From: Kiryl Shutsemau <kas@kernel.org>
To: "Lorenzo Stoakes (ARM)" <ljs@kernel.org>
Cc: Andrew Morton <akpm@linux-foundation.org>,
"Liam R. Howlett" <liam@infradead.org>,
Vlastimil Babka <vbabka@kernel.org>,
Jann Horn <jannh@google.com>, Pedro Falcato <pfalcato@suse.de>,
"David Hildenbrand (Arm)" <david@kernel.org>,
Lance Yang <lance.yang@linux.dev>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
syzbot+10e72ea5747737bcd2ad@syzkaller.appspotmail.com,
Songsen Yang <yangsongsen@insta360.com>
Subject: Re: [PATCH] mm/vma: read anon_vma carefully in validate_mm()
Date: Mon, 2 Nov 2026 10:48:45 +0000 [thread overview]
Message-ID: <auhn01dpDzK47igy@thinkstation> (raw)
In-Reply-To: <20261102-fix-validate_mm-v1-1-5c309614ef15@kernel.org>
On Mon, Nov 02, 2026 at 10:25:20AM +0000, Lorenzo Stoakes (ARM) wrote:
> Commit 38564710f443 ("mm: implement and use vma_has_anon_rmap(), silence
> KCSAN") rather clumsily converted the read of vma->anon_vma into a
> vma_has_anon_rmap() without accounting for the separation between
> assignment of anon_vma and later use of it.
>
> vma_has_anon_rmap() is intended as an optimistic check, so doesn't need to
> protect against a torn read in vma->anon_vma. The caller is expected to
> take care of this themselves if they then in fact access the field.
>
> validate_mm() does indeed then reference the field, so resolve the issue
> via READ_ONCE(), and simplify the code by separating this into a new
> function validate_mm_anon_rmap().
>
> Finally, add a vma_assert_stabilised() to make it clear the required state
> of the VMA.
>
> Fixes: 38564710f443 ("mm: implement and use vma_has_anon_rmap(), silence KCSAN")
> Reported-by: syzbot+10e72ea5747737bcd2ad@syzkaller.appspotmail.com
> Closes: https://lore.kernel.org/all/6ae78418.6552420f.2a0efc.005c.GAE@google.com/
> Reported-by: Songsen Yang <yangsongsen@insta360.com>
> Closes: https://lore.kernel.org/all/20261102074414.10902-1-yangsongsen@insta360.com/
> Signed-off-by: Lorenzo Stoakes (ARM) <ljs@kernel.org>
> ---
> mm/vma.c | 36 ++++++++++++++++++++++++------------
> 1 file changed, 24 insertions(+), 12 deletions(-)
>
> diff --git a/mm/vma.c b/mm/vma.c
> index d416ba58eb4c..9f02f19a7845 100644
> --- a/mm/vma.c
> +++ b/mm/vma.c
> @@ -669,6 +669,28 @@ static int dup_anon_vma(struct vm_area_struct *dst,
> }
>
> #ifdef CONFIG_DEBUG_VM_MAPLE_TREE
> +#ifdef CONFIG_DEBUG_VM_RB
> +static void validate_mm_anon_rmap(struct vm_area_struct *vma)
> +{
> + struct anon_vma *anon_vma = READ_ONCE(vma->anon_vma);
> + struct anon_vma_chain *avc;
> +
> + vma_assert_stabilised(vma);
> +
> + if (!anon_vma)
> + return;
> +
> + anon_vma_lock_read(anon_vma);
> + list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
> + anon_rmap_tree_verify(avc);
> + anon_vma_unlock_read(anon_vma);
> +}
> +#else
> +static void validate_mm_anon_rmap(struct vm_area_struct *vma)
> +{
> +}
> +#endif
> +
> void validate_mm(struct mm_struct *mm)
> {
> int bug = 0;
This ifdeffery seems unnecessary. What about this:
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 9f167950926a..2e43077dcd51 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -4363,9 +4363,7 @@ anon_rmap_tree_iter_first(struct anon_vma *anon_vma,
struct anon_vma_chain *
anon_rmap_tree_iter_next(struct anon_vma_chain *avc,
pgoff_t pgoff_start, pgoff_t pgoff_last);
-#ifdef CONFIG_DEBUG_VM_RB
void anon_rmap_tree_verify(struct anon_vma_chain *avc);
-#endif
#define anon_rmap_tree_foreach(avc, anon_vma, pgoff_start, pgoff_last) \
for (avc = anon_rmap_tree_iter_first(anon_vma, pgoff_start, pgoff_last); \
diff --git a/mm/vma.c b/mm/vma.c
index 1be613044b4a..30f31d3b3e26 100644
--- a/mm/vma.c
+++ b/mm/vma.c
@@ -669,6 +669,26 @@ static int dup_anon_vma(struct vm_area_struct *dst,
}
#ifdef CONFIG_DEBUG_VM_MAPLE_TREE
+static void validate_mm_anon_rmap(struct vm_area_struct *vma)
+{
+ struct anon_vma *anon_vma;
+ struct anon_vma_chain *avc;
+
+ if (!IS_ENABLED(CONFIG_DEBUG_VM_RB))
+ return;
+
+ vma_assert_stabilised(vma);
+
+ anon_vma = READ_ONCE(vma->anon_vma);
+ if (!anon_vma)
+ return;
+
+ anon_vma_lock_read(anon_vma);
+ list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
+ anon_rmap_tree_verify(avc);
+ anon_vma_unlock_read(anon_vma);
+}
+
void validate_mm(struct mm_struct *mm)
{
int bug = 0;
@@ -678,10 +698,6 @@ void validate_mm(struct mm_struct *mm)
mt_validate(&mm->mm_mt);
for_each_vma(vmi, vma) {
-#ifdef CONFIG_DEBUG_VM_RB
- struct anon_vma *anon_vma = vma->anon_vma;
- struct anon_vma_chain *avc;
-#endif
unsigned long vmi_start, vmi_end;
bool warn = 0;
@@ -702,14 +718,8 @@ void validate_mm(struct mm_struct *mm)
vma_iter_dump_tree(&vmi);
}
-#ifdef CONFIG_DEBUG_VM_RB
- if (vma_has_anon_rmap(vma)) {
- anon_vma_lock_read(anon_vma);
- list_for_each_entry(avc, &vma->anon_vma_chain, same_vma)
- anon_rmap_tree_verify(avc);
- anon_vma_unlock_read(anon_vma);
- }
-#endif
+ validate_mm_anon_rmap(vma);
+
/* Check for a infinite loop */
if (++i > mm->map_count + 10) {
i = -1;
--
Kiryl Shutsemau / Kirill A. Shutemov
next prev parent reply other threads:[~2026-11-02 10:48 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-02 10:25 Lorenzo Stoakes (ARM)
2026-11-02 10:36 ` Pedro Falcato
2026-11-02 10:48 ` Kiryl Shutsemau [this message]
2026-11-02 11:02 ` Lorenzo Stoakes (ARM)
2026-11-02 21:24 ` Andrew Morton
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=auhn01dpDzK47igy@thinkstation \
--to=kas@kernel.org \
--cc=akpm@linux-foundation.org \
--cc=david@kernel.org \
--cc=jannh@google.com \
--cc=lance.yang@linux.dev \
--cc=liam@infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=ljs@kernel.org \
--cc=pfalcato@suse.de \
--cc=syzbot+10e72ea5747737bcd2ad@syzkaller.appspotmail.com \
--cc=vbabka@kernel.org \
--cc=yangsongsen@insta360.com \
/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
all inboxes | Powered by JetHome®