On Thu 22-10-20 10:04:50, David Hildenbrand wrote: [...] > > None of that would address the original point of this thread, the global > > init_on_alloc parameter. > > Yes, but I guess we're past that: whatever leaves the buddy shall be > zeroed out. That's the whole point of that security hardening mechanism. Hugetlb can control its zeroying behavior via mount option (for MAP_HUGETLB controled by a command line parameter). If the page fault handler can recognize the pre-initialized pages then both init_on* can be implemented along with such a hugetlb specific mechanism. -- Michal Hocko SUSE Labs
