diff mbox series

[v3,1/1] page_frag: Recover from memory pressure

Message ID 20201115201029.11903-1-dongli.zhang@oracle.com (mailing list archive)
State New
Headers show
Series [v3,1/1] page_frag: Recover from memory pressure | expand

Commit Message

Dongli Zhang Nov. 15, 2020, 8:10 p.m. UTC
The ethernet driver may allocate skb (and skb->data) via napi_alloc_skb().
This ends up to page_frag_alloc() to allocate skb->data from
page_frag_cache->va.

During the memory pressure, page_frag_cache->va may be allocated as
pfmemalloc page. As a result, the skb->pfmemalloc is always true as
skb->data is from page_frag_cache->va. The skb will be dropped if the
sock (receiver) does not have SOCK_MEMALLOC. This is expected behaviour
under memory pressure.

However, once kernel is not under memory pressure any longer (suppose large
amount of memory pages are just reclaimed), the page_frag_alloc() may still
re-use the prior pfmemalloc page_frag_cache->va to allocate skb->data. As a
result, the skb->pfmemalloc is always true unless page_frag_cache->va is
re-allocated, even if the kernel is not under memory pressure any longer.

Here is how kernel runs into issue.

1. The kernel is under memory pressure and allocation of
PAGE_FRAG_CACHE_MAX_ORDER in __page_frag_cache_refill() will fail. Instead,
the pfmemalloc page is allocated for page_frag_cache->va.

2: All skb->data from page_frag_cache->va (pfmemalloc) will have
skb->pfmemalloc=true. The skb will always be dropped by sock without
SOCK_MEMALLOC. This is an expected behaviour.

3. Suppose a large amount of pages are reclaimed and kernel is not under
memory pressure any longer. We expect skb->pfmemalloc drop will not happen.

4. Unfortunately, page_frag_alloc() does not proactively re-allocate
page_frag_alloc->va and will always re-use the prior pfmemalloc page. The
skb->pfmemalloc is always true even kernel is not under memory pressure any
longer.

Fix this by freeing and re-allocating the page instead of recycling it.

References: https://lore.kernel.org/lkml/20201103193239.1807-1-dongli.zhang@oracle.com/
References: https://lore.kernel.org/linux-mm/20201105042140.5253-1-willy@infradead.org/
Suggested-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Aruna Ramakrishna <aruna.ramakrishna@oracle.com>
Cc: Bert Barbe <bert.barbe@oracle.com>
Cc: Rama Nichanamatlu <rama.nichanamatlu@oracle.com>
Cc: Venkat Venkatsubra <venkat.x.venkatsubra@oracle.com>
Cc: Manjunath Patil <manjunath.b.patil@oracle.com>
Cc: Joe Jin <joe.jin@oracle.com>
Cc: SRINIVAS <srinivas.eeda@oracle.com>
Cc: stable@vger.kernel.org
Fixes: 79930f5892e ("net: do not deplete pfmemalloc reserve")
Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
---
Changed since v1:
  - change author from Matthew to Dongli
  - Add references to all prior discussions
  - Add more details to commit message
Changed since v2:
  - add unlikely (suggested by Eric Dumazet)

 mm/page_alloc.c | 5 +++++
 1 file changed, 5 insertions(+)

