if ((op->frames) && (op->frames != &op->sframe))
kfree(op->frames);
if ((op->last_frames) && (op->last_frames != &op->last_sframe))
kfree(op->last_frames);
+ /* the last possible access to op->timer/op->thrtimer has now
+ * happened above via hrtimer_cancel() - op->sk is no longer
+ * needed by any pending timer callback, so drop our reference
+ */
+ sock_put(op->sk);
+
kfree(op);
}
+static void bcm_free_op_rcu(struct rcu_head *rcu_head)
+{
+ struct bcm_op *op = container_of(rcu_head, struct bcm_op, rcu);
+
+ INIT_WORK(&op->work, bcm_free_op_work);
+ queue_work(bcm_wq, &op->work);
+}
+
static void bcm_remove_op(struct bcm_op *op)
{
hrtimer_cancel(&op->timer);
hrtimer_cancel(&op->thrtimer);
@@ -1058,10 +1079,11 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
/* tx_ops never compare with previous received messages */
op->last_frames = NULL;
/* bcm_can_tx / bcm_tx_timeout_handler needs this */
op->sk = sk;
+ sock_hold(sk);
op->ifindex = ifindex;
/* initialize uninitialized (kzalloc) structure */
hrtimer_setup(&op->timer, bcm_tx_timeout_handler, CLOCK_MONOTONIC,
HRTIMER_MODE_REL_SOFT);
@@ -1219,10 +1241,11 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
}
}
/* bcm_can_tx / bcm_tx_timeout_handler needs this */
op->sk = sk;
+ sock_hold(sk);
op->ifindex = ifindex;
/* ifindex for timeout events w/o previous frame reception */
op->rx_ifindex = ifindex;
@@ -1837,15 +1860,19 @@ static struct notifier_block canbcm_notifier = {
static int __init bcm_module_init(void)
{
int err;
+ bcm_wq = alloc_workqueue("can-bcm-wq", WQ_UNBOUND, 0);
+ if (!bcm_wq)
+ return -ENOMEM;
+
pr_info("can: broadcast manager protocol\n");
err = register_pernet_subsys(&canbcm_pernet_ops);
if (err)
- return err;
+ goto register_pernet_failed;
err = register_netdevice_notifier(&canbcm_notifier);
if (err)
goto register_notifier_failed;
@@ -1859,17 +1886,21 @@ static int __init bcm_module_init(void)
register_proto_failed:
unregister_netdevice_notifier(&canbcm_notifier);
register_notifier_failed:
unregister_pernet_subsys(&canbcm_pernet_ops);
+register_pernet_failed:
+ destroy_workqueue(bcm_wq);
return err;
}
static void __exit bcm_module_exit(void)
{
can_proto_unregister(&bcm_can_proto);
unregister_netdevice_notifier(&canbcm_notifier);
unregister_pernet_subsys(&canbcm_pernet_ops);
+ rcu_barrier();
+ destroy_workqueue(bcm_wq);
}
module_init(bcm_module_init);
module_exit(bcm_module_exit);
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH v11 01/11] can: bcm: defer rx_op deallocation to workqueue to fix thrtimer UAF
2026-07-09 17:56 ` [PATCH v11 01/11] can: bcm: defer rx_op deallocation to workqueue to fix thrtimer UAF Oliver Hartkopp via B4 Relay
@ 2026-07-09 18:10 ` sashiko-bot
0 siblings, 0 replies; 20+ messages in thread
From: sashiko-bot @ 2026-07-09 18:10 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: Marc Kleine-Budde, Vincent Mailhol, Oleksij Rempel,
Oliver Hartkopp, linux-can
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
Pre-existing issues:
- [High] Data race when updating CAN frames of an existing BCM RX operation
- [High] Missing RCU list primitives causes data races with lockless procfs readers
--
--- Patch [1]: [PATCH v11 01/11] can: bcm: defer rx_op deallocation to workqueue to fix thrtimer UAF ---
commit ae7bb166ff5dbe1bb7ad111fcf67c490882f4b51
Author: Lee Jones <lee@kernel.org>
can: bcm: defer rx_op deallocation to workqueue to fix thrtimer UAF
This commit fixes a thrtimer UAF during BCM RX operation teardown by
deferring timer cancellation and memory deallocation to a dedicated unbound
workqueue. It also retains a reference on op->sk until the deferred work
completes to ensure the socket is not freed prematurely.
> diff --git a/net/can/bcm.c b/net/can/bcm.c
> index a4bef2c48a559..bdf53241bd7b3 100644
> --- a/net/can/bcm.c
> +++ b/net/can/bcm.c
[ ... ]
> @@ -1221,6 +1243,7 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
>
> /* bcm_can_tx / bcm_tx_timeout_handler needs this */
> op->sk = sk;
> + sock_hold(sk);
> op->ifindex = ifindex;
>
> /* ifindex for timeout events w/o previous frame reception */
[Severity: High]
This is a pre-existing issue, but does bcm_rx_setup() have a data race when
updating CAN frames for an existing BCM RX operation?
It appears memcpy_from_msg() copies directly from user space into op->frames
without holding op->bcm_rx_update_lock or using a staging buffer:
net/can/bcm.c:bcm_rx_setup() {
...
err = memcpy_from_msg(op->frames, msg, msg_head->nframes * op->cfsiz);
...
memset(op->last_frames, 0, msg_head->nframes * op->cfsiz);
...
}
Since memcpy_from_msg() can sleep, it cannot hold the lock, but this means
concurrent softirq readers like bcm_rx_handler() might observe partially
updated filter masks:
net/can/bcm.c:bcm_rx_handler() {
...
if ((get_u64(op->frames, 0) & get_u64(rxframe, 0)) ==
...
}
Could this lead to valid CAN frames being incorrectly dropped or invalid
frames accepted?
[Severity: High]
This is a pre-existing issue, but does bcm_rx_setup() lack proper RCU list
primitives when inserting into bo->rx_ops?
It currently uses the non-RCU list_add() during setup:
net/can/bcm.c:bcm_rx_setup() {
...
list_add(&op->list, &bo->rx_ops);
...
}
This lacks the smp_store_release() barrier needed for concurrent lockless
readers. For instance, bcm_proc_show() iterates these lists locklessly
under rcu_read_lock(), but it uses list_for_each_entry() instead of
list_for_each_entry_rcu():
net/can/bcm.c:bcm_proc_show() {
...
list_for_each_entry(op, &bo->tx_ops, list) {
...
}
Can this lack of memory ordering and dependency barriers allow the compiler
to generate unsafe traversals, potentially leading to a NULL pointer
dereference or invalid list traversal for concurrent procfs readers of
/proc/net/can-bcm/*?
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-bcm_fixes-v11-0-067ba11bec5c@hartkopp.net?part=1
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH v11 02/11] can: bcm: fix lockless bound/ifindex race and silent RX_SETUP failure
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
2026-07-09 17:56 ` [PATCH v11 01/11] can: bcm: defer rx_op deallocation to workqueue to fix thrtimer UAF Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 18:09 ` sashiko-bot
2026-07-09 17:56 ` [PATCH v11 03/11] can: bcm: add locking when updating filter and timer values Oliver Hartkopp via B4 Relay
` (8 subsequent siblings)
10 siblings, 1 reply; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde, Ginger
From: Oliver Hartkopp <socketcan@hartkopp.net>
bcm_sendmsg() reads bo->ifindex and checks bo->bound before taking
lock_sock(), while bcm_notify(), bcm_connect() and bcm_release() all
mutate both fields under that same lock. Because the lockless reads
and the locked writes are unordered with respect to each other, a
racing bcm_notify() (device unregister) or bcm_connect() (concurrent
bind on another thread sharing the socket) can make bcm_sendmsg()
observe an inconsistent combination, e.g. a stale bound=1 together
with the now-cleared ifindex=0, silently turning a socket bound to a
specific CAN interface into one that also matches "any" interface.
Keep the lockless bo->bound check purely as a fast-path reject, and
move the ifindex read (and a bo->bound re-check) into the locked
section, where every writer already serializes. This removes the
possibility of observing the two fields torn against each other,
rather than trying to fix it with more READ_ONCE()/WRITE_ONCE() pairs
on two independently updated fields. Annotate the now-purely-lockless
bo->bound accesses consistently across all its write sites.
Also fix bcm_rx_setup() silently returning success when the target
device disappears concurrently instead of reporting -ENODEV, so a
broken RX op is no longer left registered as if it had succeeded.
Fixes: ffd980f976e7 ("[CAN]: Add broadcast manager (bcm) protocol")
Reported-by: Ginger <ginger.jzllee@gmail.com>
Closes: https://lore.kernel.org/linux-can/CAGp+u1aBK8QVjsvAxM2Ldzep4rEbsP9x_pV3At4g=h1kVEtyhA@mail.gmail.com/
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 65 ++++++++++++++++++++++++++++++++++++++++++++++-------------
1 file changed, 51 insertions(+), 14 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index bdf53241bd7b..b612135b017d 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -1321,10 +1321,15 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
bcm_rx_handler, op,
"bcm", sk);
op->rx_reg_dev = dev;
dev_put(dev);
+ } else {
+ /* the requested device is gone - do not
+ * silently succeed without registering
+ */
+ err = -ENODEV;
}
} else
err = can_rx_register(sock_net(sk), NULL, op->can_id,
REGMASK(op->can_id),
@@ -1394,16 +1399,17 @@ static int bcm_tx_send(struct msghdr *msg, int ifindex, struct sock *sk,
*/
static int bcm_sendmsg(struct socket *sock, struct msghdr *msg, size_t size)
{
struct sock *sk = sock->sk;
struct bcm_sock *bo = bcm_sk(sk);
- int ifindex = bo->ifindex; /* default ifindex for this bcm_op */
+ int ifindex;
struct bcm_msg_head msg_head;
int cfsiz;
int ret; /* read bytes or error codes as return value */
- if (!bo->bound)
+ /* Lockless fast-path check for bound socket */
+ if (!READ_ONCE(bo->bound))
return -ENOTCONN;
/* check for valid message length from userspace */
if (size < MHSIZ)
return -EINVAL;
@@ -1415,43 +1421,65 @@ static int bcm_sendmsg(struct socket *sock, struct msghdr *msg, size_t size)
cfsiz = CFSIZ(msg_head.flags);
if ((size - MHSIZ) % cfsiz)
return -EINVAL;
+ lock_sock(sk);
+
+ /* Re-validate under the socket lock: a concurrent bcm_notify()
+ * may have unbound this socket (device removal) after the
+ * lockless fast-path check above. bo->ifindex is only ever
+ * mutated under lock_sock(), so reading it here - instead of
+ * before taking the lock - guarantees it can't be observed
+ * torn against bo->bound.
+ */
+ if (!bo->bound) {
+ ret = -ENOTCONN;
+ goto out_release;
+ }
+
+ /* default ifindex for this bcm_op */
+ ifindex = bo->ifindex;
+
/* check for alternative ifindex for this bcm_op */
if (!ifindex && msg->msg_name) {
/* no bound device as default => check msg_name */
DECLARE_SOCKADDR(struct sockaddr_can *, addr, msg->msg_name);
- if (msg->msg_namelen < BCM_MIN_NAMELEN)
- return -EINVAL;
+ if (msg->msg_namelen < BCM_MIN_NAMELEN) {
+ ret = -EINVAL;
+ goto out_release;
+ }
- if (addr->can_family != AF_CAN)
- return -EINVAL;
+ if (addr->can_family != AF_CAN) {
+ ret = -EINVAL;
+ goto out_release;
+ }
/* ifindex from sendto() */
ifindex = addr->can_ifindex;
if (ifindex) {
struct net_device *dev;
dev = dev_get_by_index(sock_net(sk), ifindex);
- if (!dev)
- return -ENODEV;
+ if (!dev) {
+ ret = -ENODEV;
+ goto out_release;
+ }
if (dev->type != ARPHRD_CAN) {
dev_put(dev);
- return -ENODEV;
+ ret = -ENODEV;
+ goto out_release;
}
dev_put(dev);
}
}
- lock_sock(sk);
-
switch (msg_head.opcode) {
case TX_SETUP:
ret = bcm_tx_setup(&msg_head, msg, ifindex, sk);
break;
@@ -1497,10 +1525,11 @@ static int bcm_sendmsg(struct socket *sock, struct msghdr *msg, size_t size)
default:
ret = -EINVAL;
break;
}
+out_release:
release_sock(sk);
return ret;
}
@@ -1533,11 +1562,16 @@ static void bcm_notify(struct bcm_sock *bo, unsigned long msg,
if (sock_net(sk)->can.bcmproc_dir && bo->bcm_proc_read) {
remove_proc_entry(bo->procname, sock_net(sk)->can.bcmproc_dir);
bo->bcm_proc_read = NULL;
}
#endif
- bo->bound = 0;
+ /* Paired with the lockless fast-path check in
+ * bcm_sendmsg(); bo->ifindex itself is only ever
+ * accessed under lock_sock() so it needs no
+ * annotation.
