On Thu, Feb 20, 2014 at 10:51:44AM +0100, Michal Hocko wrote:
> We had a report about strange OOM killer strikes on a PPC machine
> although there was a lot of swap free and a tons of anonymous memory
> which could be swapped out. In the end it turned out that the OOM was
> a side effect of zone re
On 21.02.2014 [14:07:35 -0800], Andrew Morton wrote:
> On Thu, 20 Feb 2014 10:51:44 +0100 Michal Hocko wrote:
>
> > We had a report about strange OOM killer strikes on a PPC machine
> > although there was a lot of swap free and a tons of anonymous memory
> > which could be swapped out. In the end
On Thu, 20 Feb 2014 10:51:44 +0100 Michal Hocko wrote:
> We had a report about strange OOM killer strikes on a PPC machine
> although there was a lot of swap free and a tons of anonymous memory
> which could be swapped out. In the end it turned out that the OOM was
> a side effect of zone reclaim
On Thu 20-02-14 10:51:44, Michal Hocko wrote:
> We had a report about strange OOM killer strikes on a PPC machine
> although there was a lot of swap free and a tons of anonymous memory
> which could be swapped out. In the end it turned out that the OOM was
> a side effect of zone reclaim which wasn
We had a report about strange OOM killer strikes on a PPC machine
although there was a lot of swap free and a tons of anonymous memory
which could be swapped out. In the end it turned out that the OOM was
a side effect of zone reclaim which wasn't doesn't unmap and swapp out
and so the system was p
On Wed 19-02-14 15:05:58, Nishanth Aravamudan wrote:
> On 19.02.2014 [13:56:00 -0800], David Rientjes wrote:
> > On Wed, 19 Feb 2014, Nishanth Aravamudan wrote:
> >
> > > > diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> > > > index 3e953f07edb0..4a44bdc7a8cf 100644
> > > > --- a/mm/page_alloc.c
On 19.02.2014 [13:56:00 -0800], David Rientjes wrote:
> On Wed, 19 Feb 2014, Nishanth Aravamudan wrote:
>
> > > diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> > > index 3e953f07edb0..4a44bdc7a8cf 100644
> > > --- a/mm/page_alloc.c
> > > +++ b/mm/page_alloc.c
> > > @@ -1855,7 +1855,7 @@ static vo
On Wed, 19 Feb 2014, Nishanth Aravamudan wrote:
> > diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> > index 3e953f07edb0..4a44bdc7a8cf 100644
> > --- a/mm/page_alloc.c
> > +++ b/mm/page_alloc.c
> > @@ -1855,7 +1855,7 @@ static void __paginginit init_zone_allows_reclaim(int
> > nid)
> > {
> >
On Wed 19-02-14 09:49:41, Nishanth Aravamudan wrote:
> On 19.02.2014 [18:32:59 +0100], Michal Hocko wrote:
> > On Wed 19-02-14 09:16:28, Nishanth Aravamudan wrote:
[...]
> > > I don't think this will work, because what sets N_HIGH_MEMORY (and
> > > shouldn't it be N_MEMORY?)
> >
> > This should be
On 19.02.2014 [18:03:03 +0100], Michal Hocko wrote:
> We had a report about strange OOM killer strikes on a PPC machine
> although there was a lot of swap free and a tons of anonymous memory
> which could be swapped out. In the end it turned out that the OOM was
> a side effect of zone reclaim whic
On 19.02.2014 [18:32:59 +0100], Michal Hocko wrote:
> On Wed 19-02-14 09:16:28, Nishanth Aravamudan wrote:
> > On 19.02.2014 [18:03:03 +0100], Michal Hocko wrote:
> > > We had a report about strange OOM killer strikes on a PPC machine
> > > although there was a lot of swap free and a tons of anonym
On Wed 19-02-14 09:16:28, Nishanth Aravamudan wrote:
> On 19.02.2014 [18:03:03 +0100], Michal Hocko wrote:
> > We had a report about strange OOM killer strikes on a PPC machine
> > although there was a lot of swap free and a tons of anonymous memory
> > which could be swapped out. In the end it tur
On 19.02.2014 [18:03:03 +0100], Michal Hocko wrote:
> We had a report about strange OOM killer strikes on a PPC machine
> although there was a lot of swap free and a tons of anonymous memory
> which could be swapped out. In the end it turned out that the OOM was
> a side effect of zone reclaim whic
We had a report about strange OOM killer strikes on a PPC machine
although there was a lot of swap free and a tons of anonymous memory
which could be swapped out. In the end it turned out that the OOM was
a side effect of zone reclaim which wasn't doesn't unmap and swapp out
and so the system was p
14 matches
Mail list logo