* [PATCH 0/5] arm64: IPQ6018 PCIe support
@ 2021-04-20 11:21 Baruch Siach
2021-04-20 11:21 ` [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller Baruch Siach
` (4 more replies)
0 siblings, 5 replies; 11+ messages in thread
From: Baruch Siach @ 2021-04-20 11:21 UTC (permalink / raw)
To: Andy Gross, Bjorn Andersson
Cc: Baruch Siach, Selvam Sathappan Periakaruppan, Kathiravan T,
Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul, Rob Herring,
devicetree, linux-phy, linux-pci, linux-arm-msm,
linux-arm-kernel
This series adds support for the single PCIe lane on IPQ6018 SoCs. The code is
ported from downstream Codeaurora v5.4 kernel. The main difference from
downstream code is the split of PCIe registers configuration from .init to
.post_init, since it requires phy_power_on().
Tested on IPQ6010 based hardware.
Baruch Siach (2):
dt-bindings: phy: qcom,qmp: Add IPQ60xx PCIe PHY bindings
dt-bindings: pci: qcom: Document PCIe bindings for IPQ6018 SoC
Selvam Sathappan Periakaruppan (3):
PCI: qcom: add support for IPQ60xx PCIe controller
phy: qcom-qmp: add QMP V2 PCIe PHY support for ipq60xx
arm64: dts: ipq6018: Add pcie support
.../devicetree/bindings/pci/qcom,pcie.txt | 24 ++
.../devicetree/bindings/phy/qcom,qmp-phy.yaml | 25 ++
arch/arm64/boot/dts/qcom/ipq6018.dtsi | 109 +++++++
drivers/pci/controller/dwc/pcie-qcom.c | 279 ++++++++++++++++++
drivers/phy/qualcomm/phy-qcom-qmp.c | 147 +++++++++
drivers/phy/qualcomm/phy-qcom-qmp.h | 132 +++++++++
6 files changed, 716 insertions(+)
--
2.30.2
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller
2021-04-20 11:21 [PATCH 0/5] arm64: IPQ6018 PCIe support Baruch Siach
@ 2021-04-20 11:21 ` Baruch Siach
2021-04-20 16:18 ` Rob Herring
2021-04-20 11:21 ` [PATCH 2/5] phy: qcom-qmp: add QMP V2 PCIe PHY support for ipq60xx Baruch Siach
` (3 subsequent siblings)
4 siblings, 1 reply; 11+ messages in thread
From: Baruch Siach @ 2021-04-20 11:21 UTC (permalink / raw)
To: Andy Gross, Bjorn Andersson
Cc: Selvam Sathappan Periakaruppan, Baruch Siach, Kathiravan T,
Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul, Rob Herring,
devicetree, linux-phy, linux-pci, linux-arm-msm,
linux-arm-kernel
From: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
IPQ60xx series of SoCs have one port of PCIe gen 3. Add support for that
platform.
The code is based on downstream Codeaurora kernel v5.4. Split out the
registers access part from .init into .post_init. Registers are only
accessible after phy_power_on().
Signed-off-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
---
drivers/pci/controller/dwc/pcie-qcom.c | 279 +++++++++++++++++++++++++
1 file changed, 279 insertions(+)
diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
index 8a7a300163e5..3e27de744738 100644
--- a/drivers/pci/controller/dwc/pcie-qcom.c
+++ b/drivers/pci/controller/dwc/pcie-qcom.c
@@ -41,6 +41,31 @@
#define L23_CLK_RMV_DIS BIT(2)
#define L1_CLK_RMV_DIS BIT(1)
+#define PCIE_ATU_CR1_OUTBOUND_6_GEN3 0xC00
+#define PCIE_ATU_CR2_OUTBOUND_6_GEN3 0xC04
+#define PCIE_ATU_LOWER_BASE_OUTBOUND_6_GEN3 0xC08
+#define PCIE_ATU_UPPER_BASE_OUTBOUND_6_GEN3 0xC0C
+#define PCIE_ATU_LIMIT_OUTBOUND_6_GEN3 0xC10
+#define PCIE_ATU_LOWER_TARGET_OUTBOUND_6_GEN3 0xC14
+#define PCIE_ATU_UPPER_TARGET_OUTBOUND_6_GEN3 0xC18
+
+#define PCIE_ATU_CR1_OUTBOUND_7_GEN3 0xE00
+#define PCIE_ATU_CR2_OUTBOUND_7_GEN3 0xE04
+#define PCIE_ATU_LOWER_BASE_OUTBOUND_7_GEN3 0xE08
+#define PCIE_ATU_UPPER_BASE_OUTBOUND_7_GEN3 0xE0C
+#define PCIE_ATU_LIMIT_OUTBOUND_7_GEN3 0xE10
+#define PCIE_ATU_LOWER_TARGET_OUTBOUND_7_GEN3 0xE14
+#define PCIE_ATU_UPPER_TARGET_OUTBOUND_7_GEN3 0xE18
+
+#define PCIE20_COMMAND_STATUS 0x04
+#define BUS_MASTER_EN 0x7
+#define PCIE20_DEVICE_CONTROL2_STATUS2 0x98
+#define PCIE_CAP_CPL_TIMEOUT_DISABLE 0x10
+#define PCIE30_GEN3_RELATED_OFF 0x890
+
+#define RXEQ_RGRDLESS_RXTS BIT(13)
+#define GEN3_ZRXDC_NONCOMPL BIT(0)
+
#define PCIE20_PARF_PHY_CTRL 0x40
#define PHY_CTRL_PHY_TX0_TERM_OFFSET_MASK GENMASK(20, 16)
#define PHY_CTRL_PHY_TX0_TERM_OFFSET(x) ((x) << 16)
@@ -52,6 +77,10 @@
#define PCIE20_PARF_DBI_BASE_ADDR 0x168
#define PCIE20_PARF_SLV_ADDR_SPACE_SIZE 0x16C
#define PCIE20_PARF_MHI_CLOCK_RESET_CTRL 0x174
+#define AHB_CLK_EN BIT(0)
+#define MSTR_AXI_CLK_EN BIT(1)
+#define BYPASS BIT(4)
+
#define PCIE20_PARF_AXI_MSTR_WR_ADDR_HALT 0x178
#define PCIE20_PARF_AXI_MSTR_WR_ADDR_HALT_V2 0x1A8
#define PCIE20_PARF_LTSSM 0x1B0
@@ -94,6 +123,12 @@
#define SLV_ADDR_SPACE_SZ 0x10000000
#define PCIE20_LNK_CONTROL2_LINK_STATUS2 0xa0
+#define PCIE_CAP_CURR_DEEMPHASIS BIT(16)
+#define SPEED_GEN1 0x1
+#define SPEED_GEN2 0x2
+#define SPEED_GEN3 0x3
+#define AXI_CLK_RATE 200000000
+#define RCHNG_CLK_RATE 100000000
#define DEVICE_TYPE_RC 0x4
@@ -168,6 +203,15 @@ struct qcom_pcie_resources_2_7_0 {
struct clk *pipe_clk;
};
+struct qcom_pcie_resources_2_9_0 {
+ struct clk *iface;
+ struct clk *axi_m_clk;
+ struct clk *axi_s_clk;
+ struct clk *axi_bridge_clk;
+ struct clk *rchng_clk;
+ struct reset_control *rst[8];
+};
+
union qcom_pcie_resources {
struct qcom_pcie_resources_1_0_0 v1_0_0;
struct qcom_pcie_resources_2_1_0 v2_1_0;
@@ -175,6 +219,7 @@ union qcom_pcie_resources {
struct qcom_pcie_resources_2_3_3 v2_3_3;
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;
@@ -1266,6 +1311,225 @@ static void qcom_pcie_post_deinit_2_7_0(struct qcom_pcie *pcie)
clk_disable_unprepare(res->pipe_clk);
}
+static int qcom_pcie_get_resources_2_9_0(struct qcom_pcie *pcie)
+{
+ struct qcom_pcie_resources_2_9_0 *res = &pcie->res.v2_9_0;
+ struct dw_pcie *pci = pcie->pci;
+ struct device *dev = pci->dev;
+ int i;
+ const char *rst_names[] = { "pipe", "sleep", "sticky",
+ "axi_m", "axi_s", "ahb",
+ "axi_m_sticky", "axi_s_sticky", };
+
+ res->iface = devm_clk_get(dev, "iface");
+ if (IS_ERR(res->iface))
+ return PTR_ERR(res->iface);
+
+ res->axi_m_clk = devm_clk_get(dev, "axi_m");
+ if (IS_ERR(res->axi_m_clk))
+ return PTR_ERR(res->axi_m_clk);
+
+ res->axi_s_clk = devm_clk_get(dev, "axi_s");
+ if (IS_ERR(res->axi_s_clk))
+ return PTR_ERR(res->axi_s_clk);
+
+ res->axi_bridge_clk = devm_clk_get(dev, "axi_bridge");
+ if (IS_ERR(res->axi_bridge_clk))
+ return PTR_ERR(res->axi_bridge_clk);
+
+ res->rchng_clk = devm_clk_get(dev, "rchng");
+ if (IS_ERR(res->rchng_clk))
+ return PTR_ERR(res->rchng_clk);
+
+ for (i = 0; i < ARRAY_SIZE(rst_names); i++) {
+ res->rst[i] = devm_reset_control_get(dev, rst_names[i]);
+ if (IS_ERR(res->rst[i])) {
+ return PTR_ERR(res->rst[i]);
+ }
+ }
+
+ return 0;
+}
+
+static void qcom_pcie_deinit_2_9_0(struct qcom_pcie *pcie)
+{
+ struct qcom_pcie_resources_2_9_0 *res = &pcie->res.v2_9_0;
+
+ clk_disable_unprepare(res->axi_m_clk);
+ clk_disable_unprepare(res->axi_s_clk);
+ clk_disable_unprepare(res->axi_bridge_clk);
+ clk_disable_unprepare(res->rchng_clk);
+ clk_disable_unprepare(res->iface);
+}
+
+static int qcom_pcie_init_2_9_0(struct qcom_pcie *pcie)
+{
+ struct qcom_pcie_resources_2_9_0 *res = &pcie->res.v2_9_0;
+ struct device *dev = pcie->pci->dev;
+ int i, ret;
+
+ for (i = 0; i < ARRAY_SIZE(res->rst); i++) {
+ ret = reset_control_assert(res->rst[i]);
+ if (ret) {
+ dev_err(dev, "reset #%d assert failed (%d)\n", i, ret);
+ return ret;
+ }
+ }
+
+ usleep_range(2000, 2500);
+
+ for (i = 0; i < ARRAY_SIZE(res->rst); i++) {
+ ret = reset_control_deassert(res->rst[i]);
+ if (ret) {
+ dev_err(dev, "reset #%d deassert failed (%d)\n", i,
+ ret);
+ return ret;
+ }
+ }
+
+ /*
+ * Don't have a way to see if the reset has completed.
+ * Wait for some time.
