Re: [PATCH -mmotm-2009-12-10-17-19] Fix wrong rss count of smaps

[KAMEZAWA Hiroyuki]

On Mon, 28 Dec 2009 14:31:54 +0900
Minchan Kim <minchan.kim@xxxxxxxxx> wrote:
> 
> > BTW, how about counting ZERO page in smaps? Ignoring them completely sounds
> > not very good.
> 
> I am not use it is useful. 
> 
> zero page snapshot of ongoing process is useful?
> Doesn't Admin need to know about zero page?
> Let's admins use it well. If we remove zero page again?
> How many are applications use smaps? 
> Did we have a problem without it?
> 
My concern is that hiding indormation which was exported before.
No more than that and no strong demand.


> When I think of it, there are too many qeustions. 
> Most important thing to add new statistics is just need of customer. 
> 
> Frankly speaking, I don't have good scenario of using zero page.
> Do you have any scenario it is valueable?
> 
read before write ? maybe sometimes happens.

For example. current glibc's calloc() avoids memset() if the pages are
dropped by MADVISE (without unmap). 

Before starting zero-page works, I checked "questions" in lkml and
found some reports that some applications start to go OOM after zero-page
removal.

For me, I know one of my customer's application depends on behavior of
zero page (on RHEL5). So, I tried to add again it before RHEL6 because
I think removal of zero-page corrupts compatibility.

Thanks,
-Kame

--
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/