Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1591871 > unrolled thread

[PATCH 1/2] mm: use is_migrate_highatomic() to simplify the code

Back to article view | Back to linux.kernel

Contents

  [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the code Xishi Qiu <qiuxishi@huawei.com> - 2017-03-03 12:30 +0100
    Re: [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the code Michal Hocko <mhocko@kernel.org> - 2017-03-03 17:00 +0100
      Re: [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the  code Andrew Morton <akpm@linux-foundation.org> - 2017-03-04 00:20 +0100
        Re: [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the code Michal Hocko <mhocko@kernel.org> - 2017-03-06 15:20 +0100
          Re: [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the  code Andrew Morton <akpm@linux-foundation.org> - 2017-03-06 22:00 +0100
            Re: [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the code Michal Hocko <mhocko@kernel.org> - 2017-03-08 16:00 +0100

#1591871 — [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the code

Introduce two helpers, is_migrate_highatomic() and is_migrate_highatomic_page().
Simplify the code, no functional changes.

Signed-off-by: Xishi Qiu <qiuxishi@huawei.com>
---
 include/linux/mmzone.h |  5 +++++
 mm/page_alloc.c        | 14 ++++++--------
 2 files changed, 11 insertions(+), 8 deletions(-)

diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index 8e02b37..8124440 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -66,6 +66,11 @@ enum {
 /* In mm/page_alloc.c; keep in sync also with show_migration_types() there */
 extern char * const migratetype_names[MIGRATE_TYPES];
 
+#define is_migrate_highatomic(migratetype)				\
+	(migratetype == MIGRATE_HIGHATOMIC)
+#define is_migrate_highatomic_page(_page)				\
+	(get_pageblock_migratetype(_page) == MIGRATE_HIGHATOMIC)
+
 #ifdef CONFIG_CMA
 #  define is_migrate_cma(migratetype) unlikely((migratetype) == MIGRATE_CMA)
 #  define is_migrate_cma_page(_page) (get_pageblock_migratetype(_page) == MIGRATE_CMA)
diff --git a/mm/page_alloc.c b/mm/page_alloc.c
index 9f9623d..40d79a6 100644
--- a/mm/page_alloc.c
+++ b/mm/page_alloc.c
@@ -2040,8 +2040,8 @@ static void reserve_highatomic_pageblock(struct page *page, struct zone *zone,
 
 	/* Yoink! */
 	mt = get_pageblock_migratetype(page);
-	if (mt != MIGRATE_HIGHATOMIC &&
-			!is_migrate_isolate(mt) && !is_migrate_cma(mt)) {
+	if (!is_migrate_highatomic(mt) && !is_migrate_isolate(mt)
+	    && !is_migrate_cma(mt)) {
 		zone->nr_reserved_highatomic += pageblock_nr_pages;
 		set_pageblock_migratetype(page, MIGRATE_HIGHATOMIC);
 		move_freepages_block(zone, page, MIGRATE_HIGHATOMIC);
@@ -2098,8 +2098,7 @@ static bool unreserve_highatomic_pageblock(const struct alloc_context *ac,
 			 * from highatomic to ac->migratetype. So we should
 			 * adjust the count once.
 			 */
-			if (get_pageblock_migratetype(page) ==
-							MIGRATE_HIGHATOMIC) {
+			if (is_migrate_highatomic_page(page)) {
 				/*
 				 * It should never happen but changes to
 				 * locking could inadvertently allow a per-cpu
@@ -2156,8 +2155,7 @@ static bool unreserve_highatomic_pageblock(const struct alloc_context *ac,
 
 		page = list_first_entry(&area->free_list[fallback_mt],
 						struct page, lru);
-		if (can_steal &&
-			get_pageblock_migratetype(page) != MIGRATE_HIGHATOMIC)
+		if (can_steal && !is_migrate_highatomic_page(page))
 			steal_suitable_fallback(zone, page, start_migratetype);
 