+ */
+ usleep_range(2000, 2500);
+
+ ret = clk_prepare_enable(res->iface);
+ if (ret) {
+ dev_err(dev, "cannot prepare/enable core clock\n");
+ goto err_clk_iface;
+ }
+
+ ret = clk_prepare_enable(res->axi_m_clk);
+ if (ret) {
+ dev_err(dev, "cannot prepare/enable core clock\n");
+ goto err_clk_axi_m;
+ }
+
+ ret = clk_set_rate(res->axi_m_clk, AXI_CLK_RATE);
+ if (ret) {
+ dev_err(dev, "MClk rate set failed (%d)\n", ret);
+ goto err_clk_axi_m;
+ }
+
+ ret = clk_prepare_enable(res->axi_s_clk);
+ if (ret) {
+ dev_err(dev, "cannot prepare/enable axi slave clock\n");
+ goto err_clk_axi_s;
+ }
+
+ ret = clk_set_rate(res->axi_s_clk, AXI_CLK_RATE);
+ if (ret) {
+ dev_err(dev, "SClk rate set failed (%d)\n", ret);
+ goto err_clk_axi_s;
+ }
+
+ ret = clk_prepare_enable(res->axi_bridge_clk);
+ if (ret) {
+ dev_err(dev, "cannot prepare/enable axi bridge clock\n");
+ goto err_clk_axi_bridge;
+ }
+
+ ret = clk_prepare_enable(res->rchng_clk);
+ if (ret) {
+ dev_err(dev, "cannot prepare/enable rchng clock\n");
+ goto err_clk_rchng;
+ }
+
+ ret = clk_set_rate(res->rchng_clk, RCHNG_CLK_RATE);
+ if (ret) {
+ dev_err(dev, "rchng_clk rate set failed (%d)\n", ret);
+ goto err_clk_rchng;
+ }
+
+ return 0;
+
+err_clk_rchng:
+ clk_disable_unprepare(res->rchng_clk);
+err_clk_axi_bridge:
+ clk_disable_unprepare(res->axi_bridge_clk);
+err_clk_axi_s:
+ clk_disable_unprepare(res->axi_s_clk);
+err_clk_axi_m:
+ clk_disable_unprepare(res->axi_m_clk);
+err_clk_iface:
+ clk_disable_unprepare(res->iface);
+ /*
+ * Not checking for failure, will anyway return
+ * the original failure in 'ret'.
+ */
+ for (i = 0; i < ARRAY_SIZE(res->rst); i++)
+ reset_control_assert(res->rst[i]);
+
+ return ret;
+}
+
+static int qcom_pcie_post_init_2_9_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(SLV_ADDR_SPACE_SZ,
+ pcie->parf + PCIE20_v3_PARF_SLV_ADDR_SPACE_SIZE);
+
+ 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(RXEQ_RGRDLESS_RXTS | GEN3_ZRXDC_NONCOMPL,
+ pci->dbi_base + PCIE30_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);
+
+ writel(BUS_MASTER_EN, pci->dbi_base + PCIE20_COMMAND_STATUS);
+
+ writel(DBI_RO_WR_EN, pci->dbi_base + PCIE20_MISC_CONTROL_1_REG);
+ writel(PCIE_CAP_LINK1_VAL, pci->dbi_base + offset + PCI_EXP_SLTCAP);
+
+ /* Configure PCIe link capabilities for ASPM */
+ val = readl(pci->dbi_base + offset + PCI_EXP_LNKCAP);
+ val &= ~PCI_EXP_LNKCAP_ASPMS;
+ writel(val, pci->dbi_base + offset + PCI_EXP_LNKCAP);
+
+ writel(PCIE_CAP_CPL_TIMEOUT_DISABLE, pci->dbi_base +
+ PCIE20_DEVICE_CONTROL2_STATUS2);
+
+ writel(PCIE_CAP_CURR_DEEMPHASIS | SPEED_GEN3,
+ pci->dbi_base + offset + PCI_EXP_DEVCTL2);
+
+ for (i = 0;i < 256;i++)
+ writel(0x0, pcie->parf + PCIE20_PARF_BDF_TO_SID_TABLE_N
+ + (4 * i));
+
+ writel(0x4, pci->atu_base + PCIE_ATU_CR1_OUTBOUND_6_GEN3);
+ writel(0x90000000, pci->atu_base + PCIE_ATU_CR2_OUTBOUND_6_GEN3);
+ writel(0x0, pci->atu_base + PCIE_ATU_LOWER_BASE_OUTBOUND_6_GEN3);
+ writel(0x0, pci->atu_base + PCIE_ATU_UPPER_BASE_OUTBOUND_6_GEN3);
+ writel(0x00107FFFF, pci->atu_base + PCIE_ATU_LIMIT_OUTBOUND_6_GEN3);
+ writel(0x0, pci->atu_base + PCIE_ATU_LOWER_TARGET_OUTBOUND_6_GEN3);
+ writel(0x0, pci->atu_base + PCIE_ATU_UPPER_TARGET_OUTBOUND_6_GEN3);
+ writel(0x5, pci->atu_base + PCIE_ATU_CR1_OUTBOUND_7_GEN3);
+ writel(0x90000000, pci->atu_base + PCIE_ATU_CR2_OUTBOUND_7_GEN3);
+ writel(0x200000, pci->atu_base + PCIE_ATU_LOWER_BASE_OUTBOUND_7_GEN3);
+ writel(0x0, pci->atu_base + PCIE_ATU_UPPER_BASE_OUTBOUND_7_GEN3);
+ writel(0x7FFFFF, pci->atu_base + PCIE_ATU_LIMIT_OUTBOUND_7_GEN3);
+ writel(0x0, pci->atu_base + PCIE_ATU_LOWER_TARGET_OUTBOUND_7_GEN3);
+ writel(0x0, pci->atu_base + PCIE_ATU_UPPER_TARGET_OUTBOUND_7_GEN3);
+
+ return 0;
+}
+
static int qcom_pcie_link_up(struct dw_pcie *pci)
{
u16 offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
@@ -1456,6 +1720,15 @@ static const struct qcom_pcie_ops ops_1_9_0 = {
.config_sid = qcom_pcie_config_sid_sm8250,
};
+/* Qcom IP rev.: 2.9.0 Synopsys IP rev.: 5.00a */
+static const struct qcom_pcie_ops ops_2_9_0 = {
+ .get_resources = qcom_pcie_get_resources_2_9_0,
+ .init = qcom_pcie_init_2_9_0,
+ .post_init = qcom_pcie_post_init_2_9_0,
+ .deinit = qcom_pcie_deinit_2_9_0,
+ .ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
+};
+
static const struct dw_pcie_ops dw_pcie_ops = {
.link_up = qcom_pcie_link_up,
.start_link = qcom_pcie_start_link,
@@ -1508,6 +1781,11 @@ static int qcom_pcie_probe(struct platform_device *pdev)
goto err_pm_runtime_put;
}
+ /* We need ATU for .post_init */
+ pci->atu_base = devm_platform_ioremap_resource_byname(pdev, "atu");
+ if (IS_ERR(pci->atu_base))
+ pci->atu_base = NULL;
+
pcie->phy = devm_phy_optional_get(dev, "pciephy");
if (IS_ERR(pcie->phy)) {
ret = PTR_ERR(pcie->phy);
@@ -1555,6 +1833,7 @@ static const struct of_device_id qcom_pcie_match[] = {
{ .compatible = "qcom,pcie-qcs404", .data = &ops_2_4_0 },
{ .compatible = "qcom,pcie-sdm845", .data = &ops_2_7_0 },
{ .compatible = "qcom,pcie-sm8250", .data = &ops_1_9_0 },
+ { .compatible = "qcom,pcie-ipq6018", .data = &ops_2_9_0 },
{ }
};
--
2.30.2
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH 2/5] phy: qcom-qmp: add QMP V2 PCIe PHY support for ipq60xx
2021-04-20 11:21 [PATCH 0/5] arm64: IPQ6018 PCIe support Baruch Siach
2021-04-20 11:21 ` [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller Baruch Siach
@ 2021-04-20 11:21 ` Baruch Siach
2021-04-20 11:21 ` [PATCH 3/5] arm64: dts: ipq6018: Add pcie support Baruch Siach
` (2 subsequent siblings)
4 siblings, 0 replies; 11+ messages in thread
From: Baruch Siach @ 2021-04-20 11:21 UTC (permalink / raw)
To: Andy Gross, Bjorn Andersson
Cc: Selvam Sathappan Periakaruppan, Baruch Siach, Kathiravan T,
Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul, Rob Herring,
devicetree, linux-phy, linux-pci, linux-arm-msm,
linux-arm-kernel
From: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
Based on code from downstream Codeaurora tree. The ipq60xx has one gen3
PCIe port.
Signed-off-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
---
drivers/phy/qualcomm/phy-qcom-qmp.c | 147 ++++++++++++++++++++++++++++
drivers/phy/qualcomm/phy-qcom-qmp.h | 132 +++++++++++++++++++++++++
2 files changed, 279 insertions(+)
diff --git a/drivers/phy/qualcomm/phy-qcom-qmp.c b/drivers/phy/qualcomm/phy-qcom-qmp.c
index 9cdebe7f26cb..9f3148136789 100644
--- a/drivers/phy/qualcomm/phy-qcom-qmp.c
+++ b/drivers/phy/qualcomm/phy-qcom-qmp.c
@@ -143,6 +143,13 @@ static const unsigned int msm8996_ufsphy_regs_layout[QPHY_LAYOUT_SIZE] = {
[QPHY_PCS_READY_STATUS] = 0x168,
};
+static const unsigned int ipq_pciephy_gen3_regs_layout[QPHY_LAYOUT_SIZE] = {
+ [QPHY_SW_RESET] = 0x00,
+ [QPHY_START_CTRL] = 0x44,
+ [QPHY_PCS_STATUS] = 0x14,
+ [QPHY_PCS_POWER_DOWN_CONTROL] = 0x40,
+};
+
static const unsigned int pciephy_regs_layout[QPHY_LAYOUT_SIZE] = {
[QPHY_COM_SW_RESET] = 0x400,
[QPHY_COM_POWER_DOWN_CONTROL] = 0x404,
@@ -614,6 +621,113 @@ static const struct qmp_phy_init_tbl msm8996_usb3_pcs_tbl[] = {
QMP_PHY_INIT_CFG(QPHY_POWER_STATE_CONFIG2, 0x08),
};
+static const struct qmp_phy_init_tbl ipq6018_pcie_serdes_tbl[] = {
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SSC_PER1, 0x7d),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SSC_PER2, 0x01),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SSC_STEP_SIZE1_MODE0, 0x0a),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SSC_STEP_SIZE2_MODE0, 0x05),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SSC_STEP_SIZE1_MODE1, 0x08),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SSC_STEP_SIZE2_MODE1, 0x04),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_BIAS_EN_CLKBUFLR_EN, 0x18),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_CLK_ENABLE1, 0x90),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SYS_CLK_CTRL, 0x02),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SYSCLK_BUF_ENABLE, 0x07),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_PLL_IVCO, 0x0f),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_LOCK_CMP1_MODE0, 0xd4),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_LOCK_CMP2_MODE0, 0x14),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_LOCK_CMP1_MODE1, 0xaa),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_LOCK_CMP2_MODE1, 0x29),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_BG_TRIM, 0x0f),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_CP_CTRL_MODE0, 0x09),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_CP_CTRL_MODE1, 0x09),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_PLL_RCTRL_MODE0, 0x16),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_PLL_RCTRL_MODE1, 0x16),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_PLL_CCTRL_MODE0, 0x28),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_PLL_CCTRL_MODE1, 0x28),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_BIAS_EN_CTRL_BY_PSM, 0x01),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SYSCLK_EN_SEL, 