+ */
+ WRITE_ONCE(bo->bound, 0);
bo->ifindex = 0;
notify_enodev = 1;
}
release_sock(sk);
@@ -1674,11 +1708,11 @@ static int bcm_release(struct socket *sock)
list_for_each_entry_safe(op, next, &bo->rx_ops, list)
bcm_remove_op(op);
/* remove device reference */
if (bo->bound) {
- bo->bound = 0;
+ WRITE_ONCE(bo->bound, 0);
bo->ifindex = 0;
}
sock_orphan(sk);
sock->sk = NULL;
@@ -1744,11 +1778,14 @@ static int bcm_connect(struct socket *sock, struct sockaddr_unsized *uaddr, int
goto fail;
}
}
#endif /* CONFIG_PROC_FS */
- bo->bound = 1;
+ /* bo->ifindex above is fully assigned before this point; pairs
+ * with the lockless fast-path check in bcm_sendmsg()
+ */
+ WRITE_ONCE(bo->bound, 1);
fail:
release_sock(sk);
return ret;
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH v11 02/11] can: bcm: fix lockless bound/ifindex race and silent RX_SETUP failure
2026-07-09 17:56 ` [PATCH v11 02/11] can: bcm: fix lockless bound/ifindex race and silent RX_SETUP failure Oliver Hartkopp via B4 Relay
@ 2026-07-09 18:09 ` sashiko-bot
0 siblings, 0 replies; 20+ messages in thread
From: sashiko-bot @ 2026-07-09 18:09 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: Oleksij Rempel, Marc Kleine-Budde, Vincent Mailhol,
Oliver Hartkopp, linux-can
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Lockless data race in BCM operation updates corrupts active TX/RX CAN frames.
--
--- Patch [2]: [PATCH v11 02/11] can: bcm: fix lockless bound/ifindex race and silent RX_SETUP failure ---
commit 86f16dbfed96fce9e0564ffdb9d5e63f4ec62cce
Author: Oliver Hartkopp <socketcan@hartkopp.net>
can: bcm: fix lockless bound/ifindex race and silent RX_SETUP failure
This commit fixes a race condition in bcm_sendmsg() by moving the bo->ifindex
read and bo->bound check under the socket lock. It also ensures bcm_rx_setup()
returns -ENODEV when the target device concurrently disappears.
Link: https://lore.kernel.org/linux-can/CAGp+u1aBK8QVjsvAxM2Ldzep4rEbsP9x_pV3At4g=h1kVEtyhA@mail.gmail.com/
> diff --git a/net/can/bcm.c b/net/can/bcm.c
> index bdf53241bd7b3..b612135b017d1 100644
> --- a/net/can/bcm.c
> +++ b/net/can/bcm.c
> @@ -1323,6 +1323,11 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
>
> op->rx_reg_dev = dev;
> dev_put(dev);
> + } else {
> + /* the requested device is gone - do not
> + * silently succeed without registering
> + */
> + err = -ENODEV;
> }
>
> } else
[Severity: High]
This is a pre-existing issue, but does bcm_rx_setup() modify the active
op->frames array locklessly on the RX side?
Looking at the earlier part of bcm_rx_setup():
/* update CAN frames content */
err = memcpy_from_msg(op->frames, msg,
msg_head->nframes * op->cfsiz);
Could this data race with the softirq reader bcm_rx_handler() while it is
actively processing frames?
[ ... ]
> @@ -1436,20 +1463,21 @@ static int bcm_sendmsg(struct socket *sock, struct msghdr *msg, size_t size)
> struct net_device *dev;
>
> dev = dev_get_by_index(sock_net(sk), ifindex);
> - if (!dev)
> - return -ENODEV;
> + if (!dev) {
> + ret = -ENODEV;
> + goto out_release;
> + }
>
> if (dev->type != ARPHRD_CAN) {
> dev_put(dev);
> - return -ENODEV;
> + ret = -ENODEV;
> + goto out_release;
> }
>
> dev_put(dev);
> }
> }
>
> - lock_sock(sk);
> -
> switch (msg_head.opcode) {
>
> case TX_SETUP:
[Severity: High]
This is also a pre-existing issue, but does bcm_tx_setup() write to the
active op->frames array locklessly before validating the frame length?
When TX_SETUP is called, bcm_tx_setup() copies user data:
cf = op->frames + op->cfsiz * i;
err = memcpy_from_msg((u8 *)cf, msg, op->cfsiz);
if (op->flags & CAN_FD_FRAME) {
if (cf->len > 64)
err = -EINVAL;
If an invalid length is provided from userspace, it is written to op->frames
first. The function catches the error and returns -EINVAL, but the
corrupted frame remains in the array.
Concurrently, can the high-resolution timer callback bcm_can_tx() transmit
this corrupted frame?
bcm_can_tx() reads the frame contents outside of bcm_tx_lock:
spin_lock_bh(&op->bcm_tx_lock);
cf = op->frames + op->cfsiz * op->currframe;
spin_unlock_bh(&op->bcm_tx_lock);
skb_put_data(skb, cf, op->cfsiz);
Could this pass an unvalidated cf->len down to the hardware driver and
potentially overflow device buffers? An attacker creating a cyclic
transmission via TX_SETUP, then issuing a concurrent TX_SETUP update with
an invalid length, might be able to trigger this condition.
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-bcm_fixes-v11-0-067ba11bec5c@hartkopp.net?part=2
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH v11 03/11] can: bcm: add locking when updating filter and timer values
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
2026-07-09 17:56 ` [PATCH v11 01/11] can: bcm: defer rx_op deallocation to workqueue to fix thrtimer UAF Oliver Hartkopp via B4 Relay
2026-07-09 17:56 ` [PATCH v11 02/11] can: bcm: fix lockless bound/ifindex race and silent RX_SETUP failure Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 18:18 ` sashiko-bot
2026-07-09 17:56 ` [PATCH v11 04/11] can: bcm: fix CAN frame rx/tx statistics Oliver Hartkopp via B4 Relay
` (7 subsequent siblings)
10 siblings, 1 reply; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde, syzbot+75e5e4ae00c3b4bb544e
From: Oliver Hartkopp <socketcan@hartkopp.net>
KCSAN detected a simultaneous access to timer values that can be
overwritten in bcm_rx_setup() when updating timer and filter content
while bcm_rx_handler(), bcm_rx_timeout_handler() or bcm_rx_thr_handler()
run concurrently on incoming CAN traffic.
Protect the timer (ival1/ival2/kt_ival1/kt_ival2/kt_lastmsg) and filter
(nframes/flags/frames/last_frames) updates in bcm_rx_setup() with a new
per-op bcm_rx_update_lock, taken with the matching scope in the RX
handlers. memcpy_from_msg() is staged into a temporary buffer before the
lock is taken, since it can sleep and must not run under a spinlock.
hrtimer_cancel() is always called without bcm_rx_update_lock held, since
bcm_rx_timeout_handler()/bcm_rx_thr_handler() take the same lock and a
running callback would otherwise deadlock against the canceller.
Also close a related race: bcm_rx_setup() cleared the RTR flag in the
stored reply frame's can_id as a separate, unprotected step after the
frame content was already installed, so a concurrent bcm_rx_handler()
could transmit a stale reply with CAN_RTR_FLAG still set. Fold that
normalization into the initial frame preparation instead (on the staged
buffer for updates, directly on op->frames pre-registration for new
ops), so the installed frame is always atomically self-consistent.
bcm_rx_handler()'s RX_RTR_FRAME check now takes a lock-protected
snapshot of op->flags before deciding whether to call bcm_can_tx(),
but does not hold the lock across that call.
Also take a lock-protected snapshot of the currframe in bcm_can_tx()
to avoid partly overwrites by content updates in bcm_tx_setup().
Omit calling hrtimer_forward() with zero interval in bcm_rx_thr_handler().
kt_ival2 may have been concurrently cleared by bcm_rx_setup() before it
cancels this timer, so check kt_ival2 inside the bcm_rx_update_lock.
Fixes: c2aba69d0c36 ("can: bcm: add locking for bcm_op runtime updates")
Reported-by: syzbot+75e5e4ae00c3b4bb544e@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/linux-can/6975d5cf.a00a0220.33ccc7.0022.GAE@google.com/
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 136 ++++++++++++++++++++++++++++++++++++++++++++--------------
1 file changed, 104 insertions(+), 32 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index b612135b017d..af4bb4182bc2 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -127,10 +127,11 @@ struct bcm_op {
struct canfd_frame sframe;
struct canfd_frame last_sframe;
struct sock *sk;
struct net_device *rx_reg_dev;
spinlock_t bcm_tx_lock; /* protect currframe/count in runtime updates */
+ spinlock_t bcm_rx_update_lock; /* protect filter/timer data updates */
};
struct bcm_sock {
struct sock sk;
int bound;
@@ -291,26 +292,30 @@ static int bcm_proc_show(struct seq_file *m, void *v)
/*
* bcm_can_tx - send the (next) CAN frame to the appropriate CAN interface
* of the given bcm tx op
*/
-static void bcm_can_tx(struct bcm_op *op)
+static void bcm_can_tx(struct bcm_op *op, struct canfd_frame *cf)
{
struct sk_buff *skb;
struct can_skb_ext *csx;
struct net_device *dev;
- struct canfd_frame *cf;
+ struct canfd_frame cframe;
int err;
/* no target device? => exit */
if (!op->ifindex)
return;
- /* read currframe under lock protection */
- spin_lock_bh(&op->bcm_tx_lock);
- cf = op->frames + op->cfsiz * op->currframe;
- spin_unlock_bh(&op->bcm_tx_lock);
+ if (!cf) {
+ /* read currframe under lock protection */
+ spin_lock_bh(&op->bcm_tx_lock);
+ memcpy(&cframe, op->frames + op->cfsiz * op->currframe,
+ op->cfsiz);
+ cf = &cframe;
+ spin_unlock_bh(&op->bcm_tx_lock);
+ }
dev = dev_get_by_index(sock_net(op->sk), op->ifindex);
if (!dev) {
/* RFC: should this bcm_op remove itself here? */
return;
@@ -459,11 +464,11 @@ static enum hrtimer_restart bcm_tx_timeout_handler(struct hrtimer *hrtimer)
{
struct bcm_op *op = container_of(hrtimer, struct bcm_op, timer);
struct bcm_msg_head msg_head;
if (op->kt_ival1 && (op->count > 0)) {
- bcm_can_tx(op);
+ bcm_can_tx(op, NULL);
if (!op->count && (op->flags & TX_COUNTEVT)) {
/* create notification to user */
memset(&msg_head, 0, sizeof(msg_head));
msg_head.opcode = TX_EXPIRED;
@@ -476,11 +481,11 @@ static enum hrtimer_restart bcm_tx_timeout_handler(struct hrtimer *hrtimer)
bcm_send_to_user(op, &msg_head, NULL, 0);
}
} else if (op->kt_ival2) {
- bcm_can_tx(op);
+ bcm_can_tx(op, NULL);
}
return bcm_tx_set_expiry(op, &op->timer) ?
HRTIMER_RESTART : HRTIMER_NORESTART;
}
@@ -620,10 +625,12 @@ static void bcm_rx_starttimer(struct bcm_op *op)
static enum hrtimer_restart bcm_rx_timeout_handler(struct hrtimer *hrtimer)
{
struct bcm_op *op = container_of(hrtimer, struct bcm_op, timer);
struct bcm_msg_head msg_head;
+ spin_lock_bh(&op->bcm_rx_update_lock);
+
/* if user wants to be informed, when cyclic CAN-Messages come back */
if ((op->flags & RX_ANNOUNCE_RESUME) && op->last_frames) {
/* clear received CAN frames to indicate 'nothing received' */
memset(op->last_frames, 0, op->nframes * op->cfsiz);
}
@@ -636,10 +643,12 @@ static enum hrtimer_restart bcm_rx_timeout_handler(struct hrtimer *hrtimer)
msg_head.ival1 = op->ival1;
msg_head.ival2 = op->ival2;
msg_head.can_id = op->can_id;
msg_head.nframes = 0;
+ spin_unlock_bh(&op->bcm_rx_update_lock);
+
bcm_send_to_user(op, &msg_head, NULL, 0);
return HRTIMER_NORESTART;
}
@@ -684,30 +693,43 @@ static int bcm_rx_thr_flush(struct bcm_op *op)
* Check for throttled data and send it to the userspace
*/
static enum hrtimer_restart bcm_rx_thr_handler(struct hrtimer *hrtimer)
{
struct bcm_op *op = container_of(hrtimer, struct bcm_op, thrtimer);
+ enum hrtimer_restart ret;
+
+ spin_lock_bh(&op->bcm_rx_update_lock);
- if (bcm_rx_thr_flush(op)) {
+ /* kt_ival2 may have been concurrently cleared by bcm_rx_setup()
+ * before it cancels this timer - never forward with a zero
+ * interval in that case.
