linux-kernel - Re: [PATCH] of/platform: Avoid compilation warning

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

Message-ID: <20200528130342.219f8a44@xps13>
Date:   Thu, 28 May 2020 13:03:42 +0200
From:   Miquel Raynal <miquel.raynal@...tlin.com>
To:     Rob Herring <robh@...nel.org>
Cc:     Mark Rutland <mark.rutland@....com>, devicetree@...r.kernel.org,
        Thomas Petazzoni <thomas.petazzoni@...tlin.com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] of/platform: Avoid compilation warning

Hi Rob,

Rob Herring <robh@...nel.org> wrote on Wed, 27 May 2020 12:35:55 -0600:

> On Thu, May 14, 2020 at 07:07:07PM +0200, Miquel Raynal wrote:
> > The of_find_device_by_node() helper has its dummy counterpart for when
> > CONFIG_OF is not enabled. However, it is clearly stated in the kernel
> > documentation that it "takes a reference to the embedded struct device
> > which needs to be dropped after use". Which means the of_dev_put()
> > helper might have to be called afterwards. Unfortunately, there is no
> > of_dev_put() dummy function if OF_CONFIG is not enabled which seems
> > odd in this case. The of_dev_put() helper is pretty simple, it just
> > checks the validity of the single argument and calls put_device() on
> > it. One can just call put_device() directly to avoid any build issue
> > but I find much more accurate in this case to create the dummy
> > helper.
> > 
> > With this helper, a file using of_find_device_by_node() can also call
> > of_dev_put() without triggering the following:  
> 
> IMO, you should use platform_device_put() instead. It has the NULL check 
> too.
> 
> I imagine of_dev_put() is left over from when OF devices were not 
> platform devices. 

Ok, makes sense. Perhaps we should entirely get rid of it, I don't see
a lot of users left. Or at least update the comment I was mentioning.

Cheers,
Miquèl