0x08),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_RESETSM_CNTRL, 0x20),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_LOCK_CMP_EN, 0x42),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_DEC_START_MODE0, 0x68),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_DEC_START_MODE1, 0x53),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_DIV_FRAC_START1_MODE0, 0xab),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_DIV_FRAC_START2_MODE0, 0xaa),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_DIV_FRAC_START3_MODE0, 0x02),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_DIV_FRAC_START1_MODE1, 0x55),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_DIV_FRAC_START2_MODE1, 0x55),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_DIV_FRAC_START3_MODE1, 0x05),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_INTEGLOOP_GAIN0_MODE0, 0xa0),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_INTEGLOOP_GAIN0_MODE1, 0xa0),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_VCO_TUNE1_MODE0, 0x24),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_VCO_TUNE2_MODE0, 0x02),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_VCO_TUNE1_MODE1, 0xb4),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_VCO_TUNE2_MODE1, 0x03),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_CLK_SELECT, 0x32),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_HSCLK_SEL, 0x01),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_CORE_CLK_EN, 0x00),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_CMN_CONFIG, 0x06),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_SVS_MODE_CLK_SEL, 0x05),
+ QMP_PHY_INIT_CFG(QSERDES_PLL_CORECLK_DIV_MODE1, 0x08),
+};
+
+static const struct qmp_phy_init_tbl ipq6018_pcie_tx_tbl[] = {
+ QMP_PHY_INIT_CFG(QSERDES_TX0_RES_CODE_LANE_OFFSET_TX, 0x02),
+ QMP_PHY_INIT_CFG(QSERDES_TX0_LANE_MODE_1, 0x06),
+ QMP_PHY_INIT_CFG(QSERDES_TX0_RCV_DETECT_LVL_2, 0x12),
+};
+
+static const struct qmp_phy_init_tbl ipq6018_pcie_rx_tbl[] = {
+ QMP_PHY_INIT_CFG(QSERDES_RX0_UCDR_FO_GAIN, 0x0c),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_UCDR_SO_GAIN, 0x02),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_UCDR_SO_SATURATION_AND_ENABLE, 0x7f),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_UCDR_PI_CONTROLS, 0x70),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_EQU_ADAPTOR_CNTRL2, 0x61),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_EQU_ADAPTOR_CNTRL3, 0x04),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_EQU_ADAPTOR_CNTRL4, 0x1e),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_IDAC_TSETTLE_LOW, 0xc0),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_IDAC_TSETTLE_HIGH, 0x00),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_EQ_OFFSET_ADAPTOR_CNTRL1, 0x73),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_OFFSET_ADAPTOR_CNTRL2, 0x80),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_SIGDET_ENABLES, 0x1c),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_SIGDET_CNTRL, 0x03),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_SIGDET_DEGLITCH_CNTRL, 0x14),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_00_LOW, 0xf0),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_00_HIGH, 0x01),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_00_HIGH2, 0x2f),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_00_HIGH3, 0xd3),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_00_HIGH4, 0x40),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_01_LOW, 0x01),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_01_HIGH, 0x02),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_01_HIGH2, 0xc8),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_01_HIGH3, 0x09),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_01_HIGH4, 0xb1),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_10_LOW, 0x00),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_10_HIGH, 0x02),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_10_HIGH2, 0xc8),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_10_HIGH3, 0x09),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_RX_MODE_10_HIGH4, 0xb1),
+ QMP_PHY_INIT_CFG(QSERDES_RX0_DFE_EN_TIMER, 0x04),
+};
+
+static const struct qmp_phy_init_tbl ipq6018_pcie_pcs_tbl[] = {
+ QMP_PHY_INIT_CFG(PCS_COM_FLL_CNTRL1, 0x01),
+ QMP_PHY_INIT_CFG(PCS_COM_REFGEN_REQ_CONFIG1, 0x0d),
+ QMP_PHY_INIT_CFG(PCS_COM_G12S1_TXDEEMPH_M3P5DB, 0x10),
+ QMP_PHY_INIT_CFG(PCS_COM_RX_SIGDET_LVL, 0xaa),
+ QMP_PHY_INIT_CFG(PCS_COM_P2U3_WAKEUP_DLY_TIME_AUXCLK_L, 0x01),
+ QMP_PHY_INIT_CFG(PCS_COM_RX_DCC_CAL_CONFIG, 0x01),
+ QMP_PHY_INIT_CFG(PCS_COM_EQ_CONFIG5, 0x01),
+ QMP_PHY_INIT_CFG(PCS_PCIE_POWER_STATE_CONFIG2, 0x0d),
+ QMP_PHY_INIT_CFG(PCS_PCIE_POWER_STATE_CONFIG4, 0x07),
+ QMP_PHY_INIT_CFG(PCS_PCIE_ENDPOINT_REFCLK_DRIVE, 0xc1),
+ QMP_PHY_INIT_CFG(PCS_PCIE_L1P1_WAKEUP_DLY_TIME_AUXCLK_L, 0x01),
+ QMP_PHY_INIT_CFG(PCS_PCIE_L1P2_WAKEUP_DLY_TIME_AUXCLK_L, 0x01),
+ QMP_PHY_INIT_CFG(PCS_PCIE_OSC_DTCT_ACTIONS, 0x00),
+ QMP_PHY_INIT_CFG(PCS_PCIE_EQ_CONFIG1, 0x11),
+ QMP_PHY_INIT_CFG(PCS_PCIE_PRESET_P10_PRE, 0x00),
+ QMP_PHY_INIT_CFG(PCS_PCIE_PRESET_P10_POST, 0x58),
+};
+
static const struct qmp_phy_init_tbl ipq8074_pcie_serdes_tbl[] = {
QMP_PHY_INIT_CFG(QSERDES_COM_BIAS_EN_CLKBUFLR_EN, 0x18),
QMP_PHY_INIT_CFG(QSERDES_COM_CLK_ENABLE1, 0x10),
@@ -2646,6 +2760,36 @@ static const struct qmp_phy_cfg ipq8074_pciephy_cfg = {
.pwrdn_delay_max = 1005, /* us */
};
+static const struct qmp_phy_cfg ipq6018_pciephy_cfg = {
+ .type = PHY_TYPE_PCIE,
+ .nlanes = 1,
+
+ .serdes_tbl = ipq6018_pcie_serdes_tbl,
+ .serdes_tbl_num = ARRAY_SIZE(ipq6018_pcie_serdes_tbl),
+ .tx_tbl = ipq6018_pcie_tx_tbl,
+ .tx_tbl_num = ARRAY_SIZE(ipq6018_pcie_tx_tbl),
+ .rx_tbl = ipq6018_pcie_rx_tbl,
+ .rx_tbl_num = ARRAY_SIZE(ipq6018_pcie_rx_tbl),
+ .pcs_tbl = ipq6018_pcie_pcs_tbl,
+ .pcs_tbl_num = ARRAY_SIZE(ipq6018_pcie_pcs_tbl),
+ .clk_list = ipq8074_pciephy_clk_l,
+ .num_clks = ARRAY_SIZE(ipq8074_pciephy_clk_l),
+ .reset_list = ipq8074_pciephy_reset_l,
+ .num_resets = ARRAY_SIZE(ipq8074_pciephy_reset_l),
+ .vreg_list = NULL,
+ .num_vregs = 0,
+ .regs = ipq_pciephy_gen3_regs_layout,
+
+ .start_ctrl = SERDES_START | PCS_START,
+ .pwrdn_ctrl = SW_PWRDN | REFCLK_DRV_DSBL,
+
+ .has_phy_com_ctrl = false,
+ .has_lane_rst = false,
+ .has_pwrdn_delay = true,
+ .pwrdn_delay_min = 995, /* us */
+ .pwrdn_delay_max = 1005, /* us */
+};
+
static const struct qmp_phy_cfg sdm845_qmp_pciephy_cfg = {
.type = PHY_TYPE_PCIE,
.nlanes = 1,
@@ -4532,6 +4676,9 @@ static const struct of_device_id qcom_qmp_phy_of_match_table[] = {
}, {
.compatible = "qcom,ipq8074-qmp-pcie-phy",
.data = &ipq8074_pciephy_cfg,
+ }, {
+ .compatible = "qcom,ipq6018-qmp-pcie-phy",
+ .data = &ipq6018_pciephy_cfg,
}, {
.compatible = "qcom,sc7180-qmp-usb3-phy",
.data = &sc7180_usb3phy_cfg,
diff --git a/drivers/phy/qualcomm/phy-qcom-qmp.h b/drivers/phy/qualcomm/phy-qcom-qmp.h
index 71ce3aa174ae..bd8f9637b4d0 100644
--- a/drivers/phy/qualcomm/phy-qcom-qmp.h
+++ b/drivers/phy/qualcomm/phy-qcom-qmp.h
@@ -6,6 +6,138 @@
#ifndef QCOM_PHY_QMP_H_
#define QCOM_PHY_QMP_H_
+/* QMP V2 PHY for PCIE gen3 ports - QSERDES PLL registers */
+
+#define QSERDES_PLL_BG_TIMER 0x00c
+#define QSERDES_PLL_SSC_PER1 0x01c
+#define QSERDES_PLL_SSC_PER2 0x020
+#define QSERDES_PLL_SSC_STEP_SIZE1_MODE0 0x024
+#define QSERDES_PLL_SSC_STEP_SIZE2_MODE0 0x028
+#define QSERDES_PLL_SSC_STEP_SIZE1_MODE1 0x02c
+#define QSERDES_PLL_SSC_STEP_SIZE2_MODE1 0x030
+#define QSERDES_PLL_BIAS_EN_CLKBUFLR_EN 0x03c
+#define QSERDES_PLL_CLK_ENABLE1 0x040
+#define QSERDES_PLL_SYS_CLK_CTRL 0x044
+#define QSERDES_PLL_SYSCLK_BUF_ENABLE 0x048
+#define QSERDES_PLL_PLL_IVCO 0x050
+#define QSERDES_PLL_LOCK_CMP1_MODE0 0x054
+#define QSERDES_PLL_LOCK_CMP2_MODE0 0x058
+#define QSERDES_PLL_LOCK_CMP1_MODE1 0x060
+#define QSERDES_PLL_LOCK_CMP2_MODE1 0x064
+#define QSERDES_PLL_BG_TRIM 0x074
+#define QSERDES_PLL_CLK_EP_DIV_MODE0 0x078
+#define QSERDES_PLL_CLK_EP_DIV_MODE1 0x07c
+#define QSERDES_PLL_CP_CTRL_MODE0 0x080
+#define QSERDES_PLL_CP_CTRL_MODE1 0x084
+#define QSERDES_PLL_PLL_RCTRL_MODE0 0x088
+#define QSERDES_PLL_PLL_RCTRL_MODE1 0x08C
+#define QSERDES_PLL_PLL_CCTRL_MODE0 0x090
+#define QSERDES_PLL_PLL_CCTRL_MODE1 0x094
+#define QSERDES_PLL_BIAS_EN_CTRL_BY_PSM 0x0a4
+#define QSERDES_PLL_SYSCLK_EN_SEL 0x0a8
+#define QSERDES_PLL_RESETSM_CNTRL 0x0b0
+#define QSERDES_PLL_LOCK_CMP_EN 0x0c4
+#define QSERDES_PLL_DEC_START_MODE0 0x0cc
+#define QSERDES_PLL_DEC_START_MODE1 0x0d0
+#define QSERDES_PLL_DIV_FRAC_START1_MODE0 0x0d8
+#define QSERDES_PLL_DIV_FRAC_START2_MODE0 0x0dc
+#define QSERDES_PLL_DIV_FRAC_START3_MODE0 0x0e0
+#define QSERDES_PLL_DIV_FRAC_START1_MODE1 0x0e4
+#define QSERDES_PLL_DIV_FRAC_START2_MODE1 0x0e8
+#define QSERDES_PLL_DIV_FRAC_START3_MODE1 0x0eC
+#define QSERDES_PLL_INTEGLOOP_GAIN0_MODE0 0x100
+#define QSERDES_PLL_INTEGLOOP_GAIN1_MODE0 0x104