+ */
+ if (bcm_rx_thr_flush(op) && op->kt_ival2) {
hrtimer_forward_now(hrtimer, op->kt_ival2);
- return HRTIMER_RESTART;
+ ret = HRTIMER_RESTART;
} else {
/* rearm throttle handling */
op->kt_lastmsg = 0;
- return HRTIMER_NORESTART;
+ ret = HRTIMER_NORESTART;
}
+
+ spin_unlock_bh(&op->bcm_rx_update_lock);
+
+ return ret;
}
/*
* bcm_rx_handler - handle a CAN frame reception
*/
static void bcm_rx_handler(struct sk_buff *skb, void *data)
{
struct bcm_op *op = (struct bcm_op *)data;
const struct canfd_frame *rxframe = (struct canfd_frame *)skb->data;
+ struct canfd_frame rtrframe;
unsigned int i;
unsigned char traffic_flags;
+ bool rtr_frame;
if (op->can_id != rxframe->can_id)
return;
/* make sure to handle the correct frame type (CAN / CAN FD) */
@@ -727,13 +749,22 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
/* save originator for recvfrom() */
op->rx_ifindex = skb->dev->ifindex;
/* update statistics */
op->frames_abs++;
- if (op->flags & RX_RTR_FRAME) {
+ /* snapshot the flag under lock: op->flags/op->frames may be updated
+ * concurrently by bcm_rx_setup().
+ */
+ spin_lock_bh(&op->bcm_rx_update_lock);
+ rtr_frame = op->flags & RX_RTR_FRAME;
+ if (rtr_frame)
+ memcpy(&rtrframe, op->frames, op->cfsiz);
+ spin_unlock_bh(&op->bcm_rx_update_lock);
+
+ if (rtr_frame) {
/* send reply for RTR-request (placed in op->frames[0]) */
- bcm_can_tx(op);
+ bcm_can_tx(op, &rtrframe);
return;
}
/* compute flags to distinguish between own/local/remote CAN traffic */
traffic_flags = 0;
@@ -741,10 +772,12 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
traffic_flags |= RX_LOCAL;
if (skb->sk == op->sk)
traffic_flags |= RX_OWN;
}
+ spin_lock_bh(&op->bcm_rx_update_lock);
+
if (op->flags & RX_FILTER_ID) {
/* the easiest case */
bcm_rx_update_and_send(op, op->last_frames, rxframe,
traffic_flags);
goto rx_starttimer;
@@ -776,10 +809,12 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
}
}
rx_starttimer:
bcm_rx_starttimer(op);
+
+ spin_unlock_bh(&op->bcm_rx_update_lock);
}
/*
* helpers for bcm_op handling: find & delete bcm [rx|tx] op elements
*/
@@ -1114,11 +1149,11 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
/* spec: send CAN frame when starting timer */
op->flags |= TX_ANNOUNCE;
}
if (op->flags & TX_ANNOUNCE)
- bcm_can_tx(op);
+ bcm_can_tx(op, NULL);
if (op->flags & STARTTIMER)
bcm_tx_start_timer(op);
return msg_head->nframes * op->cfsiz + MHSIZ;
@@ -1128,10 +1163,28 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
kfree(op->frames);
kfree(op);
return err;
}
+static void bcm_rx_setup_rtr_check(struct bcm_msg_head *msg_head,
+ struct bcm_op *op, void *new_frames)
+{
+ /* funny feature in RX(!)_SETUP only for RTR-mode:
+ * copy can_id into frame BUT without RTR-flag to
+ * prevent a full-load-loopback-test ... ;-]
+ * normalize this on the staged buffer, before it is
+ * ever installed into op->frames.
+ */
+ if (msg_head->flags & RX_RTR_FRAME) {
+ struct canfd_frame *frame0 = new_frames;
+
+ if ((msg_head->flags & TX_CP_CAN_ID) ||
+ frame0->can_id == op->can_id)
+ frame0->can_id = op->can_id & ~CAN_RTR_FLAG;
+ }
+}
+
/*
* bcm_rx_setup - create or update a bcm rx op (for bcm_sendmsg)
*/
static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
int ifindex, struct sock *sk)
@@ -1162,10 +1215,12 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
return -EINVAL;
/* check the given can_id */
op = bcm_find_op(&bo->rx_ops, msg_head, ifindex);
if (op) {
+ void *new_frames = NULL;
+
/* update existing BCM operation */
/*
* Do we need more space for the CAN frames than currently
* allocated? -> This is a _really_ unusual use-case and
@@ -1173,33 +1228,55 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
*/
if (msg_head->nframes > op->nframes)
return -E2BIG;
if (msg_head->nframes) {
- /* update CAN frames content */
- err = memcpy_from_msg(op->frames, msg,
+ /* get new CAN frames content before locking */
+ new_frames = kmalloc(msg_head->nframes * op->cfsiz,
+ GFP_KERNEL);
+ if (!new_frames)
+ return -ENOMEM;
+
+ err = memcpy_from_msg(new_frames, msg,
msg_head->nframes * op->cfsiz);
- if (err < 0)
+ if (err < 0) {
+ kfree(new_frames);
return err;
+ }
- /* clear last_frames to indicate 'nothing received' */
- memset(op->last_frames, 0, msg_head->nframes * op->cfsiz);
+ bcm_rx_setup_rtr_check(msg_head, op, new_frames);
}
+ spin_lock_bh(&op->bcm_rx_update_lock);
op->nframes = msg_head->nframes;
op->flags = msg_head->flags;
+ if (msg_head->nframes) {
+ /* update CAN frames content */
+ memcpy(op->frames, new_frames,
+ msg_head->nframes * op->cfsiz);
+
+ /* clear last_frames to indicate 'nothing received' */
+ memset(op->last_frames, 0,
+ msg_head->nframes * op->cfsiz);
+ }
+ spin_unlock_bh(&op->bcm_rx_update_lock);
+
+ /* free temporary frames / kfree(NULL) is safe */
+ kfree(new_frames);
+
/* Only an update -> do not call can_rx_register() */
do_rx_register = 0;
} else {
/* insert new BCM operation for the given can_id */
op = kzalloc(OPSIZ, GFP_KERNEL);
if (!op)
return -ENOMEM;
spin_lock_init(&op->bcm_tx_lock);
+ spin_lock_init(&op->bcm_rx_update_lock);
op->can_id = msg_head->can_id;
op->nframes = msg_head->nframes;
op->cfsiz = CFSIZ(msg_head->flags);
op->flags = msg_head->flags;
@@ -1237,10 +1314,12 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
if (op->last_frames != &op->last_sframe)
kfree(op->last_frames);
kfree(op);
return err;
}
+
+ bcm_rx_setup_rtr_check(msg_head, op, op->frames);
}
/* bcm_can_tx / bcm_tx_timeout_handler needs this */
op->sk = sk;
sock_hold(sk);
@@ -1264,45 +1343,38 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
} /* if ((op = bcm_find_op(&bo->rx_ops, msg_head->can_id, ifindex))) */
/* check flags */
if (op->flags & RX_RTR_FRAME) {
- struct canfd_frame *frame0 = op->frames;
-
/* no timers in RTR-mode */
hrtimer_cancel(&op->thrtimer);
hrtimer_cancel(&op->timer);
-
- /*
- * funny feature in RX(!)_SETUP only for RTR-mode:
- * copy can_id into frame BUT without RTR-flag to
- * prevent a full-load-loopback-test ... ;-]
- */
- if ((op->flags & TX_CP_CAN_ID) ||
- (frame0->can_id == op->can_id))
- frame0->can_id = op->can_id & ~CAN_RTR_FLAG;
-
} else {
if (op->flags & SETTIMER) {
/* set timer value */
+ spin_lock_bh(&op->bcm_rx_update_lock);
op->ival1 = msg_head->ival1;
op->ival2 = msg_head->ival2;
op->kt_ival1 = bcm_timeval_to_ktime(msg_head->ival1);
op->kt_ival2 = bcm_timeval_to_ktime(msg_head->ival2);
+ op->kt_lastmsg = 0;
+ spin_unlock_bh(&op->bcm_rx_update_lock);
/* disable an active timer due to zero value? */
if (!op->kt_ival1)
hrtimer_cancel(&op->timer);
/*
* In any case cancel the throttle timer, flush
* potentially blocked msgs and reset throttle handling
*/
- op->kt_lastmsg = 0;
hrtimer_cancel(&op->thrtimer);
+
+ spin_lock_bh(&op->bcm_rx_update_lock);
bcm_rx_thr_flush(op);
+ spin_unlock_bh(&op->bcm_rx_update_lock);
}
if ((op->flags & STARTTIMER) && op->kt_ival1)
hrtimer_start(&op->timer, op->kt_ival1,
HRTIMER_MODE_REL_SOFT);
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH v11 03/11] can: bcm: add locking when updating filter and timer values
2026-07-09 17:56 ` [PATCH v11 03/11] can: bcm: add locking when updating filter and timer values Oliver Hartkopp via B4 Relay
@ 2026-07-09 18:18 ` sashiko-bot
0 siblings, 0 replies; 20+ messages in thread
From: sashiko-bot @ 2026-07-09 18:18 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: Oliver Hartkopp, Marc Kleine-Budde, Vincent Mailhol, linux-can,
Oleksij Rempel
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] TX path directly overwrites active CAN frames via `memcpy_from_msg()` without a staging buffer or write-side lock, rendering the read-side lock useless and allowing torn frames or arbitrary `can_id` injection.
--
--- Patch [3]: [PATCH v11 03/11] can: bcm: add locking when updating filter and timer values ---
commit e17f7ae3aed2616decbe79bc147ce9802da8e284
Author: Oliver Hartkopp <socketcan@hartkopp.net>
can: bcm: add locking when updating filter and timer values
This commit adds locking to protect timer and filter data updates in
bcm_rx_setup() against concurrent execution of RX handlers. It also adds a
lock-protected snapshot of the currframe in bcm_can_tx() to avoid partial
overwrites by content updates in bcm_tx_setup().
Closes: https://lore.kernel.org/linux-can/6975d5cf.a00a0220.33ccc7.0022.GAE@google.com/
> diff --git a/net/can/bcm.c b/net/can/bcm.c
> index b612135b017d1..af4bb4182bc29 100644
> --- a/net/can/bcm.c
> +++ b/net/can/bcm.c
[ ... ]
> @@ -293,22 +294,26 @@ static int bcm_proc_show(struct seq_file *m, void *v)
> * bcm_can_tx - send the (next) CAN frame to the appropriate CAN interface
> * of the given bcm tx op
> */
> -static void bcm_can_tx(struct bcm_op *op)
> +static void bcm_can_tx(struct bcm_op *op, struct canfd_frame *cf)
> {
[ ... ]
> - /* read currframe under lock protection */
> - spin_lock_bh(&op->bcm_tx_lock);
> - cf = op->frames + op->cfsiz * op->currframe;
> - spin_unlock_bh(&op->bcm_tx_lock);
> + if (!cf) {
> + /* read currframe under lock protection */
> + spin_lock_bh(&op->bcm_tx_lock);
> + memcpy(&cframe, op->frames + op->cfsiz * op->currframe,
> + op->cfsiz);
> + cf = &cframe;
> + spin_unlock_bh(&op->bcm_tx_lock);
> + }
[Severity: High]
This is a pre-existing issue, but does this read-side snapshot actually
provide protection against partial overwrites from bcm_tx_setup()?
When updating an existing TX operation, bcm_tx_setup() directly overwrites
op->frames using the sleepable memcpy_from_msg() without holding
bcm_tx_lock or using a staging buffer:
net/can/bcm.c:bcm_tx_setup() {
...
err = memcpy_from_msg((u8 *)cf, msg, op->cfsiz);
...
if (msg_head->flags & TX_CP_CAN_ID) {
/* copy can_id into frame */
cf->can_id = msg_head->can_id;
}
...
}
Because the write-side completely ignores the lock, could a concurrent
timer still transmit torn CAN frames?
If TX_CP_CAN_ID is used, could the frame be transmitted after the user
payload is copied but before the enforced can_id is applied, allowing
unprivileged users to bypass ID enforcement and inject arbitrary CAN IDs?
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-bcm_fixes-v11-0-067ba11bec5c@hartkopp.net?part=3
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH v11 04/11] can: bcm: fix CAN frame rx/tx statistics
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
` (2 preceding siblings ...)
2026-07-09 17:56 ` [PATCH v11 03/11] can: bcm: add locking when updating filter and timer values Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 18:23 ` sashiko-bot
2026-07-09 17:56 ` [PATCH v11 05/11] can: bcm: add missing rcu list annotations and operations Oliver Hartkopp via B4 Relay
` (6 subsequent siblings)
10 siblings, 1 reply; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde
From: Oliver Hartkopp <socketcan@hartkopp.net>
KCSAN detected a data race within the bcm_rx_handler() when two CAN frames
have been simultaneously received and processed in a single rx op by two
different CPUs.
Use atomic operations with (signed) long data types to access the
statistics in the hot path to fix the KCSAN complaint.
Additionally simplify the update and check of statistics overflow by
using the atomic operations in separate bcm_update_[rx|tx]_stats()
functions. The rx variant runs under bcm_rx_update_lock to prevent
races when resetting the two rx counters; the tx variant runs under
bcm_tx_lock and only needs to guard its own counter's overflow.
As the rx path resets its values already at LONG_MAX / 100, there is
no conflict between the two locking domains (bcm_rx_update_lock vs.
bcm_tx_lock) even for ops that use both paths.
