sparc: get rid of superfluous __GFP_REPEAT
authorMichal Hocko <mhocko@suse.com>
Fri, 24 Jun 2016 21:49:14 +0000 (14:49 -0700)
committerLinus Torvalds <torvalds@linux-foundation.org>
Sat, 25 Jun 2016 00:23:52 +0000 (17:23 -0700)
__GFP_REPEAT has a rather weak semantic but since it has been introduced
around 2.6.12 it has been ignored for low order allocations.

{pud,pmd}_alloc_one is using __GFP_REPEAT but it always allocates from
pgtable_cache which is initialzed to PAGE_SIZE objects.  This means that
this flag has never been actually useful here because it has always been
used only for PAGE_ALLOC_COSTLY requests.

Link: http://lkml.kernel.org/r/1464599699-30131-13-git-send-email-mhocko@kernel.org
Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
arch/sparc/include/asm/pgalloc_64.h

index 5e3187185b4a8b9c3f22d4a67ea538a12fdcd9e7..3529f1378cd8d6230a0ad3c76cec74b874a94b2d 100644 (file)
@@ -41,8 +41,7 @@ static inline void __pud_populate(pud_t *pud, pmd_t *pmd)
 
 static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)
 {
-       return kmem_cache_alloc(pgtable_cache,
-                               GFP_KERNEL|__GFP_REPEAT);
+       return kmem_cache_alloc(pgtable_cache, GFP_KERNEL);
 }
 
 static inline void pud_free(struct mm_struct *mm, pud_t *pud)
@@ -52,8 +51,7 @@ static inline void pud_free(struct mm_struct *mm, pud_t *pud)
 
 static inline pmd_t *pmd_alloc_one(struct mm_struct *mm, unsigned long addr)
 {
-       return kmem_cache_alloc(pgtable_cache,
-                               GFP_KERNEL|__GFP_REPEAT);
+       return kmem_cache_alloc(pgtable_cache, GFP_KERNEL);
 }
 
 static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd)