Message-ID: <184a38a0-f2de-dd63-a8af-f4784c61365a@quicinc.com>
Date:   Mon, 20 Feb 2023 20:21:53 +0530
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:102.0) Gecko/20100101
 Thunderbird/102.7.0
Subject: Re: [PATCH 2/7] PCI: qcom: Add IPQ9574 PCIe support
Content-Language: en-US
To:     Devi Priya <quic_devipriy@quicinc.com>,
        Sricharan Ramabadhran <quic_srichara@quicinc.com>,
        <agross@kernel.org>, <andersson@kernel.org>,
        <konrad.dybcio@linaro.org>, <lpieralisi@kernel.org>,
        <kw@linux.com>, <robh@kernel.org>, <bhelgaas@google.com>,
        <krzysztof.kozlowski+dt@linaro.org>, <vkoul@kernel.org>,
        <kishon@kernel.org>, <mturquette@baylibre.com>, <sboyd@kernel.org>,
        <mani@kernel.org>, <p.zabel@pengutronix.de>,
        <svarbanov@mm-sol.com>, <linux-arm-msm@vger.kernel.org>,
        <linux-pci@vger.kernel.org>, <devicetree@vger.kernel.org>,
        <linux-kernel@vger.kernel.org>, <linux-phy@lists.infradead.org>,
        <linux-clk@vger.kernel.org>
CC:     <quic_gokulsri@quicinc.com>, <quic_sjaganat@quicinc.com>,
        <quic_arajkuma@quicinc.com>, <quic_anusha@quicinc.com>
References: <20230214164135.17039-1-quic_devipriy@quicinc.com>
 <20230214164135.17039-3-quic_devipriy@quicinc.com>
 <6ea43d8d-7b9c-5a11-097f-906e10ac3627@quicinc.com>
 <c766648f-c3a5-b842-2164-c3f480dee129@quicinc.com>
From:   Kathiravan T <quic_kathirav@quicinc.com>
In-Reply-To: <c766648f-c3a5-b842-2164-c3f480dee129@quicinc.com>
Content-Type: text/plain; charset="UTF-8"; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk


On 2/20/2023 7:11 PM, Devi Priya wrote:
> Hi Sri,
> Thanks for taking time to review the patch!
>
> On 2/16/2023 5:08 PM, Sricharan Ramabadhran wrote:
>> Hi Devi,
>>
>> On 2/14/2023 10:11 PM, Devi Priya wrote:
>>> Adding PCIe support for IPQ9574 SoC
>>>
>>> Co-developed-by: Anusha Rao <quic_anusha@quicinc.com>
>>> Signed-off-by: Anusha Rao <quic_anusha@quicinc.com>
>>> Signed-off-by: Devi Priya <quic_devipriy@quicinc.com>
>>> ---
>>>   drivers/pci/controller/dwc/pcie-qcom.c | 119 
>>> +++++++++++++++++++++++++
>>>   1 file changed, 119 insertions(+)
>>>
>>> diff --git a/drivers/pci/controller/dwc/pcie-qcom.c 
>>> b/drivers/pci/controller/dwc/pcie-qcom.c
>>> index a232b04af048..57606c113d45 100644
>>> --- a/drivers/pci/controller/dwc/pcie-qcom.c
>>> +++ b/drivers/pci/controller/dwc/pcie-qcom.c
>>> @@ -193,6 +193,12 @@ struct qcom_pcie_resources_2_9_0 {
>>>       struct reset_control *rst;
>>>   };
>>> +struct qcom_pcie_resources_1_27_0 {
>>> +    struct clk_bulk_data *clks;
>>> +    struct reset_control *rst;
>>> +    int num_clks;
>>> +};
>>> +
>>>   union qcom_pcie_resources {
>>>       struct qcom_pcie_resources_1_0_0 v1_0_0;
>>>       struct qcom_pcie_resources_2_1_0 v2_1_0;
>>> @@ -201,6 +207,7 @@ union qcom_pcie_resources {
>>>       struct qcom_pcie_resources_2_4_0 v2_4_0;
>>>       struct qcom_pcie_resources_2_7_0 v2_7_0;
>>>       struct qcom_pcie_resources_2_9_0 v2_9_0;
>>> +    struct qcom_pcie_resources_1_27_0 v1_27_0;
>>>   };
>>>   struct qcom_pcie;
>>> @@ -1409,6 +1416,104 @@ static int qcom_pcie_post_init_2_9_0(struct 
>>> qcom_pcie *pcie)
>>>       return 0;
>>>   }
>>> +static int qcom_pcie_get_resources_1_27_0(struct qcom_pcie *pcie)
>>> +{
>>> +    struct qcom_pcie_resources_1_27_0 *res = &pcie->res.v1_27_0;
>>> +    struct dw_pcie *pci = pcie->pci;
>>> +    struct device *dev = pci->dev;
>>> +
>>> +    res->num_clks = devm_clk_bulk_get_all(dev, &res->clks);
>>> +    if (res->clks < 0)
>>> +        return res->num_clks;
>>> +
>>> +    res->rst = devm_reset_control_array_get_exclusive(dev);
>>> +    if (IS_ERR(res->rst))
>>> +        return PTR_ERR(res->rst);
>>> +
>>> +    return 0;
>>> +}
>>> +
>>> +static void qcom_pcie_deinit_1_27_0(struct qcom_pcie *pcie)
>>> +{
>>> +    struct qcom_pcie_resources_1_27_0 *res = &pcie->res.v1_27_0;
>>> +
>>> +    clk_bulk_disable_unprepare(res->num_clks, res->clks);
>>> +}
>>> +
>>> +static int qcom_pcie_init_1_27_0(struct qcom_pcie *pcie)
>>> +{
>>> +    struct qcom_pcie_resources_1_27_0 *res = &pcie->res.v1_27_0;
>>> +    struct device *dev = pcie->pci->dev;
>>> +    int ret;
>>> +
>>> +    ret = reset_control_assert(res->rst);
>>> +    if (ret) {
>>> +        dev_err(dev, "reset assert failed (%d)\n", ret);
>>> +        return ret;
>>> +    }
>>> +
>>> +    /*
>>> +     * Delay periods before and after reset deassert are working 
>>> values
>>> +     * from downstream Codeaurora kernel
>>> +     */
>>> +    usleep_range(2000, 2500);
>>> +
>>> +    ret = reset_control_deassert(res->rst);
>>> +    if (ret) {
>>> +        dev_err(dev, "reset deassert failed (%d)\n", ret);
>>> +        return ret;
>>> +    }
>>> +
>>> +    usleep_range(2000, 2500);
>>> +
>>> +    return clk_bulk_prepare_enable(res->num_clks, res->clks);
>>> +}
>>> +
>>> +static int qcom_pcie_post_init_1_27_0(struct qcom_pcie *pcie)
>>> +{
>>> +    struct dw_pcie *pci = pcie->pci;
>>> +    u16 offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
>>> +    u32 val;
>>> +    int i;
>>> +
>>> +    writel(0x8000000, pcie->parf + 
>>> PCIE20_v3_PARF_SLV_ADDR_SPACE_SIZE);


Devi,


Above statement also differs. You need to consider this also when you 
use the 2_9_0 ops.


