[patch] mm: introduce a common interface for balloon pages mobility fix

[patch] mm: introduce a common interface for balloon pages mobility fix

[David Rientjes]

Fixes build failure from "mm: introduce a common interface for balloon 
pages mobility":

mm/balloon_compaction.c: In function 'balloon_page_putback':
mm/balloon_compaction.c:243: error: implicit declaration of function '__WARN'

Generic code calls WARN_ON(), not __WARN() directly.

Signed-off-by: David Rientjes <rientjes@google.com>
---
 mm/balloon_compaction.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/mm/balloon_compaction.c b/mm/balloon_compaction.c
index 32927eb..ddb0951 100644
--- a/mm/balloon_compaction.c
+++ b/mm/balloon_compaction.c
@@ -240,7 +240,7 @@ void balloon_page_putback(struct page *page)
 		put_page(page);
 		balloon_event_count(COMPACTBALLOONRETURNED);
 	} else {
-		__WARN();
+		WARN_ON(1);
 		dump_page(page);
 	}
 	unlock_page(page);

[memcg:since-3.6 365/437] warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)

[kbuild test robot]

tree:   git://git.kernel.org/pub/scm/linux/kernel/git/mhocko/mm.git since-3.6
head:   12dfb061e5fd15be23451418da01281625c0eeae
commit: 86929cfa5f751de3d8be5a846535282730865d8a [365/437] mm: introduce a common interface for balloon pages mobility
config: make ARCH=sh allyesconfig

All warnings:

warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
--
warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)

---
0-DAY kernel build testing backend         Open Source Technology Center
Fengguang Wu, Yuanhan Liu                              Intel Corporation

--
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>

[patch] mm: introduce a common interface for balloon pages mobility fix

[David Rientjes]

On Fri, 16 Nov 2012, kbuild test robot wrote:

> tree:   git://git.kernel.org/pub/scm/linux/kernel/git/mhocko/mm.git since-3.6
> head:   12dfb061e5fd15be23451418da01281625c0eeae
> commit: 86929cfa5f751de3d8be5a846535282730865d8a [365/437] mm: introduce a common interface for balloon pages mobility
> config: make ARCH=sh allyesconfig
> 
> All warnings:
> 
> warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> --
> warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> 

mm: introduce a common interface for balloon pages mobility fix

CONFIG_BALLOON_COMPACTION shouldn't be selecting options that may not be 
supported, so make it depend on memory compaction rather than selecting 
it.  CONFIG_COMPACTION is enabled by default for all configs that support 
it.
    
Signed-off-by: David Rientjes <rientjes@google.com>
---
 mm/Kconfig |    3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/mm/Kconfig b/mm/Kconfig
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -191,8 +191,7 @@ config SPLIT_PTLOCK_CPUS
 # support for memory balloon compaction
 config BALLOON_COMPACTION
 	bool "Allow for balloon memory compaction/migration"
-	select COMPACTION
-	depends on VIRTIO_BALLOON
+	depends on VIRTIO_BALLOON && COMPACTION
 	help
 	  Memory fragmentation introduced by ballooning might reduce
 	  significantly the number of 2MB contiguous memory blocks that can be

--
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>

Re: [patch] mm: introduce a common interface for balloon pages mobility fix

[Rafael Aquini]

On Fri, Nov 16, 2012 at 11:50:12AM -0800, David Rientjes wrote:
> On Fri, 16 Nov 2012, kbuild test robot wrote:
> 
> > tree:   git://git.kernel.org/pub/scm/linux/kernel/git/mhocko/mm.git since-3.6
> > head:   12dfb061e5fd15be23451418da01281625c0eeae
> > commit: 86929cfa5f751de3d8be5a846535282730865d8a [365/437] mm: introduce a common interface for balloon pages mobility
> > config: make ARCH=sh allyesconfig
> > 
> > All warnings:
> > 
> > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > --
> > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > 
> 
> mm: introduce a common interface for balloon pages mobility fix
> 
> CONFIG_BALLOON_COMPACTION shouldn't be selecting options that may not be 
> supported, so make it depend on memory compaction rather than selecting 
> it.  CONFIG_COMPACTION is enabled by default for all configs that support 
> it.
>     

This is already addressed at v12 revision which is on mmtom queue already.