The rx statistics update and the frames_filtered update in
bcm_rx_changed() were previously performed in two separate
bcm_rx_update_lock sections. For an rx op subscribed on all interfaces
(ifindex == 0), bcm_rx_handler() can run concurrently on different
CPUs, so a counter reset by one CPU between these two sections could
leave frames_filtered larger than frames_abs on another CPU, producing
a bogus (even negative) reduction percentage in procfs. Update the
statistics in the same critical section as bcm_rx_changed() to close
this gap, which also removes the now unneeded extra lock/unlock pair
around the traffic_flags calculation.
Fixes: ffd980f976e7 ("[CAN]: Add broadcast manager (bcm) protocol")
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 69 +++++++++++++++++++++++++++++++++++++++--------------------
1 file changed, 46 insertions(+), 23 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index af4bb4182bc2..38c3f690c1cd 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -110,11 +110,11 @@ struct bcm_op {
struct rcu_head rcu;
struct work_struct work;
int ifindex;
canid_t can_id;
u32 flags;
- unsigned long frames_abs, frames_filtered;
+ atomic_long_t frames_abs, frames_filtered;
struct bcm_timeval ival1, ival2;
struct hrtimer timer, thrtimer;
ktime_t rx_stamp, kt_ival1, kt_ival2, kt_lastmsg;
int rx_ifindex;
int cfsiz;
@@ -227,14 +227,17 @@ static int bcm_proc_show(struct seq_file *m, void *v)
rcu_read_lock();
list_for_each_entry_rcu(op, &bo->rx_ops, list) {
- unsigned long reduction;
+ long reduction, frames_filtered, frames_abs;
+
+ frames_filtered = atomic_long_read(&op->frames_filtered);
+ frames_abs = atomic_long_read(&op->frames_abs);
/* print only active entries & prevent division by zero */
- if (!op->frames_abs)
+ if (!frames_abs)
continue;
seq_printf(m, "rx_op: %03X %-5s ", op->can_id,
bcm_proc_getifname(net, ifname, op->ifindex));
@@ -252,13 +255,13 @@ static int bcm_proc_show(struct seq_file *m, void *v)
if (op->kt_ival2)
seq_printf(m, "thr=%lld ",
(long long)ktime_to_us(op->kt_ival2));
seq_printf(m, "# recv %ld (%ld) => reduction: ",
- op->frames_filtered, op->frames_abs);
+ frames_filtered, frames_abs);
- reduction = 100 - (op->frames_filtered * 100) / op->frames_abs;
+ reduction = 100 - (frames_filtered * 100) / frames_abs;
seq_printf(m, "%s%ld%%\n",
(reduction == 100) ? "near " : "", reduction);
}
@@ -278,20 +281,39 @@ static int bcm_proc_show(struct seq_file *m, void *v)
if (op->kt_ival2)
seq_printf(m, "t2=%lld ",
(long long)ktime_to_us(op->kt_ival2));
- seq_printf(m, "# sent %ld\n", op->frames_abs);
+ seq_printf(m, "# sent %ld\n",
+ atomic_long_read(&op->frames_abs));
}
seq_putc(m, '\n');
rcu_read_unlock();
return 0;
}
#endif /* CONFIG_PROC_FS */
+static void bcm_update_rx_stats(struct bcm_op *op)
+{
+ /* prevent overflow of the reduction% calculation in bcm_proc_show() */
+ if (atomic_long_inc_return(&op->frames_abs) > LONG_MAX / 100) {
+ atomic_long_set(&op->frames_filtered, 0);
+ atomic_long_set(&op->frames_abs, 0);
+ }
+}
+
+static void bcm_update_tx_stats(struct bcm_op *op)
+{
+ /* tx_op has no reduction% calculation - use the full range and
+ * just keep the displayed counter non-negative on overflow
+ */
+ if (atomic_long_inc_return(&op->frames_abs) == LONG_MAX)
+ atomic_long_set(&op->frames_abs, 0);
+}
+
/*
* bcm_can_tx - send the (next) CAN frame to the appropriate CAN interface
* of the given bcm tx op
*/
static void bcm_can_tx(struct bcm_op *op, struct canfd_frame *cf)
@@ -342,11 +364,11 @@ static void bcm_can_tx(struct bcm_op *op, struct canfd_frame *cf)
/* update currframe and count under lock protection */
spin_lock_bh(&op->bcm_tx_lock);
if (!err)
- op->frames_abs++;
+ bcm_update_tx_stats(op);
op->currframe++;
/* reached last frame? */
if (op->currframe >= op->nframes)
@@ -495,16 +517,13 @@ static enum hrtimer_restart bcm_tx_timeout_handler(struct hrtimer *hrtimer)
*/
static void bcm_rx_changed(struct bcm_op *op, struct canfd_frame *data)
{
struct bcm_msg_head head;
- /* update statistics */
- op->frames_filtered++;
-
- /* prevent statistics overflow */
- if (op->frames_filtered > ULONG_MAX/100)
- op->frames_filtered = op->frames_abs = 0;
+ /* update statistics (frames_filtered <= frames_abs) */
+ if (atomic_long_read(&op->frames_abs))
+ atomic_long_inc(&op->frames_filtered);
/* this element is not throttled anymore */
data->flags &= ~RX_THR;
memset(&head, 0, sizeof(head));
@@ -746,38 +765,42 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
/* save rx timestamp */
op->rx_stamp = skb->tstamp;
/* save originator for recvfrom() */
op->rx_ifindex = skb->dev->ifindex;
- /* update statistics */
- op->frames_abs++;
- /* snapshot the flag under lock: op->flags/op->frames may be updated
- * concurrently by bcm_rx_setup().
- */
+ /* op->flags/op->frames may be updated concurrently by bcm_rx_setup() */
spin_lock_bh(&op->bcm_rx_update_lock);
+
rtr_frame = op->flags & RX_RTR_FRAME;
- if (rtr_frame)
+ if (rtr_frame) {
+ bcm_update_rx_stats(op);
+ /* snapshot RTR content under lock */
memcpy(&rtrframe, op->frames, op->cfsiz);
- spin_unlock_bh(&op->bcm_rx_update_lock);
+ spin_unlock_bh(&op->bcm_rx_update_lock);
- if (rtr_frame) {
/* send reply for RTR-request (placed in op->frames[0]) */
bcm_can_tx(op, &rtrframe);
return;
}
+ /* update statistics in the same critical section as bcm_rx_changed()
+ * below: frames_filtered must never be checked/incremented against a
+ * frames_abs snapshot from a concurrent bcm_rx_handler() call on
+ * another CPU for the same (wildcard) op, or frames_filtered can end
+ * up larger than frames_abs.
+ */
+ bcm_update_rx_stats(op);
+
/* compute flags to distinguish between own/local/remote CAN traffic */
traffic_flags = 0;
if (skb->sk) {
traffic_flags |= RX_LOCAL;
if (skb->sk == op->sk)
traffic_flags |= RX_OWN;
}
- spin_lock_bh(&op->bcm_rx_update_lock);
-
if (op->flags & RX_FILTER_ID) {
/* the easiest case */
bcm_rx_update_and_send(op, op->last_frames, rxframe,
traffic_flags);
goto rx_starttimer;
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH v11 04/11] can: bcm: fix CAN frame rx/tx statistics
2026-07-09 17:56 ` [PATCH v11 04/11] can: bcm: fix CAN frame rx/tx statistics Oliver Hartkopp via B4 Relay
@ 2026-07-09 18:23 ` sashiko-bot
0 siblings, 0 replies; 20+ messages in thread
From: sashiko-bot @ 2026-07-09 18:23 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: Marc Kleine-Budde, Oleksij Rempel, linux-can, Vincent Mailhol,
Oliver Hartkopp
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] `frames_filtered` can exceed `frames_abs` when a throttle timer flushes multiple multiplex frames across a `frames_abs` reset, causing negative reduction percentages.
--
--- Patch [4]: [PATCH v11 04/11] can: bcm: fix CAN frame rx/tx statistics ---
commit 6e9ade605885c93bf1fc795169b773f4a95e422c
Author: Oliver Hartkopp <socketcan@hartkopp.net>
can: bcm: fix CAN frame rx/tx statistics
This commit fixes a KCSAN data race complaint in bcm_rx_handler() by
converting frames_abs and frames_filtered to atomic operations. It also
moves overflow updates into separate functions and aims to prevent
negative procfs reduction percentages.
> diff --git a/net/can/bcm.c b/net/can/bcm.c
> index af4bb4182bc29..38c3f690c1cd0 100644
> --- a/net/can/bcm.c
> +++ b/net/can/bcm.c
[ ... ]
> @@ -290,6 +294,24 @@ static int bcm_proc_show(struct seq_file *m, void *v)
> }
> #endif /* CONFIG_PROC_FS */
>
> +static void bcm_update_rx_stats(struct bcm_op *op)
> +{
> + /* prevent overflow of the reduction% calculation in bcm_proc_show() */
> + if (atomic_long_inc_return(&op->frames_abs) > LONG_MAX / 100) {
> + atomic_long_set(&op->frames_filtered, 0);
> + atomic_long_set(&op->frames_abs, 0);
> + }
> +}
> +
> +static void bcm_update_tx_stats(struct bcm_op *op)
> +{
[ ... ]
> @@ -497,12 +519,9 @@ static void bcm_rx_changed(struct bcm_op *op, struct canfd_frame *data)
> {
> struct bcm_msg_head head;
>
> - /* update statistics */
> - op->frames_filtered++;
> -
> - /* prevent statistics overflow */
> - if (op->frames_filtered > ULONG_MAX/100)
> - op->frames_filtered = op->frames_abs = 0;
> + /* update statistics (frames_filtered <= frames_abs) */
> + if (atomic_long_read(&op->frames_abs))
> + atomic_long_inc(&op->frames_filtered);
[Severity: Medium]
Could frames_filtered still exceed frames_abs here if a throttle timer flushes
multiple multiplex frames right after a counter reset?
If frames_abs reaches the LONG_MAX / 100 limit, bcm_update_rx_stats() resets
both counters to 0. If a subsequent frame is received, frames_abs is
incremented to 1.
If an active throttle timer then expires, bcm_rx_thr_flush() iterates and
calls bcm_rx_changed() for previously throttled frames:
net/can/bcm.c:bcm_rx_thr_flush() {
...
/* for MUX filter we start at index 1 */
for (i = 1; i < op->nframes; i++)
updated += bcm_rx_do_flush(op, i);
...
}
Since frames_abs is non-zero (1), the check in bcm_rx_changed() evaluates to
true, and it will increment frames_filtered for each flushed frame.
Because bcm_rx_thr_flush() can process multiple frames at once,
frames_filtered could quickly become larger than frames_abs, which seems to
result in the negative procfs reduction percentages this patch intended to
prevent.
> /* this element is not throttled anymore */
> data->flags &= ~RX_THR;
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-bcm_fixes-v11-0-067ba11bec5c@hartkopp.net?part=4
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH v11 05/11] can: bcm: add missing rcu list annotations and operations
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
` (3 preceding siblings ...)
2026-07-09 17:56 ` [PATCH v11 04/11] can: bcm: fix CAN frame rx/tx statistics Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 18:21 ` sashiko-bot
2026-07-09 17:56 ` [PATCH v11 06/11] can: bcm: extend bcm_tx_lock usage for data and timer updates Oliver Hartkopp via B4 Relay
` (5 subsequent siblings)
10 siblings, 1 reply; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde, sashiko-reviews
From: Oliver Hartkopp <socketcan@hartkopp.net>
sashiko-bot remarked the missing use of list_add_rcu() in
bcm_[rx|tx]_setup() to have a proper initialized bcm_op structure
when bcm_proc_show() traverses the bcm_op's under rcu_read_lock().
To cover all initial settings of the bcm_op's the list_add_rcu() calls
are moved to the end of the setup code.
While at it, also fix the mirroring removal side: bcm_release() called
bcm_remove_op() - which frees the op via call_rcu() - on ops that were
still linked in bo->tx_ops/bo->rx_ops, without list_del_rcu() first.
Unlink each op with list_del_rcu() before handing it to bcm_remove_op(),
matching the existing pattern in bcm_delete_tx_op()/bcm_delete_rx_op().
