linux-kernel - Re: [PATCH AUTOSEL 5.0 107/262] ASoC: qcom: Fix of-node refcount unbalance in apq8016_sbc_parse

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, 27 Mar 2019 19:31:15 +0000
From:   Mark Brown <broonie@...nel.org>
To:     Sasha Levin <sashal@...nel.org>
Cc:     linux-kernel@...r.kernel.org, stable@...r.kernel.org,
        Takashi Iwai <tiwai@...e.de>,
        Patrick Lai <plai@...eaurora.org>,
        Banajit Goswami <bgoswami@...eaurora.org>
Subject: Re: [PATCH AUTOSEL 5.0 107/262] ASoC: qcom: Fix of-node refcount
 unbalance in apq8016_sbc_parse_of()

On Wed, Mar 27, 2019 at 01:59:22PM -0400, Sasha Levin wrote:
> From: Takashi Iwai <tiwai@...e.de>
> 
> [ Upstream commit 8d1667200850f8753c0265fa4bd25c9a6e5f94ce ]
> 
> The apq8016 driver leaves the of-node refcount at aborting from the
> loop of for_each_child_of_node() in the error path.  Not only the
> iterator node of for_each_child_of_node(), the children nodes referred
> from it for codec and cpu have to be properly unreferenced.

Given that the device tree is entirely static on all platforms using FDT
I'm not convinced that it is sensible to backport any of the fixes for
references not being dropped that are more than one liners,
realistically nothing is ever going to notice the extra references but
there are things that will notice if we underflow references so there's
some risk we could break things somehow.  Seems like it's all risk for
zero benefit.

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)