| Message ID | 20210110124017.86750-3-songmuchun@bytedance.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Fix some bugs about HugeTLB code | expand |
On 1/10/21 4:40 AM, Muchun Song wrote: > If a new hugetlb page is allocated during fallocate it will not be > marked as active (set_page_huge_active) which will result in a later > isolate_huge_page failure when the page migration code would like to > move that page. Such a failure would be unexpected and wrong. > > Only export set_page_huge_active, just leave clear_page_huge_active > as static. Because there are no external users. > > Fixes: 70c3547e36f5 (hugetlbfs: add hugetlbfs_fallocate()) > Signed-off-by: Muchun Song <songmuchun@bytedance.com> > Cc: stable@vger.kernel.org > --- > fs/hugetlbfs/inode.c | 3 ++- > include/linux/hugetlb.h | 2 ++ > mm/hugetlb.c | 2 +- > 3 files changed, 5 insertions(+), 2 deletions(-) Thanks. Although page_huge_active is declared in page-flags.h, I much prefer the declaration of set_page_huge_active to be in hugetlb.c. Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
On Sun 10-01-21 20:40:13, Muchun Song wrote: > If a new hugetlb page is allocated during fallocate it will not be > marked as active (set_page_huge_active) which will result in a later > isolate_huge_page failure when the page migration code would like to > move that page. Such a failure would be unexpected and wrong. > > Only export set_page_huge_active, just leave clear_page_huge_active > as static. Because there are no external users. > > Fixes: 70c3547e36f5 (hugetlbfs: add hugetlbfs_fallocate()) > Signed-off-by: Muchun Song <songmuchun@bytedance.com> > Cc: stable@vger.kernel.org Acked-by: Michal Hocko <mhocko@suse.com> Thanks! > --- > fs/hugetlbfs/inode.c | 3 ++- > include/linux/hugetlb.h | 2 ++ > mm/hugetlb.c | 2 +- > 3 files changed, 5 insertions(+), 2 deletions(-) > > diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c > index b5c109703daa..21c20fd5f9ee 100644 > --- a/fs/hugetlbfs/inode.c > +++ b/fs/hugetlbfs/inode.c > @@ -735,9 +735,10 @@ static long hugetlbfs_fallocate(struct file *file, int mode, loff_t offset, > > mutex_unlock(&hugetlb_fault_mutex_table[hash]); > > + set_page_huge_active(page); > /* > * unlock_page because locked by add_to_page_cache() > - * page_put due to reference from alloc_huge_page() > + * put_page() due to reference from alloc_huge_page() > */ > unlock_page(page); > put_page(page); > diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h > index ebca2ef02212..b5807f23caf8 100644 > --- a/include/linux/hugetlb.h > +++ b/include/linux/hugetlb.h > @@ -770,6 +770,8 @@ static inline void huge_ptep_modify_prot_commit(struct vm_area_struct *vma, > } > #endif > > +void set_page_huge_active(struct page *page); > + > #else /* CONFIG_HUGETLB_PAGE */ > struct hstate {}; > > diff --git a/mm/hugetlb.c b/mm/hugetlb.c > index 1f3bf1710b66..4741d60f8955 100644 > --- a/mm/hugetlb.c > +++ b/mm/hugetlb.c > @@ -1348,7 +1348,7 @@ bool page_huge_active(struct page *page) > } > > /* never called for tail page */ > -static void set_page_huge_active(struct page *page) > +void set_page_huge_active(struct page *page) > { > VM_BUG_ON_PAGE(!PageHeadHuge(page), page); > SetPagePrivate(&page[1]); > -- > 2.11.0
diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index b5c109703daa..21c20fd5f9ee 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -735,9 +735,10 @@ static long hugetlbfs_fallocate(struct file *file, int mode, loff_t offset, mutex_unlock(&hugetlb_fault_mutex_table[hash]); + set_page_huge_active(page); /* * unlock_page because locked by add_to_page_cache() - * page_put due to reference from alloc_huge_page() + * put_page() due to reference from alloc_huge_page() */ unlock_page(page); put_page(page); diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index ebca2ef02212..b5807f23caf8 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -770,6 +770,8 @@ static inline void huge_ptep_modify_prot_commit(struct vm_area_struct *vma, } #endif +void set_page_huge_active(struct page *page); + #else /* CONFIG_HUGETLB_PAGE */ struct hstate {}; diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 1f3bf1710b66..4741d60f8955 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1348,7 +1348,7 @@ bool page_huge_active(struct page *page) } /* never called for tail page */ -static void set_page_huge_active(struct page *page) +void set_page_huge_active(struct page *page) { VM_BUG_ON_PAGE(!PageHeadHuge(page), page); SetPagePrivate(&page[1]);
If a new hugetlb page is allocated during fallocate it will not be marked as active (set_page_huge_active) which will result in a later isolate_huge_page failure when the page migration code would like to move that page. Such a failure would be unexpected and wrong. Only export set_page_huge_active, just leave clear_page_huge_active as static. Because there are no external users. Fixes: 70c3547e36f5 (hugetlbfs: add hugetlbfs_fallocate()) Signed-off-by: Muchun Song <songmuchun@bytedance.com> Cc: stable@vger.kernel.org --- fs/hugetlbfs/inode.c | 3 ++- include/linux/hugetlb.h | 2 ++ mm/hugetlb.c | 2 +- 3 files changed, 5 insertions(+), 2 deletions(-)