* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
@ 2020-01-17 9:31 ` David Rientjes
0 siblings, 0 replies; 26+ messages in thread
From: David Rientjes @ 2020-01-17 9:31 UTC (permalink / raw)
To: Michal Hocko
Cc: Kirill Tkhai, Wei Yang, hannes-druUgvl0LCNAfugRpC6u6w,
vdavydov.dev-Re5JQEeQqe8AvxtiuMwx3w,
akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b,
kirill.shutemov-VuQAYsv1563Yd54FQh9/CA,
yang.shi-KPsoFbNs7GizrGE5bRqYAgC/G2K4zDHf,
cgroups-u79uwXL29TY76Z2rM5mHXA, linux-mm-Bw31MaZKKs3YtjvyW6yDsg,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
alexander.duyck-Re5JQEeQqe8AvxtiuMwx3w,
stable-u79uwXL29TY76Z2rM5mHXA
On Fri, 17 Jan 2020, Michal Hocko wrote:
> On Thu 16-01-20 14:01:59, David Rientjes wrote:
> > On Thu, 16 Jan 2020, Kirill Tkhai wrote:
> >
> > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > > > index c5b5f74cfd4d..6450bbe394e2 100644
> > > > --- a/mm/memcontrol.c
> > > > +++ b/mm/memcontrol.c
> > > > @@ -5360,10 +5360,12 @@ static int mem_cgroup_move_account(struct page *page,
> > > > }
> > > >
> > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > > - if (compound && !list_empty(page_deferred_list(page))) {
> > > > + if (compound) {
> > > > spin_lock(&from->deferred_split_queue.split_queue_lock);
> > > > - list_del_init(page_deferred_list(page));
> > > > - from->deferred_split_queue.split_queue_len--;
> > > > + if (!list_empty(page_deferred_list(page))) {
> > > > + list_del_init(page_deferred_list(page));
> > > > + from->deferred_split_queue.split_queue_len--;
> > > > + }
> > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
> > > > }
> > > > #endif
> > > > @@ -5377,11 +5379,13 @@ static int mem_cgroup_move_account(struct page *page,
> > > > page->mem_cgroup = to;
> > > >
> > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > > - if (compound && list_empty(page_deferred_list(page))) {
> > > > + if (compound) {
> > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
> > > > - list_add_tail(page_deferred_list(page),
> > > > - &to->deferred_split_queue.split_queue);
> > > > - to->deferred_split_queue.split_queue_len++;
> > > > + if (list_empty(page_deferred_list(page))) {
> > > > + list_add_tail(page_deferred_list(page),
> > > > + &to->deferred_split_queue.split_queue);
> > > > + to->deferred_split_queue.split_queue_len++;
> > > > + }
> > > > spin_unlock(&to->deferred_split_queue.split_queue_lock);
> > > > }
> > > > #endif
> > >
> > > The patch looks OK for me. But there is another question. I forget, why we unconditionally
> > > add a page with empty deferred list to deferred_split_queue. Shouldn't we also check that
> > > it was initially in the list? Something like:
> > >
> > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > > index d4394ae4e5be..0be0136adaa6 100644
> > > --- a/mm/memcontrol.c
> > > +++ b/mm/memcontrol.c
> > > @@ -5289,6 +5289,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > struct pglist_data *pgdat;
> > > unsigned long flags;
> > > unsigned int nr_pages = compound ? hpage_nr_pages(page) : 1;
> > > + bool split = false;
> > > int ret;
> > > bool anon;
> > >
> > > @@ -5346,6 +5347,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > if (!list_empty(page_deferred_list(page))) {
> > > list_del_init(page_deferred_list(page));
> > > from->deferred_split_queue.split_queue_len--;
> > > + split = true;
> > > }
> > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
> > > }
> > > @@ -5360,7 +5362,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > page->mem_cgroup = to;
> > >
> > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > - if (compound) {
> > > + if (compound && split) {
> > > spin_lock(&to->deferred_split_queue.split_queue_lock);
> > > if (list_empty(page_deferred_list(page))) {
> > > list_add_tail(page_deferred_list(page),
> > >
> >
> > I think that's a good point, especially considering that the current code
> > appears to unconditionally place any compound page on the deferred split
> > queue of the destination memcg. The correct list that it should appear
> > on, I believe, depends on whether the pmd has been split for the process
> > being moved: note the MC_TARGET_PAGE caveat in
> > mem_cgroup_move_charge_pte_range() that does not move the charge for
> > compound pages with split pmds. So when mem_cgroup_move_account() is
> > called with compound == true, we're moving the charge of the entire
> > compound page: why would it appear on that memcg's deferred split queue?
>
> I believe Kirill asked how do we know that the page should be actually
> added to the deferred list just from the list_empty check. In other
> words what if the page hasn't been split at all?
>
Right, and I don't think that it necessarily is and the second
conditional in Wei's patch will always succeed unless we have raced. That
patch is for a lock concern but I think Kirill's question has uncovered
something more interesting.
Kirill S would definitely be best to answer Kirill T's question, but from
my understanding when mem_cgroup_move_account() is called with
compound == true that we always have an intact pmd (we never migrate
partial page charges for pages on the deferred split queue with the
current charge migration implementation) and thus the underlying page is
not eligible to be split and shouldn't be on the deferred split queue.
In other words, a page being on the deferred split queue for a memcg
should only happen when it is charged to that memcg. (This wasn't the
case when we only had per-node split queues.) I think that's currently
broken in mem_cgroup_move_account() before Wei's patch.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
@ 2020-01-17 9:31 ` David Rientjes
0 siblings, 0 replies; 26+ messages in thread
From: David Rientjes @ 2020-01-17 9:31 UTC (permalink / raw)
To: Michal Hocko
Cc: Kirill Tkhai, Wei Yang, hannes, vdavydov.dev, akpm,
kirill.shutemov, yang.shi, cgroups, linux-mm, linux-kernel,
alexander.duyck, stable
On Fri, 17 Jan 2020, Michal Hocko wrote:
> On Thu 16-01-20 14:01:59, David Rientjes wrote:
> > On Thu, 16 Jan 2020, Kirill Tkhai wrote:
> >
> > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > > > index c5b5f74cfd4d..6450bbe394e2 100644
> > > > --- a/mm/memcontrol.c
> > > > +++ b/mm/memcontrol.c
> > > > @@ -5360,10 +5360,12 @@ static int mem_cgroup_move_account(struct page *page,
> > > > }
> > > >
> > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > > - if (compound && !list_empty(page_deferred_list(page))) {
> > > > + if (compound) {
> > > > spin_lock(&from->deferred_split_queue.split_queue_lock);
> > > > - list_del_init(page_deferred_list(page));
> > > > - from->deferred_split_queue.split_queue_len--;
> > > > + if (!list_empty(page_deferred_list(page))) {
> > > > + list_del_init(page_deferred_list(page));
> > > > + from->deferred_split_queue.split_queue_len--;
> > > > + }
> > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
> > > > }
> > > > #endif
> > > > @@ -5377,11 +5379,13 @@ static int mem_cgroup_move_account(struct page *page,
> > > > page->mem_cgroup = to;
> > > >
> > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > > - if (compound && list_empty(page_deferred_list(page))) {
> > > > + if (compound) {
> > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
> > > > - list_add_tail(page_deferred_list(page),
> > > > - &to->deferred_split_queue.split_queue);
> > > > - to->deferred_split_queue.split_queue_len++;
> > > > + if (list_empty(page_deferred_list(page))) {
> > > > + list_add_tail(page_deferred_list(page),
> > > > + &to->deferred_split_queue.split_queue);
> > > > + to->deferred_split_queue.split_queue_len++;
> > > > + }
> > > > spin_unlock(&to->deferred_split_queue.split_queue_lock);
> > > > }
> > > > #endif
> > >
> > > The patch looks OK for me. But there is another question. I forget, why we unconditionally
> > > add a page with empty deferred list to deferred_split_queue. Shouldn't we also check that
> > > it was initially in the list? Something like:
> > >
> > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > > index d4394ae4e5be..0be0136adaa6 100644
> > > --- a/mm/memcontrol.c
> > > +++ b/mm/memcontrol.c
> > > @@ -5289,6 +5289,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > struct pglist_data *pgdat;
> > > unsigned long flags;
> > > unsigned int nr_pages = compound ? hpage_nr_pages(page) : 1;
> > > + bool split = false;
> > > int ret;
> > > bool anon;
> > >
> > > @@ -5346,6 +5347,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > if (!list_empty(page_deferred_list(page))) {
> > > list_del_init(page_deferred_list(page));
> > > from->deferred_split_queue.split_queue_len--;
> > > + split = true;
> > > }
> > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
> > > }
> > > @@ -5360,7 +5362,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > page->mem_cgroup = to;
> > >
> > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > - if (compound) {
> > > + if (compound && split) {
> > > spin_lock(&to->deferred_split_queue.split_queue_lock);
> > > if (list_empty(page_deferred_list(page))) {
> > > list_add_tail(page_deferred_list(page),
> > >
> >
> > I think that's a good point, especially considering that the current code
> > appears to unconditionally place any compound page on the deferred split
> > queue of the destination memcg. The correct list that it should appear
> > on, I believe, depends on whether the pmd has been split for the process
> > being moved: note the MC_TARGET_PAGE caveat in
> > mem_cgroup_move_charge_pte_range() that does not move the charge for
> > compound pages with split pmds. So when mem_cgroup_move_account() is
> > called with compound == true, we're moving the charge of the entire
> > compound page: why would it appear on that memcg's deferred split queue?
