On 4/24/26 20:36, Michael S. Tsirkin wrote: > On Fri, Apr 24, 2026 at 08:33:45PM +0200, David Hildenbrand (Arm) wrote: >> On 4/23/26 18:13, Michael S. Tsirkin wrote: >>> >>> >>> You mean, that it stays because of the reserved pool? >> >> I meant, that if we are running with init_on_free or init_on_alloc, that we >> end >> up double-zeroing right now upstream. So nobody cared to optimize that part >> so far. > > on the aliasing arches you mean? that double zeroing?
Essentially where user_alloc_needs_zeroing() would have told you that there is no need to call folio_zero_user(). -- Cheers, David
