Re: [PATCHv6 5/5] hwspinlock/omap: add support for dt nodes

[Suman Anna]

Hi Ohad,

On 11/12/2014 01:14 PM, Ohad Ben-Cohen wrote:
> Hi Suman,
> 
> On Fri, Sep 12, 2014 at 11:24 PM, Suman Anna <s-anna@xxxxxx> wrote:
>>  static int omap_hwspinlock_probe(struct platform_device *pdev)
>>  {
>> -       struct hwspinlock_pdata *pdata = pdev->dev.platform_data;
>> +       struct device_node *node = pdev->dev.of_node;
>>         struct hwspinlock_device *bank;
>>         struct hwspinlock *hwlock;
>>         struct resource *res;
>>         void __iomem *io_base;
>> -       int num_locks, i, ret;
>> +       int num_locks, i, ret, base_id;
>>
>> -       if (!pdata)
>> +       if (!node)
>>                 return -ENODEV;
>>
>> +       ret = of_hwspin_lock_get_base_id(node);
>> +       if (ret < 0 && ret != -EINVAL)
>> +               return -ENODEV;
>> +       base_id = (ret > 0 ? ret : 0);
> 
> Does this mean you allow nodes not to have the base_id property? How
> do we protect against multiple nodes not having a base_id property
> then?
> 
> Implicitly assuming a base_id value (zero in this case) may not be always safe.

None of the OMAPs have multiple IP instances, and as such the base-id is
an optional property. I have made this change to make sure we atleast
attempt to use the value if mentioned in DT and not hard-coding the
value to begin with (going by the optional property semantics). If and
when multiple instances get added and a secondary node doesn't add the
property, the node will not be registered with the core due to an
overlap failure. Here is the previous version [1] for reference.

regards
Suman

[1] https://patchwork.kernel.org/patch/4096881/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/