>
> I believe Kirill asked how do we know that the page should be actually
> added to the deferred list just from the list_empty check. In other
> words what if the page hasn't been split at all?
>
Right, and I don't think that it necessarily is and the second
conditional in Wei's patch will always succeed unless we have raced. That
patch is for a lock concern but I think Kirill's question has uncovered
something more interesting.
Kirill S would definitely be best to answer Kirill T's question, but from
my understanding when mem_cgroup_move_account() is called with
compound == true that we always have an intact pmd (we never migrate
partial page charges for pages on the deferred split queue with the
current charge migration implementation) and thus the underlying page is
not eligible to be split and shouldn't be on the deferred split queue.
In other words, a page being on the deferred split queue for a memcg
should only happen when it is charged to that memcg. (This wasn't the
case when we only had per-node split queues.) I think that's currently
broken in mem_cgroup_move_account() before Wei's patch.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
2020-01-17 9:31 ` David Rientjes
(?)
(?)
@ 2020-01-17 15:38 ` Kirill A. Shutemov
2020-01-17 19:11 ` David Rientjes
2020-01-17 19:17 ` Yang Shi
-1 siblings, 2 replies; 26+ messages in thread
From: Kirill A. Shutemov @ 2020-01-17 15:38 UTC (permalink / raw)
To: David Rientjes
Cc: Michal Hocko, Kirill Tkhai, Wei Yang, hannes, vdavydov.dev, akpm,
kirill.shutemov, yang.shi, cgroups, linux-mm, linux-kernel,
alexander.duyck, stable
On Fri, Jan 17, 2020 at 01:31:50AM -0800, David Rientjes wrote:
> On Fri, 17 Jan 2020, Michal Hocko wrote:
>
> > On Thu 16-01-20 14:01:59, David Rientjes wrote:
> > > On Thu, 16 Jan 2020, Kirill Tkhai wrote:
> > >
> > > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > > > > index c5b5f74cfd4d..6450bbe394e2 100644
> > > > > --- a/mm/memcontrol.c
> > > > > +++ b/mm/memcontrol.c
> > > > > @@ -5360,10 +5360,12 @@ static int mem_cgroup_move_account(struct page *page,
> > > > > }
> > > > >
> > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > > > - if (compound && !list_empty(page_deferred_list(page))) {
> > > > > + if (compound) {
> > > > > spin_lock(&from->deferred_split_queue.split_queue_lock);
> > > > > - list_del_init(page_deferred_list(page));
> > > > > - from->deferred_split_queue.split_queue_len--;
> > > > > + if (!list_empty(page_deferred_list(page))) {
> > > > > + list_del_init(page_deferred_list(page));
> > > > > + from->deferred_split_queue.split_queue_len--;
> > > > > + }
> > > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
> > > > > }
> > > > > #endif
> > > > > @@ -5377,11 +5379,13 @@ static int mem_cgroup_move_account(struct page *page,
> > > > > page->mem_cgroup = to;
> > > > >
> > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > > > - if (compound && list_empty(page_deferred_list(page))) {
> > > > > + if (compound) {
> > > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
> > > > > - list_add_tail(page_deferred_list(page),
> > > > > - &to->deferred_split_queue.split_queue);
> > > > > - to->deferred_split_queue.split_queue_len++;
> > > > > + if (list_empty(page_deferred_list(page))) {
> > > > > + list_add_tail(page_deferred_list(page),
> > > > > + &to->deferred_split_queue.split_queue);
> > > > > + to->deferred_split_queue.split_queue_len++;
> > > > > + }
> > > > > spin_unlock(&to->deferred_split_queue.split_queue_lock);
> > > > > }
> > > > > #endif
> > > >
> > > > The patch looks OK for me. But there is another question. I forget, why we unconditionally
> > > > add a page with empty deferred list to deferred_split_queue. Shouldn't we also check that
> > > > it was initially in the list? Something like:
> > > >
> > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> > > > index d4394ae4e5be..0be0136adaa6 100644
> > > > --- a/mm/memcontrol.c
> > > > +++ b/mm/memcontrol.c
> > > > @@ -5289,6 +5289,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > > struct pglist_data *pgdat;
> > > > unsigned long flags;
> > > > unsigned int nr_pages = compound ? hpage_nr_pages(page) : 1;
> > > > + bool split = false;
> > > > int ret;
> > > > bool anon;
> > > >
> > > > @@ -5346,6 +5347,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > > if (!list_empty(page_deferred_list(page))) {
> > > > list_del_init(page_deferred_list(page));
> > > > from->deferred_split_queue.split_queue_len--;
> > > > + split = true;
> > > > }
> > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
> > > > }
> > > > @@ -5360,7 +5362,7 @@ static int mem_cgroup_move_account(struct page *page,
> > > > page->mem_cgroup = to;
> > > >
> > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> > > > - if (compound) {
> > > > + if (compound && split) {
> > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
> > > > if (list_empty(page_deferred_list(page))) {
> > > > list_add_tail(page_deferred_list(page),
> > > >
> > >
> > > I think that's a good point, especially considering that the current code
> > > appears to unconditionally place any compound page on the deferred split
> > > queue of the destination memcg. The correct list that it should appear
> > > on, I believe, depends on whether the pmd has been split for the process
> > > being moved: note the MC_TARGET_PAGE caveat in
> > > mem_cgroup_move_charge_pte_range() that does not move the charge for
> > > compound pages with split pmds. So when mem_cgroup_move_account() is
> > > called with compound == true, we're moving the charge of the entire
> > > compound page: why would it appear on that memcg's deferred split queue?
> >
> > I believe Kirill asked how do we know that the page should be actually
> > added to the deferred list just from the list_empty check. In other
> > words what if the page hasn't been split at all?
> >
>
> Right, and I don't think that it necessarily is and the second
> conditional in Wei's patch will always succeed unless we have raced. That
> patch is for a lock concern but I think Kirill's question has uncovered
> something more interesting.
>
> Kirill S would definitely be best to answer Kirill T's question, but from
> my understanding when mem_cgroup_move_account() is called with
> compound == true that we always have an intact pmd (we never migrate
> partial page charges for pages on the deferred split queue with the
> current charge migration implementation) and thus the underlying page is
> not eligible to be split and shouldn't be on the deferred split queue.
>
> In other words, a page being on the deferred split queue for a memcg
> should only happen when it is charged to that memcg. (This wasn't the
> case when we only had per-node split queues.) I think that's currently
> broken in mem_cgroup_move_account() before Wei's patch.
Right. It's broken indeed.
We are dealing with anon page here. And it cannot be on deferred list as
long as it's mapped with PMD. We cannot get compound == true &&
!list_empty() on the (first) enter to the function. Any PMD-mapped page
will be put onto deferred by the function. This is wrong.
The fix is not obvious.
This comment got in mem_cgroup_move_charge_pte_range() my attention:
/*
* We can have a part of the split pmd here. Moving it
* can be done but it would be too convoluted so simply
* ignore such a partial THP and keep it in original
* memcg. There should be somebody mapping the head.
*/
That's exactly the case we care about: PTE-mapped THP that has to be split
under load. We don't move charge of them between memcgs and therefore we
should not move the page to different memcg.
