[PATCH][mmotm-2010-02-01-16-25] Fix wrong accouting of anon and file

[Minchan Kim]

Unfortunately, Kame said he doesn't support this series.
I am not sure we need this patch or revert patch.

Who need this?

David. Do you want to remain this patch in mmotm for your OOM patch
in future?

If anyone doesn't reply my question, Do we have to make revert patch?

== CUT_HERE ==

mm-count-lowmem-rss.patch added lowmem accouting.
But it changed file and rss accouting by mistake.
Unfortunately my review also doesn't found it.

This patch fixes it.

Signed-off-by: Minchan Kim <minch...@gmail.com>
---
mm/memory.c | 8 ++++----
1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index 3becdc3..ce8ff9d 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -210,14 +210,14 @@ void sync_mm_rss(struct task_struct *task, struct mm_struct *mm)

unsigned long get_file_rss(struct mm_struct *mm)
{
- return get_mm_counter(mm, MM_ANONPAGES)
- + get_mm_counter(mm, MM_ANON_LOWPAGES);
+ return get_mm_counter(mm, MM_FILEPAGES)
+ + get_mm_counter(mm, MM_FILE_LOWPAGES);
}

unsigned long get_anon_rss(struct mm_struct *mm)
{
- return get_mm_counter(mm, MM_FILEPAGES)
- + get_mm_counter(mm, MM_FILE_LOWPAGES);
+ return get_mm_counter(mm, MM_ANONPAGES)
+ + get_mm_counter(mm, MM_ANON_LOWPAGES);
}

unsigned long get_low_rss(struct mm_struct *mm)
--
1.6.5

--
Kind regards,
Minchan Kim

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

[David Rientjes]

On Thu, 4 Feb 2010, Minchan Kim wrote:

> Unfortunately, Kame said he doesn't support this series.
> I am not sure we need this patch or revert patch.
>

Nobody is currently using it and it adds an overhead just by doing the
extra branches, so I'd be inclined to drop mm-count-lowmem-rss.patch from
-mm and then reintroduce it later if something needs it down the line.

> Who need this?
>
> David. Do you want to remain this patch in mmotm for your OOM patch
> in future?
>

We'll need to do something for lowmem ooms so that we aren't needlessly
killing taks that don't consume it. At this point, I think it's better to
just fail GFP_DMA allocations where direct reclaim (and, later, memory
compaction) has failed unless it is __GFP_NOFAIL, which none of them are.
So this would be a change to the page allocator to defer the oom killer
and return NULL on GFP_DMA instead of needlessly killing tasks.

> If anyone doesn't reply my question, Do we have to make revert patch?
>

We won't need a revert patch, Andrew will be able to simply drop
mm-count-lowmem-rss.patch from -mm.