diff options
Diffstat (limited to 'drivers/irqchip')
52 files changed, 4297 insertions, 452 deletions
diff --git a/drivers/irqchip/Kconfig b/drivers/irqchip/Kconfig index c3928ef79344..6d12c6ab9ea4 100644 --- a/drivers/irqchip/Kconfig +++ b/drivers/irqchip/Kconfig @@ -41,10 +41,14 @@ config ARM_GIC_V3 select HAVE_ARM_SMCCC_DISCOVERY select IRQ_MSI_IOMMU +config ARM_GIC_ITS_PARENT + bool + config ARM_GIC_V3_ITS bool select GENERIC_MSI_IRQ select IRQ_MSI_LIB + select ARM_GIC_ITS_PARENT default ARM_GIC_V3 select IRQ_MSI_IOMMU @@ -54,6 +58,14 @@ config ARM_GIC_V3_ITS_FSL_MC depends on FSL_MC_BUS default ARM_GIC_V3_ITS +config ARM_GIC_V5 + bool + select IRQ_DOMAIN_HIERARCHY + select GENERIC_IRQ_EFFECTIVE_AFF_MASK + select GENERIC_MSI_IRQ + select IRQ_MSI_LIB + select ARM_GIC_ITS_PARENT + config ARM_NVIC bool select IRQ_DOMAIN_HIERARCHY @@ -80,12 +92,14 @@ config ARMADA_370_XP_IRQ bool select GENERIC_IRQ_CHIP select PCI_MSI if PCI + select IRQ_MSI_LIB if PCI select GENERIC_IRQ_EFFECTIVE_AFF_MASK if SMP config ALPINE_MSI bool depends on PCI select PCI_MSI + select IRQ_MSI_LIB select GENERIC_IRQ_CHIP config AL_FIC @@ -435,6 +449,7 @@ config LS_SCFG_MSI def_bool y if SOC_LS1021A || ARCH_LAYERSCAPE select IRQ_MSI_IOMMU depends on PCI_MSI + select IRQ_MSI_LIB config PARTITION_PERCPU bool @@ -539,6 +554,7 @@ config IMX_MU_MSI tristate "i.MX MU used as MSI controller" depends on OF && HAS_IOMEM depends on ARCH_MXC || COMPILE_TEST + depends on ARM || ARM64 default m if ARCH_MXC select IRQ_DOMAIN select IRQ_DOMAIN_HIERARCHY @@ -635,18 +651,25 @@ config STARFIVE_JH8100_INTC If you don't know what to do here, say Y. -config THEAD_C900_ACLINT_SSWI - bool "THEAD C9XX ACLINT S-mode IPI Interrupt Controller" +config ACLINT_SSWI + bool "RISC-V ACLINT S-mode IPI Interrupt Controller" depends on RISCV depends on SMP select IRQ_DOMAIN_HIERARCHY select GENERIC_IRQ_IPI_MUX help - This enables support for T-HEAD specific ACLINT SSWI device - support. + This enables support for variants of the RISC-V ACLINT-SSWI device. + Supported variants are: + - T-HEAD, with compatible "thead,c900-aclint-sswi" + - MIPS P8700, with compatible "mips,p8700-aclint-sswi" If you don't know what to do here, say Y. +# Backwards compatibility so oldconfig does not drop it. +config THEAD_C900_ACLINT_SSWI + bool + select ACLINT_SSWI + config EXYNOS_IRQ_COMBINER bool "Samsung Exynos IRQ combiner support" if COMPILE_TEST depends on (ARCH_EXYNOS && ARM) || COMPILE_TEST diff --git a/drivers/irqchip/Makefile b/drivers/irqchip/Makefile index 23ca4959e6ce..93e3ced023bb 100644 --- a/drivers/irqchip/Makefile +++ b/drivers/irqchip/Makefile @@ -33,9 +33,12 @@ obj-$(CONFIG_ARCH_REALVIEW) += irq-gic-realview.o obj-$(CONFIG_IRQ_MSI_LIB) += irq-msi-lib.o obj-$(CONFIG_ARM_GIC_V2M) += irq-gic-v2m.o obj-$(CONFIG_ARM_GIC_V3) += irq-gic-v3.o irq-gic-v3-mbi.o irq-gic-common.o -obj-$(CONFIG_ARM_GIC_V3_ITS) += irq-gic-v3-its.o irq-gic-v4.o irq-gic-v3-its-msi-parent.o +obj-$(CONFIG_ARM_GIC_ITS_PARENT) += irq-gic-its-msi-parent.o +obj-$(CONFIG_ARM_GIC_V3_ITS) += irq-gic-v3-its.o irq-gic-v4.o obj-$(CONFIG_ARM_GIC_V3_ITS_FSL_MC) += irq-gic-v3-its-fsl-mc-msi.o obj-$(CONFIG_PARTITION_PERCPU) += irq-partition-percpu.o +obj-$(CONFIG_ARM_GIC_V5) += irq-gic-v5.o irq-gic-v5-irs.o irq-gic-v5-its.o \ + irq-gic-v5-iwb.o obj-$(CONFIG_HISILICON_IRQ_MBIGEN) += irq-mbigen.o obj-$(CONFIG_ARM_NVIC) += irq-nvic.o obj-$(CONFIG_ARM_VIC) += irq-vic.o @@ -105,7 +108,7 @@ obj-$(CONFIG_RISCV_APLIC_MSI) += irq-riscv-aplic-msi.o obj-$(CONFIG_RISCV_IMSIC) += irq-riscv-imsic-state.o irq-riscv-imsic-early.o irq-riscv-imsic-platform.o obj-$(CONFIG_SIFIVE_PLIC) += irq-sifive-plic.o obj-$(CONFIG_STARFIVE_JH8100_INTC) += irq-starfive-jh8100-intc.o -obj-$(CONFIG_THEAD_C900_ACLINT_SSWI) += irq-thead-c900-aclint-sswi.o +obj-$(CONFIG_ACLINT_SSWI) += irq-aclint-sswi.o obj-$(CONFIG_IMX_IRQSTEER) += irq-imx-irqsteer.o obj-$(CONFIG_IMX_INTMUX) += irq-imx-intmux.o obj-$(CONFIG_IMX_MU_MSI) += irq-imx-mu-msi.o diff --git a/drivers/irqchip/irq-thead-c900-aclint-sswi.c b/drivers/irqchip/irq-aclint-sswi.c index 8ff6e7a1363b..93e28e9f281f 100644 --- a/drivers/irqchip/irq-thead-c900-aclint-sswi.c +++ b/drivers/irqchip/irq-aclint-sswi.c @@ -3,87 +3,77 @@ * Copyright (C) 2024 Inochi Amaoto <inochiama@gmail.com> */ -#define pr_fmt(fmt) "thead-c900-aclint-sswi: " fmt +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + #include <linux/cpu.h> #include <linux/interrupt.h> -#include <linux/io.h> -#include <linux/irq.h> #include <linux/irqchip.h> #include <linux/irqchip/chained_irq.h> -#include <linux/module.h> -#include <linux/of.h> #include <linux/of_address.h> -#include <linux/of_irq.h> -#include <linux/pci.h> #include <linux/spinlock.h> #include <linux/smp.h> #include <linux/string_choices.h> #include <asm/sbi.h> #include <asm/vendorid_list.h> -#define THEAD_ACLINT_xSWI_REGISTER_SIZE 4 - -#define THEAD_C9XX_CSR_SXSTATUS 0x5c0 -#define THEAD_C9XX_SXSTATUS_CLINTEE BIT(17) - static int sswi_ipi_virq __ro_after_init; static DEFINE_PER_CPU(void __iomem *, sswi_cpu_regs); -static void thead_aclint_sswi_ipi_send(unsigned int cpu) +static void aclint_sswi_ipi_send(unsigned int cpu) { writel(0x1, per_cpu(sswi_cpu_regs, cpu)); } -static void thead_aclint_sswi_ipi_clear(void) +static void aclint_sswi_ipi_clear(void) { writel_relaxed(0x0, this_cpu_read(sswi_cpu_regs)); } -static void thead_aclint_sswi_ipi_handle(struct irq_desc *desc) +static void aclint_sswi_ipi_handle(struct irq_desc *desc) { struct irq_chip *chip = irq_desc_get_chip(desc); chained_irq_enter(chip, desc); csr_clear(CSR_IP, IE_SIE); - thead_aclint_sswi_ipi_clear(); + aclint_sswi_ipi_clear(); ipi_mux_process(); chained_irq_exit(chip, desc); } -static int thead_aclint_sswi_starting_cpu(unsigned int cpu) +static int aclint_sswi_starting_cpu(unsigned int cpu) { enable_percpu_irq(sswi_ipi_virq, irq_get_trigger_type(sswi_ipi_virq)); return 0; } -static int thead_aclint_sswi_dying_cpu(unsigned int cpu) +static int aclint_sswi_dying_cpu(unsigned int cpu) { - thead_aclint_sswi_ipi_clear(); + aclint_sswi_ipi_clear(); disable_percpu_irq(sswi_ipi_virq); return 0; } -static int __init thead_aclint_sswi_parse_irq(struct fwnode_handle *fwnode, - void __iomem *reg) +static int __init aclint_sswi_parse_irq(struct fwnode_handle *fwnode, void __iomem *reg) { - struct of_phandle_args parent; - unsigned long hartid; - u32 contexts, i; - int rc, cpu; + u32 contexts = of_irq_count(to_of_node(fwnode)); - contexts = of_irq_count(to_of_node(fwnode)); if (!(contexts)) { pr_err("%pfwP: no ACLINT SSWI context available\n", fwnode); return -EINVAL; } - for (i = 0; i < contexts; i++) { + for (u32 i = 0; i < contexts; i++) { + struct of_phandle_args parent; + unsigned long hartid; + u32 hart_index; + int rc, cpu; + rc = of_irq_parse_one(to_of_node(fwnode), i, &parent); if (rc) return rc; @@ -97,7 +87,12 @@ static int __init thead_aclint_sswi_parse_irq(struct fwnode_handle *fwnode, cpu = riscv_hartid_to_cpuid(hartid); - per_cpu(sswi_cpu_regs, cpu) = reg + i * THEAD_ACLINT_xSWI_REGISTER_SIZE; + rc = riscv_get_hart_index(fwnode, i, &hart_index); + if (rc) { + pr_warn("%pfwP: hart index [%d] not found\n", fwnode, i); + return -EINVAL; + } + per_cpu(sswi_cpu_regs, cpu) = reg + hart_index * 4; } pr_info("%pfwP: register %u CPU%s\n", fwnode, contexts, str_plural(contexts)); @@ -105,17 +100,12 @@ static int __init thead_aclint_sswi_parse_irq(struct fwnode_handle *fwnode, return 0; } -static int __init thead_aclint_sswi_probe(struct fwnode_handle *fwnode) +static int __init aclint_sswi_probe(struct fwnode_handle *fwnode) { struct irq_domain *domain; void __iomem *reg; int virq, rc; - /* If it is T-HEAD CPU, check whether SSWI is enabled */ - if (riscv_cached_mvendorid(0) == THEAD_VENDOR_ID && - !(csr_read(THEAD_C9XX_CSR_SXSTATUS) & THEAD_C9XX_SXSTATUS_CLINTEE)) - return -ENOTSUPP; - if (!is_of_node(fwnode)) return -EINVAL; @@ -124,7 +114,7 @@ static int __init thead_aclint_sswi_probe(struct fwnode_handle *fwnode) return -ENOMEM; /* Parse SSWI setting */ - rc = thead_aclint_sswi_parse_irq(fwnode, reg); + rc = aclint_sswi_parse_irq(fwnode, reg); if (rc < 0) return rc; @@ -146,22 +136,64 @@ static int __init thead_aclint_sswi_probe(struct fwnode_handle *fwnode) } /* Register SSWI irq and handler */ - virq = ipi_mux_create(BITS_PER_BYTE, thead_aclint_sswi_ipi_send); + virq = ipi_mux_create(BITS_PER_BYTE, aclint_sswi_ipi_send); if (virq <= 0) { pr_err("unable to create muxed IPIs\n"); irq_dispose_mapping(sswi_ipi_virq); return virq < 0 ? virq : -ENOMEM; } - irq_set_chained_handler(sswi_ipi_virq, thead_aclint_sswi_ipi_handle); + irq_set_chained_handler(sswi_ipi_virq, aclint_sswi_ipi_handle); - cpuhp_setup_state(CPUHP_AP_IRQ_THEAD_ACLINT_SSWI_STARTING, - "irqchip/thead-aclint-sswi:starting", - thead_aclint_sswi_starting_cpu, - thead_aclint_sswi_dying_cpu); + cpuhp_setup_state(CPUHP_AP_IRQ_ACLINT_SSWI_STARTING, + "irqchip/aclint-sswi:starting", + aclint_sswi_starting_cpu, + aclint_sswi_dying_cpu); riscv_ipi_set_virq_range(virq, BITS_PER_BYTE); + return 0; +} + +/* generic/MIPS variant */ +static int __init generic_aclint_sswi_probe(struct fwnode_handle *fwnode) +{ + int rc; + + rc = aclint_sswi_probe(fwnode); + if (rc) + return rc; + + /* Announce that SSWI is providing IPIs */ + pr_info("providing IPIs using ACLINT SSWI\n"); + + return 0; +} + +static int __init generic_aclint_sswi_early_probe(struct device_node *node, + struct device_node *parent) +{ + return generic_aclint_sswi_probe(&node->fwnode); +} +IRQCHIP_DECLARE(generic_aclint_sswi, "mips,p8700-aclint-sswi", generic_aclint_sswi_early_probe); + +/* THEAD variant */ +#define THEAD_C9XX_CSR_SXSTATUS 0x5c0 +#define THEAD_C9XX_SXSTATUS_CLINTEE BIT(17) + +static int __init thead_aclint_sswi_probe(struct fwnode_handle *fwnode) +{ + int rc; + + /* If it is T-HEAD CPU, check whether SSWI is enabled */ + if (riscv_cached_mvendorid(0) == THEAD_VENDOR_ID && + !(csr_read(THEAD_C9XX_CSR_SXSTATUS) & THEAD_C9XX_SXSTATUS_CLINTEE)) + return -ENOTSUPP; + + rc = aclint_sswi_probe(fwnode); + if (rc) + return rc; + /* Announce that SSWI is providing IPIs */ pr_info("providing IPIs using THEAD ACLINT SSWI\n"); diff --git a/drivers/irqchip/irq-alpine-msi.c b/drivers/irqchip/irq-alpine-msi.c index a5289dc26dca..159d9ec7c0dd 100644 --- a/drivers/irqchip/irq-alpine-msi.c +++ b/drivers/irqchip/irq-alpine-msi.c @@ -14,6 +14,7 @@ #include <linux/irqchip.h> #include <linux/irqchip/arm-gic.h> +#include <linux/irqchip/irq-msi-lib.h> #include <linux/msi.h> #include <linux/of.h> #include <linux/of_address.h> @@ -29,84 +30,45 @@ #define ALPINE_MSIX_SPI_TARGET_CLUSTER0 BIT(16) struct alpine_msix_data { - spinlock_t msi_map_lock; - phys_addr_t addr; - u32 spi_first; /* The SGI number that MSIs start */ - u32 num_spis; /* The number of SGIs for MSIs */ - unsigned long *msi_map; -}; - -static void alpine_msix_mask_msi_irq(struct irq_data *d) -{ - pci_msi_mask_irq(d); - irq_chip_mask_parent(d); -} - -static void alpine_msix_unmask_msi_irq(struct irq_data *d) -{ - pci_msi_unmask_irq(d); - irq_chip_unmask_parent(d); -} - -static struct irq_chip alpine_msix_irq_chip = { - .name = "MSIx", - .irq_mask = alpine_msix_mask_msi_irq, - .irq_unmask = alpine_msix_unmask_msi_irq, - .irq_eoi = irq_chip_eoi_parent, - .irq_set_affinity = irq_chip_set_affinity_parent, + spinlock_t msi_map_lock; + phys_addr_t addr; + u32 spi_first; /* The SGI number that MSIs start */ + u32 num_spis; /* The number of SGIs for MSIs */ + unsigned long *msi_map; }; static int alpine_msix_allocate_sgi(struct alpine_msix_data *priv, int num_req) { int first; - spin_lock(&priv->msi_map_lock); - - first = bitmap_find_next_zero_area(priv->msi_map, priv->num_spis, 0, - num_req, 0); - if (first >= priv->num_spis) { - spin_unlock(&priv->msi_map_lock); + guard(spinlock)(&priv->msi_map_lock); + first = bitmap_find_next_zero_area(priv->msi_map, priv->num_spis, 0, num_req, 0); + if (first >= priv->num_spis) return -ENOSPC; - } bitmap_set(priv->msi_map, first, num_req); - - spin_unlock(&priv->msi_map_lock); - return priv->spi_first + first; } -static void alpine_msix_free_sgi(struct alpine_msix_data *priv, unsigned sgi, - int num_req) +static void alpine_msix_free_sgi(struct alpine_msix_data *priv, unsigned int sgi, int num_req) { int first = sgi - priv->spi_first; - spin_lock(&priv->msi_map_lock); - + guard(spinlock)(&priv->msi_map_lock); bitmap_clear(priv->msi_map, first, num_req); - - spin_unlock(&priv->msi_map_lock); } -static void alpine_msix_compose_msi_msg(struct irq_data *data, - struct msi_msg *msg) +static void alpine_msix_compose_msi_msg(struct irq_data *data, struct msi_msg *msg) { struct alpine_msix_data *priv = irq_data_get_irq_chip_data(data); phys_addr_t msg_addr = priv->addr; msg_addr |= (data->hwirq << 3); - msg->address_hi = upper_32_bits(msg_addr); msg->address_lo = lower_32_bits(msg_addr); msg->data = 0; } -static struct msi_domain_info alpine_msix_domain_info = { - .flags = MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS | - MSI_FLAG_PCI_MSIX, - .chip = &alpine_msix_irq_chip, -}; - static struct irq_chip middle_irq_chip = { .name = "alpine_msix_middle", .irq_mask = irq_chip_mask_parent, @@ -116,8 +78,7 @@ static struct irq_chip middle_irq_chip = { .irq_compose_msi_msg = alpine_msix_compose_msi_msg, }; -static int alpine_msix_gic_domain_alloc(struct irq_domain *domain, - unsigned int virq, int sgi) +static int alpine_msix_gic_domain_alloc(struct irq_domain *domain, unsigned int virq, int sgi) { struct irq_fwspec fwspec; struct irq_data *d; @@ -138,12 +99,10 @@ static int alpine_msix_gic_domain_alloc(struct irq_domain *domain, d = irq_domain_get_irq_data(domain->parent, virq); d->chip->irq_set_type(d, IRQ_TYPE_EDGE_RISING); - return 0; } -static int alpine_msix_middle_domain_alloc(struct irq_domain *domain, - unsigned int virq, +static int alpine_msix_middle_domain_alloc(struct irq_domain *domain, unsigned int virq, unsigned int nr_irqs, void *args) { struct alpine_msix_data *priv = domain->host_data; @@ -161,7 +120,6 @@ static int alpine_msix_middle_domain_alloc(struct irq_domain *domain, irq_domain_set_hwirq_and_chip(domain, virq + i, sgi + i, &middle_irq_chip, priv); } - return 0; err_sgi: @@ -170,8 +128,7 @@ err_sgi: return err; } -static void alpine_msix_middle_domain_free(struct irq_domain *domain, - unsigned int virq, +static void alpine_msix_middle_domain_free(struct irq_domain *domain, unsigned int virq, unsigned int nr_irqs) { struct irq_data *d = irq_domain_get_irq_data(domain, virq); @@ -182,14 +139,35 @@ static void alpine_msix_middle_domain_free(struct irq_domain *domain, } static const struct irq_domain_ops alpine_msix_middle_domain_ops = { + .select = msi_lib_irq_domain_select, .alloc = alpine_msix_middle_domain_alloc, .free = alpine_msix_middle_domain_free, }; -static int alpine_msix_init_domains(struct alpine_msix_data *priv, - struct device_node *node) +#define ALPINE_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS | \ + MSI_FLAG_USE_DEF_CHIP_OPS | \ + MSI_FLAG_PCI_MSI_MASK_PARENT) + +#define ALPINE_MSI_FLAGS_SUPPORTED (MSI_GENERIC_FLAGS_MASK | \ + MSI_FLAG_PCI_MSIX) + +static struct msi_parent_ops alpine_msi_parent_ops = { + .supported_flags = ALPINE_MSI_FLAGS_SUPPORTED, + .required_flags = ALPINE_MSI_FLAGS_REQUIRED, + .chip_flags = MSI_CHIP_FLAG_SET_EOI, + .bus_select_token = DOMAIN_BUS_NEXUS, + .bus_select_mask = MATCH_PCI_MSI, + .prefix = "ALPINE-", + .init_dev_msi_info = msi_lib_init_dev_msi_info, +}; + +static int alpine_msix_init_domains(struct alpine_msix_data *priv, struct device_node *node) { - struct irq_domain *middle_domain, *msi_domain, *gic_domain; + struct irq_domain_info info = { + .fwnode = of_fwnode_handle(node), + .ops = &alpine_msix_middle_domain_ops, + .host_data = priv, + }; struct device_node *gic_node; gic_node = of_irq_find_parent(node); @@ -198,40 +176,26 @@ static int alpine_msix_init_domains(struct alpine_msix_data *priv, return -ENODEV; } - gic_domain = irq_find_host(gic_node); + info.parent = irq_find_host(gic_node); of_node_put(gic_node); - if (!gic_domain) { + if (!info.parent) { pr_err("Failed to find the GIC domain\n"); return -ENXIO; } - middle_domain = irq_domain_create_hierarchy(gic_domain, 0, 0, NULL, - &alpine_msix_middle_domain_ops, priv); - if (!middle_domain) { - pr_err("Failed to create the MSIX middle domain\n"); - return -ENOMEM; - } - - msi_domain = pci_msi_create_irq_domain(of_fwnode_handle(node), - &alpine_msix_domain_info, - middle_domain); - if (!msi_domain) { + if (!msi_create_parent_irq_domain(&info, &alpine_msi_parent_ops)) { pr_err("Failed to create MSI domain\n"); - irq_domain_remove(middle_domain); return -ENOMEM; } - return 0; } -static int alpine_msix_init(struct device_node *node, - struct device_node *parent) +static int alpine_msix_init(struct device_node *node, struct device_node *parent) { - struct alpine_msix_data *priv; + struct alpine_msix_data *priv __free(kfree) = kzalloc(sizeof(*priv), GFP_KERNEL); struct resource res; int ret; - priv = kzalloc(sizeof(*priv), GFP_KERNEL); if (!priv) return -ENOMEM; @@ -240,7 +204,7 @@ static int alpine_msix_init(struct device_node *node, ret = of_address_to_resource(node, 0, &res); if (ret) { pr_err("Failed to allocate resource\n"); - goto err_priv; + return ret; } /* @@ -255,35 +219,28 @@ static int alpine_msix_init(struct device_node *node, if (of_property_read_u32(node, "al,msi-base-spi", &priv->spi_first)) { pr_err("Unable to parse MSI base\n"); - ret = -EINVAL; - goto err_priv; + return -EINVAL; } if (of_property_read_u32(node, "al,msi-num-spis", &priv->num_spis)) { pr_err("Unable to parse MSI numbers\n"); - ret = -EINVAL; - goto err_priv; + return -EINVAL; } - priv->msi_map = bitmap_zalloc(priv->num_spis, GFP_KERNEL); - if (!priv->msi_map) { - ret = -ENOMEM; - goto err_priv; - } + unsigned long *msi_map __free(kfree) = bitmap_zalloc(priv->num_spis, GFP_KERNEL); - pr_debug("Registering %d msixs, starting at %d\n", - priv->num_spis, priv->spi_first); + if (!msi_map) + return -ENOMEM; + priv->msi_map = msi_map; + + pr_debug("Registering %d msixs, starting at %d\n", priv->num_spis, priv->spi_first); ret = alpine_msix_init_domains(priv, node); if (ret) - goto err_map; + return ret; + retain_and_null_ptr(priv); + retain_and_null_ptr(msi_map); return 0; - -err_map: - bitmap_free(priv->msi_map); -err_priv: - kfree(priv); - return ret; } IRQCHIP_DECLARE(alpine_msix, "al,alpine-msix", alpine_msix_init); diff --git a/drivers/irqchip/irq-armada-370-xp.c b/drivers/irqchip/irq-armada-370-xp.c index 67b672a78862..a44c49e985b7 100644 --- a/drivers/irqchip/irq-armada-370-xp.c +++ b/drivers/irqchip/irq-armada-370-xp.c @@ -20,6 +20,7 @@ #include <linux/interrupt.h> #include <linux/irqchip.h> #include <linux/irqchip/chained_irq.h> +#include <linux/irqchip/irq-msi-lib.h> #include <linux/cpu.h> #include <linux/io.h> #include <linux/of_address.h> @@ -156,7 +157,6 @@ * @parent_irq: parent IRQ if MPIC is not top-level interrupt controller * @domain: MPIC main interrupt domain * @ipi_domain: IPI domain - * @msi_domain: MSI domain * @msi_inner_domain: MSI inner domain * @msi_used: bitmap of used MSI numbers * @msi_lock: mutex serializing access to @msi_used @@ -176,7 +176,6 @@ struct mpic { struct irq_domain *ipi_domain; #endif #ifdef CONFIG_PCI_MSI - struct irq_domain *msi_domain; struct irq_domain *msi_inner_domain; DECLARE_BITMAP(msi_used, PCI_MSI_FULL_DOORBELL_NR); struct mutex msi_lock; @@ -234,18 +233,6 @@ static void mpic_irq_unmask(struct irq_data *d) #ifdef CONFIG_PCI_MSI -static struct irq_chip mpic_msi_irq_chip = { - .name = "MPIC MSI", - .irq_mask = pci_msi_mask_irq, - .irq_unmask = pci_msi_unmask_irq, -}; - -static struct msi_domain_info mpic_msi_domain_info = { - .flags = (MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS | - MSI_FLAG_MULTI_PCI_MSI | MSI_FLAG_PCI_MSIX), - .chip = &mpic_msi_irq_chip, -}; - static void mpic_compose_msi_msg(struct irq_data *d, struct msi_msg *msg) { unsigned int cpu = cpumask_first(irq_data_get_effective_affinity_mask(d)); @@ -314,6 +301,7 @@ static void mpic_msi_free(struct irq_domain *domain, unsigned int virq, unsigned } static const struct irq_domain_ops mpic_msi_domain_ops = { + .select = msi_lib_irq_domain_select, .alloc = mpic_msi_alloc, .free = mpic_msi_free, }; @@ -331,6 +319,21 @@ static void mpic_msi_reenable_percpu(struct mpic *mpic) writel(1, mpic->per_cpu + MPIC_INT_CLEAR_MASK); } +#define MPIC_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS | \ + MSI_FLAG_USE_DEF_CHIP_OPS) +#define MPIC_MSI_FLAGS_SUPPORTED (MSI_FLAG_MULTI_PCI_MSI | \ + MSI_FLAG_PCI_MSIX | \ + MSI_GENERIC_FLAGS_MASK) + +static const struct msi_parent_ops mpic_msi_parent_ops = { + .required_flags = MPIC_MSI_FLAGS_REQUIRED, + .supported_flags = MPIC_MSI_FLAGS_SUPPORTED, + .bus_select_token = DOMAIN_BUS_NEXUS, + .bus_select_mask = MATCH_PCI_MSI, + .prefix = "MPIC-", + .init_dev_msi_info = msi_lib_init_dev_msi_info, +}; + static int __init mpic_msi_init(struct mpic *mpic, struct device_node *node, phys_addr_t main_int_phys_base) { @@ -348,17 +351,16 @@ static int __init mpic_msi_init(struct mpic *mpic, struct device_node *node, mpic->msi_doorbell_mask = PCI_MSI_FULL_DOORBELL_MASK; } - mpic->msi_inner_domain = irq_domain_create_linear(NULL, mpic->msi_doorbell_size, - &mpic_msi_domain_ops, mpic); - if (!mpic->msi_inner_domain) - return -ENOMEM; + struct irq_domain_info info = { + .fwnode = of_fwnode_handle(node), + .ops = &mpic_msi_domain_ops, + .host_data = mpic, + .size = mpic->msi_doorbell_size, + }; - mpic->msi_domain = pci_msi_create_irq_domain(of_fwnode_handle(node), &mpic_msi_domain_info, - mpic->msi_inner_domain); - if (!mpic->msi_domain) { - irq_domain_remove(mpic->msi_inner_domain); + mpic->msi_inner_domain = msi_create_parent_irq_domain(&info, &mpic_msi_parent_ops); + if (!mpic->msi_inner_domain) return -ENOMEM; - } mpic_msi_reenable_percpu(mpic); diff --git a/drivers/irqchip/irq-aspeed-scu-ic.c b/drivers/irqchip/irq-aspeed-scu-ic.c index 1c7045467c48..5584e0f82cce 100644 --- a/drivers/irqchip/irq-aspeed-scu-ic.c +++ b/drivers/irqchip/irq-aspeed-scu-ic.c @@ -1,61 +1,78 @@ // SPDX-License-Identifier: GPL-2.0-or-later /* - * Aspeed AST24XX, AST25XX, and AST26XX SCU Interrupt Controller + * Aspeed AST24XX, AST25XX, AST26XX, and AST27XX SCU Interrupt Controller * Copyright 2019 IBM Corporation * * Eddie James <eajames@linux.ibm.com> */ #include <linux/bitops.h> +#include <linux/io.h> #include <linux/irq.h> #include <linux/irqchip.h> #include <linux/irqchip/chained_irq.h> #include <linux/irqdomain.h> -#include <linux/mfd/syscon.h> +#include <linux/of_address.h> #include <linux/of_irq.h> -#include <linux/regmap.h> -#define ASPEED_SCU_IC_REG 0x018 -#define ASPEED_SCU_IC_SHIFT 0 -#define ASPEED_SCU_IC_ENABLE GENMASK(15, ASPEED_SCU_IC_SHIFT) -#define ASPEED_SCU_IC_NUM_IRQS 7 #define ASPEED_SCU_IC_STATUS GENMASK(28, 16) #define ASPEED_SCU_IC_STATUS_SHIFT 16 +#define AST2700_SCU_IC_STATUS GENMASK(15, 0) + +struct aspeed_scu_ic_variant { + const char *compatible; + unsigned long irq_enable; + unsigned long irq_shift; + unsigned int num_irqs; + unsigned long ier; + unsigned long isr; +}; -#define ASPEED_AST2600_SCU_IC0_REG 0x560 -#define ASPEED_AST2600_SCU_IC0_SHIFT 0 -#define ASPEED_AST2600_SCU_IC0_ENABLE \ - GENMASK(5, ASPEED_AST2600_SCU_IC0_SHIFT) -#define ASPEED_AST2600_SCU_IC0_NUM_IRQS 6 +#define SCU_VARIANT(_compat, _shift, _enable, _num, _ier, _isr) { \ + .compatible = _compat, \ + .irq_shift = _shift, \ + .irq_enable = _enable, \ + .num_irqs = _num, \ + .ier = _ier, \ + .isr = _isr, \ +} -#define ASPEED_AST2600_SCU_IC1_REG 0x570 -#define ASPEED_AST2600_SCU_IC1_SHIFT 4 -#define ASPEED_AST2600_SCU_IC1_ENABLE \ - GENMASK(5, ASPEED_AST2600_SCU_IC1_SHIFT) -#define ASPEED_AST2600_SCU_IC1_NUM_IRQS 2 +static const struct aspeed_scu_ic_variant scu_ic_variants[] __initconst = { + SCU_VARIANT("aspeed,ast2400-scu-ic", 0, GENMASK(15, 0), 7, 0x00, 0x00), + SCU_VARIANT("aspeed,ast2500-scu-ic", 0, GENMASK(15, 0), 7, 0x00, 0x00), + SCU_VARIANT("aspeed,ast2600-scu-ic0", 0, GENMASK(5, 0), 6, 0x00, 0x00), + SCU_VARIANT("aspeed,ast2600-scu-ic1", 4, GENMASK(5, 4), 2, 0x00, 0x00), + SCU_VARIANT("aspeed,ast2700-scu-ic0", 0, GENMASK(3, 0), 4, 0x00, 0x04), + SCU_VARIANT("aspeed,ast2700-scu-ic1", 0, GENMASK(3, 0), 4, 0x00, 0x04), + SCU_VARIANT("aspeed,ast2700-scu-ic2", 0, GENMASK(3, 0), 4, 0x04, 0x00), + SCU_VARIANT("aspeed,ast2700-scu-ic3", 0, GENMASK(1, 0), 2, 0x04, 0x00), +}; struct aspeed_scu_ic { - unsigned long irq_enable; - unsigned long irq_shift; - unsigned int num_irqs; - unsigned int reg; - struct regmap *scu; - struct irq_domain *irq_domain; + unsigned long irq_enable; + unsigned long irq_shift; + unsigned int num_irqs; + void __iomem *base; + struct irq_domain *irq_domain; + unsigned long ier; + unsigned long isr; }; -static void aspeed_scu_ic_irq_handler(struct irq_desc *desc) +static inline bool scu_has_split_isr(struct aspeed_scu_ic *scu) +{ + return scu->ier != scu->isr; +} + +static void aspeed_scu_ic_irq_handler_combined(struct irq_desc *desc) { - unsigned int sts; - unsigned long bit; - unsigned long enabled; - unsigned long max; - unsigned long status; struct aspeed_scu_ic *scu_ic = irq_desc_get_handler_data(desc); struct irq_chip *chip = irq_desc_get_chip(desc); - unsigned int mask = scu_ic->irq_enable << ASPEED_SCU_IC_STATUS_SHIFT; + unsigned long bit, enabled, max, status; + unsigned int sts, mask; chained_irq_enter(chip, desc); + mask = scu_ic->irq_enable << ASPEED_SCU_IC_STATUS_SHIFT; /* * The SCU IC has just one register to control its operation and read * status. The interrupt enable bits occupy the lower 16 bits of the @@ -66,7 +83,7 @@ static void aspeed_scu_ic_irq_handler(struct irq_desc *desc) * shifting the status down to get the mapping and then back up to * clear the bit. */ - regmap_read(scu_ic->scu, scu_ic->reg, &sts); + sts = readl(scu_ic->base); enabled = sts & scu_ic->irq_enable; status = (sts >> ASPEED_SCU_IC_STATUS_SHIFT) & enabled; @@ -74,43 +91,83 @@ static void aspeed_scu_ic_irq_handler(struct irq_desc *desc) max = scu_ic->num_irqs + bit; for_each_set_bit_from(bit, &status, max) { - generic_handle_domain_irq(scu_ic->irq_domain, - bit - scu_ic->irq_shift); + generic_handle_domain_irq(scu_ic->irq_domain, bit - scu_ic->irq_shift); + writel((readl(scu_ic->base) & ~mask) | BIT(bit + ASPEED_SCU_IC_STATUS_SHIFT), + scu_ic->base); + } + + chained_irq_exit(chip, desc); +} + +static void aspeed_scu_ic_irq_handler_split(struct irq_desc *desc) +{ + struct aspeed_scu_ic *scu_ic = irq_desc_get_handler_data(desc); + struct irq_chip *chip = irq_desc_get_chip(desc); + unsigned long bit, enabled, max, status; + unsigned int sts, mask; - regmap_write_bits(scu_ic->scu, scu_ic->reg, mask, - BIT(bit + ASPEED_SCU_IC_STATUS_SHIFT)); + chained_irq_enter(chip, desc); + + mask = scu_ic->irq_enable; + sts = readl(scu_ic->base + scu_ic->isr); + enabled = sts & scu_ic->irq_enable; + sts = readl(scu_ic->base + scu_ic->isr); + status = sts & enabled; + + bit = scu_ic->irq_shift; + max = scu_ic->num_irqs + bit; + + for_each_set_bit_from(bit, &status, max) { + generic_handle_domain_irq(scu_ic->irq_domain, bit - scu_ic->irq_shift); + /* Clear interrupt */ + writel(BIT(bit), scu_ic->base + scu_ic->isr); } chained_irq_exit(chip, desc); } -static void aspeed_scu_ic_irq_mask(struct irq_data *data) +static void aspeed_scu_ic_irq_mask_combined(struct irq_data *data) { struct aspeed_scu_ic *scu_ic = irq_data_get_irq_chip_data(data); - unsigned int mask = BIT(data->hwirq + scu_ic->irq_shift) | - (scu_ic->irq_enable << ASPEED_SCU_IC_STATUS_SHIFT); + unsigned int bit = BIT(data->hwirq + scu_ic->irq_shift); + unsigned int mask = bit | (scu_ic->irq_enable << ASPEED_SCU_IC_STATUS_SHIFT); /* * Status bits are cleared by writing 1. In order to prevent the mask * operation from clearing the status bits, they should be under the * mask and written with 0. */ - regmap_update_bits(scu_ic->scu, scu_ic->reg, mask, 0); + writel(readl(scu_ic->base) & ~mask, scu_ic->base); } -static void aspeed_scu_ic_irq_unmask(struct irq_data *data) +static void aspeed_scu_ic_irq_unmask_combined(struct irq_data *data) { struct aspeed_scu_ic *scu_ic = irq_data_get_irq_chip_data(data); unsigned int bit = BIT(data->hwirq + scu_ic->irq_shift); - unsigned int mask = bit | - (scu_ic->irq_enable << ASPEED_SCU_IC_STATUS_SHIFT); + unsigned int mask = bit | (scu_ic->irq_enable << ASPEED_SCU_IC_STATUS_SHIFT); /* * Status bits are cleared by writing 1. In order to prevent the unmask * operation from clearing the status bits, they should be under the * mask and written with 0. */ - regmap_update_bits(scu_ic->scu, scu_ic->reg, mask, bit); + writel((readl(scu_ic->base) & ~mask) | bit, scu_ic->base); +} + +static void aspeed_scu_ic_irq_mask_split(struct irq_data *data) +{ + struct aspeed_scu_ic *scu_ic = irq_data_get_irq_chip_data(data); + unsigned int mask = BIT(data->hwirq + scu_ic->irq_shift); + + writel(readl(scu_ic->base) & ~mask, scu_ic->base + scu_ic->ier); +} + +static void aspeed_scu_ic_irq_unmask_split(struct irq_data *data) +{ + struct aspeed_scu_ic *scu_ic = irq_data_get_irq_chip_data(data); + unsigned int bit = BIT(data->hwirq + scu_ic->irq_shift); + + writel(readl(scu_ic->base) | bit, scu_ic->base + scu_ic->ier); } static int aspeed_scu_ic_irq_set_affinity(struct irq_data *data, @@ -120,17 +177,29 @@ static int aspeed_scu_ic_irq_set_affinity(struct irq_data *data, return -EINVAL; } -static struct irq_chip aspeed_scu_ic_chip = { +static struct irq_chip aspeed_scu_ic_chip_combined = { .name = "aspeed-scu-ic", - .irq_mask = aspeed_scu_ic_irq_mask, - .irq_unmask = aspeed_scu_ic_irq_unmask, - .irq_set_affinity = aspeed_scu_ic_irq_set_affinity, + .irq_mask = aspeed_scu_ic_irq_mask_combined, + .irq_unmask = aspeed_scu_ic_irq_unmask_combined, + .irq_set_affinity = aspeed_scu_ic_irq_set_affinity, +}; + +static struct irq_chip aspeed_scu_ic_chip_split = { + .name = "ast2700-scu-ic", + .irq_mask = aspeed_scu_ic_irq_mask_split, + .irq_unmask = aspeed_scu_ic_irq_unmask_split, + .irq_set_affinity = aspeed_scu_ic_irq_set_affinity, }; static int aspeed_scu_ic_map(struct irq_domain *domain, unsigned int irq, irq_hw_number_t hwirq) { - irq_set_chip_and_handler(irq, &aspeed_scu_ic_chip, handle_level_irq); + struct aspeed_scu_ic *scu_ic = domain->host_data; + + if (scu_has_split_isr(scu_ic)) + irq_set_chip_and_handler(irq, &aspeed_scu_ic_chip_split, handle_level_irq); + else + irq_set_chip_and_handler(irq, &aspeed_scu_ic_chip_combined, handle_level_irq); irq_set_chip_data(irq, domain->host_data); return 0; @@ -143,21 +212,21 @@ static const struct irq_domain_ops aspeed_scu_ic_domain_ops = { static int aspeed_scu_ic_of_init_common(struct aspeed_scu_ic *scu_ic, struct device_node *node) { - int irq; - int rc = 0; + int irq, rc = 0; - if (!node->parent) { - rc = -ENODEV; + scu_ic->base = of_iomap(node, 0); + if (IS_ERR(scu_ic->base)) { + rc = PTR_ERR(scu_ic->base); goto err; } - scu_ic->scu = syscon_node_to_regmap(node->parent); - if (IS_ERR(scu_ic->scu)) { - rc = PTR_ERR(scu_ic->scu); - goto err; + if (scu_has_split_isr(scu_ic)) { + writel(AST2700_SCU_IC_STATUS, scu_ic->base + scu_ic->isr); + writel(0, scu_ic->base + scu_ic->ier); + } else { + writel(ASPEED_SCU_IC_STATUS, scu_ic->base); + writel(0, scu_ic->base); } - regmap_write_bits(scu_ic->scu, scu_ic->reg, ASPEED_SCU_IC_STATUS, ASPEED_SCU_IC_STATUS); - regmap_write_bits(scu_ic->scu, scu_ic->reg, ASPEED_SCU_IC_ENABLE, 0); irq = irq_of_parse_and_map(node, 0); if (!irq) { @@ -166,75 +235,60 @@ static int aspeed_scu_ic_of_init_common(struct aspeed_scu_ic *scu_ic, } scu_ic->irq_domain = irq_domain_create_linear(of_fwnode_handle(node), scu_ic->num_irqs, - &aspeed_scu_ic_domain_ops, - scu_ic); + &aspeed_scu_ic_domain_ops, scu_ic); if (!scu_ic->irq_domain) { rc = -ENOMEM; goto err; } - irq_set_chained_handler_and_data(irq, aspeed_scu_ic_irq_handler, + irq_set_chained_handler_and_data(irq, scu_has_split_isr(scu_ic) ? + aspeed_scu_ic_irq_handler_split : + aspeed_scu_ic_irq_handler_combined, scu_ic); return 0; err: kfree(scu_ic); - return rc; } -static int __init aspeed_scu_ic_of_init(struct device_node *node, - struct device_node *parent) +static const struct aspeed_scu_ic_variant *aspeed_scu_ic_find_variant(struct device_node *np) { - struct aspeed_scu_ic *scu_ic = kzalloc(sizeof(*scu_ic), GFP_KERNEL); - - if (!scu_ic) - return -ENOMEM; - - scu_ic->irq_enable = ASPEED_SCU_IC_ENABLE; - scu_ic->irq_shift = ASPEED_SCU_IC_SHIFT; - scu_ic->num_irqs = ASPEED_SCU_IC_NUM_IRQS; - scu_ic->reg = ASPEED_SCU_IC_REG; - - return aspeed_scu_ic_of_init_common(scu_ic, node); + for (int i = 0; i < ARRAY_SIZE(scu_ic_variants); i++) { + if (of_device_is_compatible(np, scu_ic_variants[i].compatible)) + return &scu_ic_variants[i]; + } + return NULL; } -static int __init aspeed_ast2600_scu_ic0_of_init(struct device_node *node, - struct device_node *parent) +static int __init aspeed_scu_ic_of_init(struct device_node *node, struct device_node *parent) { - struct aspeed_scu_ic *scu_ic = kzalloc(sizeof(*scu_ic), GFP_KERNEL); + const struct aspeed_scu_ic_variant *variant; + struct aspeed_scu_ic *scu_ic; - if (!scu_ic) - return -ENOMEM; - - scu_ic->irq_enable = ASPEED_AST2600_SCU_IC0_ENABLE; - scu_ic->irq_shift = ASPEED_AST2600_SCU_IC0_SHIFT; - scu_ic->num_irqs = ASPEED_AST2600_SCU_IC0_NUM_IRQS; - scu_ic->reg = ASPEED_AST2600_SCU_IC0_REG; - - return aspeed_scu_ic_of_init_common(scu_ic, node); -} - -static int __init aspeed_ast2600_scu_ic1_of_init(struct device_node *node, - struct device_node *parent) -{ - struct aspeed_scu_ic *scu_ic = kzalloc(sizeof(*scu_ic), GFP_KERNEL); + variant = aspeed_scu_ic_find_variant(node); + if (!variant) + return -ENODEV; + scu_ic = kzalloc(sizeof(*scu_ic), GFP_KERNEL); if (!scu_ic) return -ENOMEM; - scu_ic->irq_enable = ASPEED_AST2600_SCU_IC1_ENABLE; - scu_ic->irq_shift = ASPEED_AST2600_SCU_IC1_SHIFT; - scu_ic->num_irqs = ASPEED_AST2600_SCU_IC1_NUM_IRQS; - scu_ic->reg = ASPEED_AST2600_SCU_IC1_REG; + scu_ic->irq_enable = variant->irq_enable; + scu_ic->irq_shift = variant->irq_shift; + scu_ic->num_irqs = variant->num_irqs; + scu_ic->ier = variant->ier; + scu_ic->isr = variant->isr; return aspeed_scu_ic_of_init_common(scu_ic, node); } IRQCHIP_DECLARE(ast2400_scu_ic, "aspeed,ast2400-scu-ic", aspeed_scu_ic_of_init); IRQCHIP_DECLARE(ast2500_scu_ic, "aspeed,ast2500-scu-ic", aspeed_scu_ic_of_init); -IRQCHIP_DECLARE(ast2600_scu_ic0, "aspeed,ast2600-scu-ic0", - aspeed_ast2600_scu_ic0_of_init); -IRQCHIP_DECLARE(ast2600_scu_ic1, "aspeed,ast2600-scu-ic1", - aspeed_ast2600_scu_ic1_of_init); +IRQCHIP_DECLARE(ast2600_scu_ic0, "aspeed,ast2600-scu-ic0", aspeed_scu_ic_of_init); +IRQCHIP_DECLARE(ast2600_scu_ic1, "aspeed,ast2600-scu-ic1", aspeed_scu_ic_of_init); +IRQCHIP_DECLARE(ast2700_scu_ic0, "aspeed,ast2700-scu-ic0", aspeed_scu_ic_of_init); +IRQCHIP_DECLARE(ast2700_scu_ic1, "aspeed,ast2700-scu-ic1", aspeed_scu_ic_of_init); +IRQCHIP_DECLARE(ast2700_scu_ic2, "aspeed,ast2700-scu-ic2", aspeed_scu_ic_of_init); +IRQCHIP_DECLARE(ast2700_scu_ic3, "aspeed,ast2700-scu-ic3", aspeed_scu_ic_of_init); diff --git a/drivers/irqchip/irq-atmel-aic.c b/drivers/irqchip/irq-atmel-aic.c index 03aeed39a4d2..1dcc52760eca 100644 --- a/drivers/irqchip/irq-atmel-aic.c +++ b/drivers/irqchip/irq-atmel-aic.c @@ -188,7 +188,7 @@ static int aic_irq_domain_xlate(struct irq_domain *d, gc = dgc->gc[idx]; - guard(raw_spinlock_irq)(&gc->lock); + guard(raw_spinlock_irqsave)(&gc->lock); smr = irq_reg_readl(gc, AT91_AIC_SMR(*out_hwirq)); aic_common_set_priority(intspec[2], &smr); irq_reg_writel(gc, smr, AT91_AIC_SMR(*out_hwirq)); diff --git a/drivers/irqchip/irq-atmel-aic5.c b/drivers/irqchip/irq-atmel-aic5.c index 60b00d2c3d7a..1f14b401f71d 100644 --- a/drivers/irqchip/irq-atmel-aic5.c +++ b/drivers/irqchip/irq-atmel-aic5.c @@ -279,7 +279,7 @@ static int aic5_irq_domain_xlate(struct irq_domain *d, if (ret) return ret; - guard(raw_spinlock_irq)(&bgc->lock); + guard(raw_spinlock_irqsave)(&bgc->lock); irq_reg_writel(bgc, *out_hwirq, AT91_AIC5_SSR); smr = irq_reg_readl(bgc, AT91_AIC5_SMR); aic_common_set_priority(intspec[2], &smr); diff --git a/drivers/irqchip/irq-bcm2712-mip.c b/drivers/irqchip/irq-bcm2712-mip.c index 63de5ef6cf2d..9bd7bc0bf6d5 100644 --- a/drivers/irqchip/irq-bcm2712-mip.c +++ b/drivers/irqchip/irq-bcm2712-mip.c @@ -172,18 +172,18 @@ static const struct msi_parent_ops mip_msi_parent_ops = { static int mip_init_domains(struct mip_priv *mip, struct device_node *np) { - struct irq_domain *middle; - - middle = irq_domain_create_hierarchy(mip->parent, 0, mip->num_msis, of_fwnode_handle(np), - &mip_middle_domain_ops, mip); - if (!middle) + struct irq_domain_info info = { + .fwnode = of_fwnode_handle(np), + .ops = &mip_middle_domain_ops, + .host_data = mip, + .size = mip->num_msis, + .parent = mip->parent, + .dev = mip->dev, + }; + + if (!msi_create_parent_irq_domain(&info, &mip_msi_parent_ops)) return -ENOMEM; - irq_domain_update_bus_token(middle, DOMAIN_BUS_GENERIC_MSI); - middle->dev = mip->dev; - middle->flags |= IRQ_DOMAIN_FLAG_MSI_PARENT; - middle->msi_parent_ops = &mip_msi_parent_ops; - /* * All MSI-X unmasked for the host, masked for the VPU, and edge-triggered. */ diff --git a/drivers/irqchip/irq-gic-common.h b/drivers/irqchip/irq-gic-common.h index 020ecdf16901..710cab61d919 100644 --- a/drivers/irqchip/irq-gic-common.h +++ b/drivers/irqchip/irq-gic-common.h @@ -29,8 +29,6 @@ void gic_enable_quirks(u32 iidr, const struct gic_quirk *quirks, void gic_enable_of_quirks(const struct device_node *np, const struct gic_quirk *quirks, void *data); -extern const struct msi_parent_ops gic_v3_its_msi_parent_ops; - #define RDIST_FLAGS_PROPBASE_NEEDS_FLUSHING (1 << 0) #define RDIST_FLAGS_RD_TABLES_PREALLOCATED (1 << 1) #define RDIST_FLAGS_FORCE_NON_SHAREABLE (1 << 2) diff --git a/drivers/irqchip/irq-gic-v3-its-msi-parent.c b/drivers/irqchip/irq-gic-its-msi-parent.c index a5e110ffdd88..eb1473f1448a 100644 --- a/drivers/irqchip/irq-gic-v3-its-msi-parent.c +++ b/drivers/irqchip/irq-gic-its-msi-parent.c @@ -5,9 +5,10 @@ // Copyright (C) 2022 Intel #include <linux/acpi_iort.h> +#include <linux/of_address.h> #include <linux/pci.h> -#include "irq-gic-common.h" +#include "irq-gic-its-msi-parent.h" #include <linux/irqchip/irq-msi-lib.h> #define ITS_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS | \ @@ -18,6 +19,23 @@ MSI_FLAG_PCI_MSIX | \ MSI_FLAG_MULTI_PCI_MSI) +static int its_translate_frame_address(struct device_node *msi_node, phys_addr_t *pa) +{ + struct resource res; + int ret; + + ret = of_property_match_string(msi_node, "reg-names", "ns-translate"); + if (ret < 0) + return ret; + + ret = of_address_to_resource(msi_node, ret, &res); + if (ret) + return ret; + + *pa = res.start; + return 0; +} + #ifdef CONFIG_PCI_MSI static int its_pci_msi_vec_count(struct pci_dev *pdev, void *data) { @@ -82,8 +100,46 @@ static int its_pci_msi_prepare(struct irq_domain *domain, struct device *dev, msi_info = msi_get_domain_info(domain->parent); return msi_info->ops->msi_prepare(domain->parent, dev, nvec, info); } + +static int its_v5_pci_msi_prepare(struct irq_domain *domain, struct device *dev, + int nvec, msi_alloc_info_t *info) +{ + struct device_node *msi_node = NULL; + struct msi_domain_info *msi_info; + struct pci_dev *pdev; + phys_addr_t pa; + u32 rid; + int ret; + + if (!dev_is_pci(dev)) + return -EINVAL; + + pdev = to_pci_dev(dev); + + rid = pci_msi_map_rid_ctlr_node(pdev, &msi_node); + if (!msi_node) + return -ENODEV; + + ret = its_translate_frame_address(msi_node, &pa); + if (ret) + return -ENODEV; + + of_node_put(msi_node); + + /* ITS specific DeviceID */ + info->scratchpad[0].ul = rid; + /* ITS translate frame physical address */ + info->scratchpad[1].ul = pa; + + /* Always allocate power of two vectors */ + nvec = roundup_pow_of_two(nvec); + + msi_info = msi_get_domain_info(domain->parent); + return msi_info->ops->msi_prepare(domain->parent, dev, nvec, info); +} #else /* CONFIG_PCI_MSI */ #define its_pci_msi_prepare NULL +#define its_v5_pci_msi_prepare NULL #endif /* !CONFIG_PCI_MSI */ static int of_pmsi_get_dev_id(struct irq_domain *domain, struct device *dev, @@ -118,6 +174,53 @@ static int of_pmsi_get_dev_id(struct irq_domain *domain, struct device *dev, return ret; } +static int of_v5_pmsi_get_msi_info(struct irq_domain *domain, struct device *dev, + u32 *dev_id, phys_addr_t *pa) +{ + int ret, index = 0; + /* + * Retrieve the DeviceID and the ITS translate frame node pointer + * out of the msi-parent property. + */ + do { + struct of_phandle_args args; + + ret = of_parse_phandle_with_args(dev->of_node, + "msi-parent", "#msi-cells", + index, &args); + if (ret) + break; + /* + * The IRQ domain fwnode is the msi controller parent + * in GICv5 (where the msi controller nodes are the + * ITS translate frames). + */ + if (args.np->parent == irq_domain_get_of_node(domain)) { + if (WARN_ON(args.args_count != 1)) + return -EINVAL; + *dev_id = args.args[0]; + + ret = its_translate_frame_address(args.np, pa); + if (ret) + return -ENODEV; + break; + } + index++; + } while (!ret); + + if (ret) { + struct device_node *np = NULL; + + ret = of_map_id(dev->of_node, dev->id, "msi-map", "msi-map-mask", &np, dev_id); + if (np) { + ret = its_translate_frame_address(np, pa); + of_node_put(np); + } + } + + return ret; +} + int __weak iort_pmsi_get_dev_id(struct device *dev, u32 *dev_id) { return -1; @@ -148,6 +251,33 @@ static int its_pmsi_prepare(struct irq_domain *domain, struct device *dev, dev, nvec, info); } +static int its_v5_pmsi_prepare(struct irq_domain *domain, struct device *dev, + int nvec, msi_alloc_info_t *info) +{ + struct msi_domain_info *msi_info; + phys_addr_t pa; + u32 dev_id; + int ret; + + if (!dev->of_node) + return -ENODEV; + + ret = of_v5_pmsi_get_msi_info(domain->parent, dev, &dev_id, &pa); + if (ret) + return ret; + + /* ITS specific DeviceID */ + info->scratchpad[0].ul = dev_id; + /* ITS translate frame physical address */ + info->scratchpad[1].ul = pa; + + /* Allocate always as a power of 2 */ + nvec = roundup_pow_of_two(nvec); + + msi_info = msi_get_domain_info(domain->parent); + return msi_info->ops->msi_prepare(domain->parent, dev, nvec, info); +} + static void its_msi_teardown(struct irq_domain *domain, msi_alloc_info_t *info) { struct msi_domain_info *msi_info; @@ -199,6 +329,32 @@ static bool its_init_dev_msi_info(struct device *dev, struct irq_domain *domain, return true; } +static bool its_v5_init_dev_msi_info(struct device *dev, struct irq_domain *domain, + struct irq_domain *real_parent, struct msi_domain_info *info) +{ + if (!msi_lib_init_dev_msi_info(dev, domain, real_parent, info)) + return false; + + switch (info->bus_token) { + case DOMAIN_BUS_PCI_DEVICE_MSI: + case DOMAIN_BUS_PCI_DEVICE_MSIX: + info->ops->msi_prepare = its_v5_pci_msi_prepare; + info->ops->msi_teardown = its_msi_teardown; + break; + case DOMAIN_BUS_DEVICE_MSI: + case DOMAIN_BUS_WIRED_TO_MSI: + info->ops->msi_prepare = its_v5_pmsi_prepare; + info->ops->msi_teardown = its_msi_teardown; + break; + default: + /* Confused. How did the lib return true? */ + WARN_ON_ONCE(1); + return false; + } + + return true; +} + const struct msi_parent_ops gic_v3_its_msi_parent_ops = { .supported_flags = ITS_MSI_FLAGS_SUPPORTED, .required_flags = ITS_MSI_FLAGS_REQUIRED, @@ -208,3 +364,13 @@ const struct msi_parent_ops gic_v3_its_msi_parent_ops = { .prefix = "ITS-", .init_dev_msi_info = its_init_dev_msi_info, }; + +const struct msi_parent_ops gic_v5_its_msi_parent_ops = { + .supported_flags = ITS_MSI_FLAGS_SUPPORTED, + .required_flags = ITS_MSI_FLAGS_REQUIRED, + .chip_flags = MSI_CHIP_FLAG_SET_EOI, + .bus_select_token = DOMAIN_BUS_NEXUS, + .bus_select_mask = MATCH_PCI_MSI | MATCH_PLATFORM_MSI, + .prefix = "ITS-v5-", + .init_dev_msi_info = its_v5_init_dev_msi_info, +}; diff --git a/drivers/irqchip/irq-gic-its-msi-parent.h b/drivers/irqchip/irq-gic-its-msi-parent.h new file mode 100644 index 000000000000..df016f347337 --- /dev/null +++ b/drivers/irqchip/irq-gic-its-msi-parent.h @@ -0,0 +1,12 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ +/* + * Copyright (C) 2024 ARM Limited, All Rights Reserved. + */ + +#ifndef _IRQ_GIC_ITS_MSI_PARENT_H +#define _IRQ_GIC_ITS_MSI_PARENT_H + +extern const struct msi_parent_ops gic_v3_its_msi_parent_ops; +extern const struct msi_parent_ops gic_v5_its_msi_parent_ops; + +#endif /* _IRQ_GIC_ITS_MSI_PARENT_H */ diff --git a/drivers/irqchip/irq-gic-v2m.c b/drivers/irqchip/irq-gic-v2m.c index 24ef5af569fe..8a3410c2b7b5 100644 --- a/drivers/irqchip/irq-gic-v2m.c +++ b/drivers/irqchip/irq-gic-v2m.c @@ -153,14 +153,19 @@ static int gicv2m_irq_domain_alloc(struct irq_domain *domain, unsigned int virq, { msi_alloc_info_t *info = args; struct v2m_data *v2m = NULL, *tmp; - int hwirq, offset, i, err = 0; + int hwirq, i, err = 0; + unsigned long offset; + unsigned long align_mask = nr_irqs - 1; spin_lock(&v2m_lock); list_for_each_entry(tmp, &v2m_nodes, entry) { - offset = bitmap_find_free_region(tmp->bm, tmp->nr_spis, - get_count_order(nr_irqs)); - if (offset >= 0) { + unsigned long align_off = tmp->spi_start - (tmp->spi_start & ~align_mask); + + offset = bitmap_find_next_zero_area_off(tmp->bm, tmp->nr_spis, 