Skip to contentSkip to conversation

staging: vc04_services: Fix RCU dereference check

PATCH Superseded Padmanabha Srinivasaiah 2 participants 8 messages Dec 29, 2021 – Jan 23, 2022

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 staging: vc04_services: Fix RCU dereference check

In service_callback path RCU dereference done without rcu_read_[lock/unlock] pair, fixing same by using them.

[   32.201659] =============================
[   32.201664] WARNING: suspicious RCU usage
[   32.201670] 5.15.11-rt24-v8+ #3 Not tainted
[   32.201680] -----------------------------
[   32.201685] drivers/staging/vc04_services/interface/vchiq_arm/vchiq_core.h:529 suspicious rcu_dereference_check() usage!
[   32.201695]
[   32.201695] other info that might help us debug this:
[   32.201695]
[   32.201700]
[   32.201700] rcu_scheduler_active = 2, debug_locks = 1
[   32.201708] no locks held by vchiq-slot/0/98.
[   32.201715]
[   32.201715] stack backtrace:
[   32.201723] CPU: 1 PID: 98 Comm: vchiq-slot/0 Not tainted 5.15.11-rt24-v8+ #3
[   32.201733] Hardware name: Raspberry Pi 4 Model B Rev 1.4 (DT)
[   32.201739] Call trace:
[   32.201742]  dump_backtrace+0x0/0x1b8
[   32.201772]  show_stack+0x20/0x30
[   32.201784]  dump_stack_lvl+0x8c/0xb8
[   32.201799]  dump_stack+0x18/0x34
[   32.201808]  lockdep_rcu_suspicious+0xe4/0xf8
[   32.201817]  service_callback+0x124/0x400
[   32.201830]  slot_handler_func+0xf60/0x1e20
[   32.201839]  kthread+0x19c/0x1a8
[   32.201849]  ret_from_fork+0x10/0x20
---
 drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +3 -0
index 6759a6261500..ee1b48db9681 100644
@@ -1058,7 +1058,10 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 
 	DEBUG_TRACE(SERVICE_CALLBACK_LINE);
 
+	rcu_read_lock();
 	service = handle_to_service(handle);
+	rcu_read_unlock();
+
 	if (WARN_ON(!service))
 		return VCHIQ_SUCCESS;
 
Signature · 1 line
2.17.1
#1 lore raw

Hi Padmanabha,

Am 29.12.21 um 16:32 schrieb Padmanabha Srinivasaiah:

Quoted 39 lines

In service_callback path RCU dereference done without rcu_read_[lock/unlock] pair, fixing same by using them.

[   32.201659] =============================
[   32.201664] WARNING: suspicious RCU usage
[   32.201670] 5.15.11-rt24-v8+ #3 Not tainted
[   32.201680] -----------------------------
[   32.201685] drivers/staging/vc04_services/interface/vchiq_arm/vchiq_core.h:529 suspicious rcu_dereference_check() usage!
[   32.201695]
[   32.201695] other info that might help us debug this:
[   32.201695]
[   32.201700]
[   32.201700] rcu_scheduler_active = 2, debug_locks = 1
[   32.201708] no locks held by vchiq-slot/0/98.
[   32.201715]
[   32.201715] stack backtrace:
[   32.201723] CPU: 1 PID: 98 Comm: vchiq-slot/0 Not tainted 5.15.11-rt24-v8+ #3
[   32.201733] Hardware name: Raspberry Pi 4 Model B Rev 1.4 (DT)
[   32.201739] Call trace:
[   32.201742]  dump_backtrace+0x0/0x1b8
[   32.201772]  show_stack+0x20/0x30
[   32.201784]  dump_stack_lvl+0x8c/0xb8
[   32.201799]  dump_stack+0x18/0x34
[   32.201808]  lockdep_rcu_suspicious+0xe4/0xf8
[   32.201817]  service_callback+0x124/0x400
[   32.201830]  slot_handler_func+0xf60/0x1e20
[   32.201839]  kthread+0x19c/0x1a8
[   32.201849]  ret_from_fork+0x10/0x20
---
 drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +0 -0
index 6759a6261500..ee1b48db9681 100644

drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c service_callback(enum vchiq_reason reason, struct vchiq_header *header, · line 1067

Quoted 10 lines
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +3 -0
@@ -1058,7 +1058,10 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 
 	DEBUG_TRACE(SERVICE_CALLBACK_LINE);
 
+	rcu_read_lock();
 	service = handle_to_service(handle);
+	rcu_read_unlock();
+
 	if (WARN_ON(!service))
 		return VCHIQ_SUCCESS;
 

thanks for reporting this issue.

