Re: [PATCH] mark_page_accessed() in do_swap_page() move latter than memcg charge

[Balbir Singh <balbir@linux.vnet.ibm.com>]

* KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> [2009-01-15 20:08:36]:

> 
> sorry for late responce.
> 
> > > In this case we've hit a case where the page is valid and the pc is
> > > not. This does fix the problem, but won't this impact us getting
> > > correct reclaim stats and thus indirectly impact the working of
> > > pressure?
> > > 
> >  - If retruns NULL, only global LRU's status is updated. 
> > 
> > Because this page is not belongs to any memcg, we cannot update
> > any counters. But yes, your point is a concern.
> > 
> > Maybe moving acitvate_page() to
> > ==
> > do_swap_page()
> > {
> >     
> > - activate_page()
> >    mem_cgroup_try_charge()..
> >    ....
> >    mem_cgroup_commit_charge()....
> >    ....
> > +  activate_page()   
> > }
> > ==
> > is necessary. How do you think, kosaki ?
> 
> 
> OK. it makes sense. and my test found no bug.
> 
> ==
> 
> mark_page_accessed() update reclaim_stat statics.
> but currently, memcg charge is called after mark_page_accessed().
> 
> then, mark_page_accessed() don't update memcg statics correctly.
> 
> fixing here.
> 

Changelog needs to be a bit more elaborate, may be talk about invalid
pointer we hit if mark_page_accessed is not moved, etc.

> Signed-off-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com>
> Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> Cc: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
> Cc: Balbir Singh <balbir@linux.vnet.ibm.com>
> 
> ---
>  mm/memory.c |    4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> Index: b/mm/memory.c
> ===================================================================
> --- a/mm/memory.c
> +++ b/mm/memory.c
> @@ -2426,8 +2426,6 @@ static int do_swap_page(struct mm_struct
>  		count_vm_event(PGMAJFAULT);
>  	}
> 
> -	mark_page_accessed(page);
> -
>  	lock_page(page);
>  	delayacct_clear_flag(DELAYACCT_PF_SWAPIN);
> 
> @@ -2480,6 +2478,8 @@ static int do_swap_page(struct mm_struct
>  		try_to_free_swap(page);
>  	unlock_page(page);
> 
> +	mark_page_accessed(page);
> +
>  	if (write_access) {
>  		ret |= do_wp_page(mm, vma, address, page_table, pmd, ptl, pte);
>  		if (ret & VM_FAULT_ERROR)
> 
> 
>

Looks good to me otherwise

Reviewed-by: Balbir Singh <balbir@linux.vnet.ibm.com> 

-- 
	Balbir

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>