0, + nr_irqs, align_mask, align_off); + if (offset < tmp->nr_spis) { v2m = tmp; + bitmap_set(v2m->bm, offset, nr_irqs); break; } } diff --git a/drivers/irqchip/irq-gic-v3-its-fsl-mc-msi.c b/drivers/irqchip/irq-gic-v3-its-fsl-mc-msi.c index 11549d85f23b..b5785472765a 100644 --- a/drivers/irqchip/irq-gic-v3-its-fsl-mc-msi.c +++ b/drivers/irqchip/irq-gic-v3-its-fsl-mc-msi.c @@ -30,7 +30,7 @@ static u32 fsl_mc_msi_domain_get_msi_id(struct irq_domain *domain, u32 out_id; of_node = irq_domain_get_of_node(domain); - out_id = of_node ? of_msi_map_id(&mc_dev->dev, of_node, mc_dev->icid) : + out_id = of_node ? of_msi_xlate(&mc_dev->dev, &of_node, mc_dev->icid) : iort_msi_map_id(&mc_dev->dev, mc_dev->icid); return out_id; diff --git a/drivers/irqchip/irq-gic-v3-its.c b/drivers/irqchip/irq-gic-v3-its.c index d54fa0638dc4..467cb78435a9 100644 --- a/drivers/irqchip/irq-gic-v3-its.c +++ b/drivers/irqchip/irq-gic-v3-its.c @@ -41,6 +41,7 @@ #include <asm/exception.h> #include "irq-gic-common.h" +#include "irq-gic-its-msi-parent.h" #include <linux/irqchip/irq-msi-lib.h> #define ITS_FLAGS_CMDQ_NEEDS_FLUSHING (1ULL << 0) diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c index efc791c43d44..3de351e66ee8 100644 --- a/drivers/irqchip/irq-gic-v3.c +++ b/drivers/irqchip/irq-gic-v3.c @@ -190,12 +190,12 @@ static void __init gic_prio_init(void) /* * How priority values are used by the GIC depends on two things: - * the security state of the GIC (controlled by the GICD_CTRL.DS bit) + * the security state of the GIC (controlled by the GICD_CTLR.DS bit) * and if Group 0 interrupts can be delivered to Linux in the non-secure * world as FIQs (controlled by the SCR_EL3.FIQ bit). These affect the * way priorities are presented in ICC_PMR_EL1 and in the distributor: * - * GICD_CTRL.DS | SCR_EL3.FIQ | ICC_PMR_EL1 | Distributor + * GICD_CTLR.DS | SCR_EL3.FIQ | ICC_PMR_EL1 | Distributor * ------------------------------------------------------- * 1 | - | unchanged | unchanged * ------------------------------------------------------- @@ -223,7 +223,7 @@ static void __init gic_prio_init(void) dist_prio_nmi = __gicv3_prio_to_ns(dist_prio_nmi); } - pr_info("GICD_CTRL.DS=%d, SCR_EL3.FIQ=%d\n", + pr_info("GICD_CTLR.DS=%d, SCR_EL3.FIQ=%d\n", cpus_have_security_disabled, !cpus_have_group0); } @@ -1766,8 +1766,9 @@ static int gic_irq_domain_select(struct irq_domain *d, struct irq_fwspec *fwspec, enum irq_domain_bus_token bus_token) { - unsigned int type, ret, ppi_idx; + unsigned int type, ppi_idx; irq_hw_number_t hwirq; + int ret; /* Not for us */ if (fwspec->fwnode != d->fwnode) diff --git a/drivers/irqchip/irq-gic-v4.c b/drivers/irqchip/irq-gic-v4.c index 58c28895f8c4..8455b4a5fbb0 100644 --- a/drivers/irqchip/irq-gic-v4.c +++ b/drivers/irqchip/irq-gic-v4.c @@ -342,10 +342,10 @@ int its_get_vlpi(int irq, struct its_vlpi_map *map) return irq_set_vcpu_affinity(irq, &info); } -int its_unmap_vlpi(int irq) +void its_unmap_vlpi(int irq) { irq_clear_status_flags(irq, IRQ_DISABLE_UNLAZY); - return irq_set_vcpu_affinity(irq, NULL); + WARN_ON_ONCE(irq_set_vcpu_affinity(irq, NULL)); } int its_prop_update_vlpi(int irq, u8 config, bool inv) diff --git a/drivers/irqchip/irq-gic-v5-irs.c b/drivers/irqchip/irq-gic-v5-irs.c new file mode 100644 index 000000000000..ce2732d649a3 --- /dev/null +++ b/drivers/irqchip/irq-gic-v5-irs.c @@ -0,0 +1,827 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2024-2025 ARM Limited, All Rights Reserved. + */ + +#define pr_fmt(fmt) "GICv5 IRS: " fmt + +#include <linux/kmemleak.h> +#include <linux/log2.h> +#include <linux/of.h> +#include <linux/of_address.h> + +#include <linux/irqchip.h> +#include <linux/irqchip/arm-gic-v5.h> + +/* + * Hardcoded ID_BITS limit for systems supporting only a 1-level IST + * table. Systems supporting only a 1-level IST table aren't expected + * to require more than 2^12 LPIs. Tweak as required. + */ +#define LPI_ID_BITS_LINEAR 12 + +#define IRS_FLAGS_NON_COHERENT BIT(0) + +static DEFINE_PER_CPU_READ_MOSTLY(struct gicv5_irs_chip_data *, per_cpu_irs_data); +static LIST_HEAD(irs_nodes); + +static u32 irs_readl_relaxed(struct gicv5_irs_chip_data *irs_data, + const u32 reg_offset) +{ + return readl_relaxed(irs_data->irs_base + reg_offset); +} + +static void irs_writel_relaxed(struct gicv5_irs_chip_data *irs_data, + const u32 val, const u32 reg_offset) +{ + writel_relaxed(val, irs_data->irs_base + reg_offset); +} + +static u64 irs_readq_relaxed(struct gicv5_irs_chip_data *irs_data, + const u32 reg_offset) +{ + return readq_relaxed(irs_data->irs_base + reg_offset); +} + +static void irs_writeq_relaxed(struct gicv5_irs_chip_data *irs_data, + const u64 val, const u32 reg_offset) +{ + writeq_relaxed(val, irs_data->irs_base + reg_offset); +} + +/* + * The polling wait (in gicv5_wait_for_op_s_atomic()) on a GIC register + * provides the memory barriers (through MMIO accessors) + * required to synchronize CPU and GIC access to IST memory. + */ +static int gicv5_irs_ist_synchronise(struct gicv5_irs_chip_data *irs_data) +{ + return gicv5_wait_for_op_atomic(irs_data->irs_base, GICV5_IRS_IST_STATUSR, + GICV5_IRS_IST_STATUSR_IDLE, NULL); +} + +static int __init gicv5_irs_init_ist_linear(struct gicv5_irs_chip_data *irs_data, + unsigned int lpi_id_bits, + unsigned int istsz) +{ + size_t l2istsz; + u32 n, cfgr; + void *ist; + u64 baser; + int ret; + + /* Taken from GICv5 specifications 10.2.1.13 IRS_IST_BASER */ + n = max(5, lpi_id_bits + 1 + istsz); + + l2istsz = BIT(n + 1); + /* + * Check memory requirements. For a linear IST we cap the + * number of ID bits to a value that should never exceed + * kmalloc interface memory allocation limits, so this + * check is really belt and braces. + */ + if (l2istsz > KMALLOC_MAX_SIZE) { + u8 lpi_id_cap = ilog2(KMALLOC_MAX_SIZE) - 2 + istsz; + + pr_warn("Limiting LPI ID bits from %u to %u\n", + lpi_id_bits, lpi_id_cap); + lpi_id_bits = lpi_id_cap; + l2istsz = KMALLOC_MAX_SIZE; + } + + ist = kzalloc(l2istsz, GFP_KERNEL); + if (!ist) + return -ENOMEM; + + if (irs_data->flags & IRS_FLAGS_NON_COHERENT) + dcache_clean_inval_poc((unsigned long)ist, + (unsigned long)ist + l2istsz); + else + dsb(ishst); + + cfgr = FIELD_PREP(GICV5_IRS_IST_CFGR_STRUCTURE, + GICV5_IRS_IST_CFGR_STRUCTURE_LINEAR) | + FIELD_PREP(GICV5_IRS_IST_CFGR_ISTSZ, istsz) | + FIELD_PREP(GICV5_IRS_IST_CFGR_L2SZ, + GICV5_IRS_IST_CFGR_L2SZ_4K) | + FIELD_PREP(GICV5_IRS_IST_CFGR_LPI_ID_BITS, lpi_id_bits); + irs_writel_relaxed(irs_data, cfgr, GICV5_IRS_IST_CFGR); + + gicv5_global_data.ist.l2 = false; + + baser = (virt_to_phys(ist) & GICV5_IRS_IST_BASER_ADDR_MASK) | + FIELD_PREP(GICV5_IRS_IST_BASER_VALID, 0x1); + irs_writeq_relaxed(irs_data, baser, GICV5_IRS_IST_BASER); + + ret = gicv5_irs_ist_synchronise(irs_data); + if (ret) { + kfree(ist); + return ret; + } + kmemleak_ignore(ist); + + return 0; +} + +static int __init gicv5_irs_init_ist_two_level(struct gicv5_irs_chip_data *irs_data, + unsigned int lpi_id_bits, + unsigned int istsz, + unsigned int l2sz) +{ + __le64 *l1ist; + u32 cfgr, n; + size_t l1sz; + u64 baser; + int ret; + + /* Taken from GICv5 specifications 10.2.1.13 IRS_IST_BASER */ + n = max(5, lpi_id_bits - ((10 - istsz) + (2 * l2sz)) + 2); + + l1sz = BIT(n + 1); + + l1ist = kzalloc(l1sz, GFP_KERNEL); + if (!l1ist) + return -ENOMEM; + + if (irs_data->flags & IRS_FLAGS_NON_COHERENT) + dcache_clean_inval_poc((unsigned long)l1ist, + (unsigned long)l1ist + l1sz); + else + dsb(ishst); + + cfgr = FIELD_PREP(GICV5_IRS_IST_CFGR_STRUCTURE, + GICV5_IRS_IST_CFGR_STRUCTURE_TWO_LEVEL) | + FIELD_PREP(GICV5_IRS_IST_CFGR_ISTSZ, istsz) | + FIELD_PREP(GICV5_IRS_IST_CFGR_L2SZ, l2sz) | + FIELD_PREP(GICV5_IRS_IST_CFGR_LPI_ID_BITS, lpi_id_bits); + irs_writel_relaxed(irs_data, cfgr, GICV5_IRS_IST_CFGR); + + /* + * The L2SZ determine bits required at L2 level. Number of bytes + * required by metadata is reported through istsz - the number of bits + * covered by L2 entries scales accordingly. + */ + gicv5_global_data.ist.l2_size = BIT(11 + (2 * l2sz) + 1); + gicv5_global_data.ist.l2_bits = (10 - istsz) + (2 * l2sz); + gicv5_global_data.ist.l1ist_addr = l1ist; + gicv5_global_data.ist.l2 = true; + + baser = (virt_to_phys(l1ist) & GICV5_IRS_IST_BASER_ADDR_MASK) | + FIELD_PREP(GICV5_IRS_IST_BASER_VALID, 0x1); + irs_writeq_relaxed(irs_data, baser, GICV5_IRS_IST_BASER); + + ret = gicv5_irs_ist_synchronise(irs_data); + if (ret) { + kfree(l1ist); + return ret; + } + + return 0; +} + +/* + * Alloc L2 IST entries on demand. + * + * Locking/serialization is guaranteed by irqdomain core code by + * taking the hierarchical domain struct irq_domain.root->mutex. + */ +int gicv5_irs_iste_alloc(const u32 lpi) +{ + struct gicv5_irs_chip_data *irs_data; + unsigned int index; + u32 l2istr, l2bits; + __le64 *l1ist; + size_t l2size; + void *l2ist; + int ret; + + if (!gicv5_global_data.ist.l2) + return 0; + + irs_data = per_cpu(per_cpu_irs_data, smp_processor_id()); + if (!irs_data) + return -ENOENT; + + l2size = gicv5_global_data.ist.l2_size; + l2bits = gicv5_global_data.ist.l2_bits; + l1ist = gicv5_global_data.ist.l1ist_addr; + index = lpi >> l2bits; + + if (FIELD_GET(GICV5_ISTL1E_VALID, le64_to_cpu(l1ist[index]))) + return 0; + + l2ist = kzalloc(l2size, GFP_KERNEL); + if (!l2ist) + return -ENOMEM; + + l1ist[index] = cpu_to_le64(virt_to_phys(l2ist) & GICV5_ISTL1E_L2_ADDR_MASK); + + if (irs_data->flags & IRS_FLAGS_NON_COHERENT) { + dcache_clean_inval_poc((unsigned long)l2ist, + (unsigned long)l2ist + l2size); + dcache_clean_poc((unsigned long)(l1ist + index), + (unsigned long)(l1ist + index) + sizeof(*l1ist)); + } else { + dsb(ishst); + } + + l2istr = FIELD_PREP(GICV5_IRS_MAP_L2_ISTR_ID, lpi); + irs_writel_relaxed(irs_data, l2istr, GICV5_IRS_MAP_L2_ISTR); + + ret = gicv5_irs_ist_synchronise(irs_data); + if (ret) { + l1ist[index] = 0; + kfree(l2ist); + return ret; + } + kmemleak_ignore(l2ist); + + /* + * Make sure we invalidate the cache line pulled before the IRS + * had a chance to update the L1 entry and mark it valid. + */ + if (irs_data->flags & IRS_FLAGS_NON_COHERENT) { + /* + * gicv5_irs_ist_synchronise() includes memory + * barriers (MMIO accessors) required to guarantee that the + * following dcache invalidation is not executed before the + * IST mapping operation has completed. + */ + dcache_inval_poc((unsigned long)(l1ist + index), + (unsigned long)(l1ist + index) + sizeof(*l1ist)); + } + + return 0; +} + +/* + * Try to match the L2 IST size to the pagesize, and if this is not possible + * pick the smallest supported L2 size in order to minimise the requirement for + * physically contiguous blocks of memory as page-sized allocations are + * guaranteed to be physically contiguous, and are by definition the easiest to + * find. + * + * Fall back to the smallest supported size (in the event that the pagesize + * itself is not supported) again serves to make it easier to find physically + * contiguous blocks of memory. + */ +static unsigned int gicv5_irs_l2_sz(u32 idr2) +{ + switch (PAGE_SIZE) { + case SZ_64K: + if (GICV5_IRS_IST_L2SZ_SUPPORT_64KB(idr2)) + return GICV5_IRS_IST_CFGR_L2SZ_64K; + fallthrough; + case SZ_4K: + if (GICV5_IRS_IST_L2SZ_SUPPORT_4KB(idr2)) + return GICV5_IRS_IST_CFGR_L2SZ_4K; + fallthrough; + case SZ_16K: + if (GICV5_IRS_IST_L2SZ_SUPPORT_16KB(idr2)) + return GICV5_IRS_IST_CFGR_L2SZ_16K; + break; + } + + if (GICV5_IRS_IST_L2SZ_SUPPORT_4KB(idr2)) + return GICV5_IRS_IST_CFGR_L2SZ_4K; + + return GICV5_IRS_IST_CFGR_L2SZ_64K; +} + +static int __init gicv5_irs_init_ist(struct gicv5_irs_chip_data *irs_data) +{ + u32 lpi_id_bits, idr2_id_bits, idr2_min_lpi_id_bits, l2_iste_sz, l2sz; + u32 l2_iste_sz_split, idr2; + bool two_levels, istmd; + u64 baser; + int ret; + + baser = irs_readq_relaxed(irs_data, GICV5_IRS_IST_BASER); + if (FIELD_GET(GICV5_IRS_IST_BASER_VALID, baser)) { + pr_err("IST is marked as valid already; cannot allocate\n"); + return -EPERM; + } + + idr2 = irs_readl_relaxed(irs_data, GICV5_IRS_IDR2); + two_levels = !!FIELD_GET(GICV5_IRS_IDR2_IST_LEVELS, idr2); + + idr2_id_bits = FIELD_GET(GICV5_IRS_IDR2_ID_BITS, idr2); + idr2_min_lpi_id_bits = FIELD_GET(GICV5_IRS_IDR2_MIN_LPI_ID_BITS, idr2); + + /* + * For two level tables we are always supporting the maximum allowed + * number of IDs. + * + * For 1-level tables, we should support a number of bits that + * is >= min_lpi_id_bits but cap it to LPI_ID_BITS_LINEAR lest + * the level 1-table gets too large and its memory allocation + * may fail. + */ + if (two_levels) { + lpi_id_bits = idr2_id_bits; + } else { + lpi_id_bits = max(LPI_ID_BITS_LINEAR, idr2_min_lpi_id_bits); + lpi_id_bits = min(lpi_id_bits, idr2_id_bits); + } + + /* + * Cap the ID bits according to the CPUIF supported ID bits + */ + lpi_id_bits = min(lpi_id_bits, gicv5_global_data.cpuif_id_bits); + + if (two_levels) + l2sz = gicv5_irs_l2_sz(idr2); + + istmd = !!FIELD_GET(GICV5_IRS_IDR2_ISTMD, idr2); + + l2_iste_sz = GICV5_IRS_IST_CFGR_ISTSZ_4; + + if (istmd) { + l2_iste_sz_split = FIELD_GET(GICV5_IRS_IDR2_ISTMD_SZ, idr2); + + if (lpi_id_bits < l2_iste_sz_split) + l2_iste_sz = GICV5_IRS_IST_CFGR_ISTSZ_8; + else + l2_iste_sz = GICV5_IRS_IST_CFGR_ISTSZ_16; + } + + /* + * Follow GICv5 specification recommendation to opt in for two + * level tables (ref: 10.2.1.14 IRS_IST_CFGR). + */ + if (two_levels && (lpi_id_bits > ((10 - l2_iste_sz) + (2 * l2sz)))) { + ret = gicv5_irs_init_ist_two_level(irs_data, lpi_id_bits, + l2_iste_sz, l2sz); + } else { + ret = gicv5_irs_init_ist_linear(irs_data, lpi_id_bits, + l2_iste_sz); + } + if (ret) + return ret; + + gicv5_init_lpis(BIT(lpi_id_bits)); + + return 0; +} + +struct iaffid_entry { + u16 iaffid; + bool valid; +}; + +static DEFINE_PER_CPU(struct iaffid_entry, cpu_iaffid); + +int gicv5_irs_cpu_to_iaffid(int cpuid, u16 *iaffid) +{ + if (!per_cpu(cpu_iaffid, cpuid).valid) { + pr_err("IAFFID for CPU %d has not been initialised\n", cpuid); + return -ENODEV; + } + + *iaffid = per_cpu(cpu_iaffid, cpuid).iaffid; + + return 0; +} + +struct gicv5_irs_chip_data *gicv5_irs_lookup_by_spi_id(u32 spi_id) +{ + struct gicv5_irs_chip_data *irs_data; + u32 min, max; + + list_for_each_entry(irs_data, &irs_nodes, entry) { + if (!irs_data->spi_range) + continue; + + min = irs_data->spi_min; + max = irs_data->spi_min + irs_data->spi_range - 1; + if (spi_id >= min && spi_id <= max) + return irs_data; + } + + return NULL; +} + +static int gicv5_irs_wait_for_spi_op(struct gicv5_irs_chip_data *irs_data) +{ + u32 statusr; + int ret; + + ret = gicv5_wait_for_op_atomic(irs_data->irs_base, GICV5_IRS_SPI_STATUSR, + GICV5_IRS_SPI_STATUSR_IDLE, &statusr); + if (ret) + return ret; + + return !!FIELD_GET(GICV5_IRS_SPI_STATUSR_V, statusr) ? 0 : -EIO; +} + +static int gicv5_irs_wait_for_irs_pe(struct gicv5_irs_chip_data *irs_data, + bool selr) +{ + bool valid = true; + u32 statusr; + int ret; + + ret = gicv5_wait_for_op_atomic(irs_data->irs_base, GICV5_IRS_PE_STATUSR, + GICV5_IRS_PE_STATUSR_IDLE, &statusr); + if (ret) + return ret; + + if (selr) + valid = !!FIELD_GET(GICV5_IRS_PE_STATUSR_V, statusr); + + return valid ? 0 : -EIO; +} + +static int gicv5_irs_wait_for_pe_selr(struct gicv5_irs_chip_data *irs_data) +{ + return gicv5_irs_wait_for_irs_pe(irs_data, true); +} + +static int gicv5_irs_wait_for_pe_cr0(struct gicv5_irs_chip_data *irs_data) +{ + return gicv5_irs_wait_for_irs_pe(irs_data, false); +} + +int gicv5_spi_irq_set_type(struct irq_data *d, unsigned int type) +{ + struct gicv5_irs_chip_data *irs_data = d->chip_data; + u32 selr, cfgr; + bool level; + int ret; + + /* + * There is no distinction between HIGH/LOW for level IRQs + * and RISING/FALLING for edge IRQs in the architecture, + * hence consider them equivalent. + */ + switch (type) { + case IRQ_TYPE_EDGE_RISING: + case IRQ_TYPE_EDGE_FALLING: + level = false; + break; + case IRQ_TYPE_LEVEL_HIGH: + case IRQ_TYPE_LEVEL_LOW: + level = true; + break; + default: + return -EINVAL; + } + + guard(raw_spinlock)(&irs_data->spi_config_lock); + + selr = FIELD_PREP(GICV5_IRS_SPI_SELR_ID, d->hwirq); + irs_writel_relaxed(irs_data, selr, GICV5_IRS_SPI_SELR); + ret = gicv5_irs_wait_for_spi_op(irs_data); + if (ret) + return ret; + + cfgr = FIELD_PREP(GICV5_IRS_SPI_CFGR_TM, level); + irs_writel_relaxed(irs_data, cfgr, GICV5_IRS_SPI_CFGR); + + return gicv5_irs_wait_for_spi_op(irs_data); +} + +static int gicv5_irs_wait_for_idle(struct gicv5_irs_chip_data *irs_data) +{ + return gicv5_wait_for_op_atomic(irs_data->irs_base, GICV5_IRS_CR0, + GICV5_IRS_CR0_IDLE, NULL); +} + +void gicv5_irs_syncr(void) +{ + struct gicv5_irs_chip_data *irs_data; + u32 syncr; + + irs_data = list_first_entry_or_null(&irs_nodes, struct gicv5_irs_chip_data, entry); + if (WARN_ON_ONCE(!irs_data)) + return; + + syncr = FIELD_PREP(GICV5_IRS_SYNCR_SYNC, 1); + irs_writel_relaxed(irs_data, syncr, GICV5_IRS_SYNCR); + + gicv5_wait_for_op(irs_data->irs_base, GICV5_IRS_SYNC_STATUSR, + GICV5_IRS_SYNC_STATUSR_IDLE); +} + +int gicv5_irs_register_cpu(int cpuid) +{ + struct gicv5_irs_chip_data *irs_data; + u32 selr, cr0; + u16 iaffid; + int ret; + + ret = gicv5_irs_cpu_to_iaffid(cpuid, &iaffid); + if (ret) { + pr_err("IAFFID for CPU %d has not been initialised\n", cpuid); + return ret; + } + + irs_data = per_cpu(per_cpu_irs_data, cpuid); + if (!irs_data) { + pr_err("No IRS associated with CPU %u\n", cpuid); + return -ENXIO; + } + + selr = FIELD_PREP(GICV5_IRS_PE_SELR_IAFFID, iaffid); + irs_writel_relaxed(irs_data, selr, GICV5_IRS_PE_SELR); + + ret = gicv5_irs_wait_for_pe_selr(irs_data); + if (ret) { + pr_err("IAFFID 0x%x used in IRS_PE_SELR is invalid\n", iaffid); + return -ENXIO; + } + + cr0 = FIELD_PREP(GICV5_IRS_PE_CR0_DPS, 0x1); + irs_writel_relaxed(irs_data, cr0, GICV5_IRS_PE_CR0); + + ret = gicv5_irs_wait_for_pe_cr0(irs_data); + if (ret) + return ret; + + pr_debug("CPU %d enabled PE IAFFID 0x%x\n", cpuid, iaffid); + + return 0; +} + +static void __init gicv5_irs_init_bases(struct gicv5_irs_chip_data *irs_data, + void __iomem *irs_base, + struct fwnode_handle *handle) +{ + struct device_node *np = to_of_node(handle); + u32 cr0, cr1; + + irs_data->fwnode = handle; + irs_data->irs_base = irs_base; + + if (of_property_read_bool(np, "dma-noncoherent")) { + /* + * A non-coherent IRS implies that some cache levels cannot be + * used coherently by the cores and GIC. Our only option is to mark + * memory attributes for the GIC as non-cacheable; by default, + * non-cacheable memory attributes imply outer-shareable + * shareability, the value written into IRS_CR1_SH is ignored. + */ + cr1 = FIELD_PREP(GICV5_IRS_CR1_VPED_WA, GICV5_NO_WRITE_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VPED_RA, GICV5_NO_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VMD_WA, GICV5_NO_WRITE_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VMD_RA, GICV5_NO_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VPET_RA, GICV5_NO_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VMT_RA, GICV5_NO_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_IST_WA, GICV5_NO_WRITE_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_IST_RA, GICV5_NO_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_IC, GICV5_NON_CACHE) | + FIELD_PREP(GICV5_IRS_CR1_OC, GICV5_NON_CACHE); + irs_data->flags |= IRS_FLAGS_NON_COHERENT; + } else { + cr1 = FIELD_PREP(GICV5_IRS_CR1_VPED_WA, GICV5_WRITE_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VPED_RA, GICV5_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VMD_WA, GICV5_WRITE_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VMD_RA, GICV5_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VPET_RA, GICV5_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_VMT_RA, GICV5_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_IST_WA, GICV5_WRITE_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_IST_RA, GICV5_READ_ALLOC) | + FIELD_PREP(GICV5_IRS_CR1_IC, GICV5_WB_CACHE) | + FIELD_PREP(GICV5_IRS_CR1_OC, GICV5_WB_CACHE) | + FIELD_PREP(GICV5_IRS_CR1_SH, GICV5_INNER_SHARE); + } + + irs_writel_relaxed(irs_data, cr1, GICV5_IRS_CR1); + + cr0 = FIELD_PREP(GICV5_IRS_CR0_IRSEN, 0x1); + irs_writel_relaxed(irs_data, cr0, GICV5_IRS_CR0); + gicv5_irs_wait_for_idle(irs_data); +} + +static int __init gicv5_irs_of_init_affinity(struct device_node *node, + struct gicv5_irs_chip_data *irs_data, + u8 iaffid_bits) +{ + /* + * Detect IAFFID<->CPU mappings from the device tree and + * record IRS<->CPU topology information. + */ + u16 iaffid_mask = GENMASK(iaffid_bits - 1, 0); + int ret, i, ncpus, niaffids; + + ncpus = of_count_phandle_with_args(node, "cpus", NULL); + if (ncpus < 0) + return -EINVAL; + + niaffids = of_property_count_elems_of_size(node, "arm,iaffids", + sizeof(u16)); + if (niaffids != ncpus) + return -EINVAL; + + u16 *iaffids __free(kfree) = kcalloc(niaffids, sizeof(*iaffids), GFP_KERNEL); + if (!iaffids) + return -ENOMEM; + + ret = of_property_read_u16_array(node, "arm,iaffids", iaffids, niaffids); + if (ret) + return ret; + + for (i = 0; i < ncpus; i++) { + struct device_node *cpu_node; + int cpu; + + cpu_node = of_parse_phandle(node, "cpus", i); + if (!cpu_node) { + pr_warn(FW_BUG "Erroneous CPU node phandle\n"); + continue; + } + + cpu = of_cpu_node_to_id(cpu_node); + of_node_put(cpu_node); + if (cpu < 0) + continue; + + if (iaffids[i] & ~iaffid_mask) { + pr_warn("CPU %d iaffid 0x%x exceeds IRS iaffid bits\n", + cpu, iaffids[i]); + continue; + } + + per_cpu(cpu_iaffid, cpu).iaffid = iaffids[i]; + per_cpu(cpu_iaffid, cpu).valid = true; + + /* We also know that the CPU is connected to this IRS */ + per_cpu(per_cpu_irs_data, cpu) = irs_data; + } + + return ret; +} + +static void irs_setup_pri_bits(u32 idr1) +{ + switch (FIELD_GET(GICV5_IRS_IDR1_PRIORITY_BITS, idr1)) { + case GICV5_IRS_IDR1_PRIORITY_BITS_1BITS: + gicv5_global_data.irs_pri_bits = 1; + break; + case GICV5_IRS_IDR1_PRIORITY_BITS_2BITS: + gicv5_global_data.irs_pri_bits = 2; + break; + case GICV5_IRS_IDR1_PRIORITY_BITS_3BITS: + gicv5_global_data.irs_pri_bits = 3; + break; + case GICV5_IRS_IDR1_PRIORITY_BITS_4BITS: + gicv5_global_data.irs_pri_bits = 4; + break; + case GICV5_IRS_IDR1_PRIORITY_BITS_5BITS: + gicv5_global_data.irs_pri_bits = 5; + break; + default: + pr_warn("Detected wrong IDR priority bits value 0x%lx\n", + FIELD_GET(GICV5_IRS_IDR1_PRIORITY_BITS, idr1)); + gicv5_global_data.irs_pri_bits = 1; + break; + } +} + +static int __init gicv5_irs_init(struct device_node *node) +{ + struct gicv5_irs_chip_data *irs_data; + void __iomem *irs_base; + u32 idr, spi_count; + u8 iaffid_bits; + int ret; + + irs_data = kzalloc(sizeof(*irs_data), GFP_KERNEL); + if (!irs_data) + return -ENOMEM; + + raw_spin_lock_init(&irs_data->spi_config_lock); + + ret = of_property_match_string(node, "reg-names", "ns-config"); + if (ret < 0) { + pr_err("%pOF: ns-config reg-name not present\n", node); + goto out_err; + } + + irs_base = of_io_request_and_map(node, ret, of_node_full_name(node)); + if (IS_ERR(irs_base)) { + pr_err("%pOF: unable to map GICv5 IRS registers\n", node); + ret = PTR_ERR(irs_base); + goto out_err; + } + + gicv5_irs_init_bases(irs_data, irs_base, &node->fwnode); + + idr = irs_readl_relaxed(irs_data, GICV5_IRS_IDR1); + iaffid_bits = FIELD_GET(GICV5_IRS_IDR1_IAFFID_BITS, idr) + 1; + + ret = gicv5_irs_of_init_affinity(node, irs_data, iaffid_bits); + if (ret) { + pr_err("Failed to parse CPU IAFFIDs from the device tree!