From: sashiko-bot@kernel.org
To: "Bartosz Golaszewski" <bartosz.golaszewski@oss.qualcomm.com>
Cc: Frank.Li@kernel.org, devicetree@vger.kernel.org, robh@kernel.org,
conor+dt@kernel.org, dri-devel@lists.freedesktop.org,
imx@lists.linux.dev
Subject: Re: [PATCH v2 19/19] driver core: platform: count references to all kinds of firmware nodes
Date: Tue, 30 Jun 2026 09:18:42 +0000 [thread overview]
Message-ID: <20260630091842.BCE2F1F00A3D@smtp.kernel.org> (raw)
In-Reply-To: <20260629-pdev-fwnode-ref-v2-19-8abe2513f96e@oss.qualcomm.com>
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] Memory leak of `of_node` references in platform device release
- [High] Refcount underflow and use-after-free of software nodes on platform device release
--
commit f2aa250a52f31d368fee3251e58597913170101a
Author: Bartosz Golaszewski <bartosz.golaszewski@oss.qualcomm.com>
driver core: platform: count references to all kinds of firmware nodes
This patch switches to counting references for all kinds of firmware nodes by
replacing the of_node_put with fwnode_handle_put in platform device release.
It also updates platform device setup helpers to properly acquire fwnode
references.
> diff --git a/drivers/base/platform.c b/drivers/base/platform.c
> index f24a5f406746b..bb5f5bddd047d 100644
> --- a/drivers/base/platform.c
> +++ b/drivers/base/platform.c
> @@ -599,7 +599,7 @@ static void platform_device_release(struct device *dev)
> struct platform_object *pa = container_of(dev, struct platform_object,
> pdev.dev);
>
> - of_node_put(pa->pdev.dev.of_node);
> + fwnode_handle_put(pa->pdev.dev.fwnode);
[Severity: High]
Does replacing of_node_put() here cause a memory leak of of_node references?
Many drivers and core APIs manually acquire an of_node reference and assign
it directly to pdev->dev.of_node without setting pdev->dev.fwnode. For
example, in drivers/base/core.c:device_set_of_node_from_dev():
of_node_put(dev->of_node);
dev->of_node = of_node_get(dev2->of_node);
When a platform device set up this way is released, fwnode_handle_put() will
operate on a NULL or unrelated fwnode, failing to drop the of_node reference.
[Severity: High]
Can this fwnode_handle_put() cause a refcount underflow for software nodes?
Platform drivers often inherit their parent's fwnode without taking an
additional reference. For example, in drivers/gpio/gpio-arizona.c:
device_set_node(&pdev->dev, dev_fwnode(pdev->dev.parent));
Before this patch, calling of_node_put(dev->of_node) on a software node was
a safe no-op (as of_node was NULL). Unconditionally calling
fwnode_handle_put() here will now invoke software_node_put() and decrement
the unacquired reference, potentially leading to a use-after-free when the
device is unregistered.
> kfree(pa->pdev.dev.platform_data);
> kfree(pa->pdev.mfd_cell);
> kfree(pa->pdev.resource);
[ ... ]
--
Sashiko AI review · https://sashiko.dev/#/patchset/20260629-pdev-fwnode-ref-v2-0-8abe2513f96e@oss.qualcomm.com?part=19
next prev parent reply other threads:[~2026-06-30 9:18 UTC|newest]
Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-29 9:12 [PATCH v2 00/19] driver core: count references of the platform device's fwnode, not OF node Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 01/19] powerpc/powermac: fix OF node refcount Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 02/19] driver core: platform: provide platform_device_set_of_node() Bartosz Golaszewski
2026-06-30 9:18 ` sashiko-bot
2026-06-30 11:37 ` Manuel Ebner
2026-06-30 13:22 ` Bartosz Golaszewski
2026-07-01 15:05 ` Manuel Ebner
2026-06-29 9:12 ` [PATCH v2 03/19] driver core: platform: provide platform_device_set_fwnode() Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 04/19] driver core: platform: provide platform_device_set_of_node_from_dev() Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 05/19] of: platform: use platform_device_set_of_node() Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 06/19] powerpc/powermac: " Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 07/19] i2c: pxa-pci: " Bartosz Golaszewski
2026-06-30 9:18 ` sashiko-bot
2026-06-29 9:12 ` [PATCH v2 08/19] iommu/fsl: " Bartosz Golaszewski
2026-06-29 14:08 ` Frank Li
2026-06-29 9:12 ` [PATCH v2 09/19] net: bcmgenet: " Bartosz Golaszewski
2026-06-29 23:15 ` Jakub Kicinski
2026-06-30 9:18 ` sashiko-bot
2026-06-29 9:12 ` [PATCH v2 10/19] pmdomain: imx: " Bartosz Golaszewski
2026-06-30 9:18 ` sashiko-bot
2026-06-30 15:27 ` Frank Li
2026-06-29 9:12 ` [PATCH v2 11/19] mfd: tps6586: " Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 12/19] slimbus: qcom-ngd-ctrl: " Bartosz Golaszewski
2026-06-29 9:25 ` Konrad Dybcio
2026-06-29 9:12 ` [PATCH v2 13/19] net: mv643xx: " Bartosz Golaszewski
2026-06-29 23:16 ` Jakub Kicinski
2026-06-29 9:12 ` [PATCH v2 14/19] drm/xe/i2c: use platform_device_set_fwnode() Bartosz Golaszewski
2026-06-30 9:18 ` sashiko-bot
2026-06-29 9:12 ` [PATCH v2 15/19] platform/surface: gpe: " Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 16/19] usb: chipidea: use platform_device_set_of_node_from_dev() Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 17/19] usb: musb: " Bartosz Golaszewski
2026-06-30 9:18 ` sashiko-bot
2026-06-29 9:12 ` [PATCH v2 18/19] reset: rzg2l: " Bartosz Golaszewski
2026-06-29 9:12 ` [PATCH v2 19/19] driver core: platform: count references to all kinds of firmware nodes Bartosz Golaszewski
2026-06-29 16:21 ` Andy Shevchenko
2026-06-30 9:18 ` sashiko-bot [this message]
2026-06-29 16:23 ` [PATCH v2 00/19] driver core: count references of the platform device's fwnode, not OF node Andy Shevchenko
2026-06-29 23:20 ` ✓ CI.KUnit: success for driver core: count references of the platform device's fwnode, not OF node (rev2) Patchwork
2026-06-30 0:24 ` ✓ Xe.CI.BAT: " Patchwork
2026-06-30 12:47 ` ✓ Xe.CI.FULL: " Patchwork
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=20260630091842.BCE2F1F00A3D@smtp.kernel.org \
--to=sashiko-bot@kernel.org \
--cc=Frank.Li@kernel.org \
--cc=bartosz.golaszewski@oss.qualcomm.com \
--cc=conor+dt@kernel.org \
--cc=devicetree@vger.kernel.org \
--cc=dri-devel@lists.freedesktop.org \
--cc=imx@lists.linux.dev \
--cc=robh@kernel.org \
--cc=sashiko-reviews@lists.linux.dev \
/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.