I guess this will do the trick :P
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index c5b5f74cfd4d..e87ee4c10f6e 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -5359,14 +5359,6 @@ static int mem_cgroup_move_account(struct page *page,
__mod_lruvec_state(to_vec, NR_WRITEBACK, nr_pages);
}
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
- if (compound && !list_empty(page_deferred_list(page))) {
- spin_lock(&from->deferred_split_queue.split_queue_lock);
- list_del_init(page_deferred_list(page));
- from->deferred_split_queue.split_queue_len--;
- spin_unlock(&from->deferred_split_queue.split_queue_lock);
- }
-#endif
/*
* It is safe to change page->mem_cgroup here because the page
* is referenced, charged, and isolated - we can't race with
@@ -5376,16 +5368,6 @@ static int mem_cgroup_move_account(struct page *page,
/* caller should have done css_get */
page->mem_cgroup = to;
-#ifdef CONFIG_TRANSPARENT_HUGEPAGE
- if (compound && list_empty(page_deferred_list(page))) {
- spin_lock(&to->deferred_split_queue.split_queue_lock);
- list_add_tail(page_deferred_list(page),
- &to->deferred_split_queue.split_queue);
- to->deferred_split_queue.split_queue_len++;
- spin_unlock(&to->deferred_split_queue.split_queue_lock);
- }
-#endif
-
spin_unlock_irqrestore(&from->move_lock, flags);
ret = 0;
--
Kirill A. Shutemov
^ permalink raw reply related [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
2020-01-17 15:38 ` Kirill A. Shutemov
@ 2020-01-17 19:11 ` David Rientjes
2020-01-17 19:17 ` Yang Shi
1 sibling, 0 replies; 26+ messages in thread
From: David Rientjes @ 2020-01-17 19:11 UTC (permalink / raw)
To: Kirill A. Shutemov
Cc: Michal Hocko, Kirill Tkhai, Wei Yang, hannes, vdavydov.dev, akpm,
kirill.shutemov, yang.shi, cgroups, linux-mm, linux-kernel,
alexander.duyck, stable
On Fri, 17 Jan 2020, Kirill A. Shutemov wrote:
> > Right, and I don't think that it necessarily is and the second
> > conditional in Wei's patch will always succeed unless we have raced. That
> > patch is for a lock concern but I think Kirill's question has uncovered
> > something more interesting.
> >
> > Kirill S would definitely be best to answer Kirill T's question, but from
> > my understanding when mem_cgroup_move_account() is called with
> > compound == true that we always have an intact pmd (we never migrate
> > partial page charges for pages on the deferred split queue with the
> > current charge migration implementation) and thus the underlying page is
> > not eligible to be split and shouldn't be on the deferred split queue.
> >
> > In other words, a page being on the deferred split queue for a memcg
> > should only happen when it is charged to that memcg. (This wasn't the
> > case when we only had per-node split queues.) I think that's currently
> > broken in mem_cgroup_move_account() before Wei's patch.
>
> Right. It's broken indeed.
>
> We are dealing with anon page here. And it cannot be on deferred list as
> long as it's mapped with PMD. We cannot get compound == true &&
> !list_empty() on the (first) enter to the function. Any PMD-mapped page
> will be put onto deferred by the function. This is wrong.
>
> The fix is not obvious.
>
> This comment got in mem_cgroup_move_charge_pte_range() my attention:
>
> /*
> * We can have a part of the split pmd here. Moving it
> * can be done but it would be too convoluted so simply
> * ignore such a partial THP and keep it in original
> * memcg. There should be somebody mapping the head.
> */
>
> That's exactly the case we care about: PTE-mapped THP that has to be split
> under load. We don't move charge of them between memcgs and therefore we
> should not move the page to different memcg.
>
> I guess this will do the trick :P
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index c5b5f74cfd4d..e87ee4c10f6e 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -5359,14 +5359,6 @@ static int mem_cgroup_move_account(struct page *page,
> __mod_lruvec_state(to_vec, NR_WRITEBACK, nr_pages);
> }
>
> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> - if (compound && !list_empty(page_deferred_list(page))) {
> - spin_lock(&from->deferred_split_queue.split_queue_lock);
> - list_del_init(page_deferred_list(page));
> - from->deferred_split_queue.split_queue_len--;
> - spin_unlock(&from->deferred_split_queue.split_queue_lock);
> - }
> -#endif
> /*
> * It is safe to change page->mem_cgroup here because the page
> * is referenced, charged, and isolated - we can't race with
> @@ -5376,16 +5368,6 @@ static int mem_cgroup_move_account(struct page *page,
> /* caller should have done css_get */
> page->mem_cgroup = to;
>
> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> - if (compound && list_empty(page_deferred_list(page))) {
> - spin_lock(&to->deferred_split_queue.split_queue_lock);
> - list_add_tail(page_deferred_list(page),
> - &to->deferred_split_queue.split_queue);
> - to->deferred_split_queue.split_queue_len++;
> - spin_unlock(&to->deferred_split_queue.split_queue_lock);
> - }
> -#endif
> -
> spin_unlock_irqrestore(&from->move_lock, flags);
>
> ret = 0;
Yeah, this is what I was thinking as well. When
PageTransHuge(page) == true and there's a mapping pmd, the charge gets
moved but the page shouldn't appear on any deferred split queue; when
there isn't a mapped pmd, it should already be on a queue but the charge
doesn't get moved so no change in which queue is needed.
There was no deferred split handling in mem_cgroup_move_account() needed
for per-node deferred split queues either so this is purely an issue for
commit 87eaceb3faa5 ("mm: thp: make deferred split shrinker memcg aware")
so I think we need your patch and it should be annotated for stable 5.4+.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
@ 2020-01-17 19:11 ` David Rientjes
0 siblings, 0 replies; 26+ messages in thread
From: David Rientjes @ 2020-01-17 19:11 UTC (permalink / raw)
To: Kirill A. Shutemov
Cc: Michal Hocko, Kirill Tkhai, Wei Yang, hannes, vdavydov.dev, akpm,
kirill.shutemov, yang.shi, cgroups, linux-mm, linux-kernel,
alexander.duyck, stable
On Fri, 17 Jan 2020, Kirill A. Shutemov wrote:
> > Right, and I don't think that it necessarily is and the second
> > conditional in Wei's patch will always succeed unless we have raced. That
> > patch is for a lock concern but I think Kirill's question has uncovered
> > something more interesting.
> >
> > Kirill S would definitely be best to answer Kirill T's question, but from
> > my understanding when mem_cgroup_move_account() is called with
> > compound == true that we always have an intact pmd (we never migrate
> > partial page charges for pages on the deferred split queue with the
> > current charge migration implementation) and thus the underlying page is
> > not eligible to be split and shouldn't be on the deferred split queue.
> >
> > In other words, a page being on the deferred split queue for a memcg
> > should only happen when it is charged to that memcg. (This wasn't the
> > case when we only had per-node split queues.) I think that's currently
> > broken in mem_cgroup_move_account() before Wei's patch.
>
> Right. It's broken indeed.
>
> We are dealing with anon page here. And it cannot be on deferred list as
> long as it's mapped with PMD. We cannot get compound == true &&
> !list_empty() on the (first) enter to the function. Any PMD-mapped page
> will be put onto deferred by the function. This is wrong.
>
> The fix is not obvious.
>
> This comment got in mem_cgroup_move_charge_pte_range() my attention:
>
> /*
> * We can have a part of the split pmd here. Moving it
> * can be done but it would be too convoluted so simply
> * ignore such a partial THP and keep it in original
> * memcg. There should be somebody mapping the head.
> */
>
> That's exactly the case we care about: PTE-mapped THP that has to be split
> under load. We don't move charge of them between memcgs and therefore we
> should not move the page to different memcg.
>
> I guess this will do the trick :P
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index c5b5f74cfd4d..e87ee4c10f6e 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -5359,14 +5359,6 @@ static int mem_cgroup_move_account(struct page *page,
> __mod_lruvec_state(to_vec, NR_WRITEBACK, nr_pages);
> }
>
> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> - if (compound && !list_empty(page_deferred_list(page))) {
> - spin_lock(&from->deferred_split_queue.split_queue_lock);
> - list_del_init(page_deferred_list(page));
> - from->deferred_split_queue.split_queue_len--;
> - spin_unlock(&from->deferred_split_queue.split_queue_lock);
> - }
> -#endif
> /*
> * It is safe to change page->mem_cgroup here because the page
> * is referenced, charged, and isolated - we can't race with
> @@ -5376,16 +5368,6 @@ static int mem_cgroup_move_account(struct page *page,
> /* caller should have done css_get */
> page->mem_cgroup = to;
>
> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> - if (compound && list_empty(page_deferred_list(page))) {
> - spin_lock(&to->deferred_split_queue.split_queue_lock);
> - list_add_tail(page_deferred_list(page),
> - &to->deferred_split_queue.split_queue);
> - to->deferred_split_queue.split_queue_len++;
> - spin_unlock(&to->deferred_split_queue.split_queue_lock);
> - }
> -#endif
> -
> spin_unlock_irqrestore(&from->move_lock, flags);
>
> ret = 0;
Yeah, this is what I was thinking as well. When
PageTransHuge(page) == true and there's a mapping pmd, the charge gets
moved but the page shouldn't appear on any deferred split queue; when
there isn't a mapped pmd, it should already be on a queue but the charge
doesn't get moved so no change in which queue is needed.
There was no deferred split handling in mem_cgroup_move_account() needed
for per-node deferred split queues either so this is purely an issue for
commit 87eaceb3faa5 ("mm: thp: make deferred split shrinker memcg aware")
so I think we need your patch and it should be annotated for stable 5.4+.
