linux-kernel - Re: [PATCH RFC v2 1/3] mm: Add TAIL_MAPPING_REUSED

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <ZNp32fml0bKuFfQw@x1n>
Date:   Mon, 14 Aug 2023 14:52:09 -0400
From:   Peter Xu <peterx@...hat.com>
To:     linux-mm@...ck.org, linux-kernel@...r.kernel.org
Cc:     Mike Kravetz <mike.kravetz@...cle.com>,
        David Hildenbrand <david@...hat.com>,
        Matthew Wilcox <willy@...radead.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Yu Zhao <yuzhao@...gle.com>,
        Ryan Roberts <ryan.roberts@....com>,
        Yang Shi <shy828301@...il.com>,
        Hugh Dickins <hughd@...gle.com>,
        "Kirill A . Shutemov" <kirill@...temov.name>
Subject: Re: [PATCH RFC v2 1/3] mm: Add TAIL_MAPPING_REUSED_MAX

On Mon, Aug 14, 2023 at 02:44:09PM -0400, Peter Xu wrote:
> Tail pages have a sanity check on ->mapping fields, not all of them but
> only upon index>2.  It's because we reused ->mapping fields of the tail
> pages index=0,1 for other things.
> 
> Define a macro for "max index of tail pages that got ->mapping field
> reused" on top of folio definition, because when we grow folio tail pages
> we'd want to boost this too together.
> 
> Signed-off-by: Peter Xu <peterx@...hat.com>
> ---
>  include/linux/mm_types.h | 9 +++++++++
>  mm/huge_memory.c         | 6 +++---
>  2 files changed, 12 insertions(+), 3 deletions(-)
> 
> diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
> index 291c05cacd48..3f2b0d46f5d6 100644
> --- a/include/linux/mm_types.h
> +++ b/include/linux/mm_types.h
> @@ -248,6 +248,15 @@ static inline struct page *encoded_page_ptr(struct encoded_page *page)
>  	return (struct page *)(~ENCODE_PAGE_BITS & (unsigned long)page);
>  }
>  
> +/*
> + * This macro defines the maximum tail pages (of a folio) that can have the
> + * page->mapping field reused (offset 12 for 32bits, or 24 for 64bits).
> + *
> + * When the tail page's mapping field reused, it'll be exempted from
> + * ->mapping poisoning and checks.  Also see the macro TAIL_MAPPING.
> + */
> +#define  TAIL_MAPPING_REUSED_MAX  (2)
> +
>  /**
>   * struct folio - Represents a contiguous set of bytes.
>   * @flags: Identical to the page flags.
> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
> index 0b709d2c46c6..72f244e16dcb 100644
> --- a/mm/huge_memory.c
> +++ b/mm/huge_memory.c
> @@ -2444,9 +2444,9 @@ static void __split_huge_page_tail(struct page *head, int tail,
>  			 (1L << PG_dirty) |
>  			 LRU_GEN_MASK | LRU_REFS_MASK));
>  
> -	/* ->mapping in first and second tail page is replaced by other uses */
> -	VM_BUG_ON_PAGE(tail > 2 && page_tail->mapping != TAIL_MAPPING,
> -			page_tail);
> +	/* ->mapping in <=TAIL_MAPPING_REUSED_MAX tail pages are reused */
> +	VM_BUG_ON_PAGE(tail > TAIL_MAPPING_REUSED_MAX &&
> +		       page_tail->mapping != TAIL_MAPPING, page_tail);
>  	page_tail->mapping = head->mapping;
>  	page_tail->index = head->index + tail;

I tend to also squash this in when repost:

diff --git a/mm/internal.h b/mm/internal.h
index 02a6fd36f46e..a75df0bd1f89 100644
--- a/mm/internal.h
+++ b/mm/internal.h
@@ -428,7 +428,8 @@ static inline void prep_compound_tail(struct page *head, int tail_idx)
 {
        struct page *p = head + tail_idx;

-       p->mapping = TAIL_MAPPING;
+       if (tail_idx > TAIL_MAPPING_REUSED_MAX)
+               p->mapping = TAIL_MAPPING;
        set_compound_head(p, head);
        set_page_private(p, 0);
 }

-- 
Peter Xu