\n"); + goto out_iomem; + } + + idr = irs_readl_relaxed(irs_data, GICV5_IRS_IDR2); + if (WARN(!FIELD_GET(GICV5_IRS_IDR2_LPI, idr), + "LPI support not available - no IPIs, can't proceed\n")) { + ret = -ENODEV; + goto out_iomem; + } + + idr = irs_readl_relaxed(irs_data, GICV5_IRS_IDR7); + irs_data->spi_min = FIELD_GET(GICV5_IRS_IDR7_SPI_BASE, idr); + + idr = irs_readl_relaxed(irs_data, GICV5_IRS_IDR6); + irs_data->spi_range = FIELD_GET(GICV5_IRS_IDR6_SPI_IRS_RANGE, idr); + + if (irs_data->spi_range) { + pr_info("%s detected SPI range [%u-%u]\n", + of_node_full_name(node), + irs_data->spi_min, + irs_data->spi_min + + irs_data->spi_range - 1); + } + + /* + * Do the global setting only on the first IRS. + * Global properties (iaffid_bits, global spi count) are guaranteed to + * be consistent across IRSes by the architecture. + */ + if (list_empty(&irs_nodes)) { + + idr = irs_readl_relaxed(irs_data, GICV5_IRS_IDR1); + irs_setup_pri_bits(idr); + + idr = irs_readl_relaxed(irs_data, GICV5_IRS_IDR5); + + spi_count = FIELD_GET(GICV5_IRS_IDR5_SPI_RANGE, idr); + gicv5_global_data.global_spi_count = spi_count; + + gicv5_init_lpi_domain(); + + pr_debug("Detected %u SPIs globally\n", spi_count); + } + + list_add_tail(&irs_data->entry, &irs_nodes); + + return 0; + +out_iomem: + iounmap(irs_base); +out_err: + kfree(irs_data); + return ret; +} + +void __init gicv5_irs_remove(void) +{ + struct gicv5_irs_chip_data *irs_data, *tmp_data; + + gicv5_free_lpi_domain(); + gicv5_deinit_lpis(); + + list_for_each_entry_safe(irs_data, tmp_data, &irs_nodes, entry) { + iounmap(irs_data->irs_base); + list_del(&irs_data->entry); + kfree(irs_data); + } +} + +int __init gicv5_irs_enable(void) +{ + struct gicv5_irs_chip_data *irs_data; + int ret; + + irs_data = list_first_entry_or_null(&irs_nodes, + struct gicv5_irs_chip_data, entry); + if (!irs_data) + return -ENODEV; + + ret = gicv5_irs_init_ist(irs_data); + if (ret) { + pr_err("Failed to init IST\n"); + return ret; + } + + return 0; +} + +void __init gicv5_irs_its_probe(void) +{ + struct gicv5_irs_chip_data *irs_data; + + list_for_each_entry(irs_data, &irs_nodes, entry) + gicv5_its_of_probe(to_of_node(irs_data->fwnode)); +} + +int __init gicv5_irs_of_probe(struct device_node *parent) +{ + struct device_node *np; + int ret; + + for_each_available_child_of_node(parent, np) { + if (!of_device_is_compatible(np, "arm,gic-v5-irs")) + continue; + + ret = gicv5_irs_init(np); + if (ret) + pr_err("Failed to init IRS %s\n", np->full_name); + } + + return list_empty(&irs_nodes) ? -ENODEV : 0; +} diff --git a/drivers/irqchip/irq-gic-v5-its.c b/drivers/irqchip/irq-gic-v5-its.c new file mode 100644 index 000000000000..554485f0be1f --- /dev/null +++ b/drivers/irqchip/irq-gic-v5-its.c @@ -0,0 +1,1233 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2024-2025 ARM Limited, All Rights Reserved. + */ + +#define pr_fmt(fmt) "GICv5 ITS: " fmt + +#include <linux/bitmap.h> +#include <linux/iommu.h> +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/msi.h> +#include <linux/of.h> +#include <linux/of_address.h> +#include <linux/of_irq.h> +#include <linux/slab.h> + +#include <linux/irqchip.h> +#include <linux/irqchip/arm-gic-v5.h> +#include <linux/irqchip/irq-msi-lib.h> + +#include "irq-gic-its-msi-parent.h" + +#define ITS_FLAGS_NON_COHERENT BIT(0) + +struct gicv5_its_chip_data { + struct xarray its_devices; + struct mutex dev_alloc_lock; + struct fwnode_handle *fwnode; + struct gicv5_its_devtab_cfg devtab_cfgr; + void __iomem *its_base; + u32 flags; + unsigned int msi_domain_flags; +}; + +struct gicv5_its_dev { + struct gicv5_its_chip_data *its_node; + struct gicv5_its_itt_cfg itt_cfg; + unsigned long *event_map; + u32 device_id; + u32 num_events; + phys_addr_t its_trans_phys_base; +}; + +static u32 its_readl_relaxed(struct gicv5_its_chip_data *its_node, const u32 reg_offset) +{ + return readl_relaxed(its_node->its_base + reg_offset); +} + +static void its_writel_relaxed(struct gicv5_its_chip_data *its_node, const u32 val, + const u32 reg_offset) +{ + writel_relaxed(val, its_node->its_base + reg_offset); +} + +static void its_writeq_relaxed(struct gicv5_its_chip_data *its_node, const u64 val, + const u32 reg_offset) +{ + writeq_relaxed(val, its_node->its_base + reg_offset); +} + +static void gicv5_its_dcache_clean(struct gicv5_its_chip_data *its, void *start, + size_t sz) +{ + void *end = start + sz; + + if (its->flags & ITS_FLAGS_NON_COHERENT) + dcache_clean_inval_poc((unsigned long)start, (unsigned long)end); + else + dsb(ishst); +} + +static void its_write_table_entry(struct gicv5_its_chip_data *its, __le64 *entry, + u64 val) +{ + WRITE_ONCE(*entry, cpu_to_le64(val)); + gicv5_its_dcache_clean(its, entry, sizeof(*entry)); +} + +#define devtab_cfgr_field(its, f) \ + FIELD_GET(GICV5_ITS_DT_CFGR_##f, (its)->devtab_cfgr.cfgr) + +static int gicv5_its_cache_sync(struct gicv5_its_chip_data *its) +{ + return gicv5_wait_for_op_atomic(its->its_base, GICV5_ITS_STATUSR, + GICV5_ITS_STATUSR_IDLE, NULL); +} + +static void gicv5_its_syncr(struct gicv5_its_chip_data *its, + struct gicv5_its_dev *its_dev) +{ + u64 syncr; + + syncr = FIELD_PREP(GICV5_ITS_SYNCR_SYNC, 1) | + FIELD_PREP(GICV5_ITS_SYNCR_DEVICEID, its_dev->device_id); + + its_writeq_relaxed(its, syncr, GICV5_ITS_SYNCR); + + gicv5_wait_for_op(its->its_base, GICV5_ITS_SYNC_STATUSR, GICV5_ITS_SYNC_STATUSR_IDLE); +} + +/* Number of bits required for each L2 {device/interrupt translation} table size */ +#define ITS_L2SZ_64K_L2_BITS 13 +#define ITS_L2SZ_16K_L2_BITS 11 +#define ITS_L2SZ_4K_L2_BITS 9 + +static unsigned int gicv5_its_l2sz_to_l2_bits(unsigned int sz) +{ + switch (sz) { + case GICV5_ITS_DT_ITT_CFGR_L2SZ_64k: + return ITS_L2SZ_64K_L2_BITS; + case GICV5_ITS_DT_ITT_CFGR_L2SZ_16k: + return ITS_L2SZ_16K_L2_BITS; + case GICV5_ITS_DT_ITT_CFGR_L2SZ_4k: + default: + return ITS_L2SZ_4K_L2_BITS; + } +} + +static int gicv5_its_itt_cache_inv(struct gicv5_its_chip_data *its, u32 device_id, + u16 event_id) +{ + u32 eventr, eidr; + u64 didr; + + didr = FIELD_PREP(GICV5_ITS_DIDR_DEVICEID, device_id); + eidr = FIELD_PREP(GICV5_ITS_EIDR_EVENTID, event_id); + eventr = FIELD_PREP(GICV5_ITS_INV_EVENTR_I, 0x1); + + its_writeq_relaxed(its, didr, GICV5_ITS_DIDR); + its_writel_relaxed(its, eidr, GICV5_ITS_EIDR); + its_writel_relaxed(its, eventr, GICV5_ITS_INV_EVENTR); + + return gicv5_its_cache_sync(its); +} + +static void gicv5_its_free_itt_linear(struct gicv5_its_dev *its_dev) +{ + kfree(its_dev->itt_cfg.linear.itt); +} + +static void gicv5_its_free_itt_two_level(struct gicv5_its_dev *its_dev) +{ + unsigned int i, num_ents = its_dev->itt_cfg.l2.num_l1_ents; + + for (i = 0; i < num_ents; i++) + kfree(its_dev->itt_cfg.l2.l2ptrs[i]); + + kfree(its_dev->itt_cfg.l2.l2ptrs); + kfree(its_dev->itt_cfg.l2.l1itt); +} + +static void gicv5_its_free_itt(struct gicv5_its_dev *its_dev) +{ + if (!its_dev->itt_cfg.l2itt) + gicv5_its_free_itt_linear(its_dev); + else + gicv5_its_free_itt_two_level(its_dev); +} + +static int gicv5_its_create_itt_linear(struct gicv5_its_chip_data *its, + struct gicv5_its_dev *its_dev, + unsigned int event_id_bits) +{ + unsigned int num_ents = BIT(event_id_bits); + __le64 *itt; + + itt = kcalloc(num_ents, sizeof(*itt), GFP_KERNEL); + if (!itt) + return -ENOMEM; + + its_dev->itt_cfg.linear.itt = itt; + its_dev->itt_cfg.linear.num_ents = num_ents; + its_dev->itt_cfg.l2itt = false; + its_dev->itt_cfg.event_id_bits = event_id_bits; + + gicv5_its_dcache_clean(its, itt, num_ents * sizeof(*itt)); + + return 0; +} + +/* + * Allocate a two-level ITT. All ITT entries are allocated in one go, unlike + * with the device table. Span may be used to limit the second level table + * size, where possible. + */ +static int gicv5_its_create_itt_two_level(struct gicv5_its_chip_data *its, + struct gicv5_its_dev *its_dev, + unsigned int event_id_bits, + unsigned int itt_l2sz, + unsigned int num_events) +{ + unsigned int l1_bits, l2_bits, span, events_per_l2_table; + unsigned int complete_tables, final_span, num_ents; + __le64 *itt_l1, *itt_l2, **l2ptrs; + int i, ret; + u64 val; + + ret = gicv5_its_l2sz_to_l2_bits(itt_l2sz); + if (ret >= event_id_bits) { + pr_debug("Incorrect l2sz (0x%x) for %u EventID bits. Cannot allocate ITT\n", + itt_l2sz, event_id_bits); + return -EINVAL; + } + + l2_bits = ret; + + l1_bits = event_id_bits - l2_bits; + + num_ents = BIT(l1_bits); + + itt_l1 = kcalloc(num_ents, sizeof(*itt_l1), GFP_KERNEL); + if (!itt_l1) + return -ENOMEM; + + l2ptrs = kcalloc(num_ents, sizeof(*l2ptrs), GFP_KERNEL); + if (!l2ptrs) { + kfree(itt_l1); + return -ENOMEM; + } + + its_dev->itt_cfg.l2.l2ptrs = l2ptrs; + + its_dev->itt_cfg.l2.l2sz = itt_l2sz; + its_dev->itt_cfg.l2.l1itt = itt_l1; + its_dev->itt_cfg.l2.num_l1_ents = num_ents; + its_dev->itt_cfg.l2itt = true; + its_dev->itt_cfg.event_id_bits = event_id_bits; + + /* + * Need to determine how many entries there are per L2 - this is based + * on the number of bits in the table. + */ + events_per_l2_table = BIT(l2_bits); + complete_tables = num_events / events_per_l2_table; + final_span = order_base_2(num_events % events_per_l2_table); + + for (i = 0; i < num_ents; i++) { + size_t l2sz; + + span = i == complete_tables ? final_span : l2_bits; + + itt_l2 = kcalloc(BIT(span), sizeof(*itt_l2), GFP_KERNEL); + if (!itt_l2) { + ret = -ENOMEM; + goto out_free; + } + + its_dev->itt_cfg.l2.l2ptrs[i] = itt_l2; + + l2sz = BIT(span) * sizeof(*itt_l2); + + gicv5_its_dcache_clean(its, itt_l2, l2sz); + + val = (virt_to_phys(itt_l2) & GICV5_ITTL1E_L2_ADDR_MASK) | + FIELD_PREP(GICV5_ITTL1E_SPAN, span) | + FIELD_PREP(GICV5_ITTL1E_VALID, 0x1); + + WRITE_ONCE(itt_l1[i], cpu_to_le64(val)); + } + + gicv5_its_dcache_clean(its, itt_l1, num_ents * sizeof(*itt_l1)); + + return 0; + +out_free: + for (i = i - 1; i >= 0; i--) + kfree(its_dev->itt_cfg.l2.l2ptrs[i]); + + kfree(its_dev->itt_cfg.l2.l2ptrs); + kfree(itt_l1); + return ret; +} + +/* + * Function to check whether the device table or ITT table support + * a two-level table and if so depending on the number of id_bits + * requested, determine whether a two-level table is required. + * + * Return the 2-level size value if a two level table is deemed + * necessary. + */ +static bool gicv5_its_l2sz_two_level(bool devtab, u32 its_idr1, u8 id_bits, u8 *sz) +{ + unsigned int l2_bits, l2_sz; + + if (devtab && !FIELD_GET(GICV5_ITS_IDR1_DT_LEVELS, its_idr1)) + return false; + + if (!devtab && !FIELD_GET(GICV5_ITS_IDR1_ITT_LEVELS, its_idr1)) + return false; + + /* + * Pick an L2 size that matches the pagesize; if a match + * is not found, go for the smallest supported l2 size granule. + * + * This ensures that we will always be able to allocate + * contiguous memory at L2. + */ + switch (PAGE_SIZE) { + case SZ_64K: + if (GICV5_ITS_IDR1_L2SZ_SUPPORT_64KB(its_idr1)) { + l2_sz = GICV5_ITS_DT_ITT_CFGR_L2SZ_64k; + break; + } + fallthrough; + case SZ_4K: + if (GICV5_ITS_IDR1_L2SZ_SUPPORT_4KB(its_idr1)) { + l2_sz = GICV5_ITS_DT_ITT_CFGR_L2SZ_4k; + break; + } + fallthrough; + case SZ_16K: + if (GICV5_ITS_IDR1_L2SZ_SUPPORT_16KB(its_idr1)) { + l2_sz = GICV5_ITS_DT_ITT_CFGR_L2SZ_16k; + break; + } + if (GICV5_ITS_IDR1_L2SZ_SUPPORT_4KB(its_idr1)) { + l2_sz = GICV5_ITS_DT_ITT_CFGR_L2SZ_4k; + break; + } + if (GICV5_ITS_IDR1_L2SZ_SUPPORT_64KB(its_idr1)) { + l2_sz = GICV5_ITS_DT_ITT_CFGR_L2SZ_64k; + break; + } + + l2_sz = GICV5_ITS_DT_ITT_CFGR_L2SZ_4k; + break; + } + + l2_bits = gicv5_its_l2sz_to_l2_bits(l2_sz); + + if (l2_bits > id_bits) + return false; + + *sz = l2_sz; + + return true; +} + +static __le64 *gicv5_its_device_get_itte_ref(struct gicv5_its_dev *its_dev, + u16 event_id) +{ + unsigned int l1_idx, l2_idx, l2_bits; + __le64 *l2_itt; + + if (!its_dev->itt_cfg.l2itt) { + __le64 *itt = its_dev->itt_cfg.linear.itt; + + return &itt[event_id]; + } + + l2_bits = gicv5_its_l2sz_to_l2_bits(its_dev->itt_cfg.l2.l2sz); + l1_idx = event_id >> l2_bits; + l2_idx = event_id & GENMASK(l2_bits - 1, 0); + l2_itt = its_dev->itt_cfg.l2.l2ptrs[l1_idx]; + + return &l2_itt[l2_idx]; +} + +static int gicv5_its_device_cache_inv(struct gicv5_its_chip_data *its, + struct gicv5_its_dev *its_dev) +{ + u32 devicer; + u64 didr; + + didr = FIELD_PREP(GICV5_ITS_DIDR_DEVICEID, its_dev->device_id); + devicer = FIELD_PREP(GICV5_ITS_INV_DEVICER_I, 0x1) | + FIELD_PREP(GICV5_ITS_INV_DEVICER_EVENTID_BITS, + its_dev->itt_cfg.event_id_bits) | + FIELD_PREP(GICV5_ITS_INV_DEVICER_L1, 0x0); + its_writeq_relaxed(its, didr, GICV5_ITS_DIDR); + its_writel_relaxed(its, devicer, GICV5_ITS_INV_DEVICER); + + return gicv5_its_cache_sync(its); +} + +/* + * Allocate a level 2 device table entry, update L1 parent to reference it. + * Only used for 2-level device tables, and it is called on demand. + */ +static int gicv5_its_alloc_l2_devtab(struct gicv5_its_chip_data *its, + unsigned int l1_index) +{ + __le64 *l2devtab, *l1devtab = its->devtab_cfgr.l2.l1devtab; + u8 span, l2sz, l2_bits; + u64 l1dte; + + if (FIELD_GET(GICV5_DTL1E_VALID, le64_to_cpu(l1devtab[l1_index]))) + return 0; + + span = FIELD_GET(GICV5_DTL1E_SPAN, le64_to_cpu(l1devtab[l1_index])); + l2sz = devtab_cfgr_field(its, L2SZ); + + l2_bits = gicv5_its_l2sz_to_l2_bits(l2sz); + + /* + * Span allows us to create a smaller L2 device table. + * If it is too large, use the number of allowed L2 bits. + */ + if (span > l2_bits) + span = l2_bits; + + l2devtab = kcalloc(BIT(span), sizeof(*l2devtab), GFP_KERNEL); + if (!l2devtab) + return -ENOMEM; + + its->devtab_cfgr.l2.l2ptrs[l1_index] = l2devtab; + + l1dte = FIELD_PREP(GICV5_DTL1E_SPAN, span) | + (virt_to_phys(l2devtab) & GICV5_DTL1E_L2_ADDR_MASK) | + FIELD_PREP(GICV5_DTL1E_VALID, 0x1); + its_write_table_entry(its, &l1devtab[l1_index], l1dte); + + return 0; +} + +static __le64 *gicv5_its_devtab_get_dte_ref(struct gicv5_its_chip_data *its, + u32 device_id, bool alloc) +{ + u8 str = devtab_cfgr_field(its, STRUCTURE); + unsigned int l2sz, l2_bits, l1_idx, l2_idx; + __le64 *l2devtab; + int ret; + + if (str == GICV5_ITS_DT_ITT_CFGR_STRUCTURE_LINEAR) { + l2devtab = its->devtab_cfgr.linear.devtab; + return &l2devtab[device_id]; + } + + l2sz = devtab_cfgr_field(its, L2SZ); + l2_bits = gicv5_its_l2sz_to_l2_bits(l2sz); + l1_idx = device_id >> l2_bits; + l2_idx = device_id & GENMASK(l2_bits - 1, 0); + + if (alloc) { + /* + * Allocate a new L2 device table here before + * continuing. We make the assumption that the span in + * the L1 table has been set correctly, and blindly use + * that value. + */ + ret = gicv5_its_alloc_l2_devtab(its, l1_idx); + if (ret) + return NULL; + } + + l2devtab = its->devtab_cfgr.l2.l2ptrs[l1_idx]; + return &l2devtab[l2_idx]; +} + +/* + * Register a new device in the device table. Allocate an ITT and + * program the L2DTE entry according to the ITT structure that + * was chosen. + */ +static int gicv5_its_device_register(struct gicv5_its_chip_data *its, + struct gicv5_its_dev *its_dev) +{ + u8 event_id_bits, device_id_bits, itt_struct, itt_l2sz; + phys_addr_t itt_phys_base; + bool two_level_itt; + u32 idr1, idr2; + __le64 *dte; + u64 val; + int ret; + + device_id_bits = devtab_cfgr_field(its, DEVICEID_BITS); + + if (its_dev->device_id >= BIT(device_id_bits)) { + pr_err("Supplied DeviceID (%u) outside of Device Table range (%u)!", + its_dev->device_id, (u32)GENMASK(device_id_bits - 1, 0)); + return -EINVAL; + } + + dte = gicv5_its_devtab_get_dte_ref(its, its_dev->device_id, true); + if (!dte) + return -ENOMEM; + + if (FIELD_GET(GICV5_DTL2E_VALID, le64_to_cpu(*dte))) + return -EBUSY; + + /* + * Determine how many bits we need, validate those against the max. + * Based on these, determine if we should go for a 1- or 2-level ITT. + */ + event_id_bits = order_base_2(its_dev->num_events); + + idr2 = its_readl_relaxed(its, GICV5_ITS_IDR2); + + if (event_id_bits > FIELD_GET(GICV5_ITS_IDR2_EVENTID_BITS, idr2)) { + pr_err("Required EventID bits (%u) larger than supported bits (%u)!", + event_id_bits, + (u8)FIELD_GET(GICV5_ITS_IDR2_EVENTID_BITS, idr2)); + return -EINVAL; + } + + idr1 = its_readl_relaxed(its, GICV5_ITS_IDR1); + + /* + * L2 ITT size is programmed into the L2DTE regardless of + * whether a two-level or linear ITT is built, init it. + */ + itt_l2sz = 0; + + two_level_itt = gicv5_its_l2sz_two_level(false, idr1, event_id_bits, + &itt_l2sz); + if (two_level_itt) + ret = gicv5_its_create_itt_two_level(its, its_dev, event_id_bits, + itt_l2sz, + its_dev->num_events); + else + ret = gicv5_its_create_itt_linear(its, its_dev, event_id_bits); + if (ret) + return ret; + + itt_phys_base = two_level_itt ? virt_to_phys(its_dev->itt_cfg.l2.l1itt) : + virt_to_phys(its_dev->itt_cfg.linear.itt); + + itt_struct = two_level_itt ? GICV5_ITS_DT_ITT_CFGR_STRUCTURE_TWO_LEVEL : + GICV5_ITS_DT_ITT_CFGR_STRUCTURE_LINEAR; + + val = FIELD_PREP(GICV5_DTL2E_EVENT_ID_BITS, event_id_bits) | + FIELD_PREP(GICV5_DTL2E_ITT_STRUCTURE, itt_struct) | + (itt_phys_base & GICV5_DTL2E_ITT_ADDR_MASK) | + FIELD_PREP(GICV5_DTL2E_ITT_L2SZ, itt_l2sz) | + FIELD_PREP(GICV5_DTL2E_VALID, 0x1); + + its_write_table_entry(its, dte, val); + + ret = gicv5_its_device_cache_inv(its, its_dev); + if (ret) { + its_write_table_entry(its, dte, 0); + gicv5_its_free_itt(its_dev); + return ret; + } + + return 0; +} + +/* + * Unregister a device in the device table. Lookup the device by ID, free the + * corresponding ITT, mark the device as invalid in the device table. + */ +static int gicv5_its_device_unregister(struct gicv5_its_chip_data *its, + struct gicv5_its_dev *its_dev) +{ + __le64 *dte; + + dte = gicv5_its_devtab_get_dte_ref(its, its_dev->device_id, false); + + if (!FIELD_GET(GICV5_DTL2E_VALID, le64_to_cpu(*dte))) { + pr_debug("Device table entry for DeviceID 0x%x is not valid. Nothing to clean up!", + its_dev->device_id); + return -EINVAL; + } + + /* Zero everything - make it clear that this is an invalid entry */ + its_write_table_entry(its, dte, 0); + + gicv5_its_free_itt(its_dev); + + return gicv5_its_device_cache_inv(its, its_dev); +} + +/* + * Allocate a 1-level device table. All entries are allocated, but marked + * invalid. + */ +static int gicv5_its_alloc_devtab_linear(struct gicv5_its_chip_data *its, + u8 device_id_bits) +{ + __le64 *devtab; + size_t sz; + u64 baser; + u32 cfgr; + + /* + * We expect a GICv5 implementation requiring a large number of + * deviceID bits to support a 2-level device table. If that's not + * the case, cap the number of deviceIDs supported according to the + * kmalloc limits so that the system can chug along with a linear + * device table. + */ + sz = BIT_ULL(device_id_bits) * sizeof(*devtab); + if (sz > KMALLOC_MAX_SIZE) { + u8 device_id_cap = ilog2(KMALLOC_MAX_SIZE/sizeof(*devtab)); + + pr_warn("Limiting device ID bits from %u to %u\n", + device_id_bits, device_id_cap); + device_id_bits = device_id_cap; + } + + devtab = kcalloc(BIT(device_id_bits), sizeof(*devtab), GFP_KERNEL); + if (!devtab) + return -ENOMEM; + + gicv5_its_dcache_clean(its, devtab, sz); + + cfgr = FIELD_PREP(GICV5_ITS_DT_CFGR_STRUCTURE, + GICV5_ITS_DT_ITT_CFGR_STRUCTURE_LINEAR) | + FIELD_PREP(GICV5_ITS_DT_CFGR_L2SZ, 0) | + FIELD_PREP(GICV5_ITS_DT_CFGR_DEVICEID_BITS, device_id_bits); + its_writel_relaxed(its, cfgr, GICV5_ITS_DT_CFGR); + + baser = virt_to_phys(devtab) & GICV5_ITS_DT_BASER_ADDR_MASK; + its_writeq_relaxed(its, baser, GICV5_ITS_DT_BASER); + + its->devtab_cfgr.cfgr = cfgr; + its->devtab_cfgr.linear.devtab = devtab; + + return 0; +} + +/* + * Allocate a 2-level device table. L2 entries are not allocated, + * they are allocated on-demand. + */ +static int gicv5_its_alloc_devtab_two_level(struct gicv5_its_chip_data *its, + u8 device_id_bits, + u8 devtab_l2sz) +{ + unsigned int l1_bits, l2_bits, i; + __le64 *l1devtab, **l2ptrs; + size_t l1_sz; + u64 baser; + u32 cfgr; + + l2_bits = gicv5_its_l2sz_to_l2_bits(devtab_l2sz); + + l1_bits = device_id_bits - l2_bits; + l1_sz = BIT(l1_bits) * sizeof(*l1devtab); + /* + * With 2-level device table support it is highly unlikely + * that we are not able to allocate the required amount of + * device table memory to cover deviceID space; cap the + * deviceID space if we encounter such set-up. + * If this ever becomes a problem we could revisit the policy + * behind level 2 size selection to reduce level-1 deviceID bits. + */ + if (l1_sz > KMALLOC_MAX_SIZE) { + l1_bits = ilog2(KMALLOC_MAX_SIZE/sizeof(*l1devtab)); + + pr_warn("Limiting device ID bits from %u to %u\n", + device_id_bits, l1_bits + l2_bits); + device_id_bits = l1_bits + l2_bits; + l1_sz = KMALLOC_MAX_SIZE; + } + + l1devtab = kcalloc(BIT(l1_bits), sizeof(*l1devtab), GFP_KERNEL); + if (!l1devtab) + return -ENOMEM; + + l2ptrs = kcalloc(BIT(l1_bits), sizeof(*l2ptrs), GFP_KERNEL); + if (!l2ptrs) { + kfree(l1devtab); + return -ENOMEM; + } + + for (i = 0; i < BIT(l1_bits); i++) + l1devtab[i] = cpu_to_le64(FIELD_PREP(GICV5_DTL1E_SPAN, l2_bits)); + + gicv5_its_dcache_clean(its, l1devtab, l1_sz); + + cfgr = FIELD_PREP(GICV5_ITS_DT_CFGR_STRUCTURE, + GICV5_ITS_DT_ITT_CFGR_STRUCTURE_TWO_LEVEL) | + FIELD_PREP(GICV5_ITS_DT_CFGR_L2SZ, devtab_l2sz) | + FIELD_PREP(GICV5_ITS_DT_CFGR_DEVICEID_BITS, device_id_bits); + its_writel_relaxed(its, cfgr, GICV5_ITS_DT_CFGR); + + baser = virt_to_phys(l1devtab) & GICV5_ITS_DT_BASER_ADDR_MASK; + its_writeq_relaxed(its, baser, GICV5_ITS_DT_BASER); + + its->devtab_cfgr.cfgr = cfgr; + its->devtab_cfgr.l2.l1devtab = l1devtab; + its->devtab_cfgr.l2.l2ptrs = l2ptrs; + + return 0; +} + +/* + * Initialise the device table as either 1- or 2-level depending on what is + * supported by the hardware. + */ +static int gicv5_its_init_devtab(struct gicv5_its_chip_data *its) +{ + u8 device_id_bits, devtab_l2sz; + bool two_level_devtab; + u32 idr1; + + idr1 = its_readl_relaxed(its, GICV5_ITS_IDR1); + + device_id_bits = FIELD_GET(GICV5_ITS_IDR1_DEVICEID_BITS, idr1); + two_level_devtab = gicv5_its_l2sz_two_level(true, idr1, device_id_bits, + &devtab_l2sz); + if (two_level_devtab) + return gicv5_its_alloc_devtab_two_level(its, device_id_bits, + devtab_l2sz); + else + return gicv5_its_alloc_devtab_linear(its, device_id_bits); +} + +static void gicv5_its_deinit_devtab(struct gicv5_its_chip_data *its) +{ + u8 str = devtab_cfgr_field(its, STRUCTURE); + + if (str == GICV5_ITS_DT_ITT_CFGR_STRUCTURE_LINEAR) { + kfree(its->devtab_cfgr.linear.devtab); + } else { + kfree(its->devtab_cfgr.l2.l1devtab); + kfree(its->devtab_cfgr.l2.l2ptrs); + } +} + +static void gicv5_its_compose_msi_msg(struct irq_data *d, struct msi_msg *msg) +{ + struct gicv5_its_dev *its_dev = irq_data_get_irq_chip_data(d); + u64 addr = its_dev->its_trans_phys_base; + + msg->data = FIELD_GET(GICV5_ITS_HWIRQ_EVENT_ID, d->hwirq); + msi_msg_set_addr(irq_data_get_msi_desc(d), msg, addr); +} + +static const struct irq_chip gicv5_its_irq_chip = { + .name = "GICv5-ITS-MSI", + .irq_mask = irq_chip_mask_parent, + .irq_unmask = irq_chip_unmask_parent, + .irq_eoi = irq_chip_eoi_parent, + .irq_set_affinity = irq_chip_set_affinity_parent, + .irq_get_irqchip_state = irq_chip_get_parent_state, + .irq_set_irqchip_state = irq_chip_set_parent_state, + .irq_compose_msi_msg = gicv5_its_compose_msi_msg, +}; + +static struct gicv5_its_dev *gicv5_its_find_device(struct gicv5_its_chip_data *its, + u32 device_id) +{ + struct gicv5_its_dev *dev = xa_load(&its->its_devices, device_id); + + return dev ? dev : ERR_PTR(-ENODEV); +} + +static struct gicv5_its_dev *gicv5_its_alloc_device(struct gicv5_its_chip_data *its, int nvec, + u32 dev_id) +{ + struct gicv5_its_dev *its_dev; + void *entry; + int ret; + + its_dev = gicv5_its_find_device(its, dev_id); + if (!IS_ERR(its_dev)) { + pr_err("A device with this DeviceID (0x%x) has already been registered.