On Tue, Jun 17, 2014 at 03:47:45PM +0200, Michal Hocko wrote: > On Mon 16-06-14 15:54:23, Johannes Weiner wrote: > > Transparent huge page charges prefer falling back to regular pages > > rather than spending a lot of time in direct reclaim. > > > > Desired reclaim behavior is usually declared in the gfp mask, but THP > > charges use GFP_KERNEL and then rely on the fact that OOM is disabled > > for THP charges, and that OOM-disabled charges currently skip reclaim. > > OOM-disabled charges do one round of reclaim currently. Oops, fixed in v4. > > Needless to say, this is anything but obvious and quite error prone. > > > > Convert THP charges to use GFP_TRANSHUGE instead, which implies > > __GFP_NORETRY, to indicate the low-latency requirement. > > OK, this makes sense. It would be ideal if we could use the same gfp as > for allocation but that would be too much churn I guess because some > allocator use a allocation helper which deduces proper gfp flags without > giving them back to the caller. > > Nevertheless, I would still prefer if 05/12 was moved before > this patch because this is strictly speaking a behavior change. Yes, that's bungled up, thanks for catching that. So here is the order I put it in (reverse git history order of course): commit d0d31c8d4f4cf91edcffa704e8c65ca62af24cf8 Author: Johannes Weiner <hannes@xxxxxxxxxxx> Date: Mon Apr 14 08:16:09 2014 -0400 mm: memcontrol: retry reclaim for oom-disabled and __GFP_NOFAIL charges There is no reason why oom-disabled and __GFP_NOFAIL charges should try to reclaim only once when every other charge tries several times before giving up. Make them all retry the same number of times. Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx> commit 69f5c6c1a6553a04d7701012a73b2477df8d5a19 Author: Johannes Weiner <hannes@xxxxxxxxxxx> Date: Thu Jun 5 22:02:26 2014 -0400 mm: huge_memory: use GFP_TRANSHUGE when charging huge pages Transparent huge page charges prefer falling back to regular pages rather than spending a lot of time in direct reclaim. Desired reclaim behavior is usually declared in the gfp mask, but THP charges use GFP_KERNEL and then rely on the fact that OOM is disabled for THP charges, and that OOM-disabled charges don't retry reclaim. Needless to say, this is anything but obvious and quite error prone. Convert THP charges to use GFP_TRANSHUGE instead, which implies __GFP_NORETRY, to indicate the low-latency requirement. Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx> Acked-by: Michal Hocko <mhocko@xxxxxxx> commit d485e6b4ed62885d54c57c18c5427e2f174c9012 Author: Johannes Weiner <hannes@xxxxxxxxxxx> Date: Tue May 27 15:23:18 2014 -0400 mm: memcontrol: reclaim at least once for __GFP_NORETRY Currently, __GFP_NORETRY tries charging once and gives up before even trying to reclaim. Bring the behavior on par with the page allocator and reclaim at least once before giving up. Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx> Acked-by: Michal Hocko <mhocko@xxxxxxx> This first changes __GFP_NORETRY to provide THP-required semantics, then switches THP over to it, then fixes oom-disabled/NOFAIL charges. Does that make more sense? > > Signed-off-by: Johannes Weiner <hannes@xxxxxxxxxxx> > > Anyway > Acked-by: Michal Hocko <mhocko@xxxxxxx> Thanks! -- 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>