+#define QSERDES_PLL_INTEGLOOP_GAIN0_MODE1 0x108
+#define QSERDES_PLL_INTEGLOOP_GAIN1_MODE1 0x10c
+#define QSERDES_PLL_VCO_TUNE_MAP 0x120
+#define QSERDES_PLL_VCO_TUNE1_MODE0 0x124
+#define QSERDES_PLL_VCO_TUNE2_MODE0 0x128
+#define QSERDES_PLL_VCO_TUNE1_MODE1 0x12c
+#define QSERDES_PLL_VCO_TUNE2_MODE1 0x130
+#define QSERDES_PLL_VCO_TUNE_TIMER1 0x13c
+#define QSERDES_PLL_VCO_TUNE_TIMER2 0x140
+#define QSERDES_PLL_CLK_SELECT 0x16c
+#define QSERDES_PLL_HSCLK_SEL 0x170
+#define QSERDES_PLL_CORECLK_DIV 0x17c
+#define QSERDES_PLL_CORE_CLK_EN 0x184
+#define QSERDES_PLL_CMN_CONFIG 0x18c
+#define QSERDES_PLL_SVS_MODE_CLK_SEL 0x194
+#define QSERDES_PLL_CORECLK_DIV_MODE1 0x1b4
+
+/* QMP V2 PHY for PCIE gen3 ports - QSERDES TX registers */
+
+#define QSERDES_TX0_RES_CODE_LANE_OFFSET_TX 0x03c
+#define QSERDES_TX0_HIGHZ_DRVR_EN 0x058
+#define QSERDES_TX0_LANE_MODE_1 0x084
+#define QSERDES_TX0_RCV_DETECT_LVL_2 0x09c
+
+/* QMP V2 PHY for PCIE gen3 ports - QSERDES RX registers */
+
+#define QSERDES_RX0_UCDR_FO_GAIN 0x008
+#define QSERDES_RX0_UCDR_SO_GAIN 0x014
+#define QSERDES_RX0_UCDR_SO_SATURATION_AND_ENABLE 0x034
+#define QSERDES_RX0_UCDR_PI_CONTROLS 0x044
+#define QSERDES_RX0_RX_EQU_ADAPTOR_CNTRL2 0x0ec
+#define QSERDES_RX0_RX_EQU_ADAPTOR_CNTRL3 0x0f0
+#define QSERDES_RX0_RX_EQU_ADAPTOR_CNTRL4 0x0f4
+#define QSERDES_RX0_RX_IDAC_TSETTLE_LOW 0x0f8
+#define QSERDES_RX0_RX_IDAC_TSETTLE_HIGH 0x0fc
+#define QSERDES_RX0_RX_EQ_OFFSET_ADAPTOR_CNTRL1 0x110
+#define QSERDES_RX0_RX_OFFSET_ADAPTOR_CNTRL2 0x114
+#define QSERDES_RX0_SIGDET_ENABLES 0x118
+#define QSERDES_RX0_SIGDET_CNTRL 0x11c
+#define QSERDES_RX0_SIGDET_DEGLITCH_CNTRL 0x124
+#define QSERDES_RX0_RX_MODE_00_LOW 0x170
+#define QSERDES_RX0_RX_MODE_00_HIGH 0x174
+#define QSERDES_RX0_RX_MODE_00_HIGH2 0x178
+#define QSERDES_RX0_RX_MODE_00_HIGH3 0x17c
+#define QSERDES_RX0_RX_MODE_00_HIGH4 0x180
+#define QSERDES_RX0_RX_MODE_01_LOW 0x184
+#define QSERDES_RX0_RX_MODE_01_HIGH 0x188
+#define QSERDES_RX0_RX_MODE_01_HIGH2 0x18c
+#define QSERDES_RX0_RX_MODE_01_HIGH3 0x190
+#define QSERDES_RX0_RX_MODE_01_HIGH4 0x194
+#define QSERDES_RX0_RX_MODE_10_LOW 0x198
+#define QSERDES_RX0_RX_MODE_10_HIGH 0x19c
+#define QSERDES_RX0_RX_MODE_10_HIGH2 0x1a0
+#define QSERDES_RX0_RX_MODE_10_HIGH3 0x1a4
+#define QSERDES_RX0_RX_MODE_10_HIGH4 0x1a8
+#define QSERDES_RX0_DFE_EN_TIMER 0x1b4
+
+/* QMP V2 PHY for PCIE gen3 ports - PCS registers */
+
+#define PCS_COM_FLL_CNTRL1 0x098
+#define PCS_COM_FLL_CNTRL2 0x09c
+#define PCS_COM_FLL_CNT_VAL_L 0x0a0
+#define PCS_COM_FLL_CNT_VAL_H_TOL 0x0a4
+#define PCS_COM_FLL_MAN_CODE 0x0a8
+#define PCS_COM_REFGEN_REQ_CONFIG1 0x0dc
+#define PCS_COM_G12S1_TXDEEMPH_M3P5DB 0x16c
+#define PCS_COM_RX_SIGDET_LVL 0x188
+#define PCS_COM_P2U3_WAKEUP_DLY_TIME_AUXCLK_L 0x1a4
+#define PCS_COM_P2U3_WAKEUP_DLY_TIME_AUXCLK_H 0x1a8
+#define PCS_COM_RX_DCC_CAL_CONFIG 0x1d8
+#define PCS_COM_EQ_CONFIG5 0x1ec
+
+/* QMP V2 PHY for PCIE gen3 ports - PCS Misc registers */
+
+#define PCS_PCIE_POWER_STATE_CONFIG2 0x40c
+#define PCS_PCIE_POWER_STATE_CONFIG4 0x414
+#define PCS_PCIE_ENDPOINT_REFCLK_DRIVE 0x41c
+#define PCS_PCIE_L1P1_WAKEUP_DLY_TIME_AUXCLK_L 0x440
+#define PCS_PCIE_L1P1_WAKEUP_DLY_TIME_AUXCLK_H 0x444
+#define PCS_PCIE_L1P2_WAKEUP_DLY_TIME_AUXCLK_L 0x448
+#define PCS_PCIE_L1P2_WAKEUP_DLY_TIME_AUXCLK_H 0x44c
+#define PCS_PCIE_OSC_DTCT_CONFIG2 0x45c
+#define PCS_PCIE_OSC_DTCT_MODE2_CONFIG2 0x478
+#define PCS_PCIE_OSC_DTCT_MODE2_CONFIG4 0x480
+#define PCS_PCIE_OSC_DTCT_MODE2_CONFIG5 0x484
+#define PCS_PCIE_OSC_DTCT_ACTIONS 0x490
+#define PCS_PCIE_EQ_CONFIG1 0x4a0
+#define PCS_PCIE_EQ_CONFIG2 0x4a4
+#define PCS_PCIE_PRESET_P10_PRE 0x4bc
+#define PCS_PCIE_PRESET_P10_POST 0x4e0
+
/* Only for QMP V2 PHY - QSERDES COM registers */
#define QSERDES_COM_BG_TIMER 0x00c
#define QSERDES_COM_SSC_EN_CENTER 0x010
--
2.30.2
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH 3/5] arm64: dts: ipq6018: Add pcie support
2021-04-20 11:21 [PATCH 0/5] arm64: IPQ6018 PCIe support Baruch Siach
2021-04-20 11:21 ` [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller Baruch Siach
2021-04-20 11:21 ` [PATCH 2/5] phy: qcom-qmp: add QMP V2 PCIe PHY support for ipq60xx Baruch Siach
@ 2021-04-20 11:21 ` Baruch Siach
2021-04-20 11:21 ` [PATCH 4/5] dt-bindings: phy: qcom,qmp: Add IPQ60xx PCIe PHY bindings Baruch Siach
2021-04-20 11:21 ` [PATCH 5/5] dt-bindings: pci: qcom: Document PCIe bindings for IPQ6018 SoC Baruch Siach
4 siblings, 0 replies; 11+ messages in thread
From: Baruch Siach @ 2021-04-20 11:21 UTC (permalink / raw)
To: Andy Gross, Bjorn Andersson
Cc: Selvam Sathappan Periakaruppan, Baruch Siach, Kathiravan T,
Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul, Rob Herring,
devicetree, linux-phy, linux-pci, linux-arm-msm,
linux-arm-kernel
From: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
ipq6018 has 1 pcie gen3 port. This patch adds the support for the same.
The GICv2m reg property value is a guess based on similar SoCs
description in downstream Codeaurora kernel. It appears to work.
Signed-off-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
[baruch: adjust #address-cells/#size-cells; drop unsupported property;
use v2m msi interrupts parent; increase parf registers size]
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
---
arch/arm64/boot/dts/qcom/ipq6018.dtsi | 109 ++++++++++++++++++++++++++
1 file changed, 109 insertions(+)
diff --git a/arch/arm64/boot/dts/qcom/ipq6018.dtsi b/arch/arm64/boot/dts/qcom/ipq6018.dtsi
index 9fa5b028e4f3..a5b03b0d564b 100644
--- a/arch/arm64/boot/dts/qcom/ipq6018.dtsi
+++ b/arch/arm64/boot/dts/qcom/ipq6018.dtsi
@@ -375,6 +375,8 @@ qpic_nand: nand@79b0000 {
intc: interrupt-controller@b000000 {
compatible = "qcom,msm-qgic2";
+ #address-cells = <1>;
+ #size-cells = <1>;
interrupt-controller;
#interrupt-cells = <0x3>;
reg = <0x0 0x0b000000 0x0 0x1000>, /*GICD*/
@@ -382,6 +384,113 @@ intc: interrupt-controller@b000000 {
<0x0 0x0b001000 0x0 0x1000>, /*GICH*/
<0x0 0x0b004000 0x0 0x1000>; /*GICV*/
interrupts = <GIC_PPI 9 IRQ_TYPE_LEVEL_HIGH>;
+ ranges = <0x0 0x0 0xb00a000 0xffd>;
+
+ v2m0: v2m@0 {
+ compatible = "arm,gic-v2m-frame";
+ msi-controller;
+ reg = <0x0 0xffd>;
+ };
+ };
+
+ pcie_phy: phy@84000 {
+ compatible = "qcom,ipq6018-qmp-pcie-phy";
+ reg = <0x0 0x84000 0x0 0x1bc>; /* Serdes PLL */
+ status = "disabled";
+ #address-cells = <2>;
+ #size-cells = <2>;
+ ranges;
+
+ clocks = <&gcc GCC_PCIE0_AUX_CLK>,
+ <&gcc GCC_PCIE0_AHB_CLK>;
+ clock-names = "aux", "cfg_ahb";
+
+ resets = <&gcc GCC_PCIE0_PHY_BCR>,
+ <&gcc GCC_PCIE0PHY_PHY_BCR>;
+ reset-names = "phy",
+ "common";
+
+ pcie_phy0: lane@84200 {
+ reg = <0x0 0x84200 0x0 0x16c>, /* Serdes Tx */
+ <0x0 0x84400 0x0 0x200>, /* Serdes Rx */
+ <0x0 0x84800 0x0 0x4f4>; /* PCS: Lane0, COM, PCIE */
+ #phy-cells = <0>;
+
+ clocks = <&gcc GCC_PCIE0_PIPE_CLK>;
+ clock-names = "pipe0";
+ clock-output-names = "gcc_pcie0_pipe_clk_src";
+ #clock-cells = <0>;
+ };
+ };
+
+ pcie0: pci@20000000 {
+ compatible = "qcom,pcie-ipq6018";
+ reg = <0x0 0x20000000 0x0 0xf1d>,
+ <0x0 0x20000f20 0x0 0xa8>,
+ <0x0 0x20001000 0x0 0x1000>,
+ <0x0 0x80000 0x0 0x4000>,
+ <0x0 0x20100000 0x0 0x1000>;
+ reg-names = "dbi", "elbi", "atu", "parf", "config";
+
+ device_type = "pci";
+ linux,pci-domain = <0>;
+ bus-range = <0x00 0xff>;
+ num-lanes = <1>;
+ #address-cells = <3>;
+ #size-cells = <2>;
+
+ phys = <&pcie_phy0>;
+ phy-names = "pciephy";
+
+ ranges = <0x81000000 0 0x20200000 0 0x20200000
+ 0 0x10000>, /* downstream I/O */
+ <0x82000000 0 0x20220000 0 0x20220000
+ 0 0xfde0000>; /* non-prefetchable memory */
+
+ interrupts = <GIC_SPI 52 IRQ_TYPE_LEVEL_HIGH>;
+ interrupt-names = "msi";
+ msi-parent = <&v2m0>;
+
+ #interrupt-cells = <1>;
+ interrupt-map-mask = <0 0 0 0x7>;
+ interrupt-map = <0 0 0 1 &intc 0 75
+ IRQ_TYPE_LEVEL_HIGH>, /* int_a */
+ <0 0 0 2 &intc 0 78
+ IRQ_TYPE_LEVEL_HIGH>, /* int_b */
+ <0 0 0 3 &intc 0 79
+ IRQ_TYPE_LEVEL_HIGH>, /* int_c */
+ <0 0 0 4 &intc 0 83
+ IRQ_TYPE_LEVEL_HIGH>; /* int_d */
+
+ clocks = <&gcc GCC_SYS_NOC_PCIE0_AXI_CLK>,
+ <&gcc GCC_PCIE0_AXI_M_CLK>,
+ <&gcc GCC_PCIE0_AXI_S_CLK>,
+ <&gcc GCC_PCIE0_AXI_S_BRIDGE_CLK>,
+ <&gcc PCIE0_RCHNG_CLK>;
+ clock-names = "iface",
+ "axi_m",
+ "axi_s",
+ "axi_bridge",
+ "rchng";
+
+ resets = <&gcc GCC_PCIE0_PIPE_ARES>,
+ <&gcc GCC_PCIE0_SLEEP_ARES>,
+ <&gcc GCC_PCIE0_CORE_STICKY_ARES>,
+ <&gcc GCC_PCIE0_AXI_MASTER_ARES>,
+ <&gcc GCC_PCIE0_AXI_SLAVE_ARES>,
+ <&gcc GCC_PCIE0_AHB_ARES>,
+ <&gcc GCC_PCIE0_AXI_MASTER_STICKY_ARES>,
+ <&gcc GCC_PCIE0_AXI_SLAVE_STICKY_ARES>;
+ reset-names = "pipe",
+ "sleep",
+ "sticky",
+ "axi_m",
+ "axi_s",
+ "ahb",
+ "axi_m_sticky",
+ "axi_s_sticky";
+
+ status = "disabled";
};
watchdog@b017000 {
--
2.30.2
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH 4/5] dt-bindings: phy: qcom,qmp: Add IPQ60xx PCIe PHY bindings
2021-04-20 11:21 [PATCH 0/5] arm64: IPQ6018 PCIe support Baruch Siach
` (2 preceding siblings ...)
