Linux virtualization list
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: "Andrew Morton" <akpm@linux-foundation.org>,
	"David Hildenbrand" <david@kernel.org>,
	"Vlastimil Babka" <vbabka@kernel.org>,
	"Brendan Jackman" <jackmanb@google.com>,
	"Michal Hocko" <mhocko@suse.com>,
	"Suren Baghdasaryan" <surenb@google.com>,
	"Jason Wang" <jasowang@redhat.com>,
	"Andrea Arcangeli" <aarcange@redhat.com>,
	"Dev Jain" <dev.jain@arm.com>,
	"Geert Uytterhoeven" <geert@linux-m68k.org>,
	"Greg Ungerer" <gerg@linux-m68k.org>,
	"Gregory Price" <gourry@gourry.net>,
	"Huang, Ying" <ying.huang@linux.alibaba.com>,
	"Lance Yang" <lance.yang@linux.dev>,
	"Lorenzo Stoakes" <ljs@kernel.org>,
	"Magnus Lindholm" <linmag7@gmail.com>,
	"Matthew Wilcox" <willy@infradead.org>,
	"Miaohe Lin" <linmiaohe@huawei.com>, "Zi Yan" <ziy@nvidia.com>,
	linux-mm@kvack.org, virtualization@lists.linux.dev,
	"Jason Wang" <jasowangio@gmail.com>,
	"Eugenio Pérez" <eperezma@redhat.com>,
	"Xuan Zhuo" <xuanzhuo@linux.alibaba.com>
Subject: Re: [PATCH v11 6/7] virtio_balloon: add DEVICE_INIT_REPORTED feature
Date: Fri, 20 Nov 2026 12:03:29 +0000	[thread overview]
Message-ID: <sashiko-outbox-225115@kernel.org> (raw)
In-Reply-To: <65c518234ce31fac3f6ed436455dcb415d4b1189.1795089715.git.mst@redhat.com>

Thank you for your contribution! Sashiko AI review found 3 potential
issues to consider:
- [High] virtio_balloon: 32-bit used length overflow in
  virtballoon_free_page_report()
- [High] virtio_balloon: stale PG_zeroed flag on free pages across
  hibernation resume
- [Medium] virtio_balloon: missing feature dependency checks in
  virtballoon_validate()
--

commit 45401c714e001e18d0bb9e4b0f46a6691e93f0cb
Author: Michael S. Tsirkin <mst@redhat.com>

virtio_balloon: add DEVICE_INIT_REPORTED feature

This commit adds support for the VIRTIO_BALLOON_F_DEVICE_INIT_REPORTED
feature, allowing page reporting to mark reported pages as zeroed when the
host initializes them on reclaim.

> diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c
> index 8c1b47ec17a29..4c6630c5f9599 100644
> --- a/drivers/virtio/virtio_balloon.c
> +++ b/drivers/virtio/virtio_balloon.c
[ ... ]
> @@ -205,7 +206,9 @@ static int virtballoon_free_page_report(struct page_reporting_dev_info *pr_dev_i
>  	struct virtio_balloon *vb =
>  		container_of(pr_dev_info, struct virtio_balloon, pr_dev_info);
>  	struct virtqueue *vq = vb->reporting_vq;
> -	unsigned int unused, err;
> +	struct scatterlist *entry;
> +	unsigned int len, i;
> +	int err;
>  
>  	/* We should always be able to add these buffers to an empty queue. */
>  	err = virtqueue_add_inbuf(vq, sg, nents, vb, GFP_NOWAIT);
> @@ -221,9 +224,24 @@ static int virtballoon_free_page_report(struct page_reporting_dev_info *pr_dev_i
>  	virtqueue_kick(vq);
>  
>  	/* When host has read buffer, this completes via balloon_ack */
> -	wait_event(vb->acked, virtqueue_get_buf(vq, &unused));
> +	wait_event(vb->acked, virtqueue_get_buf(vq, &len));
>  
> -	return 0;
> +	if (!pr_dev_info->host_zeroes_pages)
> +		return 0;
> +
> +	/*
> +	 * Only a fully initialized batch may be marked reported when reported
> +	 * pages are treated as zeroed. A short completion does not identify
> +	 * which pages were initialized, so reject the entire batch.
> +	 * Subtract each entry to avoid overflowing a total-length accumulator.
> +	 */
> +	for_each_sg(sg, entry, nents, i) {
> +		if (len < entry->length)
> +			return -EIO;
> +		len -= entry->length;
> +	}
> +
> +	return len ? -EIO : 0;
>  }

[Severity: High]
Can virtqueue_get_buf() ever return a len large enough for a batch of 4 GiB
or more?

The Virtio used ring stores a 32-bit __virtio32 len field, so len is at most
U32_MAX (4 GiB - 1).  On CONFIG_ARM64_64K_PAGES, MAX_PAGE_ORDER is 13
(512 MiB per block) and capacity is up to PAGE_REPORTING_CAPACITY (32), so a
single scatterlist batch of 8 or more order-13 blocks totals 4 GiB to 16 GiB.
For any batch of 4 GiB or more, subtracting entry->length from the 32-bit len
always hits len < entry->length and returns -EIO, causing
page_reporting_cycle() to abort without marking the pages reported and retry
the same blocks on every reporting pass.

> @@ -1043,6 +1061,9 @@ static int virtballoon_probe(struct virtio_device *vdev)
>  #if defined(CONFIG_ARM64) && defined(CONFIG_ARM64_64K_PAGES)
>  		vb->pr_dev_info.order = 5;
>  #endif
> +		if (virtio_has_feature(vb->vdev,
> +				      VIRTIO_BALLOON_F_DEVICE_INIT_REPORTED))
> +			vb->pr_dev_info.host_zeroes_pages = true;

[Severity: High]
Does enabling host_zeroes_pages leave stale PG_zeroed flags on free buddy
pages across hibernation resume when want_init_on_free() and
page_poisoning_enabled() are false?

