linux-kernel - Incorrect usage of kobject_init_and

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: <CAAwk1nJOxG-2tCfTCUSgVMP2SXeNnWheT_4K6+WUMCxYVph7nA@mail.gmail.com>
Date:   Wed, 9 Jan 2019 15:14:10 -0800
From:   Machiry Aravind Kumar <makrvcs@...il.com>
To:     linux-kernel@...r.kernel.org
Subject: Incorrect usage of kobject_init_and_add?

Hi all,

I see some discrepancy in how the function: kobject_init_and_add is
used in various places of the kernel.

According to the docs, when kobject_init_and_add fails, the caller is
supposed to call kobject_put on the corresponding kobj.

We found several instances where this is indeed true:
https://elixir.bootlin.com/linux/latest/source/drivers/block/pktcdvd.c#L131

However, there are again several places where this is not the case.
For instance: https://github.com/torvalds/linux/blob/master/kernel/sched/cpufreq_schedutil.c#L754

Why is this discrepancy? Should these be fixed to add kobject_put in
their error path?

-Best,
Aravind