Potential NULL pointer deference in spi

Potential NULL pointer deference in spi

[Yizhuo Zhai]

Hi All:

drivers/spi/spi.c:

The function to_spi_device() could return NULL, but some callers
in this file does not check the return value while directly dereference
it, which seems potentially unsafe.

Such callers include spidev_release(),  spi_dev_check(),
driver_override_store(), etc.


-- 
Kind Regards,

Yizhuo Zhai

Computer Science, Graduate Student
University of California, Riverside

Re: Potential NULL pointer deference in spi

[Eric Dumazet]

On 10/9/19 10:37 PM, Yizhuo Zhai wrote:
> Hi All:
> 
> drivers/spi/spi.c:
> 
> The function to_spi_device() could return NULL, but some callers
> in this file does not check the return value while directly dereference
> it, which seems potentially unsafe.
> 
> Such callers include spidev_release(),  spi_dev_check(),
> driver_override_store(), etc.
> 
> 


Many of your reports are completely bogus.

I suggest you spend more time before sending such emails to very large audience
and risk being ignored at some point.

Thanks.

Re: Potential NULL pointer deference in spi

[Yizhuo Zhai]

Hi Eric:

My apologies for bothering, we got those report via static analysis
and haven't got a good method to verify the path to trigger them.
Therefore I sent those email to you maintainers first since you
know much better about the details. Sorry again for your time and
I take your suggestions.

On Wed, Oct 9, 2019 at 10:48 PM Eric Dumazet <eric.dumazet@gmail.com> wrote:
>
>
>
> On 10/9/19 10:37 PM, Yizhuo Zhai wrote:
> > Hi All:
> >
> > drivers/spi/spi.c:
> >
> > The function to_spi_device() could return NULL, but some callers
> > in this file does not check the return value while directly dereference
> > it, which seems potentially unsafe.
> >
> > Such callers include spidev_release(),  spi_dev_check(),
> > driver_override_store(), etc.
> >
> >
>
>
> Many of your reports are completely bogus.
>
> I suggest you spend more time before sending such emails to very large audience
> and risk being ignored at some point.
>
> Thanks.



-- 
Kind Regards,

Yizhuo Zhai

Computer Science, Graduate Student
University of California, Riverside

Re: Potential NULL pointer deference in spi

[Eric Dumazet]

On 10/10/19 10:31 PM, Yizhuo Zhai wrote:
> Hi Eric:
> 
> My apologies for bothering, we got those report via static analysis
> and haven't got a good method to verify the path to trigger them.
> Therefore I sent those email to you maintainers first since you
> know much better about the details. Sorry again for your time and
> I take your suggestions.

My suggestion is that you need to make deep investigations on your own,
before sending mails to lkml@, reaching thousands of people on the planet.

Static analysis tools having too many false positive are not worth
the time spent by humans.

I knew nothing about drivers/spi/spi.c, but after few minutes reading the code,
it was clear your report was wrong.

Do not ask us to do what you should do yourself.

Thanks.

> 
> On Wed, Oct 9, 2019 at 10:48 PM Eric Dumazet <eric.dumazet@gmail.com> wrote:
>>
>>
>>
>> On 10/9/19 10:37 PM, Yizhuo Zhai wrote:
>>> Hi All:
>>>
>>> drivers/spi/spi.c:
>>>
>>> The function to_spi_device() could return NULL, but some callers
>>> in this file does not check the return value while directly dereference
>>> it, which seems potentially unsafe.
>>>
>>> Such callers include spidev_release(),  spi_dev_check(),
>>> driver_override_store(), etc.
>>>
>>>
>>
>>
>> Many of your reports are completely bogus.
>>
>> I suggest you spend more time before sending such emails to very large audience
>> and risk being ignored at some point.
>>
>> Thanks.
> 
> 
>