linux-kernel - Re: [PATCH] scsi: ufs: ufshcd-pltfrm: check the return value of kstrdup()

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <tencent_29819C31A7E71A71105124840B2855FD970A@qq.com>
Date:   Mon, 13 Dec 2021 09:56:49 +0800
From:   "Xiaoke Wang" <xkernel.wang@...mail.com>
To:     "Bart Van Assche" <bvanassche@....org>,
        "jejb" <jejb@...ux.ibm.com>,
        "martin.petersen" <martin.petersen@...cle.com>
Cc:     "linux-scsi" <linux-scsi@...r.kernel.org>,
        "linux-kernel" <linux-kernel@...r.kernel.org>,
        "alim.akhtar" <alim.akhtar@...sung.com>,
        "avri.altman" <avri.altman@....com>
Subject: Re: [PATCH] scsi: ufs: ufshcd-pltfrm: check the return value of kstrdup()

On 12/12/21 22:32, Bart Van Assche wrote:
> Is xkernel the name of a person or the name of a robot? Patches should 
> be signed by a person even if these have been generated by a robot.
>
> > +		clki->name = kstrdup(name, GFP_KERNEL);
> > +		if (!clki->name) {
> > +			ret = -ENOMEM;
> > +			devm_kfree(dev, clki);
> > +			goto out;
> > +		}
>
> Is the devm_kfree() call necessary? Is it useful?

Note: the last mail seems fail because of the html content.

Sorry about that, xkernel is my English name I named myself, my full
name in Chinese format is "Xiaoke Wang". Thank you for your
suggestion, I will consider using my official name in the future.

clki is allocated on above of kstrdup(), line 86-90, and if clki is NULL
will return -ENOMEM which is the same return value with the patch. So
I think clki should be freed to prevent potential memory leak:
>		clki = devm_kzalloc(dev, sizeof(*clki), GFP_KERNEL);
>		if (!clki) {
>			ret = -ENOMEM;
>			goto out;
>		}

vreg is in similar case.