Reported-by: sashiko-reviews@lists.linux.dev
Closes: https://lore.kernel.org/linux-can/20260610094654.A1FFE1F00893@smtp.kernel.org/
Fixes: dac5e6249159 ("can: bcm: add missing rcu read protection for procfs content")
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 25 ++++++++++++++++---------
1 file changed, 16 insertions(+), 9 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index 38c3f690c1cd..940cd9df1cca 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -263,11 +263,11 @@ static int bcm_proc_show(struct seq_file *m, void *v)
seq_printf(m, "%s%ld%%\n",
(reduction == 100) ? "near " : "", reduction);
}
- list_for_each_entry(op, &bo->tx_ops, list) {
+ list_for_each_entry_rcu(op, &bo->tx_ops, list) {
seq_printf(m, "tx_op: %03X %s ", op->can_id,
bcm_proc_getifname(net, ifname, op->ifindex));
if (op->flags & CAN_FD_FRAME)
@@ -1007,10 +1007,11 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
int ifindex, struct sock *sk)
{
struct bcm_sock *bo = bcm_sk(sk);
struct bcm_op *op;
struct canfd_frame *cf;
+ bool add_op_to_list = false;
unsigned int i;
int err;
/* we need a real device to send frames */
if (!ifindex)
@@ -1148,12 +1149,11 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
/* currently unused in tx_ops */
hrtimer_setup(&op->thrtimer, hrtimer_dummy_timeout, CLOCK_MONOTONIC,
HRTIMER_MODE_REL_SOFT);
- /* add this bcm_op to the list of the tx_ops */
- list_add(&op->list, &bo->tx_ops);
+ add_op_to_list = true;
} /* if ((op = bcm_find_op(&bo->tx_ops, msg_head->can_id, ifindex))) */
if (op->flags & SETTIMER) {
/* set timer values */
@@ -1171,10 +1171,14 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
hrtimer_cancel(&op->timer);
/* spec: send CAN frame when starting timer */
op->flags |= TX_ANNOUNCE;
}
+ /* add this bcm_op to the list of the tx_ops? */
+ if (add_op_to_list)
+ list_add_rcu(&op->list, &bo->tx_ops);
+
if (op->flags & TX_ANNOUNCE)
bcm_can_tx(op, NULL);
if (op->flags & STARTTIMER)
bcm_tx_start_timer(op);
@@ -1355,13 +1359,10 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
hrtimer_setup(&op->timer, bcm_rx_timeout_handler, CLOCK_MONOTONIC,
HRTIMER_MODE_REL_SOFT);
hrtimer_setup(&op->thrtimer, bcm_rx_thr_handler, CLOCK_MONOTONIC,
HRTIMER_MODE_REL_SOFT);
- /* add this bcm_op to the list of the rx_ops */
- list_add(&op->list, &bo->rx_ops);
-
/* call can_rx_register() */
do_rx_register = 1;
} /* if ((op = bcm_find_op(&bo->rx_ops, msg_head->can_id, ifindex))) */
@@ -1429,14 +1430,16 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
err = can_rx_register(sock_net(sk), NULL, op->can_id,
REGMASK(op->can_id),
bcm_rx_handler, op, "bcm", sk);
if (err) {
/* this bcm rx op is broken -> remove it */
- list_del_rcu(&op->list);
bcm_remove_op(op);
return err;
}
+
+ /* add this bcm_op to the list of the rx_ops */
+ list_add_rcu(&op->list, &bo->rx_ops);
}
return msg_head->nframes * op->cfsiz + MHSIZ;
}
@@ -1766,12 +1769,14 @@ static int bcm_release(struct socket *sock)
/* remove procfs entry */
if (net->can.bcmproc_dir && bo->bcm_proc_read)
remove_proc_entry(bo->procname, net->can.bcmproc_dir);
#endif /* CONFIG_PROC_FS */
- list_for_each_entry_safe(op, next, &bo->tx_ops, list)
+ list_for_each_entry_safe(op, next, &bo->tx_ops, list) {
+ list_del_rcu(&op->list);
bcm_remove_op(op);
+ }
list_for_each_entry_safe(op, next, &bo->rx_ops, list) {
/*
* Don't care if we're bound or not (due to netdev problems)
* can_rx_unregister() is always a save thing to do here.
@@ -1798,12 +1803,14 @@ static int bcm_release(struct socket *sock)
}
synchronize_rcu();
- list_for_each_entry_safe(op, next, &bo->rx_ops, list)
+ list_for_each_entry_safe(op, next, &bo->rx_ops, list) {
+ list_del_rcu(&op->list);
bcm_remove_op(op);
+ }
/* remove device reference */
if (bo->bound) {
WRITE_ONCE(bo->bound, 0);
bo->ifindex = 0;
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH v11 05/11] can: bcm: add missing rcu list annotations and operations
2026-07-09 17:56 ` [PATCH v11 05/11] can: bcm: add missing rcu list annotations and operations Oliver Hartkopp via B4 Relay
@ 2026-07-09 18:21 ` sashiko-bot
0 siblings, 0 replies; 20+ messages in thread
From: sashiko-bot @ 2026-07-09 18:21 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: linux-can, Marc Kleine-Budde, Oliver Hartkopp, Vincent Mailhol,
Oleksij Rempel
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Unsynchronized reads of 64-bit ktime_t and struct bcm_timeval values cause torn reads on 32-bit systems, leading to corrupted hrtimer expirations.
--
--- Patch [5]: [PATCH v11 05/11] can: bcm: add missing rcu list annotations and operations ---
commit c261663cca88ad03b64923cce2abd1bd069a9797
Author: Oliver Hartkopp <socketcan@hartkopp.net>
can: bcm: add missing rcu list annotations and operations
This commit adds missing list_add_rcu() calls in bcm_tx_setup() and
bcm_rx_setup() to ensure proper initialization when bcm_proc_show()
traverses the operations under rcu_read_lock(). It also fixes a list
removal issue in bcm_release() by unlinking each operation with
list_del_rcu() before removing.
Closes: https://lore.kernel.org/linux-can/20260610094654.A1FFE1F00893@smtp.kernel.org/
> diff --git a/net/can/bcm.c b/net/can/bcm.c
> index 38c3f690c1cd0..940cd9df1cca3 100644
> --- a/net/can/bcm.c
> +++ b/net/can/bcm.c
[ ... ]
> @@ -1173,6 +1173,10 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
> op->flags |= TX_ANNOUNCE;
> }
>
> + /* add this bcm_op to the list of the tx_ops? */
> + if (add_op_to_list)
> + list_add_rcu(&op->list, &bo->tx_ops);
> +
> if (op->flags & TX_ANNOUNCE)
> bcm_can_tx(op, NULL);
>
[Severity: High]
This is a pre-existing issue, but can the 64-bit ktime_t timer interval
reads tear on 32-bit systems?
In bcm_tx_setup(), the timer interval values are updated under the lock:
net/can/bcm.c:bcm_tx_setup() {
...
if (op->flags & SETTIMER) {
spin_lock_bh(&op->bcm_tx_lock);
...
op->kt_ival1 = bcm_timeval_to_ktime(msg_head->ival1);
op->kt_ival2 = bcm_timeval_to_ktime(msg_head->ival2);
spin_unlock_bh(&op->bcm_tx_lock);
...
}
However, when the timer fires, bcm_tx_timeout_handler() and its callee
bcm_tx_set_expiry() appear to read these 64-bit values locklessly:
net/can/bcm.c:bcm_tx_set_expiry() {
...
if (op->kt_ival1 && op->count)
ival = op->kt_ival1;
else if (op->kt_ival2)
ival = op->kt_ival2;
...
}
If the softirq handler observes a torn read on a 32-bit system, could it
result in corrupted timer expiration values and lead to an immediate
re-arming loop or stalled transmission?
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-bcm_fixes-v11-0-067ba11bec5c@hartkopp.net?part=5
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH v11 06/11] can: bcm: extend bcm_tx_lock usage for data and timer updates
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
` (4 preceding siblings ...)
2026-07-09 17:56 ` [PATCH v11 05/11] can: bcm: add missing rcu list annotations and operations Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 17:56 ` [PATCH v11 07/11] can: bcm: validate frame length in bcm_rx_setup() for RTR replies Oliver Hartkopp via B4 Relay
` (4 subsequent siblings)
10 siblings, 0 replies; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde
From: Oliver Hartkopp <socketcan@hartkopp.net>
Stage new CAN frame content for an existing tx op into a kmalloc()'d
buffer and validate it there, mirroring the approach already used in
bcm_rx_setup(). Only copy the validated data into op->frames while
holding op->bcm_tx_lock, so bcm_can_tx() and bcm_tx_timeout_handler()
can no longer observe a partially updated or unvalidated frame.
Add a missing error path for memcpy_from_msg() when copying CAN frame
data from userspace.
Also move the kt_ival1/kt_ival2/ival1/ival2 updates in bcm_tx_setup()
under op->bcm_tx_lock, and read kt_ival1/kt_ival2/count under the same
lock in bcm_tx_set_expiry() and bcm_tx_timeout_handler(), closing the
torn 64-bit ktime_t read on 32-bit platforms.
Fixes: c2aba69d0c36 ("can: bcm: add locking for bcm_op runtime updates")
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 81 +++++++++++++++++++++++++++++++++++++++++------------------
1 file changed, 57 insertions(+), 24 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index 940cd9df1cca..18601d6fdf76 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -126,11 +126,11 @@ struct bcm_op {
void *last_frames;
struct canfd_frame sframe;
struct canfd_frame last_sframe;
struct sock *sk;
struct net_device *rx_reg_dev;
- spinlock_t bcm_tx_lock; /* protect currframe/count in runtime updates */
+ spinlock_t bcm_tx_lock; /* protect tx data and timer updates */
spinlock_t bcm_rx_update_lock; /* protect filter/timer data updates */
};
struct bcm_sock {
struct sock sk;
@@ -462,16 +462,22 @@ static void bcm_send_to_user(struct bcm_op *op, struct bcm_msg_head *head,
static bool bcm_tx_set_expiry(struct bcm_op *op, struct hrtimer *hrt)
{
ktime_t ival;
+ spin_lock_bh(&op->bcm_tx_lock);
+
if (op->kt_ival1 && op->count)
ival = op->kt_ival1;
- else if (op->kt_ival2)
+ else if (op->kt_ival2) {
ival = op->kt_ival2;
- else
+ } else {
+ spin_unlock_bh(&op->bcm_tx_lock);
return false;
+ }
+
+ spin_unlock_bh(&op->bcm_tx_lock);
hrtimer_set_expires(hrt, ktime_add(ktime_get(), ival));
return true;
}
@@ -484,12 +490,21 @@ static void bcm_tx_start_timer(struct bcm_op *op)
/* bcm_tx_timeout_handler - performs cyclic CAN frame transmissions */
static enum hrtimer_restart bcm_tx_timeout_handler(struct hrtimer *hrtimer)
{
struct bcm_op *op = container_of(hrtimer, struct bcm_op, timer);
struct bcm_msg_head msg_head;
+ bool tx_ival1, tx_ival2;
- if (op->kt_ival1 && (op->count > 0)) {
+ /* snapshot kt_ival1/kt_ival2/count under lock to avoid torn
+ * ktime_t reads racing with concurrent bcm_tx_setup() updates
+ */
+ spin_lock_bh(&op->bcm_tx_lock);
+ tx_ival1 = op->kt_ival1 && (op->count > 0);
+ tx_ival2 = !!op->kt_ival2;
+ spin_unlock_bh(&op->bcm_tx_lock);
+
+ if (tx_ival1) {
bcm_can_tx(op, NULL);
if (!op->count && (op->flags & TX_COUNTEVT)) {
/* create notification to user */
memset(&msg_head, 0, sizeof(msg_head));
@@ -502,11 +517,11 @@ static enum hrtimer_restart bcm_tx_timeout_handler(struct hrtimer *hrtimer)
msg_head.nframes = 0;
bcm_send_to_user(op, &msg_head, NULL, 0);
}
- } else if (op->kt_ival2) {
+ } else if (tx_ival2) {
bcm_can_tx(op, NULL);
}
return bcm_tx_set_expiry(op, &op->timer) ?
HRTIMER_RESTART : HRTIMER_NORESTART;
@@ -1026,64 +1041,80 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
return -EINVAL;
/* check the given can_id */
op = bcm_find_op(&bo->tx_ops, msg_head, ifindex);
if (op) {
+ void *new_frames;
+
/* update existing BCM operation */
/*
* Do we need more space for the CAN frames than currently
* allocated? -> This is a _really_ unusual use-case and
* therefore (complexity / locking) it is not supported.
*/
if (msg_head->nframes > op->nframes)
return -E2BIG;
- /* update CAN frames content */
+ /* get new CAN frames content into a staging buffer before
+ * locking: validate and normalize the frames there so that
+ * bcm_can_tx() / bcm_tx_timeout_handler() never observe a
+ * partially updated or unvalidated frame in op->frames
+ */
+ new_frames = kmalloc(msg_head->nframes * op->cfsiz, GFP_KERNEL);
+ if (!new_frames)
+ return -ENOMEM;
+
for (i = 0; i < msg_head->nframes; i++) {
- cf = op->frames + op->cfsiz * i;
+ cf = new_frames + op->cfsiz * i;
err = memcpy_from_msg((u8 *)cf, msg, op->cfsiz);
+ if (err < 0) {
+ kfree(new_frames);
+ return err;
+ }
if (op->flags & CAN_FD_FRAME) {
if (cf->len > 64)
err = -EINVAL;
} else {
if (cf->len > 8)
err = -EINVAL;
}
- if (err < 0)
+ if (err < 0) {
+ kfree(new_frames);
return err;
+ }
if (msg_head->flags & TX_CP_CAN_ID) {
/* copy can_id into frame */
cf->can_id = msg_head->can_id;
}
}
+
+ spin_lock_bh(&op->bcm_tx_lock);
+
+ /* update CAN frames content */
+ memcpy(op->frames, new_frames, msg_head->nframes * op->cfsiz);
+
op->flags = msg_head->flags;
- /* only lock for unlikely count/nframes/currframe changes */
if (op->nframes != msg_head->nframes ||
- op->flags & TX_RESET_MULTI_IDX ||
- op->flags & SETTIMER) {
-
- spin_lock_bh(&op->bcm_tx_lock);
+ op->flags & TX_RESET_MULTI_IDX) {
+ /* potentially update changed nframes */
+ op->nframes = msg_head->nframes;
+ /* restart multiple frame transmission */
+ op->currframe = 0;
+ }
- if (op->nframes != msg_head->nframes ||
- op->flags & TX_RESET_MULTI_IDX) {
- /* potentially update changed nframes */
- op->nframes = msg_head->nframes;
- /* restart multiple frame transmission */
- op->currframe = 0;
- }
+ if (op->flags & SETTIMER)
+ op->count = msg_head->count;
- if (op->flags & SETTIMER)
- op->count = msg_head->count;
+ spin_unlock_bh(&op->bcm_tx_lock);
- spin_unlock_bh(&op->bcm_tx_lock);
- }
+ kfree(new_frames);
} else {
/* insert new BCM operation for the given can_id */
op = kzalloc(OPSIZ, GFP_KERNEL);
@@ -1155,14 +1186,16 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
} /* if ((op = bcm_find_op(&bo->tx_ops, msg_head->can_id, ifindex))) */
if (op->flags & SETTIMER) {
/* set timer values */
+ spin_lock_bh(&op->bcm_tx_lock);
op->ival1 = msg_head->ival1;
op->ival2 = msg_head->ival2;
op->kt_ival1 = bcm_timeval_to_ktime(msg_head->ival1);
op->kt_ival2 = bcm_timeval_to_ktime(msg_head->ival2);
+ spin_unlock_bh(&op->bcm_tx_lock);
/* disable an active timer due to zero values? */
if (!op->kt_ival1 && !op->kt_ival2)
hrtimer_cancel(&op->timer);
}
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* [PATCH v11 07/11] can: bcm: validate frame length in bcm_rx_setup() for RTR replies
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
` (5 preceding siblings ...)