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
2020-01-17 15:38 ` Kirill A. Shutemov
@ 2020-01-17 19:17 ` Yang Shi
2020-01-17 19:17 ` Yang Shi
1 sibling, 0 replies; 26+ messages in thread
From: Yang Shi @ 2020-01-17 19:17 UTC (permalink / raw)
To: Kirill A. Shutemov, David Rientjes
Cc: Michal Hocko, Kirill Tkhai, Wei Yang, hannes, vdavydov.dev, akpm,
kirill.shutemov, cgroups, linux-mm, linux-kernel,
alexander.duyck, stable
On 1/17/20 7:38 AM, Kirill A. Shutemov wrote:
> On Fri, Jan 17, 2020 at 01:31:50AM -0800, David Rientjes wrote:
>> On Fri, 17 Jan 2020, Michal Hocko wrote:
>>
>>> On Thu 16-01-20 14:01:59, David Rientjes wrote:
>>>> On Thu, 16 Jan 2020, Kirill Tkhai wrote:
>>>>
>>>>>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>>>>>> index c5b5f74cfd4d..6450bbe394e2 100644
>>>>>> --- a/mm/memcontrol.c
>>>>>> +++ b/mm/memcontrol.c
>>>>>> @@ -5360,10 +5360,12 @@ static int mem_cgroup_move_account(struct page *page,
>>>>>> }
>>>>>>
>>>>>> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>>>>>> - if (compound && !list_empty(page_deferred_list(page))) {
>>>>>> + if (compound) {
>>>>>> spin_lock(&from->deferred_split_queue.split_queue_lock);
>>>>>> - list_del_init(page_deferred_list(page));
>>>>>> - from->deferred_split_queue.split_queue_len--;
>>>>>> + if (!list_empty(page_deferred_list(page))) {
>>>>>> + list_del_init(page_deferred_list(page));
>>>>>> + from->deferred_split_queue.split_queue_len--;
>>>>>> + }
>>>>>> spin_unlock(&from->deferred_split_queue.split_queue_lock);
>>>>>> }
>>>>>> #endif
>>>>>> @@ -5377,11 +5379,13 @@ static int mem_cgroup_move_account(struct page *page,
>>>>>> page->mem_cgroup = to;
>>>>>>
>>>>>> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>>>>>> - if (compound && list_empty(page_deferred_list(page))) {
>>>>>> + if (compound) {
>>>>>> spin_lock(&to->deferred_split_queue.split_queue_lock);
>>>>>> - list_add_tail(page_deferred_list(page),
>>>>>> - &to->deferred_split_queue.split_queue);
>>>>>> - to->deferred_split_queue.split_queue_len++;
>>>>>> + if (list_empty(page_deferred_list(page))) {
>>>>>> + list_add_tail(page_deferred_list(page),
>>>>>> + &to->deferred_split_queue.split_queue);
>>>>>> + to->deferred_split_queue.split_queue_len++;
>>>>>> + }
>>>>>> spin_unlock(&to->deferred_split_queue.split_queue_lock);
>>>>>> }
>>>>>> #endif
>>>>> The patch looks OK for me. But there is another question. I forget, why we unconditionally
>>>>> add a page with empty deferred list to deferred_split_queue. Shouldn't we also check that
>>>>> it was initially in the list? Something like:
>>>>>
>>>>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>>>>> index d4394ae4e5be..0be0136adaa6 100644
>>>>> --- a/mm/memcontrol.c
>>>>> +++ b/mm/memcontrol.c
>>>>> @@ -5289,6 +5289,7 @@ static int mem_cgroup_move_account(struct page *page,
>>>>> struct pglist_data *pgdat;
>>>>> unsigned long flags;
>>>>> unsigned int nr_pages = compound ? hpage_nr_pages(page) : 1;
>>>>> + bool split = false;
>>>>> int ret;
>>>>> bool anon;
>>>>>
>>>>> @@ -5346,6 +5347,7 @@ static int mem_cgroup_move_account(struct page *page,
>>>>> if (!list_empty(page_deferred_list(page))) {
>>>>> list_del_init(page_deferred_list(page));
>>>>> from->deferred_split_queue.split_queue_len--;
>>>>> + split = true;
>>>>> }
>>>>> spin_unlock(&from->deferred_split_queue.split_queue_lock);
>>>>> }
>>>>> @@ -5360,7 +5362,7 @@ static int mem_cgroup_move_account(struct page *page,
>>>>> page->mem_cgroup = to;
>>>>>
>>>>> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>>>>> - if (compound) {
>>>>> + if (compound && split) {
>>>>> spin_lock(&to->deferred_split_queue.split_queue_lock);
>>>>> if (list_empty(page_deferred_list(page))) {
>>>>> list_add_tail(page_deferred_list(page),
>>>>>
>>>> I think that's a good point, especially considering that the current code
>>>> appears to unconditionally place any compound page on the deferred split
>>>> queue of the destination memcg. The correct list that it should appear
>>>> on, I believe, depends on whether the pmd has been split for the process
>>>> being moved: note the MC_TARGET_PAGE caveat in
>>>> mem_cgroup_move_charge_pte_range() that does not move the charge for
>>>> compound pages with split pmds. So when mem_cgroup_move_account() is
>>>> called with compound == true, we're moving the charge of the entire
>>>> compound page: why would it appear on that memcg's deferred split queue?
>>> I believe Kirill asked how do we know that the page should be actually
>>> added to the deferred list just from the list_empty check. In other
>>> words what if the page hasn't been split at all?
>>>
>> Right, and I don't think that it necessarily is and the second
>> conditional in Wei's patch will always succeed unless we have raced. That
>> patch is for a lock concern but I think Kirill's question has uncovered
>> something more interesting.
>>
>> Kirill S would definitely be best to answer Kirill T's question, but from
>> my understanding when mem_cgroup_move_account() is called with
>> compound == true that we always have an intact pmd (we never migrate
>> partial page charges for pages on the deferred split queue with the
>> current charge migration implementation) and thus the underlying page is
>> not eligible to be split and shouldn't be on the deferred split queue.
>>
>> In other words, a page being on the deferred split queue for a memcg
>> should only happen when it is charged to that memcg. (This wasn't the
>> case when we only had per-node split queues.) I think that's currently
>> broken in mem_cgroup_move_account() before Wei's patch.
> Right. It's broken indeed.
Hmm... Yes, definitely. I wasn't realized this at the first place.
>
> We are dealing with anon page here. And it cannot be on deferred list as
> long as it's mapped with PMD. We cannot get compound == true &&
> !list_empty() on the (first) enter to the function. Any PMD-mapped page
> will be put onto deferred by the function. This is wrong.
>
> The fix is not obvious.
>
> This comment got in mem_cgroup_move_charge_pte_range() my attention:
>
> /*
> * We can have a part of the split pmd here. Moving it
> * can be done but it would be too convoluted so simply
> * ignore such a partial THP and keep it in original
> * memcg. There should be somebody mapping the head.
> */
>
> That's exactly the case we care about: PTE-mapped THP that has to be split
> under load. We don't move charge of them between memcgs and therefore we
> should not move the page to different memcg.
>
> I guess this will do the trick :P
It seems correct to me. In addition, memcg move charge just move PMD
mapped THP, the THP should be never on the deferred split queue of
"from" if it is PMD mapped, so actually we don't have to move it to the
deferred split queue of "to".