 		/* Remove the page from the freelists */
@@ -2494,7 +2492,7 @@ void free_hot_cold_page(struct page *page, bool cold)
 	/*
 	 * We only track unmovable, reclaimable and movable on pcp lists.
 	 * Free ISOLATE pages back to the allocator because they are being
-	 * offlined but treat RESERVE as movable pages so we can get those
+	 * offlined but treat HIGHATOMIC as movable pages so we can get those
 	 * areas back if necessary. Otherwise, we may have to free
 	 * excessively into the page allocator
 	 */
@@ -2605,7 +2603,7 @@ int __isolate_free_page(struct page *page, unsigned int order)
 		for (; page < endpage; page += pageblock_nr_pages) {
 			int mt = get_pageblock_migratetype(page);
 			if (!is_migrate_isolate(mt) && !is_migrate_cma(mt)
-				&& mt != MIGRATE_HIGHATOMIC)
+			    && !is_migrate_highatomic(mt))
 				set_pageblock_migratetype(page,
 							  MIGRATE_MOVABLE);
 		}
-- 
1.8.3.1

[toc] | [next] | [standalone]

#1592090

On Fri 03-03-17 19:10:13, Xishi Qiu wrote:
> Introduce two helpers, is_migrate_highatomic() and is_migrate_highatomic_page().
> Simplify the code, no functional changes.

static inline helpers would be nicer than macros

> Signed-off-by: Xishi Qiu <qiuxishi@huawei.com>

OK this fits with other MIGRATE_$FOO types

Acked-by: Michal Hocko <mhocko@suse.com>

> ---
>  include/linux/mmzone.h |  5 +++++
>  mm/page_alloc.c        | 14 ++++++--------
>  2 files changed, 11 insertions(+), 8 deletions(-)
> 
> diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
> index 8e02b37..8124440 100644
> --- a/include/linux/mmzone.h
> +++ b/include/linux/mmzone.h
> @@ -66,6 +66,11 @@ enum {
>  /* In mm/page_alloc.c; keep in sync also with show_migration_types() there */
>  extern char * const migratetype_names[MIGRATE_TYPES];
>  
> +#define is_migrate_highatomic(migratetype)				\
> +	(migratetype == MIGRATE_HIGHATOMIC)
> +#define is_migrate_highatomic_page(_page)				\
> +	(get_pageblock_migratetype(_page) == MIGRATE_HIGHATOMIC)
> +
>  #ifdef CONFIG_CMA
>  #  define is_migrate_cma(migratetype) unlikely((migratetype) == MIGRATE_CMA)
>  #  define is_migrate_cma_page(_page) (get_pageblock_migratetype(_page) == MIGRATE_CMA)
> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> index 9f9623d..40d79a6 100644
> --- a/mm/page_alloc.c
> +++ b/mm/page_alloc.c
> @@ -2040,8 +2040,8 @@ static void reserve_highatomic_pageblock(struct page *page, struct zone *zone,
>  
>  	/* Yoink! */
>  	mt = get_pageblock_migratetype(page);
> -	if (mt != MIGRATE_HIGHATOMIC &&
> -			!is_migrate_isolate(mt) && !is_migrate_cma(mt)) {
> +	if (!is_migrate_highatomic(mt) && !is_migrate_isolate(mt)
> +	    && !is_migrate_cma(mt)) {
>  		zone->nr_reserved_highatomic += pageblock_nr_pages;
>  		set_pageblock_migratetype(page, MIGRATE_HIGHATOMIC);
>  		move_freepages_block(zone, page, MIGRATE_HIGHATOMIC);
> @@ -2098,8 +2098,7 @@ static bool unreserve_highatomic_pageblock(const struct alloc_context *ac,
>  			 * from highatomic to ac->migratetype. So we should
>  			 * adjust the count once.
>  			 */
> -			if (get_pageblock_migratetype(page) ==
> -							MIGRATE_HIGHATOMIC) {
> +			if (is_migrate_highatomic_page(page)) {
>  				/*
>  				 * It should never happen but changes to
>  				 * locking could inadvertently allow a per-cpu
> @@ -2156,8 +2155,7 @@ static bool unreserve_highatomic_pageblock(const struct alloc_context *ac,
>  
>  		page = list_first_entry(&area->free_list[fallback_mt],
>  						struct page, lru);
> -		if (can_steal &&
> -			get_pageblock_migratetype(page) != MIGRATE_HIGHATOMIC)
> +		if (can_steal && !is_migrate_highatomic_page(page))
>  			steal_suitable_fallback(zone, page, start_migratetype);
>  
>  		/* Remove the page from the freelists */
> @@ -2494,7 +2492,7 @@ void free_hot_cold_page(struct page *page, bool cold)
>  	/*
>  	 * We only track unmovable, reclaimable and movable on pcp lists.
>  	 * Free ISOLATE pages back to the allocator because they are being
> -	 * offlined but treat RESERVE as movable pages so we can get those
> +	 * offlined but treat HIGHATOMIC as movable pages so we can get those
>  	 * areas back if necessary. Otherwise, we may have to free
>  	 * excessively into the page allocator
>  	 */
> @@ -2605,7 +2603,7 @@ int __isolate_free_page(struct page *page, unsigned int order)
>  		for (; page < endpage; page += pageblock_nr_pages) {
>  			int mt = get_pageblock_migratetype(page);
>  			if (!is_migrate_isolate(mt) && !is_migrate_cma(mt)
> -				&& mt != MIGRATE_HIGHATOMIC)
> +			    && !is_migrate_highatomic(mt))
>  				set_pageblock_migratetype(page,
>  							  MIGRATE_MOVABLE);
>  		}
> -- 
> 1.8.3.1
> 

