| Message ID | 20220616161746.3565225-1-xianting.tian@linux.alibaba.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | mm: page_alloc: validate buddy page before using | expand |
Sorry, please ignore this one. 在 2022/6/17 上午12:17, Xianting Tian 写道: > Commit 787af64d05cd ("mm: page_alloc: validate buddy before check its migratetype.") > fixes a bug in 1dd214b8f21c and there is a similar bug in d9dddbf55667 that > can be fixed in a similar way too. > > In addition, for RISC-V arch the first 2MB RAM could be reserved for opensbi, > so it would have pfn_base=512 and mem_map began with 512th PFN when > CONFIG_FLATMEM=y. > But __find_buddy_pfn algorithm thinks the start pfn 0, it could get 0 pfn or > less than the pfn_base value. We need page_is_buddy() to verify the buddy to > prevent accessing an invalid buddy. > > Fixes: d9dddbf55667 ("mm/page_alloc: prevent merging between isolated and other pageblocks") > Cc: stable@vger.kernel.org > Reported-by: zjb194813@alibaba-inc.com > Reported-by: tianhu.hh@alibaba-inc.com > Signed-off-by: Xianting Tian <xianting.tian@linux.alibaba.com> > --- > mm/page_alloc.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/mm/page_alloc.c b/mm/page_alloc.c > index a6e682569e5b..1c423faa4b62 100644 > --- a/mm/page_alloc.c > +++ b/mm/page_alloc.c > @@ -864,6 +864,9 @@ static inline void __free_one_page(struct page *page, > > buddy_idx = __find_buddy_index(page_idx, order); > buddy = page + (buddy_idx - page_idx); > + > + if (!page_is_buddy(page, buddy, order)) > + goto done_merging; > buddy_mt = get_pageblock_migratetype(buddy); > > if (migratetype != buddy_mt
On Fri, Jun 17, 2022 at 12:20:19AM +0800, Xianting Tian wrote:
> Sorry, please ignore this one.
Which "one"? This was a series :(
Sorry for the misleading, I mean just this one in the series, not the whole series, actually this patch is for 4.9, But I forget the add the Label 4.9. So I send the patch for 4.9 again in https://lkml.org/lkml/2022/6/16/782 <https://lkml.org/lkml/2022/6/16/782> 在 2022/6/17 上午12:50, Greg KH 写道: > On Fri, Jun 17, 2022 at 12:20:19AM +0800, Xianting Tian wrote: >> Sorry, please ignore this one. > Which "one"? This was a series :(
diff --git a/mm/page_alloc.c b/mm/page_alloc.c index a6e682569e5b..1c423faa4b62 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -864,6 +864,9 @@ static inline void __free_one_page(struct page *page, buddy_idx = __find_buddy_index(page_idx, order); buddy = page + (buddy_idx - page_idx); + + if (!page_is_buddy(page, buddy, order)) + goto done_merging; buddy_mt = get_pageblock_migratetype(buddy); if (migratetype != buddy_mt
Commit 787af64d05cd ("mm: page_alloc: validate buddy before check its migratetype.") fixes a bug in 1dd214b8f21c and there is a similar bug in d9dddbf55667 that can be fixed in a similar way too. In addition, for RISC-V arch the first 2MB RAM could be reserved for opensbi, so it would have pfn_base=512 and mem_map began with 512th PFN when CONFIG_FLATMEM=y. But __find_buddy_pfn algorithm thinks the start pfn 0, it could get 0 pfn or less than the pfn_base value. We need page_is_buddy() to verify the buddy to prevent accessing an invalid buddy. Fixes: d9dddbf55667 ("mm/page_alloc: prevent merging between isolated and other pageblocks") Cc: stable@vger.kernel.org Reported-by: zjb194813@alibaba-inc.com Reported-by: tianhu.hh@alibaba-inc.com Signed-off-by: Xianting Tian <xianting.tian@linux.alibaba.com> --- mm/page_alloc.c | 3 +++ 1 file changed, 3 insertions(+)