On Tue 14-01-20 09:05:57, Vlastimil Babka wrote: > On 1/14/20 3:16 AM, Linus Torvalds wrote: > > On Mon, Jan 13, 2020 at 4:29 PM Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx> wrote: > >> > >> From: Vlastimil Babka <vbabka@xxxxxxx> > >> Subject: mm, thp: tweak reclaim/compaction effort of local-only and all-node allocations > > > > I absolutely _detest_ how we've had this pattern of trying to change > > the THP logic in a late -rc game. > > > > Considering how unsuccessful some of the earlier attempts have been, > > late -rc kernels really are *not* the time to make changes like this. > > Agreed, myself would have been much more comfortable in rc1. Yeah I do agree. The patch is a result of a regression I have reported back in early October (http://lkml.kernel.org/r/20191001083743.GC15624@xxxxxxxxxxxxxx). The workload is quite trivial albeit artificial which doesn't make it super urgent. Still something to have addressed though. It has been mostly ignored for quite some time until Vlastimil came with the patch which improved the situation while not causing any other obvious problems (http://lkml.kernel.org/r/20191029151549.GO31513@xxxxxxxxxxxxxx). It has been sitting in the mmotm tree since Oct 29 and in linux-next around a similar time without any reports. 5.5-rc1 (Dec 8) would have been a good time to target but I understand that a more time in linux-next wouldn't hurt and targeting 5.6-rc1 should be sufficient. -- Michal Hocko SUSE Labs
