Re: [PATCH 11/15] PCI: pci-epf-test: Use pci_epc_get_features to get EPC features

From: Kishon Vijay Abraham I <kishon@ti.com>
To: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Bjorn Helgaas <bhelgaas@google.com>,
	Gustavo Pimentel <gustavo.pimentel@synopsys.com>,
	Alan Douglas <adouglas@cadence.com>,
	Shawn Lin <shawn.lin@rock-chips.com>,
	Jingoo Han <jingoohan1@gmail.com>,
	Heiko Stuebner <heiko@sntech.de>,
	Cyrille Pitchen <cyrille.pitchen@free-electrons.com>,
	Jia-Ju Bai <baijiaju1990@gmail.com>, <linux-omap@vger.kernel.org>,
	<linux-pci@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
	<linux-rockchip@lists.infradead.org>,
	<linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH 11/15] PCI: pci-epf-test: Use pci_epc_get_features to get EPC features
Date: Thu, 14 Feb 2019 10:36:35 +0530	[thread overview]
Message-ID: <4cb9ce03-e360-b518-96fd-f8a618a9c803@ti.com> (raw)
In-Reply-To: <20190213143646.GC25260@e107981-ln.cambridge.arm.com>

Hi Lorenzo,

On 13/02/19 8:06 PM, Lorenzo Pieralisi wrote:
> On Wed, Feb 13, 2019 at 07:08:18PM +0530, Kishon Vijay Abraham I wrote:
>> Hi Lorenzo,
>>
>> On 12/02/19 8:37 PM, Lorenzo Pieralisi wrote:
>>> On Mon, Jan 07, 2019 at 12:11:44PM +0530, Kishon Vijay Abraham I wrote:
>>>
>>> [...]
>>>
>>>>  static int pci_epf_test_bind(struct pci_epf *epf)
>>>>  {
>>>>  	int ret;
>>>>  	struct pci_epf_test *epf_test = epf_get_drvdata(epf);
>>>>  	struct pci_epf_header *header = epf->header;
>>>> +	const struct pci_epc_features *epc_features;
>>>> +	enum pci_barno test_reg_bar = BAR_0;
>>>>  	struct pci_epc *epc = epf->epc;
>>>>  	struct device *dev = &epf->dev;
>>>> +	bool linkup_notifier = false;
>>>> +	bool msix_capable = false;
>>>> +	bool msi_capable = true;
>>>>  
>>>>  	if (WARN_ON_ONCE(!epc))
>>>>  		return -EINVAL;
>>>>  
>>>> -	if (epc->features & EPC_FEATURE_NO_LINKUP_NOTIFIER)
>>>> -		epf_test->linkup_notifier = false;
>>>> -	else
>>>> -		epf_test->linkup_notifier = true;
>>>> -
>>>> -	epf_test->msix_available = epc->features & EPC_FEATURE_MSIX_AVAILABLE;
>>>> +	epc_features = pci_epc_get_features(epc, epf->func_no);
>>>
>>> I think it would work out better if struct pci_epc_features was
>>> allocated in the caller (stack) and pci_epc_get_features() take a
>>> pointer parameter to it rather than the callee and the callee would just
>>> have to fill it out, this also removes data in the driver that is not
>>> really useful.
>>>
>>> Is there any other reason behind the current design choice ?
>>
>> Some drivers are used by multiple platforms each with different features. In
>> such cases it's cleaner to have separate epc_feature table for each platform.
>>
>> I think the driver should maintain some sort of data to even populate
>> pci_epc_features allocated by EP function driver.
> 
> You mean that every EP controller driver should keep a table of
> pci_epc_features (instead of a single instance) to be matched using DT
> compatible strings to detect the platform variations ?

Yes.

Thanks
Kishon