2021-04-20 11:21 ` [PATCH 3/5] arm64: dts: ipq6018: Add pcie support Baruch Siach
@ 2021-04-20 11:21 ` Baruch Siach
2021-04-30 15:04 ` [PATCH 4/5] dt-bindings: phy: qcom, qmp: " Rob Herring
2021-04-20 11:21 ` [PATCH 5/5] dt-bindings: pci: qcom: Document PCIe bindings for IPQ6018 SoC Baruch Siach
4 siblings, 1 reply; 11+ messages in thread
From: Baruch Siach @ 2021-04-20 11:21 UTC (permalink / raw)
To: Andy Gross, Bjorn Andersson
Cc: Baruch Siach, Selvam Sathappan Periakaruppan, Kathiravan T,
Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul, Rob Herring,
devicetree, linux-phy, linux-pci, linux-arm-msm,
linux-arm-kernel
Add ipq6018 qmp phy device for the single PCIe serdes lane on IPQ60xx
SoCs.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
---
.../devicetree/bindings/phy/qcom,qmp-phy.yaml | 25 +++++++++++++++++++
1 file changed, 25 insertions(+)
diff --git a/Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml b/Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml
index 626447fee092..cb2eb7ac6d28 100644
--- a/Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml
+++ b/Documentation/devicetree/bindings/phy/qcom,qmp-phy.yaml
@@ -17,6 +17,7 @@ description:
properties:
compatible:
enum:
+ - qcom,ipq6018-qmp-pcie-phy
- qcom,ipq8074-qmp-pcie-phy
- qcom,ipq8074-qmp-usb3-phy
- qcom,msm8996-qmp-pcie-phy
@@ -294,6 +295,30 @@ allOf:
items:
- const: phy
- const: common
+ - if:
+ properties:
+ compatible:
+ contains:
+ enum:
+ - qcom,ipq6018-qmp-pcie-phy
+ then:
+ properties:
+ clocks:
+ items:
+ - description: Phy aux clock.
+ - description: Phy config clock.
+ clock-names:
+ items:
+ - const: aux
+ - const: cfg_ahb
+ resets:
+ items:
+ - description: reset of phy block.
+ - description: phy common block reset.
+ reset-names:
+ items:
+ - const: phy
+ - const: common
- if:
properties:
compatible:
--
2.30.2
^ permalink raw reply related [flat|nested] 11+ messages in thread
* [PATCH 5/5] dt-bindings: pci: qcom: Document PCIe bindings for IPQ6018 SoC
2021-04-20 11:21 [PATCH 0/5] arm64: IPQ6018 PCIe support Baruch Siach
` (3 preceding siblings ...)
2021-04-20 11:21 ` [PATCH 4/5] dt-bindings: phy: qcom,qmp: Add IPQ60xx PCIe PHY bindings Baruch Siach
@ 2021-04-20 11:21 ` Baruch Siach
2021-04-30 15:05 ` Rob Herring
4 siblings, 1 reply; 11+ messages in thread
From: Baruch Siach @ 2021-04-20 11:21 UTC (permalink / raw)
To: Andy Gross, Bjorn Andersson
Cc: Baruch Siach, Selvam Sathappan Periakaruppan, Kathiravan T,
Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul, Rob Herring,
devicetree, linux-phy, linux-pci, linux-arm-msm,
linux-arm-kernel
Document qcom,pcie-ipq6018. This is similar to the ipq8074 with a few
different clock sources, and one additional reset.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
---
.../devicetree/bindings/pci/qcom,pcie.txt | 24 +++++++++++++++++++
1 file changed, 24 insertions(+)
diff --git a/Documentation/devicetree/bindings/pci/qcom,pcie.txt b/Documentation/devicetree/bindings/pci/qcom,pcie.txt
index 0da458a051b6..25f4def468bf 100644
--- a/Documentation/devicetree/bindings/pci/qcom,pcie.txt
+++ b/Documentation/devicetree/bindings/pci/qcom,pcie.txt
@@ -14,6 +14,7 @@
- "qcom,pcie-qcs404" for qcs404
- "qcom,pcie-sdm845" for sdm845
- "qcom,pcie-sm8250" for sm8250
+ - "qcom,pcie-ipq6018" for ipq6018
- reg:
Usage: required
@@ -123,6 +124,16 @@
- "ahb" AHB clock
- "aux" Auxiliary clock
+- clock-names:
+ Usage: required for ipq6018
+ Value type: <stringlist>
+ Definition: Should contain the following entries
+ - "iface" PCIe to SysNOC BIU clock
+ - "axi_m" AXI Master clock
+ - "axi_s" AXI Slave clock
+ - "axi_bridge" AXI bridge clock
+ - "rchng"
+
- clock-names:
Usage: required for qcs404
Value type: <stringlist>
@@ -209,6 +220,19 @@
- "ahb" AHB Reset
- "axi_m_sticky" AXI Master Sticky reset
+- reset-names:
+ Usage: required for ipq6018
+ Value type: <stringlist>
+ Definition: Should contain the following entries
+ - "pipe" PIPE reset
+ - "sleep" Sleep reset
+ - "sticky" Core Sticky reset
+ - "axi_m" AXI Master reset
+ - "axi_s" AXI Slave reset
+ - "ahb" AHB Reset
+ - "axi_m_sticky" AXI Master Sticky reset
+ - "axi_s_sticky" AXI Slave Sticky reset
+
- reset-names:
Usage: required for qcs404
Value type: <stringlist>
--
2.30.2
^ permalink raw reply related [flat|nested] 11+ messages in thread
* Re: [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller
2021-04-20 11:21 ` [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller Baruch Siach
@ 2021-04-20 16:18 ` Rob Herring
2021-04-21 4:45 ` Baruch Siach
0 siblings, 1 reply; 11+ messages in thread
From: Rob Herring @ 2021-04-20 16:18 UTC (permalink / raw)
To: Baruch Siach
Cc: Andy Gross, Bjorn Andersson, Selvam Sathappan Periakaruppan,
Kathiravan T, Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul,
devicetree, linux-phy, linux-pci, linux-arm-msm,
linux-arm-kernel
On Tue, Apr 20, 2021 at 02:21:36PM +0300, Baruch Siach wrote:
> From: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
>
> IPQ60xx series of SoCs have one port of PCIe gen 3. Add support for that
> platform.
>
> The code is based on downstream Codeaurora kernel v5.4. Split out the
> registers access part from .init into .post_init. Registers are only
> accessible after phy_power_on().
>
> Signed-off-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
> Signed-off-by: Baruch Siach <baruch@tkos.co.il>
> ---
> drivers/pci/controller/dwc/pcie-qcom.c | 279 +++++++++++++++++++++++++
> 1 file changed, 279 insertions(+)
>
> diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
> index 8a7a300163e5..3e27de744738 100644
> --- a/drivers/pci/controller/dwc/pcie-qcom.c
> +++ b/drivers/pci/controller/dwc/pcie-qcom.c
> @@ -41,6 +41,31 @@
> #define L23_CLK_RMV_DIS BIT(2)
> #define L1_CLK_RMV_DIS BIT(1)
>
> +#define PCIE_ATU_CR1_OUTBOUND_6_GEN3 0xC00
> +#define PCIE_ATU_CR2_OUTBOUND_6_GEN3 0xC04
> +#define PCIE_ATU_LOWER_BASE_OUTBOUND_6_GEN3 0xC08
> +#define PCIE_ATU_UPPER_BASE_OUTBOUND_6_GEN3 0xC0C
> +#define PCIE_ATU_LIMIT_OUTBOUND_6_GEN3 0xC10
> +#define PCIE_ATU_LOWER_TARGET_OUTBOUND_6_GEN3 0xC14
> +#define PCIE_ATU_UPPER_TARGET_OUTBOUND_6_GEN3 0xC18
> +
> +#define PCIE_ATU_CR1_OUTBOUND_7_GEN3 0xE00
> +#define PCIE_ATU_CR2_OUTBOUND_7_GEN3 0xE04
> +#define PCIE_ATU_LOWER_BASE_OUTBOUND_7_GEN3 0xE08
> +#define PCIE_ATU_UPPER_BASE_OUTBOUND_7_GEN3 0xE0C
> +#define PCIE_ATU_LIMIT_OUTBOUND_7_GEN3 0xE10
> +#define PCIE_ATU_LOWER_TARGET_OUTBOUND_7_GEN3 0xE14
> +#define PCIE_ATU_UPPER_TARGET_OUTBOUND_7_GEN3 0xE18
ATU registers are standard DWC registers. Plus upstream now dynamically
detects how many ATU regions there are.
> +
> +#define PCIE20_COMMAND_STATUS 0x04
> +#define BUS_MASTER_EN 0x7
> +#define PCIE20_DEVICE_CONTROL2_STATUS2 0x98
> +#define PCIE_CAP_CPL_TIMEOUT_DISABLE 0x10
All PCI standard registers.
> +#define PCIE30_GEN3_RELATED_OFF 0x890
Looks like a DWC port logic register. The define at a minimum goes in
the common code. We probably already have one. Code touching the
register should ideally be there too (hint: look at the other drivers).
