* [PATCH net v3 0/2] nfc: st21nfca: fix SE APDU request race condition
@ 2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
0 siblings, 0 replies; 8+ messages in thread
From: Joas Antonio dos Santos @ 2026-11-12 13:39 UTC (permalink / raw)
To: david, krzk, joasantonio108; +Cc: oe-linux-nfc, netdev, linux-kernel, horms
The BWI timeout work and the EVT_TRANSMIT_DATA handler can both
complete the same SE APDU request. The callback frees its context, so
the second completion is a use-after-free followed by a double free.
1/2: Add a spinlock around bwi_active, cb and cb_context in the three
paths that use them.
2/2: Use the lock to complete a request only once and to reject
overlapping requests.
This was found with the help of an LLM-based code review tool (Claude).
I checked the code paths by hand and reviewed the fixes myself. The
series is build-tested only (W=1, arm64); I do not have the hardware.
Not addressed here, all older than this series and pointed out by the
Sashiko review of v2:
- st21nfca_se_deinit() can leave bwi_timer armed when info is freed
(it runs before nfc_hci_unregister_device() and only deletes the
timer when bwi_active is set). The remove ordering needs its own
patch.
- EVT_TRANSMIT_DATA carries no request id, so a response that arrives
after its request timed out can still be taken for the answer to the
next request.
- When nfc_hci_send_event() fails in st21nfca_hci_se_io() the request
stays pending until the BWI timeout completes it with -ETIME.
- drivers/nfc/st-nci/se.c has the same double completion. I will send
a separate patch for it once the approach here is agreed on.
---
Changes in v3:
- 2/2: free cb_context when a request is rejected with -EBUSY; it was
leaked (Sashiko).
- 2/2: still send EVT_SE_END_OF_APDU_TRANSFER for a response that
arrives after the timeout, as before the series; only the callback is
skipped (Sashiko).
- 1/2: comment on what se_lock protects (checkpatch), commit message
no longer claims every access is serialized.
- Rebased on v7.4-rc3+, target tree added to the subject.
- Link to v2: https://lore.kernel.org/r/20261110-st21nfca-v2-v2-0-cf9102b314cb@gmail.com
Changes in v2:
- Use a spinlock instead of plain bwi_active flag checks (Simon Horman).
- Split into two patches: locking, then the completion fix.
- Set bwi_active before mod_timer().
- Copy cb/cb_context under the lock, call it after unlocking.
- Reject overlapping se_io requests with -EBUSY.
- Drop the Assisted-by trailer.
- Link to v1: https://lore.kernel.org/r/179332800288.52644.8728400883396719965@gmail.com
---
Joas Antonio dos Santos (2):
nfc: st21nfca: add spinlock to serialize SE APDU request state
nfc: st21nfca: prevent double completion of SE APDU request
drivers/nfc/st21nfca/se.c | 38 ++++++++++++++++++++++++++++++++++----
drivers/nfc/st21nfca/st21nfca.h | 2 ++
2 files changed, 36 insertions(+), 4 deletions(-)
---
base-commit: 9d66f26fefacebfeed15de31c1123451a17ff119
change-id: 20261108-st21nfca-v2-f4c5f861e416
Best regards,
--
Joas Antonio dos Santos <joasantonio108@gmail.com>
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH net v3 0/2] nfc: st21nfca: fix SE APDU request race condition
@ 2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
0 siblings, 0 replies; 8+ messages in thread
From: Joas Antonio dos Santos via B4 Relay @ 2026-11-12 13:39 UTC (permalink / raw)
To: david, krzk, joasantonio108; +Cc: oe-linux-nfc, netdev, linux-kernel, horms
The BWI timeout work and the EVT_TRANSMIT_DATA handler can both
complete the same SE APDU request. The callback frees its context, so
the second completion is a use-after-free followed by a double free.
1/2: Add a spinlock around bwi_active, cb and cb_context in the three
paths that use them.
2/2: Use the lock to complete a request only once and to reject
overlapping requests.
This was found with the help of an LLM-based code review tool (Claude).
I checked the code paths by hand and reviewed the fixes myself. The
series is build-tested only (W=1, arm64); I do not have the hardware.
Not addressed here, all older than this series and pointed out by the
Sashiko review of v2:
- st21nfca_se_deinit() can leave bwi_timer armed when info is freed
(it runs before nfc_hci_unregister_device() and only deletes the
timer when bwi_active is set). The remove ordering needs its own
patch.
- EVT_TRANSMIT_DATA carries no request id, so a response that arrives
after its request timed out can still be taken for the answer to the
next request.
- When nfc_hci_send_event() fails in st21nfca_hci_se_io() the request
stays pending until the BWI timeout completes it with -ETIME.
- drivers/nfc/st-nci/se.c has the same double completion. I will send
a separate patch for it once the approach here is agreed on.
---
Changes in v3:
- 2/2: free cb_context when a request is rejected with -EBUSY; it was
leaked (Sashiko).
