On Tue, Aug 07, 2018 at 01:31:21PM +0200, Vlastimil Babka wrote: > On 08/07/2018 01:26 PM, Michal Hocko wrote: > > On Tue 07-08-18 13:19:26, Florian Westphal wrote: > >> Michal Hocko <mhocko@xxxxxxxxxx> wrote: > >>>> I can't reproduce it anymore. > >>>> If i understand correctly this way memory allocated will be > >>>> accounted to kmem of this cgroup (if inside cgroup). > >>> > >>> s@this@caller's@ > >>> > >>> Florian, is this patch acceptable > >> > >> I am no mm expert. Should all longlived GFP_KERNEL allocations set ACCOUNT? > > > > No. We should focus only on those that are under direct userspace > > control and it can be triggered by an untrusted user. > > Looks like the description in include/linux/gfp.h could use some details > to guide developers, possibly also Mike's new/improved docs (+CC). A "memory allocation guide" is definitely on my todo. If you and other mm developers have any notes, random thoughts or anything else you'd like to see there, send it my way, I'll convert them to ReST :-). > >> If so, there are more places that should get same treatment. > >> The change looks fine to me, but again, I don't know when ACCOUNT should > >> be set in the first place. > > > > see above. > > > -- Sincerely yours, Mike. -- To unsubscribe from this list: send the line "unsubscribe netfilter-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
