Re: [PATCH -mmotm 6/8] memcg: recharge charges of anonymous page

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

On Fri, 6 Nov 2009 14:14:48 +0900
Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp> wrote:

> This patch is the core part of this recharge-at-task-move feature.
> It implements functions to recharge charges of anonymous pages mapped only by
> the target task.
> 
> Implementation:
> - define struct recharge_struct and a valuable of it(recharge) to remember
>   the count of pre-charges and other information.
> - At can_attach(), parse the page table of the task and count the number of
>   mapped pages which are charged to the source mem_cgroup, and call
>   __mem_cgroup_try_charge() repeatedly and count up recharge.precharge.
> - At attach(), parse the page table again, find a target page as we did in
>   can_attach(), and call mem_cgroup_move_account() about the page.
> - Cancel all charges if recharge.precharge > 0 on failure or at the end of
>   task move.
> 

Changelog ?


> Signed-off-by: Daisuke Nishimura <nishimura@mxp.nes.nec.co.jp>
> ---
>  Documentation/cgroups/memory.txt |   36 +++++-
>  mm/memcontrol.c                  |  275 +++++++++++++++++++++++++++++++++++++-
>  2 files changed, 306 insertions(+), 5 deletions(-)
> 
> diff --git a/Documentation/cgroups/memory.txt b/Documentation/cgroups/memory.txt
> index b871f25..54281ff 100644
> --- a/Documentation/cgroups/memory.txt
> +++ b/Documentation/cgroups/memory.txt
> @@ -262,10 +262,12 @@ some of the pages cached in the cgroup (page cache pages).
>  4.2 Task migration
>  
>  When a task migrates from one cgroup to another, it's charge is not
> -carried forward. The pages allocated from the original cgroup still
> +carried forward by default. The pages allocated from the original cgroup still
>  remain charged to it, the charge is dropped when the page is freed or
>  reclaimed.
>  
> +Note: You can move charges of a task along with task migration. See 8.
> +
>  4.3 Removing a cgroup
>  
>  A cgroup can be removed by rmdir, but as discussed in sections 4.1 and 4.2, a
> @@ -414,7 +416,37 @@ NOTE1: Soft limits take effect over a long period of time, since they involve
>  NOTE2: It is recommended to set the soft limit always below the hard limit,
>         otherwise the hard limit will take precedence.
>  
> -8. TODO
> +8. Recharge at task move
> +
> +Users can move charges associated with a task along with task move, that is,
> +uncharge from the old cgroup and charge to the new cgroup.
> +
> +8.1 Interface
> +
> +This feature is disabled by default. It can be enabled(and disabled again) by
> +writing to memory.recharge_at_immigrate of the destination cgroup.
> +
> +If you want to enable it
> +
> +# echo 1 > memory.recharget_at_immigrate
> +
> +Note: A value more than 1 will be supported in futer. See 8.2.
> +
> +And if you want disable it again
> +
> +# echo 0 > memory.recharget_at_immigrate
> +
> +8.2 Type of charges which can be recharged
> +
> +We recharge a charge which meets the following conditions.