- 2/2: still send EVT_SE_END_OF_APDU_TRANSFER for a response that
arrives after the timeout, as before the series; only the callback is
skipped (Sashiko).
- 1/2: comment on what se_lock protects (checkpatch), commit message
no longer claims every access is serialized.
- Rebased on v7.4-rc3+, target tree added to the subject.
- Link to v2: https://lore.kernel.org/r/20261110-st21nfca-v2-v2-0-cf9102b314cb@gmail.com
Changes in v2:
- Use a spinlock instead of plain bwi_active flag checks (Simon Horman).
- Split into two patches: locking, then the completion fix.
- Set bwi_active before mod_timer().
- Copy cb/cb_context under the lock, call it after unlocking.
- Reject overlapping se_io requests with -EBUSY.
- Drop the Assisted-by trailer.
- Link to v1: https://lore.kernel.org/r/179332800288.52644.8728400883396719965@gmail.com
---
Joas Antonio dos Santos (2):
nfc: st21nfca: add spinlock to serialize SE APDU request state
nfc: st21nfca: prevent double completion of SE APDU request
drivers/nfc/st21nfca/se.c | 38 ++++++++++++++++++++++++++++++++++----
drivers/nfc/st21nfca/st21nfca.h | 2 ++
2 files changed, 36 insertions(+), 4 deletions(-)
---
base-commit: 9d66f26fefacebfeed15de31c1123451a17ff119
change-id: 20261108-st21nfca-v2-f4c5f861e416
Best regards,
--
Joas Antonio dos Santos <joasantonio108@gmail.com>
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH net v3 1/2] nfc: st21nfca: add spinlock to serialize SE APDU request state
2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
@ 2026-11-12 13:39 ` Joas Antonio dos Santos
-1 siblings, 0 replies; 8+ messages in thread
From: Joas Antonio dos Santos via B4 Relay @ 2026-11-12 13:39 UTC (permalink / raw)
To: david, krzk, joasantonio108; +Cc: oe-linux-nfc, netdev, linux-kernel, horms
From: Joas Antonio dos Santos <joasantonio108@gmail.com>
The bwi_active flag, callback pointer (cb), and callback context
(cb_context) in st21nfca_se_info are accessed concurrently from
multiple contexts without synchronization:
- st21nfca_hci_se_io() sets them from the netlink path
- st21nfca_se_wt_work() reads/clears them from the timeout workqueue
- st21nfca_apdu_reader_event_received() reads/clears them from the
HCI rx path
Add a spinlock (se_lock) and take it around the accesses in these
three paths. The callback pointer and context are copied under the
lock and invoked after releasing it, since both the callback and
nfc_hci_send_event() may sleep.
Also move the bwi_active = true assignment before mod_timer() so that
the flag is always set before the timeout handler could observe it.
st21nfca_se_deinit() is not changed here. It still accesses
bwi_active without the lock; the teardown ordering has problems of its
own which need a separate fix.
Found by LLM-assisted manual code review.
Fixes: 2130fb97fecf ("NFC: st21nfca: Adding support for secure element")
Cc: Simon Horman <horms@kernel.org>
Signed-off-by: Joas Antonio dos Santos <joasantonio108@gmail.com>
---
drivers/nfc/st21nfca/se.c | 24 ++++++++++++++++++++----
drivers/nfc/st21nfca/st21nfca.h | 2 ++
2 files changed, 22 insertions(+), 4 deletions(-)
diff --git a/drivers/nfc/st21nfca/se.c b/drivers/nfc/st21nfca/se.c
index 7154bc1d644a..03a30f797ff8 100644
--- a/drivers/nfc/st21nfca/se.c
+++ b/drivers/nfc/st21nfca/se.c
@@ -227,11 +227,13 @@ int st21nfca_hci_se_io(struct nfc_hci_dev *hdev, u32 se_idx,
switch (se_idx) {
case ST21NFCA_ESE_HOST_ID:
+ spin_lock(&info->se_info.se_lock);
info->se_info.cb = cb;
info->se_info.cb_context = cb_context;
+ info->se_info.bwi_active = true;
+ spin_unlock(&info->se_info.se_lock);
mod_timer(&info->se_info.bwi_timer, jiffies +
msecs_to_jiffies(info->se_info.wt_timeout));
- info->se_info.bwi_active = true;
return nfc_hci_send_event(hdev, ST21NFCA_APDU_READER_GATE,
ST21NFCA_EVT_TRANSMIT_DATA,
apdu, apdu_length);
@@ -260,11 +262,17 @@ static void st21nfca_se_wt_work(struct work_struct *work)
*/
/* hardware reset managed through VCC_UICC_OUT power supply */
