syzbot found a problem using of uinit pointer in
lg4ff_set_autocenter_default().
Reported-by: [email protected]
Tested by syzbot:
https://groups.google.com/d/msg/syzkaller-bugs/ApnMLW6sfKE/Qq0bIHGEAQAJ
Signed-off-by: Phong Tran <[email protected]>
---
drivers/hid/hid-lg4ff.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/hid/hid-lg4ff.c b/drivers/hid/hid-lg4ff.c
index 5e6a0cef2a06..44dfd08b0c32 100644
--- a/drivers/hid/hid-lg4ff.c
+++ b/drivers/hid/hid-lg4ff.c
@@ -468,10 +468,10 @@ static int lg4ff_play(struct input_dev *dev, void *data, struct ff_effect *effec
static void lg4ff_set_autocenter_default(struct input_dev *dev, u16 magnitude)
{
struct hid_device *hid = input_get_drvdata(dev);
- s32 *value;
+ s32 *value = NULL;
u32 expand_a, expand_b;
- struct lg4ff_device_entry *entry;
- struct lg_drv_data *drv_data;
+ struct lg4ff_device_entry *entry = NULL;
+ struct lg_drv_data *drv_data = NULL;
unsigned long flags;
drv_data = hid_get_drvdata(hid);
--
2.20.1
On Tue, Nov 5, 2019 at 3:18 PM Phong Tran <[email protected]> wrote:
>
> syzbot found a problem using of uinit pointer in
> lg4ff_set_autocenter_default().
>
> Reported-by: [email protected]
>
> Tested by syzbot:
>
> https://groups.google.com/d/msg/syzkaller-bugs/ApnMLW6sfKE/Qq0bIHGEAQAJ
This seems weird to me:
the syzbot link above is about `hid_get_drvdata(hid)`, and, as I read
it, the possibility that hid might not have an initialized value.
Here you are changing the initialized values of value, entry and
drv_data, all 3 are never used before their first assignment.
I have a feeling this particular syzbot check has already been fixed
upstream by d9d4b1e46d95 "HID: Fix assumption that devices have
inputs".
Cheers,
Benjamin
>
> Signed-off-by: Phong Tran <[email protected]>
> ---
> drivers/hid/hid-lg4ff.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/hid/hid-lg4ff.c b/drivers/hid/hid-lg4ff.c
> index 5e6a0cef2a06..44dfd08b0c32 100644
> --- a/drivers/hid/hid-lg4ff.c
> +++ b/drivers/hid/hid-lg4ff.c
> @@ -468,10 +468,10 @@ static int lg4ff_play(struct input_dev *dev, void *data, struct ff_effect *effec
> static void lg4ff_set_autocenter_default(struct input_dev *dev, u16 magnitude)
> {
> struct hid_device *hid = input_get_drvdata(dev);
> - s32 *value;
> + s32 *value = NULL;
> u32 expand_a, expand_b;
> - struct lg4ff_device_entry *entry;
> - struct lg_drv_data *drv_data;
> + struct lg4ff_device_entry *entry = NULL;
> + struct lg_drv_data *drv_data = NULL;
> unsigned long flags;
>
> drv_data = hid_get_drvdata(hid);
> --
> 2.20.1
>
On 11/18/19 4:43 PM, Benjamin Tissoires wrote:
> On Tue, Nov 5, 2019 at 3:18 PM Phong Tran <[email protected]> wrote:
>>
>> syzbot found a problem using of uinit pointer in
>> lg4ff_set_autocenter_default().
>>
>> Reported-by: [email protected]
>>
>> Tested by syzbot:
>>
>> https://groups.google.com/d/msg/syzkaller-bugs/ApnMLW6sfKE/Qq0bIHGEAQAJ
>
> This seems weird to me:
>
> the syzbot link above is about `hid_get_drvdata(hid)`, and, as I read
> it, the possibility that hid might not have an initialized value.
>
In the dashboard [1] shows
BUG: KMSAN: uninit-value in dev_get_drvdata include/linux/device.h:1388
[inline]
BUG: KMSAN: uninit-value in hid_get_drvdata include/linux/hid.h:628 [inline]
BUG: KMSAN: uninit-value in lg4ff_set_autocenter_default+0x23a/0xa20
drivers/hid/hid-lg4ff.c:477
base on that I did the initialization the pointer in the patch.
> Here you are changing the initialized values of value, entry and
> drv_data, all 3 are never used before their first assignment.
>
> I have a feeling this particular syzbot check has already been fixed
> upstream by d9d4b1e46d95 "HID: Fix assumption that devices have
> inputs".
>
I think the commit d9d4b1 fixed this report [2] by syzbot.
[1] https://syzkaller.appspot.com/bug?extid=1234691fec1b8ceba8b1
[2] https://syzkaller.appspot.com/bug?extid=403741a091bf41d4ae79
regards,
Phong.