Comments

Eric Dumazet Nov. 16, 2020, 8:37 a.m. UTC | #1
On Sun, Nov 15, 2020 at 9:16 PM Dongli Zhang <dongli.zhang@oracle.com> wrote:
>
> The ethernet driver may allocate skb (and skb->data) via napi_alloc_skb().
> This ends up to page_frag_alloc() to allocate skb->data from
> page_frag_cache->va.
>
> During the memory pressure, page_frag_cache->va may be allocated as
> pfmemalloc page. As a result, the skb->pfmemalloc is always true as
> skb->data is from page_frag_cache->va. The skb will be dropped if the
> sock (receiver) does not have SOCK_MEMALLOC. This is expected behaviour
> under memory pressure.
...
> References: https://lore.kernel.org/lkml/20201103193239.1807-1-dongli.zhang@oracle.com/
> References: https://lore.kernel.org/linux-mm/20201105042140.5253-1-willy@infradead.org/
> Suggested-by: Matthew Wilcox (Oracle) <willy@infradead.org>
> Cc: Aruna Ramakrishna <aruna.ramakrishna@oracle.com>
> Cc: Bert Barbe <bert.barbe@oracle.com>
> Cc: Rama Nichanamatlu <rama.nichanamatlu@oracle.com>
> Cc: Venkat Venkatsubra <venkat.x.venkatsubra@oracle.com>
> Cc: Manjunath Patil <manjunath.b.patil@oracle.com>
> Cc: Joe Jin <joe.jin@oracle.com>
> Cc: SRINIVAS <srinivas.eeda@oracle.com>
> Cc: stable@vger.kernel.org
> Fixes: 79930f5892e ("net: do not deplete pfmemalloc reserve")
> Signed-off-by: Dongli Zhang <dongli.zhang@oracle.com>
> Acked-by: Vlastimil Babka <vbabka@suse.cz>
> ---
> Changed since v1:
>   - change author from Matthew to Dongli
>   - Add references to all prior discussions
>   - Add more details to commit message
> Changed since v2:
>   - add unlikely (suggested by Eric Dumazet)
>
>  mm/page_alloc.c | 5 +++++
>  1 file changed, 5 insertions(+)
>
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 23f5066bd4a5..91129ce75ed4 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -5103,6 +5103,11 @@ void *page_frag_alloc(struct page_frag_cache *nc,
>                 if (!page_ref_sub_and_test(page, nc->pagecnt_bias))
>                         goto refill;
>
> +               if (unlikely(nc->pfmemalloc)) {
> +                       free_the_page(page, compound_order(page));
> +                       goto refill;
> +               }
> +

Reviewed-by: Eric Dumazet <edumazet@google.com>

Thanks !
Jakub Kicinski Nov. 18, 2020, 7:46 p.m. UTC | #2
On Sun, 15 Nov 2020 12:10:29 -0800 Dongli Zhang wrote:
> The ethernet driver may allocate skb (and skb->data) via napi_alloc_skb().
> This ends up to page_frag_alloc() to allocate skb->data from
> page_frag_cache->va.
> 
> During the memory pressure, page_frag_cache->va may be allocated as
> pfmemalloc page. As a result, the skb->pfmemalloc is always true as
> skb->data is from page_frag_cache->va. The skb will be dropped if the
> sock (receiver) does not have SOCK_MEMALLOC. This is expected behaviour
> under memory pressure.
> 
> However, once kernel is not under memory pressure any longer (suppose large
> amount of memory pages are just reclaimed), the page_frag_alloc() may still
> re-use the prior pfmemalloc page_frag_cache->va to allocate skb->data. As a
> result, the skb->pfmemalloc is always true unless page_frag_cache->va is
> re-allocated, even if the kernel is not under memory pressure any longer.
> 
> Here is how kernel runs into issue.
> 
> 1. The kernel is under memory pressure and allocation of
> PAGE_FRAG_CACHE_MAX_ORDER in __page_frag_cache_refill() will fail. Instead,
> the pfmemalloc page is allocated for page_frag_cache->va.
> 
> 2: All skb->data from page_frag_cache->va (pfmemalloc) will have
> skb->pfmemalloc=true. The skb will always be dropped by sock without
> SOCK_MEMALLOC. This is an expected behaviour.
> 
> 3. Suppose a large amount of pages are reclaimed and kernel is not under
> memory pressure any longer. We expect skb->pfmemalloc drop will not happen.
> 
> 4. Unfortunately, page_frag_alloc() does not proactively re-allocate
> page_frag_alloc->va and will always re-use the prior pfmemalloc page. The
> skb->pfmemalloc is always true even kernel is not under memory pressure any
> longer.
> 
> Fix this by freeing and re-allocating the page instead of recycling it.