During hibernation snapshot creation, free buddy page frames are excluded
from the image, while the vmemmap pages containing their struct page entries
(with PG_zeroed set by page_reporting_drain()) are saved.  During resume, the
boot kernel uses those free physical page frames as temporary safe pages for
struct pbe chains and image buffers, and swsusp_arch_resume() restores the
saved struct page entries with PG_zeroed still set.  Afterward,
clear_or_poison_free_pages() only zeroes free pages when
page_poisoning_enabled() or want_init_on_free() is true:

kernel/power/snapshot.c:clear_or_poison_free_pages() {
    ...
	if (page_poisoning_enabled() || want_init_on_free()) {
		memory_bm_position_reset(bm);
		pfn = memory_bm_next_pfn(bm);
		while (pfn != BM_END_OF_MAP) {
			if (pfn_valid(pfn))
				clear_or_poison_free_page(pfn_to_page(pfn));
    ...
}

In the default configuration where both are disabled, post_alloc_hook() and
vma_alloc_zeroed_movable_folio() see PG_zeroed set and skip zeroing, handing
dirty resume-kernel memory to __GFP_ZERO allocations and userspace page
faults.

> @@ -1189,6 +1210,11 @@ static int virtballoon_validate(struct virtio_device *vdev)
>  	else if (!virtio_has_feature(vdev, VIRTIO_BALLOON_F_PAGE_POISON))
>  		__virtio_clear_bit(vdev, VIRTIO_BALLOON_F_REPORTING);
>  
> +	if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT) ||
> +	    page_poisoning_enabled_static())
> +		__virtio_clear_bit(vdev,
> +				   VIRTIO_BALLOON_F_DEVICE_INIT_REPORTED);

[Severity: Medium]
Should virtballoon_validate() clear VIRTIO_BALLOON_F_DEVICE_INIT_REPORTED
when VIRTIO_BALLOON_F_REPORTING is not set, and keep VIRTIO_BALLOON_F_REPORTING
enabled when want_init_on_free() is satisfied by
VIRTIO_BALLOON_F_DEVICE_INIT_REPORTED?

If VIRTIO_BALLOON_F_REPORTING is absent or cleared above,
VIRTIO_BALLOON_F_DEVICE_INIT_REPORTED remains set in vdev->features_array,
which can cause virtio_features_ok() to fail probe with -ENODEV on a device
that requires VIRTIO_BALLOON_F_REPORTING as a prerequisite.  In addition,
when want_init_on_free() is true and page_poisoning_enabled_static() is
false, the check above clears VIRTIO_BALLOON_F_REPORTING if the device does
not offer VIRTIO_BALLOON_F_PAGE_POISON, even when
VIRTIO_BALLOON_F_DEVICE_INIT_REPORTED guarantees zeroed pages.

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1795089715.git.mst@redhat.com?part=6

  reply	other threads:[~2026-11-20 12:03 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-11-19 12:02 [PATCH v11 0/7] mm/virtio: skip redundant zeroing of host-zeroed reported pages Michael S. Tsirkin
2026-11-19 12:02 ` [PATCH v11 1/7] mm: add PG_zeroed page flag and buddy merge tracking Michael S. Tsirkin
2026-11-20 12:03   ` sashiko-bot
2026-11-19 12:02 ` [PATCH v11 2/7] mm: page_reporting: support host-zeroed reported pages Michael S. Tsirkin
2026-11-20 12:03   ` sashiko-bot
2026-11-19 12:02 ` [PATCH v11 3/7] mm: skip zeroing in vma_alloc_anon_folio_pmd for pre-zeroed pages Michael S. Tsirkin
2026-11-20 12:03   ` sashiko-bot
2026-11-19 12:02 ` [PATCH v11 4/7] mm: skip zeroing of pre-zeroed order-0 user pages Michael S. Tsirkin
2026-11-20 12:03   ` sashiko-bot
2026-11-19 12:02 ` [PATCH v11 5/7] mm: skip zeroing in alloc_anon_folio for pre-zeroed pages Michael S. Tsirkin
2026-11-20 12:03   ` sashiko-bot
2026-11-19 12:02 ` [PATCH v11 6/7] virtio_balloon: add DEVICE_INIT_REPORTED feature Michael S. Tsirkin
2026-11-20 12:03   ` sashiko-bot [this message]
2026-11-19 12:03 ` [PATCH v11 7/7] mm: page_reporting: add flush parameter with page budget Michael S. Tsirkin
2026-11-20 12:03   ` sashiko-bot

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=sashiko-outbox-225115@kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=aarcange@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=david@kernel.org \
    --cc=dev.jain@arm.com \
    --cc=eperezma@redhat.com \
    --cc=geert@linux-m68k.org \
    --cc=gerg@linux-m68k.org \
    --cc=gourry@gourry.net \
    --cc=jackmanb@google.com \
    --cc=jasowang@redhat.com \
    --cc=jasowangio@gmail.com \
    --cc=lance.yang@linux.dev \
    --cc=linmag7@gmail.com \
    --cc=linmiaohe@huawei.com \
    --cc=linux-mm@kvack.org \
    --cc=ljs@kernel.org \
    --cc=mhocko@suse.com \
    --cc=mst@redhat.com \
    --cc=sashiko-reviews@lists.linux.dev \
    --cc=surenb@google.com \
    --cc=vbabka@kernel.org \
    --cc=virtualization@lists.linux.dev \
    --cc=willy@infradead.org \
    --cc=xuanzhuo@linux.alibaba.com \
    --cc=ying.huang@linux.alibaba.com \
    --cc=ziy@nvidia.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