On Thu, 19 Nov 2015 13:30:53 +0100 Michal Hocko <mhocko@xxxxxxxxxx> wrote:
> From: Michal Hocko <mhocko@xxxxxxxx>
>
> Tetsuo Handa has reported that the system might basically livelock in OOM
> condition without triggering the OOM killer. The issue is caused by
> internal dependency of the direct reclaim on vmstat counter updates (via
> zone_reclaimable) which are performed from the workqueue context.
> If all the current workers get assigned to an allocation request,
> though, they will be looping inside the allocator trying to reclaim
> memory but zone_reclaimable can see stalled numbers so it will consider
> a zone reclaimable even though it has been scanned way too much. WQ
> concurrency logic will not consider this situation as a congested workqueue
> because it relies that worker would have to sleep in such a situation.
> This also means that it doesn't try to spawn new workers or invoke
> the rescuer thread if the one is assigned to the queue.
>
> In order to fix this issue we need to do two things. First we have to
> let wq concurrency code know that we are in trouble so we have to do
> a short sleep. In order to prevent from issues handled by 0e093d99763e
> ("writeback: do not sleep on the congestion queue if there are no
> congested BDIs or if significant congestion is not being encountered in
> the current zone") we limit the sleep only to worker threads which are
> the ones of the interest anyway.
>
> The second thing to do is to create a dedicated workqueue for vmstat and
> mark it WQ_MEM_RECLAIM to note it participates in the reclaim and to
> have a spare worker thread for it.
This vmstat update thing is being a problem. Please see Joonsoo's
"mm/vmstat: retrieve more accurate vmstat value".
Joonsoo, might this patch help with that issue?
>
> The original issue reported by Tetsuo [1] has seen multiple attempts for
> a fix. The easiest one being [2] which was targeted to the particular
> problem. There was a more general concern that looping inside the
> allocator without ever sleeping breaks the basic assumption of worker
> concurrency logic so the fix should be more general. Another attempt [3]
> therefore added a short (1 jiffy) sleep into the page allocator. This
> would, however, introduce sleeping for all callers of the page allocator
> which is not really needed. This patch tries to be a compromise and
> introduce sleeping only where it matters - for kworkers.
>
> Even though we haven't seen bug reports in the past I would suggest
> backporting this to the stable trees. The issue is present since we have
> stopped useing congestion_wait in the retry loop because WQ concurrency
> is older as well as vmstat worqueue based refresh AFAICS.
hm, I'm reluctant. If the patch fixes something that real people are
really hurting from then yes. But I suspect this is just one fly-swat
amongst many.
--
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>