All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Lorenzo Stoakes (ARM)" <ljs@kernel.org>
To: "Liam R. Howlett" <liam@infradead.org>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	 Vlastimil Babka <vbabka@kernel.org>,
	Jann Horn <jannh@google.com>, Pedro Falcato <pfalcato@suse.de>,
	 linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	stable@vger.kernel.org,  Li Jinyue <jinyue.li@linux.alibaba.com>
Subject: Re: [PATCH] mm/vma: update VMA iterator state on error in new VMA merge
Date: Mon, 16 Nov 2026 15:46:54 +0000	[thread overview]
Message-ID: <avskacTjLUyz4D_c@gremlin> (raw)
In-Reply-To: <gp4huj53rndmngvzajdfelalmh5gkhzqv73t2v6rotmr37hgpb@g4a33cixtylu>

On Thu, Nov 12, 2026 at 05:01:34PM -0500, Liam R. Howlett wrote:
> On 26/10/27 11:25AM, Lorenzo Stoakes (ARM) wrote:
>
> I'm a bit late on this one..
>
> > Commit cacded5e42b9 ("mm: avoid using vma_merge() for new VMAs") shared
> > code between vma_merge_new_range() and vma_expand().
> >
> > However in doing so, it failed to correctly restore the VMA iterator when
> > vma_expand() errors out.
> >
> > This is a problem when the immediately preceding VMA is mergeable and
> > vmg->just_expand isn't specified, at which point the VMA iterator is moved
> > to the previous VMA.
> >
> > (Note that vmg->just_expand is a special case for brk(), where the
> > iterator will already be positioned at the VMA to be expanded.)
> >
> > Resolve things by restoring the VMA iterator on error. Use
> > vma_iter_next_range() to do this, so it will correctly stop on the gap
> > after the previous VMA, rather than skip to the next VMA after it (this is
> > a new VMA merge, so is being applied at a gap).
> >
> > To isolate the error handling, return early when a merge isn't
> > possible (what the vmg->target check was doing before), then handle the
> > merge success/failure case explicitly.
> >
> > While here, update vma_expand() to jump to the nomem case when
> > dup_anon_vma() fails, so vmg->state is correctly updated.
> >
> > This doesn't impact anything in practice, but is the correct course and
> > since it's part of the error path that this patch addresses, it makes sense
> > to update it also.
> >
> > Users are very unlikely to have observed the issue this patch addresses in
> > practice, as it is near-impossible to fail an allocation for an anon_vma,
> > but a process under memory pressure that receives a fatal signal could
> > trigger it.
> >
> > This issue was originally reported by Li Jinyue for which I am grateful. He
> > indicated this was uncovered by a private syzkaller instance.
> >
> > Fixes: cacded5e42b9 ("mm: avoid using vma_merge() for new VMAs")
> > Cc: stable@vger.kernel.org
> > Reported-by: Li Jinyue <jinyue.li@linux.alibaba.com>
> > Closes: https://lore.kernel.org/all/20261025093921.1011927-1-jinyue.li@linux.alibaba.com/
> > Signed-off-by: Lorenzo Stoakes (ARM) <ljs@kernel.org>
> > ---
> >  mm/vma.c | 28 ++++++++++++++++++----------
> >  1 file changed, 18 insertions(+), 10 deletions(-)
> >
> > diff --git a/mm/vma.c b/mm/vma.c
> > index af8e65ada23b..d416ba58eb4c 100644
> > --- a/mm/vma.c
> > +++ b/mm/vma.c
> > @@ -1141,6 +1141,7 @@ struct vm_area_struct *vma_merge_new_range(struct vma_merge_struct *vmg)
> >  	struct vm_area_struct *next = vmg->next;
> >  	unsigned long end = vmg->end;
> >  	bool can_merge_left, can_merge_right;
> > +	bool move_iter_back = false;
>
> The name move_iter_back is a bit confusing because you use it to move it
> to the next entry.  I get what you mean, move it back to where it was, not
> go to prev, but the name is a bit off when used to go to the next entry.
>
> reposition_iter, maybe?

Ahh too late this is merged :P

I think it's hard to get this quite right.

So in the first instance move_iter_back = do a vma_prev() i.e. literally move
the iterator backwards.

Then, if the iterator was moved back, we want to restore it on error, I hoped
that the comment would make that clear.

So it's just separating noting that it needs doing (move_iter_back = true) and
the actual action (vma_prev()) + the restore.

So I think arguably OK as-is? But could obviously also follow up if needed.

>
> >
> >  	mmap_assert_write_locked(vmg->mm);
> >  	VM_WARN_ON_VMG(vmg->middle, vmg);
> > @@ -1182,23 +1183,30 @@ struct vm_area_struct *vma_merge_new_range(struct vma_merge_struct *vmg)
> >  			vmg->end = end;
> >
> >  		/* In expand-only case we are already positioned at prev. */
> > -		if (!vmg->just_expand) {
> > -			/* Equivalent to going to the previous range. */
> > -			vma_prev(vmg->vmi);
> > -		}
> > +		if (!vmg->just_expand)
> > +			move_iter_back = true;
> >  	}
> >
> > +	if (!can_merge_left && !can_merge_right)
> > +		return NULL;
> > +
> > +	if (move_iter_back)
> > +		vma_prev(vmg->vmi);
> > +
> >  	/*
> >  	 * Now try to expand adjacent VMA(s). This takes care of removing the
> >  	 * following VMA if we have VMAs on both sides.
> >  	 */
> > -	if (vmg->target && !vma_expand(vmg)) {
> > -		khugepaged_enter_vma(vmg->target, vmg->vm_flags);
> > -		vmg->state = VMA_MERGE_SUCCESS;
> > -		return vmg->target;
> > +	if (vma_expand(vmg)) {
> > +		/* Error occurred, restore iterator. */
> > +		if (move_iter_back)
> > +			vma_iter_next_range(vmg->vmi);
> > +		return NULL;
> >  	}
> >
> > -	return NULL;
> > +	khugepaged_enter_vma(vmg->target, vmg->vm_flags);
> > +	vmg->state = VMA_MERGE_SUCCESS;
> > +	return vmg->target;
> >  }
> >
> >  /*
> > @@ -1288,7 +1296,7 @@ int vma_expand(struct vma_merge_struct *vmg)
> >  	if (!ret && vmg->copied_from)
> >  		ret = dup_anon_vma(target, vmg->copied_from, &anon_dup);
> >  	if (ret)
> > -		return ret;
> > +		goto nomem;
> >
> >  	if (remove_next) {
> >  		vma_start_write(next);
> >
> > ---
> > base-commit: 1eb36316305767b54fac2cd40b88b3499791bdf2
> > change-id: 20261027-fix-new-merge-bug-71c2200d40fb
> >
> > Cheers,
> > --
> > Lorenzo Stoakes (ARM) <ljs@kernel.org>
> >

--
Cheers, Lorenzo

      reply	other threads:[~2026-11-16 15:46 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-27 11:25 [PATCH] mm/vma: update VMA iterator state on error in new VMA merge Lorenzo Stoakes (ARM)
2026-11-12 22:01 ` Liam R. Howlett
2026-11-16 15:46   ` Lorenzo Stoakes (ARM) [this message]

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=avskacTjLUyz4D_c@gremlin \
    --to=ljs@kernel.org \
    --cc=akpm@linux-foundation.org \
    --cc=jannh@google.com \
    --cc=jinyue.li@linux.alibaba.com \
    --cc=liam@infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=pfalcato@suse.de \
    --cc=stable@vger.kernel.org \
    --cc=vbabka@kernel.org \
    /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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.