linux-kernel - Re: [PATCH v4 07/10] PCI: PCI: Add pcie_is_link_active() to determine if the PCIe link is active

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <29809286-cc04-fbc4-60bd-460821c4bc72@quicinc.com>
Date: Tue, 25 Feb 2025 15:36:10 +0530
From: Krishna Chaitanya Chundru <quic_krichai@...cinc.com>
To: Lukas Wunner <lukas@...ner.de>,
        Krishna Chaitanya Chundru
	<krishna.chundru@....qualcomm.com>
CC: Bjorn Helgaas <bhelgaas@...gle.com>,
        Lorenzo Pieralisi
	<lpieralisi@...nel.org>,
        Krzysztof Wilczy??ski <kw@...ux.com>,
        "Manivannan
 Sadhasivam" <manivannan.sadhasivam@...aro.org>,
        Rob Herring
	<robh@...nel.org>,
        Krzysztof Kozlowski <krzk+dt@...nel.org>,
        Conor Dooley
	<conor+dt@...nel.org>,
        Bjorn Andersson <andersson@...nel.org>,
        Konrad Dybcio
	<konradybcio@...nel.org>,
        <cros-qcom-dts-watchers@...omium.org>,
        Jingoo Han
	<jingoohan1@...il.com>,
        Bartosz Golaszewski <brgl@...ev.pl>, <quic_vbadigan@...cnic.com>,
        <amitk@...nel.org>, <dmitry.baryshkov@...aro.org>,
        <linux-pci@...r.kernel.org>, <devicetree@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>, <linux-arm-msm@...r.kernel.org>,
        <jorge.ramirez@....qualcomm.com>
Subject: Re: [PATCH v4 07/10] PCI: PCI: Add pcie_is_link_active() to determine
 if the PCIe link is active



On 2/25/2025 3:24 PM, Lukas Wunner wrote:
> On Tue, Feb 25, 2025 at 03:04:04PM +0530, Krishna Chaitanya Chundru wrote:
>> Introduce a common API to check if the PCIe link is active, replacing
>> duplicate code in multiple locations.
> [...]
>> --- a/drivers/pci/hotplug/pciehp_hpc.c
>> +++ b/drivers/pci/hotplug/pciehp_hpc.c
>> @@ -234,18 +234,7 @@ static void pcie_write_cmd_nowait(struct controller *ctrl, u16 cmd, u16 mask)
>>    */
>>   int pciehp_check_link_active(struct controller *ctrl)
>>   {
>> -	struct pci_dev *pdev = ctrl_dev(ctrl);
>> -	u16 lnk_status;
>> -	int ret;
>> -
>> -	ret = pcie_capability_read_word(pdev, PCI_EXP_LNKSTA, &lnk_status);
>> -	if (ret == PCIBIOS_DEVICE_NOT_FOUND || PCI_POSSIBLE_ERROR(lnk_status))
>> -		return -ENODEV;
>> -
>> -	ret = !!(lnk_status & PCI_EXP_LNKSTA_DLLLA);
>> -	ctrl_dbg(ctrl, "%s: lnk_status = %x\n", __func__, lnk_status);
>> -
>> -	return ret;
>> +	return pcie_is_link_active(ctrl_dev(ctrl));
>>   }
> 
> Please replace all call sites of pciehp_check_link_active() with a call
> to the new function.
> 
> 
ack
>> --- a/drivers/pci/pci.c
>> +++ b/drivers/pci/pci.c
>> @@ -4923,8 +4922,7 @@ int pci_bridge_wait_for_secondary_bus(struct pci_dev *dev, char *reset_type)
>>   		if (!dev->link_active_reporting)
>>   			return -ENOTTY;
>>   
>> -		pcie_capability_read_word(dev, PCI_EXP_LNKSTA, &status);
>> -		if (!(status & PCI_EXP_LNKSTA_DLLLA))
>> +		if (pcie_is_link_active(dev))
>>   			return -ENOTTY;
> 
> Missing negation.
> 
> 
ack
>> +/**
>> + * pcie_is_link_active() - Checks if the link is active or not
>> + * @pdev: PCI device to query
>> + *
>> + * Check whether the link is active or not.
>> + *
>> + * If the config read returns error then return -ENODEV.
>> + */
>> +int pcie_is_link_active(struct pci_dev *pdev)
> 
> Why not return bool?
> 
pciehp_check_link_active is expecting int to make sure this new function
not disturbing the hotplug driver I added return type as int, I can 
change it to bool if it fine with hotplug drivers.
> I don't quite like the function name because in English the correct word
> order is subject - predicate - object, i.e. pcie_link_is_active() or
> even shorter, pcie_link_active().
> 
ack
> 
>> @@ -2094,6 +2095,10 @@ pci_alloc_irq_vectors(struct pci_dev *dev, unsigned int min_vecs,
>>   {
>>   	return -ENOSPC;
>>   }
>> +
>> +static inline int pcie_is_link_active(struct pci_dev *dev)
>> +{ return -ENODEV; }
>> +
>>   #endif /* CONFIG_PCI */
> 
> Is the empty inline really necessary?  What breaks if you leave it out?
> 
ack I will remove it.

- Krishna Chaitanya.
> Thanks,
> 
> Lukas