linux-kernel - Re: [PATCH v6 07/18] of: dynamic: change type of of_{at,de}tach

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

Date:   Wed, 7 Nov 2018 06:55:14 -0800
From:   Frank Rowand <frowand.list@...il.com>
To:     Michael Ellerman <mpe@...erman.id.au>,
        Rob Herring <robh+dt@...nel.org>,
        Pantelis Antoniou <pantelis.antoniou@...sulko.com>,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        Paul Mackerras <paulus@...ba.org>,
        Alan Tull <atull@...nel.org>, Moritz Fischer <mdf@...nel.org>
Cc:     linux-kernel@...r.kernel.org, linuxppc-dev@...ts.ozlabs.org,
        devicetree@...r.kernel.org, linux-fpga@...r.kernel.org
Subject: Re: [PATCH v6 07/18] of: dynamic: change type of
 of_{at,de}tach_node() to void

On 11/7/18 4:08 AM, Michael Ellerman wrote:
> frowand.list@...il.com writes:
> 
>> From: Frank Rowand <frank.rowand@...y.com>
>>
>> of_attach_node() and of_detach_node() always return zero, so
>> their return value is meaningless.
> 
> But should they always return zero?
> 
> At least __of_attach_node_sysfs() can fail in several ways.

Sigh.  And of_reconfig_notify() can fail.  And at one point in the
history the return value of of_reconfig_notify() was returned by
of_attach_node() if of_reconfig_notify() failed.

> And there's also this in __of_detach_node() which should probably be
> returning an error:
> 
> 	if (WARN_ON(of_node_check_flag(np, OF_DETACHED)))
>         	return;
> 
> 
> Seems to me we should instead be fixing these to propagate errors,
> rather than hiding them?

The history of how of_attach_node() stopped propagating errors is
a bit more complex than I want to dig into at the moment.  So I'll
drop this patch from the series and add investigating this onto
my todo list.  I suspect that the result of investigating will be
that error return values should not be ignored in of_attach_node()
and of_detach_node(), but should instead be propagated to the
callers, as you suggest.

-Frank

> 
> cheers
>