> +
> +#define RXEQ_RGRDLESS_RXTS BIT(13)
> +#define GEN3_ZRXDC_NONCOMPL BIT(0)
> +
> #define PCIE20_PARF_PHY_CTRL 0x40
> #define PHY_CTRL_PHY_TX0_TERM_OFFSET_MASK GENMASK(20, 16)
> #define PHY_CTRL_PHY_TX0_TERM_OFFSET(x) ((x) << 16)
> @@ -52,6 +77,10 @@
> #define PCIE20_PARF_DBI_BASE_ADDR 0x168
> #define PCIE20_PARF_SLV_ADDR_SPACE_SIZE 0x16C
> #define PCIE20_PARF_MHI_CLOCK_RESET_CTRL 0x174
> +#define AHB_CLK_EN BIT(0)
> +#define MSTR_AXI_CLK_EN BIT(1)
> +#define BYPASS BIT(4)
> +
> #define PCIE20_PARF_AXI_MSTR_WR_ADDR_HALT 0x178
> #define PCIE20_PARF_AXI_MSTR_WR_ADDR_HALT_V2 0x1A8
> #define PCIE20_PARF_LTSSM 0x1B0
> @@ -94,6 +123,12 @@
> #define SLV_ADDR_SPACE_SZ 0x10000000
>
> #define PCIE20_LNK_CONTROL2_LINK_STATUS2 0xa0
> +#define PCIE_CAP_CURR_DEEMPHASIS BIT(16)
> +#define SPEED_GEN1 0x1
> +#define SPEED_GEN2 0x2
> +#define SPEED_GEN3 0x3
> +#define AXI_CLK_RATE 200000000
> +#define RCHNG_CLK_RATE 100000000
>
> #define DEVICE_TYPE_RC 0x4
>
> @@ -168,6 +203,15 @@ struct qcom_pcie_resources_2_7_0 {
> struct clk *pipe_clk;
> };
>
> +struct qcom_pcie_resources_2_9_0 {
> + struct clk *iface;
> + struct clk *axi_m_clk;
> + struct clk *axi_s_clk;
> + struct clk *axi_bridge_clk;
> + struct clk *rchng_clk;
> + struct reset_control *rst[8];
> +};
> +
> union qcom_pcie_resources {
> struct qcom_pcie_resources_1_0_0 v1_0_0;
> struct qcom_pcie_resources_2_1_0 v2_1_0;
> @@ -175,6 +219,7 @@ union qcom_pcie_resources {
> struct qcom_pcie_resources_2_3_3 v2_3_3;
> 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;
> @@ -1266,6 +1311,225 @@ static void qcom_pcie_post_deinit_2_7_0(struct qcom_pcie *pcie)
> clk_disable_unprepare(res->pipe_clk);
> }
>
> +static int qcom_pcie_get_resources_2_9_0(struct qcom_pcie *pcie)
> +{
> + struct qcom_pcie_resources_2_9_0 *res = &pcie->res.v2_9_0;
> + struct dw_pcie *pci = pcie->pci;
> + struct device *dev = pci->dev;
> + int i;
> + const char *rst_names[] = { "pipe", "sleep", "sticky",
> + "axi_m", "axi_s", "ahb",
> + "axi_m_sticky", "axi_s_sticky", };
> +
> + res->iface = devm_clk_get(dev, "iface");
> + if (IS_ERR(res->iface))
> + return PTR_ERR(res->iface);
> +
> + res->axi_m_clk = devm_clk_get(dev, "axi_m");
> + if (IS_ERR(res->axi_m_clk))
> + return PTR_ERR(res->axi_m_clk);
> +
> + res->axi_s_clk = devm_clk_get(dev, "axi_s");
> + if (IS_ERR(res->axi_s_clk))
> + return PTR_ERR(res->axi_s_clk);
> +
> + res->axi_bridge_clk = devm_clk_get(dev, "axi_bridge");
> + if (IS_ERR(res->axi_bridge_clk))
> + return PTR_ERR(res->axi_bridge_clk);
> +
> + res->rchng_clk = devm_clk_get(dev, "rchng");
> + if (IS_ERR(res->rchng_clk))
> + return PTR_ERR(res->rchng_clk);
> +
> + for (i = 0; i < ARRAY_SIZE(rst_names); i++) {
> + res->rst[i] = devm_reset_control_get(dev, rst_names[i]);
Don't we have a bulk api for this? Clocks too.
> + if (IS_ERR(res->rst[i])) {
> + return PTR_ERR(res->rst[i]);
> + }
> + }
> +
> + return 0;
> +}
> +
> +static void qcom_pcie_deinit_2_9_0(struct qcom_pcie *pcie)
> +{
> + struct qcom_pcie_resources_2_9_0 *res = &pcie->res.v2_9_0;
> +
> + clk_disable_unprepare(res->axi_m_clk);
> + clk_disable_unprepare(res->axi_s_clk);
> + clk_disable_unprepare(res->axi_bridge_clk);
> + clk_disable_unprepare(res->rchng_clk);
> + clk_disable_unprepare(res->iface);
> +}
> +
> +static int qcom_pcie_init_2_9_0(struct qcom_pcie *pcie)
> +{
> + struct qcom_pcie_resources_2_9_0 *res = &pcie->res.v2_9_0;
> + struct device *dev = pcie->pci->dev;
> + int i, ret;
> +
> + for (i = 0; i < ARRAY_SIZE(res->rst); i++) {
> + ret = reset_control_assert(res->rst[i]);
> + if (ret) {
> + dev_err(dev, "reset #%d assert failed (%d)\n", i, ret);
> + return ret;
> + }
> + }
> +
> + usleep_range(2000, 2500);
> +
> + for (i = 0; i < ARRAY_SIZE(res->rst); i++) {
> + ret = reset_control_deassert(res->rst[i]);
> + if (ret) {
> + dev_err(dev, "reset #%d deassert failed (%d)\n", i,
> + ret);
> + return ret;
> + }
> + }
> +
> + /*
> + * Don't have a way to see if the reset has completed.
> + * Wait for some time.
> + */
> + usleep_range(2000, 2500);
> +
> + ret = clk_prepare_enable(res->iface);
> + if (ret) {
> + dev_err(dev, "cannot prepare/enable core clock\n");
> + goto err_clk_iface;
> + }
> +
> + ret = clk_prepare_enable(res->axi_m_clk);
> + if (ret) {
> + dev_err(dev, "cannot prepare/enable core clock\n");
> + goto err_clk_axi_m;
> + }
> +
> + ret = clk_set_rate(res->axi_m_clk, AXI_CLK_RATE);
> + if (ret) {
> + dev_err(dev, "MClk rate set failed (%d)\n", ret);
> + goto err_clk_axi_m;
> + }
> +
> + ret = clk_prepare_enable(res->axi_s_clk);
> + if (ret) {
> + dev_err(dev, "cannot prepare/enable axi slave clock\n");
> + goto err_clk_axi_s;
> + }
> +
> + ret = clk_set_rate(res->axi_s_clk, AXI_CLK_RATE);
> + if (ret) {
> + dev_err(dev, "SClk rate set failed (%d)\n", ret);
> + goto err_clk_axi_s;
> + }
> +
> + ret = clk_prepare_enable(res->axi_bridge_clk);
> + if (ret) {
> + dev_err(dev, "cannot prepare/enable axi bridge clock\n");
> + goto err_clk_axi_bridge;
> + }
> +
> + ret = clk_prepare_enable(res->rchng_clk);
> + if (ret) {
> + dev_err(dev, "cannot prepare/enable rchng clock\n");
> + goto err_clk_rchng;
> + }
> +
> + ret = clk_set_rate(res->rchng_clk, RCHNG_CLK_RATE);
> + if (ret) {
> + dev_err(dev, "rchng_clk rate set failed (%d)\n", ret);
> + goto err_clk_rchng;
> + }
> +
> + return 0;
> +
> +err_clk_rchng:
> + clk_disable_unprepare(res->rchng_clk);
> +err_clk_axi_bridge:
> + clk_disable_unprepare(res->axi_bridge_clk);
> +err_clk_axi_s:
> + clk_disable_unprepare(res->axi_s_clk);
> +err_clk_axi_m:
> + clk_disable_unprepare(res->axi_m_clk);
> +err_clk_iface:
> + clk_disable_unprepare(res->iface);
> + /*
> + * Not checking for failure, will anyway return
> + * the original failure in 'ret'.
> + */
> + for (i = 0; i < ARRAY_SIZE(res->rst); i++)
> + reset_control_assert(res->rst[i]);
> +
> + return ret;
> +}
> +
> +static int qcom_pcie_post_init_2_9_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(SLV_ADDR_SPACE_SZ,
> + pcie->parf + PCIE20_v3_PARF_SLV_ADDR_SPACE_SIZE);
> +
> + val = readl(pcie->parf + PCIE20_PARF_PHY_CTRL);
> + val &= ~BIT(0);
What's 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(RXEQ_RGRDLESS_RXTS | GEN3_ZRXDC_NONCOMPL,
> + pci->dbi_base + PCIE30_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);
> +
> + writel(BUS_MASTER_EN, pci->dbi_base + PCIE20_COMMAND_STATUS);
Pretty sure the DWC core or PCI core does this already.
> +
> + writel(DBI_RO_WR_EN, pci->dbi_base + PCIE20_MISC_CONTROL_1_REG);
> + writel(PCIE_CAP_LINK1_VAL, pci->dbi_base + offset + PCI_EXP_SLTCAP);
> +
> + /* Configure PCIe link capabilities for ASPM */
> + val = readl(pci->dbi_base + offset + PCI_EXP_LNKCAP);
> + val &= ~PCI_EXP_LNKCAP_ASPMS;
> + writel(val, pci->dbi_base + offset + PCI_EXP_LNKCAP);
> +
> + writel(PCIE_CAP_CPL_TIMEOUT_DISABLE, pci->dbi_base +
> + PCIE20_DEVICE_CONTROL2_STATUS2);
> +
> + writel(PCIE_CAP_CURR_DEEMPHASIS | SPEED_GEN3,
> + pci->dbi_base + offset + PCI_EXP_DEVCTL2);
This all looks like stuff that should be in the DWC core code. Maybe we
need an ASPM disable quirk or something? That's probably somewhat
common.
> +
> + for (i = 0;i < 256;i++)
> + writel(0x0, pcie->parf + PCIE20_PARF_BDF_TO_SID_TABLE_N
> + + (4 * i));
> +
> + writel(0x4, pci->atu_base + PCIE_ATU_CR1_OUTBOUND_6_GEN3);
> + writel(0x90000000, pci->atu_base + PCIE_ATU_CR2_OUTBOUND_6_GEN3);
> + writel(0x0, pci->atu_base + PCIE_ATU_LOWER_BASE_OUTBOUND_6_GEN3);
> + writel(0x0, pci->atu_base + PCIE_ATU_UPPER_BASE_OUTBOUND_6_GEN3);
> + writel(0x00107FFFF, pci->atu_base + PCIE_ATU_LIMIT_OUTBOUND_6_GEN3);
> + writel(0x0, pci->atu_base + PCIE_ATU_LOWER_TARGET_OUTBOUND_6_GEN3);
> + writel(0x0, pci->atu_base + PCIE_ATU_UPPER_TARGET_OUTBOUND_6_GEN3);
> + writel(0x5, pci->atu_base + PCIE_ATU_CR1_OUTBOUND_7_GEN3);
> + writel(0x90000000, pci->atu_base + PCIE_ATU_CR2_OUTBOUND_7_GEN3);
> + writel(0x200000, pci->atu_base + PCIE_ATU_LOWER_BASE_OUTBOUND_7_GEN3);
> + writel(0x0, pci->atu_base + PCIE_ATU_UPPER_BASE_OUTBOUND_7_GEN3);
> + writel(0x7FFFFF, pci->atu_base + PCIE_ATU_LIMIT_OUTBOUND_7_GEN3);
> + writel(0x0, pci->atu_base + PCIE_ATU_LOWER_TARGET_OUTBOUND_7_GEN3);
> + writel(0x0, pci->atu_base + PCIE_ATU_UPPER_TARGET_OUTBOUND_7_GEN3);
This should all be coming from 'ranges' in the DT. If not, why not?
If you haven't caught the theme yet, everything outside of PARF register
accesses had better have a good explanation why they can't be in common
code.