Could you please explain how to reproduce this issue?

What makes you sure that your patch fixes the issue and not just hiding the warning?

Best regards

#2 lore raw
Quoted 55 lines · Stefan Wahren

Hi Padmanabha,

Am 29.12.21 um 16:32 schrieb Padmanabha Srinivasaiah:

Quoted 39 lines

In service_callback path RCU dereference done without rcu_read_[lock/unlock] pair, fixing same by using them.

[   32.201659] =============================
[   32.201664] WARNING: suspicious RCU usage
[   32.201670] 5.15.11-rt24-v8+ #3 Not tainted
[   32.201680] -----------------------------
[   32.201685] drivers/staging/vc04_services/interface/vchiq_arm/vchiq_core.h:529 suspicious rcu_dereference_check() usage!
[   32.201695]
[   32.201695] other info that might help us debug this:
[   32.201695]
[   32.201700]
[   32.201700] rcu_scheduler_active = 2, debug_locks = 1
[   32.201708] no locks held by vchiq-slot/0/98.
[   32.201715]
[   32.201715] stack backtrace:
[   32.201723] CPU: 1 PID: 98 Comm: vchiq-slot/0 Not tainted 5.15.11-rt24-v8+ #3
[   32.201733] Hardware name: Raspberry Pi 4 Model B Rev 1.4 (DT)
[   32.201739] Call trace:
[   32.201742]  dump_backtrace+0x0/0x1b8
[   32.201772]  show_stack+0x20/0x30
[   32.201784]  dump_stack_lvl+0x8c/0xb8
[   32.201799]  dump_stack+0x18/0x34
[   32.201808]  lockdep_rcu_suspicious+0xe4/0xf8
[   32.201817]  service_callback+0x124/0x400
[   32.201830]  slot_handler_func+0xf60/0x1e20
[   32.201839]  kthread+0x19c/0x1a8
[   32.201849]  ret_from_fork+0x10/0x20
---
 drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +0 -0
index 6759a6261500..ee1b48db9681 100644

drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c service_callback(enum vchiq_reason reason, struct vchiq_header *header, · line 1067

Quoted 10 lines
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +3 -0
@@ -1058,7 +1058,10 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 
 	DEBUG_TRACE(SERVICE_CALLBACK_LINE);
 
+	rcu_read_lock();
 	service = handle_to_service(handle);
+	rcu_read_unlock();
+
 	if (WARN_ON(!service))
 		return VCHIQ_SUCCESS;
 

thanks for reporting this issue.

Quoted 1 line

Could you please explain how to reproduce this issue?

Have enabled lock dependency check in 5.15 plus rt patchset have been applied. And to catch this warning not loading any modules from default build.

Quoted 3 lines

What makes you sure that your patch fixes the issue and not just hiding the warning?

Thank you stefan for the question. Yes, you are right dereferenced pointer access should be inside the rcu read-side critical section. Will regenerate the patch and will share for the review.

Quoted 2 lines

Best regards

#3 lore raw

PATCH v2 1/1 staging: vc04_services: Fix RCU dereference check

In service_callback path RCU dereferenced pointer struct vchiq_service need to be accessed inside rcu read-critical section.

Accessing same with rcu_read_[lock/unlock] fixes the issue.

