diff mbox series

mm/vmscan: drop checking if _deferred_list is empty before using TTU_SYNC

Message ID 20240629234155.53524-1-21cnbao@gmail.com (mailing list archive)
State New
Headers show
Series mm/vmscan: drop checking if _deferred_list is empty before using TTU_SYNC | expand

Commit Message

Barry Song June 29, 2024, 11:41 p.m. UTC
From: Barry Song <v-songbaohua@oppo.com>

The optimization of list_empty(&folio->_deferred_list) aimed to prevent
increasing the PTL duration when a large folio is partially unmapped,
for example, from subpage 0 to subpage (nr - 2).

But Ryan's commit 5ed890ce5147 ("mm: vmscan: avoid split during
shrink_folio_list()") actually splits this kind of large folios. This
makes the "optimization" useless.

Signed-off-by: Barry Song <v-songbaohua@oppo.com>
Cc: David Hildenbrand <david@redhat.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
---
 -v1
 * I remember David and Ryan once suggested that this check could be dropped
   while the patch was being pulled into mm-unstable. However, for some reason,
   I forgot to request squashing this change.

 mm/vmscan.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Ryan Roberts July 1, 2024, 8:48 a.m. UTC | #1
On 30/06/2024 00:41, Barry Song wrote:
> From: Barry Song <v-songbaohua@oppo.com>
> 
> The optimization of list_empty(&folio->_deferred_list) aimed to prevent
> increasing the PTL duration when a large folio is partially unmapped,
> for example, from subpage 0 to subpage (nr - 2).
> 
> But Ryan's commit 5ed890ce5147 ("mm: vmscan: avoid split during
> shrink_folio_list()") actually splits this kind of large folios. This
> makes the "optimization" useless.
> 
> Signed-off-by: Barry Song <v-songbaohua@oppo.com>
> Cc: David Hildenbrand <david@redhat.com>
> Cc: Ryan Roberts <ryan.roberts@arm.com>

Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>

> ---
>  -v1
>  * I remember David and Ryan once suggested that this check could be dropped
>    while the patch was being pulled into mm-unstable. However, for some reason,
>    I forgot to request squashing this change.
> 
>  mm/vmscan.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 3d4c681c6d40..0761f91b407f 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -1291,7 +1291,7 @@ static unsigned int shrink_folio_list(struct list_head *folio_list,
>  			 * try_to_unmap acquire PTL from the first PTE,
>  			 * eliminating the influence of temporary PTE values.
>  			 */
> -			if (folio_test_large(folio) && list_empty(&folio->_deferred_list))
> +			if (folio_test_large(folio))

As it stands, the list_empty() technically needs a data_race() annotation. I
think your original patch went into v6.10-rc1? If so, perhaps it makes sense to
try to get this into the next rc?

>  				flags |= TTU_SYNC;
>  
>  			try_to_unmap(folio, flags);
David Hildenbrand July 1, 2024, 8:51 a.m. UTC | #2
On 30.06.24 01:41, Barry Song wrote:
> From: Barry Song <v-songbaohua@oppo.com>
> 
> The optimization of list_empty(&folio->_deferred_list) aimed to prevent
> increasing the PTL duration when a large folio is partially unmapped,
> for example, from subpage 0 to subpage (nr - 2).
> 
> But Ryan's commit 5ed890ce5147 ("mm: vmscan: avoid split during
> shrink_folio_list()") actually splits this kind of large folios. This
> makes the "optimization" useless.
> 
> Signed-off-by: Barry Song <v-songbaohua@oppo.com>
> Cc: David Hildenbrand <david@redhat.com>
> Cc: Ryan Roberts <ryan.roberts@arm.com>
> ---
>   -v1
>   * I remember David and Ryan once suggested that this check could be dropped
>     while the patch was being pulled into mm-unstable. However, for some reason,
>     I forgot to request squashing this change.
> 
>   mm/vmscan.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/mm/vmscan.c b/mm/vmscan.c
> index 3d4c681c6d40..0761f91b407f 100644
> --- a/mm/vmscan.c
> +++ b/mm/vmscan.c
> @@ -1291,7 +1291,7 @@ static unsigned int shrink_folio_list(struct list_head *folio_list,
>   			 * try_to_unmap acquire PTL from the first PTE,
>   			 * eliminating the influence of temporary PTE values.
>   			 */
> -			if (folio_test_large(folio) && list_empty(&folio->_deferred_list))
> +			if (folio_test_large(folio))
>   				flags |= TTU_SYNC;
>   
>   			try_to_unmap(folio, flags);

