On 2/15/26 11:55 PM, Jonathan Cameron wrote:
> On Sun, 15 Feb 2026 22:06:28 +0200
> Erikas Bitovtas <[email protected]> wrote:
>
>> On 2/15/26 9:31 PM, Jonathan Cameron wrote:
>>> On Sun, 15 Feb 2026 19:28:56 +0200
>>> Erikas Bitovtas <[email protected]> wrote:
>>>
>>>> On 2/14/26 8:09 PM, Jonathan Cameron wrote:
>>>>>> ---
>>>>>> drivers/iio/light/vcnl4000.c | 40
>>>>>> ++++++++++++++++++++++++++++++++++++++++
>>>>>> 1 file changed, 40 insertions(+)
>>>>>>
>>>>>> diff --git a/drivers/iio/light/vcnl4000.c b/drivers/iio/light/vcnl4000.c
>>>>>> index a36c23813679..1f8f4e4586f4 100644
>>>>>> --- a/drivers/iio/light/vcnl4000.c
>>>>>> +++ b/drivers/iio/light/vcnl4000.c
>>>>>> @@ -185,6 +185,7 @@ static const int vcnl4040_ps_oversampling_ratio[] =
>>>>>> {1, 2, 4, 8};
>>>>>> #define VCNL4000_SLEEP_DELAY_MS 2000 /* before we enter
>>>>>> pm_runtime_suspend */
>>>>>>
>>>>>> enum vcnl4000_device_ids {
>>>>>> + CM36672P,
>>>>>> VCNL4000,
>>>>>> VCNL4010,
>>>>>> VCNL4040,
>>>>>> @@ -235,6 +236,8 @@ struct vcnl4000_chip_spec {
>>>>>> };
>>>>>>
>>>>>> static const struct i2c_device_id vcnl4000_id[] = {
>>>>>> + { "cm36672p", CM36672P },
>>>>>> + { "cm36686", VCNL4040 },
>>>>>> { "vcnl4000", VCNL4000 },
>>>>>> { "vcnl4010", VCNL4010 },
>>>>>> { "vcnl4020", VCNL4010 },
>>>>>> @@ -1842,6 +1845,22 @@ static const struct iio_chan_spec
>>>>>> vcnl4040_channels[] = {
>>>>>> }
>>>>>> };
>>>>>
>>>>> ...
>>>>>
>>>>>> [VCNL4000] = {
>>>>>> .prod = "VCNL4000",
>>>>>> .init = vcnl4000_init,
>>>>>> @@ -2033,6 +2065,14 @@ static int vcnl4000_probe(struct i2c_client
>>>>>> *client)
>>>>>> }
>>>>>>
>>>>>> static const struct of_device_id vcnl_4000_of_match[] = {
>>>>>> + {
>>>>>> + .compatible = "capella,cm36672p",
>>>>>> + .data = (void *)CM36672P,
>>>>>> + },
>>>>>> + {
>>>>>> + .compatible = "capella,cm36686",
>>>>>> + .data = (void *)VCNL4040,
>>>>>
>>>>> Is this necessary? I 'think' if you drop it we'll match instead
>>>>> on the vcnl4040 fallback and then the access to the data will be
>>>>> through the stripped name only bit of the compatible (first entry, not
>>>>> the fallback so cm36686 in this case). So you do need the cm36686
>>>>> entry in the i2c_device_id table above. Probably better to keep
>>>>> this here to avoid having to reason this out - but perhaps a
>>>>> comment to that affect would be useful (assuming you verify my
>>>>> reasoning).
>>>>>
>>>> After I removed the entry for "capella,cm36686", I received the "Unable
>>>> to handle kernel NULL pointer dereference" error in dmesg. And at least
>>>> stk3310 driver includes a compatible entry both for the device (stk3013)
>>>> and for the fallback (stk3310). So my assumption is that this entry is
>>>> needed.
>>>> I could include a comment explaining that cm36686 is fully compatible
>>>> with vcnl4040, however, if that is necessary.
>>>
>>> Thanks for checking.
>>>
>>> What did you get as the backtrace? I'm hoping it'll explain what I'm
>>> misunderstanding! The hacks around using the wrong table for compatible
>>> matches have tripped me up before.
>>>
>>> Jonathan
>>>
>>
>> I am attaching a link to the dmesg. There were quite a lot of lines in
>> the stack trace and I am not sure what is the right way to post logs in
>> the mailing list.
>>
>> https://pastebin.com/QgeTdNEP
>
> Thanks. only relevant bit is probably:
>
> [ 15.566076] vcnl4000_probe+0x54/0x288 [vcnl4000] (P)
> [ 15.566102] i2c_device_probe+0x2b0/0x358
> [ 15.566121] really_probe+0x154/0x448
>
> My guess is my understanding of i2c_client_get_device_id() is wrong and that
> is returning NULL. That can only happen if client->name is not a match for
> anything the i2_device_id table. If you have a chance, can you dump
> what client->name is in this case? I thought it ended up as
> cm36686 (stripped first entry in compatible) but seems I'm probably wrong on
> that :(
>
> The path I thought worked was via info->type (which gets copied to
> client->name)
> set via of_alias_from_compatible() here.
> https://elixir.bootlin.com/linux/v6.19-rc4/source/drivers/i2c/i2c-core-of.c#L30
> Which should just return the first compatible without that vendor prefix.
>
> Meh, this doesn't really matter anyway as once we refactor to actually use
> the data in the of_device_id table, we will need the entry and in the meantime
> it's sort of documentation.
>
> J
>
Apparently I just had commented out the i2c_device_id entry for cm36686
as well, when I had to comment out only of_device_id entry. After adding
i2c_device_id entry back, it works, just as you said.
I will submit a v5 with of_device_id entry removed if that is necessary.
