Re: [PATCH v2 3/3] vmscan: implement swap token priority aging

[KOSAKI Motohiro]

Hi,

(2011/05/21 4:30), Andrew Morton wrote:
> On Thu, 19 May 2011 11:34:15 +0900
> KOSAKI Motohiro<kosaki.motohiro@xxxxxxxxxxxxxx>  wrote:
>
> > While testing for memcg aware swap token, I observed a swap token
> > was often grabbed an intermittent running process (eg init, auditd)
> > and they never release a token.
> >
> > Why?
> >
> > Some processes (eg init, auditd, audispd) wake up when a process
> > exiting. And swap token can be get first page-in process when
> > a process exiting makes no swap token owner. Thus such above
> > intermittent running process often get a token.
> >
> > And currently, swap token priority is only decreased at page fault
> > path. Then, if the process sleep immediately after to grab swap
> > token, the swap token priority never be decreased. That's obviously
> > undesirable.
> >
> > This patch implement very poor (and lightweight) priority aging.
> > It only be affect to the above corner case and doesn't change swap
> > tendency workload performance (eg multi process qsbench load)
> >
> > ...
> >
> > --- a/mm/thrash.c
> > +++ b/mm/thrash.c
> > @@ -25,10 +25,13 @@
> >
> >   #include<trace/events/vmscan.h>
> >
> > +#define TOKEN_AGING_INTERVAL	(0xFF)
>
> Needs a comment describing its units and what it does, please.
> Sufficient for readers to understand why this value was chosen and what
> effect they could expect to see from changing it.

Will do.


> >   static DEFINE_SPINLOCK(swap_token_lock);
> >   struct mm_struct *swap_token_mm;
> >   struct mem_cgroup *swap_token_memcg;
> >   static unsigned int global_faults;
> > +static unsigned int last_aging;
>
> Is this a good name?  Would something like prev_global_faults be better?

Current code is "next-aging = last-aging + 256global-fault". so,
prev_global_faults is slightly misleading to me.


> `global_faults' and `last_aging' could be made static local in
> grab_swap_token().

OK. even though I don't like static in a function.


> >   void grab_swap_token(struct mm_struct *mm)
> >   {
> > @@ -47,6 +50,11 @@ void grab_swap_token(struct mm_struct *mm)
> >   	if (!swap_token_mm)
> >   		goto replace_token;
> >
> > +	if ((global_faults - last_aging)>  TOKEN_AGING_INTERVAL) {
> > +		swap_token_mm->token_priority /= 2;
> > +		last_aging = global_faults;
> > +	}
>
> It's really hard to reverse-engineer the design decisions from the
> implementation here, therefore...  ?
>
> >   	if (mm == swap_token_mm) {
> >   		mm->token_priority += 2;
> >   		goto update_priority;
> > @@ -64,7 +72,7 @@ void grab_swap_token(struct mm_struct *mm)
> >   		goto replace_token;
> >
> >   update_priority:
> > -	trace_update_swap_token_priority(mm, old_prio);
> > +	trace_update_swap_token_priority(mm, old_prio, swap_token_mm);
> >
> >   out:
> >   	mm->faultstamp = global_faults;
> > @@ -80,6 +88,7 @@ replace_token:
> >   	trace_replace_swap_token(swap_token_mm, mm);
> >   	swap_token_mm = mm;
> >   	swap_token_memcg = memcg;
> > +	last_aging = global_faults;
> >   	goto out;
> >   }
>
> In fact all of grab_swap_token() and the thrash-detection code in
> general are pretty tricky and unobvious stuff.  So we left it
> undocumented :(

Well, original swap token (designed by Rik) is pretty straight forward
implementation on the theory. Therefore we didn't need too verbose doc.
The paper give us good well documentation.

# Oh, now http://www.cs.wm.edu/~sjiang/token.pdf is dead link.
# we should fix it to http://www.cse.ohio-state.edu/hpcs/WWW/HTML/publications/papers/TR-05-1.pdf,
# maybe, or do anybody know better url?

But following commit rewrite almost all code. and we lost good documentation.
It's a bit sad.


commit 7602bdf2fd14a40dd9b104e516fdc05e1bd17952
Author: Ashwin Chaugule <ashwin.chaugule@xxxxxxxxxxxx>
Date:   Wed Dec 6 20:31:57 2006 -0800

    [PATCH] new scheme to preempt swap token





--
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/