From: Dongliang Mu <mudongliangabcd@gmail.com>

There is a slab-out-of-bounds Write bug in hid-bigbenff driver.
The problem is the driver assumes the device must have an input but
some malicious devices violate this assumption.

Fix this by checking hid_device's input is non-empty before its usage.

Reported-by: syzkaller <syzkaller@googlegroups.com>
Signed-off-by: Dongliang Mu <mudongliangabcd@gmail.com>
---
 drivers/hid/hid-bigbenff.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/hid/hid-bigbenff.c b/drivers/hid/hid-bigbenff.c
index 74ad8bf98bfd..c14d1774101d 100644
--- a/drivers/hid/hid-bigbenff.c
+++ b/drivers/hid/hid-bigbenff.c
@@ -347,6 +347,11 @@ static int bigben_probe(struct hid_device *hid,
 	bigben->report = list_entry(report_list->next,
 		struct hid_report, list);
 
+	if (list_empty(&hid->inputs)) {
+		hid_err(hid, "no inputs found\n");
+		return -ENODEV;
+	}
+
 	hidinput = list_first_entry(&hid->inputs, struct hid_input, list);
 	set_bit(FF_RUMBLE, hidinput->input->ffbit);
 
-- 
2.25.1

On Fri, 6 May 2022, Dongliang Mu wrote:

> From: Dongliang Mu <mudongliangabcd@gmail.com>
> 
> There is a slab-out-of-bounds Write bug in hid-bigbenff driver.
> The problem is the driver assumes the device must have an input but
> some malicious devices violate this assumption.
> 
> Fix this by checking hid_device's input is non-empty before its usage.
> 
> Reported-by: syzkaller <syzkaller@googlegroups.com>
> Signed-off-by: Dongliang Mu <mudongliangabcd@gmail.com>
> ---
>  drivers/hid/hid-bigbenff.c | 5 +++++
>  1 file changed, 5 insertions(+)
> 
> diff --git a/drivers/hid/hid-bigbenff.c b/drivers/hid/hid-bigbenff.c
> index 74ad8bf98bfd..c14d1774101d 100644
> --- a/drivers/hid/hid-bigbenff.c
> +++ b/drivers/hid/hid-bigbenff.c
> @@ -347,6 +347,11 @@ static int bigben_probe(struct hid_device *hid,
>  	bigben->report = list_entry(report_list->next,
>  		struct hid_report, list);
>  
> +	if (list_empty(&hid->inputs)) {
> +		hid_err(hid, "no inputs found\n");
> +		return -ENODEV;
> +	}
> +

Thanks for the fix. It doesn't seemt o be fully correct though -- as you'd 
be returning -ENODEV here in the situation when hid_hw_start() has already 
happened. So I believe better thing to do here is to do error = -ENODEV; 
goto error_hw_stop;

Could you please fix that up and resend? Thanks,

-- 
Jiri Kosina
SUSE Labs

On Fri, May 6, 2022 at 2:53 PM Jiri Kosina <jikos@kernel.org> wrote:
>
> On Fri, 6 May 2022, Dongliang Mu wrote:
>
> > From: Dongliang Mu <mudongliangabcd@gmail.com>
> >
> > There is a slab-out-of-bounds Write bug in hid-bigbenff driver.
> > The problem is the driver assumes the device must have an input but
> > some malicious devices violate this assumption.
> >
> > Fix this by checking hid_device's input is non-empty before its usage.
> >
> > Reported-by: syzkaller <syzkaller@googlegroups.com>
> > Signed-off-by: Dongliang Mu <mudongliangabcd@gmail.com>
> > ---
> >  drivers/hid/hid-bigbenff.c | 5 +++++
> >  1 file changed, 5 insertions(+)
> >
> > diff --git a/drivers/hid/hid-bigbenff.c b/drivers/hid/hid-bigbenff.c
> > index 74ad8bf98bfd..c14d1774101d 100644
> > --- a/drivers/hid/hid-bigbenff.c
> > +++ b/drivers/hid/hid-bigbenff.c
> > @@ -347,6 +347,11 @@ static int bigben_probe(struct hid_device *hid,
> >       bigben->report = list_entry(report_list->next,
> >               struct hid_report, list);
> >
> > +     if (list_empty(&hid->inputs)) {
> > +             hid_err(hid, "no inputs found\n");
> > +             return -ENODEV;
> > +     }
> > +
>
> Thanks for the fix. It doesn't seemt o be fully correct though -- as you'd
> be returning -ENODEV here in the situation when hid_hw_start() has already
> happened. So I believe better thing to do here is to do error = -ENODEV;
> goto error_hw_stop;
>
> Could you please fix that up and resend? Thanks,

Oh, yes. I will send a v2 version ASAP.
>
> --
> Jiri Kosina
> SUSE Labs
>