[   32.201659] =============================
[   32.201664] WARNING: suspicious RCU usage
[   32.201670] 5.15.11-rt24-v8+ #3 Not tainted
[   32.201680] -----------------------------
[   32.201685] drivers/staging/vc04_services/interface/vchiq_arm/vchiq_core.h:529 suspicious rcu_dereference_check() usage!
[   32.201695]
[   32.201695] other info that might help us debug this:
[   32.201695]
[   32.201700]
[   32.201700] rcu_scheduler_active = 2, debug_locks = 1
[   32.201708] no locks held by vchiq-slot/0/98.
[   32.201715]
[   32.201715] stack backtrace:
[   32.201723] CPU: 1 PID: 98 Comm: vchiq-slot/0 Not tainted 5.15.11-rt24-v8+ #3
[   32.201733] Hardware name: Raspberry Pi 4 Model B Rev 1.4 (DT)
[   32.201739] Call trace:
[   32.201742]  dump_backtrace+0x0/0x1b8
[   32.201772]  show_stack+0x20/0x30
[   32.201784]  dump_stack_lvl+0x8c/0xb8
[   32.201799]  dump_stack+0x18/0x34
[   32.201808]  lockdep_rcu_suspicious+0xe4/0xf8
[   32.201817]  service_callback+0x124/0x400
[   32.201830]  slot_handler_func+0xf60/0x1e20
[   32.201839]  kthread+0x19c/0x1a8
[   32.201849]  ret_from_fork+0x10/0x20

---
Changes in v2:
RCU dereferenced pointer need to be accessed inside rcu
read-side critical section.

 .../vc04_services/interface/vchiq_arm/vchiq_arm.c      | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +8 -2
index 6759a6261500..8ddd400ab2c3 100644
@@ -1053,24 +1053,30 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 	struct vchiq_service *service;
 	struct vchiq_instance *instance;
 	bool skip_completion = false;
+	unsigned int localport;
 
 	DEBUG_INITIALISE(g_state.local);
 
 	DEBUG_TRACE(SERVICE_CALLBACK_LINE);
 
+	rcu_read_lock();
 	service = handle_to_service(handle);
-	if (WARN_ON(!service))
+	if (WARN_ON(!service)) {
+		rcu_read_unlock();
 		return VCHIQ_SUCCESS;
+	}
 
 	user_service = (struct user_service *)service->base.userdata;
 	instance = user_service->instance;
+	localport = service->localport;
+	rcu_read_unlock();
 
 	if (!instance || instance->closing)
 		return VCHIQ_SUCCESS;
 
 	vchiq_log_trace(vchiq_arm_log_level,
 			"%s - service %lx(%d,%p), reason %d, header %lx, instance %lx, bulk_userdata %lx",
-			__func__, (unsigned long)user_service, service->localport,
+			__func__, (unsigned long)user_service, (int)localport,
 			user_service->userdata, reason, (unsigned long)header,
 			(unsigned long)instance, (unsigned long)bulk_userdata);
 
Signature · 1 line
2.17.1
#4 lore raw

Hi Padmanabha,

Am 30.12.21 um 15:54 schrieb Padmanabha Srinivasaiah:

Quoted 53 lines

In service_callback path RCU dereferenced pointer struct vchiq_service need to be accessed inside rcu read-critical section.

Accessing same with rcu_read_[lock/unlock] fixes the issue.

[   32.201659] =============================
[   32.201664] WARNING: suspicious RCU usage
[   32.201670] 5.15.11-rt24-v8+ #3 Not tainted
[   32.201680] -----------------------------
[   32.201685] drivers/staging/vc04_services/interface/vchiq_arm/vchiq_core.h:529 suspicious rcu_dereference_check() usage!
[   32.201695]
[   32.201695] other info that might help us debug this:
[   32.201695]
[   32.201700]
[   32.201700] rcu_scheduler_active = 2, debug_locks = 1
[   32.201708] no locks held by vchiq-slot/0/98.
[   32.201715]
[   32.201715] stack backtrace:
[   32.201723] CPU: 1 PID: 98 Comm: vchiq-slot/0 Not tainted 5.15.11-rt24-v8+ #3
[   32.201733] Hardware name: Raspberry Pi 4 Model B Rev 1.4 (DT)
[   32.201739] Call trace:
[   32.201742]  dump_backtrace+0x0/0x1b8
[   32.201772]  show_stack+0x20/0x30
[   32.201784]  dump_stack_lvl+0x8c/0xb8
[   32.201799]  dump_stack+0x18/0x34
[   32.201808]  lockdep_rcu_suspicious+0xe4/0xf8
[   32.201817]  service_callback+0x124/0x400
[   32.201830]  slot_handler_func+0xf60/0x1e20
[   32.201839]  kthread+0x19c/0x1a8
[   32.201849]  ret_from_fork+0x10/0x20

---
Changes in v2:
RCU dereferenced pointer need to be accessed inside rcu
read-side critical section.

