On 3/6/26 12:10 PM, Guenter Roeck wrote:
> On Fri, Mar 06, 2026 at 11:16:51AM -0600, Andrew Davis wrote:
>> The function i2c_match_id() is used to fetch the matching ID from
>> the i2c_device_id table. This can instead be done with
>> i2c_client_get_device_id(). For this driver functionality should
>> not change. Switch over to remove the last couple users of the
>> i2c_match_id() function from kernel.
>>
>> Signed-off-by: Andrew Davis <afd@ti.com>
>> ---
>> drivers/hwmon/pmbus/ltc2978.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/hwmon/pmbus/ltc2978.c b/drivers/hwmon/pmbus/ltc2978.c
>> index 8f5be520a15db..d69a5e675e80e 100644
>> --- a/drivers/hwmon/pmbus/ltc2978.c
>> +++ b/drivers/hwmon/pmbus/ltc2978.c
>> @@ -733,7 +733,7 @@ static int ltc2978_probe(struct i2c_client *client)
>> return chip_id;
>>
>> data->id = chip_id;
>> - id = i2c_match_id(ltc2978_id, client);
>> + id = i2c_client_get_device_id(client);
>
> AI feedback:
>
> Is `id` guaranteed to be non-NULL here?
>
> If the device is instantiated via ACPI `PRP0001` or using a fallback DT
> compatible string where the first compatible string is not in the
> `ltc2978_id` table, `i2c_client_get_device_id()` will return `NULL`.
> This leads to a NULL pointer dereference when accessing `id->driver_data`.
>
> While this vulnerability existed in the old code with `i2c_match_id()`,
> adding a NULL check here might be a good idea while the code is being
> refactored.
>
> I never know if this is real. Any idea ?
>
The AI is right on both parts, the second being the important one that
this was preexisting. i2c_match_id() should never return NULL in
practice as we must have matched on something to have gotten probe()'d
in the first place. And for the same reason i2c_client_get_device_id()
shouldn't ever return NULL either. So no change here.
But I see how this would confuse an AI as both functions have a
"return NULL;" statement in them. This problem of returning the matched
ID in-band with returning 0 for errors (which could also be a valid ID)
is one of the motivating reasons for me removing i2c_match_id(). Even
though its replacement has the same issue, once we get everyone over
to the same single API then switching that one API over to something
safer becomes possible.
Andrew
> Thanks,
> Guenter
>
>> if (data->id != id->driver_data)
>> dev_warn(&client->dev,
>> "Device mismatch: Configured %s (%d), detected %d\n",
>> --
>> 2.39.2
>>