> +
> +a. It must be charged to the old cgroup.
> +b. A charge of an anonymous page used by the target task. The page must be used
> +   only by the target task.
> +
> +Note: More type of pages(e.g. file cache, shmem,) will be supported in future.
> +
> +9. TODO
>  
>  1. Add support for accounting huge pages (as a separate controller)
>  2. Make per-cgroup scanner reclaim not-shared pages first
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index afa1179..f4b7116 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -21,6 +21,8 @@
>  #include <linux/memcontrol.h>
>  #include <linux/cgroup.h>
>  #include <linux/mm.h>
> +#include <linux/migrate.h>
> +#include <linux/hugetlb.h>
>  #include <linux/pagemap.h>
>  #include <linux/smp.h>
>  #include <linux/page-flags.h>
> @@ -239,6 +241,18 @@ struct mem_cgroup {
>  };
>  
>  /*
> + * A data structure and a valiable for recharging charges at task move.
> + * "recharge" and its members are protected by cgroup_lock
> + */
> +struct recharge_struct {
> +	struct mem_cgroup *from;
> +	struct mem_cgroup *to;
> +	struct task_struct *target;	/* the target task being moved */
> +	unsigned long precharge;
> +};
> +static struct recharge_struct recharge;
> +
> +/*
>   * Maximum loops in mem_cgroup_hierarchical_reclaim(), used for soft
>   * limit reclaim to prevent infinite loops, if they ever occur.
>   */
> @@ -1496,7 +1510,7 @@ charged:
>  	 * Insert ancestor (and ancestor's ancestors), to softlimit RB-tree.
>  	 * if they exceeds softlimit.
>  	 */
> -	if (mem_cgroup_soft_limit_check(mem))
> +	if (page && mem_cgroup_soft_limit_check(mem))
>  		mem_cgroup_update_tree(mem, page);
>  done:
>  	return 0;
> @@ -3416,10 +3430,170 @@ static int mem_cgroup_populate(struct cgroup_subsys *ss,
>  }
>  
>  /* Handlers for recharge at task move. */
> +/**
> + * is_target_pte_for_recharge - check a pte whether it is valid for recharge
> + * @vma: the vma the pte to be checked belongs
> + * @addr: the address corresponding to the pte to be checked
> + * @ptent: the pte to be checked
> + * @target: the pointer the target page will be stored(can be NULL)
> + *
> + * Returns
> + *   0(RECHARGE_TARGET_NONE): if the pte is not a target for recharge.
> + *   1(RECHARGE_TARGET_PAGE): if the page corresponding to this pte is a target
> + *     for recharge. if @target is not NULL, the page is stored in target->page
> + *     with extra refcnt got(Callers should handle it).
> + *
> + * Called with pte lock held.
> + */
> +/* We add a new member later. */
> +union recharge_target {
> +	struct page	*page;
> +};
> +
> +/* We add a new type later. */
> +enum recharge_target_type {
> +	RECHARGE_TARGET_NONE,	/* not used */
> +	RECHARGE_TARGET_PAGE,
> +};
> +
> +static int is_target_pte_for_recharge(struct vm_area_struct *vma,
> +		unsigned long addr, pte_t ptent, union recharge_target *target)
> +{
> +	struct page *page;
> +	struct page_cgroup *pc;
> +	int ret = 0;
> +
> +	if (!pte_present(ptent))
> +		return 0;
> +
> +	page = vm_normal_page(vma, addr, ptent);
