From: leixiang <leixiang@kylinos.cn>
Cc: leixiang <leixiang@kylinos.cn>,
Bjorn Helgaas <bhelgaas@google.com>,
Logan Gunthorpe <logang@deltatee.com>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
Chaitanya Kulkarni <kch@nvidia.com>, Jens Axboe <axboe@kernel.dk>,
linux-pci@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: [PATCH] PCI/P2PDMA: Fix use-after-free in pci_p2pdma_add_resource() error path
Date: Fri, 17 Jul 2026 17:51:17 +0800 [thread overview]
Message-ID: <20260717095120.51111-1-leixiang@kylinos.cn> (raw)
In pci_p2pdma_add_resource(), a devm action is registered to call
pci_p2pdma_unmap_mappings() with the 'p2p_pgmap' context. If the subsequent
call to gen_pool_add_owner() fails, the code jumps to the 'pages_free'
error path.
In this error path, 'p2p_pgmap' is explicitly freed via devm_kfree().
However, the previously registered devm action is never removed. This
leaves a dangling pointer in the device's devres list. When the device
is unbound and devres is cleaning up resources, the unmap action will be
called with the already-freed 'p2p_pgmap', resulting in a use-after-free
bug.
Fix this by properly calling devm_remove_action() in the error path before
freeing 'p2p_pgmap'.
Fixes: 7e9c7ef83d78 ("PCI/P2PDMA: Allow userspace VMA allocations through sysfs")
Assisted-by: Antigravity:Gemini [tools]
Signed-off-by: leixiang <leixiang@kylinos.cn>
---
drivers/pci/p2pdma.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/pci/p2pdma.c b/drivers/pci/p2pdma.c
index b2d5266f8653..7f54532932ab 100644
--- a/drivers/pci/p2pdma.c
+++ b/drivers/pci/p2pdma.c
@@ -451,13 +451,15 @@ int pci_p2pdma_add_resource(struct pci_dev *pdev, int bar, size_t size,
range_len(&pgmap->range), dev_to_node(&pdev->dev),
&pgmap->ref);
if (error)
- goto pages_free;
+ goto unmap_mappings;
pci_info(pdev, "added peer-to-peer DMA memory %#llx-%#llx\n",
pgmap->range.start, pgmap->range.end);
return 0;
+unmap_mappings:
+ devm_remove_action(&pdev->dev, pci_p2pdma_unmap_mappings, p2p_pgmap);
pages_free:
devm_memunmap_pages(&pdev->dev, pgmap);
pgmap_free:
--
2.43.0
next reply other threads:[~2026-07-17 9:51 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-07-17 9:51 leixiang [this message]
2026-07-17 10:48 ` Greg Kroah-Hartman
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=20260717095120.51111-1-leixiang@kylinos.cn \
--to=leixiang@kylinos.cn \
--cc=axboe@kernel.dk \
--cc=bhelgaas@google.com \
--cc=gregkh@linuxfoundation.org \
--cc=kch@nvidia.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-pci@vger.kernel.org \
--cc=logang@deltatee.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