> On Jan 11, 2020, at 6:03 AM, David Hildenbrand <david@xxxxxxxxxx> wrote: > > So I just remember why I think this (and the previously reported done > for ACPI DIMMs) are false positives. The actual locking order is > > onlining/offlining from user space: > > kn->count -> device_hotplug_lock -> cpu_hotplug_lock -> mem_hotplug_lock > > memory removal: > > device_hotplug_lock -> cpu_hotplug_lock -> mem_hotplug_lock -> kn->count > > > This looks like a locking inversion - but it's not. Whenever we come via > user space we do a mutex_trylock(), which resolves this issue by backing > up. The device_hotplug_lock will prevent > > I have no clue why the device_hotplug_lock does not pop up in the > lockdep report here. Sounds wrong to me. > > I think this is a false positive and not stable material. The point is that there are other paths does kn->count —> cpu_hotplug_lock without needing device_hotplug_lock to race with memory removal. kmem_cache_shrink_all+0x50/0x100 (cpu_hotplug_lock.rw_sem/mem_hotplug_lock.rw_sem) shrink_store+0x34/0x60 slab_attr_store+0x6c/0x170 sysfs_kf_write+0x70/0xb0 kernfs_fop_write+0x11c/0x270 ((kn->count) __vfs_write+0x3c/0x70 vfs_write+0xcc/0x200 ksys_write+0x7c/0x140 system_call+0x5c/0x6
