| Message ID | 824c61d13fe2731d812df8a0a878ca1a36399e76.1501122135.git.honghui.zhang@mediatek.com |
|---|---|
| State | Changes Requested |
| Headers | show |
On Thu, Jul 27, 2017 at 10:58:38AM +0800, honghui.zhang@mediatek.com wrote: > From: Ryder Lee <ryder.lee@mediatek.com> > > Add support for new Gen2 controller which has two root ports and shares > the probing flow with legacy controller. Currently this IP block can be > found on MT7622/MT2712. > > Signed-off-by: Ryder Lee <ryder.lee@mediatek.com> > Signed-off-by: Honghui Zhang <honghui.zhang@mediatek.com> > --- > drivers/pci/host/Kconfig | 5 +- > drivers/pci/host/pcie-mediatek.c | 490 ++++++++++++++++++++++++++++++++++++++- > 2 files changed, 489 insertions(+), 6 deletions(-) > > diff --git a/drivers/pci/host/Kconfig b/drivers/pci/host/Kconfig > index 89d61c2..5b1ae9f 100644 > --- a/drivers/pci/host/Kconfig > +++ b/drivers/pci/host/Kconfig > @@ -182,14 +182,13 @@ config PCIE_ROCKCHIP > > config PCIE_MEDIATEK > bool "MediaTek PCIe controller" > - depends on ARM && (ARCH_MEDIATEK || COMPILE_TEST) > + depends on (ARM || ARM64) && (ARCH_MEDIATEK || COMPILE_TEST) > depends on OF > depends on PCI > select PCIEPORTBUS > help > Say Y here if you want to enable PCIe controller support on > - MT7623 series SoCs. There is one single root complex with 3 root > - ports available. Each port supports Gen2 lane x1. > + MediaTek SoCs. > > config PCIE_TANGO_SMP8759 > bool "Tango SMP8759 PCIe controller (DANGEROUS)" > diff --git a/drivers/pci/host/pcie-mediatek.c b/drivers/pci/host/pcie-mediatek.c > index f8e2f68..5fe92ff 100644 > --- a/drivers/pci/host/pcie-mediatek.c > +++ b/drivers/pci/host/pcie-mediatek.c > @@ -3,6 +3,7 @@ > * > * Copyright (c) 2017 MediaTek Inc. > * Author: Ryder Lee <ryder.lee@mediatek.com> > + * Honghui Zhang <honghui.zhang@mediatek.com> > * > * This program is free software; you can redistribute it and/or modify > * it under the terms of the GNU General Public License version 2 as > @@ -17,6 +18,8 @@ > #include <linux/clk.h> > #include <linux/delay.h> > #include <linux/iopoll.h> > +#include <linux/irq.h> > +#include <linux/irqdomain.h> > #include <linux/kernel.h> > #include <linux/of_address.h> > #include <linux/of_pci.h> > @@ -64,16 +67,77 @@ > #define PCIE_FC_CREDIT_MASK (GENMASK(31, 31) | GENMASK(28, 16)) > #define PCIE_FC_CREDIT_VAL(x) ((x) << 16) > > +/* PCIe V2 share registers */ > +#define PCIE_SYS_CFG_V2 0x0 > +#define PCIE_CSR_LTSSM_EN(x) BIT(0 + (x) * 8) > +#define PCIE_CSR_ASPM_L1_EN(x) BIT(1 + (x) * 8) > + > +/* PCIe V2 per-port registers */ > +#define PCIE_INT_MASK 0x420 > +#define INTX_MASK GENMASK(19, 16) > +#define INTX_SHIFT 16 > +#define INTX_NUM 4 > +#define PCIE_INT_STATUS 0x424 > + > +#define PCIE_AHB_TRANS_BASE0_L 0x438 > +#define PCIE_AHB_TRANS_BASE0_H 0x43c > +#define AHB2PCIE_BASEL(x) ((x) & GENMASK(31, 0)) > +#define AHB2PCIE_BASEH(x) (((u64)(x) >> 32) & GENMASK(31, 0)) > +#define AHB2PCIE_SIZE(x) ((x) & GENMASK(4, 0)) > +#define PCIE_AXI_WINDOW0 0x448 > +#define WIN_ENABLE BIT(7) > + > +/* PCIe V2 Configuration Transaction Header */ > +#define PCIE_CFG_HEADER0 0x460 > +#define PCIE_CFG_HEADER1 0x464 > +#define PCIE_CFG_HEADER2 0x468 > +#define PCIE_CFG_WDATA 0x470 > +#define PCIE_APP_TLP_REQ 0x488 > +#define PCIE_CFG_RDATA 0x48c > +#define APP_CFG_REQ BIT(0) > +#define APP_CPL_STATUS GENMASK(7, 5) > + > +#define CFG_WRRD_TYPE_0 4 > +#define CFG_WR_FMT 2 > +#define CFG_RD_FMT 0 > + > +#define CFG_DW0_LENGTH(length) ((length) & GENMASK(9, 0)) > +#define CFG_DW0_TYPE(type) (((type) << 24) & GENMASK(28, 24)) > +#define CFG_DW0_FMT(fmt) (((fmt) << 29) & GENMASK(31, 29)) > +#define CFG_DW2_REGN(regn) ((regn) & GENMASK(11, 2)) > +#define CFG_DW2_FUN(fun) (((fun) << 16) & GENMASK(18, 16)) > +#define CFG_DW2_DEV(dev) (((dev) << 19) & GENMASK(23, 19)) > +#define CFG_DW2_BUS(bus) (((bus) << 24) & GENMASK(31, 24)) > +#define CFG_HEADER_DW0(type, fmt) \ > + (CFG_DW0_LENGTH(1) | CFG_DW0_TYPE(type) | CFG_DW0_FMT(fmt)) > +#define CFG_HEADER_DW1(where, size) \ > + (GENMASK(((size) - 1), 0) << ((where) & 0x3)) > +#define CFG_HEADER_DW2(regn, fun, dev, bus) \ > + (CFG_DW2_REGN(regn) | CFG_DW2_FUN(fun) | \ > + CFG_DW2_DEV(dev) | CFG_DW2_BUS(bus)) > + > +#define PCIE_RST_CTRL 0x510 > +#define PCIE_PHY_RSTB BIT(0) > +#define PCIE_PIPE_SRSTB BIT(1) > +#define PCIE_MAC_SRSTB BIT(2) > +#define PCIE_CRSTB BIT(3) > +#define PCIE_PERSTB BIT(8) > +#define PCI_LINKDOWN_RST_EN GENMASK(15, 13) > +#define PCIE_LINK_STATUS_V2 0x804 > +#define PCIE_PORT_LINKUP_V2 BIT(10) > + > struct mtk_pcie_port; > > /** > * struct mtk_pcie_soc - differentiate between host generations > * @ops: pointer to configuration access functions > * @startup: pointer to controller setting functions > + * @setup_irq: pointer to initialize irq functions s/ irq / IRQ / > */ > struct mtk_pcie_soc { > struct pci_ops *ops; > int (*startup)(struct mtk_pcie_port *port); > + int (*setup_irq)(struct mtk_pcie_port *port, struct device_node *node); > }; > > /** > @@ -82,10 +146,18 @@ struct mtk_pcie_soc { > * @list: port list > * @pcie: pointer to PCIe host info > * @reset: pointer to port reset control > - * @sys_ck: pointer to bus clock > + * @sys_ck: pointer to transaction/data link layer clock > + * @ahb_ck: pinter to AHB slave interface operating clock for CSR access s/pinter/pointer/ > + * and RC initiated MMIO access > + * @axi_ck: pointer to application layer MMIO channel operating clock > + * @aux_ck: pointer to pe2_mac_bridge and pe2_mac_core operating clock > + * when pcie_mac_ck/pcie_pipe_ck is turned off > + * @obff_ck: pointer to OBFF functional block operating clock > + * @pipe_ck: pointer to LTSSM and phy/mac layer operating clock s/phy/PHY/ s/mac/MAC/ > * @phy: pointer to phy control block s/ phy / PHY / > * @lane: lane count > * @index: port index > + * @irq_domain: legacy intx irq domain s/intx irq/INTx IRQ/ > */ > struct mtk_pcie_port { > void __iomem *base; > @@ -93,9 +165,15 @@ struct mtk_pcie_port { > struct mtk_pcie *pcie; > struct reset_control *reset; > struct clk *sys_ck; > + struct clk *ahb_ck; > + struct clk *axi_ck; > + struct clk *aux_ck; > + struct clk *obff_ck; > + struct clk *pipe_ck; > struct phy *phy; > u32 lane; > u32 index; > + struct irq_domain *irq_domain; > }; > > /** > @@ -156,6 +234,12 @@ static void mtk_pcie_put_resources(struct mtk_pcie *pcie) > > list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > phy_power_off(port->phy); > + phy_exit(port->phy); > + clk_disable_unprepare(port->pipe_ck); > + clk_disable_unprepare(port->obff_ck); > + clk_disable_unprepare(port->axi_ck); > + clk_disable_unprepare(port->aux_ck); > + clk_disable_unprepare(port->ahb_ck); > clk_disable_unprepare(port->sys_ck); > mtk_pcie_port_free(port); > } > @@ -163,6 +247,298 @@ static void mtk_pcie_put_resources(struct mtk_pcie *pcie) > mtk_pcie_subsys_powerdown(pcie); > } > > +static int mtk_pcie_check_cfg_cpld(struct mtk_pcie_port *port) > +{ > + u32 val; > + int err; > + > + err = readl_poll_timeout_atomic(port->base + PCIE_APP_TLP_REQ, val, > + !(val & APP_CFG_REQ), 10, > + 100 * USEC_PER_MSEC); > + if (err) > + return PCIBIOS_SET_FAILED; > + > + if (readl(port->base + PCIE_APP_TLP_REQ) & APP_CPL_STATUS) > + return PCIBIOS_SET_FAILED; > + > + return PCIBIOS_SUCCESSFUL; > +} > + > +static int mtk_pcie_hw_rd_cfg(struct mtk_pcie_port *port, u32 bus, u32 devfn, > + int where, int size, u32 *val) > +{ > + int reg, shift = 8 * (where & 3); > + > + /* Write PCIe Configuration Transaction Header for cfgrd */ > + writel(CFG_HEADER_DW0(CFG_WRRD_TYPE_0, CFG_RD_FMT), > + port->base + PCIE_CFG_HEADER0); > + writel(CFG_HEADER_DW1(where, size), port->base + PCIE_CFG_HEADER1); > + writel(CFG_HEADER_DW2(where, PCI_FUNC(devfn), PCI_SLOT(devfn), bus), > + port->base + PCIE_CFG_HEADER2); > + > + /* Triget h/w to transmit Cfgrd TLP */ s/Triget/Trigger/ > + reg = readl(port->base + PCIE_APP_TLP_REQ); > + writel(reg | APP_CFG_REQ, port->base + PCIE_APP_TLP_REQ); > + > + /* Check complete condition */ s/complete condition/completion status/ ? > + if (mtk_pcie_check_cfg_cpld(port)) > + return PCIBIOS_SET_FAILED; > + > + /* Read cpld payload of Cfgrd */ > + *val = readl(port->base + PCIE_CFG_RDATA); > + > + switch (size) { > + case 4: > + break; > + case 3: > + *val = (*val >> shift) & 0xffffff; > + break; > + case 2: > + *val = (*val >> shift) & 0xffff; > + break; > + case 1: > + *val = (*val >> shift) & 0xff; > + break; > + default: > + return PCIBIOS_BAD_REGISTER_NUMBER; > + } > + > + return PCIBIOS_SUCCESSFUL; > +} > + > +static int mtk_pcie_hw_wr_cfg(struct mtk_pcie_port *port, u32 bus, u32 devfn, > + int where, int size, u32 val) > +{ > + /* Write PCIe Configuration Transaction Header for Cfgwr */ > + writel(CFG_HEADER_DW0(CFG_WRRD_TYPE_0, CFG_WR_FMT), > + port->base + PCIE_CFG_HEADER0); > + writel(CFG_HEADER_DW1(where, size), port->base + PCIE_CFG_HEADER1); > + writel(CFG_HEADER_DW2(where, PCI_FUNC(devfn), PCI_SLOT(devfn), bus), > + port->base + PCIE_CFG_HEADER2); > + > + /* Write cfgwr data */ > + val = val << 8 * (where & 3); > + writel(val, port->base + PCIE_CFG_WDATA); > + > + /* Triget h/w to transmit Cfgwr TLP */ s/Triget/Trigger/ > + val = readl(port->base + PCIE_APP_TLP_REQ); > + val |= APP_CFG_REQ; > + writel(val, port->base + PCIE_APP_TLP_REQ); > + > + /* Check complete condition */ s/complete condition/completion status/ ? > + return mtk_pcie_check_cfg_cpld(port); > +} > + > +static struct mtk_pcie_port *mtk_pcie_find_port(struct mtk_pcie *pcie, > + struct pci_bus *bus, int devfn) > +{ > + struct pci_dev *dev; > + struct pci_bus *pbus; > + struct mtk_pcie_port *port, *tmp; > + > + list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > + if (bus->number == 0 && port->index == PCI_SLOT(devfn)) { > + return port; > + } else if (bus->number != 0) { > + pbus = bus; > + do { > + dev = pbus->self; > + if (port->index == PCI_SLOT(dev->devfn)) > + return port; > + pbus = dev->bus; > + } while (dev->bus->number != 0); > + } > + } > + > + return NULL; You should be able to use sysdata to avoid searching the list. See drivers/pci/host/pci-aardvark.c, for example. > +} > + > +static int mtk_pcie_config_read(struct pci_bus *bus, unsigned int devfn, > + int where, int size, u32 *val) > +{ > + struct mtk_pcie_port *port; > + struct pci_host_bridge *host = pci_find_host_bridge(bus); > + struct mtk_pcie *pcie = pci_host_bridge_priv(host); Sysdata should make this very simple; see advk_pcie_rd_conf(). > + u32 bn = bus->number; > + int ret; > + > + port = mtk_pcie_find_port(pcie, bus, devfn); > + if (!port) { > + *val = ~0; > + return PCIBIOS_DEVICE_NOT_FOUND; > + } > + > + ret = mtk_pcie_hw_rd_cfg(port, bn, devfn, where, size, val); > + if (ret) > + *val = ~0; > + > + return ret; > +} > + > +static int mtk_pcie_config_write(struct pci_bus *bus, unsigned int devfn, > + int where, int size, u32 val) > +{ > + u32 bn = bus->number; > + struct pci_host_bridge *host = pci_find_host_bridge(bus); > + struct mtk_pcie *pcie = pci_host_bridge_priv(host); > + struct mtk_pcie_port *port; > + > + port = mtk_pcie_find_port(pcie, bus, devfn); > + if (!port) > + return PCIBIOS_DEVICE_NOT_FOUND; > + > + return mtk_pcie_hw_wr_cfg(port, bn, devfn, where, size, val); > +} > + > +static struct pci_ops mtk_pcie_ops_v2 = { > + .read = mtk_pcie_config_read, > + .write = mtk_pcie_config_write, > +}; > + > +static int mtk_pcie_startup_ports_v2(struct mtk_pcie_port *port) > +{ > + struct mtk_pcie *pcie = port->pcie; > + struct resource *mem = &pcie->mem; > + u32 val; > + size_t size; > + int err; > + > + /* mt7622 platforms need to enable LTSSM and ASPM from PCIe subsys */ > + if (pcie->base) { > + val = readl(pcie->base + PCIE_SYS_CFG_V2); > + val |= PCIE_CSR_LTSSM_EN(port->index) | > + PCIE_CSR_ASPM_L1_EN(port->index); > + writel(val, pcie->base + PCIE_SYS_CFG_V2); > + } > + > + /* Assert all reset signals */ > + writel(0, port->base + PCIE_RST_CTRL); > + > + /* > + * Enable rc internal reset. > + * The reset will work when the link is from link up to link down. ? That sentence doesn't parse for me. > + */ > + writel(PCI_LINKDOWN_RST_EN, port->base + PCIE_RST_CTRL); > + > + /* De-assert phy, pe, pipe, mac and configuration reset */ > + val = readl(port->base + PCIE_RST_CTRL); > + val |= PCIE_PHY_RSTB | PCIE_PERSTB | PCIE_PIPE_SRSTB | > + PCIE_MAC_SRSTB | PCIE_CRSTB; > + writel(val, port->base + PCIE_RST_CTRL); > + > + /* PCIe v2.0 need at least 100ms delay to train from Gen1 to Gen2 */ > + err = readl_poll_timeout(port->base + PCIE_LINK_STATUS_V2, val, > + !!(val & PCIE_PORT_LINKUP_V2), 20, > + 100 * USEC_PER_MSEC); > + if (err) > + return -ETIMEDOUT; > + > + /* Set INTx mask */ > + val = readl(port->base + PCIE_INT_MASK); > + val &= ~INTX_MASK; > + writel(val, port->base + PCIE_INT_MASK); > + > + /* Set AHB to PCIe translation windows */ > + size = mem->end - mem->start; > + val = AHB2PCIE_BASEL(mem->start) | AHB2PCIE_SIZE(fls(size)); > + writel(val, port->base + PCIE_AHB_TRANS_BASE0_L); > + > + val = AHB2PCIE_BASEH(mem->start); > + writel(val, port->base + PCIE_AHB_TRANS_BASE0_H); > + > + /* Set PCIe to axi translation memory space.*/ s/axi/AXI/ > + val = fls(0xffffffff) | WIN_ENABLE; > + writel(val, port->base + PCIE_AXI_WINDOW0); > + > + return 0; > +} > + > +static int mtk_pcie_intx_map(struct irq_domain *domain, unsigned int irq, > + irq_hw_number_t hwirq) > +{ > + irq_set_chip_and_handler(irq, &dummy_irq_chip, handle_simple_irq); > + irq_set_chip_data(irq, domain->host_data); > + > + return 0; > +} > + > +static const struct irq_domain_ops intx_domain_ops = { > + .map = mtk_pcie_intx_map, > +}; > + > +static int mtk_pcie_init_irq_domain(struct mtk_pcie_port *port, > + struct device_node *node) > +{ > + struct device *dev = port->pcie->dev; > + struct device_node *pcie_intc_node; > + > + /* Setup INTx */ > + pcie_intc_node = of_get_next_child(node, NULL); > + if (!pcie_intc_node) { > + dev_err(dev, "No PCIe Intc node found\n"); > + return PTR_ERR(pcie_intc_node); > + } > + > + port->irq_domain = irq_domain_add_linear(pcie_intc_node, INTX_NUM, > + &intx_domain_ops, port); I think there's an issue here with a 4-element IRQ domain and the hwirq numbers 1-4 from the of_irq_parse_and_map_pci() path, so INTD may not work correctly. See http://lkml.kernel.org/r/20170801212931.GA26498@bhelgaas-glaptop.roam.corp.google.com and related discussion. > + if (!port->irq_domain) { > + dev_err(dev, "Failed to get INTx IRQ domain\n"); > + return PTR_ERR(port->irq_domain); > + } > + > + return 0; > +} > + > +static irqreturn_t mtk_pcie_intr_handler(int irq, void *data) > +{ > + struct mtk_pcie_port *port = (struct mtk_pcie_port *)data; > + struct device *dev = port->pcie->dev; > + unsigned long status; > + u32 virq; > + u32 bit = INTX_SHIFT; > + > + while ((status = readl(port->base + PCIE_INT_STATUS)) & INTX_MASK) { > + for_each_set_bit_from(bit, &status, INTX_NUM + INTX_SHIFT) { > + /* Clear the INTx */ > + writel(1 << bit, port->base + PCIE_INT_STATUS); > + virq = irq_find_mapping(port->irq_domain, > + bit - INTX_SHIFT); > + if (virq) > + generic_handle_irq(virq); > + else > + dev_err(dev, "unexpected IRQ, INT%d\n", > + bit - INTX_SHIFT); PCI INTx are conventionally INTA, INTB, INTC, INTD (not INT1, INT2, etc). > + } > + } > + > + return IRQ_HANDLED; > +} > + > +static int mtk_pcie_setup_irq(struct mtk_pcie_port *port, > + struct device_node *node) > +{ > + struct mtk_pcie *pcie = port->pcie; > + struct device *dev = pcie->dev; > + struct platform_device *pdev = to_platform_device(dev); > + int err, irq; > + > + irq = platform_get_irq(pdev, port->index); > + err = devm_request_irq(dev, irq, mtk_pcie_intr_handler, > + IRQF_SHARED, "mtk-pcie", port); > + if (err) { > + dev_err(dev, "unable to request irq %d\n", irq); s/irq/IRQ/ > + return err; > + } > + > + err = mtk_pcie_init_irq_domain(port, node); > + if (err) { > + dev_err(dev, "failed to init pcie lagecy irq domain\n"); s/lagecy/legacy/ s/irq/IRQ/ s/pcie/PCIe/ > + return -ENODEV; > + } > + > + return 0; > +} > + > static void __iomem *mtk_pcie_map_bus(struct pci_bus *bus, > unsigned int devfn, int where) > { > @@ -249,13 +625,49 @@ static void mtk_pcie_enable_ports(struct mtk_pcie_port *port) > > err = clk_prepare_enable(port->sys_ck); > if (err) { > - dev_err(dev, "failed to enable port%d clock\n", port->index); > + dev_err(dev, "failed to enable sys_ck%d\n", port->index); > goto err_sys_clk; > } > > + err = clk_prepare_enable(port->ahb_ck); > + if (err) { > + dev_err(dev, "failed to enable ahb_ck%d\n", port->index); > + goto err_ahb_clk; > + } > + > + err = clk_prepare_enable(port->aux_ck); > + if (err) { > + dev_err(dev, "failed to enable aux_ck%d\n", port->index); > + goto err_aux_clk; > + } > + > + err = clk_prepare_enable(port->axi_ck); > + if (err) { > + dev_err(dev, "failed to enable axi_ck%d\n", port->index); > + goto err_axi_clk; > + } > + > + err = clk_prepare_enable(port->obff_ck); > + if (err) { > + dev_err(dev, "failed to enable obff_ck%d\n", port->index); > + goto err_obff_clk; > + } > + > + err = clk_prepare_enable(port->pipe_ck); > + if (err) { > + dev_err(dev, "failed to enable pipe_ck%d\n", port->index); > + goto err_pipe_clk; > + } > + > reset_control_assert(port->reset); > reset_control_deassert(port->reset); > > + err = phy_init(port->phy); > + if (err) { > + dev_err(dev, "failed to initialize port%d phy\n", port->index); > + goto err_phy_init; > + } > + > err = phy_power_on(port->phy); > if (err) { > dev_err(dev, "failed to power on port%d phy\n", port->index); > @@ -269,6 +681,18 @@ static void mtk_pcie_enable_ports(struct mtk_pcie_port *port) > > phy_power_off(port->phy); > err_phy_on: > + phy_exit(port->phy); > +err_phy_init: > + clk_disable_unprepare(port->pipe_ck); > +err_pipe_clk: > + clk_disable_unprepare(port->obff_ck); > +err_obff_clk: > + clk_disable_unprepare(port->axi_ck); > +err_axi_clk: > + clk_disable_unprepare(port->aux_ck); > +err_aux_clk: > + clk_disable_unprepare(port->ahb_ck); > +err_ahb_clk: > clk_disable_unprepare(port->sys_ck); > err_sys_clk: > mtk_pcie_port_free(port); > @@ -306,10 +730,56 @@ static int mtk_pcie_parse_ports(struct mtk_pcie *pcie, > snprintf(name, sizeof(name), "sys_ck%d", index); > port->sys_ck = devm_clk_get(dev, name); > if (IS_ERR(port->sys_ck)) { > - dev_err(dev, "failed to get port%d clock\n", index); > + dev_err(dev, "failed to get sys_ck%d\n", index); > return PTR_ERR(port->sys_ck); > } > > + /* sys_ck might be divided into the following parts in some chips */ > + snprintf(name, sizeof(name), "ahb_ck%d", index); > + port->ahb_ck = devm_clk_get(dev, name); > + if (IS_ERR(port->ahb_ck)) { > + if (PTR_ERR(port->ahb_ck) == -EPROBE_DEFER) > + return -EPROBE_DEFER; > + > + port->ahb_ck = NULL; > + } > + > + snprintf(name, sizeof(name), "axi_ck%d", index); > + port->axi_ck = devm_clk_get(dev, name); > + if (IS_ERR(port->axi_ck)) { > + if (PTR_ERR(port->axi_ck) == -EPROBE_DEFER) > + return -EPROBE_DEFER; > + > + port->axi_ck = NULL; > + } > + > + snprintf(name, sizeof(name), "aux_ck%d", index); > + port->aux_ck = devm_clk_get(dev, name); > + if (IS_ERR(port->aux_ck)) { > + if (PTR_ERR(port->aux_ck) == -EPROBE_DEFER) > + return -EPROBE_DEFER; > + > + port->aux_ck = NULL; > + } > + > + snprintf(name, sizeof(name), "obff_ck%d", index); > + port->obff_ck = devm_clk_get(dev, name); > + if (IS_ERR(port->obff_ck)) { > + if (PTR_ERR(port->obff_ck) == -EPROBE_DEFER) > + return -EPROBE_DEFER; > + > + port->obff_ck = NULL; > + } > + > + snprintf(name, sizeof(name), "pipe_ck%d", index); > + port->pipe_ck = devm_clk_get(dev, name); > + if (IS_ERR(port->pipe_ck)) { > + if (PTR_ERR(port->pipe_ck) == -EPROBE_DEFER) > + return -EPROBE_DEFER; > + > + port->pipe_ck = NULL; > + } > + > snprintf(name, sizeof(name), "pcie-rst%d", index); > port->reset = devm_reset_control_get_optional(dev, name); > if (PTR_ERR(port->reset) == -EPROBE_DEFER) > @@ -324,6 +794,12 @@ static int mtk_pcie_parse_ports(struct mtk_pcie *pcie, > port->index = index; > port->pcie = pcie; > > + if (pcie->soc->setup_irq) { > + err = pcie->soc->setup_irq(port, node); > + if (err) > + return err; > + } > + > INIT_LIST_HEAD(&port->list); > list_add_tail(&port->list, &pcie->ports); > > @@ -553,9 +1029,17 @@ static struct mtk_pcie_soc mtk_pcie_soc_v1 = { > .startup = mtk_pcie_startup_ports, > }; > > +static struct mtk_pcie_soc mtk_pcie_soc_v2 = { > + .ops = &mtk_pcie_ops_v2, > + .startup = mtk_pcie_startup_ports_v2, > + .setup_irq = mtk_pcie_setup_irq, > +}; > + > static const struct of_device_id mtk_pcie_ids[] = { > { .compatible = "mediatek,mt2701-pcie", .data = &mtk_pcie_soc_v1 }, > { .compatible = "mediatek,mt7623-pcie", .data = &mtk_pcie_soc_v1 }, > + { .compatible = "mediatek,mt2712-pcie", .data = &mtk_pcie_soc_v2 }, > + { .compatible = "mediatek,mt7622-pcie", .data = &mtk_pcie_soc_v2 }, > {}, > }; > > -- > 2.6.4 >
Hi, Bjorn, Thanks very much for your reviews. The mis-spells will be fixed in next version. On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: ...... > > +} > > + > > +static struct mtk_pcie_port *mtk_pcie_find_port(struct mtk_pcie *pcie, > > + struct pci_bus *bus, int devfn) > > +{ > > + struct pci_dev *dev; > > + struct pci_bus *pbus; > > + struct mtk_pcie_port *port, *tmp; > > + > > + list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > + if (bus->number == 0 && port->index == PCI_SLOT(devfn)) { > > + return port; > > + } else if (bus->number != 0) { > > + pbus = bus; > > + do { > > + dev = pbus->self; > > + if (port->index == PCI_SLOT(dev->devfn)) > > + return port; > > + pbus = dev->bus; > > + } while (dev->bus->number != 0); > > + } > > + } > > + > > + return NULL; > > You should be able to use sysdata to avoid searching the list. > See drivers/pci/host/pci-aardvark.c, for example. > I could put the mtk_pcie * in sysdata, but still need to searching the list to get the mtk_pcie_port *, how about: list_for_each_entry_safe(port, tmp, &pcie->ports, list) { if (port->index == PCI_SLOT(devfn)) return port; } > > +} > > + > > +static int mtk_pcie_config_read(struct pci_bus *bus, unsigned int devfn, > > + int where, int size, u32 *val) > > +{ > > + struct mtk_pcie_port *port; > > + struct pci_host_bridge *host = pci_find_host_bridge(bus); > > + struct mtk_pcie *pcie = pci_host_bridge_priv(host); > > Sysdata should make this very simple; see advk_pcie_rd_conf(). thanks. > > > + u32 bn = bus->number; > > + int ret; > > + > > + port = mtk_pcie_find_port(pcie, bus, devfn); > > + if (!port) { > > + *val = ~0; > > + return PCIBIOS_DEVICE_NOT_FOUND; > > + } > > + > > + ret = mtk_pcie_hw_rd_cfg(port, bn, devfn, where, size, val); > > + if (ret) > > + *val = ~0; > > + > > + return ret; > > +} > > + > > +static int mtk_pcie_config_write(struct pci_bus *bus, unsigned int devfn, > > + int where, int size, u32 val) > > +{ > > + u32 bn = bus->number; > > + struct pci_host_bridge *host = pci_find_host_bridge(bus); > > + struct mtk_pcie *pcie = pci_host_bridge_priv(host); > > + struct mtk_pcie_port *port; > > + > > + port = mtk_pcie_find_port(pcie, bus, devfn); > > + if (!port) > > + return PCIBIOS_DEVICE_NOT_FOUND; > > + > > + return mtk_pcie_hw_wr_cfg(port, bn, devfn, where, size, val); > > +} > > + > > +static struct pci_ops mtk_pcie_ops_v2 = { > > + .read = mtk_pcie_config_read, > > + .write = mtk_pcie_config_write, > > +}; > > + > > +static int mtk_pcie_startup_ports_v2(struct mtk_pcie_port *port) > > +{ > > + struct mtk_pcie *pcie = port->pcie; > > + struct resource *mem = &pcie->mem; > > + u32 val; > > + size_t size; > > + int err; > > + > > + /* mt7622 platforms need to enable LTSSM and ASPM from PCIe subsys */ > > + if (pcie->base) { > > + val = readl(pcie->base + PCIE_SYS_CFG_V2); > > + val |= PCIE_CSR_LTSSM_EN(port->index) | > > + PCIE_CSR_ASPM_L1_EN(port->index); > > + writel(val, pcie->base + PCIE_SYS_CFG_V2); > > + } > > + > > + /* Assert all reset signals */ > > + writel(0, port->base + PCIE_RST_CTRL); > > + > > + /* > > + * Enable rc internal reset. > > + * The reset will work when the link is from link up to link down. > > ? That sentence doesn't parse for me. What about: /* * Enable PCIe link down reset, if link status changed from link up to * link down, this will reset MAC control registers and configuration * space. */ > > > + */ > > + writel(PCI_LINKDOWN_RST_EN, port->base + PCIE_RST_CTRL); > > + > > + /* De-assert phy, pe, pipe, mac and configuration reset */ > > + val = readl(port->base + PCIE_RST_CTRL); > > + val |= PCIE_PHY_RSTB | PCIE_PERSTB | PCIE_PIPE_SRSTB | > > + PCIE_MAC_SRSTB | PCIE_CRSTB; > > + writel(val, port->base + PCIE_RST_CTRL); > > + > > + /* PCIe v2.0 need at least 100ms delay to train from Gen1 to Gen2 */ > > + err = readl_poll_timeout(port->base + PCIE_LINK_STATUS_V2, val, > > + !!(val & PCIE_PORT_LINKUP_V2), 20, > > + 100 * USEC_PER_MSEC); > > + if (err) > > + return -ETIMEDOUT; > > + > > + /* Set INTx mask */ > > + val = readl(port->base + PCIE_INT_MASK); > > + val &= ~INTX_MASK; > > + writel(val, port->base + PCIE_INT_MASK); > > + > > + /* Set AHB to PCIe translation windows */ > > + size = mem->end - mem->start; > > + val = AHB2PCIE_BASEL(mem->start) | AHB2PCIE_SIZE(fls(size)); > > + writel(val, port->base + PCIE_AHB_TRANS_BASE0_L); > > + > > + val = AHB2PCIE_BASEH(mem->start); > > + writel(val, port->base + PCIE_AHB_TRANS_BASE0_H); > > + > > + /* Set PCIe to axi translation memory space.*/ > > s/axi/AXI/ > > > + val = fls(0xffffffff) | WIN_ENABLE; > > + writel(val, port->base + PCIE_AXI_WINDOW0); > > + > > + return 0; > > +} > > + > > +static int mtk_pcie_intx_map(struct irq_domain *domain, unsigned int irq, > > + irq_hw_number_t hwirq) > > +{ > > + irq_set_chip_and_handler(irq, &dummy_irq_chip, handle_simple_irq); > > + irq_set_chip_data(irq, domain->host_data); > > + > > + return 0; > > +} > > + > > +static const struct irq_domain_ops intx_domain_ops = { > > + .map = mtk_pcie_intx_map, > > +}; > > + > > +static int mtk_pcie_init_irq_domain(struct mtk_pcie_port *port, > > + struct device_node *node) > > +{ > > + struct device *dev = port->pcie->dev; > > + struct device_node *pcie_intc_node; > > + > > + /* Setup INTx */ > > + pcie_intc_node = of_get_next_child(node, NULL); > > + if (!pcie_intc_node) { > > + dev_err(dev, "No PCIe Intc node found\n"); > > + return PTR_ERR(pcie_intc_node); > > + } > > + > > + port->irq_domain = irq_domain_add_linear(pcie_intc_node, INTX_NUM, > > + &intx_domain_ops, port); > > I think there's an issue here with a 4-element IRQ domain and the > hwirq numbers 1-4 from the of_irq_parse_and_map_pci() path, so INTD > may not work correctly. > > See > http://lkml.kernel.org/r/20170801212931.GA26498@bhelgaas-glaptop.roam.corp.google.com > and related discussion. > Sorry, I did not get this, I do some test with an intel E350T4 PCIe NICs, it's a x1 lane multi-function device. What I got from the log is below: ->of_irq_parse_and_map_pci ->of_irq_parse_pci ->irq_create_of_mapping ->irq_create_fwspec_mapping ->irq_domain_translate which will go through d->ops->translate #the hwirq really start from 0 And I tested every NIC port of the Intel E350T4 with tftp transfer data, seems all are OK with this code. What I got from the proc is as below: cat /proc/interrupts CPU0 CPU1 CPU2 1: 0 0 0 GICv2 25 Level vgic 3: 5042 224 206 GICv2 30 Level arch_timer 4: 0 0 0 GICv2 27 Level kvm guest timer 6: 201 0 0 MT_SYSIRQ 91 Level ttyS0 7: 57 0 0 MT_SYSIRQ 115 Level mtk-pcie 8: 0 0 0 MT_SYSIRQ 117 Level mtk-pcie 9: 9 0 0 dummy 0 Edge eth0 10: 40 0 0 dummy 1 Edge eth1 11: 5 0 0 dummy 2 Edge eth2 12: 3 0 0 dummy 3 Edge eth3 IPI0: 314 507 1164 Rescheduling interrupts or did I missed something? > > + if (!port->irq_domain) { > > + dev_err(dev, "Failed to get INTx IRQ domain\n"); > > + return PTR_ERR(port->irq_domain); > > + } > > + > > + return 0; > > +} > > + > > +static irqreturn_t mtk_pcie_intr_handler(int irq, void *data) > > +{ > > + struct mtk_pcie_port *port = (struct mtk_pcie_port *)data; > > + struct device *dev = port->pcie->dev; > > + unsigned long status; > > + u32 virq; > > + u32 bit = INTX_SHIFT; > > + > > + while ((status = readl(port->base + PCIE_INT_STATUS)) & INTX_MASK) { > > + for_each_set_bit_from(bit, &status, INTX_NUM + INTX_SHIFT) { > > + /* Clear the INTx */ > > + writel(1 << bit, port->base + PCIE_INT_STATUS); > > + virq = irq_find_mapping(port->irq_domain, > > + bit - INTX_SHIFT); > > + if (virq) > > + generic_handle_irq(virq); > > + else > > + dev_err(dev, "unexpected IRQ, INT%d\n", > > + bit - INTX_SHIFT); > > PCI INTx are conventionally INTA, INTB, INTC, INTD (not INT1, INT2, > etc). > > > + } > > + } > > + > > + return IRQ_HANDLED; > > +} > > + > > +static int mtk_pcie_setup_irq(struct mtk_pcie_port *port, > > + struct device_node *node) > > +{ > > + struct mtk_pcie *pcie = port->pcie; > > + struct device *dev = pcie->dev; > > + struct platform_device *pdev = to_platform_device(dev); > > + int err, irq; > > + > > + irq = platform_get_irq(pdev, port->index); > > + err = devm_request_irq(dev, irq, mtk_pcie_intr_handler, > > + IRQF_SHARED, "mtk-pcie", port); > > + if (err) { > > + dev_err(dev, "unable to request irq %d\n", irq); > > s/irq/IRQ/ > > > + return err; > > + } > > + > > + err = mtk_pcie_init_irq_domain(port, node); > > + if (err) { > > + dev_err(dev, "failed to init pcie lagecy irq domain\n"); > > s/lagecy/legacy/ > s/irq/IRQ/ > s/pcie/PCIe/ > > > + return -ENODEV; > > + } > > + > > + return 0; > > +} > > + > > static void __iomem *mtk_pcie_map_bus(struct pci_bus *bus, > > unsigned int devfn, int where) > > { > > @@ -249,13 +625,49 @@ static void mtk_pcie_enable_ports(struct mtk_pcie_port *port) > > > > err = clk_prepare_enable(port->sys_ck); > > if (err) { > > - dev_err(dev, "failed to enable port%d clock\n", port->index); > > + dev_err(dev, "failed to enable sys_ck%d\n", port->index); > > goto err_sys_clk; > > } > > > > + err = clk_prepare_enable(port->ahb_ck); > > + if (err) { > > + dev_err(dev, "failed to enable ahb_ck%d\n", port->index); > > + goto err_ahb_clk; > > + } > > + > > + err = clk_prepare_enable(port->aux_ck); > > + if (err) { > > + dev_err(dev, "failed to enable aux_ck%d\n", port->index); > > + goto err_aux_clk; > > + } > > + > > + err = clk_prepare_enable(port->axi_ck); > > + if (err) { > > + dev_err(dev, "failed to enable axi_ck%d\n", port->index); > > + goto err_axi_clk; > > + } > > + > > + err = clk_prepare_enable(port->obff_ck); > > + if (err) { > > + dev_err(dev, "failed to enable obff_ck%d\n", port->index); > > + goto err_obff_clk; > > + } > > + > > + err = clk_prepare_enable(port->pipe_ck); > > + if (err) { > > + dev_err(dev, "failed to enable pipe_ck%d\n", port->index); > > + goto err_pipe_clk; > > + } > > + > > reset_control_assert(port->reset); > > reset_control_deassert(port->reset); > > > > + err = phy_init(port->phy); > > + if (err) { > > + dev_err(dev, "failed to initialize port%d phy\n", port->index); > > + goto err_phy_init; > > + } > > + > > err = phy_power_on(port->phy); > > if (err) { > > dev_err(dev, "failed to power on port%d phy\n", port->index); > > @@ -269,6 +681,18 @@ static void mtk_pcie_enable_ports(struct mtk_pcie_port *port) > > > > phy_power_off(port->phy); > > err_phy_on: > > + phy_exit(port->phy); > > +err_phy_init: > > + clk_disable_unprepare(port->pipe_ck); > > +err_pipe_clk: > > + clk_disable_unprepare(port->obff_ck); > > +err_obff_clk: > > + clk_disable_unprepare(port->axi_ck); > > +err_axi_clk: > > + clk_disable_unprepare(port->aux_ck); > > +err_aux_clk: > > + clk_disable_unprepare(port->ahb_ck); > > +err_ahb_clk: > > clk_disable_unprepare(port->sys_ck); > > err_sys_clk: > > mtk_pcie_port_free(port); > > @@ -306,10 +730,56 @@ static int mtk_pcie_parse_ports(struct mtk_pcie *pcie, > > snprintf(name, sizeof(name), "sys_ck%d", index); > > port->sys_ck = devm_clk_get(dev, name); > > if (IS_ERR(port->sys_ck)) { > > - dev_err(dev, "failed to get port%d clock\n", index); > > + dev_err(dev, "failed to get sys_ck%d\n", index); > > return PTR_ERR(port->sys_ck); > > } > > > > + /* sys_ck might be divided into the following parts in some chips */ > > + snprintf(name, sizeof(name), "ahb_ck%d", index); > > + port->ahb_ck = devm_clk_get(dev, name); > > + if (IS_ERR(port->ahb_ck)) { > > + if (PTR_ERR(port->ahb_ck) == -EPROBE_DEFER) > > + return -EPROBE_DEFER; > > + > > + port->ahb_ck = NULL; > > + } > > + > > + snprintf(name, sizeof(name), "axi_ck%d", index); > > + port->axi_ck = devm_clk_get(dev, name); > > + if (IS_ERR(port->axi_ck)) { > > + if (PTR_ERR(port->axi_ck) == -EPROBE_DEFER) > > + return -EPROBE_DEFER; > > + > > + port->axi_ck = NULL; > > + } > > + > > + snprintf(name, sizeof(name), "aux_ck%d", index); > > + port->aux_ck = devm_clk_get(dev, name); > > + if (IS_ERR(port->aux_ck)) { > > + if (PTR_ERR(port->aux_ck) == -EPROBE_DEFER) > > + return -EPROBE_DEFER; > > + > > + port->aux_ck = NULL; > > + } > > + > > + snprintf(name, sizeof(name), "obff_ck%d", index); > > + port->obff_ck = devm_clk_get(dev, name); > > + if (IS_ERR(port->obff_ck)) { > > + if (PTR_ERR(port->obff_ck) == -EPROBE_DEFER) > > + return -EPROBE_DEFER; > > + > > + port->obff_ck = NULL; > > + } > > + > > + snprintf(name, sizeof(name), "pipe_ck%d", index); > > + port->pipe_ck = devm_clk_get(dev, name); > > + if (IS_ERR(port->pipe_ck)) { > > + if (PTR_ERR(port->pipe_ck) == -EPROBE_DEFER) > > + return -EPROBE_DEFER; > > + > > + port->pipe_ck = NULL; > > + } > > + > > snprintf(name, sizeof(name), "pcie-rst%d", index); > > port->reset = devm_reset_control_get_optional(dev, name); > > if (PTR_ERR(port->reset) == -EPROBE_DEFER) > > @@ -324,6 +794,12 @@ static int mtk_pcie_parse_ports(struct mtk_pcie *pcie, > > port->index = index; > > port->pcie = pcie; > > > > + if (pcie->soc->setup_irq) { > > + err = pcie->soc->setup_irq(port, node); > > + if (err) > > + return err; > > + } > > + > > INIT_LIST_HEAD(&port->list); > > list_add_tail(&port->list, &pcie->ports); > > > > @@ -553,9 +1029,17 @@ static struct mtk_pcie_soc mtk_pcie_soc_v1 = { > > .startup = mtk_pcie_startup_ports, > > }; > > > > +static struct mtk_pcie_soc mtk_pcie_soc_v2 = { > > + .ops = &mtk_pcie_ops_v2, > > + .startup = mtk_pcie_startup_ports_v2, > > + .setup_irq = mtk_pcie_setup_irq, > > +}; > > + > > static const struct of_device_id mtk_pcie_ids[] = { > > { .compatible = "mediatek,mt2701-pcie", .data = &mtk_pcie_soc_v1 }, > > { .compatible = "mediatek,mt7623-pcie", .data = &mtk_pcie_soc_v1 }, > > + { .compatible = "mediatek,mt2712-pcie", .data = &mtk_pcie_soc_v2 }, > > + { .compatible = "mediatek,mt7622-pcie", .data = &mtk_pcie_soc_v2 }, > > {}, > > }; > > > > -- > > 2.6.4 > >
On Fri, Aug 04, 2017 at 04:39:36PM +0800, Honghui Zhang wrote: > On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: > > > ...... > > > +} > > > + > > > +static struct mtk_pcie_port *mtk_pcie_find_port(struct mtk_pcie *pcie, > > > + struct pci_bus *bus, int devfn) > > > +{ > > > + struct pci_dev *dev; > > > + struct pci_bus *pbus; > > > + struct mtk_pcie_port *port, *tmp; > > > + > > > + list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > > + if (bus->number == 0 && port->index == PCI_SLOT(devfn)) { > > > + return port; > > > + } else if (bus->number != 0) { > > > + pbus = bus; > > > + do { > > > + dev = pbus->self; > > > + if (port->index == PCI_SLOT(dev->devfn)) > > > + return port; > > > + pbus = dev->bus; > > > + } while (dev->bus->number != 0); > > > + } > > > + } > > > + > > > + return NULL; > > > > You should be able to use sysdata to avoid searching the list. > > See drivers/pci/host/pci-aardvark.c, for example. > > > > I could put the mtk_pcie * in sysdata, but still need to searching the > list to get the mtk_pcie_port *, how about: > > list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > if (port->index == PCI_SLOT(devfn)) > return port; > } No. Other drivers don't need to search the list. Please take a look at them and see how they solve this problem. I don't think your hardware is fundamentally different in a way that means you need to search when the others don't. > > > + * Enable rc internal reset. > > > + * The reset will work when the link is from link up to link down. > > > > ? That sentence doesn't parse for me. > > What about: > > /* > * Enable PCIe link down reset, if link status changed from link up to > * link down, this will reset MAC control registers and configuration > * space. > */ That at least parses as a sentence. > > > + port->irq_domain = irq_domain_add_linear(pcie_intc_node, INTX_NUM, > > > + &intx_domain_ops, port); > > > > I think there's an issue here with a 4-element IRQ domain and the > > hwirq numbers 1-4 from the of_irq_parse_and_map_pci() path, so INTD > > may not work correctly. > > > > See > > http://lkml.kernel.org/r/20170801212931.GA26498@bhelgaas-glaptop.roam.corp.google.com > > and related discussion. > > Sorry, I did not get this, > I do some test with an intel E350T4 PCIe NICs, it's a x1 lane > multi-function device. > What I got from the log is below: > ->of_irq_parse_and_map_pci > ->of_irq_parse_pci > ->irq_create_of_mapping > ->irq_create_fwspec_mapping > ->irq_domain_translate > which will go through > d->ops->translate #the hwirq really start from 0 > > And I tested every NIC port of the Intel E350T4 with tftp transfer data, > seems all are OK with this code. OK. I don't know what d->ops->translate is involved here, but if it works, I guess this is OK for now. We're trying to clean this up and make it consistent across all the drivers. Many of them allocate a 5-element IRQ domain, some make a 4-element domain, and on some of them INTD doesn't work. It's a mess. Bjorn
