> object_init_and_add() should be handled when it return an error, > because kobject_init_and_add() takes reference even when it fails. I find this wording approach improvable. > Previous commit "b8eb718348b8" fixed a similar problem. Would you like to indicate with such information (according to the topic “net-sysfs: Fix reference count leak in rx|netdev_queue_add_kobject”) that your issue was detected also with the help of the kernel fuzzing tool “syzkaller”? How do you think about to add an imperative wording to the change description? https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?id=b0c3ba31be3e45a130e13b278cf3b90f69bda6f6#n151 Will it be helpful to combine this update step together with “edac: Fix reference count leak in edac_pci_main_kobj_setup” into a small patch series? https://lore.kernel.org/lkml/20200528202238.18078-1-wu000273@xxxxxxx/ https://lore.kernel.org/patchwork/patch/1249351/ Regards, Markus