-- 
Michal Hocko
SUSE Labs

[toc] | [prev] | [next] | [standalone]

#1592338 — Re: [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the code

On Fri, 3 Mar 2017 14:18:08 +0100 Michal Hocko <mhocko@kernel.org> wrote:

> On Fri 03-03-17 19:10:13, Xishi Qiu wrote:
> > Introduce two helpers, is_migrate_highatomic() and is_migrate_highatomic_page().
> > Simplify the code, no functional changes.
> 
> static inline helpers would be nicer than macros

Always.

We made a big dependency mess in mmzone.h.  internal.h works.

--- a/include/linux/mmzone.h~mm-use-is_migrate_highatomic-to-simplify-the-code-fix
+++ a/include/linux/mmzone.h
@@ -35,7 +35,7 @@
  */
 #define PAGE_ALLOC_COSTLY_ORDER 3
 
-enum {
+enum migratetype {
 	MIGRATE_UNMOVABLE,
 	MIGRATE_MOVABLE,
 	MIGRATE_RECLAIMABLE,
@@ -66,11 +66,6 @@ enum {
 /* In mm/page_alloc.c; keep in sync also with show_migration_types() there */
 extern char * const migratetype_names[MIGRATE_TYPES];
 
-#define is_migrate_highatomic(migratetype)				\
-	(migratetype == MIGRATE_HIGHATOMIC)
-#define is_migrate_highatomic_page(_page)				\
-	(get_pageblock_migratetype(_page) == MIGRATE_HIGHATOMIC)
-
 #ifdef CONFIG_CMA
 #  define is_migrate_cma(migratetype) unlikely((migratetype) == MIGRATE_CMA)
 #  define is_migrate_cma_page(_page) (get_pageblock_migratetype(_page) == MIGRATE_CMA)
diff -puN mm/page_alloc.c~mm-use-is_migrate_highatomic-to-simplify-the-code-fix mm/page_alloc.c
diff -puN mm/internal.h~mm-use-is_migrate_highatomic-to-simplify-the-code-fix mm/internal.h
--- a/mm/internal.h~mm-use-is_migrate_highatomic-to-simplify-the-code-fix
+++ a/mm/internal.h
@@ -503,4 +503,14 @@ extern const struct trace_print_flags pa
 extern const struct trace_print_flags vmaflag_names[];
 extern const struct trace_print_flags gfpflag_names[];
 