> Signed-off-by: David Rientjes <rientjes@google.com>
> ---
>  mm/Kconfig |    3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/mm/Kconfig b/mm/Kconfig
> --- a/mm/Kconfig
> +++ b/mm/Kconfig
> @@ -191,8 +191,7 @@ config SPLIT_PTLOCK_CPUS
>  # support for memory balloon compaction
>  config BALLOON_COMPACTION
>  	bool "Allow for balloon memory compaction/migration"
> -	select COMPACTION
> -	depends on VIRTIO_BALLOON
> +	depends on VIRTIO_BALLOON && COMPACTION
>  	help
>  	  Memory fragmentation introduced by ballooning might reduce
>  	  significantly the number of 2MB contiguous memory blocks that can be

--
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>

Re: [patch] mm: introduce a common interface for balloon pages mobility fix

[Rafael Aquini]

On Fri, Nov 16, 2012 at 11:50:12AM -0800, David Rientjes wrote:
> On Fri, 16 Nov 2012, kbuild test robot wrote:
> 
> > tree:   git://git.kernel.org/pub/scm/linux/kernel/git/mhocko/mm.git since-3.6
> > head:   12dfb061e5fd15be23451418da01281625c0eeae
> > commit: 86929cfa5f751de3d8be5a846535282730865d8a [365/437] mm: introduce a common interface for balloon pages mobility
> > config: make ARCH=sh allyesconfig
> > 
> > All warnings:
> > 
> > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > --
> > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > 
> 
> mm: introduce a common interface for balloon pages mobility fix
> 
> CONFIG_BALLOON_COMPACTION shouldn't be selecting options that may not be 
> supported, so make it depend on memory compaction rather than selecting 
> it.  CONFIG_COMPACTION is enabled by default for all configs that support 
> it.
>     

This is addressed by the submitted v12 review, which is on mmtom queue already.


> Signed-off-by: David Rientjes <rientjes@google.com>
> ---
>  mm/Kconfig |    3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/mm/Kconfig b/mm/Kconfig
> --- a/mm/Kconfig
> +++ b/mm/Kconfig
> @@ -191,8 +191,7 @@ config SPLIT_PTLOCK_CPUS
>  # support for memory balloon compaction
>  config BALLOON_COMPACTION
>  	bool "Allow for balloon memory compaction/migration"
> -	select COMPACTION
> -	depends on VIRTIO_BALLOON
> +	depends on VIRTIO_BALLOON && COMPACTION
>  	help
>  	  Memory fragmentation introduced by ballooning might reduce
>  	  significantly the number of 2MB contiguous memory blocks that can be

--
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>

Re: [patch] mm: introduce a common interface for balloon pages mobility fix

[David Rientjes]

On Fri, 16 Nov 2012, Rafael Aquini wrote:

> > > tree:   git://git.kernel.org/pub/scm/linux/kernel/git/mhocko/mm.git since-3.6
> > > head:   12dfb061e5fd15be23451418da01281625c0eeae
> > > commit: 86929cfa5f751de3d8be5a846535282730865d8a [365/437] mm: introduce a common interface for balloon pages mobility
> > > config: make ARCH=sh allyesconfig
> > > 
> > > All warnings:
> > > 
> > > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > > --
> > > warning: (BALLOON_COMPACTION && TRANSPARENT_HUGEPAGE) selects COMPACTION which has unmet direct dependencies (MMU)
> > > 
> > 
> > mm: introduce a common interface for balloon pages mobility fix
> > 
> > CONFIG_BALLOON_COMPACTION shouldn't be selecting options that may not be 
> > supported, so make it depend on memory compaction rather than selecting 
> > it.  CONFIG_COMPACTION is enabled by default for all configs that support 
> > it.
> >     
> 
> This is already addressed at v12 revision which is on mmtom queue already.
> 

Ah, it was merged yesterday but it depends on TRANSPARENT_HUGEPAGE rather 
than COMPACTION.  Why?  Surely thp is not the only thing that benefits 
from keeping memory defragmented for high-order allocs.

--
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>

Re: [patch] mm: introduce a common interface for balloon pages mobility fix

[David Rientjes]

On Fri, 16 Nov 2012, David Rientjes wrote:

> > This is already addressed at v12 revision which is on mmtom queue already.
> > 
> 
> Ah, it was merged yesterday but it depends on TRANSPARENT_HUGEPAGE rather 
> than COMPACTION.  Why?  Surely thp is not the only thing that benefits 
> from keeping memory defragmented for high-order allocs.
> 

Ping on this?  The direct question was why depend on TRANSPARENT_HUGEPAGE 
rather than COMPACTION, i.e. why do only order-9 allocations benefit from 
defragmented memory rather than all high order allocations?

Thanks.

