mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Suyuan Wang via B4 Relay <devnull+shyckless.gmail.com@kernel.org>
To: tony.luck@intel.com, bp@alien8.de, shubhrajyoti.datta@amd.com
Cc: linux-edac@vger.kernel.org, linux-kernel@vger.kernel.org,
	 radhey.shyam.pandey@amd.com
Subject: [PATCH] EDAC/versalnet: Fix device-name use-after-free on removal
Date: Wed, 18 Nov 2026 13:30:32 +0000	[thread overview]
Message-ID: <20261118-edac-versalnet-v1-1-cee5b2d3cb25@gmail.com> (raw)

From: Suyuan Wang <stevewong@hust.edu.cn>

mc_init() stores a pointer to the backing device's name in mci->dev_name
without copying the string:

	mci->dev_name = dev_name(dev);

remove_one_mc() then unregisters that device before calling
edac_mc_del_mc():

	device_unregister(mci->pdev);
	edac_mc_del_mc(mci->pdev);

device_unregister() can drop the last reference, and kobject_cleanup()
frees the device's name along with the device. edac_mc_del_mc() then
logs the controller removal through edac_dev_name(mci), i.e.
mci->dev_name, reading the freed name string.

Call edac_mc_del_mc() before device_unregister(), so the device and its
name stay alive until EDAC has logged and removed the controller.

Found with a static analysis tool and checked by hand.

Fixes: d5fe2fec6c40 ("EDAC: Add a driver for the AMD Versal NET DDR controller")
Link: https://lore.kernel.org/linux-edac/20260724171945.2812749-9-shubhrajyoti.datta@amd.com/
Cc: stable@vger.kernel.org
Co-developed-by: Kaixuan Li <kaixuan.li@ntu.edu.sg>
Signed-off-by: Kaixuan Li <kaixuan.li@ntu.edu.sg>
Signed-off-by: Suyuan Wang <stevewong@hust.edu.cn>
---
We found this independently and later found the earlier patch linked
above, which fixes the same bug as part of a larger series; this is a
standalone two-line version. The bad ordering predates the refactor in
62a9fc50e8d9 -- it was already in the original driver, in both the normal
remove path and the init-failure rollback path.

The change only reorders two cleanup calls. Checked by inspection; I
couldn't build it here (it needs an arm64 Zynq config and my arm64 cross
toolchain fails at an unrelated vdso step), and I don't have the hardware.
---
 drivers/edac/versalnet_edac.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/edac/versalnet_edac.c b/drivers/edac/versalnet_edac.c
index d59ba3904cda..075b0105c427 100644
--- a/drivers/edac/versalnet_edac.c
+++ b/drivers/edac/versalnet_edac.c
@@ -767,8 +767,8 @@ static void remove_one_mc(struct mc_priv *priv, int i)
 	struct mem_ctl_info *mci;
 
 	mci = priv->mci[i];
-	device_unregister(mci->pdev);
 	edac_mc_del_mc(mci->pdev);
+	device_unregister(mci->pdev);
 	edac_mc_free(mci);
 }
 





                 reply	other threads:[~2026-11-18 13:30 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=20261118-edac-versalnet-v1-1-cee5b2d3cb25@gmail.com \
    --to=devnull+shyckless.gmail.com@kernel.org \
    --cc=bp@alien8.de \
    --cc=linux-edac@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=radhey.shyam.pandey@amd.com \
    --cc=shubhrajyoti.datta@amd.com \
    --cc=shyckless@gmail.com \
    --cc=tony.luck@intel.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®