u8 param = 0x01;
+ se_io_cb_t cb;
+ void *cb_context;
struct st21nfca_hci_info *info = container_of(work,
struct st21nfca_hci_info,
se_info.timeout_work);
+ spin_lock(&info->se_info.se_lock);
info->se_info.bwi_active = false;
+ cb = info->se_info.cb;
+ cb_context = info->se_info.cb_context;
+ spin_unlock(&info->se_info.se_lock);
if (!info->se_info.xch_error) {
info->se_info.xch_error = true;
@@ -275,7 +283,7 @@ static void st21nfca_se_wt_work(struct work_struct *work)
nfc_hci_send_event(info->hdev, ST21NFCA_DEVICE_MGNT_GATE,
ST21NFCA_EVT_SE_HARD_RESET, ¶m, 1);
}
- info->se_info.cb(info->se_info.cb_context, NULL, 0, -ETIME);
+ cb(cb_context, NULL, 0, -ETIME);
}
static void st21nfca_se_wt_timeout(struct timer_list *t)
@@ -375,6 +383,8 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
u8 event, struct sk_buff *skb)
{
int r = 0;
+ se_io_cb_t cb;
+ void *cb_context;
struct st21nfca_hci_info *info = nfc_hci_get_clientdata(hdev);
pr_debug("apdu reader gate event: %x\n", event);
@@ -383,14 +393,19 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
case ST21NFCA_EVT_TRANSMIT_DATA:
timer_delete_sync(&info->se_info.bwi_timer);
cancel_work_sync(&info->se_info.timeout_work);
+
+ spin_lock(&info->se_info.se_lock);
info->se_info.bwi_active = false;
+ cb = info->se_info.cb;
+ cb_context = info->se_info.cb_context;
+ spin_unlock(&info->se_info.se_lock);
+
r = nfc_hci_send_event(hdev, ST21NFCA_DEVICE_MGNT_GATE,
ST21NFCA_EVT_SE_END_OF_APDU_TRANSFER, NULL, 0);
if (r < 0)
goto exit;
- info->se_info.cb(info->se_info.cb_context,
- skb->data, skb->len, 0);
+ cb(cb_context, skb->data, skb->len, 0);
break;
case ST21NFCA_EVT_WTX_REQUEST:
mod_timer(&info->se_info.bwi_timer, jiffies +
@@ -416,6 +431,7 @@ void st21nfca_se_init(struct nfc_hci_dev *hdev)
/* initialize timers */
timer_setup(&info->se_info.bwi_timer, st21nfca_se_wt_timeout, 0);
info->se_info.bwi_active = false;
+ spin_lock_init(&info->se_info.se_lock);
timer_setup(&info->se_info.se_active_timer,
st21nfca_se_activation_timeout, 0);
diff --git a/drivers/nfc/st21nfca/st21nfca.h b/drivers/nfc/st21nfca/st21nfca.h
index ae6771cc9894..03e41fadb5ed 100644
--- a/drivers/nfc/st21nfca/st21nfca.h
+++ b/drivers/nfc/st21nfca/st21nfca.h
@@ -6,6 +6,7 @@
#ifndef __LOCAL_ST21NFCA_H_
#define __LOCAL_ST21NFCA_H_
+#include <linux/spinlock.h>
#include <net/nfc/hci.h>
#include <linux/skbuff.h>
#include <linux/workqueue.h>
@@ -131,6 +132,7 @@ struct st21nfca_se_info {
struct timer_list bwi_timer;
int wt_timeout; /* in msecs */
bool bwi_active;
+ spinlock_t se_lock; /* protects bwi_active, cb and cb_context */
struct timer_list se_active_timer;
bool se_active;
--
2.43.0
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net v3 1/2] nfc: st21nfca: add spinlock to serialize SE APDU request state
@ 2026-11-12 13:39 ` Joas Antonio dos Santos
0 siblings, 0 replies; 8+ messages in thread
From: Joas Antonio dos Santos @ 2026-11-12 13:39 UTC (permalink / raw)
To: david, krzk, joasantonio108; +Cc: oe-linux-nfc, netdev, linux-kernel, horms
The bwi_active flag, callback pointer (cb), and callback context
(cb_context) in st21nfca_se_info are accessed concurrently from
multiple contexts without synchronization:
- st21nfca_hci_se_io() sets them from the netlink path
- st21nfca_se_wt_work() reads/clears them from the timeout workqueue
- st21nfca_apdu_reader_event_received() reads/clears them from the
HCI rx path
Add a spinlock (se_lock) and take it around the accesses in these
three paths. The callback pointer and context are copied under the
lock and invoked after releasing it, since both the callback and
nfc_hci_send_event() may sleep.
Also move the bwi_active = true assignment before mod_timer() so that
the flag is always set before the timeout handler could observe it.
st21nfca_se_deinit() is not changed here. It still accesses
bwi_active without the lock; the teardown ordering has problems of its
own which need a separate fix.
Found by LLM-assisted manual code review.
