Re: [PATCH 4.19 016/125] media: pci: ttpci: av7110: fix possible buffer overflow caused by bad DMA value in debiirq()

[Jia-Ju Bai]

On 2020/9/2 5:16, Pavel Machek wrote:
> On Tue 2020-09-01 18:35:23, Greg Kroah-Hartman wrote:
> > On Tue, Sep 01, 2020 at 05:25:12PM +0100, Sean Young wrote:
> > > Greg,
> > >
> > > On Tue, Sep 01, 2020 at 05:09:31PM +0200, Greg Kroah-Hartman wrote:
> > > > From: Jia-Ju Bai <baijiaju@xxxxxxxxxxxxxxx>
> > > >
> > > > [ Upstream commit 6499a0db9b0f1e903d52f8244eacc1d4be00eea2 ]
> > > >
> > > > The value av7110->debi_virt is stored in DMA memory, and it is assigned
> > > > to data, and thus data[0] can be modified at any time by malicious
> > > > hardware. In this case, "if (data[0] < 2)" can be passed, but then
> > > > data[0] can be changed into a large number, which may cause buffer
> > > > overflow when the code "av7110->ci_slot[data[0]]" is used.
> > > >
> > > > To fix this possible bug, data[0] is assigned to a local variable, which
> > > > replaces the use of data[0].
> > > See the discussion here:
> > >
> > > https://lkml.org/lkml/2020/8/31/479
> > >
> > > It does not seem worthwhile merging to the stable trees.
> > It doesn't hurt either :)
> Update stable kernel rules.
>
> If "patch does not match description and is pretty obviously useless"
> but "does not hurt" is acceptable for stable tree, people should know.
>
> You are pushing known junk into stable. Stop that.

Sorry for my useless patch...

Recently I submitted a new patch wiith READ_ONCE() to fix the problem 
that Pavel said:
https://lkml.org/lkml/2020/8/30/67

If you think this new patch is still useless, reverting the code is fine 
to me.


Best wishes,
Jia-Ju Bai