Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Thu, 27 Jul 2023 17:20:13 +0200
From:   Miquel Raynal <miquel.raynal@bootlin.com>
To:     Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc:     Alexander Usyskin <alexander.usyskin@intel.com>,
        Richard Weinberger <richard@nod.at>,
        Vignesh Raghavendra <vigneshr@ti.com>,
        linux-mtd@lists.infradead.org, linux-kernel@vger.kernel.org,
        Tomas Winkler <tomas.winkler@intel.com>,
        Vitaly Lubart <vitaly.lubart@intel.com>,
        Zhang Xiaoxu <zhangxiaoxu5@huawei.com>
Subject: Re: [PATCH] mtd: fix use-after-free in mtd release
Message-ID: <20230727172013.7c85c05d@xps-13>
In-Reply-To: <ZMKJRNDoQV8p0DH4@smile.fi.intel.com>
References: <20230727145758.3880967-1-alexander.usyskin@intel.com>
        <ZMKJRNDoQV8p0DH4@smile.fi.intel.com>
Organization: Bootlin
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

Hi Andy,

andriy.shevchenko@linux.intel.com wrote on Thu, 27 Jul 2023 18:12:04
+0300:

> On Thu, Jul 27, 2023 at 05:57:58PM +0300, Alexander Usyskin wrote:
> > I case of partition device_unregister in mtd_device_release =20
>=20
> In
>=20
> device_unregister()
> mtd_device_release()
>=20
> > calls mtd_release which frees mtd_info structure for partition. =20
>=20
> mtd_release()
>=20
> > All code after device_unregister in mtd_device_release thus =20
>=20
> device_unregister()
> mtd_device_release()
>=20
> > works already freed memory. =20
>=20
> uses?
>=20
> > Move part of code to mtd_release and restict mtd->dev cleanup =20
>=20
> mtd_release()

Yup, thanks for all these suggestions, I agree with them.

> > to non-partion object.
> > For partition object such cleanup have no sense as partition
> > mtd_info is removed.
> >=20
> > Cc: Miquel Raynal <miquel.raynal@bootlin.com>
> > Cc: Zhang Xiaoxu <zhangxiaoxu5@huawei.com>
> > Fixes: 19bfa9ebebb5 ("mtd: use refcount to prevent corruption") =20
>=20
> Closes: ?

Did I miss a recent update on the use of Fixes? I thought Closes was
supposed to point at a bug report while Fixes would point to the faulty
commit. Right now I feel like Fixes is the right tag, but if you have a
source explaining why we should not longer do it like I am used to,
I would appreciate a link.

Thanks,
Miqu=C3=A8l