Fixes: 2130fb97fecf ("NFC: st21nfca: Adding support for secure element")
Cc: Simon Horman <horms@kernel.org>
Signed-off-by: Joas Antonio dos Santos <joasantonio108@gmail.com>
---
drivers/nfc/st21nfca/se.c | 24 ++++++++++++++++++++----
drivers/nfc/st21nfca/st21nfca.h | 2 ++
2 files changed, 22 insertions(+), 4 deletions(-)
diff --git a/drivers/nfc/st21nfca/se.c b/drivers/nfc/st21nfca/se.c
index 7154bc1d644a..03a30f797ff8 100644
--- a/drivers/nfc/st21nfca/se.c
+++ b/drivers/nfc/st21nfca/se.c
@@ -227,11 +227,13 @@ int st21nfca_hci_se_io(struct nfc_hci_dev *hdev, u32 se_idx,
switch (se_idx) {
case ST21NFCA_ESE_HOST_ID:
+ spin_lock(&info->se_info.se_lock);
info->se_info.cb = cb;
info->se_info.cb_context = cb_context;
+ info->se_info.bwi_active = true;
+ spin_unlock(&info->se_info.se_lock);
mod_timer(&info->se_info.bwi_timer, jiffies +
msecs_to_jiffies(info->se_info.wt_timeout));
- info->se_info.bwi_active = true;
return nfc_hci_send_event(hdev, ST21NFCA_APDU_READER_GATE,
ST21NFCA_EVT_TRANSMIT_DATA,
apdu, apdu_length);
@@ -260,11 +262,17 @@ static void st21nfca_se_wt_work(struct work_struct *work)
*/
/* hardware reset managed through VCC_UICC_OUT power supply */
u8 param = 0x01;
+ se_io_cb_t cb;
+ void *cb_context;
struct st21nfca_hci_info *info = container_of(work,
struct st21nfca_hci_info,
se_info.timeout_work);
+ spin_lock(&info->se_info.se_lock);
info->se_info.bwi_active = false;
+ cb = info->se_info.cb;
+ cb_context = info->se_info.cb_context;
+ spin_unlock(&info->se_info.se_lock);
if (!info->se_info.xch_error) {
info->se_info.xch_error = true;
@@ -275,7 +283,7 @@ static void st21nfca_se_wt_work(struct work_struct *work)
nfc_hci_send_event(info->hdev, ST21NFCA_DEVICE_MGNT_GATE,
ST21NFCA_EVT_SE_HARD_RESET, ¶m, 1);
}
- info->se_info.cb(info->se_info.cb_context, NULL, 0, -ETIME);
+ cb(cb_context, NULL, 0, -ETIME);
}
static void st21nfca_se_wt_timeout(struct timer_list *t)
@@ -375,6 +383,8 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
u8 event, struct sk_buff *skb)
{
int r = 0;
+ se_io_cb_t cb;
+ void *cb_context;
struct st21nfca_hci_info *info = nfc_hci_get_clientdata(hdev);
pr_debug("apdu reader gate event: %x\n", event);
@@ -383,14 +393,19 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
case ST21NFCA_EVT_TRANSMIT_DATA:
timer_delete_sync(&info->se_info.bwi_timer);
cancel_work_sync(&info->se_info.timeout_work);
+
+ spin_lock(&info->se_info.se_lock);
info->se_info.bwi_active = false;
+ cb = info->se_info.cb;
+ cb_context = info->se_info.cb_context;
+ spin_unlock(&info->se_info.se_lock);
+
r = nfc_hci_send_event(hdev, ST21NFCA_DEVICE_MGNT_GATE,
ST21NFCA_EVT_SE_END_OF_APDU_TRANSFER, NULL, 0);
if (r < 0)
goto exit;
- info->se_info.cb(info->se_info.cb_context,
- skb->data, skb->len, 0);
+ cb(cb_context, skb->data, skb->len, 0);
break;
case ST21NFCA_EVT_WTX_REQUEST:
mod_timer(&info->se_info.bwi_timer, jiffies +
@@ -416,6 +431,7 @@ void st21nfca_se_init(struct nfc_hci_dev *hdev)
/* initialize timers */
timer_setup(&info->se_info.bwi_timer, st21nfca_se_wt_timeout, 0);
info->se_info.bwi_active = false;
+ spin_lock_init(&info->se_info.se_lock);
timer_setup(&info->se_info.se_active_timer,
st21nfca_se_activation_timeout, 0);
diff --git a/drivers/nfc/st21nfca/st21nfca.h b/drivers/nfc/st21nfca/st21nfca.h
index ae6771cc9894..03e41fadb5ed 100644
--- a/drivers/nfc/st21nfca/st21nfca.h
+++ b/drivers/nfc/st21nfca/st21nfca.h
@@ -6,6 +6,7 @@
#ifndef __LOCAL_ST21NFCA_H_
#define __LOCAL_ST21NFCA_H_
+#include <linux/spinlock.h>
#include <net/nfc/hci.h>
#include <linux/skbuff.h>
#include <linux/workqueue.h>
@@ -131,6 +132,7 @@ struct st21nfca_se_info {
struct timer_list bwi_timer;
int wt_timeout; /* in msecs */
bool bwi_active;
+ spinlock_t se_lock; /* protects bwi_active, cb and cb_context */
struct timer_list se_active_timer;
bool se_active;
--
2.43.0
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net v3 2/2] nfc: st21nfca: prevent double completion of SE APDU request
2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
@ 2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
-1 siblings, 0 replies; 8+ messages in thread
From: Joas Antonio dos Santos @ 2026-11-12 13:39 UTC (permalink / raw)
To: david, krzk, joasantonio108; +Cc: oe-linux-nfc, netdev, linux-kernel, horms
Use the spinlock added in the previous commit to prevent completing an
SE APDU request more than once. Without this, the timeout work handler
(st21nfca_se_wt_work) and the EVT_TRANSMIT_DATA event handler can both
invoke the callback for the same request. The callback frees its
context, so the second call is a use-after-free followed by a double
free.
