Received-SPF: pass (google.com: domain of linux-kernel+bounces-130842-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.48.161 as permitted sender) client-ip=147.75.48.161;
Message-ID: <62ae8487-b768-424e-b6a5-a5f31b3b55bb@kernel.org>
Date: Thu, 4 Apr 2024 11:40:37 +0900
Precedence: bulk
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: [PATCH v2 02/10] PCI: endpoint: Decouple EPC and PCIe bus
 specific events
To: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Cc: Lorenzo Pieralisi <lpieralisi@kernel.org>,
 =?UTF-8?Q?Krzysztof_Wilczy=C5=84ski?= <kw@linux.com>,
 Rob Herring <robh@kernel.org>, Bjorn Helgaas <bhelgaas@google.com>,
 Kishon Vijay Abraham I <kishon@kernel.org>,
 Thierry Reding <thierry.reding@gmail.com>,
 Jonathan Hunter <jonathanh@nvidia.com>, Jingoo Han <jingoohan1@gmail.com>,
 linux-pci@vger.kernel.org, linux-arm-msm@vger.kernel.org,
 linux-kernel@vger.kernel.org, mhi@lists.linux.dev,
 linux-tegra@vger.kernel.org, Niklas Cassel <cassel@kernel.org>
References: <20240401-pci-epf-rework-v2-0-970dbe90b99d@linaro.org>
 <20240401-pci-epf-rework-v2-2-970dbe90b99d@linaro.org>
 <45b2db99-2d03-469b-aa37-bc6c63cef141@kernel.org>
 <20240403142650.GA72531@thinkpad>
From: Damien Le Moal <dlemoal@kernel.org>
Content-Language: en-US
Organization: Western Digital Research
In-Reply-To: <20240403142650.GA72531@thinkpad>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit

On 4/3/24 23:26, Manivannan Sadhasivam wrote:
> On Tue, Apr 02, 2024 at 09:14:20AM +0900, Damien Le Moal wrote:
>> On 4/2/24 00:50, Manivannan Sadhasivam wrote:
>>> Currently, 'struct pci_epc_event_ops' has a bunch of events that are sent
>>> from the EPC driver to EPF driver. But those events are a mix of EPC
>>> specific events like core_init and PCIe bus specific events like LINK_UP,
>>> LINK_DOWN, BME etc...
>>>
>>> Let's decouple them to respective structs (pci_epc_event_ops,
>>> pci_epc_bus_event_ops) to make the separation clear.
>>
>> I fail to see the benefits here. The event operation names are quite clear and,
>> in my opinion, it is clear if an event op applies to the controller or to the
>> bus/link. If anything, "core_init" could a little more clear, so renaming that
>> "ep_controller_init" or something like that (clearly spelling out what is being
>> initialized) seems enough to me. Similarly, the "bme" op name is very criptic.
>> Renaming that to "bus_master_enable" would go a long way clarifying the code.
>> For link events, "link_up", "link_down" are clear. So I think there is no need
>> to split the event op struct like this. Renaming the ops is better.
>>
>> Note that I am not opposed to this patch, but I think it is just code churn
>> that does not really bring any fundamental improvement. Regardless, renaming
>> "core_init" and "bme" ops is I think desired.
>>
> 
> Niklas shared the same view during v1, but I hate to see the events being mixed
> in a single ops. Especially that it will confuse the developers who are not
> familiar with the EP subsystem.
> 
> But since the argument is coming twice, I've decided to drop this for now and
> just rename the 'core_init' callback to 'epc_init' and name the deinit callback
> as 'epc_deinit'.

Sounds good. Please also rename the completely unclear "bme" operation. Spell it
out to be clear.

-- 
Damien Le Moal
Western Digital Research