2026-07-09 17:56 ` [PATCH v11 06/11] can: bcm: extend bcm_tx_lock usage for data and timer updates Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 17:56 ` [PATCH v11 08/11] can: bcm: add missing device refcount for CAN filter removal Oliver Hartkopp via B4 Relay
` (3 subsequent siblings)
10 siblings, 0 replies; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde
From: Oliver Hartkopp <socketcan@hartkopp.net>
bcm_tx_setup() validates cf->len against the CAN/CAN FD DLC limits
before installing frames for TX_SETUP, but bcm_rx_setup() never did
the same for the RTR-reply frame configured via RX_SETUP with
RX_RTR_FRAME.
Fixes: ffd980f976e7 ("[CAN]: Add broadcast manager (bcm) protocol")
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 59 +++++++++++++++++++++++++++++++++++++++++------------------
1 file changed, 41 insertions(+), 18 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index 18601d6fdf76..3cc3deb98a98 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -1223,26 +1223,41 @@ static int bcm_tx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
kfree(op->frames);
kfree(op);
return err;
}
-static void bcm_rx_setup_rtr_check(struct bcm_msg_head *msg_head,
- struct bcm_op *op, void *new_frames)
+static int bcm_rx_setup_rtr_check(struct bcm_msg_head *msg_head,
+ struct bcm_op *op, void *new_frames)
{
+ struct canfd_frame *frame0 = new_frames;
+
+ if (!(msg_head->flags & RX_RTR_FRAME))
+ return 0;
+
+ /* this frame is sent out as-is by bcm_can_tx() whenever a matching
+ * remote request is received, so validate its length the same way
+ * bcm_tx_setup() validates TX_SETUP frames before installing it
+ */
+ if (msg_head->flags & CAN_FD_FRAME) {
+ if (frame0->len > 64)
+ return -EINVAL;
+ } else {
+ if (frame0->len > 8)
+ return -EINVAL;
+ }
+
/* funny feature in RX(!)_SETUP only for RTR-mode:
* copy can_id into frame BUT without RTR-flag to
* prevent a full-load-loopback-test ... ;-]
* normalize this on the staged buffer, before it is
* ever installed into op->frames.
*/
- if (msg_head->flags & RX_RTR_FRAME) {
- struct canfd_frame *frame0 = new_frames;
+ if ((msg_head->flags & TX_CP_CAN_ID) ||
+ frame0->can_id == op->can_id)
+ frame0->can_id = op->can_id & ~CAN_RTR_FLAG;
- if ((msg_head->flags & TX_CP_CAN_ID) ||
- frame0->can_id == op->can_id)
- frame0->can_id = op->can_id & ~CAN_RTR_FLAG;
- }
+ return 0;
}
/*
* bcm_rx_setup - create or update a bcm rx op (for bcm_sendmsg)
*/
@@ -1301,11 +1316,15 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
if (err < 0) {
kfree(new_frames);
return err;
}
- bcm_rx_setup_rtr_check(msg_head, op, new_frames);
+ err = bcm_rx_setup_rtr_check(msg_head, op, new_frames);
+ if (err < 0) {
+ kfree(new_frames);
+ return err;
+ }
}
spin_lock_bh(&op->bcm_rx_update_lock);
op->nframes = msg_head->nframes;
op->flags = msg_head->flags;
@@ -1366,20 +1385,16 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
}
if (msg_head->nframes) {
err = memcpy_from_msg(op->frames, msg,
msg_head->nframes * op->cfsiz);
- if (err < 0) {
- if (op->frames != &op->sframe)
- kfree(op->frames);
- if (op->last_frames != &op->last_sframe)
- kfree(op->last_frames);
- kfree(op);
- return err;
- }
+ if (err < 0)
+ goto free_op;
- bcm_rx_setup_rtr_check(msg_head, op, op->frames);
+ err = bcm_rx_setup_rtr_check(msg_head, op, op->frames);
+ if (err < 0)
+ goto free_op;
}
/* bcm_can_tx / bcm_tx_timeout_handler needs this */
op->sk = sk;
sock_hold(sk);
@@ -1472,10 +1487,18 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
/* add this bcm_op to the list of the rx_ops */
list_add_rcu(&op->list, &bo->rx_ops);
}
return msg_head->nframes * op->cfsiz + MHSIZ;
+
+free_op:
+ if (op->frames != &op->sframe)
+ kfree(op->frames);
+ if (op->last_frames != &op->last_sframe)
+ kfree(op->last_frames);
+ kfree(op);
+ return err;
}
/*
* bcm_tx_send - send a single CAN frame to the CAN interface (for bcm_sendmsg)
*/
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* [PATCH v11 08/11] can: bcm: add missing device refcount for CAN filter removal
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
` (6 preceding siblings ...)
2026-07-09 17:56 ` [PATCH v11 07/11] can: bcm: validate frame length in bcm_rx_setup() for RTR replies Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 17:56 ` [PATCH v11 09/11] can: bcm: fix stale rx/tx ops after device removal Oliver Hartkopp via B4 Relay
` (2 subsequent siblings)
10 siblings, 0 replies; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde, sashiko-bot
From: Oliver Hartkopp <socketcan@hartkopp.net>
sashiko-bot remarked a problem with a concurrent device unregistration
in isotp.c which also is present in the bcm.c code. A former fix for raw.c
commit c275a176e4b6 ("can: raw: add missing refcount for memory leak fix")
introduced a netdevice_tracker which solves the issue for bcm.c too.
bcm_release(), bcm_delete_rx_op() and bcm_notifier() relied on
dev_get_by_index(ifindex) to re-find the device for an rx_op before
unregistering its filter. If a concurrent NETDEV_UNREGISTER has already
unlisted the device from the ifindex table, that lookup fails and
can_rx_unregister() is silently skipped, leaving a stale CAN filter
pointing at the soon-to-be-freed bcm_op/socket.
Hold a netdev_hold()/netdev_put() tracked reference on op->rx_reg_dev
from the moment the rx filter is registered in bcm_rx_setup() until it
is unregistered in bcm_rx_unreg(), and use that reference directly in
bcm_release() and bcm_delete_rx_op() instead of re-looking the device
up by ifindex.
Reported-by: sashiko-bot@kernel.org
Link: https://sashiko.dev/#/patchset/20260707094716.63578-1-socketcan@hartkopp.net
Fixes: ffd980f976e7 ("[CAN]: Add broadcast manager (bcm) protocol")
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 47 +++++++++++++++++++++++++++--------------------
1 file changed, 27 insertions(+), 20 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index 3cc3deb98a98..5660165cdf0e 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -126,10 +126,11 @@ struct bcm_op {
void *last_frames;
struct canfd_frame sframe;
struct canfd_frame last_sframe;
struct sock *sk;
struct net_device *rx_reg_dev;
+ netdevice_tracker rx_reg_dev_tracker;
spinlock_t bcm_tx_lock; /* protect tx data and timer updates */
spinlock_t bcm_rx_update_lock; /* protect filter/timer data updates */
};
struct bcm_sock {
@@ -914,10 +915,11 @@ static void bcm_rx_unreg(struct net_device *dev, struct bcm_op *op)
can_rx_unregister(dev_net(dev), dev, op->can_id,
REGMASK(op->can_id), bcm_rx_handler, op);
/* mark as removed subscription */
op->rx_reg_dev = NULL;
+ netdev_put(dev, &op->rx_reg_dev_tracker);
} else
printk(KERN_ERR "can-bcm: bcm_rx_unreg: registered device "
"mismatch %p %p\n", op->rx_reg_dev, dev);
}
@@ -944,21 +946,18 @@ static int bcm_delete_rx_op(struct list_head *ops, struct bcm_msg_head *mh,
if (op->ifindex) {
/*
* Only remove subscriptions that had not
* been removed due to NETDEV_UNREGISTER
* in bcm_notifier()
+ *
+ * op->rx_reg_dev is a tracked reference taken
+ * when the subscription was registered, so it
+ * stays valid here even if a concurrent
+ * NETDEV_UNREGISTER already unlisted the dev.
*/
- if (op->rx_reg_dev) {
- struct net_device *dev;
-
- dev = dev_get_by_index(sock_net(op->sk),
- op->ifindex);
- if (dev) {
- bcm_rx_unreg(dev, op);
- dev_put(dev);
- }
- }
+ if (op->rx_reg_dev)
+ bcm_rx_unreg(op->rx_reg_dev, op);
} else
can_rx_unregister(sock_net(op->sk), NULL,
op->can_id,
REGMASK(op->can_id),
bcm_rx_handler, op);
@@ -1463,11 +1462,21 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
op->can_id,
REGMASK(op->can_id),
bcm_rx_handler, op,
"bcm", sk);
- op->rx_reg_dev = dev;
+ /* keep a tracked reference so that a later
+ * unregister can safely reach the device even
+ * if a concurrent NETDEV_UNREGISTER has
+ * already unlisted it by ifindex
+ */
+ if (!err) {
+ op->rx_reg_dev = dev;
+ netdev_hold(dev,
+ &op->rx_reg_dev_tracker,
+ GFP_KERNEL);
+ }
dev_put(dev);
} else {
/* the requested device is gone - do not
* silently succeed without registering
*/
@@ -1840,20 +1849,18 @@ static int bcm_release(struct socket *sock)
if (op->ifindex) {
/*
* Only remove subscriptions that had not
* been removed due to NETDEV_UNREGISTER
* in bcm_notifier()
+ *
+ * op->rx_reg_dev is a tracked reference taken
+ * when the subscription was registered, so it
+ * stays valid here even if a concurrent
+ * NETDEV_UNREGISTER already unlisted the device.
*/
- if (op->rx_reg_dev) {
- struct net_device *dev;
-
- dev = dev_get_by_index(net, op->ifindex);
- if (dev) {
- bcm_rx_unreg(dev, op);
- dev_put(dev);
- }
- }
+ if (op->rx_reg_dev)
+ bcm_rx_unreg(op->rx_reg_dev, op);
} else
can_rx_unregister(net, NULL, op->can_id,
REGMASK(op->can_id),
bcm_rx_handler, op);
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* [PATCH v11 09/11] can: bcm: fix stale rx/tx ops after device removal
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
` (7 preceding siblings ...)
2026-07-09 17:56 ` [PATCH v11 08/11] can: bcm: add missing device refcount for CAN filter removal Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 18:39 ` sashiko-bot
2026-07-09 17:56 ` [PATCH v11 10/11] can: bcm: fix data race on rx_stamp/rx_ifindex in bcm_rx_handler() Oliver Hartkopp via B4 Relay
2026-07-09 17:56 ` [PATCH v11 11/11] can: bcm: track a single source interface for ANYDEV timeout/throttle ops Oliver Hartkopp via B4 Relay
10 siblings, 1 reply; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde, sashiko-bot
From: Oliver Hartkopp <socketcan@hartkopp.net>
RX: an RX_SETUP update(!) for an existing op skipped can_rx_register()
unconditionally, even when a concurrent NETDEV_UNREGISTER had already
torn down its registration (op->rx_reg_dev == NULL). This silently
did not re-enable frame delivery for that updated filter. bcm_rx_setup()
now re-registers in that case, while leaving rx_ops with ifindex = 0
(all CAN devices) which never carry a tracked rx_reg_dev registered as-is.
TX: bcm_notify() only handled bo->rx_ops on NETDEV_UNREGISTER, leaving
tx_ops with an active cyclic transmission re-arming its hrtimer
indefinitely to execute bcm_tx_timeout_handler(). Cancelling the hrtimer
prevents the runaway timer and any injection into a later reused ifindex,
since nothing else calls bcm_can_tx() for the op until an explicit
TX_SETUP update re-arms it.