\n", + dev_id); + + return ERR_PTR(-EBUSY); + } + + its_dev = kzalloc(sizeof(*its_dev), GFP_KERNEL); + if (!its_dev) + return ERR_PTR(-ENOMEM); + + its_dev->device_id = dev_id; + its_dev->num_events = nvec; + + ret = gicv5_its_device_register(its, its_dev); + if (ret) { + pr_err("Failed to register the device\n"); + goto out_dev_free; + } + + its_dev->its_node = its; + + its_dev->event_map = (unsigned long *)bitmap_zalloc(its_dev->num_events, GFP_KERNEL); + if (!its_dev->event_map) { + ret = -ENOMEM; + goto out_unregister; + } + + entry = xa_store(&its->its_devices, dev_id, its_dev, GFP_KERNEL); + if (xa_is_err(entry)) { + ret = xa_err(entry); + goto out_bitmap_free; + } + + return its_dev; + +out_bitmap_free: + bitmap_free(its_dev->event_map); +out_unregister: + gicv5_its_device_unregister(its, its_dev); +out_dev_free: + kfree(its_dev); + return ERR_PTR(ret); +} + +static int gicv5_its_msi_prepare(struct irq_domain *domain, struct device *dev, + int nvec, msi_alloc_info_t *info) +{ + u32 dev_id = info->scratchpad[0].ul; + struct msi_domain_info *msi_info; + struct gicv5_its_chip_data *its; + struct gicv5_its_dev *its_dev; + + msi_info = msi_get_domain_info(domain); + its = msi_info->data; + + guard(mutex)(&its->dev_alloc_lock); + + its_dev = gicv5_its_alloc_device(its, nvec, dev_id); + if (IS_ERR(its_dev)) + return PTR_ERR(its_dev); + + its_dev->its_trans_phys_base = info->scratchpad[1].ul; + info->scratchpad[0].ptr = its_dev; + + return 0; +} + +static void gicv5_its_msi_teardown(struct irq_domain *domain, msi_alloc_info_t *info) +{ + struct gicv5_its_dev *its_dev = info->scratchpad[0].ptr; + struct msi_domain_info *msi_info; + struct gicv5_its_chip_data *its; + + msi_info = msi_get_domain_info(domain); + its = msi_info->data; + + guard(mutex)(&its->dev_alloc_lock); + + if (WARN_ON_ONCE(!bitmap_empty(its_dev->event_map, its_dev->num_events))) + return; + + xa_erase(&its->its_devices, its_dev->device_id); + bitmap_free(its_dev->event_map); + gicv5_its_device_unregister(its, its_dev); + kfree(its_dev); +} + +static struct msi_domain_ops gicv5_its_msi_domain_ops = { + .msi_prepare = gicv5_its_msi_prepare, + .msi_teardown = gicv5_its_msi_teardown, +}; + +static int gicv5_its_map_event(struct gicv5_its_dev *its_dev, u16 event_id, u32 lpi) +{ + struct gicv5_its_chip_data *its = its_dev->its_node; + u64 itt_entry; + __le64 *itte; + + itte = gicv5_its_device_get_itte_ref(its_dev, event_id); + + if (FIELD_GET(GICV5_ITTL2E_VALID, *itte)) + return -EEXIST; + + itt_entry = FIELD_PREP(GICV5_ITTL2E_LPI_ID, lpi) | + FIELD_PREP(GICV5_ITTL2E_VALID, 0x1); + + its_write_table_entry(its, itte, itt_entry); + + gicv5_its_itt_cache_inv(its, its_dev->device_id, event_id); + + return 0; +} + +static void gicv5_its_unmap_event(struct gicv5_its_dev *its_dev, u16 event_id) +{ + struct gicv5_its_chip_data *its = its_dev->its_node; + u64 itte_val; + __le64 *itte; + + itte = gicv5_its_device_get_itte_ref(its_dev, event_id); + + itte_val = le64_to_cpu(*itte); + itte_val &= ~GICV5_ITTL2E_VALID; + + its_write_table_entry(its, itte, itte_val); + + gicv5_its_itt_cache_inv(its, its_dev->device_id, event_id); +} + +static int gicv5_its_alloc_eventid(struct gicv5_its_dev *its_dev, msi_alloc_info_t *info, + unsigned int nr_irqs, u32 *eventid) +{ + int event_id_base; + + if (!(info->flags & MSI_ALLOC_FLAGS_FIXED_MSG_DATA)) { + event_id_base = bitmap_find_free_region(its_dev->event_map, + its_dev->num_events, + get_count_order(nr_irqs)); + if (event_id_base < 0) + return event_id_base; + } else { + /* + * We want to have a fixed EventID mapped for hardcoded + * message data allocations. + */ + if (WARN_ON_ONCE(nr_irqs != 1)) + return -EINVAL; + + event_id_base = info->hwirq; + + if (event_id_base >= its_dev->num_events) { + pr_err("EventID ouside of ITT range; cannot allocate an ITT entry!\n"); + + return -EINVAL; + } + + if (test_and_set_bit(event_id_base, its_dev->event_map)) { + pr_warn("Can't reserve event_id bitmap\n"); + return -EINVAL; + + } + } + + *eventid = event_id_base; + + return 0; +} + +static void gicv5_its_free_eventid(struct gicv5_its_dev *its_dev, u32 event_id_base, + unsigned int nr_irqs) +{ + bitmap_release_region(its_dev->event_map, event_id_base, + get_count_order(nr_irqs)); +} + +static int gicv5_its_irq_domain_alloc(struct irq_domain *domain, unsigned int virq, + unsigned int nr_irqs, void *arg) +{ + u32 device_id, event_id_base, lpi; + struct gicv5_its_dev *its_dev; + msi_alloc_info_t *info = arg; + irq_hw_number_t hwirq; + struct irq_data *irqd; + int ret, i; + + its_dev = info->scratchpad[0].ptr; + + ret = gicv5_its_alloc_eventid(its_dev, info, nr_irqs, &event_id_base); + if (ret) + return ret; + + ret = iommu_dma_prepare_msi(info->desc, its_dev->its_trans_phys_base); + if (ret) + goto out_eventid; + + device_id = its_dev->device_id; + + for (i = 0; i < nr_irqs; i++) { + ret = gicv5_alloc_lpi(); + if (ret < 0) { + pr_debug("Failed to find free LPI!\n"); + goto out_free_irqs; + } + lpi = ret; + + ret = irq_domain_alloc_irqs_parent(domain, virq + i, 1, &lpi); + if (ret) { + gicv5_free_lpi(lpi); + goto out_free_irqs; + } + + /* + * Store eventid and deviceid into the hwirq for later use. + * + * hwirq = event_id << 32 | device_id + */ + hwirq = FIELD_PREP(GICV5_ITS_HWIRQ_DEVICE_ID, device_id) | + FIELD_PREP(GICV5_ITS_HWIRQ_EVENT_ID, (u64)event_id_base + i); + irq_domain_set_info(domain, virq + i, hwirq, + &gicv5_its_irq_chip, its_dev, + handle_fasteoi_irq, NULL, NULL); + + irqd = irq_get_irq_data(virq + i); + irqd_set_single_target(irqd); + irqd_set_affinity_on_activate(irqd); + } + + return 0; + +out_free_irqs: + while (--i >= 0) { + irqd = irq_domain_get_irq_data(domain, virq + i); + gicv5_free_lpi(irqd->parent_data->hwirq); + irq_domain_reset_irq_data(irqd); + irq_domain_free_irqs_parent(domain, virq + i, 1); + } +out_eventid: + gicv5_its_free_eventid(its_dev, event_id_base, nr_irqs); + return ret; +} + +static void gicv5_its_irq_domain_free(struct irq_domain *domain, unsigned int virq, + unsigned int nr_irqs) +{ + struct irq_data *d = irq_domain_get_irq_data(domain, virq); + struct gicv5_its_chip_data *its; + struct gicv5_its_dev *its_dev; + u16 event_id_base; + unsigned int i; + + its_dev = irq_data_get_irq_chip_data(d); + its = its_dev->its_node; + + event_id_base = FIELD_GET(GICV5_ITS_HWIRQ_EVENT_ID, d->hwirq); + + bitmap_release_region(its_dev->event_map, event_id_base, + get_count_order(nr_irqs)); + + /* Hierarchically free irq data */ + for (i = 0; i < nr_irqs; i++) { + d = irq_domain_get_irq_data(domain, virq + i); + + gicv5_free_lpi(d->parent_data->hwirq); + irq_domain_reset_irq_data(d); + irq_domain_free_irqs_parent(domain, virq + i, 1); + } + + gicv5_its_syncr(its, its_dev); + gicv5_irs_syncr(); +} + +static int gicv5_its_irq_domain_activate(struct irq_domain *domain, struct irq_data *d, + bool reserve) +{ + struct gicv5_its_dev *its_dev = irq_data_get_irq_chip_data(d); + u16 event_id; + u32 lpi; + + event_id = FIELD_GET(GICV5_ITS_HWIRQ_EVENT_ID, d->hwirq); + lpi = d->parent_data->hwirq; + + return gicv5_its_map_event(its_dev, event_id, lpi); +} + +static void gicv5_its_irq_domain_deactivate(struct irq_domain *domain, + struct irq_data *d) +{ + struct gicv5_its_dev *its_dev = irq_data_get_irq_chip_data(d); + u16 event_id; + + event_id = FIELD_GET(GICV5_ITS_HWIRQ_EVENT_ID, d->hwirq); + + gicv5_its_unmap_event(its_dev, event_id); +} + +static const struct irq_domain_ops gicv5_its_irq_domain_ops = { + .alloc = gicv5_its_irq_domain_alloc, + .free = gicv5_its_irq_domain_free, + .activate = gicv5_its_irq_domain_activate, + .deactivate = gicv5_its_irq_domain_deactivate, + .select = msi_lib_irq_domain_select, +}; + +static int gicv5_its_write_cr0(struct gicv5_its_chip_data *its, bool enable) +{ + u32 cr0 = FIELD_PREP(GICV5_ITS_CR0_ITSEN, enable); + + its_writel_relaxed(its, cr0, GICV5_ITS_CR0); + return gicv5_wait_for_op_atomic(its->its_base, GICV5_ITS_CR0, + GICV5_ITS_CR0_IDLE, NULL); +} + +static int gicv5_its_enable(struct gicv5_its_chip_data *its) +{ + return gicv5_its_write_cr0(its, true); +} + +static int gicv5_its_disable(struct gicv5_its_chip_data *its) +{ + return gicv5_its_write_cr0(its, false); +} + +static void gicv5_its_print_info(struct gicv5_its_chip_data *its_node) +{ + bool devtab_linear; + u8 device_id_bits; + u8 str; + + device_id_bits = devtab_cfgr_field(its_node, DEVICEID_BITS); + + str = devtab_cfgr_field(its_node, STRUCTURE); + devtab_linear = (str == GICV5_ITS_DT_ITT_CFGR_STRUCTURE_LINEAR); + + pr_info("ITS %s enabled using %s device table device_id_bits %u\n", + fwnode_get_name(its_node->fwnode), + devtab_linear ? "linear" : "2-level", + device_id_bits); +} + +static int gicv5_its_init_domain(struct gicv5_its_chip_data *its, struct irq_domain *parent) +{ + struct irq_domain_info dom_info = { + .fwnode = its->fwnode, + .ops = &gicv5_its_irq_domain_ops, + .domain_flags = its->msi_domain_flags, + .parent = parent, + }; + struct msi_domain_info *info; + + info = kzalloc(sizeof(*info), GFP_KERNEL); + if (!info) + return -ENOMEM; + + info->ops = &gicv5_its_msi_domain_ops; + info->data = its; + dom_info.host_data = info; + + if (!msi_create_parent_irq_domain(&dom_info, &gic_v5_its_msi_parent_ops)) { + kfree(info); + return -ENOMEM; + } + + return 0; +} + +static int __init gicv5_its_init_bases(void __iomem *its_base, struct fwnode_handle *handle, + struct irq_domain *parent_domain) +{ + struct device_node *np = to_of_node(handle); + struct gicv5_its_chip_data *its_node; + u32 cr0, cr1; + bool enabled; + int ret; + + its_node = kzalloc(sizeof(*its_node), GFP_KERNEL); + if (!its_node) + return -ENOMEM; + + mutex_init(&its_node->dev_alloc_lock); + xa_init(&its_node->its_devices); + its_node->fwnode = handle; + its_node->its_base = its_base; + its_node->msi_domain_flags = IRQ_DOMAIN_FLAG_ISOLATED_MSI | + IRQ_DOMAIN_FLAG_FWNODE_PARENT; + + cr0 = its_readl_relaxed(its_node, GICV5_ITS_CR0); + enabled = FIELD_GET(GICV5_ITS_CR0_ITSEN, cr0); + if (WARN(enabled, "ITS %s enabled, disabling it before proceeding\n", np->full_name)) { + ret = gicv5_its_disable(its_node); + if (ret) + goto out_free_node; + } + + if (of_property_read_bool(np, "dma-noncoherent")) { + /* + * A non-coherent ITS implies that some cache levels cannot be + * used coherently by the cores and GIC. Our only option is to mark + * memory attributes for the GIC as non-cacheable; by default, + * non-cacheable memory attributes imply outer-shareable + * shareability, the value written into ITS_CR1_SH is ignored. + */ + cr1 = FIELD_PREP(GICV5_ITS_CR1_ITT_RA, GICV5_NO_READ_ALLOC) | + FIELD_PREP(GICV5_ITS_CR1_DT_RA, GICV5_NO_READ_ALLOC) | + FIELD_PREP(GICV5_ITS_CR1_IC, GICV5_NON_CACHE) | + FIELD_PREP(GICV5_ITS_CR1_OC, GICV5_NON_CACHE); + its_node->flags |= ITS_FLAGS_NON_COHERENT; + } else { + cr1 = FIELD_PREP(GICV5_ITS_CR1_ITT_RA, GICV5_READ_ALLOC) | + FIELD_PREP(GICV5_ITS_CR1_DT_RA, GICV5_READ_ALLOC) | + FIELD_PREP(GICV5_ITS_CR1_IC, GICV5_WB_CACHE) | + FIELD_PREP(GICV5_ITS_CR1_OC, GICV5_WB_CACHE) | + FIELD_PREP(GICV5_ITS_CR1_SH, GICV5_INNER_SHARE); + } + + its_writel_relaxed(its_node, cr1, GICV5_ITS_CR1); + + ret = gicv5_its_init_devtab(its_node); + if (ret) + goto out_free_node; + + ret = gicv5_its_enable(its_node); + if (ret) + goto out_free_devtab; + + ret = gicv5_its_init_domain(its_node, parent_domain); + if (ret) + goto out_disable_its; + + gicv5_its_print_info(its_node); + + return 0; + +out_disable_its: + gicv5_its_disable(its_node); +out_free_devtab: + gicv5_its_deinit_devtab(its_node); +out_free_node: + kfree(its_node); + return ret; +} + +static int __init gicv5_its_init(struct device_node *node) +{ + void __iomem *its_base; + int ret, idx; + + idx = of_property_match_string(node, "reg-names", "ns-config"); + if (idx < 0) { + pr_err("%pOF: ns-config reg-name not present\n", node); + return -ENODEV; + } + + its_base = of_io_request_and_map(node, idx, of_node_full_name(node)); + if (IS_ERR(its_base)) { + pr_err("%pOF: unable to map GICv5 ITS_CONFIG_FRAME\n", node); + return PTR_ERR(its_base); + } + + ret = gicv5_its_init_bases(its_base, of_fwnode_handle(node), + gicv5_global_data.lpi_domain); + if (ret) + goto out_unmap; + + return 0; + +out_unmap: + iounmap(its_base); + return ret; +} + +void __init gicv5_its_of_probe(struct device_node *parent) +{ + struct device_node *np; + + for_each_available_child_of_node(parent, np) { + if (!of_device_is_compatible(np, "arm,gic-v5-its")) + continue; + + if (gicv5_its_init(np)) + pr_err("Failed to init ITS %s\n", np->full_name); + } +} diff --git a/drivers/irqchip/irq-gic-v5-iwb.c b/drivers/irqchip/irq-gic-v5-iwb.c new file mode 100644 index 000000000000..ad9fdc14d1c6 --- /dev/null +++ b/drivers/irqchip/irq-gic-v5-iwb.c @@ -0,0 +1,277 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2024-2025 ARM Limited, All Rights Reserved. + */ +#define pr_fmt(fmt) "GICv5 IWB: " fmt + +#include <linux/init.h> +#include <linux/kernel.h> +#include <linux/msi.h> +#include <linux/of.h> +#include <linux/of_address.h> +#include <linux/of_platform.h> + +#include <linux/irqchip.h> +#include <linux/irqchip/arm-gic-v5.h> + +struct gicv5_iwb_chip_data { + void __iomem *iwb_base; + u16 nr_regs; +}; + +static u32 iwb_readl_relaxed(struct gicv5_iwb_chip_data *iwb_node, const u32 reg_offset) +{ + return readl_relaxed(iwb_node->iwb_base + reg_offset); +} + +static void iwb_writel_relaxed(struct gicv5_iwb_chip_data *iwb_node, const u32 val, + const u32 reg_offset) +{ + writel_relaxed(val, iwb_node->iwb_base + reg_offset); +} + +static int gicv5_iwb_wait_for_wenabler(struct gicv5_iwb_chip_data *iwb_node) +{ + return gicv5_wait_for_op_atomic(iwb_node->iwb_base, GICV5_IWB_WENABLE_STATUSR, + GICV5_IWB_WENABLE_STATUSR_IDLE, NULL); +} + +static int __gicv5_iwb_set_wire_enable(struct gicv5_iwb_chip_data *iwb_node, + u32 iwb_wire, bool enable) +{ + u32 n = iwb_wire / 32; + u8 i = iwb_wire % 32; + u32 val; + + if (n >= iwb_node->nr_regs) { + pr_err("IWB_WENABLER<n> is invalid for n=%u\n", n); + return -EINVAL; + } + + /* + * Enable IWB wire/pin at this point + * Note: This is not the same as enabling the interrupt + */ + val = iwb_readl_relaxed(iwb_node, GICV5_IWB_WENABLER + (4 * n)); + if (enable) + val |= BIT(i); + else + val &= ~BIT(i); + iwb_writel_relaxed(iwb_node, val, GICV5_IWB_WENABLER + (4 * n)); + + return gicv5_iwb_wait_for_wenabler(iwb_node); +} + +static int gicv5_iwb_enable_wire(struct gicv5_iwb_chip_data *iwb_node, + u32 iwb_wire) +{ + return __gicv5_iwb_set_wire_enable(iwb_node, iwb_wire, true); +} + +static int gicv5_iwb_disable_wire(struct gicv5_iwb_chip_data *iwb_node, + u32 iwb_wire) +{ + return __gicv5_iwb_set_wire_enable(iwb_node, iwb_wire, false); +} + +static void gicv5_iwb_irq_disable(struct irq_data *d) +{ + struct gicv5_iwb_chip_data *iwb_node = irq_data_get_irq_chip_data(d); + + gicv5_iwb_disable_wire(iwb_node, d->hwirq); + irq_chip_disable_parent(d); +} + +static void gicv5_iwb_irq_enable(struct irq_data *d) +{ + struct gicv5_iwb_chip_data *iwb_node = irq_data_get_irq_chip_data(d); + + gicv5_iwb_enable_wire(iwb_node, d->hwirq); + irq_chip_enable_parent(d); +} + +static int gicv5_iwb_set_type(struct irq_data *d, unsigned int type) +{ + struct gicv5_iwb_chip_data *iwb_node = irq_data_get_irq_chip_data(d); + u32 iwb_wire, n, wtmr; + u8 i; + + iwb_wire = d->hwirq; + i = iwb_wire % 32; + n = iwb_wire / 32; + + if (n >= iwb_node->nr_regs) { + pr_err_once("reg %u out of range\n", n); + return -EINVAL; + } + + wtmr = iwb_readl_relaxed(iwb_node, GICV5_IWB_WTMR + (4 * n)); + + switch (type) { + case IRQ_TYPE_LEVEL_HIGH: + case IRQ_TYPE_LEVEL_LOW: + wtmr |= BIT(i); + break; + case IRQ_TYPE_EDGE_RISING: + case IRQ_TYPE_EDGE_FALLING: + wtmr &= ~BIT(i); + break; + default: + pr_debug("unexpected wire trigger mode"); + return -EINVAL; + } + + iwb_writel_relaxed(iwb_node, wtmr, GICV5_IWB_WTMR + (4 * n)); + + return 0; +} + +static void gicv5_iwb_domain_set_desc(msi_alloc_info_t *alloc_info, struct msi_desc *desc) +{ + alloc_info->desc = desc; + alloc_info->hwirq = (u32)desc->data.icookie.value; +} + +static int gicv5_iwb_irq_domain_translate(struct irq_domain *d, struct irq_fwspec *fwspec, + irq_hw_number_t *hwirq, + unsigned int *type) +{ + if (!is_of_node(fwspec->fwnode)) + return -EINVAL; + + if (fwspec->param_count < 2) + return -EINVAL; + + /* + * param[0] is be the wire + * param[1] is the interrupt type + */ + *hwirq = fwspec->param[0]; + *type = fwspec->param[1] & IRQ_TYPE_SENSE_MASK; + + return 0; +} + +static void gicv5_iwb_write_msi_msg(struct irq_data *d, struct msi_msg *msg) {} + +static const struct msi_domain_template iwb_msi_template = { + .chip = { + .name = "GICv5-IWB", + .irq_mask = irq_chip_mask_parent, + .irq_unmask = irq_chip_unmask_parent, + .irq_enable = gicv5_iwb_irq_enable, + .irq_disable = gicv5_iwb_irq_disable, + .irq_eoi = irq_chip_eoi_parent, + .irq_set_type = gicv5_iwb_set_type, + .irq_write_msi_msg = gicv5_iwb_write_msi_msg, + .irq_set_affinity = irq_chip_set_affinity_parent, + .irq_get_irqchip_state = irq_chip_get_parent_state, + .irq_set_irqchip_state = irq_chip_set_parent_state, + .flags = IRQCHIP_SET_TYPE_MASKED | + IRQCHIP_SKIP_SET_WAKE | + IRQCHIP_MASK_ON_SUSPEND, + }, + + .ops = { + .set_desc = gicv5_iwb_domain_set_desc, + .msi_translate = gicv5_iwb_irq_domain_translate, + }, + + .info = { + .bus_token = DOMAIN_BUS_WIRED_TO_MSI, + .flags = MSI_FLAG_USE_DEV_FWNODE, + }, + + .alloc_info = { + .flags = MSI_ALLOC_FLAGS_FIXED_MSG_DATA, + }, +}; + +static bool gicv5_iwb_create_device_domain(struct device *dev, unsigned int size, + struct gicv5_iwb_chip_data *iwb_node) +{ + if (WARN_ON_ONCE(!dev->msi.domain)) + return false; + + return msi_create_device_irq_domain(dev, MSI_DEFAULT_DOMAIN, + &iwb_msi_template, size, + NULL, iwb_node); +} + +static struct gicv5_iwb_chip_data * +gicv5_iwb_init_bases(void __iomem *iwb_base, struct platform_device *pdev) +{ + u32 nr_wires, idr0, cr0; + unsigned int n; + int ret; + + struct gicv5_iwb_chip_data *iwb_node __free(kfree) = kzalloc(sizeof(*iwb_node), + GFP_KERNEL); + if (!iwb_node) + return ERR_PTR(-ENOMEM); + + iwb_node->iwb_base = iwb_base; + + idr0 = iwb_readl_relaxed(iwb_node, GICV5_IWB_IDR0); + nr_wires = (FIELD_GET(GICV5_IWB_IDR0_IW_RANGE, idr0) + 1) * 32; + + cr0 = iwb_readl_relaxed(iwb_node, GICV5_IWB_CR0); + if (!FIELD_GET(GICV5_IWB_CR0_IWBEN, cr0)) { + dev_err(&pdev->dev, "IWB must be enabled in firmware\n"); + return ERR_PTR(-EINVAL); + } + + iwb_node->nr_regs = FIELD_GET(GICV5_IWB_IDR0_IW_RANGE, idr0) + 1; + + for (n = 0; n < iwb_node->nr_regs; n++) + iwb_writel_relaxed(iwb_node, 0, GICV5_IWB_WENABLER + (sizeof(u32) * n)); + + ret = gicv5_iwb_wait_for_wenabler(iwb_node); + if (ret) + return ERR_PTR(ret); + + if (!gicv5_iwb_create_device_domain(&pdev->dev, nr_wires, iwb_node)) + return ERR_PTR(-ENOMEM); + + return_ptr(iwb_node); +} + +static int gicv5_iwb_device_probe(struct platform_device *pdev) +{ + struct gicv5_iwb_chip_data *iwb_node; + void __iomem *iwb_base; + struct resource *res; + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!res) + return -EINVAL; + + iwb_base = devm_ioremap(&pdev->dev, res->start, resource_size(res)); + if (!iwb_base) { + dev_err(&pdev->dev, "failed to ioremap %pR\n", res); + return -ENOMEM; + } + + iwb_node = gicv5_iwb_init_bases(iwb_base, pdev); + if (IS_ERR(iwb_node)) + return PTR_ERR(iwb_node); + + return 0; +} + +static const struct of_device_id gicv5_iwb_of_match[] = { + { .compatible = "arm,gic-v5-iwb" }, + { /* END */ } +}; +MODULE_DEVICE_TABLE(of, gicv5_iwb_of_match); + +static struct platform_driver gicv5_iwb_platform_driver = { + .driver = { + .name = "GICv5 IWB", + .of_match_table = gicv5_iwb_of_match, + .suppress_bind_attrs = true, + }, + .probe = gicv5_iwb_device_probe, +}; + +module_platform_driver(gicv5_iwb_platform_driver); diff --git a/drivers/irqchip/irq-gic-v5.c b/drivers/irqchip/irq-gic-v5.c new file mode 100644 index 000000000000..4bd224f359a7 --- /dev/null +++ b/drivers/irqchip/irq-gic-v5.c @@ -0,0 +1,1137 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2024-2025 ARM Limited, All Rights Reserved. + */ + +#define pr_fmt(fmt) "GICv5: " fmt + +#include <linux/cpuhotplug.h> +#include <linux/idr.h> +#include <linux/irqdomain.h> +#include <linux/slab.h> +#include <linux/wordpart.h> + +#include <linux/irqchip.h> +#include <linux/irqchip/arm-gic-v5.h> +#include <linux/irqchip/arm-vgic-info.h> + +#include <asm/cpufeature.h> +#include <asm/exception.h> + +static u8 pri_bits __ro_after_init = 5; + +#define GICV5_IRQ_PRI_MASK 0x1f +#define GICV5_IRQ_PRI_MI (GICV5_IRQ_PRI_MASK & GENMASK(4, 5 - pri_bits)) + +#define PPI_NR 128 + +static bool gicv5_cpuif_has_gcie(void) +{ + return this_cpu_has_cap(ARM64_HAS_GICV5_CPUIF); +} + +struct gicv5_chip_data gicv5_global_data __read_mostly; + +static DEFINE_IDA(lpi_ida); +static