Check bwi_active under the lock in both completion paths and only
complete a request that is still pending. A response that arrives
after the timeout still gets its EVT_SE_END_OF_APDU_TRANSFER, as
before, but no callback.
Also reject a new SE APDU request in st21nfca_hci_se_io() with -EBUSY
while one is in progress, instead of overwriting the callback and
context of the pending request. The callback is not called in that
case, so free the context like the -ENODEV path does; nfc_se_io() does
not free it once the driver has been called.
Found by LLM-assisted manual code review.
Fixes: 2130fb97fecf ("NFC: st21nfca: Adding support for secure element")
Cc: Simon Horman <horms@kernel.org>
Signed-off-by: Joas Antonio dos Santos <joasantonio108@gmail.com>
---
drivers/nfc/st21nfca/se.c | 16 +++++++++++++++-
1 file changed, 15 insertions(+), 1 deletion(-)
diff --git a/drivers/nfc/st21nfca/se.c b/drivers/nfc/st21nfca/se.c
index 03a30f797ff8..63a88a61d310 100644
--- a/drivers/nfc/st21nfca/se.c
+++ b/drivers/nfc/st21nfca/se.c
@@ -228,6 +228,12 @@ int st21nfca_hci_se_io(struct nfc_hci_dev *hdev, u32 se_idx,
switch (se_idx) {
case ST21NFCA_ESE_HOST_ID:
spin_lock(&info->se_info.se_lock);
+ if (info->se_info.bwi_active) {
+ spin_unlock(&info->se_info.se_lock);
+ /* The callback is not called, see the comment below */
+ kfree(cb_context);
+ return -EBUSY;
+ }
info->se_info.cb = cb;
info->se_info.cb_context = cb_context;
info->se_info.bwi_active = true;
@@ -269,6 +275,10 @@ static void st21nfca_se_wt_work(struct work_struct *work)
se_info.timeout_work);
spin_lock(&info->se_info.se_lock);
+ if (!info->se_info.bwi_active) {
+ spin_unlock(&info->se_info.se_lock);
+ return;
+ }
info->se_info.bwi_active = false;
cb = info->se_info.cb;
cb_context = info->se_info.cb_context;
@@ -383,6 +393,7 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
u8 event, struct sk_buff *skb)
{
int r = 0;
+ bool active;
se_io_cb_t cb;
void *cb_context;
struct st21nfca_hci_info *info = nfc_hci_get_clientdata(hdev);
@@ -395,6 +406,7 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
cancel_work_sync(&info->se_info.timeout_work);
spin_lock(&info->se_info.se_lock);
+ active = info->se_info.bwi_active;
info->se_info.bwi_active = false;
cb = info->se_info.cb;
cb_context = info->se_info.cb_context;
@@ -405,7 +417,9 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
if (r < 0)
goto exit;
- cb(cb_context, skb->data, skb->len, 0);
+ /* The request may have been completed by the timeout already */
+ if (active)
+ cb(cb_context, skb->data, skb->len, 0);
break;
case ST21NFCA_EVT_WTX_REQUEST:
mod_timer(&info->se_info.bwi_timer, jiffies +
--
2.43.0
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH net v3 2/2] nfc: st21nfca: prevent double completion of SE APDU request
@ 2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
0 siblings, 0 replies; 8+ messages in thread
From: Joas Antonio dos Santos via B4 Relay @ 2026-11-12 13:39 UTC (permalink / raw)
To: david, krzk, joasantonio108; +Cc: oe-linux-nfc, netdev, linux-kernel, horms
From: Joas Antonio dos Santos <joasantonio108@gmail.com>
Use the spinlock added in the previous commit to prevent completing an
SE APDU request more than once. Without this, the timeout work handler
(st21nfca_se_wt_work) and the EVT_TRANSMIT_DATA event handler can both
invoke the callback for the same request. The callback frees its
context, so the second call is a use-after-free followed by a double
free.
Check bwi_active under the lock in both completion paths and only
complete a request that is still pending. A response that arrives
after the timeout still gets its EVT_SE_END_OF_APDU_TRANSFER, as
before, but no callback.
Also reject a new SE APDU request in st21nfca_hci_se_io() with -EBUSY
while one is in progress, instead of overwriting the callback and
context of the pending request. The callback is not called in that
case, so free the context like the -ENODEV path does; nfc_se_io() does
not free it once the driver has been called.