> Cheers,
> Benjamin
>
>>
>> Signed-off-by: Phong Tran <[email protected]>
>> ---
>> drivers/hid/hid-lg4ff.c | 6 +++---
>> 1 file changed, 3 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/hid/hid-lg4ff.c b/drivers/hid/hid-lg4ff.c
>> index 5e6a0cef2a06..44dfd08b0c32 100644
>> --- a/drivers/hid/hid-lg4ff.c
>> +++ b/drivers/hid/hid-lg4ff.c
>> @@ -468,10 +468,10 @@ static int lg4ff_play(struct input_dev *dev, void *data, struct ff_effect *effec
>> static void lg4ff_set_autocenter_default(struct input_dev *dev, u16 magnitude)
>> {
>> struct hid_device *hid = input_get_drvdata(dev);
>> - s32 *value;
>> + s32 *value = NULL;
>> u32 expand_a, expand_b;
>> - struct lg4ff_device_entry *entry;
>> - struct lg_drv_data *drv_data;
>> + struct lg4ff_device_entry *entry = NULL;
>> + struct lg_drv_data *drv_data = NULL;
>> unsigned long flags;
>>
>> drv_data = hid_get_drvdata(hid);
>> --
>> 2.20.1
>>
>
On Tue, Nov 19, 2019 at 2:30 PM Phong Tran <[email protected]> wrote:
>
> On 11/18/19 4:43 PM, Benjamin Tissoires wrote:
> > On Tue, Nov 5, 2019 at 3:18 PM Phong Tran <[email protected]> wrote:
> >>
> >> syzbot found a problem using of uinit pointer in
> >> lg4ff_set_autocenter_default().
> >>
> >> Reported-by: [email protected]
> >>
> >> Tested by syzbot:
> >>
> >> https://groups.google.com/d/msg/syzkaller-bugs/ApnMLW6sfKE/Qq0bIHGEAQAJ
> >
> > This seems weird to me:
> >
> > the syzbot link above is about `hid_get_drvdata(hid)`, and, as I read
> > it, the possibility that hid might not have an initialized value.
> >
>
> In the dashboard [1] shows
> BUG: KMSAN: uninit-value in dev_get_drvdata include/linux/device.h:1388
> [inline]
> BUG: KMSAN: uninit-value in hid_get_drvdata include/linux/hid.h:628 [inline]
> BUG: KMSAN: uninit-value in lg4ff_set_autocenter_default+0x23a/0xa20
> drivers/hid/hid-lg4ff.c:477
> base on that I did the initialization the pointer in the patch.
>
> > Here you are changing the initialized values of value, entry and
> > drv_data, all 3 are never used before their first assignment.
> >
> > I have a feeling this particular syzbot check has already been fixed
> > upstream by d9d4b1e46d95 "HID: Fix assumption that devices have
> > inputs".
> >
>
> I think the commit d9d4b1 fixed this report [2] by syzbot.
>
> [1] https://syzkaller.appspot.com/bug?extid=1234691fec1b8ceba8b1
> [2] https://syzkaller.appspot.com/bug?extid=403741a091bf41d4ae79
As you can see in my long discussion with syzbot today, d9d4b1 also
fixed that one.
https://groups.google.com/forum/#!msg/syzkaller-bugs/ApnMLW6sfKE/Qq0bIHGEAQAJ
Cheers,
Benjamin
>
> regards,
> Phong.
> > Cheers,
> > Benjamin
> >
> >>
> >> Signed-off-by: Phong Tran <[email protected]>
> >> ---
> >> drivers/hid/hid-lg4ff.c | 6 +++---
> >> 1 file changed, 3 insertions(+), 3 deletions(-)
> >>
> >> diff --git a/drivers/hid/hid-lg4ff.c b/drivers/hid/hid-lg4ff.c
> >> index 5e6a0cef2a06..44dfd08b0c32 100644
> >> --- a/drivers/hid/hid-lg4ff.c
> >> +++ b/drivers/hid/hid-lg4ff.c
> >> @@ -468,10 +468,10 @@ static int lg4ff_play(struct input_dev *dev, void *data, struct ff_effect *effec
> >> static void lg4ff_set_autocenter_default(struct input_dev *dev, u16 magnitude)
> >> {
> >> struct hid_device *hid = input_get_drvdata(dev);
> >> - s32 *value;
> >> + s32 *value = NULL;
> >> u32 expand_a, expand_b;
> >> - struct lg4ff_device_entry *entry;
> >> - struct lg_drv_data *drv_data;
> >> + struct lg4ff_device_entry *entry = NULL;
> >> + struct lg_drv_data *drv_data = NULL;
> >> unsigned long flags;
> >>
> >> drv_data = hid_get_drvdata(hid);
> >> --
> >> 2.20.1
> >>
> >
>