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index c5b5f74cfd4d..e87ee4c10f6e 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -5359,14 +5359,6 @@ static int mem_cgroup_move_account(struct page *page,
> __mod_lruvec_state(to_vec, NR_WRITEBACK, nr_pages);
> }
>
> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> - if (compound && !list_empty(page_deferred_list(page))) {
> - spin_lock(&from->deferred_split_queue.split_queue_lock);
> - list_del_init(page_deferred_list(page));
> - from->deferred_split_queue.split_queue_len--;
> - spin_unlock(&from->deferred_split_queue.split_queue_lock);
> - }
> -#endif
> /*
> * It is safe to change page->mem_cgroup here because the page
> * is referenced, charged, and isolated - we can't race with
> @@ -5376,16 +5368,6 @@ static int mem_cgroup_move_account(struct page *page,
> /* caller should have done css_get */
> page->mem_cgroup = to;
>
> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> - if (compound && list_empty(page_deferred_list(page))) {
> - spin_lock(&to->deferred_split_queue.split_queue_lock);
> - list_add_tail(page_deferred_list(page),
> - &to->deferred_split_queue.split_queue);
> - to->deferred_split_queue.split_queue_len++;
> - spin_unlock(&to->deferred_split_queue.split_queue_lock);
> - }
> -#endif
> -
> spin_unlock_irqrestore(&from->move_lock, flags);
>
> ret = 0;
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
@ 2020-01-17 19:17 ` Yang Shi
0 siblings, 0 replies; 26+ messages in thread
From: Yang Shi @ 2020-01-17 19:17 UTC (permalink / raw)
To: Kirill A. Shutemov, David Rientjes
Cc: Michal Hocko, Kirill Tkhai, Wei Yang,
hannes-druUgvl0LCNAfugRpC6u6w,
vdavydov.dev-Re5JQEeQqe8AvxtiuMwx3w,
akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b,
kirill.shutemov-VuQAYsv1563Yd54FQh9/CA,
cgroups-u79uwXL29TY76Z2rM5mHXA, linux-mm-Bw31MaZKKs3YtjvyW6yDsg,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
alexander.duyck-Re5JQEeQqe8AvxtiuMwx3w,
stable-u79uwXL29TY76Z2rM5mHXA
On 1/17/20 7:38 AM, Kirill A. Shutemov wrote:
> On Fri, Jan 17, 2020 at 01:31:50AM -0800, David Rientjes wrote:
>> On Fri, 17 Jan 2020, Michal Hocko wrote:
>>
>>> On Thu 16-01-20 14:01:59, David Rientjes wrote:
>>>> On Thu, 16 Jan 2020, Kirill Tkhai wrote:
>>>>
>>>>>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>>>>>> index c5b5f74cfd4d..6450bbe394e2 100644
>>>>>> --- a/mm/memcontrol.c
>>>>>> +++ b/mm/memcontrol.c
>>>>>> @@ -5360,10 +5360,12 @@ static int mem_cgroup_move_account(struct page *page,
>>>>>> }
>>>>>>
>>>>>> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>>>>>> - if (compound && !list_empty(page_deferred_list(page))) {
>>>>>> + if (compound) {
>>>>>> spin_lock(&from->deferred_split_queue.split_queue_lock);
>>>>>> - list_del_init(page_deferred_list(page));
>>>>>> - from->deferred_split_queue.split_queue_len--;
>>>>>> + if (!list_empty(page_deferred_list(page))) {
>>>>>> + list_del_init(page_deferred_list(page));
>>>>>> + from->deferred_split_queue.split_queue_len--;
>>>>>> + }
>>>>>> spin_unlock(&from->deferred_split_queue.split_queue_lock);
>>>>>> }
>>>>>> #endif
>>>>>> @@ -5377,11 +5379,13 @@ static int mem_cgroup_move_account(struct page *page,
>>>>>> page->mem_cgroup = to;
>>>>>>
>>>>>> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>>>>>> - if (compound && list_empty(page_deferred_list(page))) {
>>>>>> + if (compound) {
>>>>>> spin_lock(&to->deferred_split_queue.split_queue_lock);
>>>>>> - list_add_tail(page_deferred_list(page),
>>>>>> - &to->deferred_split_queue.split_queue);
>>>>>> - to->deferred_split_queue.split_queue_len++;
>>>>>> + if (list_empty(page_deferred_list(page))) {
>>>>>> + list_add_tail(page_deferred_list(page),
>>>>>> + &to->deferred_split_queue.split_queue);
>>>>>> + to->deferred_split_queue.split_queue_len++;
>>>>>> + }
>>>>>> spin_unlock(&to->deferred_split_queue.split_queue_lock);
>>>>>> }
>>>>>> #endif
>>>>> The patch looks OK for me. But there is another question. I forget, why we unconditionally
>>>>> add a page with empty deferred list to deferred_split_queue. Shouldn't we also check that
>>>>> it was initially in the list? Something like:
>>>>>
>>>>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>>>>> index d4394ae4e5be..0be0136adaa6 100644
>>>>> --- a/mm/memcontrol.c
>>>>> +++ b/mm/memcontrol.c
>>>>> @@ -5289,6 +5289,7 @@ static int mem_cgroup_move_account(struct page *page,
>>>>> struct pglist_data *pgdat;
>>>>> unsigned long flags;
>>>>> unsigned int nr_pages = compound ? hpage_nr_pages(page) : 1;
>>>>> + bool split = false;
>>>>> int ret;
>>>>> bool anon;
>>>>>
>>>>> @@ -5346,6 +5347,7 @@ static int mem_cgroup_move_account(struct page *page,
>>>>> if (!list_empty(page_deferred_list(page))) {
>>>>> list_del_init(page_deferred_list(page));
>>>>> from->deferred_split_queue.split_queue_len--;
>>>>> + split = true;
>>>>> }
>>>>> spin_unlock(&from->deferred_split_queue.split_queue_lock);
>>>>> }
>>>>> @@ -5360,7 +5362,7 @@ static int mem_cgroup_move_account(struct page *page,
>>>>> page->mem_cgroup = to;
>>>>>
>>>>> #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>>>>> - if (compound) {
>>>>> + if (compound && split) {
>>>>> spin_lock(&to->deferred_split_queue.split_queue_lock);
>>>>> if (list_empty(page_deferred_list(page))) {
>>>>> list_add_tail(page_deferred_list(page),
>>>>>
>>>> I think that's a good point, especially considering that the current code
>>>> appears to unconditionally place any compound page on the deferred split
>>>> queue of the destination memcg. The correct list that it should appear
>>>> on, I believe, depends on whether the pmd has been split for the process
>>>> being moved: note the MC_TARGET_PAGE caveat in
>>>> mem_cgroup_move_charge_pte_range() that does not move the charge for
>>>> compound pages with split pmds. So when mem_cgroup_move_account() is
>>>> called with compound == true, we're moving the charge of the entire
>>>> compound page: why would it appear on that memcg's deferred split queue?
>>> I believe Kirill asked how do we know that the page should be actually
>>> added to the deferred list just from the list_empty check. In other
>>> words what if the page hasn't been split at all?
>>>
>> Right, and I don't think that it necessarily is and the second
>> conditional in Wei's patch will always succeed unless we have raced. That
>> patch is for a lock concern but I think Kirill's question has uncovered
>> something more interesting.
>>
>> Kirill S would definitely be best to answer Kirill T's question, but from
>> my understanding when mem_cgroup_move_account() is called with
>> compound == true that we always have an intact pmd (we never migrate
>> partial page charges for pages on the deferred split queue with the
>> current charge migration implementation) and thus the underlying page is
>> not eligible to be split and shouldn't be on the deferred split queue.
>>
>> In other words, a page being on the deferred split queue for a memcg
>> should only happen when it is charged to that memcg. (This wasn't the
>> case when we only had per-node split queues.) I think that's currently
>> broken in mem_cgroup_move_account() before Wei's patch.
> Right. It's broken indeed.
Hmm... Yes, definitely. I wasn't realized this at the first place.
>
> We are dealing with anon page here. And it cannot be on deferred list as
> long as it's mapped with PMD. We cannot get compound == true &&
> !list_empty() on the (first) enter to the function. Any PMD-mapped page
> will be put onto deferred by the function. This is wrong.
>
> The fix is not obvious.
>
> This comment got in mem_cgroup_move_charge_pte_range() my attention:
>
> /*
> * We can have a part of the split pmd here. Moving it
> * can be done but it would be too convoluted so simply
> * ignore such a partial THP and keep it in original
> * memcg. There should be somebody mapping the head.
> */
>
> That's exactly the case we care about: PTE-mapped THP that has to be split
> under load. We don't move charge of them between memcgs and therefore we
> should not move the page to different memcg.
>
> I guess this will do the trick :P
It seems correct to me. In addition, memcg move charge just move PMD
mapped THP, the THP should be never on the deferred split queue of
"from" if it is PMD mapped, so actually we don't have to move it to the
deferred split queue of "to".