+static inline bool is_migrate_highatomic(enum migratetype migratetype)
+{
+	return migratetype == MIGRATE_HIGHATOMIC;
+}
+
+static inline bool is_migrate_highatomic_page(struct page *page)
+{
+	return get_pageblock_migratetype(page) == MIGRATE_HIGHATOMIC;
+}
+
 #endif	/* __MM_INTERNAL_H */
_

[toc] | [prev] | [next] | [standalone]

#1593394

On Fri 03-03-17 15:06:19, Andrew Morton wrote:
> On Fri, 3 Mar 2017 14:18:08 +0100 Michal Hocko <mhocko@kernel.org> wrote:
> 
> > On Fri 03-03-17 19:10:13, Xishi Qiu wrote:
> > > Introduce two helpers, is_migrate_highatomic() and is_migrate_highatomic_page().
> > > Simplify the code, no functional changes.
> > 
> > static inline helpers would be nicer than macros
> 
> Always.
> 
> We made a big dependency mess in mmzone.h.  internal.h works.

Just too bad we have three different header files for
is_migrate_isolate{_page} - include/linux/page-isolation.h
is_migrate_cma{_page} - include/linux/mmzone.h
is_migrate_highatomic{_page} - mm/internal.h

I guess we want all of them in internal.h?

-- 
Michal Hocko
SUSE Labs

[toc] | [prev] | [next] | [standalone]

#1593728 — Re: [PATCH 1/2] mm: use is_migrate_highatomic() to simplify the code

On Mon, 6 Mar 2017 14:38:33 +0100 Michal Hocko <mhocko@kernel.org> wrote:

> On Fri 03-03-17 15:06:19, Andrew Morton wrote:
> > On Fri, 3 Mar 2017 14:18:08 +0100 Michal Hocko <mhocko@kernel.org> wrote:
> > 
> > > On Fri 03-03-17 19:10:13, Xishi Qiu wrote:
> > > > Introduce two helpers, is_migrate_highatomic() and is_migrate_highatomic_page().
> > > > Simplify the code, no functional changes.
> > > 
> > > static inline helpers would be nicer than macros
> > 
> > Always.
> > 
> > We made a big dependency mess in mmzone.h.  internal.h works.
> 
> Just too bad we have three different header files for
> is_migrate_isolate{_page} - include/linux/page-isolation.h
> is_migrate_cma{_page} - include/linux/mmzone.h
> is_migrate_highatomic{_page} - mm/internal.h
> 
> I guess we want all of them in internal.h?

I suppose so.  arch/powerpc/mm/mmu_context_iommu.c is using
is_migrate_cma_page which would need some attention.  

[toc] | [prev] | [next] | [standalone]

#1595230

On Mon 06-03-17 12:43:18, Andrew Morton wrote:
> On Mon, 6 Mar 2017 14:38:33 +0100 Michal Hocko <mhocko@kernel.org> wrote:
> 
> > On Fri 03-03-17 15:06:19, Andrew Morton wrote:
> > > On Fri, 3 Mar 2017 14:18:08 +0100 Michal Hocko <mhocko@kernel.org> wrote:
> > > 
> > > > On Fri 03-03-17 19:10:13, Xishi Qiu wrote:
> > > > > Introduce two helpers, is_migrate_highatomic() and is_migrate_highatomic_page().
> > > > > Simplify the code, no functional changes.
> > > > 
> > > > static inline helpers would be nicer than macros
> > > 
> > > Always.
> > > 
> > > We made a big dependency mess in mmzone.h.  internal.h works.
> > 
> > Just too bad we have three different header files for
> > is_migrate_isolate{_page} - include/linux/page-isolation.h
> > is_migrate_cma{_page} - include/linux/mmzone.h
> > is_migrate_highatomic{_page} - mm/internal.h
> > 
> > I guess we want all of them in internal.h?
> 
> I suppose so.  arch/powerpc/mm/mmu_context_iommu.c is using
> is_migrate_cma_page which would need some attention.  

Ble. I am really wondering whether these helpers actually make the end
result really worth all the troubles and the mess.

-- 
Michal Hocko
SUSE Labs

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web