 .../vc04_services/interface/vchiq_arm/vchiq_arm.c      | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +1 -0
index 6759a6261500..8ddd400ab2c3 100644
@@ -1053,24 +1053,30 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 	struct vchiq_service *service;
 	struct vchiq_instance *instance;
 	bool skip_completion = false;
+	unsigned int localport;
 
 	DEBUG_INITIALISE(g_state.local);
 
 	DEBUG_TRACE(SERVICE_CALLBACK_LINE);

drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c service_callback(enum vchiq_reason reason, struct vchiq_header *header, · line 1069

Quoted 10 lines
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +4 -1
 
+	rcu_read_lock();
 	service = handle_to_service(handle);
-	if (WARN_ON(!service))
+	if (WARN_ON(!service)) {
+		rcu_read_unlock();
 		return VCHIQ_SUCCESS;
+	}
 
 	user_service = (struct user_service *)service->base.userdata;

user_service is part of the service struct and it's modification below in this function is protected by a spinlock ( msg_queue_spinlock ). So i would expected that all read accesses to user_service before the spinlock are protected by RCU. After applying this patch there would be still the check for "user_service->is_vchi" unprotected. But i'm not sure about this.

Best regards

Quoted 14 lines
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +3 -1
 	instance = user_service->instance;
+	localport = service->localport;
+	rcu_read_unlock();
 
 	if (!instance || instance->closing)
 		return VCHIQ_SUCCESS;
 
 	vchiq_log_trace(vchiq_arm_log_level,
 			"%s - service %lx(%d,%p), reason %d, header %lx, instance %lx, bulk_userdata %lx",
-			__func__, (unsigned long)user_service, service->localport,
+			__func__, (unsigned long)user_service, (int)localport,
 			user_service->userdata, reason, (unsigned long)header,
 			(unsigned long)instance, (unsigned long)bulk_userdata);
 
#5 lore raw
Quoted 65 lines · Stefan Wahren

Hi Padmanabha,

Am 30.12.21 um 15:54 schrieb Padmanabha Srinivasaiah:

Quoted 62 lines

In service_callback path RCU dereferenced pointer struct vchiq_service need to be accessed inside rcu read-critical section.

Accessing same with rcu_read_[lock/unlock] fixes the issue.

[   32.201659] =============================
[   32.201664] WARNING: suspicious RCU usage
[   32.201670] 5.15.11-rt24-v8+ #3 Not tainted
[   32.201680] -----------------------------
[   32.201685] drivers/staging/vc04_services/interface/vchiq_arm/vchiq_core.h:529 suspicious rcu_dereference_check() usage!
[   32.201695]
[   32.201695] other info that might help us debug this:
[   32.201695]
[   32.201700]
[   32.201700] rcu_scheduler_active = 2, debug_locks = 1
[   32.201708] no locks held by vchiq-slot/0/98.
[   32.201715]
[   32.201715] stack backtrace:
[   32.201723] CPU: 1 PID: 98 Comm: vchiq-slot/0 Not tainted 5.15.11-rt24-v8+ #3
[   32.201733] Hardware name: Raspberry Pi 4 Model B Rev 1.4 (DT)
[   32.201739] Call trace:
[   32.201742]  dump_backtrace+0x0/0x1b8
[   32.201772]  show_stack+0x20/0x30
[   32.201784]  dump_stack_lvl+0x8c/0xb8
[   32.201799]  dump_stack+0x18/0x34
[   32.201808]  lockdep_rcu_suspicious+0xe4/0xf8
[   32.201817]  service_callback+0x124/0x400
[   32.201830]  slot_handler_func+0xf60/0x1e20
[   32.201839]  kthread+0x19c/0x1a8
[   32.201849]  ret_from_fork+0x10/0x20

---
Changes in v2:
RCU dereferenced pointer need to be accessed inside rcu
read-side critical section.