Found by LLM-assisted manual code review.
Fixes: 2130fb97fecf ("NFC: st21nfca: Adding support for secure element")
Cc: Simon Horman <horms@kernel.org>
Signed-off-by: Joas Antonio dos Santos <joasantonio108@gmail.com>
---
drivers/nfc/st21nfca/se.c | 16 +++++++++++++++-
1 file changed, 15 insertions(+), 1 deletion(-)
diff --git a/drivers/nfc/st21nfca/se.c b/drivers/nfc/st21nfca/se.c
index 03a30f797ff8..63a88a61d310 100644
--- a/drivers/nfc/st21nfca/se.c
+++ b/drivers/nfc/st21nfca/se.c
@@ -228,6 +228,12 @@ int st21nfca_hci_se_io(struct nfc_hci_dev *hdev, u32 se_idx,
switch (se_idx) {
case ST21NFCA_ESE_HOST_ID:
spin_lock(&info->se_info.se_lock);
+ if (info->se_info.bwi_active) {
+ spin_unlock(&info->se_info.se_lock);
+ /* The callback is not called, see the comment below */
+ kfree(cb_context);
+ return -EBUSY;
+ }
info->se_info.cb = cb;
info->se_info.cb_context = cb_context;
info->se_info.bwi_active = true;
@@ -269,6 +275,10 @@ static void st21nfca_se_wt_work(struct work_struct *work)
se_info.timeout_work);
spin_lock(&info->se_info.se_lock);
+ if (!info->se_info.bwi_active) {
+ spin_unlock(&info->se_info.se_lock);
+ return;
+ }
info->se_info.bwi_active = false;
cb = info->se_info.cb;
cb_context = info->se_info.cb_context;
@@ -383,6 +393,7 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
u8 event, struct sk_buff *skb)
{
int r = 0;
+ bool active;
se_io_cb_t cb;
void *cb_context;
struct st21nfca_hci_info *info = nfc_hci_get_clientdata(hdev);
@@ -395,6 +406,7 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
cancel_work_sync(&info->se_info.timeout_work);
spin_lock(&info->se_info.se_lock);
+ active = info->se_info.bwi_active;
info->se_info.bwi_active = false;
cb = info->se_info.cb;
cb_context = info->se_info.cb_context;
@@ -405,7 +417,9 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
if (r < 0)
goto exit;
- cb(cb_context, skb->data, skb->len, 0);
+ /* The request may have been completed by the timeout already */
+ if (active)
+ cb(cb_context, skb->data, skb->len, 0);
break;
case ST21NFCA_EVT_WTX_REQUEST:
mod_timer(&info->se_info.bwi_timer, jiffies +
--
2.43.0
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH net v3 1/2] nfc: st21nfca: add spinlock to serialize SE APDU request state
2026-11-12 13:39 ` Joas Antonio dos Santos
(?)
@ 2026-11-18 10:05 ` Simon Horman
-1 siblings, 0 replies; 8+ messages in thread
From: Simon Horman @ 2026-11-18 10:05 UTC (permalink / raw)
To: joasantonio108; +Cc: david, krzk, oe-linux-nfc, netdev, linux-kernel
On Thu, Nov 12, 2026 at 01:39:29PM +0000, Joas Antonio dos Santos via B4 Relay wrote:
> From: Joas Antonio dos Santos <joasantonio108@gmail.com>
>
> The bwi_active flag, callback pointer (cb), and callback context
> (cb_context) in st21nfca_se_info are accessed concurrently from
> multiple contexts without synchronization:
>
> - st21nfca_hci_se_io() sets them from the netlink path
> - st21nfca_se_wt_work() reads/clears them from the timeout workqueue
> - st21nfca_apdu_reader_event_received() reads/clears them from the
> HCI rx path
>
> Add a spinlock (se_lock) and take it around the accesses in these
> three paths. The callback pointer and context are copied under the
> lock and invoked after releasing it, since both the callback and
> nfc_hci_send_event() may sleep.
>
> Also move the bwi_active = true assignment before mod_timer() so that
> the flag is always set before the timeout handler could observe it.
>
> st21nfca_se_deinit() is not changed here. It still accesses
> bwi_active without the lock; the teardown ordering has problems of its
> own which need a separate fix.
>
> Found by LLM-assisted manual code review.
>
> Fixes: 2130fb97fecf ("NFC: st21nfca: Adding support for secure element")
> Cc: Simon Horman <horms@kernel.org>
> Signed-off-by: Joas Antonio dos Santos <joasantonio108@gmail.com>
Hi Joas,
sashiko.dev has provided an AI-generated review of this patch,
which seems to raise issues not covered so far.
Can releasing se_lock before mod_timer() leave bwi_timer armed while
bwi_active is false?
