Re: [PATCH 3/5] mt76: fix a leaked reference by adding a missingof_node_put

[Markus Elfring]

>>> @@ -54,22 +54,30 @@  mt76_get_of_eeprom(struct mt76_dev *dev, int len)
>>>          part = np->name;
>>>
>>>      mtd = get_mtd_device_nm(part);
>> â
>>> +    if (retlen < len) {
>>> +        ret = -EINVAL;
>>> +        goto out_put_node;
>>
>> I find a jump to an immediately following source code place unnecessary.
>> Would you like to delete it?
>>
>>
>>> +    }
>>>
>>> -    return 0;
>>> +out_put_node:
>>> +    of_node_put(np);
>>> +    return ret;
>>>  #else
>>
>> Can another bit of fine-tuning matter here?
>> https://git.kernel.org/pub/scm/linux/kernel/git/next/linux-next.git/tree/drivers/net/wireless/mediatek/mt76/eeprom.c?id=34e022d8b780a03902d82fb3997ba7c7b1f40c81#n73
â
> We may have some different opinions here.

Obviously, yes for this implementation detail.


> Deleting the goto statement may not be good.

I find such an adjustment helpful here.


> If the code further up is changed it's easy enough to miss
> that a goto statement needs to be added here.

There are the usual consequences to consider for every change.


> Better to set ret to zero explicitly, this is the success path after all.

I disagree to this information because the variable was set to
the return value from a call of the function âmtd_readâ already.

Regards,
Markus