>
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index c5b5f74cfd4d..e87ee4c10f6e 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -5359,14 +5359,6 @@ static int mem_cgroup_move_account(struct page *page,
> __mod_lruvec_state(to_vec, NR_WRITEBACK, nr_pages);
> }
>
> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> - if (compound && !list_empty(page_deferred_list(page))) {
> - spin_lock(&from->deferred_split_queue.split_queue_lock);
> - list_del_init(page_deferred_list(page));
> - from->deferred_split_queue.split_queue_len--;
> - spin_unlock(&from->deferred_split_queue.split_queue_lock);
> - }
> -#endif
> /*
> * It is safe to change page->mem_cgroup here because the page
> * is referenced, charged, and isolated - we can't race with
> @@ -5376,16 +5368,6 @@ static int mem_cgroup_move_account(struct page *page,
> /* caller should have done css_get */
> page->mem_cgroup = to;
>
> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
> - if (compound && list_empty(page_deferred_list(page))) {
> - spin_lock(&to->deferred_split_queue.split_queue_lock);
> - list_add_tail(page_deferred_list(page),
> - &to->deferred_split_queue.split_queue);
> - to->deferred_split_queue.split_queue_len++;
> - spin_unlock(&to->deferred_split_queue.split_queue_lock);
> - }
> -#endif
> -
> spin_unlock_irqrestore(&from->move_lock, flags);
>
> ret = 0;
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
@ 2020-01-17 22:18 ` Wei Yang
0 siblings, 0 replies; 26+ messages in thread
From: Wei Yang @ 2020-01-17 22:18 UTC (permalink / raw)
To: Yang Shi
Cc: Kirill A. Shutemov, David Rientjes, Michal Hocko, Kirill Tkhai,
Wei Yang, hannes, vdavydov.dev, akpm, kirill.shutemov, cgroups,
linux-mm, linux-kernel, alexander.duyck, stable
On Fri, Jan 17, 2020 at 11:17:38AM -0800, Yang Shi wrote:
>
>
>On 1/17/20 7:38 AM, Kirill A. Shutemov wrote:
>> On Fri, Jan 17, 2020 at 01:31:50AM -0800, David Rientjes wrote:
>> > On Fri, 17 Jan 2020, Michal Hocko wrote:
>> >
>> > > On Thu 16-01-20 14:01:59, David Rientjes wrote:
>> > > > On Thu, 16 Jan 2020, Kirill Tkhai wrote:
>> > > >
>> > > > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> > > > > > index c5b5f74cfd4d..6450bbe394e2 100644
>> > > > > > --- a/mm/memcontrol.c
>> > > > > > +++ b/mm/memcontrol.c
>> > > > > > @@ -5360,10 +5360,12 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > > }
>> > > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> > > > > > - if (compound && !list_empty(page_deferred_list(page))) {
>> > > > > > + if (compound) {
>> > > > > > spin_lock(&from->deferred_split_queue.split_queue_lock);
>> > > > > > - list_del_init(page_deferred_list(page));
>> > > > > > - from->deferred_split_queue.split_queue_len--;
>> > > > > > + if (!list_empty(page_deferred_list(page))) {
>> > > > > > + list_del_init(page_deferred_list(page));
>> > > > > > + from->deferred_split_queue.split_queue_len--;
>> > > > > > + }
>> > > > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
>> > > > > > }
>> > > > > > #endif
>> > > > > > @@ -5377,11 +5379,13 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > > page->mem_cgroup = to;
>> > > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> > > > > > - if (compound && list_empty(page_deferred_list(page))) {
>> > > > > > + if (compound) {
>> > > > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
>> > > > > > - list_add_tail(page_deferred_list(page),
>> > > > > > - &to->deferred_split_queue.split_queue);
>> > > > > > - to->deferred_split_queue.split_queue_len++;
>> > > > > > + if (list_empty(page_deferred_list(page))) {
>> > > > > > + list_add_tail(page_deferred_list(page),
>> > > > > > + &to->deferred_split_queue.split_queue);
>> > > > > > + to->deferred_split_queue.split_queue_len++;
>> > > > > > + }
>> > > > > > spin_unlock(&to->deferred_split_queue.split_queue_lock);
>> > > > > > }
>> > > > > > #endif
>> > > > > The patch looks OK for me. But there is another question. I forget, why we unconditionally
>> > > > > add a page with empty deferred list to deferred_split_queue. Shouldn't we also check that
>> > > > > it was initially in the list? Something like:
>> > > > >
>> > > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> > > > > index d4394ae4e5be..0be0136adaa6 100644
>> > > > > --- a/mm/memcontrol.c
>> > > > > +++ b/mm/memcontrol.c
>> > > > > @@ -5289,6 +5289,7 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > struct pglist_data *pgdat;
>> > > > > unsigned long flags;
>> > > > > unsigned int nr_pages = compound ? hpage_nr_pages(page) : 1;
>> > > > > + bool split = false;
>> > > > > int ret;
>> > > > > bool anon;
>> > > > > @@ -5346,6 +5347,7 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > if (!list_empty(page_deferred_list(page))) {
>> > > > > list_del_init(page_deferred_list(page));
>> > > > > from->deferred_split_queue.split_queue_len--;
>> > > > > + split = true;
>> > > > > }
>> > > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
>> > > > > }
>> > > > > @@ -5360,7 +5362,7 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > page->mem_cgroup = to;
>> > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> > > > > - if (compound) {
>> > > > > + if (compound && split) {
>> > > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
>> > > > > if (list_empty(page_deferred_list(page))) {
>> > > > > list_add_tail(page_deferred_list(page),
>> > > > >
>> > > > I think that's a good point, especially considering that the current code
>> > > > appears to unconditionally place any compound page on the deferred split
>> > > > queue of the destination memcg. The correct list that it should appear
>> > > > on, I believe, depends on whether the pmd has been split for the process
>> > > > being moved: note the MC_TARGET_PAGE caveat in
>> > > > mem_cgroup_move_charge_pte_range() that does not move the charge for
>> > > > compound pages with split pmds. So when mem_cgroup_move_account() is
>> > > > called with compound == true, we're moving the charge of the entire
>> > > > compound page: why would it appear on that memcg's deferred split queue?
>> > > I believe Kirill asked how do we know that the page should be actually
>> > > added to the deferred list just from the list_empty check. In other
>> > > words what if the page hasn't been split at all?
>> > >
>> > Right, and I don't think that it necessarily is and the second
>> > conditional in Wei's patch will always succeed unless we have raced. That
>> > patch is for a lock concern but I think Kirill's question has uncovered
>> > something more interesting.
>> >
>> > Kirill S would definitely be best to answer Kirill T's question, but from
>> > my understanding when mem_cgroup_move_account() is called with
>> > compound == true that we always have an intact pmd (we never migrate
>> > partial page charges for pages on the deferred split queue with the
>> > current charge migration implementation) and thus the underlying page is
>> > not eligible to be split and shouldn't be on the deferred split queue.
>> >
>> > In other words, a page being on the deferred split queue for a memcg
>> > should only happen when it is charged to that memcg. (This wasn't the
>> > case when we only had per-node split queues.) I think that's currently
>> > broken in mem_cgroup_move_account() before Wei's patch.
>> Right. It's broken indeed.
>
>Hmm... Yes, definitely. I wasn't realized this at the first place.
>
>>
>> We are dealing with anon page here. And it cannot be on deferred list as
>> long as it's mapped with PMD. We cannot get compound == true &&
>> !list_empty() on the (first) enter to the function. Any PMD-mapped page
>> will be put onto deferred by the function. This is wrong.
>>
>> The fix is not obvious.
>>
>> This comment got in mem_cgroup_move_charge_pte_range() my attention:
>>
>> /*
>> * We can have a part of the split pmd here. Moving it
>> * can be done but it would be too convoluted so simply
>> * ignore such a partial THP and keep it in original
>> * memcg. There should be somebody mapping the head.
>> */
>>
>> That's exactly the case we care about: PTE-mapped THP that has to be split
>> under load. We don't move charge of them between memcgs and therefore we
>> should not move the page to different memcg.
>>
>> I guess this will do the trick :P
>
>It seems correct to me. In addition, memcg move charge just move PMD mapped
>THP, the THP should be never on the deferred split queue of "from" if it is
>PMD mapped, so actually we don't have to move it to the deferred split queue
>of "to".
>
Well, I got the point.
Since Kirill S found the correct solution, should I prepare v3 or Kirill will
send it?
