linux-kernel - Re: [PATCH] HSI: core: fix resource leaks in hsi_add_client_from

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20210409173425.GB7405@nvidia.com>
Date:   Fri, 9 Apr 2021 14:34:25 -0300
From:   Jason Gunthorpe <jgg@...dia.com>
To:     Dan Carpenter <dan.carpenter@...cle.com>
Cc:     Sebastian Reichel <sre@...nel.org>, Pavel Machek <pavel@....cz>,
        linux-kernel@...r.kernel.org, kernel-janitors@...r.kernel.org
Subject: Re: [PATCH] HSI: core: fix resource leaks in hsi_add_client_from_dt()

On Fri, Apr 09, 2021 at 02:08:17PM +0300, Dan Carpenter wrote:
> If some of the allocations fail between the dev_set_name() and the
> device_register() then the name will not be freed.  Fix this by
> moving dev_set_name() directly in front of the call to device_register().
> 
> Fixes: a2aa24734d9d ("HSI: Add common DT binding for HSI client devices")
> Signed-off-by: Dan Carpenter <dan.carpenter@...cle.com>
> ---
> Jason, this is the most common type of error I see with device_register().
> Is there a downside to calling dev_set_name() later?  Presumably it's
> printed out somewhere, but I feel like just moving the dev_set_name() is
> almost always the best and simplest fix.

It is hard to tell without detailed analysis.. ie a dev_err()/etc call
will use the name. It is why I don't like this design pattern of
avoiding device_initialize() and using device_registrR()

This movement looks OK though

Reviewed-by: Jason Gunthorpe <jgg@...dia.com>

Jason