u32 num_lpis __ro_after_init; + +void __init gicv5_init_lpis(u32 lpis) +{ + num_lpis = lpis; +} + +void __init gicv5_deinit_lpis(void) +{ + num_lpis = 0; +} + +static int alloc_lpi(void) +{ + if (!num_lpis) + return -ENOSPC; + + return ida_alloc_max(&lpi_ida, num_lpis - 1, GFP_KERNEL); +} + +static void release_lpi(u32 lpi) +{ + ida_free(&lpi_ida, lpi); +} + +int gicv5_alloc_lpi(void) +{ + return alloc_lpi(); +} + +void gicv5_free_lpi(u32 lpi) +{ + release_lpi(lpi); +} + +static void gicv5_ppi_priority_init(void) +{ + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR0_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR1_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR2_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR3_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR4_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR5_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR6_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR7_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR8_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR9_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR10_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR11_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR12_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR13_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR14_EL1); + write_sysreg_s(REPEAT_BYTE(GICV5_IRQ_PRI_MI), SYS_ICC_PPI_PRIORITYR15_EL1); + + /* + * Context syncronization required to make sure system register writes + * effects are synchronised. + */ + isb(); +} + +static void gicv5_hwirq_init(irq_hw_number_t hwirq, u8 priority, u8 hwirq_type) +{ + u64 cdpri, cdaff; + u16 iaffid; + int ret; + + if (hwirq_type == GICV5_HWIRQ_TYPE_LPI || hwirq_type == GICV5_HWIRQ_TYPE_SPI) { + cdpri = FIELD_PREP(GICV5_GIC_CDPRI_PRIORITY_MASK, priority) | + FIELD_PREP(GICV5_GIC_CDPRI_TYPE_MASK, hwirq_type) | + FIELD_PREP(GICV5_GIC_CDPRI_ID_MASK, hwirq); + gic_insn(cdpri, CDPRI); + + ret = gicv5_irs_cpu_to_iaffid(smp_processor_id(), &iaffid); + + if (WARN_ON_ONCE(ret)) + return; + + cdaff = FIELD_PREP(GICV5_GIC_CDAFF_IAFFID_MASK, iaffid) | + FIELD_PREP(GICV5_GIC_CDAFF_TYPE_MASK, hwirq_type) | + FIELD_PREP(GICV5_GIC_CDAFF_ID_MASK, hwirq); + gic_insn(cdaff, CDAFF); + } +} + +static void gicv5_ppi_irq_mask(struct irq_data *d) +{ + u64 hwirq_id_bit = BIT_ULL(d->hwirq % 64); + + if (d->hwirq < 64) + sysreg_clear_set_s(SYS_ICC_PPI_ENABLER0_EL1, hwirq_id_bit, 0); + else + sysreg_clear_set_s(SYS_ICC_PPI_ENABLER1_EL1, hwirq_id_bit, 0); + + /* + * We must ensure that the disable takes effect immediately to + * guarantee that the lazy-disabled IRQ mechanism works. + * A context synchronization event is required to guarantee it. + * Reference: I_ZLTKB/R_YRGMH GICv5 specification - section 2.9.1. + */ + isb(); +} + +static void gicv5_iri_irq_mask(struct irq_data *d, u8 hwirq_type) +{ + u64 cddis; + + cddis = FIELD_PREP(GICV5_GIC_CDDIS_ID_MASK, d->hwirq) | + FIELD_PREP(GICV5_GIC_CDDIS_TYPE_MASK, hwirq_type); + + gic_insn(cddis, CDDIS); + /* + * We must make sure that GIC CDDIS write effects are propagated + * immediately to make sure the disable takes effect to guarantee + * that the lazy-disabled IRQ mechanism works. + * Rule R_XCLJC states that the effects of a GIC system instruction + * complete in finite time. + * The GSB ensures completion of the GIC instruction and prevents + * loads, stores and GIC instructions from executing part of their + * functionality before the GSB SYS. + */ + gsb_sys(); +} + +static void gicv5_spi_irq_mask(struct irq_data *d) +{ + gicv5_iri_irq_mask(d, GICV5_HWIRQ_TYPE_SPI); +} + +static void gicv5_lpi_irq_mask(struct irq_data *d) +{ + gicv5_iri_irq_mask(d, GICV5_HWIRQ_TYPE_LPI); +} + +static void gicv5_ppi_irq_unmask(struct irq_data *d) +{ + u64 hwirq_id_bit = BIT_ULL(d->hwirq % 64); + + if (d->hwirq < 64) + sysreg_clear_set_s(SYS_ICC_PPI_ENABLER0_EL1, 0, hwirq_id_bit); + else + sysreg_clear_set_s(SYS_ICC_PPI_ENABLER1_EL1, 0, hwirq_id_bit); + /* + * We must ensure that the enable takes effect in finite time - a + * context synchronization event is required to guarantee it, we + * can not take for granted that would happen (eg a core going straight + * into idle after enabling a PPI). + * Reference: I_ZLTKB/R_YRGMH GICv5 specification - section 2.9.1. + */ + isb(); +} + +static void gicv5_iri_irq_unmask(struct irq_data *d, u8 hwirq_type) +{ + u64 cden; + + cden = FIELD_PREP(GICV5_GIC_CDEN_ID_MASK, d->hwirq) | + FIELD_PREP(GICV5_GIC_CDEN_TYPE_MASK, hwirq_type); + /* + * Rule R_XCLJC states that the effects of a GIC system instruction + * complete in finite time and that's the only requirement when + * unmasking an SPI/LPI IRQ. + */ + gic_insn(cden, CDEN); +} + +static void gicv5_spi_irq_unmask(struct irq_data *d) +{ + gicv5_iri_irq_unmask(d, GICV5_HWIRQ_TYPE_SPI); +} + +static void gicv5_lpi_irq_unmask(struct irq_data *d) +{ + gicv5_iri_irq_unmask(d, GICV5_HWIRQ_TYPE_LPI); +} + +static void gicv5_hwirq_eoi(u32 hwirq_id, u8 hwirq_type) +{ + u64 cddi; + + cddi = FIELD_PREP(GICV5_GIC_CDDI_ID_MASK, hwirq_id) | + FIELD_PREP(GICV5_GIC_CDDI_TYPE_MASK, hwirq_type); + + gic_insn(cddi, CDDI); + + gic_insn(0, CDEOI); +} + +static void gicv5_ppi_irq_eoi(struct irq_data *d) +{ + /* Skip deactivate for forwarded PPI interrupts */ + if (irqd_is_forwarded_to_vcpu(d)) { + gic_insn(0, CDEOI); + return; + } + + gicv5_hwirq_eoi(d->hwirq, GICV5_HWIRQ_TYPE_PPI); +} + +static void gicv5_spi_irq_eoi(struct irq_data *d) +{ + gicv5_hwirq_eoi(d->hwirq, GICV5_HWIRQ_TYPE_SPI); +} + +static void gicv5_lpi_irq_eoi(struct irq_data *d) +{ + gicv5_hwirq_eoi(d->hwirq, GICV5_HWIRQ_TYPE_LPI); +} + +static int gicv5_iri_irq_set_affinity(struct irq_data *d, + const struct cpumask *mask_val, + bool force, u8 hwirq_type) +{ + int ret, cpuid; + u16 iaffid; + u64 cdaff; + + if (force) + cpuid = cpumask_first(mask_val); + else + cpuid = cpumask_any_and(mask_val, cpu_online_mask); + + ret = gicv5_irs_cpu_to_iaffid(cpuid, &iaffid); + if (ret) + return ret; + + cdaff = FIELD_PREP(GICV5_GIC_CDAFF_IAFFID_MASK, iaffid) | + FIELD_PREP(GICV5_GIC_CDAFF_TYPE_MASK, hwirq_type) | + FIELD_PREP(GICV5_GIC_CDAFF_ID_MASK, d->hwirq); + gic_insn(cdaff, CDAFF); + + irq_data_update_effective_affinity(d, cpumask_of(cpuid)); + + return IRQ_SET_MASK_OK_DONE; +} + +static int gicv5_spi_irq_set_affinity(struct irq_data *d, + const struct cpumask *mask_val, + bool force) +{ + return gicv5_iri_irq_set_affinity(d, mask_val, force, + GICV5_HWIRQ_TYPE_SPI); +} + +static int gicv5_lpi_irq_set_affinity(struct irq_data *d, + const struct cpumask *mask_val, + bool force) +{ + return gicv5_iri_irq_set_affinity(d, mask_val, force, + GICV5_HWIRQ_TYPE_LPI); +} + +enum ppi_reg { + PPI_PENDING, + PPI_ACTIVE, + PPI_HM +}; + +static __always_inline u64 read_ppi_sysreg_s(unsigned int irq, + const enum ppi_reg which) +{ + switch (which) { + case PPI_PENDING: + return irq < 64 ? read_sysreg_s(SYS_ICC_PPI_SPENDR0_EL1) : + read_sysreg_s(SYS_ICC_PPI_SPENDR1_EL1); + case PPI_ACTIVE: + return irq < 64 ? read_sysreg_s(SYS_ICC_PPI_SACTIVER0_EL1) : + read_sysreg_s(SYS_ICC_PPI_SACTIVER1_EL1); + case PPI_HM: + return irq < 64 ? read_sysreg_s(SYS_ICC_PPI_HMR0_EL1) : + read_sysreg_s(SYS_ICC_PPI_HMR1_EL1); + default: + BUILD_BUG_ON(1); + } +} + +static __always_inline void write_ppi_sysreg_s(unsigned int irq, bool set, + const enum ppi_reg which) +{ + u64 bit = BIT_ULL(irq % 64); + + switch (which) { + case PPI_PENDING: + if (set) { + if (irq < 64) + write_sysreg_s(bit, SYS_ICC_PPI_SPENDR0_EL1); + else + write_sysreg_s(bit, SYS_ICC_PPI_SPENDR1_EL1); + } else { + if (irq < 64) + write_sysreg_s(bit, SYS_ICC_PPI_CPENDR0_EL1); + else + write_sysreg_s(bit, SYS_ICC_PPI_CPENDR1_EL1); + } + return; + case PPI_ACTIVE: + if (set) { + if (irq < 64) + write_sysreg_s(bit, SYS_ICC_PPI_SACTIVER0_EL1); + else + write_sysreg_s(bit, SYS_ICC_PPI_SACTIVER1_EL1); + } else { + if (irq < 64) + write_sysreg_s(bit, SYS_ICC_PPI_CACTIVER0_EL1); + else + write_sysreg_s(bit, SYS_ICC_PPI_CACTIVER1_EL1); + } + return; + default: + BUILD_BUG_ON(1); + } +} + +static int gicv5_ppi_irq_get_irqchip_state(struct irq_data *d, + enum irqchip_irq_state which, + bool *state) +{ + u64 hwirq_id_bit = BIT_ULL(d->hwirq % 64); + + switch (which) { + case IRQCHIP_STATE_PENDING: + *state = !!(read_ppi_sysreg_s(d->hwirq, PPI_PENDING) & hwirq_id_bit); + return 0; + case IRQCHIP_STATE_ACTIVE: + *state = !!(read_ppi_sysreg_s(d->hwirq, PPI_ACTIVE) & hwirq_id_bit); + return 0; + default: + pr_debug("Unexpected PPI irqchip state\n"); + return -EINVAL; + } +} + +static int gicv5_iri_irq_get_irqchip_state(struct irq_data *d, + enum irqchip_irq_state which, + bool *state, u8 hwirq_type) +{ + u64 icsr, cdrcfg; + + cdrcfg = d->hwirq | FIELD_PREP(GICV5_GIC_CDRCFG_TYPE_MASK, hwirq_type); + + gic_insn(cdrcfg, CDRCFG); + isb(); + icsr = read_sysreg_s(SYS_ICC_ICSR_EL1); + + if (FIELD_GET(ICC_ICSR_EL1_F, icsr)) { + pr_err("ICSR_EL1 is invalid\n"); + return -EINVAL; + } + + switch (which) { + case IRQCHIP_STATE_PENDING: + *state = !!(FIELD_GET(ICC_ICSR_EL1_Pending, icsr)); + return 0; + + case IRQCHIP_STATE_ACTIVE: + *state = !!(FIELD_GET(ICC_ICSR_EL1_Active, icsr)); + return 0; + + default: + pr_debug("Unexpected irqchip_irq_state\n"); + return -EINVAL; + } +} + +static int gicv5_spi_irq_get_irqchip_state(struct irq_data *d, + enum irqchip_irq_state which, + bool *state) +{ + return gicv5_iri_irq_get_irqchip_state(d, which, state, + GICV5_HWIRQ_TYPE_SPI); +} + +static int gicv5_lpi_irq_get_irqchip_state(struct irq_data *d, + enum irqchip_irq_state which, + bool *state) +{ + return gicv5_iri_irq_get_irqchip_state(d, which, state, + GICV5_HWIRQ_TYPE_LPI); +} + +static int gicv5_ppi_irq_set_irqchip_state(struct irq_data *d, + enum irqchip_irq_state which, + bool state) +{ + switch (which) { + case IRQCHIP_STATE_PENDING: + write_ppi_sysreg_s(d->hwirq, state, PPI_PENDING); + return 0; + case IRQCHIP_STATE_ACTIVE: + write_ppi_sysreg_s(d->hwirq, state, PPI_ACTIVE); + return 0; + default: + pr_debug("Unexpected PPI irqchip state\n"); + return -EINVAL; + } +} + +static void gicv5_iri_irq_write_pending_state(struct irq_data *d, bool state, + u8 hwirq_type) +{ + u64 cdpend; + + cdpend = FIELD_PREP(GICV5_GIC_CDPEND_TYPE_MASK, hwirq_type) | + FIELD_PREP(GICV5_GIC_CDPEND_ID_MASK, d->hwirq) | + FIELD_PREP(GICV5_GIC_CDPEND_PENDING_MASK, state); + + gic_insn(cdpend, CDPEND); +} + +static void gicv5_spi_irq_write_pending_state(struct irq_data *d, bool state) +{ + gicv5_iri_irq_write_pending_state(d, state, GICV5_HWIRQ_TYPE_SPI); +} + +static void gicv5_lpi_irq_write_pending_state(struct irq_data *d, bool state) +{ + gicv5_iri_irq_write_pending_state(d, state, GICV5_HWIRQ_TYPE_LPI); +} + +static int gicv5_spi_irq_set_irqchip_state(struct irq_data *d, + enum irqchip_irq_state which, + bool state) +{ + switch (which) { + case IRQCHIP_STATE_PENDING: + gicv5_spi_irq_write_pending_state(d, state); + break; + default: + pr_debug("Unexpected irqchip_irq_state\n"); + return -EINVAL; + } + + return 0; +} + +static int gicv5_lpi_irq_set_irqchip_state(struct irq_data *d, + enum irqchip_irq_state which, + bool state) +{ + switch (which) { + case IRQCHIP_STATE_PENDING: + gicv5_lpi_irq_write_pending_state(d, state); + break; + + default: + pr_debug("Unexpected irqchip_irq_state\n"); + return -EINVAL; + } + + return 0; +} + +static int gicv5_spi_irq_retrigger(struct irq_data *data) +{ + return !gicv5_spi_irq_set_irqchip_state(data, IRQCHIP_STATE_PENDING, + true); +} + +static int gicv5_lpi_irq_retrigger(struct irq_data *data) +{ + return !gicv5_lpi_irq_set_irqchip_state(data, IRQCHIP_STATE_PENDING, + true); +} + +static void gicv5_ipi_send_single(struct irq_data *d, unsigned int cpu) +{ + /* Mark the LPI pending */ + irq_chip_retrigger_hierarchy(d); +} + +static bool gicv5_ppi_irq_is_level(irq_hw_number_t hwirq) +{ + u64 bit = BIT_ULL(hwirq % 64); + + return !!(read_ppi_sysreg_s(hwirq, PPI_HM) & bit); +} + +static int gicv5_ppi_irq_set_vcpu_affinity(struct irq_data *d, void *vcpu) +{ + if (vcpu) + irqd_set_forwarded_to_vcpu(d); + else + irqd_clr_forwarded_to_vcpu(d); + + return 0; +} + +static const struct irq_chip gicv5_ppi_irq_chip = { + .name = "GICv5-PPI", + .irq_mask = gicv5_ppi_irq_mask, + .irq_unmask = gicv5_ppi_irq_unmask, + .irq_eoi = gicv5_ppi_irq_eoi, + .irq_get_irqchip_state = gicv5_ppi_irq_get_irqchip_state, + .irq_set_irqchip_state = gicv5_ppi_irq_set_irqchip_state, + .irq_set_vcpu_affinity = gicv5_ppi_irq_set_vcpu_affinity, + .flags = IRQCHIP_SKIP_SET_WAKE | + IRQCHIP_MASK_ON_SUSPEND, +}; + +static const struct irq_chip gicv5_spi_irq_chip = { + .name = "GICv5-SPI", + .irq_mask = gicv5_spi_irq_mask, + .irq_unmask = gicv5_spi_irq_unmask, + .irq_eoi = gicv5_spi_irq_eoi, + .irq_set_type = gicv5_spi_irq_set_type, + .irq_set_affinity = gicv5_spi_irq_set_affinity, + .irq_retrigger = gicv5_spi_irq_retrigger, + .irq_get_irqchip_state = gicv5_spi_irq_get_irqchip_state, + .irq_set_irqchip_state = gicv5_spi_irq_set_irqchip_state, + .flags = IRQCHIP_SET_TYPE_MASKED | + IRQCHIP_SKIP_SET_WAKE | + IRQCHIP_MASK_ON_SUSPEND, +}; + +static const struct irq_chip gicv5_lpi_irq_chip = { + .name = "GICv5-LPI", + .irq_mask = gicv5_lpi_irq_mask, + .irq_unmask = gicv5_lpi_irq_unmask, + .irq_eoi = gicv5_lpi_irq_eoi, + .irq_set_affinity = gicv5_lpi_irq_set_affinity, + .irq_retrigger = gicv5_lpi_irq_retrigger, + .irq_get_irqchip_state = gicv5_lpi_irq_get_irqchip_state, + .irq_set_irqchip_state = gicv5_lpi_irq_set_irqchip_state, + .flags = IRQCHIP_SKIP_SET_WAKE | + IRQCHIP_MASK_ON_SUSPEND, +}; + +static const struct irq_chip gicv5_ipi_irq_chip = { + .name = "GICv5-IPI", + .irq_mask = irq_chip_mask_parent, + .irq_unmask = irq_chip_unmask_parent, + .irq_eoi = irq_chip_eoi_parent, + .irq_set_affinity = irq_chip_set_affinity_parent, + .irq_get_irqchip_state = irq_chip_get_parent_state, + .irq_set_irqchip_state = irq_chip_set_parent_state, + .ipi_send_single = gicv5_ipi_send_single, + .flags = IRQCHIP_SKIP_SET_WAKE | + IRQCHIP_MASK_ON_SUSPEND, +}; + +static __always_inline int gicv5_irq_domain_translate(struct irq_domain *d, + struct irq_fwspec *fwspec, + irq_hw_number_t *hwirq, + unsigned int *type, + const u8 hwirq_type) +{ + if (!is_of_node(fwspec->fwnode)) + return -EINVAL; + + if (fwspec->param_count < 3) + return -EINVAL; + + if (fwspec->param[0] != hwirq_type) + return -EINVAL; + + *hwirq = fwspec->param[1]; + + switch (hwirq_type) { + case GICV5_HWIRQ_TYPE_PPI: + /* + * Handling mode is hardcoded for PPIs, set the type using + * HW reported value. + */ + *type = gicv5_ppi_irq_is_level(*hwirq) ? IRQ_TYPE_LEVEL_LOW : + IRQ_TYPE_EDGE_RISING; + break; + case GICV5_HWIRQ_TYPE_SPI: + *type = fwspec->param[2] & IRQ_TYPE_SENSE_MASK; + break; + default: + BUILD_BUG_ON(1); + } + + return 0; +} + +static int gicv5_irq_ppi_domain_translate(struct irq_domain *d, + struct irq_fwspec *fwspec, + irq_hw_number_t *hwirq, + unsigned int *type) +{ + return gicv5_irq_domain_translate(d, fwspec, hwirq, type, + GICV5_HWIRQ_TYPE_PPI); +} + +static int gicv5_irq_ppi_domain_alloc(struct irq_domain *domain, unsigned int virq, + unsigned int nr_irqs, void *arg) +{ + unsigned int type = IRQ_TYPE_NONE; + struct irq_fwspec *fwspec = arg; + irq_hw_number_t hwirq; + int ret; + + if (WARN_ON_ONCE(nr_irqs != 1)) + return -EINVAL; + + ret = gicv5_irq_ppi_domain_translate(domain, fwspec, &hwirq, &type); + if (ret) + return ret; + + if (type & IRQ_TYPE_LEVEL_MASK) + irq_set_status_flags(virq, IRQ_LEVEL); + + irq_set_percpu_devid(virq); + irq_domain_set_info(domain, virq, hwirq, &gicv5_ppi_irq_chip, NULL, + handle_percpu_devid_irq, NULL, NULL); + + return 0; +} + +static void gicv5_irq_domain_free(struct irq_domain *domain, unsigned int virq, + unsigned int nr_irqs) +{ + struct irq_data *d; + + if (WARN_ON_ONCE(nr_irqs != 1)) + return; + + d = irq_domain_get_irq_data(domain, virq); + + irq_set_handler(virq, NULL); + irq_domain_reset_irq_data(d); +} + +static int gicv5_irq_ppi_domain_select(struct irq_domain *d, struct irq_fwspec *fwspec, + enum irq_domain_bus_token bus_token) +{ + if (fwspec->fwnode != d->fwnode) + return 0; + + if (fwspec->param[0] != GICV5_HWIRQ_TYPE_PPI) + return 0; + + return (d == gicv5_global_data.ppi_domain); +} + +static const struct irq_domain_ops gicv5_irq_ppi_domain_ops = { + .translate = gicv5_irq_ppi_domain_translate, + .alloc = gicv5_irq_ppi_domain_alloc, + .free = gicv5_irq_domain_free, + .select = gicv5_irq_ppi_domain_select +}; + +static int gicv5_irq_spi_domain_translate(struct irq_domain *d, + struct irq_fwspec *fwspec, + irq_hw_number_t *hwirq, + unsigned int *type) +{ + return gicv5_irq_domain_translate(d, fwspec, hwirq, type, + GICV5_HWIRQ_TYPE_SPI); +} + +static int gicv5_irq_spi_domain_alloc(struct irq_domain *domain, unsigned int virq, + unsigned int nr_irqs, void *arg) +{ + struct gicv5_irs_chip_data *chip_data; + unsigned int type = IRQ_TYPE_NONE; + struct irq_fwspec *fwspec = arg; + struct irq_data *irqd; + irq_hw_number_t hwirq; + int ret; + + if (WARN_ON_ONCE(nr_irqs != 1)) + return -EINVAL; + + ret = gicv5_irq_spi_domain_translate(domain, fwspec, &hwirq, &type); + if (ret) + return ret; + + irqd = irq_desc_get_irq_data(irq_to_desc(virq)); + chip_data = gicv5_irs_lookup_by_spi_id(hwirq); + + irq_domain_set_info(domain, virq, hwirq, &gicv5_spi_irq_chip, chip_data, + handle_fasteoi_irq, NULL, NULL); + irq_set_probe(virq); + irqd_set_single_target(irqd); + + gicv5_hwirq_init(hwirq, GICV5_IRQ_PRI_MI, GICV5_HWIRQ_TYPE_SPI); + + return 0; +} + +static int gicv5_irq_spi_domain_select(struct irq_domain *d, struct irq_fwspec *fwspec, + enum irq_domain_bus_token bus_token) +{ + if (fwspec->fwnode != d->fwnode) + return 0; + + if (fwspec->param[0] != GICV5_HWIRQ_TYPE_SPI) + return 0; + + return (d == gicv5_global_data.spi_domain); +} + +static const struct irq_domain_ops gicv5_irq_spi_domain_ops = { + .translate = gicv5_irq_spi_domain_translate, + .alloc = gicv5_irq_spi_domain_alloc, + .free = gicv5_irq_domain_free, + .select = gicv5_irq_spi_domain_select +}; + +static void gicv5_lpi_config_reset(struct irq_data *d) +{ + u64 cdhm; + + /* + * Reset LPIs handling mode to edge by default and clear pending + * state to make sure we start the LPI with a clean state from + * previous incarnations. + */ + cdhm = FIELD_PREP(GICV5_GIC_CDHM_HM_MASK, 0) | + FIELD_PREP(GICV5_GIC_CDHM_TYPE_MASK, GICV5_HWIRQ_TYPE_LPI) | + FIELD_PREP(GICV5_GIC_CDHM_ID_MASK, d->hwirq); + gic_insn(cdhm, CDHM); + + gicv5_lpi_irq_write_pending_state(d, false); +} + +static int gicv5_irq_lpi_domain_alloc(struct irq_domain *domain, unsigned int virq, + unsigned int nr_irqs, void *arg) +{ + irq_hw_number_t hwirq; + struct irq_data *irqd; + u32 *lpi = arg; + int ret; + + if (WARN_ON_ONCE(nr_irqs != 1)) + return -EINVAL; + + hwirq = *lpi; + + irqd = irq_domain_get_irq_data(domain, virq); + + irq_domain_set_info(domain, virq, hwirq, &gicv5_lpi_irq_chip, NULL, + handle_fasteoi_irq, NULL, NULL); + irqd_set_single_target(irqd); + + ret = gicv5_irs_iste_alloc(hwirq); + if (ret < 0) + return ret; + + gicv5_hwirq_init(hwirq, GICV5_IRQ_PRI_MI, GICV5_HWIRQ_TYPE_LPI); + gicv5_lpi_config_reset(irqd); + + return 0; +} + +static const struct irq_domain_ops gicv5_irq_lpi_domain_ops = { + .alloc = gicv5_irq_lpi_domain_alloc, + .free = gicv5_irq_domain_free, +}; + +void __init gicv5_init_lpi_domain(void) +{ + struct irq_domain *d; + + d = irq_domain_create_tree(NULL, &gicv5_irq_lpi_domain_ops, NULL); + gicv5_global_data.lpi_domain = d; +} + +void __init gicv5_free_lpi_domain(void) +{ + irq_domain_remove(gicv5_global_data.lpi_domain); + gicv5_global_data.lpi_domain = NULL; +} + +static int gicv5_irq_ipi_domain_alloc(struct irq_domain *domain, unsigned int virq, + unsigned int nr_irqs, void *arg) +{ + struct irq_data *irqd; + int ret, i; + u32 lpi; + + for (i = 0; i < nr_irqs; i++) { + ret = gicv5_alloc_lpi(); + if (ret < 0) + return ret; + + lpi = ret; + + ret = irq_domain_alloc_irqs_parent(domain, virq + i, 1, &lpi); + if (ret) { + gicv5_free_lpi(lpi); + return ret; + } + + irqd = irq_domain_get_irq_data(domain, virq + i); + + irq_domain_set_hwirq_and_chip(domain, virq + i, i, + &gicv5_ipi_irq_chip, NULL); + + irqd_set_single_target(irqd); + + irq_set_handler(virq + i, handle_percpu_irq); + } + + return 0; +} + +static void gicv5_irq_ipi_domain_free(struct irq_domain *domain, unsigned int virq, + unsigned int nr_irqs) +{ + struct irq_data *d; + unsigned int i; + + for (i = 0; i < nr_irqs; i++) { + d = irq_domain_get_irq_data(domain, virq + i); + + if (!d) + return; + + gicv5_free_lpi(d->parent_data->hwirq); + + irq_set_handler(virq + i, NULL); + irq_domain_reset_irq_data(d); + irq_domain_free_irqs_parent(domain, virq + i, 1); + } +} + +static const struct irq_domain_ops gicv5_irq_ipi_domain_ops = { + .alloc = gicv5_irq_ipi_domain_alloc, + .free = gicv5_irq_ipi_domain_free, +}; + +static void handle_irq_per_domain(u32 hwirq) +{ + u8 hwirq_type = FIELD_GET(GICV5_HWIRQ_TYPE, hwirq); + u32 hwirq_id = FIELD_GET(GICV5_HWIRQ_ID, hwirq); + struct irq_domain *domain; + + switch (hwirq_type) { + case GICV5_HWIRQ_TYPE_PPI: + domain = gicv5_global_data.ppi_domain; + break; + case GICV5_HWIRQ_TYPE_SPI: + domain = gicv5_global_data.spi_domain; + break; + case GICV5_HWIRQ_TYPE_LPI: + domain = gicv5_global_data.lpi_domain; + break; + default: + pr_err_once("Unknown IRQ type, bail out\n"); + return; + } + + if (generic_handle_domain_irq(domain, hwirq_id)) { + pr_err_once("Could not handle, hwirq = 0x%x", hwirq_id); + gicv5_hwirq_eoi(hwirq_id, hwirq_type); + } +} + +static void __exception_irq_entry gicv5_handle_irq(struct pt_regs *regs) +{ + bool valid; + u32 hwirq; + u64 ia; + + ia = gicr_insn(CDIA); + valid = GICV5_GICR_CDIA_VALID(ia); + + if (!valid) + return; + + /* + * Ensure that the CDIA instruction effects (ie IRQ activation) are + * completed before handling the interrupt. + */ + gsb_ack(); + + /* + * Ensure instruction ordering between an acknowledgment and subsequent + * instructions in the IRQ handler using an ISB. + */ + isb(); + + hwirq = FIELD_GET(GICV5_HWIRQ_INTID, ia); + + handle_irq_per_domain(hwirq); +} + +static void gicv5_cpu_disable_interrupts(void) +{ + u64 cr0; + + cr0 = FIELD_PREP(ICC_CR0_EL1_EN, 0); + write_sysreg_s(cr0, SYS_ICC_CR0_EL1); +} + +static void gicv5_cpu_enable_interrupts(void) +{ + u64 cr0, pcr; + + write_sysreg_s(0, SYS_ICC_PPI_ENABLER0_EL1); + write_sysreg_s(0, SYS_ICC_PPI_ENABLER1_EL1); + + gicv5_ppi_priority_init(); + + pcr = FIELD_PREP(ICC_PCR_EL1_PRIORITY, GICV5_IRQ_PRI_MI); + write_sysreg_s(pcr, SYS_ICC_PCR_EL1); + + cr0 = FIELD_PREP(ICC_CR0_EL1_EN, 1); + write_sysreg_s(cr0, SYS_ICC_CR0_EL1); +} + +static int base_ipi_virq; + +static int gicv5_starting_cpu(unsigned int cpu) +{ + if (WARN(!gicv5_cpuif_has_gcie(), + "GICv5 system components present but CPU does not have FEAT_GCIE")) + return -ENODEV; + + gicv5_cpu_enable_interrupts(); + + return gicv5_irs_register_cpu(cpu); +} + +static void __init gicv5_smp_init(void) +{ + unsigned int num_ipis = GICV5_IPIS_PER_CPU * nr_cpu_ids; + + cpuhp_setup_state_nocalls(CPUHP_AP_IRQ_GIC_STARTING, + "irqchip/arm/gicv5:starting", + gicv5_starting_cpu, NULL); + + base_ipi_virq = irq_domain_alloc_irqs(gicv5_global_data.ipi_domain, + num_ipis, NUMA_NO_NODE, NULL); + if (WARN(base_ipi_virq <= 0, "IPI IRQ allocation was not successful")) + return; + + set_smp_ipi_range_percpu(base_ipi_virq, GICV5_IPIS_PER_CPU, nr_cpu_ids); +} + +static void __init gicv5_free_domains(void) +{ + if (gicv5_global_data.ppi_domain) + irq_domain_remove(gicv5_global_data.ppi_domain); + if (gicv5_global_data.spi_domain) + irq_domain_remove(gicv5_global_data.spi_domain); + if (gicv5_global_data.ipi_domain) + irq_domain_remove(gicv5_global_data.ipi_domain); + + gicv5_global_data.ppi_domain = NULL; + gicv5_global_data.spi_domain = NULL; + gicv5_global_data.ipi_domain = NULL; +} + +static int __init gicv5_init_domains(struct fwnode_handle *handle) +{ + u32 spi_count = gicv5_global_data.global_spi_count; + struct irq_domain *d; + + d = irq_domain_create_linear(handle, PPI_NR, &gicv5_irq_ppi_domain_ops, NULL); + if (!d) + return -ENOMEM; + + irq_domain_update_bus_token(d, DOMAIN_BUS_WIRED); + gicv5_global_data.ppi_domain = d; + + if (spi_count) { + d = irq_domain_create_linear(handle, spi_count, + &gicv5_irq_spi_domain_ops, NULL); + + if (!d) { + gicv5_free_domains(); + return -ENOMEM; + } + + gicv5_global_data.spi_domain = d; + irq_domain_update_bus_token(d, DOMAIN_BUS_WIRED); + } + + if (!WARN(!gicv5_global_data.lpi_domain, + "LPI domain uninitialized, can't set up IPIs")) { + d = irq_domain_create_hierarchy(gicv5_global_data.lpi_domain, + 0, GICV5_IPIS_PER_CPU * nr_cpu_ids, + NULL, &gicv5_irq_ipi_domain_ops, + NULL); + + if (!d) { + gicv5_free_domains(); + return -ENOMEM; + } + gicv5_global_data.ipi_domain = d; + } + gicv5_global_data.fwnode = handle; + + return 0; +} + +static void gicv5_set_cpuif_pribits(void) +{ + u64 icc_idr0 = read_sysreg_s(SYS_ICC_IDR0_EL1); + + switch (FIELD_GET(ICC_IDR0_EL1_PRI_BITS, icc_idr0)) { + case ICC_IDR0_EL1_PRI_BITS_4BITS: + gicv5_global_data.cpuif_pri_bits = 4; + break; + case ICC_IDR0_EL1_PRI_BITS_5BITS: + gicv5_global_data.cpuif_pri_bits = 5; + break; + default: + pr_err("Unexpected ICC_IDR0_EL1_PRI_BITS value, default to 4"); + gicv5_global_data.cpuif_pri_bits = 4; + break; + } +} + +static void gicv5_set_cpuif_idbits(void) +{ + u32 icc_idr0 = read_sysreg_s(SYS_ICC_IDR0_EL1); + + switch (FIELD_GET(ICC_IDR0_EL1_ID_BITS, icc_idr0)) { + case ICC_IDR0_EL1_ID_BITS_16BITS: + gicv5_global_data.cpuif_id_bits = 16; + break; + case ICC_IDR0_EL1_ID_BITS_24BITS: + gicv5_global_data.cpuif_id_bits = 24; + break; + default: + pr_err("Unexpected ICC_IDR0_EL1_ID_BITS value, default to 16"); + gicv5_global_data.cpuif_id_bits = 16; + break; + } +} + +#ifdef CONFIG_KVM +static struct gic_kvm_info gic_v5_kvm_info __initdata; + +static bool __init gicv5_cpuif_has_gcie_legacy(void) +{ + u64 idr0 = read_sysreg_s(SYS_ICC_IDR0_EL1); + return !!FIELD_GET(ICC_IDR0_EL1_GCIE_LEGACY, idr0); +} + +static void __init gic_of_setup_kvm_info(struct device_node *node) +{ + gic_v5_kvm_info.type = GIC_V5; + gic_v5_kvm_info.has_gcie_v3_compat = gicv5_cpuif_has_gcie_legacy(); + + /* GIC Virtual CPU interface maintenance interrupt */ + gic_v5_kvm_info.no_maint_irq_mask = false; + gic_v5_kvm_info.maint_irq = irq_of_parse_and_map(node, 0); + if (!gic_v5_kvm_info.maint_irq) { + pr_warn("cannot find GICv5 virtual CPU interface maintenance interrupt\n"); + return; + } + + vgic_set_kvm_info(&gic_v5_kvm_info); +} +#else +static inline void __init gic_of_setup_kvm_info(struct device_node *node) +{ +} +#endif // CONFIG_KVM + +static int __init gicv5_of_init(struct device_node *node, struct device_node *parent) +{ + int ret = gicv5_irs_of_probe(node); + if (ret) + return ret; + + ret = gicv5_init_domains(of_fwnode_handle(node)); + if (ret) + goto out_irs; + + gicv5_set_cpuif_pribits(); + gicv5_set_cpuif_idbits(); + + pri_bits = min_not_zero(gicv5_global_data.cpuif_pri_bits, + gicv5_global_data.irs_pri_bits); + + ret = gicv5_starting_cpu(smp_processor_id()); + if (ret) + goto out_dom; + + ret = set_handle_irq(gicv5_handle_irq); + if (ret) + goto out_int; + + ret = gicv5_irs_enable(); + if (ret) + goto out_int; + + gicv5_smp_init(); + + gicv5_irs_its_probe(); + + gic_of_setup_kvm_info(node); + + return 0; + +out_int: + gicv5_cpu_disable_interrupts(); +out_dom: + gicv5_free_domains(); +out_irs: + gicv5_irs_remove(); + + return ret; +} +IRQCHIP_DECLARE(gic_v5, "arm,gic-v5", gicv5_of_init); diff --git a/drivers/irqchip/irq-gic.c b/drivers/irqchip/irq-gic.c index 6503573557fd..1269ab8eb726 100644 --- a/drivers/irqchip/irq-gic.c +++ b/drivers/irqchip/irq-gic.c @@ -54,7 +54,7 @@ static void gic_check_cpu_features(void) { - WARN_TAINT_ONCE(this_cpu_has_cap(ARM64_HAS_GIC_CPUIF_SYSREGS), + WARN_TAINT_ONCE(this_cpu_has_cap(ARM64_HAS_GICV3_CPUIF), TAINT_CPU_OUT_OF_SPEC, "GICv3 system registers enabled, broken firmware!\n"); } diff --git a/drivers/irqchip/irq-imgpdc.c b/drivers/irqchip/irq-imgpdc.c index f0410d5d7315..e9ef2f5a7207 100644 --- a/drivers/irqchip/irq-imgpdc.c +++ b/drivers/irqchip/irq-imgpdc.c @@ -372,8 +372,8 @@ static int pdc_intc_probe(struct platform_device *pdev) priv->syswake_irq = irq; /* Set up an IRQ domain */ - priv->domain = irq_domain_create_linear(of_fwnode_handle(node), 16, &irq_generic_chip_ops, - priv); + priv->domain = irq_domain_create_linear(dev_fwnode(&pdev->dev), 16, &irq_generic_chip_ops, + priv); if (unlikely(!priv->domain)) { dev_err(&pdev->dev, "cannot add IRQ domain\n"); return -ENOMEM; diff --git a/drivers/irqchip/irq-imx-irqsteer.c b/drivers/irqchip/irq-imx-irqsteer.c index 6dc9ac48fee5..4682ce5bf8d3 100644 --- a/drivers/irqchip/irq-imx-irqsteer.c +++ b/drivers/irqchip/irq-imx-irqsteer.c @@ -212,8 +212,8 @@ static int imx_irqsteer_probe(struct platform_device *pdev) /* steer all IRQs into configured channel */ writel_relaxed(BIT(data->channel), data->regs + CHANCTRL); - data->domain = irq_domain_create_linear(of_fwnode_handle(np), data->reg_num * 32, - &imx_irqsteer_domain_ops, data); + data->domain = irq_domain_create_linear(dev_fwnode(&pdev->dev), data->reg_num * 32, + &imx_irqsteer_domain_ops, data); if (!data->domain) { dev_err(&pdev->dev, "failed to create IRQ domain\n"); ret = -ENOMEM; diff --git a/drivers/irqchip/irq-imx-mu-msi.c b/drivers/irqchip/irq-imx-mu-msi.c index 137da1927d14..d2a4e8a61a42 100644 --- a/drivers/irqchip/irq-imx-mu-msi.c +++ b/drivers/irqchip/irq-imx-mu-msi.c @@ -223,21 +223,21 @@ static const struct msi_parent_ops imx_mu_msi_parent_ops = { static int imx_mu_msi_domains_init(struct imx_mu_msi *msi_data, struct device *dev) { - struct fwnode_handle *fwnodes = dev_fwnode(dev); + struct irq_domain_info info = { + .ops = &imx_mu_msi_domain_ops, + .fwnode = dev_fwnode(dev), + .size = IMX_MU_CHANS, + .host_data = msi_data, + }; struct irq_domain *parent; /* Initialize MSI domain parent */ - parent = irq_domain_create_linear(fwnodes, IMX_MU_CHANS, - &imx_mu_msi_domain_ops, msi_data); + parent = msi_create_parent_irq_domain(&info, &imx_mu_msi_parent_ops); if (!parent) { dev_err(dev, "failed to create IRQ domain\n"); return -ENOMEM; } - - irq_domain_update_bus_token(parent, DOMAIN_BUS_NEXUS); parent->dev = parent->pm_dev = dev; - parent->flags |= IRQ_DOMAIN_FLAG_MSI_PARENT; - parent->msi_parent_ops = &imx_mu_msi_parent_ops; return 0; } diff --git a/drivers/irqchip/irq-keystone.c b/drivers/irqchip/irq-keystone.c index c9e902b7bf48..922fff09354f 100644 --- a/drivers/irqchip/irq-keystone.c +++ b/drivers/irqchip/irq-keystone.c @@ -157,8 +157,8 @@ static int keystone_irq_probe(struct platform_device *pdev) kirq->chip.irq_mask = keystone_irq_setmask; kirq->chip.irq_unmask = keystone_irq_unmask; - kirq->irqd = irq_domain_create_linear(of_fwnode_handle(np), KEYSTONE_N_IRQ, - &keystone_irq_ops, kirq); + kirq->irqd = irq_domain_create_linear(dev_fwnode(dev), KEYSTONE_N_IRQ, &keystone_irq_ops, + kirq); if (!kirq->irqd) { dev_err(dev, "IRQ domain registration failed\n"); return -ENODEV; diff --git a/drivers/irqchip/irq-loongson-eiointc.c b/drivers/irqchip/irq-loongson-eiointc.c index b2860eb2d32c..39e5a72ccd3c 100644 --- a/drivers/irqchip/irq-loongson-eiointc.c +++ b/drivers/irqchip/irq-loongson-eiointc.c @@ -46,6 +46,7 @@ #define EIOINTC_ALL_ENABLE_VEC_MASK(vector) (EIOINTC_ALL_ENABLE & ~BIT(vector & 0x1f)) #define EIOINTC_REG_ENABLE_VEC(vector) (EIOINTC_REG_ENABLE + ((vector >> 5) << 2)) #define EIOINTC_USE_CPU_ENCODE BIT(0) +#define EIOINTC_ROUTE_MULT_IP BIT(1) #define MAX_EIO_NODES (NR_CPUS / CORES_PER_EIO_NODE) @@ -59,6 +60,14 @@ #define EIOINTC_REG_ROUTE_VEC_MASK(vector) (0xff << EIOINTC_REG_ROUTE_VEC_SHIFT(vector)) static int nr_pics; +struct eiointc_priv; + +struct eiointc_ip_route { + struct eiointc_priv *priv; + /* Offset Routed destination IP */ + int start; + int end; +}; struct eiointc_priv { u32 node; @@ -68,6 +77,8 @@ struct eiointc_priv { struct fwnode_handle *domain_handle; struct irq_domain *eiointc_domain; int flags; + irq_hw_number_t parent_hwirq; + struct eiointc_ip_route route_info[VEC_REG_COUNT]; }; static struct eiointc_priv *eiointc_priv[MAX_IO_PICS]; @@ -188,6 +199,7 @@ static int eiointc_router_init(unsigned int cpu) { int i, bit, cores, index, node; unsigned int data; + int hwirq, mask; node = cpu_to_eio_node(cpu); index = eiointc_index(node); @@ -197,6 +209,13 @@ static int eiointc_router_init(unsigned int cpu) return -EINVAL; } + /* Enable cpu interrupt pin from eiointc */ + hwirq = eiointc_priv[index]->parent_hwirq; + mask = BIT(hwirq); + if (eiointc_priv[index]->flags & EIOINTC_ROUTE_MULT_IP) + mask |= BIT(hwirq + 1) | BIT(hwirq + 2) | BIT(hwirq + 3); + set_csr_ecfg(mask); + if (!(eiointc_priv[index]->flags & EIOINTC_USE_CPU_ENCODE)) cores = CORES_PER_EIO_NODE; else @@ -211,8 +230,31 @@ static int eiointc_router_init(unsigned int cpu) } for (i = 0; i < eiointc_priv[0]->vec_count / 32 / 4; i++) { - bit = BIT(1 + index); /* Route to IP[1 + index] */ - data = bit | (bit << 8) | (bit << 16) | (bit << 24); + /* + * Route to interrupt pin, relative offset used here + * Offset 0 means routing to IP0 and so on + * + * If flags is set with EIOINTC_ROUTE_MULT_IP, + * every 64 vector routes to different consecutive + * IPs, otherwise all vector routes to the same IP + */ + if (eiointc_priv[index]->flags & EIOINTC_ROUTE_MULT_IP) { + /* The first 64 vectors route to hwirq */ + bit = BIT(hwirq++ - INT_HWI0); + data = bit | (bit << 8); + + /* The second 64 vectors route to hwirq + 1 */ + bit = BIT(hwirq++ - INT_HWI0); + data |= (bit << 16) | (bit << 24); + + /* + * Route to hwirq + 2/hwirq + 3 separately + * in next loop + */ + } else { + bit = BIT(hwirq - INT_HWI0); + data = bit | (bit << 8) | (bit << 16) | (bit << 24); + } iocsr_write32(data, EIOINTC_REG_IPMAP + i * 4); } @@ -241,15 +283,22 @@ static int eiointc_router_init(unsigned int cpu) static void eiointc_irq_dispatch(struct irq_desc *desc) { - int i; - u64 pending; - bool handled = false; + struct eiointc_ip_route *info = irq_desc_get_handler_data(desc); struct irq_chip *chip = irq_desc_get_chip(desc); - struct eiointc_priv *priv = irq_desc_get_handler_data(desc); + bool handled = false; + u64 pending; + int i; chained_irq_enter(chip, desc); - for (i = 0; i < eiointc_priv[0]->vec_count / VEC_COUNT_PER_REG; i++) { + /* + * If EIOINTC_ROUTE_MULT_IP is set, every 64 interrupt vectors in + * eiointc interrupt controller routes to different cpu interrupt pins + * + * Every cpu interrupt pin has its own irq handler, it is ok to + * read ISR for these 64 interrupt vectors rather than all vectors + */ + for (i = info->start; i < info->end; i++) { pending = iocsr_read64(EIOINTC_REG_ISR + (i << 3)); /* Skip handling if pending bitmap is zero */ @@ -262,7 +311,7 @@ static void eiointc_irq_dispatch(struct irq_desc *desc) int bit = __ffs(pending); int irq = bit + VEC_COUNT_PER_REG * i; - generic_handle_domain_irq(priv->eiointc_domain, irq); + generic_handle_domain_irq(info->priv->eiointc_domain, irq); pending &= ~BIT(bit); handled = true; } @@ -462,8 +511,33 @@ static int __init eiointc_init(struct eiointc_priv *priv, int parent_irq, } eiointc_priv[nr_pics++] = priv; + /* + * Only the first eiointc device on VM supports routing to + * different CPU interrupt pins. The later eiointc devices use + * generic method if there are multiple eiointc devices in future + */ + if (cpu_has_hypervisor && (nr_pics == 1)) { + priv->flags |= EIOINTC_ROUTE_MULT_IP; + priv->parent_hwirq = INT_HWI0; + } + + if (priv->flags & EIOINTC_ROUTE_MULT_IP) { + for (i = 0; i < priv->vec_count / VEC_COUNT_PER_REG; i++) { + priv->route_info[i].start = priv->parent_hwirq - INT_HWI0 + i; + priv->route_info[i].end = priv->route_info[i].start + 1; + priv->route_info[i].priv = priv; + parent_irq = get_percpu_irq(priv->parent_hwirq + i); + irq_set_chained_handler_and_data(parent_irq, eiointc_irq_dispatch, + &priv->route_info[i]); + } + } else { + priv->route_info[0].start = 0; + priv->route_info[0].end = priv->vec_count / VEC_COUNT_PER_REG; + priv->route_info[0].priv = priv; + irq_set_chained_handler_and_data(parent_irq, eiointc_irq_dispatch, + &priv->route_info[0]); + } eiointc_router_init(0); - irq_set_chained_handler_and_data(parent_irq, eiointc_irq_dispatch, priv); if (nr_pics == 1) { register_syscore_ops(&eiointc_syscore_ops); @@ -495,7 +569,7 @@ int __init eiointc_acpi_init(struct irq_domain *parent, priv->vec_count = VEC_COUNT; priv->node = acpi_eiointc->node; - + priv->parent_hwirq = acpi_eiointc->cascade; parent_irq = irq_create_mapping(parent, acpi_eiointc->cascade); ret = eiointc_init(priv, parent_irq, acpi_eiointc->node_map); @@ -527,8 +601,9 @@ out_free_priv: static int __init eiointc_of_init(struct device_node *of_node, struct device_node *parent) { - int parent_irq, ret; struct eiointc_priv *priv; + struct irq_data *irq_data; + int parent_irq, ret; priv = kzalloc(sizeof(*priv), GFP_KERNEL); if (!priv) @@ -544,6 +619,12 @@ static int __init eiointc_of_init(struct device_node *of_node, if (ret < 0) goto out_free_priv; + irq_data = irq_get_irq_data(parent_irq); + if (!irq_data) { + ret = -ENODEV; + goto out_free_priv; + } + /* * In particular, the number of devices supported by the LS2K0500 * extended I/O interrupt vector is 128. @@ -552,7 +633,7 @@ static int __init eiointc_of_init(struct device_node *of_node, priv->vec_count = 128; else priv->vec_count = VEC_COUNT; - + priv->parent_hwirq = irqd_to_hwirq(irq_data); priv->node = 0; priv->domain_handle = of_fwnode_handle(of_node); diff --git a/drivers/irqchip/irq-loongson-pch-lpc.c b/drivers/irqchip/irq-loongson-pch-lpc.c index 2d4c3ec128b8..912bf50a5c7c 100644 --- a/drivers/irqchip/irq-loongson-pch-lpc.c +++ b/drivers/irqchip/irq-loongson-pch-lpc.c @@ -200,8 +200,13 @@ int __init pch_lpc_acpi_init(struct irq_domain *parent, goto iounmap_base; } - priv->lpc_domain = irq_domain_create_linear(irq_handle, LPC_COUNT, - &pch_lpc_domain_ops, priv); + /* + * The LPC interrupt controller is a legacy i8259-compatible device, + * which requires a static 1:1 mapping for IRQs 0-15. + * Use irq_domain_create_legacy to establish this static mapping early. + */ + priv->lpc_domain = irq_domain_create_legacy(irq_handle, LPC_COUNT, 0, 0, + &pch_lpc_domain_ops, priv); if (!priv->lpc_domain) { pr_err("Failed to create IRQ domain\n"); goto free_irq_handle; diff --git a/drivers/irqchip/irq-loongson-pch-msi.c b/drivers/irqchip/irq-loongson-pch-msi.c index a0257c7bef10..4aedc9b90ff7 100644 --- a/drivers/irqchip/irq-loongson-pch-msi.c +++ b/drivers/irqchip/irq-loongson-pch-msi.c @@ -153,26 +153,21 @@ static struct msi_parent_ops pch_msi_parent_ops = { .init_dev_msi_info = msi_lib_init_dev_msi_info, }; -static int pch_msi_init_domains(struct pch_msi_data *priv, - struct irq_domain *parent, +static int pch_msi_init_domains(struct pch_msi_data *priv, struct irq_domain *parent, struct fwnode_handle *domain_handle) { - struct irq_domain *middle_domain; - - middle_domain = irq_domain_create_hierarchy(parent, 0, priv->num_irqs, - domain_handle, - &pch_msi_middle_domain_ops, - priv); - if (!middle_domain) { + struct irq_domain_info info = { + .ops = &pch_msi_middle_domain_ops, + .size = priv->num_irqs, + .parent = parent, + .host_data = priv, + .fwnode = domain_handle, + }; + + if (!msi_create_parent_irq_domain(&info, &pch_msi_parent_ops)) { pr_err("Failed to create the MSI middle domain\n"); return -ENOMEM; } - - irq_domain_update_bus_token(middle_domain, DOMAIN_BUS_NEXUS); - - middle_domain->flags |= IRQ_DOMAIN_FLAG_MSI_PARENT; - middle_domain->msi_parent_ops = &pch_msi_parent_ops; - return 0; } diff --git a/drivers/irqchip/irq-ls-scfg-msi.c b/drivers/irqchip/irq-ls-scfg-msi.c index 84bc5e4b47cf..4910f364e568 100644 --- a/drivers/irqchip/irq-ls-scfg-msi.c +++ b/drivers/irqchip/irq-ls-scfg-msi.c @@ -14,6 +14,7 @@ #include <linux/iommu.h> #include <linux/irq.h> #include <linux/irqchip/chained_irq.h> +#include <linux/irqchip/irq-msi-lib.h> #include <linux/irqdomain.h> #include <linux/of_irq.h> #include <linux/of_pci.h> @@ -47,7 +48,6 @@ struct ls_scfg_msi { spinlock_t lock; struct platform_device *pdev; struct irq_domain *parent; - struct irq_domain *msi_domain; void __iomem *regs; phys_addr_t msiir_addr; struct ls_scfg_msi_cfg *cfg; @@ -57,17 +57,18 @@ struct ls_scfg_msi { unsigned long *used; }; -static struct irq_chip ls_scfg_msi_irq_chip = { - .name = "MSI", - .irq_mask = pci_msi_mask_irq, - .irq_unmask = pci_msi_unmask_irq, -}; - -static struct msi_domain_info ls_scfg_msi_domain_info = { - .flags = (MSI_FLAG_USE_DEF_DOM_OPS | - MSI_FLAG_USE_DEF_CHIP_OPS | - MSI_FLAG_PCI_MSIX), - .chip = &ls_scfg_msi_irq_chip, +#define MPIC_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS | \ + MSI_FLAG_USE_DEF_CHIP_OPS) +#define MPIC_MSI_FLAGS_SUPPORTED (MSI_FLAG_PCI_MSIX | \ + MSI_GENERIC_FLAGS_MASK) + +static const struct msi_parent_ops ls_scfg_msi_parent_ops = { + .required_flags = MPIC_MSI_FLAGS_REQUIRED, + .supported_flags = MPIC_MSI_FLAGS_SUPPORTED, + .bus_select_token = DOMAIN_BUS_NEXUS, + .bus_select_mask = MATCH_PCI_MSI, + .prefix = "MSI-", + .init_dev_msi_info = msi_lib_init_dev_msi_info, }; static int msi_affinity_flag = 1; @@ -185,6 +186,7 @@ static void ls_scfg_msi_domain_irq_free(struct irq_domain *domain, } static const struct irq_domain_ops ls_scfg_msi_domain_ops = { + .select = msi_lib_irq_domain_select, .alloc = ls_scfg_msi_domain_irq_alloc, .free = ls_scfg_msi_domain_irq_free, }; @@ -214,23 +216,16 @@ static void ls_scfg_msi_irq_handler(struct irq_desc *desc) static int ls_scfg_msi_domains_init(struct ls_scfg_msi *msi_data) { - /* Initialize MSI domain parent */ - msi_data->parent = irq_domain_create_linear(NULL, - msi_data->irqs_num, - &ls_scfg_msi_domain_ops, - msi_data); + struct irq_domain_info info = { + .fwnode = of_fwnode_handle(msi_data->pdev->dev.of_node), + .ops = &ls_scfg_msi_domain_ops, + .host_data = msi_data, + .size = msi_data->irqs_num, + }; + + msi_data->parent = msi_create_parent_irq_domain(&info, &ls_scfg_msi_parent_ops); if (!msi_data->parent) { - dev_err(&msi_data->pdev->dev, "failed to create IRQ domain\n"); - return -ENOMEM; - } - - msi_data->msi_domain = pci_msi_create_irq_domain( - of_fwnode_handle(msi_data->pdev->dev.of_node), - &ls_scfg_msi_domain_info, - msi_data->parent); - if (!msi_data->msi_domain) { dev_err(&msi_data->pdev->dev, "failed to create MSI domain\n"); - irq_domain_remove(msi_data->parent); return -ENOMEM; } @@ -405,7 +400,6 @@ static void ls_scfg_msi_remove(struct platform_device *pdev) for (i = 0; i < msi_data->msir_num; i++) ls_scfg_msi_teardown_hwirq(&msi_data->msir[i]); - irq_domain_remove(msi_data->msi_domain); irq_domain_remove(msi_data->parent); platform_set_drvdata(pdev, NULL); diff --git a/drivers/irqchip/irq-mips-gic.c b/drivers/irqchip/irq-mips-gic.c index 34e8d09c12a0..19a57c5e2b2e 100644 --- a/drivers/irqchip/irq-mips-gic.c +++ b/drivers/irqchip/irq-mips-gic.c @@ -375,9 +375,13 @@ static int gic_set_affinity(struct irq_data *d, const struct cpumask *cpumask, /* * The GIC specifies that we can only route an interrupt to one VP(E), * ie. CPU in Linux parlance, at a time. Therefore we always route to - * the first online CPU in the mask. + * the first forced or online CPU in the mask. */ - cpu = cpumask_first_and(cpumask, cpu_online_mask); + if (force) + cpu = cpumask_first(cpumask); + else + cpu = cpumask_first_and(cpumask, cpu_online_mask); + if (cpu >= NR_CPUS) return -EINVAL; diff --git a/drivers/irqchip/irq-msi-lib.c b/drivers/irqchip/irq-msi-lib.c index 246c30205af4..d5eefc3d7215 100644 --- a/drivers/irqchip/irq-msi-lib.c +++ b/drivers/irqchip/irq-msi-lib.c @@ -112,6 +112,20 @@ bool msi_lib_init_dev_msi_info(struct device *dev, struct irq_domain *domain, */ if (!chip->irq_set_affinity && !(info->flags & MSI_FLAG_NO_AFFINITY)) chip->irq_set_affinity = msi_domain_set_affinity; + + /* + * If the parent domain insists on being in charge of masking, obey + * blindly. The interrupt is un-masked at the PCI level on startup + * and masked on shutdown to prevent rogue interrupts after the + * driver freed the interrupt. Not masking it at the PCI level + * speeds up operation for disable/enable_irq() as it avoids + * getting all the way out to the PCI device. + */ + if (info->flags & MSI_FLAG_PCI_MSI_MASK_PARENT) { + chip->irq_mask = irq_chip_mask_parent; + chip->irq_unmask = irq_chip_unmask_parent; + } + return true; } EXPORT_SYMBOL_GPL(msi_lib_init_dev_msi_info); @@ -137,7 +151,10 @@ int msi_lib_irq_domain_select(struct irq_domain *d, struct irq_fwspec *fwspec, if (!ops) return 0; - if (fwspec->fwnode != d->fwnode || fwspec->param_count != 0) + struct fwnode_handle *fwh __free(fwnode_handle) = + d->flags & IRQ_DOMAIN_FLAG_FWNODE_PARENT ? fwnode_get_parent(fwspec->fwnode) + : fwnode_handle_get(fwspec->fwnode); + if (fwh != d->fwnode || fwspec->param_count != 0) return 0; /* Handle pure domain searches */ diff --git a/drivers/irqchip/irq-mvebu-gicp.c b/drivers/irqchip/irq-mvebu-gicp.c index d3232d6d8dce..667bde3c651f 100644 --- a/drivers/irqchip/irq-mvebu-gicp.c +++ b/drivers/irqchip/irq-mvebu-gicp.c @@ -177,6 +177,7 @@ static int mvebu_gicp_probe(struct platform_device *pdev) .ops = &gicp_domain_ops, }; struct mvebu_gicp *gicp; + void __iomem *base; int ret, i; gicp = devm_kzalloc(&pdev->dev, sizeof(*gicp), GFP_KERNEL); @@ -236,6 +237,15 @@ static int mvebu_gicp_probe(struct platform_device *pdev) return -ENODEV; } + base = ioremap(gicp->res->start, resource_size(gicp->res)); + if (!base) { + dev_err(&pdev->dev, "ioremap() failed. Unable to clear pending interrupts.