Skip to contentSkip to conversation

mm/page_alloc.c: clean code by merging two functions

RFC PATCH Applied [email protected] 3 participants 3 messages Sep 16, 2020 – Sep 21, 2020

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

RFC PATCH 1/1 mm/page_alloc.c: clean code by merging two functions

From: Mateusz Nosek <[email protected]>

The 'finalise_ac' function is just 'epilogue' for 'prepare_alloc_pages'. Therefore there is no need to keep them both so 'finalise_ac' content can be merged into 'prepare_alloc_pages' code. It would make '__alloc_pages_nodemask' cleaner when it comes to readability.

---
 mm/page_alloc.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
mm/page_alloc.c +2 -8
index b9bd75cacf02..a094e297c88f 100644
@@ -4824,12 +4824,6 @@ static inline bool prepare_alloc_pages(gfp_t gfp_mask, unsigned int order,
 
 	*alloc_flags = current_alloc_flags(gfp_mask, *alloc_flags);
 
-	return true;
-}
-
-/* Determine whether to spread dirty pages and what the first usable zone */
-static inline void finalise_ac(gfp_t gfp_mask, struct alloc_context *ac)
-{
 	/* Dirty zone balancing only done in the fast path */
 	ac->spread_dirty_pages = (gfp_mask & __GFP_WRITE);
 
@@ -4840,6 +4834,8 @@ static inline void finalise_ac(gfp_t gfp_mask, struct alloc_context *ac)
 	 */
 	ac->preferred_zoneref = first_zones_zonelist(ac->zonelist,
 					ac->highest_zoneidx, ac->nodemask);
+
+	return true;
 }
 
 /*
@@ -4868,8 +4864,6 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, int preferred_nid,
 	if (!prepare_alloc_pages(gfp_mask, order, preferred_nid, nodemask, &ac, &alloc_mask, &alloc_flags))
 		return NULL;
 
-	finalise_ac(gfp_mask, &ac);
-
 	/*
 	 * Forbid the first pass from falling back to types that fragment
 	 * memory until all local zones are considered.
Signature · 1 line
2.20.1
#1 lore raw

Hi,

(added Mel)

Quoted 6 lines · On Wed, Sep 16, 2020 at 01:01:18PM +0200, [email protected] wrote:

From: Mateusz Nosek <[email protected]>

The 'finalise_ac' function is just 'epilogue' for 'prepare_alloc_pages'. Therefore there is no need to keep them both so 'finalise_ac' content can be merged into 'prepare_alloc_pages' code. It would make '__alloc_pages_nodemask' cleaner when it comes to readability.

These functions were split from alloc_pages_nodemask() by commit 9cd7555875bb ("mm, page_alloc: split alloc_pages_nodemask()"), so at the very least I'd add Mel to 'cc.

Quoted 44 lines
---
 mm/page_alloc.c | 10 ++--------
 1 file changed, 2 insertions(+), 8 deletions(-)
mm/page_alloc.c +2 -8
index b9bd75cacf02..a094e297c88f 100644
@@ -4824,12 +4824,6 @@ static inline bool prepare_alloc_pages(gfp_t gfp_mask, unsigned int order,
 
 	*alloc_flags = current_alloc_flags(gfp_mask, *alloc_flags);
 
-	return true;
-}
-
-/* Determine whether to spread dirty pages and what the first usable zone */
-static inline void finalise_ac(gfp_t gfp_mask, struct alloc_context *ac)
-{
 	/* Dirty zone balancing only done in the fast path */
 	ac->spread_dirty_pages = (gfp_mask & __GFP_WRITE);
 
@@ -4840,6 +4834,8 @@ static inline void finalise_ac(gfp_t gfp_mask, struct alloc_context *ac)
 	 */
 	ac->preferred_zoneref = first_zones_zonelist(ac->zonelist,
 					ac->highest_zoneidx, ac->nodemask);
+
+	return true;
 }
 
 /*
@@ -4868,8 +4864,6 @@ __alloc_pages_nodemask(gfp_t gfp_mask, unsigned int order, int preferred_nid,
 	if (!prepare_alloc_pages(gfp_mask, order, preferred_nid, nodemask, &ac, &alloc_mask, &alloc_flags))
 		return NULL;
 
-	finalise_ac(gfp_mask, &ac);
-
 	/*
 	 * Forbid the first pass from falling back to types that fragment
 	 * memory until all local zones are considered.
Signature · 1 line
2.20.1
Signature · 2 lines
Sincerely yours,
Mike.
#2 lore raw
Quoted 7 lines · Mike Rapoport

Hi,

(added Mel)

Quoted 2 lines · On Wed, Sep 16, 2020 at 01:01:18PM +0200, [email protected] wrote:

From: Mateusz Nosek <[email protected]>

Quoted 9 lines
Quoted 4 lines

The 'finalise_ac' function is just 'epilogue' for 'prepare_alloc_pages'. Therefore there is no need to keep them both so 'finalise_ac' content can be merged into 'prepare_alloc_pages' code. It would make '__alloc_pages_nodemask' cleaner when it comes to readability.

These functions were split from alloc_pages_nodemask() by commit 9cd7555875bb ("mm, page_alloc: split alloc_pages_nodemask()"), so at the very least I'd add Mel to 'cc.

The changelog explains the intent. At the time, there was demand for a bulk page allocator for networking support and one was prototyped
https://lore.kernel.org/lkml/[email protected]/
but the networking side to consume the interface was never finished.

Signature · 2 lines
Mel Gorman
SUSE Labs