Hi Honghui, Bjorn, On Fri, 2017-08-04 at 08:18 -0500, Bjorn Helgaas wrote: > On Fri, Aug 04, 2017 at 04:39:36PM +0800, Honghui Zhang wrote: > > On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: > > > > + > > > > +static struct mtk_pcie_port *mtk_pcie_find_port(struct mtk_pcie *pcie, > > > > + struct pci_bus *bus, int devfn) > > > > +{ > > > > + struct pci_dev *dev; > > > > + struct pci_bus *pbus; > > > > + struct mtk_pcie_port *port, *tmp; > > > > + > > > > + list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > > > + if (bus->number == 0 && port->index == PCI_SLOT(devfn)) { > > > > + return port; > > > > + } else if (bus->number != 0) { > > > > + pbus = bus; > > > > + do { > > > > + dev = pbus->self; > > > > + if (port->index == PCI_SLOT(dev->devfn)) > > > > + return port; > > > > + pbus = dev->bus; > > > > + } while (dev->bus->number != 0); > > > > + } > > > > + } > > > > + > > > > + return NULL; > > > > > > You should be able to use sysdata to avoid searching the list. > > > See drivers/pci/host/pci-aardvark.c, for example. > > > > > > > I could put the mtk_pcie * in sysdata, but still need to searching the > > list to get the mtk_pcie_port *, how about: > > > > list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > if (port->index == PCI_SLOT(devfn)) > > return port; > > } > > No. Other drivers don't need to search the list. Please take a look > at them and see how they solve this problem. I don't think your > hardware is fundamentally different in a way that means you need to > search when the others don't. > I'm not directly involved in this generation, but I guess the main reason why Honghui need to do that is just because this hardware access configuration space via per-port registers, not just for the guard. Currently, We had a host bridge with two ports (two subnodes in binding text), thus he tried to tells them apart so that he can get the correct registers. Some platforms don't need to do that since they just have a single port (no more subnodes), the others might have specific/shared registers to access configuration space. (e.g. Tegra, MTK legacy IP block). Or, he can split them into two independent nodes, but it will break common probing flow by doing so. (I'd prefer to use subnodes.) Ryder
On Sat, 2017-08-05 at 12:52 +0800, Ryder Lee wrote: > Hi Honghui, Bjorn, > > On Fri, 2017-08-04 at 08:18 -0500, Bjorn Helgaas wrote: > > On Fri, Aug 04, 2017 at 04:39:36PM +0800, Honghui Zhang wrote: > > > On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: > > > > > + > > > > > +static struct mtk_pcie_port *mtk_pcie_find_port(struct mtk_pcie *pcie, > > > > > + struct pci_bus *bus, int devfn) > > > > > +{ > > > > > + struct pci_dev *dev; > > > > > + struct pci_bus *pbus; > > > > > + struct mtk_pcie_port *port, *tmp; > > > > > + > > > > > + list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > > > > + if (bus->number == 0 && port->index == PCI_SLOT(devfn)) { > > > > > + return port; > > > > > + } else if (bus->number != 0) { > > > > > + pbus = bus; > > > > > + do { > > > > > + dev = pbus->self; > > > > > + if (port->index == PCI_SLOT(dev->devfn)) > > > > > + return port; > > > > > + pbus = dev->bus; > > > > > + } while (dev->bus->number != 0); > > > > > + } > > > > > + } > > > > > + > > > > > + return NULL; > > > > > > > > You should be able to use sysdata to avoid searching the list. > > > > See drivers/pci/host/pci-aardvark.c, for example. > > > > > > > > > > I could put the mtk_pcie * in sysdata, but still need to searching the > > > list to get the mtk_pcie_port *, how about: > > > > > > list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > > if (port->index == PCI_SLOT(devfn)) > > > return port; > > > } > > > > No. Other drivers don't need to search the list. Please take a look > > at them and see how they solve this problem. I don't think your > > hardware is fundamentally different in a way that means you need to > > search when the others don't. > > > > I'm not directly involved in this generation, but I guess the main reason why Honghui need to do that is just because this hardware access configuration space via per-port registers, not just for the guard. > Currently, We had a host bridge with two ports (two subnodes in binding text), thus he tried to tells them apart so that he can get the correct registers. > > Some platforms don't need to do that since they just have a single port (no more subnodes), the others might have specific/shared registers to access configuration space. (e.g. Tegra, MTK legacy IP block). > Or, he can split them into two independent nodes, but it will break common probing flow by doing so. (I'd prefer to use subnodes.) > > Ryder > Sorry for the typesetting in previous mail and noise again, I've took a look at pci-rcar-gen2.c, this is a similar case I can found for Honghui's case. It gathers two ports reg regions into one, and uses the "slot id" to calculate the cfg base of each port. Perhaps this is a example for those who need to use subnodes and use port registers for cfg operation. Not sure whether it's worthwhile doing that since we need to changes ports/host structures. Ryder.
On Sat, 2017-08-05 at 14:16 +0800, Ryder Lee wrote: > On Sat, 2017-08-05 at 12:52 +0800, Ryder Lee wrote: > > Hi Honghui, Bjorn, > > > > On Fri, 2017-08-04 at 08:18 -0500, Bjorn Helgaas wrote: > > > On Fri, Aug 04, 2017 at 04:39:36PM +0800, Honghui Zhang wrote: > > > > On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: > > > > > > + > > > > > > +static struct mtk_pcie_port *mtk_pcie_find_port(struct mtk_pcie *pcie, > > > > > > + struct pci_bus *bus, int devfn) > > > > > > +{ > > > > > > + struct pci_dev *dev; > > > > > > + struct pci_bus *pbus; > > > > > > + struct mtk_pcie_port *port, *tmp; > > > > > > + > > > > > > + list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > > > > > + if (bus->number == 0 && port->index == PCI_SLOT(devfn)) { > > > > > > + return port; > > > > > > + } else if (bus->number != 0) { > > > > > > + pbus = bus; > > > > > > + do { > > > > > > + dev = pbus->self; > > > > > > + if (port->index == PCI_SLOT(dev->devfn)) > > > > > > + return port; > > > > > > + pbus = dev->bus; > > > > > > + } while (dev->bus->number != 0); > > > > > > + } > > > > > > + } > > > > > > + > > > > > > + return NULL; > > > > > > > > > > You should be able to use sysdata to avoid searching the list. > > > > > See drivers/pci/host/pci-aardvark.c, for example. > > > > > > > > > > > > > I could put the mtk_pcie * in sysdata, but still need to searching the > > > > list to get the mtk_pcie_port *, how about: > > > > > > > > list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > > > if (port->index == PCI_SLOT(devfn)) > > > > return port; > > > > } > > > > > > No. Other drivers don't need to search the list. Please take a look > > > at them and see how they solve this problem. I don't think your > > > hardware is fundamentally different in a way that means you need to > > > search when the others don't. > > > > > > > I'm not directly involved in this generation, but I guess the main reason why Honghui need to do that is just because this hardware access configuration space via per-port registers, not just for the guard. > > Currently, We had a host bridge with two ports (two subnodes in binding text), thus he tried to tells them apart so that he can get the correct registers. > > > > Some platforms don't need to do that since they just have a single port (no more subnodes), the others might have specific/shared registers to access configuration space. (e.g. Tegra, MTK legacy IP block). > > Or, he can split them into two independent nodes, but it will break common probing flow by doing so. (I'd prefer to use subnodes.) > > > > Ryder > > > > Sorry for the typesetting in previous mail and noise again, > > I've took a look at pci-rcar-gen2.c, this is a similar case I can found > for Honghui's case. It gathers two ports reg regions into one, and uses > the "slot id" to calculate the cfg base of each port. > > Perhaps this is a example for those who need to use subnodes and use > port registers for cfg operation. Not sure whether it's worthwhile doing > that since we need to changes ports/host structures. > > Ryder. > As Ryder's description, Mediatek's new generation HW blocks has two separate ports, they have separate control register base address. We must touch the per-port control register to access the EP's configuration space. One port's control register is the only way to access the EP's configuration space(the EP which is connect under this very port). Given an EP device, we need to determine which ports it's been connected, and get the base address for that port. It's a bit like pci-tegra/pci-mvebu. Seems list is not forbidden, pci-tegra search the list to identify the ports[1], mvebu use point array to search the ports[2], they have the same functionality through different approach. I may propose another patch to make the code like mvebu[2] if you insist, but I'm prefer the current list way. [1]http://elixir.free-electrons.com/linux/v4.13-rc4/source/drivers/pci/host/pci-tegra.c#L456 [2]http://elixir.free-electrons.com/linux/v4.13-rc4/source/drivers/pci/host/pci-mvebu.c#L780 thanks. >
