From: Vastargazing <vebohr@gmail.com>
To: vebohr@gmail.com
Cc: stable@vger.kernel.org
Subject: [PATCH 3/5] mtd: maps: physmap: fix reference leak on failed device registration
Date: Mon, 4 May 2026 13:08:18 +0300 [thread overview]
Message-ID: <6717e0b2a6244ee4e691dba03eb8c790c202e89e.1777889235.git.vebohr@gmail.com> (raw)
In-Reply-To: <cover.1777889235.git.vebohr@gmail.com>
When platform_device_register() fails in physmap_init(), the embedded
struct device has already been initialized by device_initialize() inside
platform_device_register(). The error path unregisters the driver but
returns without dropping the device reference:
physmap_init()
-> platform_device_register(&physmap_flash)
-> device_initialize(&physmap_flash.dev) /* kref = 1 */
-> platform_device_add(&physmap_flash) /* fails */
<- platform_driver_unregister() called, but kref still 1
Per platform_device_register() kernel-doc:
NOTE: _Never_ directly free @pdev after calling this function, even if
it returned an error! Always use platform_device_put() to give up the
reference initialised in this function instead.
Fix this by calling platform_device_put() before unregistering the driver.
Fixes: 73566edf9b91 ("[MTD] Convert physmap to platform driver")
Cc: stable@vger.kernel.org
Assisted-by: GitHub Copilot (Claude Sonnet 4.5)
Signed-off-by: Vastargazing <vebohr@gmail.com>
---
drivers/mtd/maps/physmap-core.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/mtd/maps/physmap-core.c b/drivers/mtd/maps/physmap-core.c
index dcda7685fc99..45d79ca622c1 100644
--- a/drivers/mtd/maps/physmap-core.c
+++ b/drivers/mtd/maps/physmap-core.c
@@ -654,8 +654,10 @@ static int __init physmap_init(void)
#ifdef CONFIG_MTD_PHYSMAP_COMPAT
if (err == 0) {
err = platform_device_register(&physmap_flash);
- if (err)
+ if (err) {
+ platform_device_put(&physmap_flash);
platform_driver_unregister(&physmap_flash_driver);
+ }
}
#endif
--
2.51.0
WARNING: multiple messages have this Message-ID (diff)
From: Vastargazing <vebohr@gmail.com>
To: linux-kernel@vger.kernel.org
Cc: Vastargazing <vebohr@gmail.com>,
stable@vger.kernel.org, Miquel Raynal <miquel.raynal@bootlin.com>,
Richard Weinberger <richard@nod.at>,
Vignesh Raghavendra <vigneshr@ti.com>,
David Woodhouse <dwmw2@infradead.org>,
Lennert Buytenhek <buytenh@wantstofly.org>,
linux-mtd@lists.infradead.org
Subject: [PATCH 3/5] mtd: maps: physmap: fix reference leak on failed device registration
Date: Mon, 4 May 2026 13:08:45 +0300 [thread overview]
Message-ID: <6717e0b2a6244ee4e691dba03eb8c790c202e89e.1777889235.git.vebohr@gmail.com> (raw)
Message-ID: <20260504100845.1iFxGQcZbBvg-BiM15JnM20wHG_PaaJlEFMHg_f2M0A@z> (raw)
In-Reply-To: <cover.1777889235.git.vebohr@gmail.com>
When platform_device_register() fails in physmap_init(), the embedded
struct device has already been initialized by device_initialize() inside
platform_device_register(). The error path unregisters the driver but
returns without dropping the device reference:
physmap_init()
-> platform_device_register(&physmap_flash)
-> device_initialize(&physmap_flash.dev) /* kref = 1 */
-> platform_device_add(&physmap_flash) /* fails */
<- platform_driver_unregister() called, but kref still 1
Per platform_device_register() kernel-doc:
NOTE: _Never_ directly free @pdev after calling this function, even if
it returned an error! Always use platform_device_put() to give up the
reference initialised in this function instead.
Fix this by calling platform_device_put() before unregistering the driver.
Fixes: 73566edf9b91 ("[MTD] Convert physmap to platform driver")
Cc: stable@vger.kernel.org
Assisted-by: GitHub Copilot (Claude Sonnet 4.5)
Signed-off-by: Vastargazing <vebohr@gmail.com>
---
drivers/mtd/maps/physmap-core.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/mtd/maps/physmap-core.c b/drivers/mtd/maps/physmap-core.c
index dcda7685fc99..45d79ca622c1 100644
--- a/drivers/mtd/maps/physmap-core.c
+++ b/drivers/mtd/maps/physmap-core.c
@@ -654,8 +654,10 @@ static int __init physmap_init(void)
#ifdef CONFIG_MTD_PHYSMAP_COMPAT
if (err == 0) {
err = platform_device_register(&physmap_flash);
- if (err)
+ if (err) {
+ platform_device_put(&physmap_flash);
platform_driver_unregister(&physmap_flash_driver);
+ }
}
#endif
--
2.51.0
next prev parent reply other threads:[~2026-05-04 10:08 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <cover.1777889235.git.vebohr@gmail.com>
2026-05-04 10:08 ` [PATCH 1/5] misc: eeprom: digsy_mtc: fix reference leak on failed device registration Vastargazing
2026-05-04 10:08 ` Vastargazing
2026-05-04 10:08 ` [PATCH 2/5] platform/chrome: cros_ec_lpc: " Vastargazing
2026-05-04 10:08 ` Vastargazing
2026-05-05 2:40 ` Tzung-Bi Shih
2026-05-04 10:08 ` Vastargazing [this message]
2026-05-04 10:08 ` [PATCH 3/5] mtd: maps: physmap: " Vastargazing
2026-05-04 13:58 ` Miquel Raynal
2026-05-04 10:08 ` [PATCH 4/5] perf: arm: pmu: " Vastargazing
2026-05-04 10:08 ` Vastargazing
2026-05-05 8:50 ` Sudeep Holla
2026-05-04 10:08 ` [PATCH 5/5] mfd: sm501: " Vastargazing
2026-05-04 10:08 ` Vastargazing
2026-05-04 12:48 ` [PATCH v2] " Valery Borovsky
2026-05-04 13:52 ` Miquel Raynal
2026-05-05 15:00 ` Lee Jones
2026-05-05 15:12 ` Miquel Raynal
2026-05-06 15:40 ` [PATCH v3] " Valery Borovsky
2026-05-06 15:43 ` [PATCH 3/5] mtd: maps: physmap-core: " Valery Borovsky
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=6717e0b2a6244ee4e691dba03eb8c790c202e89e.1777889235.git.vebohr@gmail.com \
--to=vebohr@gmail.com \
--cc=stable@vger.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox