Re: [PATCH] memcg: reduce lock time at move charge (Was Re: [PATCH 04/10] memcg: disable local interrupts in lock_page_cgroup()

[KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>]

On Thu, 7 Oct 2010 16:28:11 +0900
Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> wrote:

> On Thu, 7 Oct 2010 15:21:11 +0900
> KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> 
> > On Thu, 7 Oct 2010 11:17:43 +0900
> > KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> >  
> > > > hmm, if we'll do that, I think we need to do that under pte_lock in
> > > > mem_cgroup_move_charge_pte_range(). But, we can't do wait_on_page_writeback()
> > > > under pte_lock, right? Or, we need re-organize current move-charge implementation.
> > > > 
> > > Nice catch. I think releaseing pte_lock() is okay. (and it should be released)
> > > 
> > > IIUC, task's css_set() points to new cgroup when "move" is called. Then,
> > > it's not necessary to take pte_lock, I guess.
> > > (And taking pte_lock too long is not appreciated..)
> > > 
> > > I'll write a sample patch today.
> > > 
> > Here.
> Great!
> 
> > ==
> > From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> > 
> > Now, at task migration among cgroup, memory cgroup scans page table and moving
> > account if flags are properly set.
> > 
> > The core code, mem_cgroup_move_charge_pte_range() does
> > 
> >  	pte_offset_map_lock();
> > 	for all ptes in a page table:
> > 		1. look into page table, find_and_get a page
> > 		2. remove it from LRU.
> > 		3. move charge.
> > 		4. putback to LRU. put_page()
> > 	pte_offset_map_unlock();
> > 
> > for pte entries on a 3rd(2nd) level page table.
> > 
> > This pte_offset_map_lock seems a bit long. This patch modifies a rountine as
> > 
> > 	pte_offset_map_lock()
> > 	for 32 pages:
> > 		      find_and_get a page
> > 		      record it
> > 	pte_offset_map_unlock()
> > 	for all recorded pages
> > 		      isolate it from LRU.
> > 		      move charge
> > 		      putback to LRU
> > 	for all recorded pages
> > 		      put_page()
> > 
> > Note: newly-charged pages while we move account are charged to the new group.
> > 
> > Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> > ---
> >  mm/memcontrol.c |   92 ++++++++++++++++++++++++++++++++++++--------------------
> >  1 file changed, 60 insertions(+), 32 deletions(-)
> > 
> > Index: mmotm-0928/mm/memcontrol.c
> > ===================================================================
> > --- mmotm-0928.orig/mm/memcontrol.c
> > +++ mmotm-0928/mm/memcontrol.c
> > @@ -4475,17 +4475,22 @@ one_by_one:
> >   *
> >   * Called with pte lock held.
> >   */
> > -union mc_target {
> > -	struct page	*page;
> > -	swp_entry_t	ent;
> > -};
> >  
> >  enum mc_target_type {
> > -	MC_TARGET_NONE,	/* not used */
> > +	MC_TARGET_NONE, /* used as failure code(0) */
> >  	MC_TARGET_PAGE,
> >  	MC_TARGET_SWAP,
> >  };
> >  
> > +struct mc_target {
> > +	enum mc_target_type type;
> > +	union {
> > +		struct page	*page;
> > +		swp_entry_t	ent;
> > +	} val;
> > +};
> > +
> > +
> >  static struct page *mc_handle_present_pte(struct vm_area_struct *vma,
> >  						unsigned long addr, pte_t ptent)
> >  {
> > @@ -4561,7 +4566,7 @@ static struct page *mc_handle_file_pte(s
> >  }
> >  
> >  static int is_target_pte_for_mc(struct vm_area_struct *vma,
> > -		unsigned long addr, pte_t ptent, union mc_target *target)
> > +		unsigned long addr, pte_t ptent, struct mc_target *target)
> >  {
> >  	struct page *page = NULL;
> >  	struct page_cgroup *pc;
> > @@ -4587,7 +4592,7 @@ static int is_target_pte_for_mc(struct v
> >  		if (PageCgroupUsed(pc) && pc->mem_cgroup == mc.from) {
> >  			ret = MC_TARGET_PAGE;
> >  			if (target)
> > -				target->page = page;
> > +				target->val.page = page;
> >  		}
> >  		if (!ret || !target)
> >  			put_page(page);
> > @@ -4597,8 +4602,10 @@ static int is_target_pte_for_mc(struct v
> >  			css_id(&mc.from->css) == lookup_swap_cgroup(ent)) {
> >  		ret = MC_TARGET_SWAP;
> >  		if (target)
> > -			target->ent = ent;
> > +			target->val.ent = ent;
> >  	}
> > +	if (target)
> > +		target->type = ret;
> >  	return ret;
> >  }
> >  
> > @@ -4751,6 +4758,9 @@ static void mem_cgroup_cancel_attach(str
> >  	mem_cgroup_clear_mc();
> >  }
> >  
> > +
> > +#define MC_MOVE_ONCE		(32)
> > +
> >  static int mem_cgroup_move_charge_pte_range(pmd_t *pmd,
> >  				unsigned long addr, unsigned long end,
> >  				struct mm_walk *walk)
> > @@ -4759,26 +4769,47 @@ static int mem_cgroup_move_charge_pte_ra
> >  	struct vm_area_struct *vma = walk->private;
> >  	pte_t *pte;
> >  	spinlock_t *ptl;
> > +	struct mc_target *target;
> > +	int index, num;
> > +
> > +	target = kzalloc(sizeof(struct mc_target) *MC_MOVE_ONCE, GFP_KERNEL);
> hmm? I can't see it freed anywhere.
> 
leaked..


