The ath11k active pdevs are protected by RCU but the temperature event
handling code calling ath11k_mac_get_ar_by_pdev_id() was not marked as a
read-side critical section as reported by RCU lockdep:
=============================
WARNING: suspicious RCU usage
6.6.0-rc6 #7 Not tainted
-----------------------------
drivers/net/wireless/ath/ath11k/mac.c:638 suspicious rcu_dereference_check() usage!
other info that might help us debug this:
rcu_scheduler_active = 2, debug_locks = 1
no locks held by swapper/0/0.
...
Call trace:
...
lockdep_rcu_suspicious+0x16c/0x22c
ath11k_mac_get_ar_by_pdev_id+0x194/0x1b0 [ath11k]
ath11k_wmi_tlv_op_rx+0xa84/0x2c1c [ath11k]
ath11k_htc_rx_completion_handler+0x388/0x510 [ath11k]
Mark the code in question as an RCU read-side critical section to avoid
any potential use-after-free issues.
Fixes: a41d10348b01 ("ath11k: add thermal sensor device support")
Cc: [email protected] # 5.7
Signed-off-by: Johan Hovold <[email protected]>
---
drivers/net/wireless/ath/ath11k/wmi.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/net/wireless/ath/ath11k/wmi.c b/drivers/net/wireless/ath/ath11k/wmi.c
index 23ad6825e5be..980ff588325d 100644
--- a/drivers/net/wireless/ath/ath11k/wmi.c
+++ b/drivers/net/wireless/ath/ath11k/wmi.c
@@ -8383,6 +8383,8 @@ ath11k_wmi_pdev_temperature_event(struct ath11k_base *ab,
ath11k_dbg(ab, ATH11K_DBG_WMI, "event pdev temperature ev temp %d pdev_id %d\n",
ev->temp, ev->pdev_id);
+ rcu_read_lock();
+
ar = ath11k_mac_get_ar_by_pdev_id(ab, ev->pdev_id);
if (!ar) {
ath11k_warn(ab, "invalid pdev id in pdev temperature ev %d", ev->pdev_id);
@@ -8392,6 +8394,8 @@ ath11k_wmi_pdev_temperature_event(struct ath11k_base *ab,
ath11k_thermal_event_temperature(ar, ev->temp);
+ rcu_read_unlock();
+
kfree(tb);
}
--
2.41.0
On 10/19/2023 8:04 AM, Jeff Johnson wrote:
> perhaps have a goto cleanup that does both the unlock() and the kfree()?
or goto exit to be consistent with others, including the DFS radar event
On Thu, Oct 19, 2023 at 08:04:55AM -0700, Jeff Johnson wrote:
> On 10/19/2023 3:42 AM, Johan Hovold wrote:
> > The ath11k active pdevs are protected by RCU but the temperature event
> > handling code calling ath11k_mac_get_ar_by_pdev_id() was not marked as a
> > read-side critical section as reported by RCU lockdep:
>
> How do I enable this? Just enable CONFIG_PROVE_RCU?
Yeah, via CONFIG_PROVE_LOCKING.
> Of course I'd also need to exercise the code path...
Right, this one I hit when reading out the sensor temperature (e.g.
using lm_sensors).
> > + rcu_read_lock();
> > +
> > ar = ath11k_mac_get_ar_by_pdev_id(ab, ev->pdev_id);
> > if (!ar) {
> > ath11k_warn(ab, "invalid pdev id in pdev temperature ev %d", ev->pdev_id);
>
> aren't you missing an unlock() in this error path?
>
> perhaps have a goto cleanup that does both the unlock() and the kfree()?
Bah, I am, thanks for catching that.
Spent too much time on scanning for further instances that I didn't
check the first one properly. Sorry about that. Will send a v2.
Johan