linux-kernel - Re: [PATCH 4/4] extcon: extcon-max8997: Fix IRQ freeing at error path

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]

Date:   Fri, 14 May 2021 08:12:34 +0000
From:   "Vaittinen, Matti" <Matti.Vaittinen@...rohmeurope.com>
To:     "krzysztof.kozlowski@...onical.com" 
        <krzysztof.kozlowski@...onical.com>
CC:     "hdegoede@...hat.com" <hdegoede@...hat.com>,
        "m.szyprowski@...sung.com" <m.szyprowski@...sung.com>,
        "cw00.choi@...sung.com" <cw00.choi@...sung.com>,
        "myungjoo.ham@...sung.com" <myungjoo.ham@...sung.com>,
        "b.zolnierkie@...sung.com" <b.zolnierkie@...sung.com>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 4/4] extcon: extcon-max8997: Fix IRQ freeing at error path


On Tue, 2021-05-11 at 12:27 -0400, Krzysztof Kozlowski wrote:
> On 10/05/2021 23:32, Matti Vaittinen wrote:
> > Hi Krzysztof,
> > 
> > On Mon, 2021-05-10 at 10:21 -0400, Krzysztof Kozlowski wrote:
> > > On 10/05/2021 04:12, Matti Vaittinen wrote:
> > > > If reading MAX8997_MUIC_REG_STATUS1 fails at probe the driver
> > > > exits
> > > > without freeing the requested IRQs.
> > > 
> > > The driver frees IRQ on probe failure, so maybe you meant missing
> > > IRQ
> > > mapping dispose?
> > 
> > No. The commit 3e34c8198960 ("extcon: max8997: Avoid forcing UART
> > path
> > on drive probe") introduced a return w/o IRQ freeing if reading the
> > MAX8997_MUIC_REG_STATUS1 fails at the end of the probe. This is not
> > visible in the patch though - as the return is Ok after the IRQs
> > and
> > work-queue cancellation are managed by devm.
> 
> I see it now, right. The fix is big and changes too much to be
> backportable. I would prefer to simply fix the problem with "goto
> err_irq" and convert to devm in next patch.

Agree. Backporting devm WQs to stable just to fix this seems like an
overkill. I'll respin the series.

Best Regards
--Matti