Linux-mm Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "David Hildenbrand (Arm)" <david@kernel.org>
To: Hui Zhu <hui.zhu@linux.dev>,
	Andrew Morton <akpm@linux-foundation.org>,
	Lorenzo Stoakes <ljs@kernel.org>,
	"Liam R. Howlett" <liam@infradead.org>,
	Vlastimil Babka <vbabka@kernel.org>,
	Mike Rapoport <rppt@kernel.org>,
	Suren Baghdasaryan <surenb@google.com>,
	Michal Hocko <mhocko@suse.com>,
	linux-mm@kvack.org, linux-kernel@vger.kernel.org
Cc: Hui Zhu <zhuhui@kylinos.cn>
Subject: Re: [PATCH] mm: avoid KCSAN false positive in page_to_nid()
Date: Tue, 23 Jun 2026 09:59:45 +0200	[thread overview]
Message-ID: <97f2c4ad-c302-424d-a7c8-fdfa0fc0d811@kernel.org> (raw)
In-Reply-To: <20260623074157.578113-1-hui.zhu@linux.dev>

On 6/23/26 09:41, Hui Zhu wrote:
> From: Hui Zhu <zhuhui@kylinos.cn>
> 
> KCSAN reports a data race between page_to_nid() reading page->flags
> and folio_trylock()/folio_lock() doing test_and_set_bit_lock(PG_locked,
> ...) on the same word from another CPU, e.g.:
> 
>   BUG: KCSAN: data-race in __lruvec_stat_mod_folio / shmem_get_folio_gfp
> 
> The node id occupies a fixed, high bit-range of page->flags that is
> set once when the page is initialized and never modified afterwards,
> so it can never overlap with the low PG_locked/PG_waiters bits touched
> by the folio lock path. The race is therefore harmless: page_to_nid()
> always returns a consistent value regardless of how the read
> interleaves with the lock bit ops.
> 
> Wrap the flags read with data_race() to tell KCSAN this race is
> intentional and benign, consistent with how page->page_type is
> already annotated for similar packed-field accesses.
> 
> Signed-off-by: Hui Zhu <zhuhui@kylinos.cn>
> ---
>  include/linux/mm.h | 9 ++++++++-
>  1 file changed, 8 insertions(+), 1 deletion(-)
> 
> diff --git a/include/linux/mm.h b/include/linux/mm.h
> index 485df9c2dbdd..122d3b39369f 100644
> --- a/include/linux/mm.h
> +++ b/include/linux/mm.h
> @@ -2296,7 +2296,14 @@ static inline int memdesc_nid(memdesc_flags_t mdf)
>  
>  static inline int page_to_nid(const struct page *page)
>  {
> -	return memdesc_nid(PF_POISONED_CHECK(page)->flags);
> +	/*
> +	 * The node id occupies a fixed high bit-range of page->flags
> +	 * that is set once at page init and never changed afterwards.
> +	 * It cannot overlap with the low PG_locked/PG_waiters bits
> +	 * that folio_lock()/folio_unlock() concurrently update, so
> +	 * this data race is benign.
> +	 */

Do we really need this excessive comment?

> +	return memdesc_nid(data_race(PF_POISONED_CHECK(page)->flags));

In memdesc_zonenum() we use ASSERT_EXCLUSIVE_BITS.

Can we do the same here inside memdesc_nid?

diff --git a/include/linux/mm.h b/include/linux/mm.h
index 69daeeab7fe8f..76d3bb54be844 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -2290,6 +2290,7 @@ int memdesc_nid(memdesc_flags_t mdf);
 #else
 static inline int memdesc_nid(memdesc_flags_t mdf)
 {
+       ASSERT_EXCLUSIVE_BITS(mdf.f, NODES_MASK << NODES_PGSHIFT);
        return (mdf.f >> NODES_PGSHIFT) & NODES_MASK;
 }
 #endif


-- 
Cheers,

David


  reply	other threads:[~2026-06-23  7:59 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-06-23  7:41 [PATCH] mm: avoid KCSAN false positive in page_to_nid() Hui Zhu
2026-06-23  7:59 ` David Hildenbrand (Arm) [this message]
2026-06-23 10:25   ` Lorenzo Stoakes

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=97f2c4ad-c302-424d-a7c8-fdfa0fc0d811@kernel.org \
    --to=david@kernel.org \
    --cc=akpm@linux-foundation.org \
    --cc=hui.zhu@linux.dev \
    --cc=liam@infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=ljs@kernel.org \
    --cc=mhocko@suse.com \
    --cc=rppt@kernel.org \
    --cc=surenb@google.com \
    --cc=vbabka@kernel.org \
    --cc=zhuhui@kylinos.cn \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox