Re: [PATCH 14/14] page-replace-kswapd-incmin.patch

From: Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>
Date: Sat, 31 Dec 2005 10:40:40 +0100

On Fri, 2005-12-30 at 23:15 -0200, Marcelo Tosatti wrote:
> On Fri, Dec 30, 2005 at 11:42:34PM +0100, Peter Zijlstra wrote:
> >
> > From: Nick Piggin <npiggin@xxxxxxx>
> >
> > Explicitly teach kswapd about the incremental min logic instead of just scanning
> > all zones under the first low zone. This should keep more even pressure applied
> > on the zones.
> >
> > The new shrink_zone() logic exposes the very worst side of the current
> > balance_pgdat() function. Without this patch reclaim is limited to ZONE_DMA.
>
> Can you please describe the issue with over protection of DMA zone you experienced?
>
> I'll see if I can reproduce it with Nick's standalone patch on top of vanilla, what
> load was that?

With the mdb bench the following behaviour was observed:
(mem=128M)

- PageCache would fill zone_normal
- PageCache would fill zone_dma
- reclaim starts
- initially things look right
- after a while zone_dma is reclaimed so fast that it frequently gets a
full eviction (nr_resident == 0).
- from this point onward zone_normal practiaclly sits idle and zone_dma
goes wild with all the action.

--
Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

References:
- [PATCH] vm: page-replace and clockpro
  - From: Peter Zijlstra
- [PATCH 14/14] page-replace-kswapd-incmin.patch
  - From: Peter Zijlstra
- Re: [PATCH 14/14] page-replace-kswapd-incmin.patch
  - From: Marcelo Tosatti

Prev by Date: Re: [PATCH] Fix user data corrupted by old value return of sysctl
Next by Date: Re: [PATCH] Fix user data corrupted by old value return of sysctl
Previous by thread: Re: [PATCH 14/14] page-replace-kswapd-incmin.patch
Next by thread: [PATCH 3/9] clockpro-PG_test.patch
Index(es):
- Date
- Thread

Relevant Pages

Re: Linux-VServer example results for sharing vs. separate mappings ...
... memory hunks and process groups with cpusets. ... is mostly decoupled from the
present "hardware zones". ... A container uses a software zone. ... Or change
the reclaim code ... (Linux-Kernel)
Re: shrink_all_memory tweaks (was: Re: Userland swsusp failure (mm-related))
... shrinking of the active list until we decide it's worth it) -- or we can ...
experence as much as shrinking the active/inactive lists. ... they are most likely to have
enough pages to reclaim. ... balance_pgdatwill work across all this node's zones
until ... (Linux-Kernel)
[PATCH -mm] swsusp: rework memory shrinker (rev. 2)
... Rework the swsusp's memory shrinker in the following way: ... inactive lists
are next (the zones with more inactive pages go first) etc. ... Each time shrink_all_memoryattempts
to shrink the active and inactive lists ... The aim of this is to alter the reclaim
logic to choose the best pages to keep ... (Linux-Kernel)
[RFC][PATCH -mm] swsusp: reworked memory shrinker
... inactive lists are next (the zones with more inactive pages go first) etc. ...
Each time shrink_all_memoryattempts to shrink the active and inactive lists ... The aim
of this is to alter the reclaim logic to choose the best pages to keep ... int all_zones_ok;
... (Linux-Kernel)
[PATCH -mm] swsusp: rework memory shrinker
... inactive lists are next (the zones with more inactive pages go first) etc. ...
Each time shrink_all_memoryattempts to shrink the active and inactive lists ... The aim
of this is to alter the reclaim logic to choose the best pages to keep ... int all_zones_ok;
... (Linux-Kernel)