From: Sebastian Frias <sf84@laposte.net>
Subject: Re: [PATCH v2] net: ethernet: support "fixed-link" DT key/node on
 nb8800 driver
Date: Fri, 05 Feb 2016 15:08:04 +0100
Message-ID: <56B4ACC4.1000607@laposte.net>
References: <56B4A445.7080402@laposte.net>	<yw1x1t8ri8dz.fsf@unicorn.mansr.com> <56B4A877.4020800@laposte.net> <yw1xsi17gsxl.fsf@unicorn.mansr.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=windows-1252;
	format=flowed
Content-Transfer-Encoding: QUOTED-PRINTABLE
Cc: "David S. Miller" <davem@davemloft.net>, netdev@vger.kernel.org,
	LKML <linux-kernel@vger.kernel.org>, mason <slash.tmp@free.fr>
To: =?windows-1252?Q?M=E5ns_Rullg=E5rd?= <mans@mansr.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
In-Reply-To: <yw1xsi17gsxl.fsf@unicorn.mansr.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: netdev.vger.kernel.org

On 02/05/2016 02:58 PM, M=E5ns Rullg=E5rd wrote:
> Sebastian Frias <sf84@laposte.net> writes:
>
>> Signed-off-by: Sebastian Frias <sf84@laposte.net>
>> ---
>>   drivers/net/ethernet/aurora/nb8800.c | 15 ++++++++++++---
>>   1 file changed, 12 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/aurora/nb8800.c
>> b/drivers/net/ethernet/aurora/nb8800.c
>> index ecc4a33..dd7bedc 100644
>> --- a/drivers/net/ethernet/aurora/nb8800.c
>> +++ b/drivers/net/ethernet/aurora/nb8800.c
>> @@ -1462,9 +1462,18 @@ static int nb8800_probe(struct platform_devic=
e *pdev)
>>
>>   	priv->phy_node =3D of_parse_phandle(pdev->dev.of_node, "phy-handl=
e", 0);
>>   	if (!priv->phy_node) {
>> -		dev_err(&pdev->dev, "no PHY specified\n");
>> -		ret =3D -ENODEV;
>> -		goto err_free_bus;
>> +		if (of_phy_is_fixed_link(pdev->dev.of_node)) {
>> +			ret =3D of_phy_register_fixed_link(pdev->dev.of_node);
>> +			if (ret < 0) {
>> +				dev_err(&pdev->dev, "bad fixed-link spec\n");
>> +				goto err_free_bus;
>> +			}
>> +			priv->phy_node =3D of_node_get(pdev->dev.of_node);
>> +		} else {
>> +			dev_err(&pdev->dev, "no PHY specified\n");
>> +			ret =3D -ENODEV;
>> +			goto err_free_bus;
>> +		}
>>   	}
>
> Maybe it would be clearer to reduce the if() nesting a bit, like this
> for instance:
>
> 	if (of_phy_is_fixed_link(pdev->dev.of_node)) {
> 		ret =3D of_phy_register_fixed_link(pdev->dev.of_node);
> 		if (ret < 0) {
> 			dev_err(&pdev->dev, "bad fixed-link spec\n");
> 			goto err_free_bus;
> 		}
> 		priv->phy_node =3D of_node_get(pdev->dev.of_node);
> 	}
>
> 	if (!priv->phy_node)
> 		priv->phy_node =3D of_parse_phandle(pdev->dev.of_node,
> 						  "phy-handle", 0);
>
> 	if (!priv->phy_node) {
> 		dev_err(&pdev->dev, "no PHY specified\n");
> 		ret =3D -ENODEV;
> 		goto err_free_bus;
> 	}
>
>

Thanks M=E5ns for your comments.
With old code + my patch, we only hit 1 comparison in the general case,=
=20
and a 2nd one in "fixed-link" case.
With your suggestion above, it would mean that we hit 3 comparisons all=
=20
the time.
If you are ok with the 3 comparisons, I can post a v3.