destroy_workqueue seems necessary before return from scmi_notification_init in the error handling case when fails to do devm_kcalloc(). Fix this by simply moving devm_kcalloc to the front.
Fixes: bd31b249692e ("firmware: arm_scmi: Add notification dispatch and delivery") Suggested-by: Cristian Marussi <cristian.maru...@arm.com> Signed-off-by: Qinglang Miao <miaoqingl...@huawei.com> --- v2: fix this problem by simply moving codes. drivers/firmware/arm_scmi/notify.c | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/drivers/firmware/arm_scmi/notify.c b/drivers/firmware/arm_scmi/notify.c index 2754f9d01636..fdb2cc95dfde 100644 --- a/drivers/firmware/arm_scmi/notify.c +++ b/drivers/firmware/arm_scmi/notify.c @@ -1468,17 +1468,17 @@ int scmi_notification_init(struct scmi_handle *handle) ni->gid = gid; ni->handle = handle; + ni->registered_protocols = devm_kcalloc(handle->dev, SCMI_MAX_PROTO, + sizeof(char *), GFP_KERNEL); + if (!ni->registered_protocols) + goto err; + ni->notify_wq = alloc_workqueue("scmi_notify", WQ_UNBOUND | WQ_FREEZABLE | WQ_SYSFS, 0); if (!ni->notify_wq) goto err; - ni->registered_protocols = devm_kcalloc(handle->dev, SCMI_MAX_PROTO, - sizeof(char *), GFP_KERNEL); - if (!ni->registered_protocols) - goto err; - mutex_init(&ni->pending_mtx); hash_init(ni->pending_events_handlers); -- 2.23.0