CPU0 CPU1
st21nfca_hci_se_io()
spin_lock(&se_lock)
bwi_active = true
spin_unlock(&se_lock)
st21nfca_apdu_reader_event_received()
timer_delete_sync(&bwi_timer)
...
spin_lock(&se_lock)
bwi_active = false
spin_unlock(&se_lock)
mod_timer(&bwi_timer, ...)
st21nfca_hci_remove()
st21nfca_se_deinit()
/* bwi_active is false, skips
* timer_delete_sync() */
kfree(info)
<softirq>
st21nfca_se_wt_timeout()
schedule_work(&info->...) <-- use-after-free
The ST21NFCA_EVT_WTX_REQUEST branch in st21nfca_apdu_reader_event_received()
also calls mod_timer(&info->se_info.bwi_timer, ...) without holding se_lock or
checking bwi_active, which likewise re-arms bwi_timer after a timeout or on an
unsolicited event while bwi_active is false.
In addition, st21nfca_se_deinit() reads and clears bwi_active without se_lock,
only calls timer_delete_sync(&info->se_info.bwi_timer) when bwi_active is
true, and leaks cb_context if it cancels an in-flight request:
drivers/nfc/st21nfca/se.c:st21nfca_se_deinit() {
if (info->se_info.bwi_active)
timer_delete_sync(&info->se_info.bwi_timer);
if (info->se_info.se_active)
timer_delete_sync(&info->se_info.se_active_timer);
cancel_work_sync(&info->se_info.timeout_work);
info->se_info.bwi_active = false;
info->se_info.se_active = false;
}
These gaps are not addressed in the follow-up commit "nfc: st21nfca: prevent
double completion of SE APDU request" either.
Link: https://sashiko.dev/#/patchset/20261112-st21nfca-v2-v3-0-c81702269e01%40gmail.com?part=1
I actually think there are two issues here.
1. The headline issue relating to races in teardown. This strikes me as
a pre-existing problem. And suitable for follow-up.
2. This part: " In addition, st21nfca_se_deinit() reads and clears
bwi_active without se_lock". And it seems to me it would be worth
addressing that in an updated version of this patch for completeness.
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH net v3 2/2] nfc: st21nfca: prevent double completion of SE APDU request
2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
(?)
@ 2026-11-18 10:05 ` Simon Horman
-1 siblings, 0 replies; 8+ messages in thread
From: Simon Horman @ 2026-11-18 10:05 UTC (permalink / raw)
To: joasantonio108; +Cc: david, krzk, oe-linux-nfc, netdev, linux-kernel
On Thu, Nov 12, 2026 at 01:39:30PM +0000, Joas Antonio dos Santos via B4 Relay wrote:
> From: Joas Antonio dos Santos <joasantonio108@gmail.com>
>
> Use the spinlock added in the previous commit to prevent completing an
> SE APDU request more than once. Without this, the timeout work handler
> (st21nfca_se_wt_work) and the EVT_TRANSMIT_DATA event handler can both
> invoke the callback for the same request. The callback frees its
> context, so the second call is a use-after-free followed by a double
> free.
>
> Check bwi_active under the lock in both completion paths and only
> complete a request that is still pending. A response that arrives
> after the timeout still gets its EVT_SE_END_OF_APDU_TRANSFER, as
> before, but no callback.
>
> Also reject a new SE APDU request in st21nfca_hci_se_io() with -EBUSY
> while one is in progress, instead of overwriting the callback and
> context of the pending request. The callback is not called in that
> case, so free the context like the -ENODEV path does; nfc_se_io() does
> not free it once the driver has been called.
>
> Found by LLM-assisted manual code review.
>
> Fixes: 2130fb97fecf ("NFC: st21nfca: Adding support for secure element")
> Cc: Simon Horman <horms@kernel.org>
> Signed-off-by: Joas Antonio dos Santos <joasantonio108@gmail.com>
Hi Joas,
Sashiko.dev has provided an AI-generated review of this patch at
https://sashiko.dev/#/patchset/20261112-st21nfca-v2-v3-0-c81702269e01%40gmail.com?part=2
It raises two issues.
Issue 1:
> diff --git a/drivers/nfc/st21nfca/se.c b/drivers/nfc/st21nfca/se.c
> index 03a30f797ff8f..63a88a61d310f 100644
> --- a/drivers/nfc/st21nfca/se.c
> +++ b/drivers/nfc/st21nfca/se.c
> @@ -228,6 +228,12 @@ int st21nfca_hci_se_io(struct nfc_hci_dev *hdev, u32 se_idx,
> switch (se_idx) {
> case ST21NFCA_ESE_HOST_ID:
> spin_lock(&info->se_info.se_lock);
> + if (info->se_info.bwi_active) {
> + spin_unlock(&info->se_info.se_lock);
> + /* The callback is not called, see the comment below */
> + kfree(cb_context);
> + return -EBUSY;
> + }
> info->se_info.cb = cb;
> info->se_info.cb_context = cb_context;
> info->se_info.bwi_active = true;
What happens if nfc_hci_send_event() fails at the end of this case?
drivers/nfc/st21nfca/se.c:st21nfca_hci_se_io() {
...
info->se_info.bwi_active = true;
spin_unlock(&info->se_info.se_lock);
mod_timer(&info->se_info.bwi_timer, jiffies +
msecs_to_jiffies(info->se_info.wt_timeout));
return nfc_hci_send_event(hdev, ST21NFCA_APDU_READER_GATE,
ST21NFCA_EVT_TRANSMIT_DATA,
apdu, apdu_length);
...
