Re: Potential null-pointer-dereference problem due to missing null-checking for ata_timing_find_mode

[Damien Le Moal <damien.lemoal@opensource.wdc.com>]

On 2021/10/25 3:08, YE Chengfeng wrote:
> Hi,
> 
> I found that the function *ata_timing_find_mode *could return a null pointer in
> some situattions, but some call sites of this function don't check whether the
> return value is a null pointer. Could it be a potential null-pointer-dereference
> problem? 
> https://github.com/torvalds/linux/blob/master/drivers/ata/pata_acpi.c#L145
> <https://github.com/torvalds/linux/blob/master/drivers/ata/pata_acpi.c#L145>
> <https://github.com/torvalds/linux/blob/master/drivers/ata/pata_acpi.c#L145>
> 	
> linux/pata_acpi.c at master · torvalds/linux
> <https://github.com/torvalds/linux/blob/master/drivers/ata/pata_acpi.c#L145>
> Linux kernel source tree. Contribute to torvalds/linux development by creating
> an account on GitHub.
> github.com
> 
> 
> 
> Best Regards,
> - Chengfeng

The ata_timing array last element is 0xff, which is the initial value of pio/dma
mode. So an entry will always be returned. I do not think the NULL return ever
triggers.

-- 
Damien Le Moal
Western Digital Research