On 12/21/18 11:03 PM, Andrew Morton wrote: > On Fri, 21 Dec 2018 16:13:34 +0100 Michal Hocko <mhocko@xxxxxxxxxx> wrote: > >> On Fri 21-12-18 11:06:25, Vlastimil Babka wrote: >> [...] >>> There's also Michal's series to expand THP eligibility reporting. If it >>> would be feasible for you to switch to that implementation in your >>> userspace, and perhaps only locally and temporarily add your patch to >>> your currently used older kernel versions, then the benefit would be one >>> less obscure exception in the kernel API. >> >> Agreed. With this patch there is no way to check for the actual state of >> the madvise on the mapping reliably. So the only reason the flag is >> actually exported. > > So where does this leave us? Between a rock and a hard place? Seems like Google can't change their userspace easily, so we'll have to live with no way to check the actual state of the madvise. Maybe CRIU folks will care and have to add new flags e.g. 'nh2' and 'hg2' that ignore the prctl, or we can try again in 10 years hoping everyone meanwhile adopted Michal's new way to recognize the prctl.