Unlike bcm_rx_unreg(), which clears the tracked rx_reg_dev for rx_ops,
the ifindex is intentionally left unchanged for tx_ops. bcm_tx_setup()
always rejects ifindex 0, so clearing it would strand the op: neither a
later TX_SETUP (bcm_find_op()) nor TX_DELETE (bcm_delete_tx_op()) could
ever find it again, since both require an exact ifindex match.
Reported-by: sashiko-bot@kernel.org
Link: https://lore.kernel.org/linux-can/20260708094536.DDF821F00A3A@smtp.kernel.org/
Link: https://lore.kernel.org/linux-can/20260708154039.347ED1F000E9@smtp.kernel.org/
Fixes: ffd980f976e7 ("[CAN]: Add broadcast manager (bcm) protocol")
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 47 ++++++++++++++++++++++++++++++++++++++---------
1 file changed, 38 insertions(+), 9 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index 5660165cdf0e..fbf1cbb7c99f 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -1264,10 +1264,11 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
int ifindex, struct sock *sk)
{
struct bcm_sock *bo = bcm_sk(sk);
struct bcm_op *op;
int do_rx_register;
+ int new_op = 0;
int err = 0;
if ((msg_head->flags & RX_FILTER_ID) || (!(msg_head->nframes))) {
/* be robust against wrong usage ... */
msg_head->flags |= RX_FILTER_ID;
@@ -1340,12 +1341,19 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
spin_unlock_bh(&op->bcm_rx_update_lock);
/* free temporary frames / kfree(NULL) is safe */
kfree(new_frames);
- /* Only an update -> do not call can_rx_register() */
- do_rx_register = 0;
+ /* Don't register a new CAN filter for the rx_op update unless
+ * a concurrent NETDEV_UNREGISTER notifier already tore down
+ * the previous registration. In this case the receiver needs
+ * to be re-registered here so that this update doesn't
+ * silently stop delivering frames for the given ifindex.
+ * Ops with ifindex = 0 (all CAN interfaces) never carry a
+ * tracked rx_reg_dev and stay registered as-is.
+ */
+ do_rx_register = (ifindex && !op->rx_reg_dev) ? 1 : 0;
} else {
/* insert new BCM operation for the given can_id */
op = kzalloc(OPSIZ, GFP_KERNEL);
if (!op)
@@ -1408,10 +1416,11 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
hrtimer_setup(&op->thrtimer, bcm_rx_thr_handler, CLOCK_MONOTONIC,
HRTIMER_MODE_REL_SOFT);
/* call can_rx_register() */
do_rx_register = 1;
+ new_op = 1;
} /* if ((op = bcm_find_op(&bo->rx_ops, msg_head->can_id, ifindex))) */
/* check flags */
@@ -1449,11 +1458,14 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
if ((op->flags & STARTTIMER) && op->kt_ival1)
hrtimer_start(&op->timer, op->kt_ival1,
HRTIMER_MODE_REL_SOFT);
}
- /* now we can register for can_ids, if we added a new bcm_op */
+ /* now we can register for can_ids, if we added a new bcm_op
+ * or need to re-register after a NETDEV_UNREGISTER tore down
+ * the previous registration of an existing op
+ */
if (do_rx_register) {
if (ifindex) {
struct net_device *dev;
dev = dev_get_by_index(sock_net(sk), ifindex);
@@ -1481,22 +1493,31 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
* silently succeed without registering
*/
err = -ENODEV;
}
- } else
+ } else {
err = can_rx_register(sock_net(sk), NULL, op->can_id,
REGMASK(op->can_id),
bcm_rx_handler, op, "bcm", sk);
+ }
+
if (err) {
- /* this bcm rx op is broken -> remove it */
- bcm_remove_op(op);
+ /* newly created bcm rx op is broken -> remove it */
+ if (new_op)
+ bcm_remove_op(op);
+
+ /* an existing op just stays unregistered until the
+ * next successful RX_SETUP - it is already part of
+ * bo->rx_ops and must not be freed here
+ */
return err;
}
- /* add this bcm_op to the list of the rx_ops */
- list_add_rcu(&op->list, &bo->rx_ops);
+ /* add a new bcm_op to the list of the rx_ops */
+ if (new_op)
+ list_add_rcu(&op->list, &bo->rx_ops);
}
return msg_head->nframes * op->cfsiz + MHSIZ;
free_op:
@@ -1712,15 +1733,23 @@ static void bcm_notify(struct bcm_sock *bo, unsigned long msg,
switch (msg) {
case NETDEV_UNREGISTER:
lock_sock(sk);
- /* remove device specific receive entries */
+ /* rx_ops: remove device specific receive entries */
list_for_each_entry(op, &bo->rx_ops, list)
if (op->rx_reg_dev == dev)
bcm_rx_unreg(dev, op);
+ /* tx_ops: stop device specific cyclic transmissions on the
+ * vanishing ifindex. Cancelling the timer is enough to stop
+ * cyclic bcm_can_tx() calls as there is no re-arming.
+ */
+ list_for_each_entry(op, &bo->tx_ops, list)
+ if (op->ifindex == dev->ifindex)
+ hrtimer_cancel(&op->timer);
+
/* remove device reference, if this is our bound device */
if (bo->bound && bo->ifindex == dev->ifindex) {
#if IS_ENABLED(CONFIG_PROC_FS)
if (sock_net(sk)->can.bcmproc_dir && bo->bcm_proc_read) {
remove_proc_entry(bo->procname, sock_net(sk)->can.bcmproc_dir);
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH v11 09/11] can: bcm: fix stale rx/tx ops after device removal
2026-07-09 17:56 ` [PATCH v11 09/11] can: bcm: fix stale rx/tx ops after device removal Oliver Hartkopp via B4 Relay
@ 2026-07-09 18:39 ` sashiko-bot
0 siblings, 0 replies; 20+ messages in thread
From: sashiko-bot @ 2026-07-09 18:39 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: Marc Kleine-Budde, Oleksij Rempel, Oliver Hartkopp,
Vincent Mailhol, linux-can
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Data race on op->rx_stamp and op->rx_ifindex in bcm_rx_handler causes incorrect metadata to be delivered to userspace.
--
--- Patch [9]: [PATCH v11 09/11] can: bcm: fix stale rx/tx ops after device removal ---
commit ef4eaedce01b041fb9a737217fc61cc4150a0e37
Author: Oliver Hartkopp <socketcan@hartkopp.net>
can: bcm: fix stale rx/tx ops after device removal
This commit fixes issues with stale rx and tx operations after a CAN device
is removed. It ensures rx operations are re-registered if they were torn
down by a concurrent NETDEV_UNREGISTER, and cancels the hrtimer for tx
operations to prevent runaway timers.
Link: https://lore.kernel.org/linux-can/20260708094536.DDF821F00A3A@smtp.kernel.org/
Link: https://lore.kernel.org/linux-can/20260708154039.347ED1F000E9@smtp.kernel.org/
> diff --git a/net/can/bcm.c b/net/can/bcm.c
> index 5660165cdf0eb..fbf1cbb7c99fe 100644
> --- a/net/can/bcm.c
> +++ b/net/can/bcm.c
[ ... ]
> @@ -1714,11 +1735,19 @@ static void bcm_notify(struct bcm_sock *bo, unsigned long msg,
> case NETDEV_UNREGISTER:
> lock_sock(sk);
>
> - /* remove device specific receive entries */
> + /* rx_ops: remove device specific receive entries */
> list_for_each_entry(op, &bo->rx_ops, list)
> if (op->rx_reg_dev == dev)
> bcm_rx_unreg(dev, op);
[Severity: High]
This is a pre-existing issue, but is there a data race on op->rx_stamp and
op->rx_ifindex in bcm_rx_handler() that causes incorrect metadata to be
delivered to userspace?
If a CAN frame matching the same ANYDEV filter is received concurrently on
different interfaces, bcm_rx_handler() can execute simultaneously on
different CPUs. Could both CPUs write to the shared op->rx_stamp and
op->rx_ifindex without holding a lock?
net/can/bcm.c:bcm_rx_handler() {
/* save rx timestamp */
op->rx_stamp = skb->tstamp;
/* save originator for recvfrom() */
op->rx_ifindex = skb->dev->ifindex;
/* op->flags/op->frames may be updated concurrently by bcm_rx_setup() */
spin_lock_bh(&op->bcm_rx_update_lock);
}
When CPU 1 subsequently acquires op->bcm_rx_update_lock and calls
bcm_send_to_user(), will it read the values overwritten by CPU 2?
net/can/bcm.c:bcm_send_to_user() {
if (has_timestamp) {
/* restore rx timestamp */
skb->tstamp = op->rx_stamp;
}
/* ... */
addr = (struct sockaddr_can *)skb->cb;
memset(addr, 0, sizeof(*addr));
addr->can_family = AF_CAN;
addr->can_ifindex = op->rx_ifindex;
}
For ANYDEV sockets, the ifindex returned in sockaddr_can is the sole source
of truth for identifying the origin bus, and delivering the wrong ifindex
can cause userspace gateway applications to misroute CAN frames.
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-bcm_fixes-v11-0-067ba11bec5c@hartkopp.net?part=9
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH v11 10/11] can: bcm: fix data race on rx_stamp/rx_ifindex in bcm_rx_handler()
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
` (8 preceding siblings ...)
2026-07-09 17:56 ` [PATCH v11 09/11] can: bcm: fix stale rx/tx ops after device removal Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 18:38 ` sashiko-bot
2026-07-09 17:56 ` [PATCH v11 11/11] can: bcm: track a single source interface for ANYDEV timeout/throttle ops Oliver Hartkopp via B4 Relay
10 siblings, 1 reply; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde, sashiko-bot
From: Oliver Hartkopp <socketcan@hartkopp.net>
For an rx op subscribed on all interfaces (ifindex == 0), the same op
is registered once in the shared per-netns wildcard filter list, so
bcm_rx_handler() can run concurrently on different CPUs for frames
arriving on different net devices.
op->rx_stamp and op->rx_ifindex were written before bcm_rx_update_lock was
taken, allowing concurrent writers to race each other - including a torn
store of the 64-bit rx_stamp on 32-bit platforms.
Beyond a torn store bcm_send_to_user() must report the timestamp/ifindex
of the very same frame whose content it is delivering. So the assignment
is placed in the same unbroken bcm_rx_update_lock section as the content
comparison.
As a side effect, the RTR-request frame feature (which never reach
bcm_send_to_user()) no longer updates rx_stamp/rx_ifindex, since only
the notification path needs them.
Fixes: ffd980f976e7 ("[CAN]: Add broadcast manager (bcm) protocol")
Reported-by: sashiko-bot@kernel.org
Link: https://lore.kernel.org/linux-can/20260707145135.5BC831F00A3A@smtp.kernel.org/
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 13 ++++++++-----
1 file changed, 8 insertions(+), 5 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index fbf1cbb7c99f..10a883b50fc6 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -777,15 +777,10 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
}
/* disable timeout */
hrtimer_cancel(&op->timer);
- /* save rx timestamp */
- op->rx_stamp = skb->tstamp;
- /* save originator for recvfrom() */
- op->rx_ifindex = skb->dev->ifindex;
-
/* op->flags/op->frames may be updated concurrently by bcm_rx_setup() */
spin_lock_bh(&op->bcm_rx_update_lock);
rtr_frame = op->flags & RX_RTR_FRAME;
if (rtr_frame) {
@@ -813,10 +808,18 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
traffic_flags |= RX_LOCAL;
if (skb->sk == op->sk)
traffic_flags |= RX_OWN;
}
+ /* save rx timestamp and originator for recvfrom() under lock.
+ * For an op subscribed on all interfaces (ifindex == 0)
+ * bcm_rx_handler() can run concurrently on different CPUs so
+ * the CAN content and the meta data must be bundled correctly.
+ */
+ op->rx_stamp = skb->tstamp;
+ op->rx_ifindex = skb->dev->ifindex;
+
if (op->flags & RX_FILTER_ID) {
/* the easiest case */
bcm_rx_update_and_send(op, op->last_frames, rxframe,
traffic_flags);
goto rx_starttimer;
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH v11 10/11] can: bcm: fix data race on rx_stamp/rx_ifindex in bcm_rx_handler()
2026-07-09 17:56 ` [PATCH v11 10/11] can: bcm: fix data race on rx_stamp/rx_ifindex in bcm_rx_handler() Oliver Hartkopp via B4 Relay
@ 2026-07-09 18:38 ` sashiko-bot
0 siblings, 0 replies; 20+ messages in thread
From: sashiko-bot @ 2026-07-09 18:38 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: Vincent Mailhol, Marc Kleine-Budde, linux-can, Oleksij Rempel,
Oliver Hartkopp
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Unconditional overwrite of `op->rx_stamp` corrupts timestamp metadata for throttled or multiplexed CAN frames.
--
--- Patch [10]: [PATCH v11 10/11] can: bcm: fix data race on rx_stamp/rx_ifindex in bcm_rx_handler() ---
commit 962a73c290cc743e31470d3c9ed820b166d58a79
Author: Oliver Hartkopp <socketcan@hartkopp.net>
can: bcm: fix data race on rx_stamp/rx_ifindex in bcm_rx_handler()
This commit moves the update of rx_stamp and rx_ifindex under the
bcm_rx_update_lock to prevent concurrent writers from racing each
other, avoiding torn stores on 32-bit platforms. It also ensures
metadata is bundled correctly with the frame content delivered to
userspace.
