On Wed 22-11-17 15:31:14, Paolo Bonzini wrote: > On 22/11/2017 14:06, Tetsuo Handa wrote: > >> I am not sure we want to overcomplicate the code too much. Most > >> architectures do not have that many numa nodes to care. If we really > >> need to care maybe we should rethink and get rid of the per numa > >> deferred count altogether. > > the amount of changes needed for checking for an error will exceed the amount of > > changes needed for making register_shrinker() not to return an error. > > Do we want to overcomplicate register_shrinker() callers? > > For KVM it's not a big deal, fixing kvm_mmu_module_init to check the > return value is trivial. I suspect others will be in a similar situation. I've tried to do so for sget_userns [1] and it didn't look terrible either. [1] http://lkml.kernel.org/r/20171121140500.bgkpwcdk2dxesao4@xxxxxxxxxxxxxx -- Michal Hocko SUSE Labs -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>
