score: get rid of superfluous __GFP_REPEAT
authorMichal Hocko <[email protected]>
Fri, 24 Jun 2016 21:49:09 +0000 (14:49 -0700)
committerLinus Torvalds <[email protected]>
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.

pte_alloc_one{_kernel} allocate PTE_ORDER which is 0.  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/[email protected]
Signed-off-by: Michal Hocko <[email protected]>
Cc: Chen Liqin <[email protected]>
Cc: Lennox Wu <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
arch/score/include/asm/pgalloc.h

index 2e067657db988d66d8fbd01912fbb4729d934ad3..49b012d78c1ac9a6868a1a08f88c54e415c76028 100644 (file)
@@ -42,8 +42,7 @@ static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm,
 {
        pte_t *pte;
 
-       pte = (pte_t *) __get_free_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO,
-                                       PTE_ORDER);
+       pte = (pte_t *) __get_free_pages(GFP_KERNEL|__GFP_ZERO, PTE_ORDER);
 
        return pte;
 }
@@ -53,7 +52,7 @@ static inline struct page *pte_alloc_one(struct mm_struct *mm,
 {
        struct page *pte;
 
-       pte = alloc_pages(GFP_KERNEL | __GFP_REPEAT, PTE_ORDER);
+       pte = alloc_pages(GFP_KERNEL, PTE_ORDER);
        if (!pte)
                return NULL;
        clear_highpage(pte);