 .../vc04_services/interface/vchiq_arm/vchiq_arm.c      | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +5 -1
index 6759a6261500..8ddd400ab2c3 100644
@@ -1053,24 +1053,30 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 	struct vchiq_service *service;
 	struct vchiq_instance *instance;
 	bool skip_completion = false;
+	unsigned int localport;
 
 	DEBUG_INITIALISE(g_state.local);
 
 	DEBUG_TRACE(SERVICE_CALLBACK_LINE);
 
+	rcu_read_lock();
 	service = handle_to_service(handle);
-	if (WARN_ON(!service))
+	if (WARN_ON(!service)) {
+		rcu_read_unlock();
 		return VCHIQ_SUCCESS;
+	}
 
Quoted 9 lines

drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c service_callback(enum vchiq_reason reason, struct vchiq_header *header, · line 1069

Quoted 1 line
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +0 -0
 	user_service = (struct user_service *)service->base.userdata;

user_service is part of the service struct and it's modification below in this function is protected by a spinlock ( msg_queue_spinlock ). So i would expected that all read accesses to user_service before the spinlock are protected by RCU. After applying this patch there would be still the check for "user_service->is_vchi" unprotected. But i'm not sure about this.

Thank you stefan for identfying it. Yes, userdata/user_service can be potentially released after graceperiod.

Also here pointer is used around different synchronization mechanism, taking an extra reference will keep semantics simpler and will not prolong the graceperiod. Will regenerate the patch again.

Quoted 17 lines

Best regards

Quoted 14 lines
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +3 -1
 	instance = user_service->instance;
+	localport = service->localport;
+	rcu_read_unlock();
 
 	if (!instance || instance->closing)
 		return VCHIQ_SUCCESS;
 
 	vchiq_log_trace(vchiq_arm_log_level,
 			"%s - service %lx(%d,%p), reason %d, header %lx, instance %lx, bulk_userdata %lx",
-			__func__, (unsigned long)user_service, service->localport,
+			__func__, (unsigned long)user_service, (int)localport,
 			user_service->userdata, reason, (unsigned long)header,
 			(unsigned long)instance, (unsigned long)bulk_userdata);
 
#6 lore raw

PATCH v3 1/1 staging: vc04_services: Fix RCU dereference check

In service_callback path RCU dereferenced pointer struct vchiq_service need to be accessed inside rcu read-critical section.

Also userdata/user_service part of vchiq_service is accessed around different synchronization mechanism, getting an extra reference to a pointer keeps sematics simpler and avoids prolonged graceperiod.

Accessing vchiq_service with rcu_read_[lock/unlock] fixes below issue.

[   32.201659] =============================
[   32.201664] WARNING: suspicious RCU usage
[   32.201670] 5.15.11-rt24-v8+ #3 Not tainted
[   32.201680] -----------------------------
[   32.201685] drivers/staging/vc04_services/interface/vchiq_arm/vchiq_core.h:529 suspicious rcu_dereference_check() usage!
[   32.201695]
[   32.201695] other info that might help us debug this:
[   32.201695]
[   32.201700]
[   32.201700] rcu_scheduler_active = 2, debug_locks = 1
[   32.201708] no locks held by vchiq-slot/0/98.
[   32.201715]
[   32.201715] stack backtrace:
[   32.201723] CPU: 1 PID: 98 Comm: vchiq-slot/0 Not tainted 5.15.11-rt24-v8+ #3
[   32.201733] Hardware name: Raspberry Pi 4 Model B Rev 1.4 (DT)
[   32.201739] Call trace:
[   32.201742]  dump_backtrace+0x0/0x1b8
[   32.201772]  show_stack+0x20/0x30
[   32.201784]  dump_stack_lvl+0x8c/0xb8
[   32.201799]  dump_stack+0x18/0x34
[   32.201808]  lockdep_rcu_suspicious+0xe4/0xf8
[   32.201817]  service_callback+0x124/0x400
[   32.201830]  slot_handler_func+0xf60/0x1e20
[   32.201839]  kthread+0x19c/0x1a8
[   32.201849]  ret_from_fork+0x10/0x20

---
Change in v3:
Taking an extra reference for service pointer to keep semantics
simpler to accesses in different synchronization mechanism.

Changes in v2:
RCU dereferenced pointer need to be accessed inside rcu
read-side critical section.

 .../interface/vchiq_arm/vchiq_arm.c           | 20 +++++++++++++++++--
 1 file changed, 18 insertions(+), 2 deletions(-)
drivers/staging/vc04_services/interface/vchiq_arm/vchiq_arm.c +18 -2
index 6759a6261500..3a2e4582db8e 100644
@@ -1058,15 +1058,27 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 
 	DEBUG_TRACE(SERVICE_CALLBACK_LINE);
 
+	rcu_read_lock();
 	service = handle_to_service(handle);
-	if (WARN_ON(!service))
+	if (WARN_ON(!service)) {
+		rcu_read_unlock();
 		return VCHIQ_SUCCESS;
+	}
 
 	user_service = (struct user_service *)service->base.userdata;
 	instance = user_service->instance;
 
-	if (!instance || instance->closing)
+	if (!instance || instance->closing) {
+		rcu_read_unlock();
 		return VCHIQ_SUCCESS;
+	}
+
+	/*
+	 * As hopping around different synchronization mechanism,
+	 * taking an extra reference results in simpler implementation.
+	 */
+	vchiq_service_get(service);
+	rcu_read_unlock();
 
 	vchiq_log_trace(vchiq_arm_log_level,
 			"%s - service %lx(%d,%p), reason %d, header %lx, instance %lx, bulk_userdata %lx",
@@ -1097,6 +1109,7 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 							bulk_userdata);
 				if (status != VCHIQ_SUCCESS) {
 					DEBUG_TRACE(SERVICE_CALLBACK_LINE);
+					vchiq_service_put(service);
 					return status;
 				}
 			}
@@ -1105,10 +1118,12 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 			if (wait_for_completion_interruptible(&user_service->remove_event)) {
 				vchiq_log_info(vchiq_arm_log_level, "%s interrupted", __func__);
 				DEBUG_TRACE(SERVICE_CALLBACK_LINE);
+				vchiq_service_put(service);
 				return VCHIQ_RETRY;
 			} else if (instance->closing) {
 				vchiq_log_info(vchiq_arm_log_level, "%s closing", __func__);
 				DEBUG_TRACE(SERVICE_CALLBACK_LINE);
+				vchiq_service_put(service);
 				return VCHIQ_ERROR;
 			}
 			DEBUG_TRACE(SERVICE_CALLBACK_LINE);
@@ -1137,6 +1152,7 @@ service_callback(enum vchiq_reason reason, struct vchiq_header *header,
 		header = NULL;
 	}
 	DEBUG_TRACE(SERVICE_CALLBACK_LINE);
+	vchiq_service_put(service);
 
 	if (skip_completion)
 		return VCHIQ_SUCCESS;
Signature · 1 line
2.17.1
Tested-by on v3 1/1
#7 lore raw

Am 31.12.21 um 20:54 schrieb Padmanabha Srinivasaiah:

Quoted 35 lines

In service_callback path RCU dereferenced pointer struct vchiq_service need to be accessed inside rcu read-critical section.

Also userdata/user_service part of vchiq_service is accessed around different synchronization mechanism, getting an extra reference to a pointer keeps sematics simpler and avoids prolonged graceperiod.

Accessing vchiq_service with rcu_read_[lock/unlock] fixes below issue.

[   32.201659] =============================
[   32.201664] WARNING: suspicious RCU usage
[   32.201670] 5.15.11-rt24-v8+ #3 Not tainted
[   32.201680] -----------------------------
[   32.201685] drivers/staging/vc04_services/interface/vchiq_arm/vchiq_core.h:529 suspicious rcu_dereference_check() usage!
[   32.201695]
[   32.201695] other info that might help us debug this:
[   32.201695]
[   32.201700]
[   32.201700] rcu_scheduler_active = 2, debug_locks = 1
[   32.201708] no locks held by vchiq-slot/0/98.
[   32.201715]
[   32.201715] stack backtrace:
[   32.201723] CPU: 1 PID: 98 Comm: vchiq-slot/0 Not tainted 5.15.11-rt24-v8+ #3
[   32.201733] Hardware name: Raspberry Pi 4 Model B Rev 1.4 (DT)
[   32.201739] Call trace:
[   32.201742]  dump_backtrace+0x0/0x1b8
[   32.201772]  show_stack+0x20/0x30
[   32.201784]  dump_stack_lvl+0x8c/0xb8
[   32.201799]  dump_stack+0x18/0x34
[   32.201808]  lockdep_rcu_suspicious+0xe4/0xf8
[   32.201817]  service_callback+0x124/0x400
[   32.201830]  slot_handler_func+0xf60/0x1e20
[   32.201839]  kthread+0x19c/0x1a8
[   32.201849]  ret_from_fork+0x10/0x20
Quoted 1 line

Thanks