--
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>

Re: [patch] mm: introduce a common interface for balloon pages mobility fix

[Rafael Aquini]

On Tue, Nov 27, 2012 at 03:50:07PM -0800, David Rientjes wrote:
> On Fri, 16 Nov 2012, David Rientjes wrote:
> 
> > > This is already addressed at v12 revision which is on mmtom queue already.
> > > 
> > 
> > Ah, it was merged yesterday but it depends on TRANSPARENT_HUGEPAGE rather 
> > than COMPACTION.  Why?  Surely thp is not the only thing that benefits 
> > from keeping memory defragmented for high-order allocs.
> > 
> 
> Ping on this?  The direct question was why depend on TRANSPARENT_HUGEPAGE 
> rather than COMPACTION, i.e. why do only order-9 allocations benefit from 
> defragmented memory rather than all high order allocations?
> 
> Thanks.

Ugh, I missed this one on the mail pile.

That's a nice feature for all system, indeed. I'd say you're 100% right. Would
you mind in submitting the change?

Thanks David, and sorry sorry for the late reply.

Rafael

--
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>

[patch] mm: introduce a common interface for balloon pages mobility fix

[David Rientjes]

It's useful to keep memory defragmented so that all high-order page 
allocations have a chance to succeed, not simply transparent hugepages.  
Thus, allow balloon compaction for any system with memory compaction 
enabled, which is the defconfig.

Signed-off-by: David Rientjes <rientjes@google.com>
---
 mm/Kconfig |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/mm/Kconfig b/mm/Kconfig
--- a/mm/Kconfig
+++ b/mm/Kconfig
@@ -200,7 +200,7 @@ config SPLIT_PTLOCK_CPUS
 config BALLOON_COMPACTION
 	bool "Allow for balloon memory compaction/migration"
 	def_bool y
-	depends on TRANSPARENT_HUGEPAGE && VIRTIO_BALLOON
+	depends on COMPACTION && VIRTIO_BALLOON
 	help
 	  Memory fragmentation introduced by ballooning might reduce
 	  significantly the number of 2MB contiguous memory blocks that can be

--
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>

Re: [patch] mm: introduce a common interface for balloon pages mobility fix

[Michal Hocko]

On Tue 27-11-12 16:15:59, David Rientjes wrote:
> It's useful to keep memory defragmented so that all high-order page 
> allocations have a chance to succeed, not simply transparent hugepages.  
> Thus, allow balloon compaction for any system with memory compaction 
> enabled, which is the defconfig.
> 
> Signed-off-by: David Rientjes <rientjes@google.com>

Yes, makes sense.
Acked-by: Michal Hocko <mhocko@suse.cz>

Thanks

> ---
>  mm/Kconfig |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/mm/Kconfig b/mm/Kconfig
> --- a/mm/Kconfig
> +++ b/mm/Kconfig
> @@ -200,7 +200,7 @@ config SPLIT_PTLOCK_CPUS
>  config BALLOON_COMPACTION
>  	bool "Allow for balloon memory compaction/migration"
>  	def_bool y
> -	depends on TRANSPARENT_HUGEPAGE && VIRTIO_BALLOON
> +	depends on COMPACTION && VIRTIO_BALLOON
>  	help
>  	  Memory fragmentation introduced by ballooning might reduce
>  	  significantly the number of 2MB contiguous memory blocks that can be

-- 
Michal Hocko
SUSE Labs

--
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>

Re: [patch] mm: introduce a common interface for balloon pages mobility fix

[Rafael Aquini]

On Tue, Nov 27, 2012 at 04:15:59PM -0800, David Rientjes wrote:
> It's useful to keep memory defragmented so that all high-order page 
> allocations have a chance to succeed, not simply transparent hugepages.  
> Thus, allow balloon compaction for any system with memory compaction 
> enabled, which is the defconfig.
> 
> Signed-off-by: David Rientjes <rientjes@google.com>
> ---
Thanks David

Acked-by: Rafael Aquini <aquini@redhat.com>


>  mm/Kconfig |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/mm/Kconfig b/mm/Kconfig
> --- a/mm/Kconfig
> +++ b/mm/Kconfig
> @@ -200,7 +200,7 @@ config SPLIT_PTLOCK_CPUS
>  config BALLOON_COMPACTION
>  	bool "Allow for balloon memory compaction/migration"
>  	def_bool y
> -	depends on TRANSPARENT_HUGEPAGE && VIRTIO_BALLOON
> +	depends on COMPACTION && VIRTIO_BALLOON
>  	help
>  	  Memory fragmentation introduced by ballooning might reduce
>  	  significantly the number of 2MB contiguous memory blocks that can be

--
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>