| Message ID | 20210605010626.1459873-1-almasrymina@google.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | mm, hugetlb: fix allocation error check and copy func name | expand |
On Sat, Jun 5, 2021 at 9:07 AM Mina Almasry <almasrymina@google.com> wrote: > > The alloc_huge_page_vma() doesn't return error pointers, it returns > NULL. > > There is an existing migrate_page_copy() which not only copies the page > contents, but also page state/metadata. People could get confused that > 'migrate_page_copy' and 'migrate_copy_huge_page' do not have the same > functionality. Rename to 'copy_huge_page()' to make it less likely > folks get confused. > > Fixes: 6a0badc266df ("mm, hugetlb: fix racy resv_huge_pages underflow on > UFFDIO_COPY") I think the commit id is not stable. So this really should squash the commit of "mm, hugetlb: fix racy resv_huge_pages underflow on UFFDIO_COPY". But anyway, Reviewed-by: Muchun Song <songmuchun@bytedance.com> > > Signed-off-by: Mina Almasry <almasrymina@google.com> > Cc: linux-mm@kvack.org > Cc: Mike Kravetz <mike.kravetz@oracle.com> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: linux-kernel@vger.kernel.org
On 6/4/21 6:06 PM, Mina Almasry wrote: > The alloc_huge_page_vma() doesn't return error pointers, it returns > NULL. > > There is an existing migrate_page_copy() which not only copies the page > contents, but also page state/metadata. People could get confused that > 'migrate_page_copy' and 'migrate_copy_huge_page' do not have the same > functionality. Rename to 'copy_huge_page()' to make it less likely > folks get confused. > > Fixes: 6a0badc266df ("mm, hugetlb: fix racy resv_huge_pages underflow on > UFFDIO_COPY") > > Signed-off-by: Mina Almasry <almasrymina@google.com> > Cc: linux-mm@kvack.org > Cc: Mike Kravetz <mike.kravetz@oracle.com> > Cc: Andrew Morton <akpm@linux-foundation.org> > Cc: linux-kernel@vger.kernel.org > > --- > include/linux/migrate.h | 4 ++-- > mm/hugetlb.c | 4 ++-- > mm/migrate.c | 4 ++-- > 3 files changed, 6 insertions(+), 6 deletions(-) As mentioned by Muchun, Andrew should squash this and original patch. With the combined patches, Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
diff --git a/include/linux/migrate.h b/include/linux/migrate.h index e0b98876a52e..9b7b7cd3bae9 100644 --- a/include/linux/migrate.h +++ b/include/linux/migrate.h @@ -51,7 +51,7 @@ extern int migrate_huge_page_move_mapping(struct address_space *mapping, struct page *newpage, struct page *page); extern int migrate_page_move_mapping(struct address_space *mapping, struct page *newpage, struct page *page, int extra_count); -extern void migrate_copy_huge_page(struct page *dst, struct page *src); +extern void copy_huge_page(struct page *dst, struct page *src); #else static inline void putback_movable_pages(struct list_head *l) {} @@ -78,7 +78,7 @@ static inline int migrate_huge_page_move_mapping(struct address_space *mapping, return -ENOSYS; } -static inline void migrate_copy_huge_page(struct page *dst, struct page *src) +static inline void copy_huge_page(struct page *dst, struct page *src) { } #endif /* CONFIG_MIGRATION */ diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 64a460ff2d5b..637395907908 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5130,7 +5130,7 @@ int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm, * contents. */ page = alloc_huge_page_vma(h, dst_vma, dst_addr); - if (IS_ERR(page)) { + if (!page) { ret = -ENOMEM; goto out; } @@ -5156,7 +5156,7 @@ int hugetlb_mcopy_atomic_pte(struct mm_struct *dst_mm, *pagep = NULL; goto out; } - migrate_copy_huge_page(page, *pagep); + copy_huge_page(page, *pagep); put_page(*pagep); *pagep = NULL; } diff --git a/mm/migrate.c b/mm/migrate.c index 2515210a9747..51f355757490 100644 --- a/mm/migrate.c +++ b/mm/migrate.c @@ -557,7 +557,7 @@ static void __copy_gigantic_page(struct page *dst, struct page *src, } } -void migrate_copy_huge_page(struct page *dst, struct page *src) +void copy_huge_page(struct page *dst, struct page *src) { int i; int nr_pages; @@ -662,7 +662,7 @@ EXPORT_SYMBOL(migrate_page_states); void migrate_page_copy(struct page *newpage, struct page *page) { if (PageHuge(page) || PageTransHuge(page)) - migrate_copy_huge_page(newpage, page); + copy_huge_page(newpage, page); else copy_highpage(newpage, page);
The alloc_huge_page_vma() doesn't return error pointers, it returns NULL. There is an existing migrate_page_copy() which not only copies the page contents, but also page state/metadata. People could get confused that 'migrate_page_copy' and 'migrate_copy_huge_page' do not have the same functionality. Rename to 'copy_huge_page()' to make it less likely folks get confused. Fixes: 6a0badc266df ("mm, hugetlb: fix racy resv_huge_pages underflow on UFFDIO_COPY") Signed-off-by: Mina Almasry <almasrymina@google.com> Cc: linux-mm@kvack.org Cc: Mike Kravetz <mike.kravetz@oracle.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: linux-kernel@vger.kernel.org --- include/linux/migrate.h | 4 ++-- mm/hugetlb.c | 4 ++-- mm/migrate.c | 4 ++-- 3 files changed, 6 insertions(+), 6 deletions(-) -- 2.32.0.rc1.229.g3e70b5a671-goog