Re: [PATCH v2] iio: adc: ad7923: Fix buffer overflow for tx_buf and ring_xfer

[Zicheng Qu <quzicheng@huawei.com>]

Hi Jonathan and Nuno,

Thank you for pointing that out. I included Nuno's name because I think 
the final correct solution came from Nuno, and I wanted to acknowledge 
the contribution. However, I didn't realize I needed to confirm with 
Nuno before adding the sign-off.

In the future, I will ensure to discuss with Nuno or anyone else 
involved to avoid similar issues, or use "suggested by" instead. I 
apologize to Nuno for any confusion this may have caused.

Thanks for the guidance and apologize again.

Best regards,
Zicheng Qu


On 2024/11/1 5:05, Jonathan Cameron wrote:
> On Thu, 31 Oct 2024 15:20:24 +0100
> Nuno Sá <noname.nuno@gmail.com> wrote:
>
>> On Tue, 2024-10-29 at 13:46 +0000, Zicheng Qu wrote:
>>> The AD7923 was updated to support devices with 8 channels, but the size
>>> of tx_buf and ring_xfer was not increased accordingly, leading to a
>>> potential buffer overflow in ad7923_update_scan_mode().
>>>
>>> Fixes: 851644a60d20 ("iio: adc: ad7923: Add support for the ad7908/ad7918/ad7928")
>>> Cc: <stable@vger.kernel.org>
>>> Signed-off-by: Nuno Sá <noname.nuno@gmail.com>
>>> Signed-off-by: Zicheng Qu <quzicheng@huawei.com>
>>> ---
>> Reviewed-by: Nuno Sa <nuno.sa@analog.com>
>>
> Confusing one. I'll fix the authorship up for your analog address
>
> Zicheng, usually a Suggested-by after checking with the author if it's
> a patch in a review thread.
>
> You can't really give someone elses' SoB without them explicitly sending it.
> If Nuno let you know that was fine off the list, then just mention that under
> ---
>
> This time I'm going to take Nuno's RB as fine to indicate no objection
> to the SoB. Nuno, feel free to shout if you want to handle this differently.
>
> Applied.
>
> Jonathan
>
>
>>> v2:
>>> - Fixed: Addressed buffer overflow in ad7923_update_scan_mode() due to
>>> insufficient tx_buf and ring_xfer size for 8-channel devices.
>>> - Issue: Original patch attempted to fix the overflow by limiting the
>>> length, but did not address the root cause of buffer size mismatch.
>>> - Solution: Increased tx_buf and ring_xfer sizes recommended by Nuno to
>>> support all 8 channels, ensuring adequate buffer capacity.
>>> - Previous patch link:
>>> https://lore.kernel.org/linux-iio/20241028142357.1032380-1-quzicheng@huawei.com/T/#u
>>>   drivers/iio/adc/ad7923.c | 4 ++--
>>>   1 file changed, 2 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/iio/adc/ad7923.c b/drivers/iio/adc/ad7923.c
>>> index 09680015a7ab..acc44cb34f82 100644
>>> --- a/drivers/iio/adc/ad7923.c
>>> +++ b/drivers/iio/adc/ad7923.c
>>> @@ -48,7 +48,7 @@
>>>   
>>>   struct ad7923_state {
>>>   	struct spi_device		*spi;
>>> -	struct spi_transfer		ring_xfer[5];
>>> +	struct spi_transfer		ring_xfer[9];
>>>   	struct spi_transfer		scan_single_xfer[2];
>>>   	struct spi_message		ring_msg;
>>>   	struct spi_message		scan_single_msg;
>>> @@ -64,7 +64,7 @@ struct ad7923_state {
>>>   	 * Length = 8 channels + 4 extra for 8 byte timestamp
>>>   	 */
>>>   	__be16				rx_buf[12] __aligned(IIO_DMA_MINALIGN);
>>> -	__be16				tx_buf[4];
>>> +	__be16				tx_buf[8];
>>>   };
>>>   
>>>   struct ad7923_chip_info {