> +
> + return 0;
> +}
> +
> static int qcom_pcie_link_up(struct dw_pcie *pci)
> {
> u16 offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
> @@ -1456,6 +1720,15 @@ static const struct qcom_pcie_ops ops_1_9_0 = {
> .config_sid = qcom_pcie_config_sid_sm8250,
> };
>
> +/* Qcom IP rev.: 2.9.0 Synopsys IP rev.: 5.00a */
> +static const struct qcom_pcie_ops ops_2_9_0 = {
> + .get_resources = qcom_pcie_get_resources_2_9_0,
> + .init = qcom_pcie_init_2_9_0,
> + .post_init = qcom_pcie_post_init_2_9_0,
> + .deinit = qcom_pcie_deinit_2_9_0,
> + .ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
> +};
> +
> static const struct dw_pcie_ops dw_pcie_ops = {
> .link_up = qcom_pcie_link_up,
> .start_link = qcom_pcie_start_link,
> @@ -1508,6 +1781,11 @@ static int qcom_pcie_probe(struct platform_device *pdev)
> goto err_pm_runtime_put;
> }
>
> + /* We need ATU for .post_init */
> + pci->atu_base = devm_platform_ioremap_resource_byname(pdev, "atu");
The DWC core handles this now.
> + if (IS_ERR(pci->atu_base))
> + pci->atu_base = NULL;
> +
> pcie->phy = devm_phy_optional_get(dev, "pciephy");
> if (IS_ERR(pcie->phy)) {
> ret = PTR_ERR(pcie->phy);
> @@ -1555,6 +1833,7 @@ static const struct of_device_id qcom_pcie_match[] = {
> { .compatible = "qcom,pcie-qcs404", .data = &ops_2_4_0 },
> { .compatible = "qcom,pcie-sdm845", .data = &ops_2_7_0 },
> { .compatible = "qcom,pcie-sm8250", .data = &ops_1_9_0 },
> + { .compatible = "qcom,pcie-ipq6018", .data = &ops_2_9_0 },
> { }
> };
>
> --
> 2.30.2
>
>
> _______________________________________________
> linux-arm-kernel mailing list
> linux-arm-kernel@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller
2021-04-20 16:18 ` Rob Herring
@ 2021-04-21 4:45 ` Baruch Siach
2021-04-21 18:45 ` Rob Herring
0 siblings, 1 reply; 11+ messages in thread
From: Baruch Siach @ 2021-04-21 4:45 UTC (permalink / raw)
To: Rob Herring
Cc: Andy Gross, Bjorn Andersson, Selvam Sathappan Periakaruppan,
Kathiravan T, Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul,
devicetree, linux-phy, linux-pci, linux-arm-msm,
linux-arm-kernel
Hi Rob,
Thanks for your review.
I have a few comments below.
On Tue, Apr 20 2021, Rob Herring wrote:
> On Tue, Apr 20, 2021 at 02:21:36PM +0300, Baruch Siach wrote:
>> From: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
>>
>> IPQ60xx series of SoCs have one port of PCIe gen 3. Add support for that
>> platform.
>>
>> The code is based on downstream Codeaurora kernel v5.4. Split out the
>> registers access part from .init into .post_init. Registers are only
>> accessible after phy_power_on().
>>
>> Signed-off-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
>> Signed-off-by: Baruch Siach <baruch@tkos.co.il>
>> ---
>> drivers/pci/controller/dwc/pcie-qcom.c | 279 +++++++++++++++++++++++++
>> 1 file changed, 279 insertions(+)
>>
>> diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
>> index 8a7a300163e5..3e27de744738 100644
>> --- a/drivers/pci/controller/dwc/pcie-qcom.c
>> +++ b/drivers/pci/controller/dwc/pcie-qcom.c
>> @@ -41,6 +41,31 @@
>> #define L23_CLK_RMV_DIS BIT(2)
>> #define L1_CLK_RMV_DIS BIT(1)
>>
>> +#define PCIE_ATU_CR1_OUTBOUND_6_GEN3 0xC00
>> +#define PCIE_ATU_CR2_OUTBOUND_6_GEN3 0xC04
>> +#define PCIE_ATU_LOWER_BASE_OUTBOUND_6_GEN3 0xC08
>> +#define PCIE_ATU_UPPER_BASE_OUTBOUND_6_GEN3 0xC0C
>> +#define PCIE_ATU_LIMIT_OUTBOUND_6_GEN3 0xC10
>> +#define PCIE_ATU_LOWER_TARGET_OUTBOUND_6_GEN3 0xC14
>> +#define PCIE_ATU_UPPER_TARGET_OUTBOUND_6_GEN3 0xC18
>> +
>> +#define PCIE_ATU_CR1_OUTBOUND_7_GEN3 0xE00
>> +#define PCIE_ATU_CR2_OUTBOUND_7_GEN3 0xE04
>> +#define PCIE_ATU_LOWER_BASE_OUTBOUND_7_GEN3 0xE08
>> +#define PCIE_ATU_UPPER_BASE_OUTBOUND_7_GEN3 0xE0C
>> +#define PCIE_ATU_LIMIT_OUTBOUND_7_GEN3 0xE10
>> +#define PCIE_ATU_LOWER_TARGET_OUTBOUND_7_GEN3 0xE14
>> +#define PCIE_ATU_UPPER_TARGET_OUTBOUND_7_GEN3 0xE18
>
> ATU registers are standard DWC registers. Plus upstream now dynamically
> detects how many ATU regions there are.
>
>> +#define PCIE20_COMMAND_STATUS 0x04
>> +#define BUS_MASTER_EN 0x7
>> +#define PCIE20_DEVICE_CONTROL2_STATUS2 0x98
>> +#define PCIE_CAP_CPL_TIMEOUT_DISABLE 0x10
>
> All PCI standard registers.
PCIE20_COMMAND_STATUS is indeed the common PCI_COMMAND. I could not find
anything that matches PCIE20_DEVICE_CONTROL2_STATUS2. Where should I
look?
>
>> +#define PCIE30_GEN3_RELATED_OFF 0x890
>
> Looks like a DWC port logic register. The define at a minimum goes in
> the common code. We probably already have one. Code touching the
> register should ideally be there too (hint: look at the other drivers).
pcie-tegra194.c uses the equivalent GEN3_RELATED_OFF. So I can move the
definition to a common header. As for the code, I don't know. The tegra
configuration sequence involves other registers as well.
[snip]
>> +static int qcom_pcie_post_init_2_9_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(SLV_ADDR_SPACE_SZ,
>> + pcie->parf + PCIE20_v3_PARF_SLV_ADDR_SPACE_SIZE);
>> +
>> + val = readl(pcie->parf + PCIE20_PARF_PHY_CTRL);
>> + val &= ~BIT(0);
>
> What's BIT(0)?
I have no idea. I have no access to hardware documentation. I'm just
porting working code from the Codeaurora tree.
>
>> + 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(RXEQ_RGRDLESS_RXTS | GEN3_ZRXDC_NONCOMPL,
>> + pci->dbi_base + PCIE30_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);
>> +
>> + writel(BUS_MASTER_EN, pci->dbi_base + PCIE20_COMMAND_STATUS);
>
> Pretty sure the DWC core or PCI core does this already.
>
>> +
>> + writel(DBI_RO_WR_EN, pci->dbi_base + PCIE20_MISC_CONTROL_1_REG);
>> + writel(PCIE_CAP_LINK1_VAL, pci->dbi_base + offset + PCI_EXP_SLTCAP);
>> +
>> + /* Configure PCIe link capabilities for ASPM */
>> + val = readl(pci->dbi_base + offset + PCI_EXP_LNKCAP);
>> + val &= ~PCI_EXP_LNKCAP_ASPMS;
>> + writel(val, pci->dbi_base + offset + PCI_EXP_LNKCAP);
>> +
>> + writel(PCIE_CAP_CPL_TIMEOUT_DISABLE, pci->dbi_base +
>> + PCIE20_DEVICE_CONTROL2_STATUS2);
>> +
>> + writel(PCIE_CAP_CURR_DEEMPHASIS | SPEED_GEN3,
>> + pci->dbi_base + offset + PCI_EXP_DEVCTL2);
>
> This all looks like stuff that should be in the DWC core code. Maybe we
> need an ASPM disable quirk or something? That's probably somewhat
> common.
Where in common code should that be? Which part is quirky?
>> +
>> + for (i = 0;i < 256;i++)
>> + writel(0x0, pcie->parf + PCIE20_PARF_BDF_TO_SID_TABLE_N
>> + + (4 * i));
>> +
>
>> + writel(0x4, pci->atu_base + PCIE_ATU_CR1_OUTBOUND_6_GEN3);
>> + writel(0x90000000, pci->atu_base + PCIE_ATU_CR2_OUTBOUND_6_GEN3);
>> + writel(0x0, pci->atu_base + PCIE_ATU_LOWER_BASE_OUTBOUND_6_GEN3);
>> + writel(0x0, pci->atu_base + PCIE_ATU_UPPER_BASE_OUTBOUND_6_GEN3);
>> + writel(0x00107FFFF, pci->atu_base + PCIE_ATU_LIMIT_OUTBOUND_6_GEN3);
>> + writel(0x0, pci->atu_base + PCIE_ATU_LOWER_TARGET_OUTBOUND_6_GEN3);
>> + writel(0x0, pci->atu_base + PCIE_ATU_UPPER_TARGET_OUTBOUND_6_GEN3);
>> + writel(0x5, pci->atu_base + PCIE_ATU_CR1_OUTBOUND_7_GEN3);
>> + writel(0x90000000, pci->atu_base + PCIE_ATU_CR2_OUTBOUND_7_GEN3);
>> + writel(0x200000, pci->atu_base + PCIE_ATU_LOWER_BASE_OUTBOUND_7_GEN3);
>> + writel(0x0, pci->atu_base + PCIE_ATU_UPPER_BASE_OUTBOUND_7_GEN3);
>> + writel(0x7FFFFF, pci->atu_base + PCIE_ATU_LIMIT_OUTBOUND_7_GEN3);
>> + writel(0x0, pci->atu_base + PCIE_ATU_LOWER_TARGET_OUTBOUND_7_GEN3);
>> + writel(0x0, pci->atu_base + PCIE_ATU_UPPER_TARGET_OUTBOUND_7_GEN3);
>
> This should all be coming from 'ranges' in the DT. If not, why not?
I'll try to drop it and see if it works. I see that common code
overwrites this area anyway.
>
> If you haven't caught the theme yet, everything outside of PARF register
> accesses had better have a good explanation why they can't be in common
> code.
>
>> +
>> + return 0;
>> +}
>> +
>> static int qcom_pcie_link_up(struct dw_pcie *pci)
>> {
>> u16 offset = dw_pcie_find_capability(pci, PCI_CAP_ID_EXP);
>> @@ -1456,6 +1720,15 @@ static const struct qcom_pcie_ops ops_1_9_0 = {
>> .config_sid = qcom_pcie_config_sid_sm8250,
>> };
>>
>> +/* Qcom IP rev.: 2.9.0 Synopsys IP rev.: 5.00a */
>> +static const struct qcom_pcie_ops ops_2_9_0 = {
>> + .get_resources = qcom_pcie_get_resources_2_9_0,
>> + .init = qcom_pcie_init_2_9_0,
>> + .post_init = qcom_pcie_post_init_2_9_0,
>> + .deinit = qcom_pcie_deinit_2_9_0,
>> + .ltssm_enable = qcom_pcie_2_3_2_ltssm_enable,
>> +};
>> +
>> static const struct dw_pcie_ops dw_pcie_ops = {
>> .link_up = qcom_pcie_link_up,
>> .start_link = qcom_pcie_start_link,
>> @@ -1508,6 +1781,11 @@ static int qcom_pcie_probe(struct platform_device *pdev)
>> goto err_pm_runtime_put;
>> }
>>
>> + /* We need ATU for .post_init */
>> + pci->atu_base = devm_platform_ioremap_resource_byname(pdev, "atu");
>
> The DWC core handles this now.