>>
>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> index c5b5f74cfd4d..e87ee4c10f6e 100644
>> --- a/mm/memcontrol.c
>> +++ b/mm/memcontrol.c
>> @@ -5359,14 +5359,6 @@ static int mem_cgroup_move_account(struct page *page,
>> __mod_lruvec_state(to_vec, NR_WRITEBACK, nr_pages);
>> }
>> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> - if (compound && !list_empty(page_deferred_list(page))) {
>> - spin_lock(&from->deferred_split_queue.split_queue_lock);
>> - list_del_init(page_deferred_list(page));
>> - from->deferred_split_queue.split_queue_len--;
>> - spin_unlock(&from->deferred_split_queue.split_queue_lock);
>> - }
>> -#endif
>> /*
>> * It is safe to change page->mem_cgroup here because the page
>> * is referenced, charged, and isolated - we can't race with
>> @@ -5376,16 +5368,6 @@ static int mem_cgroup_move_account(struct page *page,
>> /* caller should have done css_get */
>> page->mem_cgroup = to;
>> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> - if (compound && list_empty(page_deferred_list(page))) {
>> - spin_lock(&to->deferred_split_queue.split_queue_lock);
>> - list_add_tail(page_deferred_list(page),
>> - &to->deferred_split_queue.split_queue);
>> - to->deferred_split_queue.split_queue_len++;
>> - spin_unlock(&to->deferred_split_queue.split_queue_lock);
>> - }
>> -#endif
>> -
>> spin_unlock_irqrestore(&from->move_lock, flags);
>> ret = 0;
--
Wei Yang
Help you, Help me
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
@ 2020-01-17 22:18 ` Wei Yang
0 siblings, 0 replies; 26+ messages in thread
From: Wei Yang @ 2020-01-17 22:18 UTC (permalink / raw)
To: Yang Shi
Cc: Kirill A. Shutemov, David Rientjes, Michal Hocko, Kirill Tkhai,
Wei Yang, hannes-druUgvl0LCNAfugRpC6u6w,
vdavydov.dev-Re5JQEeQqe8AvxtiuMwx3w,
akpm-de/tnXTf+JLsfHDXvbKv3WD2FQJk+8+b,
kirill.shutemov-VuQAYsv1563Yd54FQh9/CA,
cgroups-u79uwXL29TY76Z2rM5mHXA, linux-mm-Bw31MaZKKs3YtjvyW6yDsg,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
alexander.duyck-Re5JQEeQqe8AvxtiuMwx3w,
stable-u79uwXL29TY76Z2rM5mHXA
On Fri, Jan 17, 2020 at 11:17:38AM -0800, Yang Shi wrote:
>
>
>On 1/17/20 7:38 AM, Kirill A. Shutemov wrote:
>> On Fri, Jan 17, 2020 at 01:31:50AM -0800, David Rientjes wrote:
>> > On Fri, 17 Jan 2020, Michal Hocko wrote:
>> >
>> > > On Thu 16-01-20 14:01:59, David Rientjes wrote:
>> > > > On Thu, 16 Jan 2020, Kirill Tkhai wrote:
>> > > >
>> > > > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> > > > > > index c5b5f74cfd4d..6450bbe394e2 100644
>> > > > > > --- a/mm/memcontrol.c
>> > > > > > +++ b/mm/memcontrol.c
>> > > > > > @@ -5360,10 +5360,12 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > > }
>> > > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> > > > > > - if (compound && !list_empty(page_deferred_list(page))) {
>> > > > > > + if (compound) {
>> > > > > > spin_lock(&from->deferred_split_queue.split_queue_lock);
>> > > > > > - list_del_init(page_deferred_list(page));
>> > > > > > - from->deferred_split_queue.split_queue_len--;
>> > > > > > + if (!list_empty(page_deferred_list(page))) {
>> > > > > > + list_del_init(page_deferred_list(page));
>> > > > > > + from->deferred_split_queue.split_queue_len--;
>> > > > > > + }
>> > > > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
>> > > > > > }
>> > > > > > #endif
>> > > > > > @@ -5377,11 +5379,13 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > > page->mem_cgroup = to;
>> > > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> > > > > > - if (compound && list_empty(page_deferred_list(page))) {
>> > > > > > + if (compound) {
>> > > > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
>> > > > > > - list_add_tail(page_deferred_list(page),
>> > > > > > - &to->deferred_split_queue.split_queue);
>> > > > > > - to->deferred_split_queue.split_queue_len++;
>> > > > > > + if (list_empty(page_deferred_list(page))) {
>> > > > > > + list_add_tail(page_deferred_list(page),
>> > > > > > + &to->deferred_split_queue.split_queue);
>> > > > > > + to->deferred_split_queue.split_queue_len++;
>> > > > > > + }
>> > > > > > spin_unlock(&to->deferred_split_queue.split_queue_lock);
>> > > > > > }
>> > > > > > #endif
>> > > > > The patch looks OK for me. But there is another question. I forget, why we unconditionally
>> > > > > add a page with empty deferred list to deferred_split_queue. Shouldn't we also check that
>> > > > > it was initially in the list? Something like:
>> > > > >
>> > > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> > > > > index d4394ae4e5be..0be0136adaa6 100644
>> > > > > --- a/mm/memcontrol.c
>> > > > > +++ b/mm/memcontrol.c
>> > > > > @@ -5289,6 +5289,7 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > struct pglist_data *pgdat;
>> > > > > unsigned long flags;
>> > > > > unsigned int nr_pages = compound ? hpage_nr_pages(page) : 1;
>> > > > > + bool split = false;
>> > > > > int ret;
>> > > > > bool anon;
>> > > > > @@ -5346,6 +5347,7 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > if (!list_empty(page_deferred_list(page))) {
>> > > > > list_del_init(page_deferred_list(page));
>> > > > > from->deferred_split_queue.split_queue_len--;
>> > > > > + split = true;
>> > > > > }
>> > > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
>> > > > > }
>> > > > > @@ -5360,7 +5362,7 @@ static int mem_cgroup_move_account(struct page *page,
>> > > > > page->mem_cgroup = to;
>> > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> > > > > - if (compound) {
>> > > > > + if (compound && split) {
>> > > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
>> > > > > if (list_empty(page_deferred_list(page))) {
>> > > > > list_add_tail(page_deferred_list(page),
>> > > > >
>> > > > I think that's a good point, especially considering that the current code
>> > > > appears to unconditionally place any compound page on the deferred split
>> > > > queue of the destination memcg. The correct list that it should appear
>> > > > on, I believe, depends on whether the pmd has been split for the process
>> > > > being moved: note the MC_TARGET_PAGE caveat in
>> > > > mem_cgroup_move_charge_pte_range() that does not move the charge for
>> > > > compound pages with split pmds. So when mem_cgroup_move_account() is
>> > > > called with compound == true, we're moving the charge of the entire
>> > > > compound page: why would it appear on that memcg's deferred split queue?
>> > > I believe Kirill asked how do we know that the page should be actually
>> > > added to the deferred list just from the list_empty check. In other
>> > > words what if the page hasn't been split at all?
>> > >
>> > Right, and I don't think that it necessarily is and the second
>> > conditional in Wei's patch will always succeed unless we have raced. That
>> > patch is for a lock concern but I think Kirill's question has uncovered
>> > something more interesting.
>> >
>> > Kirill S would definitely be best to answer Kirill T's question, but from
>> > my understanding when mem_cgroup_move_account() is called with
>> > compound == true that we always have an intact pmd (we never migrate
>> > partial page charges for pages on the deferred split queue with the
>> > current charge migration implementation) and thus the underlying page is
>> > not eligible to be split and shouldn't be on the deferred split queue.
>> >
>> > In other words, a page being on the deferred split queue for a memcg
>> > should only happen when it is charged to that memcg. (This wasn't the
>> > case when we only had per-node split queues.) I think that's currently
>> > broken in mem_cgroup_move_account() before Wei's patch.
>> Right. It's broken indeed.
>
>Hmm... Yes, definitely. I wasn't realized this at the first place.
>
>>
>> We are dealing with anon page here. And it cannot be on deferred list as
>> long as it's mapped with PMD. We cannot get compound == true &&
>> !list_empty() on the (first) enter to the function. Any PMD-mapped page
>> will be put onto deferred by the function. This is wrong.
>>
>> The fix is not obvious.
>>
>> This comment got in mem_cgroup_move_charge_pte_range() my attention:
>>
>> /*
>> * We can have a part of the split pmd here. Moving it
>> * can be done but it would be too convoluted so simply
>> * ignore such a partial THP and keep it in original
>> * memcg. There should be somebody mapping the head.
>> */
>>
>> That's exactly the case we care about: PTE-mapped THP that has to be split
>> under load. We don't move charge of them between memcgs and therefore we
>> should not move the page to different memcg.
>>
>> I guess this will do the trick :P
>
>It seems correct to me. In addition, memcg move charge just move PMD mapped
>THP, the THP should be never on the deferred split queue of "from" if it is
>PMD mapped, so actually we don't have to move it to the deferred split queue
>of "to".
>
Well, I got the point.
Since Kirill S found the correct solution, should I prepare v3 or Kirill will
send it?
>>
>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> index c5b5f74cfd4d..e87ee4c10f6e 100644
>> --- a/mm/memcontrol.c
>> +++ b/mm/memcontrol.c
>> @@ -5359,14 +5359,6 @@ static int mem_cgroup_move_account(struct page *page,
>> __mod_lruvec_state(to_vec, NR_WRITEBACK, nr_pages);
>> }
>> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> - if (compound && !list_empty(page_deferred_list(page))) {
>> - spin_lock(&from->deferred_split_queue.split_queue_lock);
>> - list_del_init(page_deferred_list(page));
>> - from->deferred_split_queue.split_queue_len--;
>> - spin_unlock(&from->deferred_split_queue.split_queue_lock);
>> - }
>> -#endif
>> /*
>> * It is safe to change page->mem_cgroup here because the page
>> * is referenced, charged, and isolated - we can't race with
>> @@ -5376,16 +5368,6 @@ static int mem_cgroup_move_account(struct page *page,
>> /* caller should have done css_get */
>> page->mem_cgroup = to;
>> -#ifdef CONFIG_TRANSPARENT_HUGEPAGE
>> - if (compound && list_empty(page_deferred_list(page))) {
>> - spin_lock(&to->deferred_split_queue.split_queue_lock);
>> - list_add_tail(page_deferred_list(page),
>> - &to->deferred_split_queue.split_queue);
>> - to->deferred_split_queue.split_queue_len++;
>> - spin_unlock(&to->deferred_split_queue.split_queue_lock);
>> - }
>> -#endif
>> -
>> spin_unlock_irqrestore(&from->move_lock, flags);
>> ret = 0;
--
Wei Yang
Help you, Help me
^ permalink raw reply [flat|nested] 26+ messages in thread
* Re: [Patch v3] mm: thp: grab the lock before manipulation defer list
2020-01-17 22:18 ` Wei Yang
(?)