> +	if (!page || !page_mapped(page))
> +		return 0;
> +	/* TODO: We don't recharge file(including shmem/tmpfs) pages for now. */
> +	if (!PageAnon(page))
> +		return 0;
> +	/*
> +	 * TODO: We don't recharge shared(used by multiple processes) pages
> +	 * for now.
> +	 */
> +	if (page_mapcount(page) > 1)
> +		return 0;
> +	if (!get_page_unless_zero(page))
> +		return 0;
> +
> +	pc = lookup_page_cgroup(page);
> +	lock_page_cgroup(pc);
> +	if (PageCgroupUsed(pc) && pc->mem_cgroup == recharge.from) {
> +		ret = RECHARGE_TARGET_PAGE;
> +		if (target)
> +			target->page = page;
> +	}
> +	unlock_page_cgroup(pc);
> +
> +	if (!ret || !target)
> +		put_page(page);
> +
> +	return ret;
> +}
> +
> +static int mem_cgroup_recharge_do_precharge(void)
> +{
> +	int ret = -ENOMEM;
> +	struct mem_cgroup *mem = recharge.to;
> +
> +	ret = __mem_cgroup_try_charge(NULL, GFP_KERNEL, &mem, false, NULL);
> +	if (ret || !mem)
> +		return -ENOMEM;
> +
> +	recharge.precharge++;
> +	return ret;
> +}
> +
> +static int mem_cgroup_recharge_prepare_pte_range(pmd_t *pmd,
> +					unsigned long addr, unsigned long end,
> +					struct mm_walk *walk)
> +{
> +	int ret = 0;
> +	unsigned long count = 0;
> +	struct vm_area_struct *vma = walk->private;
> +	pte_t *pte;
> +	spinlock_t *ptl;
> +
> +	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
> +	for (; addr != end; pte++, addr += PAGE_SIZE)
> +		if (is_target_pte_for_recharge(vma, addr, *pte, NULL))
> +			count++;
> +	pte_unmap_unlock(pte - 1, ptl);
> +
> +	while (count-- && !ret)
> +		ret = mem_cgroup_recharge_do_precharge();
> +
> +	return ret;
> +}
> +
> +static int mem_cgroup_recharge_prepare(void)
> +{
> +	int ret = 0;
> +	struct mm_struct *mm;
> +	struct vm_area_struct *vma;
> +
> +	mm = get_task_mm(recharge.target);
> +	if (!mm)
> +		return 0;
> +
> +	down_read(&mm->mmap_sem);
> +	for (vma = mm->mmap; vma; vma = vma->vm_next) {
> +		struct mm_walk mem_cgroup_recharge_prepare_walk = {
> +			.pmd_entry = mem_cgroup_recharge_prepare_pte_range,
> +			.mm = mm,
> +			.private = vma,
> +		};
> +		if (signal_pending(current)) {
> +			ret = -EINTR;
> +			break;
> +		}
> +		if (is_vm_hugetlb_page(vma))
> +			continue;
> +		ret = walk_page_range(vma->vm_start, vma->vm_end,
> +					&mem_cgroup_recharge_prepare_walk);
> +		if (ret)
> +			break;
> +		cond_resched();
> +	}
> +	up_read(&mm->mmap_sem);
> +
> +	mmput(mm);
> +	return ret;
> +}
> +
> +static void mem_cgroup_clear_recharge(void)
> +{
> +	while (recharge.precharge--)
> +		mem_cgroup_cancel_charge(recharge.to);
> +	recharge.from = NULL;
> +	recharge.to = NULL;
> +	recharge.target = NULL;
> +}
> +
>  static int mem_cgroup_can_recharge(struct mem_cgroup *mem,
>  					struct task_struct *p)
>  {
> -	return 0;
> +	int ret;
> +	struct mem_cgroup *from = mem_cgroup_from_task(p);
> +
> +	if (from == mem)
> +		return 0;
> +
> +	recharge.from = from;
> +	recharge.to = mem;
> +	recharge.target = p;
> +	recharge.precharge = 0;
> +
> +	ret = mem_cgroup_recharge_prepare();
> +
> +	if (ret)
> +		mem_cgroup_clear_recharge();
> +	return ret;
>  }
>  

