[PATCH v6 03/16] firmware: arm_scmi: Quiesce notifications before teardown

Sudeep Holla <[email protected]> Tue, 14 Jul 2026 13:56:22 +0100
Newsgroups org.kernel.vger.arm-scmi,org.infradead.lists.linux-arm-kernel
Message-ID <[email protected]>
scmi_notification_exit() clears and releases the notification instance,
but transport callbacks can still deliver incoming notifications until
the TX/RX channels are freed. During remove, an RX interrupt in that
window can enter scmi_notify() while notification state is being torn
down and then dereference freed memory. The same ordering exists on the
probe error path after notification initialization.

The notification late-init worker has a separate lifetime issue: protocol
event registration queues ni->init_work on the system workqueue, so
destroying ni->notify_wq does not drain that work. If the devres group is
released while init_work is still pending or running, the late-init worker
can dereference the freed notification instance.

Quiesce the notification core before TX/RX channels are torn down, then
clean up the channels before releasing the notification core resources.
Use disable_work_sync() so future late-init queueing is rejected and any
already queued or running late-init work has completed before channel
teardown starts.

Fixes: 1e7cbfaa66d3 ("firmware: arm_scmi: Free mailbox channels if probe fails")
Reported-by: Sashiko <[email protected]>
Signed-off-by: Sudeep Holla <[email protected]>
---
 drivers/firmware/arm_scmi/driver.c | 13 +++++++------
 drivers/firmware/arm_scmi/notify.c | 21 +++++++++++++++++++++
 drivers/firmware/arm_scmi/notify.h |  1 +
 3 files changed, 29 insertions(+), 6 deletions(-)

diff --git a/drivers/firmware/arm_scmi/driver.c b/drivers/firmware/arm_scmi/driver.c
index 52aff3aded4c..ec373595f955 100644
--- a/drivers/firmware/arm_scmi/driver.c
+++ b/drivers/firmware/arm_scmi/driver.c
@@ -3325,7 +3325,7 @@ static int scmi_probe(struct platform_device *pdev)
 			dev_err(dev, "%s", err_str);
 			return 0;
 		}
-		goto notification_exit;
+		goto raw_mode_cleanup;
 	}
 
 	mutex_lock(&scmi_list_mutex);
@@ -3367,17 +3367,18 @@ static int scmi_probe(struct platform_device *pdev)
 
 	return 0;
 
-notification_exit:
+raw_mode_cleanup:
 	if (IS_ENABLED(CONFIG_ARM_SCMI_RAW_MODE_SUPPORT))
 		scmi_raw_mode_cleanup(info->raw);
-	scmi_notification_exit(&info->handle);
 clear_dev_req_notifier:
 	blocking_notifier_chain_unregister(&scmi_requested_devices_nh,
 					   &info->dev_req_nb);
 clear_bus_notifier:
 	bus_unregister_notifier(&scmi_bus_type, &info->bus_nb);
 clear_txrx_setup:
+	scmi_notification_quiesce(&info->handle);
 	scmi_cleanup_txrx_channels(info);
+	scmi_notification_exit(&info->handle);
 clear_ida:
 	ida_free(&scmi_id, info->id);
 
@@ -3404,6 +3405,9 @@ static void scmi_remove(struct platform_device *pdev)
 	blocking_notifier_chain_unregister(&scmi_requested_devices_nh,
 					   &info->dev_req_nb);
 
+	/* Stop transport callbacks before tearing down notifications. */
+	scmi_notification_quiesce(&info->handle);
+	scmi_cleanup_txrx_channels(info);
 	scmi_notification_exit(&info->handle);
 
 	mutex_lock(&info->protocols_mtx);
@@ -3416,9 +3420,6 @@ static void scmi_remove(struct platform_device *pdev)
 
 	bus_unregister_notifier(&scmi_bus_type, &info->bus_nb);
 
-	/* Safe to free channels since no more users */
-	scmi_cleanup_txrx_channels(info);
-
 	ida_free(&scmi_id, info->id);
 }
 
diff --git a/drivers/firmware/arm_scmi/notify.c b/drivers/firmware/arm_scmi/notify.c
index 40ec184eedae..3239c28d6fa9 100644
--- a/drivers/firmware/arm_scmi/notify.c
+++ b/drivers/firmware/arm_scmi/notify.c
@@ -1706,6 +1706,25 @@ int scmi_notification_init(struct scmi_handle *handle)
 	return -ENOMEM;
 }
 
+/**
+ * scmi_notification_quiesce()  - Stop notification late initialization
+ * @handle: The handle identifying the platform instance to quiesce
+ *
+ * Prevent new late-init work from being queued and wait for any already queued
+ * or running late-init work to complete before transport channels are torn
+ * down.
+ */
+void scmi_notification_quiesce(struct scmi_handle *handle)
+{
+	struct scmi_notify_instance *ni;
+
+	ni = scmi_notification_instance_data_get(handle);
+	if (!ni)
+		return;
+
+	disable_work_sync(&ni->init_work);
+}
+
 /**
  * scmi_notification_exit()  - Shutdown and clean Notification core
  * @handle: The handle identifying the platform instance to shutdown
@@ -1717,6 +1736,8 @@ void scmi_notification_exit(struct scmi_handle *handle)
 	ni = scmi_notification_instance_data_get(handle);
 	if (!ni)
 		return;
+
+	scmi_notification_quiesce(handle);
 	scmi_notification_instance_data_set(handle, NULL);
 
 	/* Destroy while letting pending work complete */
diff --git a/drivers/firmware/arm_scmi/notify.h b/drivers/firmware/arm_scmi/notify.h
index 76758a736cf4..f18f98c5ab3b 100644
--- a/drivers/firmware/arm_scmi/notify.h
+++ b/drivers/firmware/arm_scmi/notify.h
@@ -82,6 +82,7 @@ struct scmi_protocol_events {
 };
 
 int scmi_notification_init(struct scmi_handle *handle);
+void scmi_notification_quiesce(struct scmi_handle *handle);
 void scmi_notification_exit(struct scmi_handle *handle);
 int scmi_register_protocol_events(const struct scmi_handle *handle, u8 proto_id,
 				  const struct scmi_protocol_handle *ph,

-- 
2.43.0