Reviewed-by: David Hildenbrand <david@redhat.com>
Barry Song July 1, 2024, 9:35 a.m. UTC | #3
On Mon, Jul 1, 2024 at 8:48 PM Ryan Roberts <ryan.roberts@arm.com> wrote:
>
> On 30/06/2024 00:41, Barry Song wrote:
> > From: Barry Song <v-songbaohua@oppo.com>
> >
> > The optimization of list_empty(&folio->_deferred_list) aimed to prevent
> > increasing the PTL duration when a large folio is partially unmapped,
> > for example, from subpage 0 to subpage (nr - 2).
> >
> > But Ryan's commit 5ed890ce5147 ("mm: vmscan: avoid split during
> > shrink_folio_list()") actually splits this kind of large folios. This
> > makes the "optimization" useless.
> >
> > Signed-off-by: Barry Song <v-songbaohua@oppo.com>
> > Cc: David Hildenbrand <david@redhat.com>
> > Cc: Ryan Roberts <ryan.roberts@arm.com>
>
> Reviewed-by: Ryan Roberts <ryan.roberts@arm.com>
>
> > ---
> >  -v1
> >  * I remember David and Ryan once suggested that this check could be dropped
> >    while the patch was being pulled into mm-unstable. However, for some reason,
> >    I forgot to request squashing this change.
> >
> >  mm/vmscan.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/mm/vmscan.c b/mm/vmscan.c
> > index 3d4c681c6d40..0761f91b407f 100644
> > --- a/mm/vmscan.c
> > +++ b/mm/vmscan.c
> > @@ -1291,7 +1291,7 @@ static unsigned int shrink_folio_list(struct list_head *folio_list,
> >                        * try_to_unmap acquire PTL from the first PTE,
> >                        * eliminating the influence of temporary PTE values.
> >                        */
> > -                     if (folio_test_large(folio) && list_empty(&folio->_deferred_list))
> > +                     if (folio_test_large(folio))
>
> As it stands, the list_empty() technically needs a data_race() annotation. I
> think your original patch went into v6.10-rc1? If so, perhaps it makes sense to
> try to get this into the next rc?

Either way is fine. Hi Andrew, if you include it for the next RC, could you
please add this tag?

Fixes: 73bc32875ee9 ("mm: hold PTL from the first PTE while reclaiming
a large folio")

And an additional changelog:

Additionally, the list_empty() technically requires a data_race() annotation.

>
> >                               flags |= TTU_SYNC;
> >
> >                       try_to_unmap(folio, flags);
>

Thanks
Barry
Andrew Morton July 2, 2024, 6:32 a.m. UTC | #4
On Mon, 1 Jul 2024 21:35:41 +1200 Barry Song <21cnbao@gmail.com> wrote:

> > > +++ b/mm/vmscan.c
> > > @@ -1291,7 +1291,7 @@ static unsigned int shrink_folio_list(struct list_head *folio_list,
> > >                        * try_to_unmap acquire PTL from the first PTE,
> > >                        * eliminating the influence of temporary PTE values.
> > >                        */
> > > -                     if (folio_test_large(folio) && list_empty(&folio->_deferred_list))
> > > +                     if (folio_test_large(folio))
> >
> > As it stands, the list_empty() technically needs a data_race() annotation. I
> > think your original patch went into v6.10-rc1? If so, perhaps it makes sense to
> > try to get this into the next rc?
> 
> Either way is fine. Hi Andrew, if you include it for the next RC, could you
> please add this tag?
> 
> Fixes: 73bc32875ee9 ("mm: hold PTL from the first PTE while reclaiming
> a large folio")

Nope!  That invites people to backport this change into kernels which
don't contain Ryan's 5ed890ce5147 ("mm: vmscan: avoid split during
shrink_folio_list()").

I'll simply queue this in the normal fashion.

> And an additional changelog:
> 
> Additionally, the list_empty() technically requires a data_race() annotation.

Done, thanks.
diff mbox series

Patch

diff --git a/mm/vmscan.c b/mm/vmscan.c
index 3d4c681c6d40..0761f91b407f 100644
--- a/mm/vmscan.c
+++ b/mm/vmscan.c
@@ -1291,7 +1291,7 @@  static unsigned int shrink_folio_list(struct list_head *folio_list,
 			 * try_to_unmap acquire PTL from the first PTE,
 			 * eliminating the influence of temporary PTE values.
 			 */
-			if (folio_test_large(folio) && list_empty(&folio->_deferred_list))
+			if (folio_test_large(folio))
 				flags |= TTU_SYNC;
 
 			try_to_unmap(folio, flags);