On Sat, Aug 05, 2017 at 12:52:43PM +0800, Ryder Lee wrote: > Hi Honghui, Bjorn, > > On Fri, 2017-08-04 at 08:18 -0500, Bjorn Helgaas wrote: > > On Fri, Aug 04, 2017 at 04:39:36PM +0800, Honghui Zhang wrote: > > > On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: > > > > > + > > > > > +static struct mtk_pcie_port *mtk_pcie_find_port(struct mtk_pcie *pcie, > > > > > + struct pci_bus *bus, int devfn) > > > > > +{ > > > > > + struct pci_dev *dev; > > > > > + struct pci_bus *pbus; > > > > > + struct mtk_pcie_port *port, *tmp; > > > > > + > > > > > + list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > > > > + if (bus->number == 0 && port->index == PCI_SLOT(devfn)) { > > > > > + return port; > > > > > + } else if (bus->number != 0) { > > > > > + pbus = bus; > > > > > + do { > > > > > + dev = pbus->self; > > > > > + if (port->index == PCI_SLOT(dev->devfn)) > > > > > + return port; > > > > > + pbus = dev->bus; > > > > > + } while (dev->bus->number != 0); > > > > > + } > > > > > + } > > > > > + > > > > > + return NULL; > > > > > > > > You should be able to use sysdata to avoid searching the list. > > > > See drivers/pci/host/pci-aardvark.c, for example. > > > > > > > > > > I could put the mtk_pcie * in sysdata, but still need to searching the > > > list to get the mtk_pcie_port *, how about: > > > > > > list_for_each_entry_safe(port, tmp, &pcie->ports, list) { > > > if (port->index == PCI_SLOT(devfn)) > > > return port; > > > } > > > > No. Other drivers don't need to search the list. Please take a look > > at them and see how they solve this problem. I don't think your > > hardware is fundamentally different in a way that means you need to > > search when the others don't. > > > > I'm not directly involved in this generation, but I guess the main > reason why Honghui need to do that is just because this hardware > access configuration space via per-port registers, not just for the > guard. Currently, We had a host bridge with two ports (two subnodes > in binding text), thus he tried to tells them apart so that he can > get the correct registers. > > Some platforms don't need to do that since they just have a single > port (no more subnodes), the others might have specific/shared > registers to access configuration space. (e.g. Tegra, MTK legacy IP > block). Or, he can split them into two independent nodes, but it > will break common probing flow by doing so. (I'd prefer to use > subnodes.) The PCI core interface (pci_scan_root_bus_bridge()) starts with sysdata in the struct pci_host_bridge, so every PCI bus under that host bridge has the same sysdata. You have multiple root ports (sounds like two ports in this case) under that host bridge. Each port has independent interrupt mappings, but the MMIO address space routed to the ports is described in the upper-level host bridge (the "pcie" node in DT). I assume the I/O and MMIO routing through the root ports works as described in the PCI-to-PCI bridge spec, using PCI_MEMORY_BASE, PCI_PREF_MEMORY_BASE, etc. If that's the case, I think your current DT "pcie" node is appropriate, and both ports should have the same sysdata (as they do in your current patch), and you do need some additional way to get from that sysdata (the struct mtk_pcie) to the per-port data (the struct mtk_pcie_port). Apparently the two root ports of this MT7622/MT2712 controller are hardwired at device (PCI_SLOT) 0 and 1. Using a list to look that up seems like a little overkill, since you could index an array by PCI_SLOT(), but I guess you could do it either way. I would probably rename "port->index" to "port->slot" or similar to make it more obvious that it's not merely the Nth port we found; it's the one that is hardwired at PCI_SLOT N. Also, s/mtk_pcie_parse_ports/mtk_pcie_parse_port/, since it parses one port each time you call it. And factor out the lookup as Ryder suggested so it's not duplicated in the mtk_pcie_config_read()/mtk_pcie_config_write() paths. Bjorn
On Fri, Aug 04, 2017 at 08:18:09AM -0500, Bjorn Helgaas wrote: > On Fri, Aug 04, 2017 at 04:39:36PM +0800, Honghui Zhang wrote: > > On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: > > > > + port->irq_domain = irq_domain_add_linear(pcie_intc_node, INTX_NUM, > > > > + &intx_domain_ops, port); > > > > > > I think there's an issue here with a 4-element IRQ domain and the > > > hwirq numbers 1-4 from the of_irq_parse_and_map_pci() path, so INTD > > > may not work correctly. > > > > > > See > > > http://lkml.kernel.org/r/20170801212931.GA26498@bhelgaas-glaptop.roam.corp.google.com > > > and related discussion. > > > > Sorry, I did not get this, > > I do some test with an intel E350T4 PCIe NICs, it's a x1 lane > > multi-function device. > > What I got from the log is below: > > ->of_irq_parse_and_map_pci > > ->of_irq_parse_pci > > ->irq_create_of_mapping > > ->irq_create_fwspec_mapping > > ->irq_domain_translate > > which will go through > > d->ops->translate #the hwirq really start from 0 > > > > And I tested every NIC port of the Intel E350T4 with tftp transfer data, > > seems all are OK with this code. > > OK. I don't know what d->ops->translate is involved here, but if it > works, I guess this is OK for now. We're trying to clean this up and > make it consistent across all the drivers. Many of them allocate a > 5-element IRQ domain, some make a 4-element domain, and on some of > them INTD doesn't work. It's a mess. Paul Burton is cleaning this up. Can you point out the d->ops->translate function that's involved here?
On Tue, 2017-08-08 at 15:19 -0500, Bjorn Helgaas wrote: > On Fri, Aug 04, 2017 at 08:18:09AM -0500, Bjorn Helgaas wrote: > > On Fri, Aug 04, 2017 at 04:39:36PM +0800, Honghui Zhang wrote: > > > On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: > > > > > > + port->irq_domain = irq_domain_add_linear(pcie_intc_node, INTX_NUM, > > > > > + &intx_domain_ops, port); > > > > > > > > I think there's an issue here with a 4-element IRQ domain and the > > > > hwirq numbers 1-4 from the of_irq_parse_and_map_pci() path, so INTD > > > > may not work correctly. > > > > > > > > See > > > > http://lkml.kernel.org/r/20170801212931.GA26498@bhelgaas-glaptop.roam.corp.google.com > > > > and related discussion. > > > > > > Sorry, I did not get this, > > > I do some test with an intel E350T4 PCIe NICs, it's a x1 lane > > > multi-function device. > > > What I got from the log is below: > > > ->of_irq_parse_and_map_pci > > > ->of_irq_parse_pci > > > ->irq_create_of_mapping > > > ->irq_create_fwspec_mapping > > > ->irq_domain_translate > > > which will go through > > > d->ops->translate #the hwirq really start from 0 > > > > > > And I tested every NIC port of the Intel E350T4 with tftp transfer data, > > > seems all are OK with this code. > > > > OK. I don't know what d->ops->translate is involved here, but if it > > works, I guess this is OK for now. We're trying to clean this up and > > make it consistent across all the drivers. Many of them allocate a > > 5-element IRQ domain, some make a 4-element domain, and on some of > > them INTD doesn't work. It's a mess. > > Paul Burton is cleaning this up. Can you point out the d->ops->translate > function that's involved here? Hi, Bjorn, Sorry for my last reply, I was tracking the wrong logs. The real trick is here: ->of_irq_parse_and_map_pci ->of_irq_parse_pci #out_irq->args[0] start from 1(1 == INTA) ->of_irq_parse_raw After of_irq_parse_raw finished it's work, the out_irq->args[0] will be remapped as "interrupt-map" property defines[1], which in my case, it's start from 0, and then fwspec->param[0] is start from 0 (0 == INTA). My "interrupt-map" property is defined as below: interrupt-map = <0 0 0 1 &pcie_intc0 0>, <0 0 0 2 &pcie_intc0 1>, <0 0 0 3 &pcie_intc0 2>, <0 0 0 4 &pcie_intc0 3>; I do some test with the changes of property defined as below: interrupt-map = <0 0 0 1 &pcie_intc0 1>, <0 0 0 2 &pcie_intc0 2>, <0 0 0 3 &pcie_intc0 3>, <0 0 0 4 &pcie_intc0 4>; Then I got the same running complain as Paul have got[2] So I guess it's the "interrupt-map" property defined in dtsi node play the key role in this. [1]http://elixir.free-electrons.com/linux/v4.13-rc4/source/drivers/of/irq.c#L265 [2]https://patchwork.kernel.org/patch/9794355 thanks.
Hi Honghui & Bjorn, On Tuesday, 8 August 2017 23:49:52 PDT Honghui Zhang wrote: > On Tue, 2017-08-08 at 15:19 -0500, Bjorn Helgaas wrote: > > On Fri, Aug 04, 2017 at 08:18:09AM -0500, Bjorn Helgaas wrote: > > > On Fri, Aug 04, 2017 at 04:39:36PM +0800, Honghui Zhang wrote: > > > > On Thu, 2017-08-03 at 17:42 -0500, Bjorn Helgaas wrote: > > > > > > + port->irq_domain = irq_domain_add_linear(pcie_intc_node, > > > > > > INTX_NUM, > > > > > > + &intx_domain_ops, port); > > > > > > > > > > I think there's an issue here with a 4-element IRQ domain and the > > > > > hwirq numbers 1-4 from the of_irq_parse_and_map_pci() path, so INTD > > > > > may not work correctly. > > > > > > > > > > See > > > > > http://lkml.kernel.org/r/20170801212931.GA26498@bhelgaas-glaptop.roa > > > > > m.corp.google.com and related discussion. > > > > > > > > Sorry, I did not get this, > > > > I do some test with an intel E350T4 PCIe NICs, it's a x1 lane > > > > multi-function device. > > > > What I got from the log is below: > > > > ->of_irq_parse_and_map_pci > > > > > > > > ->of_irq_parse_pci > > > > > > > > ->irq_create_of_mapping > > > > > > > > ->irq_create_fwspec_mapping > > > > > > > > ->irq_domain_translate > > > > which will go through > > > > d->ops->translate #the hwirq really start from 0 > > > > > > > > And I tested every NIC port of the Intel E350T4 with tftp transfer > > > > data, > > > > seems all are OK with this code. > > > > > > OK. I don't know what d->ops->translate is involved here, but if it > > > works, I guess this is OK for now. We're trying to clean this up and > > > make it consistent across all the drivers. Many of them allocate a > > > 5-element IRQ domain, some make a 4-element domain, and on some of > > > them INTD doesn't work. It's a mess. > > > > Paul Burton is cleaning this up. Can you point out the d->ops->translate > > function that's involved here? > > Hi, Bjorn, > > Sorry for my last reply, I was tracking the wrong logs. The real trick > is here: > > ->of_irq_parse_and_map_pci > ->of_irq_parse_pci #out_irq->args[0] start from 1(1 == INTA) > ->of_irq_parse_raw > > After of_irq_parse_raw finished it's work, the out_irq->args[0] will be > remapped as "interrupt-map" property defines[1], which in my case, it's > start from 0, and then fwspec->param[0] is start from 0 (0 == INTA). > > My "interrupt-map" property is defined as below: > interrupt-map = <0 0 0 1 &pcie_intc0 0>, > <0 0 0 2 &pcie_intc0 1>, > <0 0 0 3 &pcie_intc0 2>, > <0 0 0 4 &pcie_intc0 3>; > > I do some test with the changes of property defined as below: > interrupt-map = <0 0 0 1 &pcie_intc0 1>, > <0 0 0 2 &pcie_intc0 2>, > <0 0 0 3 &pcie_intc0 3>, > <0 0 0 4 &pcie_intc0 4>; > Then I got the same running complain as Paul have got[2] > > So I guess it's the "interrupt-map" property defined in dtsi node play > the key role in this. > > [1]http://elixir.free-electrons.com/linux/v4.13-rc4/source/drivers/of/irq.c# > L265 [2]https://patchwork.kernel.org/patch/9794355 > > thanks. That seems like a possibly more sensible way to handle it, so long as you have control over all the device trees that the driver may be exposed to. Bjorn: is this something you want to deal with on a driver by driver basis? ie. new drivers just use interrupt-map as above and older ones with existing DT bindings use the xlate function? Looking at the drivers & device trees we have in-tree it seems we already have a mix of 0-3 & 1-4 ranges in use, so we'd just need to use the xlate function for those which currently use 1-4. I had originally done the same thing with interrupt-map on the MIPS Boston board[1] and the Xilinx PCIe driver[2], though that change would break any pre-existing device trees that use 1-4 in the interrupt-map property (which is sadly what the driver's binding document shows...). Thanks, Paul [1] https://www.linux-mips.org/archives/linux-mips/2016-08/msg00425.html [2] https://patchwork.kernel.org/patch/9763191/
