Hi Andy,


On 09/01/2016 06:17 AM, Andy Shevchenko wrote:
> On Tue, 2016-08-30 at 11:18 -0700, Sathyanarayanan Kuppuswamy wrote:
>
>> IMO, Main problem here is, In some scenarios get_platform_data
>> returns
>> NULL on error case and in some cases it return NULL for no platform
>> data
>> case.
> We have possibility of the following scenarios:
> 1) fatal error (should return error code and fail booting)
> 2) non-fatal error (prevents certain device to be enumerated, pdata =
> NULL)
> 3) no pdata for the device (not an error! pdata is optional to the
> certain driver)
> 4) pdata != NULL (fully armed device driver)
>
> According to above I doubt we will have many 1) cases. Otherwise pdata =
> NULL is okay.
I agree with 2,3 and 4 scenarios. But I am not sure about the first 
case. Since these are peripheral devices, any failure in them should not 
stop the device boot. Do you have any examples for this case ?

Attached patch fixes the return value issues in get_platform_data code 
in device_libs directory. Please check and let me know your comments.

>
>> For example, in following function, returning NULL on gpio_base == -1
>> and when nr >= PCAL9555A_NUM is incorrect. I think in both these
>> scenarios valid behavior is to stop the device creation process.
> Yes, but do _not_ stop processing and booting!
>
>> Creating an i2c device without validi irq number or gpio number is
>> incorrect.
>>
>> So after investigating it bit more, I think its more appropriate to
>> fix
>> these error cases in device_libs dir. Let me know your comments.
> I doubt that is a proper fix. It might be part of the fix in sfi.c where
> you somehow distinguish fatal errors from non-fatal.
>
>

-- 
Sathyanarayanan Kuppuswamy
Android kernel developer