[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <Y4nqrcw3LGfj3OZR@sirena.org.uk>
Date: Fri, 2 Dec 2022 12:08:13 +0000
From: Mark Brown <broonie@...nel.org>
To: ChiYuan Huang <u0084500@...il.com>
Cc: lgirdwood@...il.com, lee@...nel.org, matthias.bgg@...il.com,
yangyingliang@...wei.com, chiaen_wu@...htek.com,
linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
linux-mediatek@...ts.infradead.org,
ChiYuan Huang <cy_huang@...htek.com>
Subject: Re: [PATCH 1/2] regulator: mt6370: Fix potential UAF issue
On Fri, Dec 02, 2022 at 11:35:35AM +0800, ChiYuan Huang wrote:
> Mark Brown <broonie@...nel.org> 於 2022年12月1日 週四 晚上7:43寫道:
> > > The original code uses i2c dev as the parent in order to reuse
> > > the 'regulator_of_get_init_data'. But this will cause regulation
> > > constraint devres attached to i2c dev, not the mfd cell platform
> > > device.
> > This is a general issue which will affect a lot of MFDs, we would be
> > better to fix this by changing the API to provide a device to be used
> > for the devres allocations separately to the one used for looking up the
> > DT.
> Not to affect too much, the better way may change the 'regulator_register' API.
> Append it as regulator_register(dev, .....
> This could separate device object with devres allocation and DT lookup.
Yes, I think so - a new optional argument.
Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)
Powered by blists - more mailing lists