\n"); + } else { + for (i = 0; i < 64; i++) + writel(i, base + GICP_CLRSPI_NSR_OFFSET); + iounmap(base); + } + return msi_create_parent_irq_domain(&info, &gicp_msi_parent_ops) ? 0 : -ENOMEM; } diff --git a/drivers/irqchip/irq-mvebu-pic.c b/drivers/irqchip/irq-mvebu-pic.c index 8db638aa21d2..cd8b73482b9f 100644 --- a/drivers/irqchip/irq-mvebu-pic.c +++ b/drivers/irqchip/irq-mvebu-pic.c @@ -150,7 +150,7 @@ static int mvebu_pic_probe(struct platform_device *pdev) return -EINVAL; } - pic->domain = irq_domain_create_linear(of_fwnode_handle(node), PIC_MAX_IRQS, + pic->domain = irq_domain_create_linear(dev_fwnode(&pdev->dev), PIC_MAX_IRQS, &mvebu_pic_domain_ops, pic); if (!pic->domain) { dev_err(&pdev->dev, "Failed to allocate irq domain\n"); diff --git a/drivers/irqchip/irq-nvic.c b/drivers/irqchip/irq-nvic.c index 76e11cac9631..2191a2b79578 100644 --- a/drivers/irqchip/irq-nvic.c +++ b/drivers/irqchip/irq-nvic.c @@ -73,8 +73,9 @@ static int __init nvic_of_init(struct device_node *node, struct device_node *parent) { unsigned int clr = IRQ_NOREQUEST | IRQ_NOPROBE | IRQ_NOAUTOEN; - unsigned int irqs, i, ret, numbanks; + unsigned int irqs, i, numbanks; void __iomem *nvic_base; + int ret; numbanks = (readl_relaxed(V7M_SCS_ICTR) & V7M_SCS_ICTR_INTLINESNUM_MASK) + 1; diff --git a/drivers/irqchip/irq-pruss-intc.c b/drivers/irqchip/irq-pruss-intc.c index 87a5813fd835..81078d56f38d 100644 --- a/drivers/irqchip/irq-pruss-intc.c +++ b/drivers/irqchip/irq-pruss-intc.c @@ -555,7 +555,7 @@ static int pruss_intc_probe(struct platform_device *pdev) mutex_init(&intc->lock); - intc->domain = irq_domain_create_linear(of_fwnode_handle(dev->of_node), max_system_events, + intc->domain = irq_domain_create_linear(dev_fwnode(dev), max_system_events, &pruss_intc_irq_domain_ops, intc); if (!intc->domain) return -ENOMEM; diff --git a/drivers/irqchip/irq-renesas-intc-irqpin.c b/drivers/irqchip/irq-renesas-intc-irqpin.c index 0959ed43b1a9..7951292d2d9b 100644 --- a/drivers/irqchip/irq-renesas-intc-irqpin.c +++ b/drivers/irqchip/irq-renesas-intc-irqpin.c @@ -513,10 +513,8 @@ static int intc_irqpin_probe(struct platform_device *pdev) irq_chip->irq_set_wake = intc_irqpin_irq_set_wake; irq_chip->flags = IRQCHIP_MASK_ON_SUSPEND; - p->irq_domain = irq_domain_create_simple(of_fwnode_handle(dev->of_node), - nirqs, 0, - &intc_irqpin_irq_domain_ops, - p); + p->irq_domain = irq_domain_create_simple(dev_fwnode(dev), nirqs, 0, + &intc_irqpin_irq_domain_ops, p); if (!p->irq_domain) { ret = -ENXIO; dev_err(dev, "cannot initialize irq domain\n"); @@ -572,7 +570,7 @@ static void intc_irqpin_remove(struct platform_device *pdev) pm_runtime_disable(&pdev->dev); } -static int __maybe_unused intc_irqpin_suspend(struct device *dev) +static int intc_irqpin_suspend(struct device *dev) { struct intc_irqpin_priv *p = dev_get_drvdata(dev); @@ -582,7 +580,7 @@ static int __maybe_unused intc_irqpin_suspend(struct device *dev) return 0; } -static SIMPLE_DEV_PM_OPS(intc_irqpin_pm_ops, intc_irqpin_suspend, NULL); +static DEFINE_SIMPLE_DEV_PM_OPS(intc_irqpin_pm_ops, intc_irqpin_suspend, NULL); static struct platform_driver intc_irqpin_device_driver = { .probe = intc_irqpin_probe, @@ -590,7 +588,7 @@ static struct platform_driver intc_irqpin_device_driver = { .driver = { .name = "renesas_intc_irqpin", .of_match_table = intc_irqpin_dt_ids, - .pm = &intc_irqpin_pm_ops, + .pm = pm_sleep_ptr(&intc_irqpin_pm_ops), } }; diff --git a/drivers/irqchip/irq-renesas-irqc.c b/drivers/irqchip/irq-renesas-irqc.c index 5c3196e5a437..a20a6471b0e4 100644 --- a/drivers/irqchip/irq-renesas-irqc.c +++ b/drivers/irqchip/irq-renesas-irqc.c @@ -168,7 +168,7 @@ static int irqc_probe(struct platform_device *pdev) p->cpu_int_base = p->iomem + IRQC_INT_CPU_BASE(0); /* SYS-SPI */ - p->irq_domain = irq_domain_create_linear(of_fwnode_handle(dev->of_node), p->number_of_irqs, + p->irq_domain = irq_domain_create_linear(dev_fwnode(dev), p->number_of_irqs, &irq_generic_chip_ops, p); if (!p->irq_domain) { ret = -ENXIO; @@ -227,7 +227,7 @@ static void irqc_remove(struct platform_device *pdev) pm_runtime_disable(&pdev->dev); } -static int __maybe_unused irqc_suspend(struct device *dev) +static int irqc_suspend(struct device *dev) { struct irqc_priv *p = dev_get_drvdata(dev); @@ -237,7 +237,7 @@ static int __maybe_unused irqc_suspend(struct device *dev) return 0; } -static SIMPLE_DEV_PM_OPS(irqc_pm_ops, irqc_suspend, NULL); +static DEFINE_SIMPLE_DEV_PM_OPS(irqc_pm_ops, irqc_suspend, NULL); static const struct of_device_id irqc_dt_ids[] = { { .compatible = "renesas,irqc", }, @@ -251,7 +251,7 @@ static struct platform_driver irqc_device_driver = { .driver = { .name = "renesas_irqc", .of_match_table = irqc_dt_ids, - .pm = &irqc_pm_ops, + .pm = pm_sleep_ptr(&irqc_pm_ops), } }; diff --git a/drivers/irqchip/irq-renesas-rza1.c b/drivers/irqchip/irq-renesas-rza1.c index 0a9640ba0adb..6047a524ac77 100644 --- a/drivers/irqchip/irq-renesas-rza1.c +++ b/drivers/irqchip/irq-renesas-rza1.c @@ -142,11 +142,12 @@ static const struct irq_domain_ops rza1_irqc_domain_ops = { static int rza1_irqc_parse_map(struct rza1_irqc_priv *priv, struct device_node *gic_node) { - unsigned int imaplen, i, j, ret; struct device *dev = priv->dev; + unsigned int imaplen, i, j; struct device_node *ipar; const __be32 *imap; u32 intsize; + int ret; imap = of_get_property(dev->of_node, "interrupt-map", &imaplen); if (!imap) @@ -231,9 +232,8 @@ static int rza1_irqc_probe(struct platform_device *pdev) priv->chip.irq_set_type = rza1_irqc_set_type; priv->chip.flags = IRQCHIP_MASK_ON_SUSPEND | IRQCHIP_SKIP_SET_WAKE; - priv->irq_domain = irq_domain_create_hierarchy(parent, 0, IRQC_NUM_IRQ, - of_fwnode_handle(np), &rza1_irqc_domain_ops, - priv); + priv->irq_domain = irq_domain_create_hierarchy(parent, 0, IRQC_NUM_IRQ, dev_fwnode(dev), + &rza1_irqc_domain_ops, priv); if (!priv->irq_domain) { dev_err(dev, "cannot initialize irq domain\n"); ret = -ENOMEM; diff --git a/drivers/irqchip/irq-renesas-rzg2l.c b/drivers/irqchip/irq-renesas-rzg2l.c index 1e861bd64f97..2a54adeb4cc7 100644 --- a/drivers/irqchip/irq-renesas-rzg2l.c +++ b/drivers/irqchip/irq-renesas-rzg2l.c @@ -574,12 +574,11 @@ static int rzg2l_irqc_common_init(struct device_node *node, struct device_node * raw_spin_lock_init(&rzg2l_irqc_data->lock); - irq_domain = irq_domain_create_hierarchy(parent_domain, 0, IRQC_NUM_IRQ, - of_fwnode_handle(node), &rzg2l_irqc_domain_ops, - rzg2l_irqc_data); + irq_domain = irq_domain_create_hierarchy(parent_domain, 0, IRQC_NUM_IRQ, dev_fwnode(dev), + &rzg2l_irqc_domain_ops, rzg2l_irqc_data); if (!irq_domain) { pm_runtime_put(dev); - return dev_err_probe(dev, -ENOMEM, "failed to add irq domain\n"); + return -ENOMEM; } register_syscore_ops(&rzg2l_irqc_syscore_ops); diff --git a/drivers/irqchip/irq-renesas-rzv2h.c b/drivers/irqchip/irq-renesas-rzv2h.c index 69b32c19e8ff..9018d9c3911e 100644 --- a/drivers/irqchip/irq-renesas-rzv2h.c +++ b/drivers/irqchip/irq-renesas-rzv2h.c @@ -11,18 +11,15 @@ #include <linux/bitfield.h> #include <linux/cleanup.h> -#include <linux/clk.h> #include <linux/err.h> #include <linux/io.h> #include <linux/irqchip.h> #include <linux/irqchip/irq-renesas-rzv2h.h> #include <linux/irqdomain.h> -#include <linux/of_address.h> #include <linux/of_platform.h> #include <linux/pm_runtime.h> #include <linux/reset.h> #include <linux/spinlock.h> -#include <linux/syscore_ops.h> /* DT "interrupts" indexes */ #define ICU_IRQ_START 1 @@ -427,7 +424,9 @@ static const struct irq_chip rzv2h_icu_chip = { .irq_retrigger = irq_chip_retrigger_hierarchy, .irq_set_type = rzv2h_icu_set_type, .irq_set_affinity = irq_chip_set_affinity_parent, - .flags = IRQCHIP_SET_TYPE_MASKED, + .flags = IRQCHIP_MASK_ON_SUSPEND | + IRQCHIP_SET_TYPE_MASKED | + IRQCHIP_SKIP_SET_WAKE, }; static int rzv2h_icu_alloc(struct irq_domain *domain, unsigned int virq, unsigned int nr_irqs, @@ -558,7 +557,7 @@ static int rzv2h_icu_init_common(struct device_node *node, struct device_node *p raw_spin_lock_init(&rzv2h_icu_data->lock); irq_domain = irq_domain_create_hierarchy(parent_domain, 0, ICU_NUM_IRQ, - of_fwnode_handle(node), &rzv2h_icu_domain_ops, + dev_fwnode(&pdev->dev), &rzv2h_icu_domain_ops, rzv2h_icu_data); if (!irq_domain) { dev_err(&pdev->dev, "failed to add irq domain\n"); diff --git a/drivers/irqchip/irq-riscv-aplic-direct.c b/drivers/irqchip/irq-riscv-aplic-direct.c index 205ad61d15e4..c2a75bf3d20c 100644 --- a/drivers/irqchip/irq-riscv-aplic-direct.c +++ b/drivers/irqchip/irq-riscv-aplic-direct.c @@ -219,20 +219,6 @@ static int aplic_direct_parse_parent_hwirq(struct device *dev, u32 index, return 0; } -static int aplic_direct_get_hart_index(struct device *dev, u32 logical_index, - u32 *hart_index) -{ - const char *prop_hart_index = "riscv,hart-indexes"; - struct device_node *np = to_of_node(dev->fwnode); - - if (!np || !of_property_present(np, prop_hart_index)) { - *hart_index = logical_index; - return 0; - } - - return of_property_read_u32_index(np, prop_hart_index, logical_index, hart_index); -} - int aplic_direct_setup(struct device *dev, void __iomem *regs) { int i, j, rc, cpu, current_cpu, setup_count = 0; @@ -279,7 +265,7 @@ int aplic_direct_setup(struct device *dev, void __iomem *regs) cpumask_set_cpu(cpu, &direct->lmask); idc = per_cpu_ptr(&aplic_idcs, cpu); - rc = aplic_direct_get_hart_index(dev, i, &idc->hart_index); + rc = riscv_get_hart_index(dev->fwnode, i, &idc->hart_index); if (rc) { dev_warn(dev, "hart index not found for IDC%d\n", i); continue; diff --git a/drivers/irqchip/irq-riscv-imsic-early.c b/drivers/irqchip/irq-riscv-imsic-early.c index d9ae87808651..2709cacf4855 100644 --- a/drivers/irqchip/irq-riscv-imsic-early.c +++ b/drivers/irqchip/irq-riscv-imsic-early.c @@ -8,6 +8,7 @@ #include <linux/acpi.h> #include <linux/cpu.h> #include <linux/interrupt.h> +#include <linux/init.h> #include <linux/io.h> #include <linux/irq.h> #include <linux/irqchip.h> @@ -21,6 +22,14 @@ #include "irq-riscv-imsic-state.h" static int imsic_parent_irq; +bool imsic_noipi __ro_after_init; + +static int __init imsic_noipi_cfg(char *buf) +{ + imsic_noipi = true; + return 0; +} +early_param("irqchip.riscv_imsic_noipi", imsic_noipi_cfg); #ifdef CONFIG_SMP static void imsic_ipi_send(unsigned int cpu) @@ -32,12 +41,18 @@ static void imsic_ipi_send(unsigned int cpu) static void imsic_ipi_starting_cpu(void) { + if (imsic_noipi) + return; + /* Enable IPIs for current CPU. */ __imsic_id_set_enable(IMSIC_IPI_ID); } static void imsic_ipi_dying_cpu(void) { + if (imsic_noipi) + return; + /* Disable IPIs for current CPU. */ __imsic_id_clear_enable(IMSIC_IPI_ID); } @@ -46,6 +61,9 @@ static int __init imsic_ipi_domain_init(void) { int virq; + if (imsic_noipi) + return 0; + /* Create IMSIC IPI multiplexing */ virq = ipi_mux_create(IMSIC_NR_IPI, imsic_ipi_send); if (virq <= 0) @@ -88,7 +106,7 @@ static void imsic_handle_irq(struct irq_desc *desc) while ((local_id = csr_swap(CSR_TOPEI, 0))) { local_id >>= TOPEI_ID_SHIFT; - if (local_id == IMSIC_IPI_ID) { + if (!imsic_noipi && local_id == IMSIC_IPI_ID) { if (IS_ENABLED(CONFIG_SMP)) ipi_mux_process(); continue; diff --git a/drivers/irqchip/irq-riscv-imsic-platform.c b/drivers/irqchip/irq-riscv-imsic-platform.c index 1b9fbfce9581..643c8e459611 100644 --- a/drivers/irqchip/irq-riscv-imsic-platform.c +++ b/drivers/irqchip/irq-riscv-imsic-platform.c @@ -307,6 +307,10 @@ static const struct msi_parent_ops imsic_msi_parent_ops = { int imsic_irqdomain_init(void) { + struct irq_domain_info info = { + .ops = &imsic_base_domain_ops, + .host_data = imsic, + }; struct imsic_global_config *global; if (!imsic || !imsic->fwnode) { @@ -320,16 +324,12 @@ int imsic_irqdomain_init(void) } /* Create Base IRQ domain */ - imsic->base_domain = irq_domain_create_tree(imsic->fwnode, - &imsic_base_domain_ops, imsic); + info.fwnode = imsic->fwnode, + imsic->base_domain = msi_create_parent_irq_domain(&info, &imsic_msi_parent_ops); if (!imsic->base_domain) { pr_err("%pfwP: failed to create IMSIC base domain\n", imsic->fwnode); return -ENOMEM; } - imsic->base_domain->flags |= IRQ_DOMAIN_FLAG_MSI_PARENT; - imsic->base_domain->msi_parent_ops = &imsic_msi_parent_ops; - - irq_domain_update_bus_token(imsic->base_domain, DOMAIN_BUS_NEXUS); global = &imsic->global; pr_info("%pfwP: hart-index-bits: %d, guest-index-bits: %d\n", diff --git a/drivers/irqchip/irq-riscv-imsic-state.c b/drivers/irqchip/irq-riscv-imsic-state.c index 77670dd645ac..dc95ad856d80 100644 --- a/drivers/irqchip/irq-riscv-imsic-state.c +++ b/drivers/irqchip/irq-riscv-imsic-state.c @@ -134,7 +134,7 @@ static bool __imsic_local_sync(struct imsic_local_priv *lpriv) lockdep_assert_held(&lpriv->lock); for_each_set_bit(i, lpriv->dirty_bitmap, imsic->global.nr_ids + 1) { - if (!i || i == IMSIC_IPI_ID) + if (!i || (!imsic_noipi && i == IMSIC_IPI_ID)) goto skip; vec = &lpriv->vectors[i]; @@ -419,7 +419,7 @@ void imsic_vector_debug_show(struct seq_file *m, struct imsic_vector *vec, int i seq_printf(m, "%*starget_cpu : %5u\n", ind, "", vec->cpu); seq_printf(m, "%*starget_local_id : %5u\n", ind, "", vec->local_id); seq_printf(m, "%*sis_reserved : %5u\n", ind, "", - (vec->local_id <= IMSIC_IPI_ID) ? 1 : 0); + (!imsic_noipi && vec->local_id <= IMSIC_IPI_ID) ? 1 : 0); seq_printf(m, "%*sis_enabled : %5u\n", ind, "", is_enabled ? 1 : 0); seq_printf(m, "%*sis_move_pending : %5u\n", ind, "", mvec ? 1 : 0); if (mvec) { @@ -583,7 +583,8 @@ static int __init imsic_matrix_init(void) irq_matrix_assign_system(imsic->matrix, 0, false); /* Reserve IPI ID because it is special and used internally */ - irq_matrix_assign_system(imsic->matrix, IMSIC_IPI_ID, false); + if (!imsic_noipi) + irq_matrix_assign_system(imsic->matrix, IMSIC_IPI_ID, false); return 0; } diff --git a/drivers/irqchip/irq-riscv-imsic-state.h b/drivers/irqchip/irq-riscv-imsic-state.h index 3202ffa4e849..57f951952b0c 100644 --- a/drivers/irqchip/irq-riscv-imsic-state.h +++ b/drivers/irqchip/irq-riscv-imsic-state.h @@ -61,6 +61,7 @@ struct imsic_priv { struct irq_domain *base_domain; }; +extern bool imsic_noipi; extern struct imsic_priv *imsic; void __imsic_eix_update(unsigned long base_id, unsigned long num_id, bool pend, bool val); diff --git a/drivers/irqchip/irq-sg2042-msi.c b/drivers/irqchip/irq-sg2042-msi.c index af16bc5a3c8b..f7cf0dc72eab 100644 --- a/drivers/irqchip/irq-sg2042-msi.c +++ b/drivers/irqchip/irq-sg2042-msi.c @@ -30,6 +30,7 @@ struct sg204x_msi_chip_info { * @doorbell_addr: see TRM, 10.1.32, GP_INTR0_SET * @irq_first: First vectors number that MSIs starts * @num_irqs: Number of vectors for MSIs + * @irq_type: IRQ type for MSIs * @msi_map: mapping for allocated MSI vectors. * @msi_map_lock: Lock for msi_map * @chip_info: chip specific infomations @@ -41,6 +42,7 @@ struct sg204x_msi_chipdata { u32 irq_first; u32 num_irqs; + unsigned int irq_type; unsigned long *msi_map; struct mutex msi_map_lock; @@ -85,6 +87,8 @@ static void sg2042_msi_irq_compose_msi_msg(struct irq_data *d, struct msi_msg *m static const struct irq_chip sg2042_msi_middle_irq_chip = { .name = "SG2042 MSI", + .irq_startup = irq_chip_startup_parent, + .irq_shutdown = irq_chip_shutdown_parent, .irq_ack = sg2042_msi_irq_ack, .irq_mask = irq_chip_mask_parent, .irq_unmask = irq_chip_unmask_parent, @@ -114,6 +118,8 @@ static void sg2044_msi_irq_compose_msi_msg(struct irq_data *d, struct msi_msg *m static struct irq_chip sg2044_msi_middle_irq_chip = { .name = "SG2044 MSI", + .irq_startup = irq_chip_startup_parent, + .irq_shutdown = irq_chip_shutdown_parent, .irq_ack = sg2044_msi_irq_ack, .irq_mask = irq_chip_mask_parent, .irq_unmask = irq_chip_unmask_parent, @@ -133,14 +139,14 @@ static int sg204x_msi_parent_domain_alloc(struct irq_domain *domain, unsigned in fwspec.fwnode = domain->parent->fwnode; fwspec.param_count = 2; fwspec.param[0] = data->irq_first + hwirq; - fwspec.param[1] = IRQ_TYPE_EDGE_RISING; + fwspec.param[1] = data->irq_type; ret = irq_domain_alloc_irqs_parent(domain, virq, 1, &fwspec); if (ret) return ret; d = irq_domain_get_irq_data(domain->parent, virq); - return d->chip->irq_set_type(d, IRQ_TYPE_EDGE_RISING); + return d->chip->irq_set_type(d, data->irq_type); } static int sg204x_msi_middle_domain_alloc(struct irq_domain *domain, unsigned int virq, @@ -185,8 +191,10 @@ static const struct irq_domain_ops sg204x_msi_middle_domain_ops = { .select = msi_lib_irq_domain_select, }; -#define SG2042_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS | \ - MSI_FLAG_USE_DEF_CHIP_OPS) +#define SG2042_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS | \ + MSI_FLAG_USE_DEF_CHIP_OPS | \ + MSI_FLAG_PCI_MSI_MASK_PARENT | \ + MSI_FLAG_PCI_MSI_STARTUP_PARENT) #define SG2042_MSI_FLAGS_SUPPORTED MSI_GENERIC_FLAGS_MASK @@ -200,10 +208,13 @@ static const struct msi_parent_ops sg2042_msi_parent_ops = { .init_dev_msi_info = msi_lib_init_dev_msi_info, }; -#define SG2044_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS | \ - MSI_FLAG_USE_DEF_CHIP_OPS) +#define SG2044_MSI_FLAGS_REQUIRED (MSI_FLAG_USE_DEF_DOM_OPS | \ + MSI_FLAG_USE_DEF_CHIP_OPS | \ + MSI_FLAG_PCI_MSI_MASK_PARENT | \ + MSI_FLAG_PCI_MSI_STARTUP_PARENT) -#define SG2044_MSI_FLAGS_SUPPORTED (MSI_GENERIC_FLAGS_MASK | \ +#define SG2044_MSI_FLAGS_SUPPORTED (MSI_GENERIC_FLAGS_MASK | \ + MSI_FLAG_MULTI_PCI_MSI | \ MSI_FLAG_PCI_MSIX) static const struct msi_parent_ops sg2044_msi_parent_ops = { @@ -219,20 +230,18 @@ static const struct msi_parent_ops sg2044_msi_parent_ops = { static int sg204x_msi_init_domains(struct sg204x_msi_chipdata *data, struct irq_domain *plic_domain, struct device *dev) { - struct fwnode_handle *fwnode = dev_fwnode(dev); - struct irq_domain *middle_domain; - - middle_domain = irq_domain_create_hierarchy(plic_domain, 0, data->num_irqs, fwnode, - &sg204x_msi_middle_domain_ops, data); - if (!middle_domain) { + struct irq_domain_info info = { + .ops = &sg204x_msi_middle_domain_ops, + .parent = plic_domain, + .size = data->num_irqs, + .fwnode = dev_fwnode(dev), + .host_data = data, + }; + + if (!msi_create_parent_irq_domain(&info, data->chip_info->parent_ops)) { pr_err("Failed to create the MSI middle domain\n"); return -ENOMEM; } - - irq_domain_update_bus_token(middle_domain, DOMAIN_BUS_NEXUS); - - middle_domain->flags |= IRQ_DOMAIN_FLAG_MSI_PARENT; - middle_domain->msi_parent_ops = data->chip_info->parent_ops; return 0; } @@ -291,6 +300,7 @@ static int sg2042_msi_probe(struct platform_device *pdev) } data->irq_first = (u32)args.args[0]; + data->irq_type = (unsigned int)args.args[1]; data->num_irqs = (u32)args.args[args.nargs - 1]; mutex_init(&data->msi_map_lock); diff --git a/drivers/irqchip/irq-sifive-plic.c b/drivers/irqchip/irq-sifive-plic.c index bf69a4802b71..559fda8fb3a8 100644 --- a/drivers/irqchip/irq-sifive-plic.c +++ b/drivers/irqchip/irq-sifive-plic.c @@ -179,12 +179,14 @@ static int plic_set_affinity(struct irq_data *d, if (cpu >= nr_cpu_ids) return -EINVAL; - plic_irq_disable(d); + /* Invalidate the original routing entry */ + plic_irq_toggle(irq_data_get_effective_affinity_mask(d), d, 0); irq_data_update_effective_affinity(d, cpumask_of(cpu)); + /* Setting the new routing entry if irq is enabled */ if (!irqd_irq_disabled(d)) - plic_irq_enable(d); + plic_irq_toggle(irq_data_get_effective_affinity_mask(d), d, 1); return IRQ_SET_MASK_OK_DONE; } @@ -257,7 +259,7 @@ static int plic_irq_suspend(void) readl(priv->regs + PRIORITY_BASE + i * PRIORITY_PER_ID)); } - for_each_cpu(cpu, cpu_present_mask) { + for_each_present_cpu(cpu) { struct plic_handler *handler = per_cpu_ptr(&plic_handlers, cpu); if (!handler->present) @@ -289,7 +291,7 @@ static void plic_irq_resume(void) priv->regs + PRIORITY_BASE + i * PRIORITY_PER_ID); } - for_each_cpu(cpu, cpu_present_mask) { + for_each_present_cpu(cpu) { struct plic_handler *handler = per_cpu_ptr(&plic_handlers, cpu); if (!handler->present) diff --git a/drivers/irqchip/irq-stm32mp-exti.c b/drivers/irqchip/irq-stm32mp-exti.c index c6b4407d05f9..a24f4f1a4f8f 100644 --- a/drivers/irqchip/irq-stm32mp-exti.c +++ b/drivers/irqchip/irq-stm32mp-exti.c @@ -683,9 +683,7 @@ static int stm32mp_exti_probe(struct platform_device *pdev) } domain = irq_domain_create_hierarchy(parent_domain, 0, drv_data->bank_nr * IRQS_PER_BANK, - of_fwnode_handle(np), &stm32mp_exti_domain_ops, - host_data); - + dev_fwnode(dev), &stm32mp_exti_domain_ops, host_data); if (!domain) { dev_err(dev, "Could not register exti domain\n"); return -ENOMEM; diff --git a/drivers/irqchip/irq-ti-sci-inta.c b/drivers/irqchip/irq-ti-sci-inta.c index 7de59238e6b0..01963d36cfaf 100644 --- a/drivers/irqchip/irq-ti-sci-inta.c +++ b/drivers/irqchip/irq-ti-sci-inta.c @@ -701,8 +701,7 @@ static int ti_sci_inta_irq_domain_probe(struct platform_device *pdev) if (ret) return ret; - domain = irq_domain_create_linear(of_fwnode_handle(dev_of_node(dev)), - ti_sci_get_num_resources(inta->vint), + domain = irq_domain_create_linear(dev_fwnode(dev), ti_sci_get_num_resources(inta->vint), &ti_sci_inta_irq_domain_ops, inta); if (!domain) { dev_err(dev, "Failed to allocate IRQ domain\n"); diff --git a/drivers/irqchip/irq-ti-sci-intr.c b/drivers/irqchip/irq-ti-sci-intr.c index 07fff5ae5ce0..354613e74ad0 100644 --- a/drivers/irqchip/irq-ti-sci-intr.c +++ b/drivers/irqchip/irq-ti-sci-intr.c @@ -274,8 +274,7 @@ static int ti_sci_intr_irq_domain_probe(struct platform_device *pdev) return PTR_ERR(intr->out_irqs); } - domain = irq_domain_create_hierarchy(parent_domain, 0, 0, - of_fwnode_handle(dev_of_node(dev)), + domain = irq_domain_create_hierarchy(parent_domain, 0, 0, dev_fwnode(dev), &ti_sci_intr_irq_domain_ops, intr); if (!domain) { dev_err(dev, "Failed to allocate IRQ domain\n"); diff --git a/drivers/irqchip/irq-ts4800.c b/drivers/irqchip/irq-ts4800.c index e625f4fb2bb8..1e236d5b7516 100644 --- a/drivers/irqchip/irq-ts4800.c +++ b/drivers/irqchip/irq-ts4800.c @@ -125,7 +125,7 @@ static int ts4800_ic_probe(struct platform_device *pdev) return -EINVAL; } - data->domain = irq_domain_create_linear(of_fwnode_handle(node), 8, &ts4800_ic_ops, data); + data->domain = irq_domain_create_linear(dev_fwnode(&pdev->dev), 8, &ts4800_ic_ops, data); if (!data->domain) { dev_err(&pdev->dev, "cannot add IRQ domain\n"); return -ENOMEM; |