Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
IronPort-SDR: yKHBdykJS8cE/gHmO8vawlE4qOKOD11ZMgi8l9HgJGIiOErlQpiS/WbbWa0K28owapGB+HQ9cF
 7DURQizwIfvGErdPAXhBRPxJ0VVlYgR/97rWcQaBgYEedozbMzDGfqT0DNX3uqxBYDp9H/S6W7
 lnCKfG1R1Sp8m/ECq9ADEXEd89nAET7ideI+dv3BtME5WXJLGwMcUXgies+R8EsZviTq7SkG5U
 /+4OgoSvgNVPOaro+rfh0K2BZ1Ww8iQ8PyHJrADbEK0DR9JtRa/gppM85TmIBx1/u2IqTOxK3Y
 889Cqb6pjyVZJk5MU8ABzMst
IronPort-SDR: GWxNxY1xDj5O+reACcUJUlLDOtQ47w3Bn/ieqQKHWASogK+WRnycnnI5OJg1jIeaFntjDENoTR
 Hl7QMbqUFqgZj0hBEmfwxE6cg3Zy6u0WRB2wCUSzZPfI0K3DCRnc+b47ZHUxrhDhMkOXlyLonP
 RSjEV+MeEFrwriWCnrgDmOuN+TeVVbtmtnx9c5AvyHh5dLmArii9UycIzVjOx3dLGxn+o5N0XS
 xWvxClap6XzWO3S1uB1hDY6BZGhpYG7x/81see3r/3sNZcd4j3MV4d8upULwyG/8C52WYDyoId
 ucY=
WDCIronportException: Internal
Message-ID: <0dcebae2-5e4e-a0d3-181d-37bb9b40d564@opensource.wdc.com>
Date:   Thu, 16 Jun 2022 09:23:28 +0900
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:91.0)
 Gecko/20100101 Thunderbird/91.10.0
Subject: Re: [PATCH v4 07/23] ata: libahci_platform: Convert to using devm
 bulk clocks API
Content-Language: en-US
To:     Serge Semin <fancer.lancer@gmail.com>
Cc:     Serge Semin <Sergey.Semin@baikalelectronics.ru>,
        Hans de Goede <hdegoede@redhat.com>,
        Jens Axboe <axboe@kernel.dk>, Hannes Reinecke <hare@suse.de>,
        Alexey Malahov <Alexey.Malahov@baikalelectronics.ru>,
        Pavel Parkhomenko <Pavel.Parkhomenko@baikalelectronics.ru>,
        Rob Herring <robh+dt@kernel.org>, linux-ide@vger.kernel.org,
        linux-kernel@vger.kernel.org, devicetree@vger.kernel.org
References: <20220610081801.11854-1-Sergey.Semin@baikalelectronics.ru>
 <20220610081801.11854-8-Sergey.Semin@baikalelectronics.ru>
 <3bf20887-6e2f-41f4-e4ec-5c2278f6cb18@opensource.wdc.com>
 <20220615204509.siz54h4vbgvb3zkm@mobilestation>
From:   Damien Le Moal <damien.lemoal@opensource.wdc.com>
Organization: Western Digital Research
In-Reply-To: <20220615204509.siz54h4vbgvb3zkm@mobilestation>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 2022/06/16 5:45, Serge Semin wrote:
[...]
>>> +		hpriv->clks = devm_kzalloc(dev, sizeof(*hpriv->clks), GFP_KERNEL);
>>> +		if (!hpriv->clks) {
>>> +			rc = -ENOMEM;
>>> +			goto err_out;
>>> +		}
>>> +		hpriv->clks->clk = devm_clk_get_optional(dev, NULL);
> 
>>> +		if (IS_ERR(hpriv->clks->clk)) {
>>> +			rc = PTR_ERR(hpriv->clks->clk);
>>> +			goto err_out;
>>> +		} else if (hpriv->clks->clk) {
>>
>> Nit: the else is not needed here.
> 
> Well, it depends on what you see behind it. I see many reasons to keep
> it and only one tiny reason to drop it. Keeping it will improve the
> code readability and maintainability like having a more natural
> execution flow representation, thus clearer read-flow (else part as
> exception to the if part), less modifications should the goto part is
> changed/removed, a more exact program flow representation can be used
> by the compiler for some internal optimizations, it's one line shorter
> than the case we no 'else' here. On the other hand indeed we can drop
> it since if the conditional statement is true, the code afterwards
> won't be executed due to the goto operator. But as I see it dropping
> the else operator won't improve anything, but vise-versa will worsen
> the code instead. So if I get to miss something please justify why you
> want it being dropped, otherwise I would rather preserve it.

An else after a goto or return is never necessary and in my opinion makes the
code harder to read. I am not interested in debating this in general anyway. For
this particular case, the code would be:

		hpriv->clks->clk = devm_clk_get_optional(dev, NULL);
		if (IS_ERR(hpriv->clks->clk)) {
			/* Error path */
			rc = PTR_ERR(hpriv->clks->clk);
			goto err_out;
		}

		/* Normal path */
		if (hpriv->clks->clk) {
			...
		}

Which in my opinion is a lot easier to understand compared to having to parse
the if/else if and figure out which case in that sequence is normal vs error.

As noted, this is a nit. If you really insist, keep that else if.

> 
> -Sergey
> 
>>
>>> +			hpriv->clks->id = __clk_get_name(hpriv->clks->clk);
>>> +			hpriv->n_clks = 1;
>>>  		}
>>> -		hpriv->clks[i] = clk;
>>>  	}
>>>  
>>>  	hpriv->ahci_regulator = devm_regulator_get(dev, "ahci");
>>
>>
>> -- 
>> Damien Le Moal
>> Western Digital Research


-- 
Damien Le Moal
Western Digital Research