Re: [PATCH] thermal: mtk: Allocate enough space for mtk_thermal.
From: Zhang Rui
Date: Wed Jan 30 2019 - 08:38:48 EST
On ä, 2019-01-30 at 11:04 +0100, Daniel Lezcano wrote:
> On 30/01/2019 10:25, Pi-Hsun Shih wrote:
> >
> > On Wed, Jan 30, 2019 at 3:44 PM Daniel Lezcano
> > <daniel.lezcano@xxxxxxxxxx> wrote:
> > >
> > >
> > > On 30/01/2019 07:04, Peter Shih wrote:
> > > >
> > > > Adding Michael Kao to cc list.
> > > >
> > > > On Wed, Jan 9, 2019 at 1:57 PM Pi-Hsun Shih <pihsun@xxxxxxxxxxx
> > > > g> wrote:
> > > > >
> > > > >
> > > > > The mtk_thermal struct contains a 'struct mtk_thermal_bank
> > > > > banks[];',
> > > > > but the allocation only allocates sizeof(struct mtk_thermal)
> > > > > bytes,
> > > > > which cause out of bound access with the ->banks[] member.
> > > > > Change it to
> > > > > a fixed size array instead.
> > > Even if the fix is correct, it pushes back the bug later in time
> > > if a
> > > new board containing more than MAX_NUM_ZONES is introduced. I
> > > suggest to
> > > dynamically allocate the array with the 'num_banks' value.
> > >
> > For the current code structure, those mtk_thermal_data are
> > statically declared,
> > so if there's new board containing more than MAX_NUM_ZONES of
> > bank_data, it
> > would actually be a compile error.
> >
> > I'm fine with either way, but feel like that this is simpler than
> > manually
> > calculating the size needed for allocation.
> Right, I missed it can be caught at compile time.
>
> Reviewed-by: Daniel Lezcano <daniel.lezcano@xxxxxxxxxx>
>
As this is a bugfix, I will take it and queue it for next -rc.
thanks,
rui
>
>