> Considering you reset target[]->type to MC_TARGET_NONE, do you intended to
> reuse targe[] while walking the page table ?
yes.

> If so, how about adding a new member(struct mc_target *targe) to move_charge_struct,
> and allocate/free it at mem_cgroup_move_charge() ?
> 
Hmm, sounds nice.

I'll do.

Thanks,
-Kame

> Thanks,
> Daisuke Nishimura.
> 
> > +	if (!target)
> > +		return -ENOMEM;
> >  
> >  retry:
> >  	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
> > -	for (; addr != end; addr += PAGE_SIZE) {
> > +	for (num = 0; num < MC_MOVE_ONCE && addr != end; addr += PAGE_SIZE) {
> >  		pte_t ptent = *(pte++);
> > -		union mc_target target;
> > -		int type;
> > +		ret = is_target_pte_for_mc(vma, addr, ptent, &target[num]);
> > +		if (!ret)
> > +			continue;
> > +		target[num++].type = ret;
> > +	}
> > +	pte_unmap_unlock(pte - 1, ptl);
> > +	cond_resched();
> > +
> > +	ret = 0;
> > +	index = 0;
> > +	do {
> > +		struct mc_target *mt;
> >  		struct page *page;
> >  		struct page_cgroup *pc;
> >  		swp_entry_t ent;
> >  
> > -		if (!mc.precharge)
> > -			break;
> > +		if (!mc.precharge) {
> > +			ret = mem_cgroup_do_precharge(1);
> > +			if (ret)
> > +				goto out;
> > +			continue;
> > +		}
> > +
> > +		mt = &target[index++];
> >  
> > -		type = is_target_pte_for_mc(vma, addr, ptent, &target);
> > -		switch (type) {
> > +		switch (mt->type) {
> >  		case MC_TARGET_PAGE:
> > -			page = target.page;
> > +			page = mt->val.page;
> >  			if (isolate_lru_page(page))
> > -				goto put;
> > +				break;
> >  			pc = lookup_page_cgroup(page);
> >  			if (!mem_cgroup_move_account(pc,
> >  						mc.from, mc.to, false)) {
> > @@ -4787,11 +4818,9 @@ retry:
> >  				mc.moved_charge++;
> >  			}
> >  			putback_lru_page(page);
> > -put:			/* is_target_pte_for_mc() gets the page */
> > -			put_page(page);
> >  			break;
> >  		case MC_TARGET_SWAP:
> > -			ent = target.ent;
> > +			ent = mt->val.ent;
> >  			if (!mem_cgroup_move_swap_account(ent,
> >  						mc.from, mc.to, false)) {
> >  				mc.precharge--;
> > @@ -4802,21 +4831,20 @@ put:			/* is_target_pte_for_mc() gets th
> >  		default:
> >  			break;
> >  		}
> > +	} while (index < num);
> > +out:
> > +	for (index = 0; index < num; index++) {
> > +		if (target[index].type == MC_TARGET_PAGE)
> > +			put_page(target[index].val.page);
> > +		target[index].type = MC_TARGET_NONE;
> >  	}
> > -	pte_unmap_unlock(pte - 1, ptl);
> > +
> > +	if (ret)
> > +		return ret;
> >  	cond_resched();
> >  
> > -	if (addr != end) {
> > -		/*
> > -		 * We have consumed all precharges we got in can_attach().
> > -		 * We try charge one by one, but don't do any additional
> > -		 * charges to mc.to if we have failed in charge once in attach()
> > -		 * phase.
> > -		 */
> > -		ret = mem_cgroup_do_precharge(1);
> > -		if (!ret)
> > -			goto retry;
> > -	}
> > +	if (addr != end)
> > +		goto retry;
> >  
> >  	return ret;
> >  }
> > 
> 

--
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>