Re: [PATCH RFC 2/2] mm, hugetlb: do not rely on overcommit limit during migration

[Michal Hocko]

On Tue 28-11-17 15:12:11, Michal Hocko wrote:
[...]
> +/*
> + * Internal hugetlb specific page flag. Do not use outside of the hugetlb
> + * code
> + */
> +static inline bool PageHugeTemporary(struct page *page)
> +{
> +     if (!PageHuge(page))
> +             return false;
> +
> +     return page[2].flags == -1U;
> +}
> +
> +static inline void SetPageHugeTemporary(struct page *page)
> +{
> +     page[2].flags = -1U;
> +}
> +
> +static inline void ClearPageHugeTemporary(struct page *page)
> +{
> +     page[2].flags = 0;
> +}

Ups, this is obviously not OK. I was just lucky to not hit BUG_ONs
because I am clearly overwriting node/zone data in flags. I will have to
find something else to abuse. I will go with my favorite mapping pointer
which is not used at all.
---
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 1be43563e226..db7544a0b7b6 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1259,17 +1259,17 @@ static inline bool PageHugeTemporary(struct page *page)
        if (!PageHuge(page))
                return false;
 
-       return page[2].flags == -1U;
+       return page[2].mapping == -1U;
 }
 
 static inline void SetPageHugeTemporary(struct page *page)
 {
-       page[2].flags = -1U;
+       page[2].mapping = -1U;
 }
 
 static inline void ClearPageHugeTemporary(struct page *page)
 {
-       page[2].flags = 0;
+       page[2].mapping = NULL;
 }
 
 void free_huge_page(struct page *page)
-- 
Michal Hocko
SUSE Labs