diff --git a/drivers/pci/host/Kconfig b/drivers/pci/host/Kconfig index 89d61c2..5b1ae9f 100644 --- a/drivers/pci/host/Kconfig +++ b/drivers/pci/host/Kconfig @@ -182,14 +182,13 @@ config PCIE_ROCKCHIP config PCIE_MEDIATEK bool "MediaTek PCIe controller" - depends on ARM && (ARCH_MEDIATEK || COMPILE_TEST) + depends on (ARM || ARM64) && (ARCH_MEDIATEK || COMPILE_TEST) depends on OF depends on PCI select PCIEPORTBUS help Say Y here if you want to enable PCIe controller support on - MT7623 series SoCs. There is one single root complex with 3 root - ports available. Each port supports Gen2 lane x1. + MediaTek SoCs. config PCIE_TANGO_SMP8759 bool "Tango SMP8759 PCIe controller (DANGEROUS)" diff --git a/drivers/pci/host/pcie-mediatek.c b/drivers/pci/host/pcie-mediatek.c index f8e2f68..5fe92ff 100644 --- a/drivers/pci/host/pcie-mediatek.c +++ b/drivers/pci/host/pcie-mediatek.c @@ -3,6 +3,7 @@ * * Copyright (c) 2017 MediaTek Inc. * Author: Ryder Lee <ryder.lee@mediatek.com> + * Honghui Zhang <honghui.zhang@mediatek.com> * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License version 2 as @@ -17,6 +18,8 @@ #include <linux/clk.h> #include <linux/delay.h> #include <linux/iopoll.h> +#include <linux/irq.h> +#include <linux/irqdomain.h> #include <linux/kernel.h> #include <linux/of_address.h> #include <linux/of_pci.h> @@ -64,16 +67,77 @@ #define PCIE_FC_CREDIT_MASK (GENMASK(31, 31) | GENMASK(28, 16)) #define PCIE_FC_CREDIT_VAL(x) ((x) << 16) +/* PCIe V2 share registers */ +#define PCIE_SYS_CFG_V2 0x0 +#define PCIE_CSR_LTSSM_EN(x) BIT(0 + (x) * 8) +#define PCIE_CSR_ASPM_L1_EN(x) BIT(1 + (x) * 8) + +/* PCIe V2 per-port registers */ +#define PCIE_INT_MASK 0x420 +#define INTX_MASK GENMASK(19, 16) +#define INTX_SHIFT 16 +#define INTX_NUM 4 +#define PCIE_INT_STATUS 0x424 + +#define PCIE_AHB_TRANS_BASE0_L 0x438 +#define PCIE_AHB_TRANS_BASE0_H 0x43c +#define AHB2PCIE_BASEL(x) ((x) & GENMASK(31, 0)) +#define AHB2PCIE_BASEH(x) (((u64)(x) >> 32) & GENMASK(31, 0)) +#define AHB2PCIE_SIZE(x) ((x) & GENMASK(4, 0)) +#define PCIE_AXI_WINDOW0 0x448 +#define WIN_ENABLE BIT(7) + +/* PCIe V2 Configuration Transaction Header */ +#define PCIE_CFG_HEADER0 0x460 +#define PCIE_CFG_HEADER1 0x464 +#define PCIE_CFG_HEADER2 0x468 +#define PCIE_CFG_WDATA 0x470 +#define PCIE_APP_TLP_REQ 0x488 +#define PCIE_CFG_RDATA 0x48c +#define APP_CFG_REQ BIT(0) +#define APP_CPL_STATUS GENMASK(7, 5) + +#define CFG_WRRD_TYPE_0 4 +#define CFG_WR_FMT 2 +#define CFG_RD_FMT 0 + +#define CFG_DW0_LENGTH(length) ((length) & GENMASK(9, 0)) +#define CFG_DW0_TYPE(type) (((type) << 24) & GENMASK(28, 24)) +#define CFG_DW0_FMT(fmt) (((fmt) << 29) & GENMASK(31, 29)) +#define CFG_DW2_REGN(regn) ((regn) & GENMASK(11, 2)) +#define CFG_DW2_FUN(fun) (((fun) << 16) & GENMASK(18, 16)) +#define CFG_DW2_DEV(dev) (((dev) << 19) & GENMASK(23, 19)) +#define CFG_DW2_BUS(bus) (((bus) << 24) & GENMASK(31, 24)) +#define CFG_HEADER_DW0(type, fmt) \ + (CFG_DW0_LENGTH(1) | CFG_DW0_TYPE(type) | CFG_DW0_FMT(fmt)) +#define CFG_HEADER_DW1(where, size) \ + (GENMASK(((size) - 1), 0) << ((where) & 0x3)) +#define CFG_HEADER_DW2(regn, fun, dev, bus) \ + (CFG_DW2_REGN(regn) | CFG_DW2_FUN(fun) | \ + CFG_DW2_DEV(dev) | CFG_DW2_BUS(bus)) + +#define PCIE_RST_CTRL 0x510 +#define PCIE_PHY_RSTB BIT(0) +#define PCIE_PIPE_SRSTB BIT(1) +#define PCIE_MAC_SRSTB BIT(2) +#define PCIE_CRSTB BIT(3) +#define PCIE_PERSTB BIT(8) +#define PCI_LINKDOWN_RST_EN GENMASK(15, 13) +#define PCIE_LINK_STATUS_V2 0x804 +#define PCIE_PORT_LINKUP_V2 BIT(10) + struct mtk_pcie_port; /** * struct mtk_pcie_soc - differentiate between host generations * @ops: pointer to configuration access functions * @startup: pointer to controller setting functions + * @setup_irq: pointer to initialize irq functions */ struct mtk_pcie_soc { struct pci_ops *ops; int (*startup)(struct mtk_pcie_port *port); + int (*setup_irq)(struct mtk_pcie_port *port, struct device_node *node); }; /** @@ -82,10 +146,18 @@ struct mtk_pcie_soc { * @list: port list * @pcie: pointer to PCIe host info * @reset: pointer to port reset control - * @sys_ck: pointer to bus clock + * @sys_ck: pointer to transaction/data link layer clock + * @ahb_ck: pinter to AHB slave interface operating clock for CSR access + * and RC initiated MMIO access + * @axi_ck: pointer to application layer MMIO channel operating clock + * @aux_ck: pointer to pe2_mac_bridge and pe2_mac_core operating clock + * when pcie_mac_ck/pcie_pipe_ck is turned off + * @obff_ck: pointer to OBFF functional block operating clock + * @pipe_ck: pointer to LTSSM and phy/mac layer operating clock * @phy: pointer to phy control block * @lane: lane count * @index: port index + * @irq_domain: legacy intx irq domain */ struct mtk_pcie_port { void __iomem *base; @@ -93,9 +165,15 @@ struct mtk_pcie_port { struct mtk_pcie *pcie; struct reset_control *reset; struct clk *sys_ck; + struct clk *ahb_ck; + struct clk *axi_ck; + struct clk *aux_ck; + struct clk *obff_ck; + struct clk *pipe_ck; struct phy *phy; u32 lane; u32 index; + struct irq_domain *irq_domain; }; /** @@ -156,6 +234,12 @@ static void mtk_pcie_put_resources(struct mtk_pcie *pcie) list_for_each_entry_safe(port, tmp, &pcie->ports, list) { phy_power_off(port->phy); + phy_exit(port->phy); + clk_disable_unprepare(port->pipe_ck); + clk_disable_unprepare(port->obff_ck); + clk_disable_unprepare(port->axi_ck); + clk_disable_unprepare(port->aux_ck); + clk_disable_unprepare(port->ahb_ck); clk_disable_unprepare(port->sys_ck); mtk_pcie_port_free(port); } @@ -163,6 +247,298 @@ static void mtk_pcie_put_resources(struct mtk_pcie *pcie) mtk_pcie_subsys_powerdown(pcie); } +static int mtk_pcie_check_cfg_cpld(struct mtk_pcie_port *port) +{ + u32 val; + int err; + + err = readl_poll_timeout_atomic(port->base + PCIE_APP_TLP_REQ, val, + !(val & APP_CFG_REQ), 10, + 100 * USEC_PER_MSEC); + if (err) + return PCIBIOS_SET_FAILED; + + if (readl(port->base + PCIE_APP_TLP_REQ) & APP_CPL_STATUS) + return PCIBIOS_SET_FAILED; + + return PCIBIOS_SUCCESSFUL; +} + +static int mtk_pcie_hw_rd_cfg(struct mtk_pcie_port *port, u32 bus, u32 devfn, + int where, int size, u32 *val) +{ + int reg, shift = 8 * (where & 3); + + /* Write PCIe Configuration Transaction Header for cfgrd */ + writel(CFG_HEADER_DW0(CFG_WRRD_TYPE_0, CFG_RD_FMT), + port->base + PCIE_CFG_HEADER0); + writel(CFG_HEADER_DW1(where, size), port->base + PCIE_CFG_HEADER1); + writel(CFG_HEADER_DW2(where, PCI_FUNC(devfn), PCI_SLOT(devfn), bus), + port->base + PCIE_CFG_HEADER2); + + /* Triget h/w to transmit Cfgrd TLP */ + reg = readl(port->base + PCIE_APP_TLP_REQ); + writel(reg | APP_CFG_REQ, port->base + PCIE_APP_TLP_REQ); + + /* Check complete condition */ + if (mtk_pcie_check_cfg_cpld(port)) + return PCIBIOS_SET_FAILED; + + /* Read cpld payload of Cfgrd */ + *val = readl(port->base + PCIE_CFG_RDATA); + + switch (size) { + case 4: + break; + case 3: + *val = (*val >> shift) & 0xffffff; + break; + case 2: + *val = (*val >> shift) & 0xffff; + break; + case 1: + *val = (*val >> shift) & 0xff; + break; + default: + return PCIBIOS_BAD_REGISTER_NUMBER; + } + + return PCIBIOS_SUCCESSFUL; +} + +static int mtk_pcie_hw_wr_cfg(struct mtk_pcie_port *port, u32 bus, u32 devfn, + int where, int size, u32 val) +{ + /* Write PCIe Configuration Transaction Header for Cfgwr */ + writel(CFG_HEADER_DW0(CFG_WRRD_TYPE_0, CFG_WR_FMT), + port->base + PCIE_CFG_HEADER0); + writel(CFG_HEADER_DW1(where, size), port->base + PCIE_CFG_HEADER1); + writel(CFG_HEADER_DW2(where, PCI_FUNC(devfn), PCI_SLOT(devfn), bus), + port->base + PCIE_CFG_HEADER2); + + /* Write cfgwr data */ + val = val << 8 * (where & 3); + writel(val, port->base + PCIE_CFG_WDATA); + + /* Triget h/w to transmit Cfgwr TLP */ + val = readl(port->base + PCIE_APP_TLP_REQ); + val |= APP_CFG_REQ; + writel(val, port->base + PCIE_APP_TLP_REQ); + + /* Check complete condition */ + return mtk_pcie_check_cfg_cpld(port); +} + +static struct mtk_pcie_port *mtk_pcie_find_port(struct mtk_pcie *pcie, + struct pci_bus *bus, int devfn) +{ + struct pci_dev *dev; + struct pci_bus *pbus; + struct mtk_pcie_port *port, *tmp; + + list_for_each_entry_safe(port, tmp, &pcie->ports, list) { + if (bus->number == 0 && port->index == PCI_SLOT(devfn)) { + return port; + } else if (bus->number != 0) { + pbus = bus; + do { + dev = pbus->self; + if (port->index == PCI_SLOT(dev->devfn)) + return port; + pbus = dev->bus; + } while (dev->bus->number != 0); + } + } + + return NULL; +} + +static int mtk_pcie_config_read(struct pci_bus *bus, unsigned int devfn, + int where, int size, u32 *val) +{ + struct mtk_pcie_port *port; + struct pci_host_bridge *host = pci_find_host_bridge(bus); + struct mtk_pcie *pcie = pci_host_bridge_priv(host); + u32 bn = bus->number; + int ret; + + port = mtk_pcie_find_port(pcie, bus, devfn); + if (!port) { + *val = ~0; + return PCIBIOS_DEVICE_NOT_FOUND; + } + + ret = mtk_pcie_hw_rd_cfg(port, bn, devfn, where, size, val); + if (ret) + *val = ~0; + + return ret; +} + +static int mtk_pcie_config_write(struct pci_bus *bus, unsigned int devfn, + int where, int size, u32 val) +{ + u32 bn = bus->number; + struct pci_host_bridge *host = pci_find_host_bridge(bus); + struct mtk_pcie *pcie = pci_host_bridge_priv(host); + struct mtk_pcie_port *port; + + port = mtk_pcie_find_port(pcie, bus, devfn); + if (!port) + return PCIBIOS_DEVICE_NOT_FOUND; + + return mtk_pcie_hw_wr_cfg(port, bn, devfn, where, size, val); +} + +static struct pci_ops mtk_pcie_ops_v2 = { + .read = mtk_pcie_config_read, + .write = mtk_pcie_config_write, +}; + +static int mtk_pcie_startup_ports_v2(struct mtk_pcie_port *port) +{ + struct mtk_pcie *pcie = port->pcie; + struct resource *mem = &pcie->mem; + u32 val; + size_t size; + int err; + + /* mt7622 platforms need to enable LTSSM and ASPM from PCIe subsys */ + if (pcie->base) { + val = readl(pcie->base + PCIE_SYS_CFG_V2); + val |= PCIE_CSR_LTSSM_EN(port->index) | + PCIE_CSR_ASPM_L1_EN(port->index); + writel(val, pcie->base + PCIE_SYS_CFG_V2); + } + + /* Assert all reset signals */ + writel(0, port->base + PCIE_RST_CTRL); + + /* + * Enable rc internal reset. + * The reset will work when the link is from link up to link down. + */ + writel(PCI_LINKDOWN_RST_EN, port->base + PCIE_RST_CTRL); + + /* De-assert phy, pe, pipe, mac and configuration reset */ + val = readl(port->base + PCIE_RST_CTRL); + val |= PCIE_PHY_RSTB | PCIE_PERSTB | PCIE_PIPE_SRSTB | + PCIE_MAC_SRSTB | PCIE_CRSTB; + writel(val, port->base + PCIE_RST_CTRL); + + /* PCIe v2.0 need at least 100ms delay to train from Gen1 to Gen2 */ + err = readl_poll_timeout(port->base + PCIE_LINK_STATUS_V2, val, + !!(val & PCIE_PORT_LINKUP_V2), 20, + 100 * USEC_PER_MSEC); + if (err) + return -ETIMEDOUT; + + /* Set INTx mask */ + val = readl(port->base + PCIE_INT_MASK); + val &= ~INTX_MASK; + writel(val, port->base + PCIE_INT_MASK); + + /* Set AHB to PCIe translation windows */ + size = mem->end - mem->start; + val = AHB2PCIE_BASEL(mem->start) | AHB2PCIE_SIZE(fls(size)); + writel(val, port->base + PCIE_AHB_TRANS_BASE0_L); + + val = AHB2PCIE_BASEH(mem->start); + writel(val, port->base + PCIE_AHB_TRANS_BASE0_H); + + /* Set PCIe to axi translation memory space.*/ + val = fls(0xffffffff) | WIN_ENABLE; + writel(val, port->base + PCIE_AXI_WINDOW0); + + return 0; +} + +static int mtk_pcie_intx_map(struct irq_domain *domain, unsigned int irq, + irq_hw_number_t hwirq) +{ + irq_set_chip_and_handler(irq, &dummy_irq_chip, handle_simple_irq); + irq_set_chip_data(irq, domain->host_data); + + return 0; +} + +static const struct irq_domain_ops intx_domain_ops = { + .map = mtk_pcie_intx_map, +}; + +static int mtk_pcie_init_irq_domain(struct mtk_pcie_port *port, + struct device_node *node) +{ + struct device *dev = port->pcie->dev; + struct device_node *pcie_intc_node; + + /* Setup INTx */ + pcie_intc_node = of_get_next_child(node, NULL); + if (!pcie_intc_node) { + dev_err(dev, "No PCIe Intc node found\n"); + return PTR_ERR(pcie_intc_node); + } + + port->irq_domain = irq_domain_add_linear(pcie_intc_node, INTX_NUM, + &intx_domain_ops, port); + if (!port->irq_domain) { + dev_err(dev, "Failed to get INTx IRQ domain\n"); + return PTR_ERR(port->irq_domain); + } + + return 0; +} + +static irqreturn_t mtk_pcie_intr_handler(int irq, void *data) +{ + struct mtk_pcie_port *port = (struct mtk_pcie_port *)data; + struct device *dev = port->pcie->dev; + unsigned long status; + u32 virq; + u32 bit = INTX_SHIFT; + + while ((status = readl(port->base + PCIE_INT_STATUS)) & INTX_MASK) { + for_each_set_bit_from(bit, &status, INTX_NUM + INTX_SHIFT) { + /* Clear the INTx */ + writel(1 << bit, port->base + PCIE_INT_STATUS); + virq = irq_find_mapping(port->irq_domain, + bit - INTX_SHIFT); + if (virq) + generic_handle_irq(virq); + else + dev_err(dev, "unexpected IRQ, INT%d\n", + bit - INTX_SHIFT); + } + } + + return IRQ_HANDLED; +} + +static int mtk_pcie_setup_irq(struct mtk_pcie_port *port, + struct device_node *node) +{ + struct mtk_pcie *pcie = port->pcie; + struct device *dev = pcie->dev; + struct platform_device *pdev = to_platform_device(dev); + int err, irq; + + irq = platform_get_irq(pdev, port->index); + err = devm_request_irq(dev, irq, mtk_pcie_intr_handler, + IRQF_SHARED, "mtk-pcie", port); + if (err) { + dev_err(dev, "unable to request irq %d\n", irq); + return err; + } + + err = mtk_pcie_init_irq_domain(port, node); + if (err) { + dev_err(dev, "failed to init pcie lagecy irq domain\n"); + return -ENODEV; + } + + return 0; +} + static void __iomem *mtk_pcie_map_bus(struct pci_bus *bus, unsigned int devfn, int where) { @@ -249,13 +625,49 @@ static void mtk_pcie_enable_ports(struct mtk_pcie_port *port) err = clk_prepare_enable(port->sys_ck); if (err) { - dev_err(dev, "failed to enable port%d clock\n", port->index); + dev_err(dev, "failed to enable sys_ck%d\n", port->index); goto err_sys_clk; } + err = clk_prepare_enable(port->ahb_ck); + if (err) { + dev_err(dev, "failed to enable ahb_ck%d\n", port->index); + goto err_ahb_clk; + } + + err = clk_prepare_enable(port->aux_ck); + if (err) { + dev_err(dev, "failed to enable aux_ck%d\n", port->index); + goto err_aux_clk; + } + + err = clk_prepare_enable(port->axi_ck); + if (err) { + dev_err(dev, "failed to enable axi_ck%d\n", port->index); + goto err_axi_clk; + } + + err = clk_prepare_enable(port->obff_ck); + if (err) { + dev_err(dev, "failed to enable obff_ck%d\n", port->index); + goto err_obff_clk; + } + + err = clk_prepare_enable(port->pipe_ck); + if (err) { + dev_err(dev, "failed to enable pipe_ck%d\n", port->index); + goto err_pipe_clk; + } + reset_control_assert(port->reset); reset_control_deassert(port->reset); + err = phy_init(port->phy); + if (err) { + dev_err(dev, "failed to initialize port%d phy\n", port->index); + goto err_phy_init; + } + err = phy_power_on(port->phy); if (err) { dev_err(dev, "failed to power on port%d phy\n", port->index); @@ -269,6 +681,18 @@ static void mtk_pcie_enable_ports(struct mtk_pcie_port *port) phy_power_off(port->phy); err_phy_on: + phy_exit(port->phy); +err_phy_init: + clk_disable_unprepare(port->pipe_ck); +err_pipe_clk: + clk_disable_unprepare(port->obff_ck); +err_obff_clk: + clk_disable_unprepare(port->axi_ck); +err_axi_clk: + clk_disable_unprepare(port->aux_ck); +err_aux_clk: + clk_disable_unprepare(port->ahb_ck); +err_ahb_clk: clk_disable_unprepare(port->sys_ck); err_sys_clk: mtk_pcie_port_free(port); @@ -306,10 +730,56 @@ static int mtk_pcie_parse_ports(struct mtk_pcie *pcie, snprintf(name, sizeof(name), "sys_ck%d", index); port->sys_ck = devm_clk_get(dev, name); if (IS_ERR(port->sys_ck)) { - dev_err(dev, "failed to get port%d clock\n", index); + dev_err(dev, "failed to get sys_ck%d\n", index); return PTR_ERR(port->sys_ck); } + /* sys_ck might be divided into the following parts in some chips */ + snprintf(name, sizeof(name), "ahb_ck%d", index); + port->ahb_ck = devm_clk_get(dev, name); + if (IS_ERR(port->ahb_ck)) { + if (PTR_ERR(port->ahb_ck) == -EPROBE_DEFER) + return -EPROBE_DEFER; + + port->ahb_ck = NULL; + } + + snprintf(name, sizeof(name), "axi_ck%d", index); + port->axi_ck = devm_clk_get(dev, name); + if (IS_ERR(port->axi_ck)) { + if (PTR_ERR(port->axi_ck) == -EPROBE_DEFER) + return -EPROBE_DEFER; + + port->axi_ck = NULL; + } + + snprintf(name, sizeof(name), "aux_ck%d", index); + port->aux_ck = devm_clk_get(dev, name); + if (IS_ERR(port->aux_ck)) { + if (PTR_ERR(port->aux_ck) == -EPROBE_DEFER) + return -EPROBE_DEFER; + + port->aux_ck = NULL; + } + + snprintf(name, sizeof(name), "obff_ck%d", index); + port->obff_ck = devm_clk_get(dev, name); + if (IS_ERR(port->obff_ck)) { + if (PTR_ERR(port->obff_ck) == -EPROBE_DEFER) + return -EPROBE_DEFER; + + port->obff_ck = NULL; + } + + snprintf(name, sizeof(name), "pipe_ck%d", index); + port->pipe_ck = devm_clk_get(dev, name); + if (IS_ERR(port->pipe_ck)) { + if (PTR_ERR(port->pipe_ck) == -EPROBE_DEFER) + return -EPROBE_DEFER; + + port->pipe_ck = NULL; + } + snprintf(name, sizeof(name), "pcie-rst%d", index); port->reset = devm_reset_control_get_optional(dev, name); if (PTR_ERR(port->reset) == -EPROBE_DEFER) @@ -324,6 +794,12 @@ static int mtk_pcie_parse_ports(struct mtk_pcie *pcie, port->index = index; port->pcie = pcie; + if (pcie->soc->setup_irq) { + err = pcie->soc->setup_irq(port, node); + if (err) + return err; + } + INIT_LIST_HEAD(&port->list); list_add_tail(&port->list, &pcie->ports); @@ -553,9 +1029,17 @@ static struct mtk_pcie_soc mtk_pcie_soc_v1 = { .startup = mtk_pcie_startup_ports, }; +static struct mtk_pcie_soc mtk_pcie_soc_v2 = { + .ops = &mtk_pcie_ops_v2, + .startup = mtk_pcie_startup_ports_v2, + .setup_irq = mtk_pcie_setup_irq, +}; + static const struct of_device_id mtk_pcie_ids[] = { { .compatible = "mediatek,mt2701-pcie", .data = &mtk_pcie_soc_v1 }, { .compatible = "mediatek,mt7623-pcie", .data = &mtk_pcie_soc_v1 }, + { .compatible = "mediatek,mt2712-pcie", .data = &mtk_pcie_soc_v2 }, + { .compatible = "mediatek,mt7622-pcie", .data = &mtk_pcie_soc_v2 }, {}, };