Hi all, I notice that most of the usage of kobject_init_and_add in drivers are wrong, and now some drivers code has maken it right, please see commit dfb5394f804e (https://lkml.org/lkml/2020/4/11/282). function qib_create_port_files() in /drivers/infiniband/hw/qib/qib_sysfs.c may have the similar issue and leak kobject. if kobject_init_and_add() failed, the ppd->pport_kobj->kobj may already increased it's refcnt and allocated memory to store it's name, so a kobject_put is need before return. int qib_create_port_files(struct ib_device *ibdev, u8 port_num, struct kobject *kobj) { struct qib_pportdata *ppd; struct qib_devdata *dd = dd_from_ibdev(ibdev); int ret; if (!port_num || port_num > dd->num_pports) { qib_dev_err(dd, "Skipping infiniband class with invalid port %u\n", port_num); ret = -ENODEV; goto bail; } ppd = &dd->pport[port_num - 1]; ret = kobject_init_and_add(&ppd->pport_kobj, &qib_port_ktype, kobj, "linkcontrol"); if (ret) { qib_dev_err(dd, "Skipping linkcontrol sysfs info, (err %d) port %u\n", ret, port_num); goto bail; } ... bail: return ret; } Best regards, Lin Yi
