linux-kernel - Re: [PATCH v12 2/6] PCI: qcom: Add ICC bandwidth vote for CPU to PCIe path

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <825d76be-1e97-ee3f-55e6-baecdc1bded8@quicinc.com>
Date: Tue, 14 May 2024 15:07:55 +0530
From: Krishna Chaitanya Chundru <quic_krichai@...cinc.com>
To: Mayank Rana <quic_mrana@...cinc.com>,
        Bjorn Andersson
	<andersson@...nel.org>,
        Konrad Dybcio <konrad.dybcio@...aro.org>,
        Rob Herring
	<robh@...nel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Conor Dooley <conor+dt@...nel.org>,
        Manivannan Sadhasivam
	<manivannan.sadhasivam@...aro.org>,
        Lorenzo Pieralisi
	<lpieralisi@...nel.org>,
        Krzysztof Wilczyński
	<kw@...ux.com>,
        Bjorn Helgaas <bhelgaas@...gle.com>, <johan+linaro@...nel.org>,
        <bmasney@...hat.com>, <djakov@...nel.org>
CC: <linux-arm-msm@...r.kernel.org>, <devicetree@...r.kernel.org>,
        <linux-kernel@...r.kernel.org>, <linux-pci@...r.kernel.org>,
        <vireshk@...nel.org>, <quic_vbadigan@...cinc.com>,
        <quic_skananth@...cinc.com>, <quic_nitegupt@...cinc.com>,
        <quic_parass@...cinc.com>, <krzysztof.kozlowski@...aro.org>,
        Bryan O'Donoghue
	<bryan.odonoghue@...aro.org>
Subject: Re: [PATCH v12 2/6] PCI: qcom: Add ICC bandwidth vote for CPU to PCIe
 path



On 5/10/2024 12:04 AM, Mayank Rana wrote:
> Hi Krishna
> 
> On 4/26/2024 6:52 PM, Krishna chaitanya chundru wrote:
>> To access the host controller registers of the host controller and the
>> endpoint BAR/config space, the CPU-PCIe ICC (interconnect) path should
>> be voted otherwise it may lead to NoC (Network on chip) timeout.
>> We are surviving because of other driver voting for this path.
>>
>> As there is less access on this path compared to PCIe to mem path
>> add minimum vote i.e 1KBps bandwidth always which is sufficient enough
>> to keep the path active and is recommended by HW team.
>>
>> During S2RAM (Suspend-to-RAM), the DBI access can happen very late (while
>> disabling the boot CPU). So do not disable the CPU-PCIe interconnect path
>> during S2RAM as that may lead to NoC error.
>>
>> Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@...aro.org>
>> Signed-off-by: Krishna chaitanya chundru <quic_krichai@...cinc.com>
>> ---
>>   drivers/pci/controller/dwc/pcie-qcom.c | 44 
>> ++++++++++++++++++++++++++++++----
>>   1 file changed, 40 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/pci/controller/dwc/pcie-qcom.c 
>> b/drivers/pci/controller/dwc/pcie-qcom.c
>> index 14772edcf0d3..465d63b4be1c 100644
>> --- a/drivers/pci/controller/dwc/pcie-qcom.c
>> +++ b/drivers/pci/controller/dwc/pcie-qcom.c
>> @@ -245,6 +245,7 @@ struct qcom_pcie {
>>       struct phy *phy;
>>       struct gpio_desc *reset;
>>       struct icc_path *icc_mem;
>> +    struct icc_path *icc_cpu;
>>       const struct qcom_pcie_cfg *cfg;
>>       struct dentry *debugfs;
>>       bool suspended;
>> @@ -1409,6 +1410,9 @@ static int qcom_pcie_icc_init(struct qcom_pcie 
>> *pcie)
>>       if (IS_ERR(pcie->icc_mem))
>>           return PTR_ERR(pcie->icc_mem);
>> +    pcie->icc_cpu = devm_of_icc_get(pci->dev, "cpu-pcie");
>> +    if (IS_ERR(pcie->icc_cpu))
>> +        return PTR_ERR(pcie->icc_cpu);
>>       /*
>>        * Some Qualcomm platforms require interconnect bandwidth 
>> constraints
>>        * to be set before enabling interconnect clocks.
>> @@ -1418,7 +1422,20 @@ static int qcom_pcie_icc_init(struct qcom_pcie 
>> *pcie)
>>        */
>>       ret = icc_set_bw(pcie->icc_mem, 0, QCOM_PCIE_LINK_SPEED_TO_BW(1));
>>       if (ret) {
>> -        dev_err(pci->dev, "failed to set interconnect bandwidth: %d\n",
>> +        dev_err(pci->dev, "Failed to set bandwidth for PCIe-MEM 
>> interconnect path: %d\n",
>> +            ret);
>> +        return ret;
>> +    }
>> +
>> +    /*
>> +     * Since the CPU-PCIe path is only used for activities like register
>> +     * access of the host controller and endpoint Config/BAR space 
>> access,
>> +     * HW team has recommended to use a minimal bandwidth of 1KBps 
>> just to
>> +     * keep the path active.
>> +     */
>> +    ret = icc_set_bw(pcie->icc_cpu, 0, kBps_to_icc(1));
>> +    if (ret) {
>> +        dev_err(pci->dev, "Failed to set bandwidth for CPU-PCIe 
>> interconnect path: %d\n",
>>               ret);
> Is it needed to undo icc_mem related bus bandwidth vote here ?
I will add the logic to remove icc_mem path while returning from here,
in the next patch series.

- Krishna Chaitanya.
>>           return ret;
>>       }
>> @@ -1448,7 +1465,7 @@ static void qcom_pcie_icc_update(struct 
>> qcom_pcie *pcie)
>>       ret = icc_set_bw(pcie->icc_mem, 0, width * 
>> QCOM_PCIE_LINK_SPEED_TO_BW(speed));
>>       if (ret) {
>> -        dev_err(pci->dev, "failed to set interconnect bandwidth: %d\n",
>> +        dev_err(pci->dev, "Failed to set bandwidth for PCIe-MEM 
>> interconnect path: %d\n",
>>               ret);
>>       }
>>   }
>