@ 2020-01-17 22:57 ` Kirill A. Shutemov
-1 siblings, 0 replies; 26+ messages in thread
From: Kirill A. Shutemov @ 2020-01-17 22:57 UTC (permalink / raw)
To: Wei Yang
Cc: Yang Shi, Kirill A. Shutemov, David Rientjes, Michal Hocko,
Kirill Tkhai, hannes, vdavydov.dev, akpm, cgroups, linux-mm,
linux-kernel, alexander.duyck, stable
On Fri, Jan 17, 2020 at 10:18:59PM +0000, Wei Yang wrote:
> On Fri, Jan 17, 2020 at 11:17:38AM -0800, Yang Shi wrote:
> >
> >
> >On 1/17/20 7:38 AM, Kirill A. Shutemov wrote:
> >> On Fri, Jan 17, 2020 at 01:31:50AM -0800, David Rientjes wrote:
> >> > On Fri, 17 Jan 2020, Michal Hocko wrote:
> >> >
> >> > > On Thu 16-01-20 14:01:59, David Rientjes wrote:
> >> > > > On Thu, 16 Jan 2020, Kirill Tkhai wrote:
> >> > > >
> >> > > > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> >> > > > > > index c5b5f74cfd4d..6450bbe394e2 100644
> >> > > > > > --- a/mm/memcontrol.c
> >> > > > > > +++ b/mm/memcontrol.c
> >> > > > > > @@ -5360,10 +5360,12 @@ static int mem_cgroup_move_account(struct page *page,
> >> > > > > > }
> >> > > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> >> > > > > > - if (compound && !list_empty(page_deferred_list(page))) {
> >> > > > > > + if (compound) {
> >> > > > > > spin_lock(&from->deferred_split_queue.split_queue_lock);
> >> > > > > > - list_del_init(page_deferred_list(page));
> >> > > > > > - from->deferred_split_queue.split_queue_len--;
> >> > > > > > + if (!list_empty(page_deferred_list(page))) {
> >> > > > > > + list_del_init(page_deferred_list(page));
> >> > > > > > + from->deferred_split_queue.split_queue_len--;
> >> > > > > > + }
> >> > > > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
> >> > > > > > }
> >> > > > > > #endif
> >> > > > > > @@ -5377,11 +5379,13 @@ static int mem_cgroup_move_account(struct page *page,
> >> > > > > > page->mem_cgroup = to;
> >> > > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> >> > > > > > - if (compound && list_empty(page_deferred_list(page))) {
> >> > > > > > + if (compound) {
> >> > > > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
> >> > > > > > - list_add_tail(page_deferred_list(page),
> >> > > > > > - &to->deferred_split_queue.split_queue);
> >> > > > > > - to->deferred_split_queue.split_queue_len++;
> >> > > > > > + if (list_empty(page_deferred_list(page))) {
> >> > > > > > + list_add_tail(page_deferred_list(page),
> >> > > > > > + &to->deferred_split_queue.split_queue);
> >> > > > > > + to->deferred_split_queue.split_queue_len++;
> >> > > > > > + }
> >> > > > > > spin_unlock(&to->deferred_split_queue.split_queue_lock);
> >> > > > > > }
> >> > > > > > #endif
> >> > > > > The patch looks OK for me. But there is another question. I forget, why we unconditionally
> >> > > > > add a page with empty deferred list to deferred_split_queue. Shouldn't we also check that
> >> > > > > it was initially in the list? Something like:
> >> > > > >
> >> > > > > diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> >> > > > > index d4394ae4e5be..0be0136adaa6 100644
> >> > > > > --- a/mm/memcontrol.c
> >> > > > > +++ b/mm/memcontrol.c
> >> > > > > @@ -5289,6 +5289,7 @@ static int mem_cgroup_move_account(struct page *page,
> >> > > > > struct pglist_data *pgdat;
> >> > > > > unsigned long flags;
> >> > > > > unsigned int nr_pages = compound ? hpage_nr_pages(page) : 1;
> >> > > > > + bool split = false;
> >> > > > > int ret;
> >> > > > > bool anon;
> >> > > > > @@ -5346,6 +5347,7 @@ static int mem_cgroup_move_account(struct page *page,
> >> > > > > if (!list_empty(page_deferred_list(page))) {
> >> > > > > list_del_init(page_deferred_list(page));
> >> > > > > from->deferred_split_queue.split_queue_len--;
> >> > > > > + split = true;
> >> > > > > }
> >> > > > > spin_unlock(&from->deferred_split_queue.split_queue_lock);
> >> > > > > }
> >> > > > > @@ -5360,7 +5362,7 @@ static int mem_cgroup_move_account(struct page *page,
> >> > > > > page->mem_cgroup = to;
> >> > > > > #ifdef CONFIG_TRANSPARENT_HUGEPAGE
> >> > > > > - if (compound) {
> >> > > > > + if (compound && split) {
> >> > > > > spin_lock(&to->deferred_split_queue.split_queue_lock);
> >> > > > > if (list_empty(page_deferred_list(page))) {
> >> > > > > list_add_tail(page_deferred_list(page),
> >> > > > >
> >> > > > I think that's a good point, especially considering that the current code
> >> > > > appears to unconditionally place any compound page on the deferred split
> >> > > > queue of the destination memcg. The correct list that it should appear
> >> > > > on, I believe, depends on whether the pmd has been split for the process
> >> > > > being moved: note the MC_TARGET_PAGE caveat in
> >> > > > mem_cgroup_move_charge_pte_range() that does not move the charge for
> >> > > > compound pages with split pmds. So when mem_cgroup_move_account() is
> >> > > > called with compound == true, we're moving the charge of the entire
> >> > > > compound page: why would it appear on that memcg's deferred split queue?
> >> > > I believe Kirill asked how do we know that the page should be actually
> >> > > added to the deferred list just from the list_empty check. In other
> >> > > words what if the page hasn't been split at all?
> >> > >
> >> > Right, and I don't think that it necessarily is and the second
> >> > conditional in Wei's patch will always succeed unless we have raced. That
> >> > patch is for a lock concern but I think Kirill's question has uncovered
> >> > something more interesting.
> >> >
> >> > Kirill S would definitely be best to answer Kirill T's question, but from
> >> > my understanding when mem_cgroup_move_account() is called with
> >> > compound == true that we always have an intact pmd (we never migrate
> >> > partial page charges for pages on the deferred split queue with the
> >> > current charge migration implementation) and thus the underlying page is
> >> > not eligible to be split and shouldn't be on the deferred split queue.
> >> >
> >> > In other words, a page being on the deferred split queue for a memcg
> >> > should only happen when it is charged to that memcg. (This wasn't the
> >> > case when we only had per-node split queues.) I think that's currently
> >> > broken in mem_cgroup_move_account() before Wei's patch.
> >> Right. It's broken indeed.
> >
> >Hmm... Yes, definitely. I wasn't realized this at the first place.
> >
> >>
> >> We are dealing with anon page here. And it cannot be on deferred list as
> >> long as it's mapped with PMD. We cannot get compound == true &&
> >> !list_empty() on the (first) enter to the function. Any PMD-mapped page
> >> will be put onto deferred by the function. This is wrong.
> >>
> >> The fix is not obvious.
> >>
> >> This comment got in mem_cgroup_move_charge_pte_range() my attention:
> >>
> >> /*
> >> * We can have a part of the split pmd here. Moving it
> >> * can be done but it would be too convoluted so simply
> >> * ignore such a partial THP and keep it in original
> >> * memcg. There should be somebody mapping the head.
> >> */
> >>
> >> That's exactly the case we care about: PTE-mapped THP that has to be split
> >> under load. We don't move charge of them between memcgs and therefore we
> >> should not move the page to different memcg.
> >>
> >> I guess this will do the trick :P
> >
> >It seems correct to me. In addition, memcg move charge just move PMD mapped
> >THP, the THP should be never on the deferred split queue of "from" if it is
> >PMD mapped, so actually we don't have to move it to the deferred split queue
> >of "to".
> >
>
> Well, I got the point.
>
> Since Kirill S found the correct solution, should I prepare v3 or Kirill will
> send it?
Go ahead. With my Suggested-by.
--
Kirill A. Shutemov
^ permalink raw reply [flat|nested] 26+ messages in thread