With ATU code in .post_init gone, we can remove this as well.
baruch
--
~. .~ Tk Open Systems
=}------------------------------------------------ooO--U--Ooo------------{=
- baruch@tkos.co.il - tel: +972.52.368.4656, http://www.tkos.co.il -
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller
2021-04-21 4:45 ` Baruch Siach
@ 2021-04-21 18:45 ` Rob Herring
0 siblings, 0 replies; 11+ messages in thread
From: Rob Herring @ 2021-04-21 18:45 UTC (permalink / raw)
To: Baruch Siach
Cc: Andy Gross, Bjorn Andersson, Selvam Sathappan Periakaruppan,
Kathiravan T, Bjorn Helgaas, Kishon Vijay Abraham I, Vinod Koul,
devicetree, linux-phy, PCI, linux-arm-msm, linux-arm-kernel
On Tue, Apr 20, 2021 at 11:45 PM Baruch Siach <baruch@tkos.co.il> wrote:
>
> Hi Rob,
>
> Thanks for your review.
>
> I have a few comments below.
>
> On Tue, Apr 20 2021, Rob Herring wrote:
> > On Tue, Apr 20, 2021 at 02:21:36PM +0300, Baruch Siach wrote:
> >> From: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
> >>
> >> IPQ60xx series of SoCs have one port of PCIe gen 3. Add support for that
> >> platform.
> >>
> >> The code is based on downstream Codeaurora kernel v5.4. Split out the
> >> registers access part from .init into .post_init. Registers are only
> >> accessible after phy_power_on().
> >>
> >> Signed-off-by: Selvam Sathappan Periakaruppan <speriaka@codeaurora.org>
> >> Signed-off-by: Baruch Siach <baruch@tkos.co.il>
> >> ---
> >> drivers/pci/controller/dwc/pcie-qcom.c | 279 +++++++++++++++++++++++++
> >> 1 file changed, 279 insertions(+)
> >>
> >> diff --git a/drivers/pci/controller/dwc/pcie-qcom.c b/drivers/pci/controller/dwc/pcie-qcom.c
> >> index 8a7a300163e5..3e27de744738 100644
> >> --- a/drivers/pci/controller/dwc/pcie-qcom.c
> >> +++ b/drivers/pci/controller/dwc/pcie-qcom.c
> >> @@ -41,6 +41,31 @@
> >> #define L23_CLK_RMV_DIS BIT(2)
> >> #define L1_CLK_RMV_DIS BIT(1)
> >>
> >> +#define PCIE_ATU_CR1_OUTBOUND_6_GEN3 0xC00
> >> +#define PCIE_ATU_CR2_OUTBOUND_6_GEN3 0xC04
> >> +#define PCIE_ATU_LOWER_BASE_OUTBOUND_6_GEN3 0xC08
> >> +#define PCIE_ATU_UPPER_BASE_OUTBOUND_6_GEN3 0xC0C
> >> +#define PCIE_ATU_LIMIT_OUTBOUND_6_GEN3 0xC10
> >> +#define PCIE_ATU_LOWER_TARGET_OUTBOUND_6_GEN3 0xC14
> >> +#define PCIE_ATU_UPPER_TARGET_OUTBOUND_6_GEN3 0xC18
> >> +
> >> +#define PCIE_ATU_CR1_OUTBOUND_7_GEN3 0xE00
> >> +#define PCIE_ATU_CR2_OUTBOUND_7_GEN3 0xE04
> >> +#define PCIE_ATU_LOWER_BASE_OUTBOUND_7_GEN3 0xE08
> >> +#define PCIE_ATU_UPPER_BASE_OUTBOUND_7_GEN3 0xE0C
> >> +#define PCIE_ATU_LIMIT_OUTBOUND_7_GEN3 0xE10
> >> +#define PCIE_ATU_LOWER_TARGET_OUTBOUND_7_GEN3 0xE14
> >> +#define PCIE_ATU_UPPER_TARGET_OUTBOUND_7_GEN3 0xE18
> >
> > ATU registers are standard DWC registers. Plus upstream now dynamically
> > detects how many ATU regions there are.
> >
> >> +#define PCIE20_COMMAND_STATUS 0x04
> >> +#define BUS_MASTER_EN 0x7
> >> +#define PCIE20_DEVICE_CONTROL2_STATUS2 0x98
> >> +#define PCIE_CAP_CPL_TIMEOUT_DISABLE 0x10
> >
> > All PCI standard registers.
>
> PCIE20_COMMAND_STATUS is indeed the common PCI_COMMAND. I could not find
> anything that matches PCIE20_DEVICE_CONTROL2_STATUS2. Where should I
> look?
Looks like PCI_EXP_DEVCTL2 and PCI_EXP_DEVSTA2 to me. The register bit
looks like PCI_EXP_DEVCTL2_COMP_TMOUT_DIS. Those are extended config
registers so their offset is variable.
>
> >
> >> +#define PCIE30_GEN3_RELATED_OFF 0x890
> >
> > Looks like a DWC port logic register. The define at a minimum goes in
> > the common code. We probably already have one. Code touching the
> > register should ideally be there too (hint: look at the other drivers).
>
> pcie-tegra194.c uses the equivalent GEN3_RELATED_OFF. So I can move the
> definition to a common header. As for the code, I don't know. The tegra
> configuration sequence involves other registers as well.
I'm sure the Tegra folks will be happy to tell you if anything breaks.
>
> [snip]
>
> >> +static int qcom_pcie_post_init_2_9_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(SLV_ADDR_SPACE_SZ,
> >> + pcie->parf + PCIE20_v3_PARF_SLV_ADDR_SPACE_SIZE);
> >> +
> >> + val = readl(pcie->parf + PCIE20_PARF_PHY_CTRL);
> >> + val &= ~BIT(0);
> >
> > What's BIT(0)?
>
> I have no idea. I have no access to hardware documentation. I'm just
> porting working code from the Codeaurora tree.
Based on the 7 other existing modifications to that bit, it's 'enable
PCIe clocks and resets'. Looks like we need some refactoring at least
so there's not yet another copy.
> >> + 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(RXEQ_RGRDLESS_RXTS | GEN3_ZRXDC_NONCOMPL,
> >> + pci->dbi_base + PCIE30_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);
> >> +
> >> + writel(BUS_MASTER_EN, pci->dbi_base + PCIE20_COMMAND_STATUS);
> >
> > Pretty sure the DWC core or PCI core does this already.
> >
> >> +
> >> + writel(DBI_RO_WR_EN, pci->dbi_base + PCIE20_MISC_CONTROL_1_REG);
> >> + writel(PCIE_CAP_LINK1_VAL, pci->dbi_base + offset + PCI_EXP_SLTCAP);
I have to wonder if all the bits being set here are really true.
Hotplug is really supported? There's an attention button? Power
indicator? If anything, that's all board specific and would need to
come from firmware (DT).
> >> +
> >> + /* Configure PCIe link capabilities for ASPM */
> >> + val = readl(pci->dbi_base + offset + PCI_EXP_LNKCAP);
> >> + val &= ~PCI_EXP_LNKCAP_ASPMS;
> >> + writel(val, pci->dbi_base + offset + PCI_EXP_LNKCAP);
> >> +
> >> + writel(PCIE_CAP_CPL_TIMEOUT_DISABLE, pci->dbi_base +
> >> + PCIE20_DEVICE_CONTROL2_STATUS2);
> >> +
> >> + writel(PCIE_CAP_CURR_DEEMPHASIS | SPEED_GEN3,
SPEED_GEN3 does not look right for PCI_EXP_DEVCTL2.
> >> + pci->dbi_base + offset + PCI_EXP_DEVCTL2);
> >
> > This all looks like stuff that should be in the DWC core code. Maybe we
> > need an ASPM disable quirk or something? That's probably somewhat
> > common.
>
> Where in common code should that be?
If these registers are initialized elsewhere, in the same place.
Otherwise, probably in dw_pcie_setup_rc().
> Which part is quirky?
Disabling ASPM. It's a bit strange that some of this is needed at all
considering no other platform using the same IP needs it.
Rob
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH 4/5] dt-bindings: phy: qcom, qmp: Add IPQ60xx PCIe PHY bindings
2021-04-20 11:21 ` [PATCH 4/5] dt-bindings: phy: qcom,qmp: Add IPQ60xx PCIe PHY bindings Baruch Siach
@ 2021-04-30 15:04 ` Rob Herring
0 siblings, 0 replies; 11+ messages in thread
From: Rob Herring @ 2021-04-30 15:04 UTC (permalink / raw)
To: Baruch Siach
Cc: Kishon Vijay Abraham I, linux-arm-msm, Vinod Koul, Kathiravan T,
devicetree, linux-arm-kernel, Rob Herring, Andy Gross, linux-pci,
linux-phy, Bjorn Helgaas, Selvam Sathappan Periakaruppan,
Bjorn Andersson
On Tue, 20 Apr 2021 14:21:39 +0300, Baruch Siach wrote:
> Add ipq6018 qmp phy device for the single PCIe serdes lane on IPQ60xx
> SoCs.
>
> Signed-off-by: Baruch Siach <baruch@tkos.co.il>
> ---
> .../devicetree/bindings/phy/qcom,qmp-phy.yaml | 25 +++++++++++++++++++
> 1 file changed, 25 insertions(+)
>
Reviewed-by: Rob Herring <robh@kernel.org>
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH 5/5] dt-bindings: pci: qcom: Document PCIe bindings for IPQ6018 SoC
2021-04-20 11:21 ` [PATCH 5/5] dt-bindings: pci: qcom: Document PCIe bindings for IPQ6018 SoC Baruch Siach
@ 2021-04-30 15:05 ` Rob Herring
0 siblings, 0 replies; 11+ messages in thread
From: Rob Herring @ 2021-04-30 15:05 UTC (permalink / raw)
To: Baruch Siach
Cc: linux-pci, Kishon Vijay Abraham I, Bjorn Helgaas, Vinod Koul,
devicetree, linux-arm-msm, linux-arm-kernel, Bjorn Andersson,
linux-phy, Rob Herring, Selvam Sathappan Periakaruppan,
Kathiravan T, Andy Gross
On Tue, 20 Apr 2021 14:21:40 +0300, Baruch Siach wrote:
> Document qcom,pcie-ipq6018. This is similar to the ipq8074 with a few
> different clock sources, and one additional reset.
>
> Signed-off-by: Baruch Siach <baruch@tkos.co.il>
> ---
> .../devicetree/bindings/pci/qcom,pcie.txt | 24 +++++++++++++++++++
> 1 file changed, 24 insertions(+)
>
Reviewed-by: Rob Herring <robh@kernel.org>
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2021-04-30 15:05 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-04-20 11:21 [PATCH 0/5] arm64: IPQ6018 PCIe support Baruch Siach
2021-04-20 11:21 ` [PATCH 1/5] PCI: qcom: add support for IPQ60xx PCIe controller Baruch Siach
2021-04-20 16:18 ` Rob Herring
2021-04-21 4:45 ` Baruch Siach
2021-04-21 18:45 ` Rob Herring
2021-04-20 11:21 ` [PATCH 2/5] phy: qcom-qmp: add QMP V2 PCIe PHY support for ipq60xx Baruch Siach
2021-04-20 11:21 ` [PATCH 3/5] arm64: dts: ipq6018: Add pcie support Baruch Siach
2021-04-20 11:21 ` [PATCH 4/5] dt-bindings: phy: qcom,qmp: Add IPQ60xx PCIe PHY bindings Baruch Siach
2021-04-30 15:04 ` [PATCH 4/5] dt-bindings: phy: qcom, qmp: " Rob Herring
2021-04-20 11:21 ` [PATCH 5/5] dt-bindings: pci: qcom: Document PCIe bindings for IPQ6018 SoC Baruch Siach
2021-04-30 15:05 ` Rob Herring
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).