Andrew, are you taking this via -mm or should I put it in net? 
I'm sending a PR to Linus tomorrow.
Andrew Morton Nov. 18, 2020, 9:13 p.m. UTC | #3
On Wed, 18 Nov 2020 11:46:54 -0800 Jakub Kicinski <kuba@kernel.org> wrote:

> > 1. The kernel is under memory pressure and allocation of
> > PAGE_FRAG_CACHE_MAX_ORDER in __page_frag_cache_refill() will fail. Instead,
> > the pfmemalloc page is allocated for page_frag_cache->va.
> > 
> > 2: All skb->data from page_frag_cache->va (pfmemalloc) will have
> > skb->pfmemalloc=true. The skb will always be dropped by sock without
> > SOCK_MEMALLOC. This is an expected behaviour.
> > 
> > 3. Suppose a large amount of pages are reclaimed and kernel is not under
> > memory pressure any longer. We expect skb->pfmemalloc drop will not happen.
> > 
> > 4. Unfortunately, page_frag_alloc() does not proactively re-allocate
> > page_frag_alloc->va and will always re-use the prior pfmemalloc page. The
> > skb->pfmemalloc is always true even kernel is not under memory pressure any
> > longer.
> > 
> > Fix this by freeing and re-allocating the page instead of recycling it.
> 
> Andrew, are you taking this via -mm or should I put it in net? 
> I'm sending a PR to Linus tomorrow.

Please go ahead - if/when it appears in mainline or linux-next, I'll
drop the -mm copy.
Jakub Kicinski Nov. 18, 2020, 11:23 p.m. UTC | #4
On Wed, 18 Nov 2020 13:13:35 -0800 Andrew Morton wrote:
> On Wed, 18 Nov 2020 11:46:54 -0800 Jakub Kicinski <kuba@kernel.org> wrote:
> 
> > > 1. The kernel is under memory pressure and allocation of
> > > PAGE_FRAG_CACHE_MAX_ORDER in __page_frag_cache_refill() will fail. Instead,
> > > the pfmemalloc page is allocated for page_frag_cache->va.
> > > 
> > > 2: All skb->data from page_frag_cache->va (pfmemalloc) will have
> > > skb->pfmemalloc=true. The skb will always be dropped by sock without
> > > SOCK_MEMALLOC. This is an expected behaviour.
> > > 
> > > 3. Suppose a large amount of pages are reclaimed and kernel is not under
> > > memory pressure any longer. We expect skb->pfmemalloc drop will not happen.
> > > 
> > > 4. Unfortunately, page_frag_alloc() does not proactively re-allocate
> > > page_frag_alloc->va and will always re-use the prior pfmemalloc page. The
> > > skb->pfmemalloc is always true even kernel is not under memory pressure any
> > > longer.
> > > 
> > > Fix this by freeing and re-allocating the page instead of recycling it.  
> > 
> > Andrew, are you taking this via -mm or should I put it in net? 
> > I'm sending a PR to Linus tomorrow.  
> 
> Please go ahead - if/when it appears in mainline or linux-next, I'll
> drop the -mm copy.  

Okay, applied, thank you!
diff mbox series

Patch

diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 23f5066bd4a5..91129ce75ed4 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -5103,6 +5103,11 @@  void *page_frag_alloc(struct page_frag_cache *nc,
 		if (!page_ref_sub_and_test(page, nc->pagecnt_bias))
 			goto refill;
 
+		if (unlikely(nc->pfmemalloc)) {
+			free_the_page(page, compound_order(page));
+			goto refill;
+		}
+
 #if (PAGE_SIZE < PAGE_FRAG_CACHE_MAX_SIZE)
 		/* if size can vary use size else just use PAGE_SIZE */
 		size = nc->size;