On 28/02/2016 16:13, Parav Pandit wrote:
> diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
> index 00da80e..54ea8ce 100644
> --- a/drivers/infiniband/core/device.c
> +++ b/drivers/infiniband/core/device.c
> @@ -343,28 +343,38 @@ int ib_register_device(struct ib_device *device,
>
> ret = read_port_immutable(device);
> if (ret) {
> - printk(KERN_WARNING "Couldn't create per port immutable data %s\n",
> - device->name);
> + pr_warn("Couldn't create per port immutable data %s\n",
> + device->name);
> goto out;
This change doesn't belong in the patch.
On Tue, Mar 1, 2016 at 2:42 PM, Haggai Eran <[email protected]> wrote:
> On 28/02/2016 16:13, Parav Pandit wrote:
>> diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
>> index 00da80e..54ea8ce 100644
>> --- a/drivers/infiniband/core/device.c
>> +++ b/drivers/infiniband/core/device.c
>> @@ -343,28 +343,38 @@ int ib_register_device(struct ib_device *device,
>>
>> ret = read_port_immutable(device);
>> if (ret) {
>> - printk(KERN_WARNING "Couldn't create per port immutable data %s\n",
>> - device->name);
>> + pr_warn("Couldn't create per port immutable data %s\n",
>> + device->name);
>> goto out;
>
> This change doesn't belong in the patch.
I agree, but few warnings are with pr_warn and few with printk just
make code look uneven.
So I changed printk to pr_warn in same function instead of spinning
complete new patch.
On 01/03/2016 11:22, Parav Pandit wrote:
> On Tue, Mar 1, 2016 at 2:42 PM, Haggai Eran <[email protected]> wrote:
>> On 28/02/2016 16:13, Parav Pandit wrote:
>>> diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
>>> index 00da80e..54ea8ce 100644
>>> --- a/drivers/infiniband/core/device.c
>>> +++ b/drivers/infiniband/core/device.c
>>> @@ -343,28 +343,38 @@ int ib_register_device(struct ib_device *device,
>>>
>>> ret = read_port_immutable(device);
>>> if (ret) {
>>> - printk(KERN_WARNING "Couldn't create per port immutable data %s\n",
>>> - device->name);
>>> + pr_warn("Couldn't create per port immutable data %s\n",
>>> + device->name);
>>> goto out;
>>
>> This change doesn't belong in the patch.
> I agree, but few warnings are with pr_warn and few with printk just
> make code look uneven.
> So I changed printk to pr_warn in same function instead of spinning
> complete new patch.
Still, I think it would be better to have such cosmetic changes in a
separate patch, so that we have a cleaner git history. You can send
this extra patch separately from this patchset so that Doug can take
it independently.
On Tue, Mar 1, 2016 at 4:36 PM, Haggai Eran <[email protected]> wrote:
> On 01/03/2016 11:22, Parav Pandit wrote:
>> On Tue, Mar 1, 2016 at 2:42 PM, Haggai Eran <[email protected]> wrote:
>>> On 28/02/2016 16:13, Parav Pandit wrote:
>>>> diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c
>>>> index 00da80e..54ea8ce 100644
>>>> --- a/drivers/infiniband/core/device.c
>>>> +++ b/drivers/infiniband/core/device.c
>>>> @@ -343,28 +343,38 @@ int ib_register_device(struct ib_device *device,
>>>>
>>>> ret = read_port_immutable(device);
>>>> if (ret) {
>>>> - printk(KERN_WARNING "Couldn't create per port immutable data %s\n",
>>>> - device->name);
>>>> + pr_warn("Couldn't create per port immutable data %s\n",
>>>> + device->name);
>>>> goto out;
>>>
>>> This change doesn't belong in the patch.
>> I agree, but few warnings are with pr_warn and few with printk just
>> make code look uneven.
>> So I changed printk to pr_warn in same function instead of spinning
>> complete new patch.
> Still, I think it would be better to have such cosmetic changes in a
> separate patch, so that we have a cleaner git history. You can send
> this extra patch separately from this patchset so that Doug can take
> it independently.
o.k.