}
When nfc_hci_send_event() returns a synchronous error such as
-EADDRNOTAVAIL, info->se_info.bwi_active stays true and
info->se_info.bwi_timer remains armed. With the new -EBUSY check, every
later st21nfca_hci_se_io() call is rejected until the timer expires, at
which point st21nfca_se_wt_work() sends a spurious reset to the secure
element and invokes cb(cb_context, NULL, 0, -ETIME) for a request that
already failed synchronously.
It seems to me that while this is worth checking into, it may well
be theoretical or unreachable. In which case I think any solution
need not be included in this patch-set.
Issue 2:
> @@ -269,6 +275,10 @@ static void st21nfca_se_wt_work(struct work_struct *work)
> se_info.timeout_work);
>
> spin_lock(&info->se_info.se_lock);
> + if (!info->se_info.bwi_active) {
> + spin_unlock(&info->se_info.se_lock);
> + return;
> + }
> info->se_info.bwi_active = false;
> cb = info->se_info.cb;
> cb_context = info->se_info.cb_context;
> @@ -383,6 +393,7 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
> u8 event, struct sk_buff *skb)
> {
> int r = 0;
> + bool active;
> se_io_cb_t cb;
> void *cb_context;
> struct st21nfca_hci_info *info = nfc_hci_get_clientdata(hdev);
> @@ -395,6 +406,7 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
> cancel_work_sync(&info->se_info.timeout_work);
>
> spin_lock(&info->se_info.se_lock);
> + active = info->se_info.bwi_active;
> info->se_info.bwi_active = false;
> cb = info->se_info.cb;
> cb_context = info->se_info.cb_context;
> @@ -405,7 +417,9 @@ int st21nfca_apdu_reader_event_received(struct nfc_hci_dev *hdev,
> if (r < 0)
> goto exit;
>
> - cb(cb_context, skb->data, skb->len, 0);
> + /* The request may have been completed by the timeout already */
> + if (active)
> + cb(cb_context, skb->data, skb->len, 0);
> break;
Can a late ST21NFCA_EVT_TRANSMIT_DATA from a timed-out request complete a
newly submitted request instead?
CPU0 (timeout_work) CPU1 (msg_rx_work) CPU2 (syscall)
st21nfca_se_wt_work() st21nfca_apdu_reader_event_received()
spin_lock(&se_lock) timer_delete_sync()
bwi_active = false cancel_work_sync()
spin_unlock(&se_lock)
st21nfca_hci_se_io()
spin_lock(&se_lock)
cb = req2_cb
cb_context = req2_ctx
bwi_active = true
spin_unlock(&se_lock)
mod_timer(&bwi_timer)
nfc_hci_send_event(reset)
req1_cb(req1_ctx, -ETIME)
spin_lock(&se_lock)
active = bwi_active (true)
bwi_active = false
cb = req2_cb
cb_context = req2_ctx
spin_unlock(&se_lock)
...
req2_cb(req2_ctx, req1_skb) <-- wrong request
Once st21nfca_se_wt_work() clears info->se_info.bwi_active and drops
info->se_info.se_lock, a new st21nfca_hci_se_io() call can install a second
request before st21nfca_apdu_reader_event_received() acquires se_lock (or
before a late response from the first request arrives). The first
request's response then completes and frees the second request's
cb_context with stale data, while st21nfca_se_wt_work() can also queue its
reset after the new request.
In addition, because both st21nfca_apdu_reader_event_received() above and
st21nfca_hci_se_io() call timer_delete_sync() / mod_timer() outside
se_lock, info->se_info.bwi_timer can remain armed after bwi_active is set
to false; st21nfca_se_deinit() only deletes bwi_timer when bwi_active is
true, so removing the device before the timer expires can then trigger a
use-after-free.
This seems to be a problem, but it also seems to be related to the
tear-down race flagged in patch 1/2. And as per my opinion there,
I think this is likely out-of scope for this patch-set.
^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2026-11-18 10:05 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-12 13:39 [PATCH net v3 0/2] nfc: st21nfca: fix SE APDU request race condition Joas Antonio dos Santos
2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
2026-11-12 13:39 ` [PATCH net v3 1/2] nfc: st21nfca: add spinlock to serialize SE APDU request state Joas Antonio dos Santos via B4 Relay
2026-11-12 13:39 ` Joas Antonio dos Santos
2026-11-18 10:05 ` Simon Horman
2026-11-12 13:39 ` [PATCH net v3 2/2] nfc: st21nfca: prevent double completion of SE APDU request Joas Antonio dos Santos
2026-11-12 13:39 ` Joas Antonio dos Santos via B4 Relay
2026-11-18 10:05 ` Simon Horman
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.