Thanks,


>>> +
>>> +    val = readl(pcie->parf + PCIE20_PARF_PHY_CTRL);
>>> +    val &= ~BIT(0);
>>> +    writel(val, pcie->parf + PCIE20_PARF_PHY_CTRL);
>>> +
>>> +    writel(0, pcie->parf + PCIE20_PARF_DBI_BASE_ADDR);
>>> +
>>> +    writel(DEVICE_TYPE_RC, pcie->parf + PCIE20_PARF_DEVICE_TYPE);
>>> +    writel(BYPASS | MSTR_AXI_CLK_EN | AHB_CLK_EN,
>>> +           pcie->parf + PCIE20_PARF_MHI_CLOCK_RESET_CTRL);
>>> +    writel(GEN3_RELATED_OFF_RXEQ_RGRDLESS_RXTS |
>>> +           GEN3_RELATED_OFF_GEN3_ZRXDC_NONCOMPL,
>>> +           pci->dbi_base + GEN3_RELATED_OFF);
>>> +
>>> +    writel(MST_WAKEUP_EN | SLV_WAKEUP_EN | MSTR_ACLK_CGC_DIS |
>>> +           SLV_ACLK_CGC_DIS | CORE_CLK_CGC_DIS |
>>> +           AUX_PWR_DET | L23_CLK_RMV_DIS | L1_CLK_RMV_DIS,
>>> +           pcie->parf + PCIE20_PARF_SYS_CTRL);
>>> +
>>> +    writel(0, pcie->parf + PCIE20_PARF_Q2A_FLUSH);
>>> +
>>> +    dw_pcie_dbi_ro_wr_en(pci);
>>> +    writel(PCIE_CAP_SLOT_VAL, pci->dbi_base + offset + 
>>> PCI_EXP_SLTCAP);
>>> +
>>> +    val = readl(pci->dbi_base + offset + PCI_EXP_LNKCAP);
>>> +    val &= ~PCI_EXP_LNKCAP_ASPMS;
>>> +    writel(val, pci->dbi_base + offset + PCI_EXP_LNKCAP);
>>> +
>>> +    writel(PCI_EXP_DEVCTL2_COMP_TMOUT_DIS, pci->dbi_base + offset +
>>> +           PCI_EXP_DEVCTL2);
>>> +
>>> +    for (i = 0; i < 256; i++)
>>> +        writel(0, pcie->parf + PCIE20_PARF_BDF_TO_SID_TABLE_N + (4 
>>> * i));
>>> +
>>> +    return 0;
>>> +}
>>> +
>>>   static int qcom_pcie_link_up(struct dw_pcie *pci)
>>>   {
>>>       u16 offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
>>> @@ -1620,6 +1725,15 @@ static const struct qcom_pcie_ops ops_2_9_0 = {
>>>       .ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
>>>   };
>>> +/* Qcom IP rev.: 1.27.0 Synopsys IP rev.: 5.80a */
>>> +static const struct qcom_pcie_ops ops_1_27_0 = {
>>> +    .get_resources = qcom_pcie_get_resources_1_27_0,
>>> +    .init = qcom_pcie_init_1_27_0,
>>> +    .post_init = qcom_pcie_post_init_1_27_0,
>>> +    .deinit = qcom_pcie_deinit_1_27_0,
>>> +    .ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
>>> +};
>>> +
>>>   static const struct qcom_pcie_cfg cfg_1_0_0 = {
>>>       .ops = &ops_1_0_0,
>>>   };
>>> @@ -1652,6 +1766,10 @@ static const struct qcom_pcie_cfg cfg_2_9_0 = {
>>>       .ops = &ops_2_9_0,
>>>   };
>>> +static const struct qcom_pcie_cfg cfg_1_27_0 = {
>>> +    .ops = &ops_1_27_0,
>>> +};
>>> +
>>>   static const struct dw_pcie_ops dw_pcie_ops = {
>>>       .link_up = qcom_pcie_link_up,
>>>       .start_link = qcom_pcie_start_link,
>>> @@ -1829,6 +1947,7 @@ static const struct of_device_id 
>>> qcom_pcie_match[] = {
>>>       { .compatible = "qcom,pcie-ipq8064-v2", .data = &cfg_2_1_0 },
>>>       { .compatible = "qcom,pcie-ipq8074", .data = &cfg_2_3_3 },
>>>       { .compatible = "qcom,pcie-ipq8074-gen3", .data = &cfg_2_9_0 },
>>> +    { .compatible = "qcom,pcie-ipq9574", .data = &cfg_1_27_0 },
>>
>>    I do not see much difference between 2_9_0 and 1_27_0. Is this patch
>>    really required. Can you check if it works with 2_9_0 itself ?
> Yes right Sri, Only the clocks seem to differ between 2_9_0 and 1_27_0.
> Will update 2_9_0 ops to get the clocks from the DT and use the same 
> for ipq9574 in the next spin.
>
> Best Regards,
> Devi Priya
>>
>> Regards,
>>   Sricharan