Link: https://lore.kernel.org/linux-can/20260707145135.5BC831F00A3A@smtp.kernel.org/
> diff --git a/net/can/bcm.c b/net/can/bcm.c
> index fbf1cbb7c99fe..10a883b50fc63 100644
> --- a/net/can/bcm.c
> +++ b/net/can/bcm.c
[ ... ]
> @@ -815,6 +810,14 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
> traffic_flags |= RX_OWN;
> }
>
> + /* save rx timestamp and originator for recvfrom() under lock.
> + * For an op subscribed on all interfaces (ifindex == 0)
> + * bcm_rx_handler() can run concurrently on different CPUs so
> + * the CAN content and the meta data must be bundled correctly.
> + */
> + op->rx_stamp = skb->tstamp;
> + op->rx_ifindex = skb->dev->ifindex;
[Severity: High]
This is a pre-existing issue, but does this unconditional update of
op->rx_stamp corrupt the timestamp metadata for throttled or multiplexed
CAN frames?
The commit message states that bcm_send_to_user() must report the timestamp
and ifindex of the very same frame whose content it is delivering. However, if
a BCM RX op is configured with a throttle timer, a CAN frame might match a
mask and get queued by the timer. If a second CAN frame arrives on the same
interface, it will unconditionally overwrite the shared op->rx_stamp here
before checking if the frame matches any multiplex index or if it will be
throttled.
When the throttle timer later flushes the first frame:
net/can/bcm.c:bcm_rx_thr_flush() {
...
for (i = 1; i < op->nframes; i++)
updated += bcm_rx_do_flush(op, i);
...
}
will all notifications use the single, overwritten op->rx_stamp from the
subsequent frame, rather than the timestamp of the frame actually being
delivered?
> +
> if (op->flags & RX_FILTER_ID) {
> /* the easiest case */
> bcm_rx_update_and_send(op, op->last_frames, rxframe,
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-bcm_fixes-v11-0-067ba11bec5c@hartkopp.net?part=10
^ permalink raw reply [flat|nested] 20+ messages in thread
* [PATCH v11 11/11] can: bcm: track a single source interface for ANYDEV timeout/throttle ops
2026-07-09 17:56 [PATCH v11 00/11] can: bcm: collected fixes - final6 \o/ Oliver Hartkopp via B4 Relay
` (9 preceding siblings ...)
2026-07-09 17:56 ` [PATCH v11 10/11] can: bcm: fix data race on rx_stamp/rx_ifindex in bcm_rx_handler() Oliver Hartkopp via B4 Relay
@ 2026-07-09 17:56 ` Oliver Hartkopp via B4 Relay
2026-07-09 18:43 ` sashiko-bot
10 siblings, 1 reply; 20+ messages in thread
From: Oliver Hartkopp via B4 Relay @ 2026-07-09 17:56 UTC (permalink / raw)
To: linux-can; +Cc: Oliver Hartkopp, Marc Kleine-Budde, sashiko-bot
From: Oliver Hartkopp <socketcan@hartkopp.net>
An ANYDEV rx op (ifindex == 0) with an active RX timeout and/or
throttle timer has no defined semantics when matching frames arrive
from several interfaces: bcm_rx_handler() can run concurrently for
the same op on different CPUs, racing hrtimer_cancel()/
bcm_rx_starttimer() against bcm_rx_timeout_handler() and causing
spurious RX_TIMEOUT notifications and last_frames corruption. The
same concurrency lets throttled multiplex frames from different
interfaces clobber the single rx_ifindex/rx_stamp fields shared by
the op.
Add op->if_detected to track the first interface that delivers a
matching frame while a timeout/throttle timer is configured, and
reject frames from any other interface for that op. The claim is
decided in bcm_rx_handler() before hrtimer_cancel() touches
op->timer, so a rejected frame can never disturb the claimed
interface's watchdog. RTR-mode ops are excluded via RX_RTR_FRAME,
independent of kt_ival1/kt_ival2, since those may briefly hold a
stale value from an earlier non-RTR configuration.
The claim is released in bcm_notify() on NETDEV_UNREGISTER and in
bcm_rx_setup() when SETTIMER reconfigures the timer values.
Fixes: ffd980f976e7 ("[CAN]: Add broadcast manager (bcm) protocol")
Reported-by: sashiko-bot@kernel.org
Link: https://lore.kernel.org/linux-can/20260709105031.1A39C1F000E9@smtp.kernel.org/
Signed-off-by: Oliver Hartkopp <socketcan@hartkopp.net>
---
net/can/bcm.c | 44 +++++++++++++++++++++++++++++++++++++++-----
1 file changed, 39 insertions(+), 5 deletions(-)
diff --git a/net/can/bcm.c b/net/can/bcm.c
index 10a883b50fc6..8cc0d2267582 100644
--- a/net/can/bcm.c
+++ b/net/can/bcm.c
@@ -115,10 +115,11 @@ struct bcm_op {
atomic_long_t frames_abs, frames_filtered;
struct bcm_timeval ival1, ival2;
struct hrtimer timer, thrtimer;
ktime_t rx_stamp, kt_ival1, kt_ival2, kt_lastmsg;
int rx_ifindex;
+ int if_detected; /* first received ifindex in ANYDEV rx_op mode */
int cfsiz;
u32 count;
u32 nframes;
u32 currframe;
/* void pointers to arrays of struct can[fd]_frame */
@@ -774,10 +775,32 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
} else {
if (!can_is_can_skb(skb))
return;
}
+ /* An ANYDEV op with an active RX timeout and/or throttle timer
+ * tracks a single source interface: claim the first interface that
+ * delivers a matching frame and reject frames from any other one,
+ * before hrtimer_cancel() below can touch op->timer - this avoids
+ * racing bcm_rx_timeout_handler() across concurrent interfaces.
+ * RX_RTR_FRAME ops are excluded, as kt_ival1/kt_ival2 may briefly
+ * hold a stale value from an earlier non-RTR configuration.
+ */
+ if (!op->ifindex && !(op->flags & RX_RTR_FRAME) &&
+ (op->kt_ival1 || op->kt_ival2)) {
+ bool foreign;
+
+ spin_lock_bh(&op->bcm_rx_update_lock);
+ if (!op->if_detected)
+ op->if_detected = skb->dev->ifindex;
+ foreign = op->if_detected != skb->dev->ifindex;
+ spin_unlock_bh(&op->bcm_rx_update_lock);
+
+ if (foreign)
+ return;
+ }
+
/* disable timeout */
hrtimer_cancel(&op->timer);
/* op->flags/op->frames may be updated concurrently by bcm_rx_setup() */
spin_lock_bh(&op->bcm_rx_update_lock);
@@ -808,14 +831,13 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
traffic_flags |= RX_LOCAL;
if (skb->sk == op->sk)
traffic_flags |= RX_OWN;
}
- /* save rx timestamp and originator for recvfrom() under lock.
- * For an op subscribed on all interfaces (ifindex == 0)
- * bcm_rx_handler() can run concurrently on different CPUs so
- * the CAN content and the meta data must be bundled correctly.
+ /* save rx timestamp and originator for recvfrom() under lock: an
+ * ANYDEV op without an active timer can still run concurrently on
+ * different CPUs, so content and meta data must be bundled here.
*/
op->rx_stamp = skb->tstamp;
op->rx_ifindex = skb->dev->ifindex;
if (op->flags & RX_FILTER_ID) {
@@ -1439,10 +1461,11 @@ static int bcm_rx_setup(struct bcm_msg_head *msg_head, struct msghdr *msg,
op->ival1 = msg_head->ival1;
op->ival2 = msg_head->ival2;
op->kt_ival1 = bcm_timeval_to_ktime(msg_head->ival1);
op->kt_ival2 = bcm_timeval_to_ktime(msg_head->ival2);
op->kt_lastmsg = 0;
+ op->if_detected = 0; /* reclaim ifindex in ANYDEV mode */
spin_unlock_bh(&op->bcm_rx_update_lock);
/* disable an active timer due to zero value? */
if (!op->kt_ival1)
hrtimer_cancel(&op->timer);
@@ -1737,14 +1760,25 @@ static void bcm_notify(struct bcm_sock *bo, unsigned long msg,
case NETDEV_UNREGISTER:
lock_sock(sk);
/* rx_ops: remove device specific receive entries */
- list_for_each_entry(op, &bo->rx_ops, list)
+ list_for_each_entry(op, &bo->rx_ops, list) {
if (op->rx_reg_dev == dev)
bcm_rx_unreg(dev, op);
+ /* release an ANYDEV op's claim (see bcm_rx_handler())
+ * on this now confirmed-gone interface.
+ */
+ if (!op->ifindex) {
+ spin_lock_bh(&op->bcm_rx_update_lock);
+ if (op->if_detected == dev->ifindex)
+ op->if_detected = 0;
+ spin_unlock_bh(&op->bcm_rx_update_lock);
+ }
+ }
+
/* tx_ops: stop device specific cyclic transmissions on the
* vanishing ifindex. Cancelling the timer is enough to stop
* cyclic bcm_can_tx() calls as there is no re-arming.
*/
list_for_each_entry(op, &bo->tx_ops, list)
--
2.53.0
^ permalink raw reply related [flat|nested] 20+ messages in thread* Re: [PATCH v11 11/11] can: bcm: track a single source interface for ANYDEV timeout/throttle ops
2026-07-09 17:56 ` [PATCH v11 11/11] can: bcm: track a single source interface for ANYDEV timeout/throttle ops Oliver Hartkopp via B4 Relay
@ 2026-07-09 18:43 ` sashiko-bot
0 siblings, 0 replies; 20+ messages in thread
From: sashiko-bot @ 2026-07-09 18:43 UTC (permalink / raw)
To: Oliver Hartkopp
Cc: Oliver Hartkopp, linux-can, Vincent Mailhol, Oleksij Rempel,
Marc Kleine-Budde
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] Race condition between interface unregistration and in-flight packets permanently breaks ANYDEV sockets.
- [Medium] Data race and torn read on 64-bit ktime_t variables bypassing ANYDEV concurrency protection.
--
--- Patch [11]: [PATCH v11 11/11] can: bcm: track a single source interface for ANYDEV timeout/throttle ops ---
commit ed6fe9660233297822929009c213120acb00229f
Author: Oliver Hartkopp <socketcan@hartkopp.net>
can: bcm: track a single source interface for ANYDEV timeout/throttle ops
An ANYDEV rx op with an active RX timeout/throttle timer has no defined
semantics when matching frames arrive from multiple interfaces, leading to
data corruption. This patch adds op->if_detected to track the first interface
that delivers a matching frame, rejecting frames from other interfaces.
The claim is released on NETDEV_UNREGISTER or timer reconfiguration.
Link: https://lore.kernel.org/linux-can/20260709105031.1A39C1F000E9@smtp.kernel.org/
> diff --git a/net/can/bcm.c b/net/can/bcm.c
> index 10a883b50fc63..8cc0d2267582f 100644
> --- a/net/can/bcm.c
> +++ b/net/can/bcm.c
[ ... ]
> @@ -776,6 +777,28 @@ static void bcm_rx_handler(struct sk_buff *skb, void *data)
> return;
> }
>
> + /* An ANYDEV op with an active RX timeout and/or throttle timer
> + * tracks a single source interface: claim the first interface that
> + * delivers a matching frame and reject frames from any other one,
> + * before hrtimer_cancel() below can touch op->timer - this avoids
> + * racing bcm_rx_timeout_handler() across concurrent interfaces.
> + * RX_RTR_FRAME ops are excluded, as kt_ival1/kt_ival2 may briefly
> + * hold a stale value from an earlier non-RTR configuration.
> + */
> + if (!op->ifindex && !(op->flags & RX_RTR_FRAME) &&
> + (op->kt_ival1 || op->kt_ival2)) {
[Severity: Medium]
Can the lockless read of these 64-bit ktime_t variables cause a torn read on
32-bit platforms?
If the torn read evaluates to zero while the timers are actually configured,
wouldn't it bypass the interface-tracking logic and allow concurrent frames
to clobber shared metadata?
> + bool foreign;
> +
> + spin_lock_bh(&op->bcm_rx_update_lock);
> + if (!op->if_detected)
> + op->if_detected = skb->dev->ifindex;
[Severity: High]
Can an in-flight packet spuriously re-claim the stale ifindex here?
In bcm_notify(), the claim is released when an interface goes down. Since
bcm_notify() executes before synchronize_net(), could an already dequeued
packet arrive here, see op->if_detected == 0, and re-claim the dead
interface's index?
This seems like it would permanently block future packets from other valid
interfaces.
> + foreign = op->if_detected != skb->dev->ifindex;
> + spin_unlock_bh(&op->bcm_rx_update_lock);
> +
> + if (foreign)
> + return;
> + }
> +
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260709-bcm_fixes-v11-0-067ba11bec5c@hartkopp.net?part=11
^ permalink raw reply [flat|nested] 20+ messages in thread