Hmm...Hmm...looks nicer. But can I have another suggestion ?

==
static int mem_cgroup_can_recharge(struct mem_cgroup *mem,
				struct task_struct *p)
{
	unsigned long rss;
	struct mm_struct *mm;

	mm = get_task_mm(p);
	if (!mm)
		return;

	rss = get_mm_counter(mm, anon_rss);
	precharge(rss);
	mmput(mm);
}
==
Do you think anonymous memory are so shared at "move" as that
we need page table scan ?

If typical sequence is
==
	fork()
		-> exec()

	move child
==
No problem will happen.

>  static int mem_cgroup_can_attach(struct cgroup_subsys *ss,
> @@ -3442,11 +3616,104 @@ static void mem_cgroup_cancel_attach(struct cgroup_subsys *ss,
>  				struct task_struct *p,
>  				bool threadgroup)
>  {
> +	struct mem_cgroup *mem = mem_cgroup_from_cont(cgroup);
> +
>  	mutex_unlock(&memcg_tasklist);
> +	if (mem->recharge_at_immigrate && thread_group_leader(p))
> +		mem_cgroup_clear_recharge();
> +}
> +
> +static int mem_cgroup_recharge_pte_range(pmd_t *pmd,
> +				unsigned long addr, unsigned long end,
> +				struct mm_walk *walk)
> +{
> +	int ret = 0;
> +	struct vm_area_struct *vma = walk->private;
> +	pte_t *pte;
> +	spinlock_t *ptl;
> +
> +retry:
> +	pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
> +	for (; addr != end; addr += PAGE_SIZE) {
> +		pte_t ptent = *(pte++);
> +		union recharge_target target;
> +		int type;
> +		struct page *page;
> +		struct page_cgroup *pc;
> +
> +		if (!recharge.precharge)
> +			break;
> +
> +		type = is_target_pte_for_recharge(vma, addr, ptent, &target);
> +		switch (type) {
> +		case RECHARGE_TARGET_PAGE:
> +			page = target.page;
> +			if (isolate_lru_page(page))
> +				goto put;
> +			pc = lookup_page_cgroup(page);
> +			if (!mem_cgroup_move_account(pc,
> +						recharge.from, recharge.to)) {
> +				css_put(&recharge.to->css);
> +				recharge.precharge--;
> +			}
> +			putback_lru_page(page);
> +put:			/* is_target_pte_for_recharge() gets the page */
> +			put_page(page);
> +			break;
> +		default:
> +			continue;

continue for what ?

And we forget "failed to move" pte. This "move" is best-effort service.
Right ?

> +		}
> +	}
> +	pte_unmap_unlock(pte - 1, ptl);
> +
> +	if (addr != end) {
> +		/*
> +		 * We have consumed all precharges we got in can_attach().
> +		 * We try precharge one by one, but don't do any additional
> +		 * precharges nor recharges to recharge.to if we have failed in
> +		 * precharge once in attach() phase.
> +		 */
> +		ret = mem_cgroup_recharge_do_precharge();
> +		if (!ret)
> +			goto retry;
> +	}
> +
> +	return ret;
>  }
>  



>  static void mem_cgroup_recharge(void)
>  {
> +	struct mm_struct *mm;
> +	struct vm_area_struct *vma;
> +
> +	mm = get_task_mm(recharge.target);
> +	if (!mm)
> +		return;
> +
> +	lru_add_drain_all();
> +	down_read(&mm->mmap_sem);
> +	for (vma = mm->mmap; vma; vma = vma->vm_next) {
> +		int ret;
> +		struct mm_walk mem_cgroup_recharge_walk = {
> +			.pmd_entry = mem_cgroup_recharge_pte_range,
> +			.mm = mm,
> +			.private = vma,
> +		};
> +		if (is_vm_hugetlb_page(vma))
> +			continue;
> +		ret = walk_page_range(vma->vm_start, vma->vm_end,
> +						&mem_cgroup_recharge_walk);

At _this_ point, check VM_SHARED and skip scan is a sane operation.
Could you add checks ?


> +		if (ret)
> +			/*
> +			 * means we have consumed all precharges and failed in
> +			 * doing additional precharge. Just abandon here.
> +			 */
> +			break;
> +		cond_resched();
> +	}
> +	up_read(&mm->mmap_sem);
> +
> +	mmput(mm);
>  }
>  
>  static void mem_cgroup_move_task(struct cgroup_subsys *ss,
> @@ -3458,8 +3725,10 @@ static void mem_cgroup_move_task(struct cgroup_subsys *ss,
>  	struct mem_cgroup *mem = mem_cgroup_from_cont(cont);
>  
>  	mutex_unlock(&memcg_tasklist);
> -	if (mem->recharge_at_immigrate && thread_group_leader(p))
> +	if (mem->recharge_at_immigrate && thread_group_leader(p)) {
>  		mem_cgroup_recharge();
> +		mem_cgroup_clear_recharge();
> +	}

Is it guranteed that thread_group_leader(p) is true if this is true at
can_attach() ?
If no,
	if (.....) {
		mem_cgroup_recharge()
	}
	mem_cgroup_cleare_recharge()

is better.


Thanks,
-Kame

--
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>