diff options
Diffstat (limited to 'arch/x86/kernel/apic/io_apic.c')
| -rw-r--r-- | arch/x86/kernel/apic/io_apic.c | 1560 |
1 files changed, 753 insertions, 807 deletions
diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c index 2953bbf05c08..28f934f05a85 100644 --- a/arch/x86/kernel/apic/io_apic.c +++ b/arch/x86/kernel/apic/io_apic.c @@ -48,6 +48,7 @@ #include <linux/jiffies.h> /* time_after() */ #include <linux/slab.h> #include <linux/memblock.h> +#include <linux/msi.h> #include <asm/irqdomain.h> #include <asm/io.h> @@ -58,12 +59,14 @@ #include <asm/acpi.h> #include <asm/dma.h> #include <asm/timer.h> +#include <asm/time.h> #include <asm/i8259.h> #include <asm/setup.h> #include <asm/irq_remapping.h> #include <asm/hw_irq.h> - #include <asm/apic.h> +#include <asm/pgtable.h> +#include <asm/x86_init.h> #define for_each_ioapic(idx) \ for ((idx) = 0; (idx) < nr_ioapics; (idx)++) @@ -83,17 +86,17 @@ static unsigned int ioapic_dynirq_base; static int ioapic_initialized; struct irq_pin_list { - struct list_head list; - int apic, pin; + struct list_head list; + int apic, pin; }; struct mp_chip_data { - struct list_head irq_2_pin; - struct IO_APIC_route_entry entry; - int trigger; - int polarity; - u32 count; - bool isa_irq; + struct list_head irq_2_pin; + struct IO_APIC_route_entry entry; + bool is_level; + bool active_low; + bool isa_irq; + u32 count; }; struct mp_ioapic_gsi { @@ -102,21 +105,17 @@ struct mp_ioapic_gsi { }; static struct ioapic { - /* - * # of IRQ routing registers - */ - int nr_registers; - /* - * Saved state during suspend/resume, or while enabling intr-remap. - */ - struct IO_APIC_route_entry *saved_registers; + /* # of IRQ routing registers */ + int nr_registers; + /* Saved state during suspend/resume, or while enabling intr-remap. */ + struct IO_APIC_route_entry *saved_registers; /* I/O APIC config */ - struct mpc_ioapic mp_config; + struct mpc_ioapic mp_config; /* IO APIC gsi routing info */ - struct mp_ioapic_gsi gsi_config; - struct ioapic_domain_cfg irqdomain_cfg; - struct irq_domain *irqdomain; - struct resource *iomem_res; + struct mp_ioapic_gsi gsi_config; + struct ioapic_domain_cfg irqdomain_cfg; + struct irq_domain *irqdomain; + struct resource *iomem_res; } ioapics[MAX_IO_APICS]; #define mpc_ioapic_ver(ioapic_idx) ioapics[ioapic_idx].mp_config.apicver @@ -153,19 +152,6 @@ static inline bool mp_is_legacy_irq(int irq) return irq >= 0 && irq < nr_legacy_irqs(); } -/* - * Initialize all legacy IRQs and all pins on the first IOAPIC - * if we have legacy interrupt controller. Kernel boot option "pirq=" - * may rely on non-legacy pins on the first IOAPIC. - */ -static inline int mp_init_irq_at_boot(int ioapic, int irq) -{ - if (!nr_legacy_irqs()) - return 0; - - return ioapic == 0 || mp_is_legacy_irq(irq); -} - static inline struct irq_domain *mp_ioapic_irqdomain(int ioapic) { return ioapics[ioapic].irqdomain; @@ -188,7 +174,7 @@ int mp_bus_id_to_type[MAX_MP_BUSSES]; DECLARE_BITMAP(mp_bus_not_pci, MAX_MP_BUSSES); -int skip_ioapic_setup; +bool ioapic_is_disabled __ro_after_init; /** * disable_ioapic_support() - disables ioapic support at runtime @@ -199,7 +185,7 @@ void disable_ioapic_support(void) noioapicquirk = 1; noioapicreroute = -1; #endif - skip_ioapic_setup = 1; + ioapic_is_disabled = true; } static int __init parse_noapic(char *str) @@ -210,15 +196,14 @@ static int __init parse_noapic(char *str) } early_param("noapic", parse_noapic); -/* Will be called in mpparse/acpi/sfi codes for saving IRQ info */ +/* Will be called in mpparse/ACPI codes for saving IRQ info */ void mp_save_irq(struct mpc_intsrc *m) { int i; - apic_printk(APIC_VERBOSE, "Int: type %d, pol %d, trig %d, bus %02x," - " IRQ %02x, APIC ID %x, APIC INT %02x\n", - m->irqtype, m->irqflag & 3, (m->irqflag >> 2) & 3, m->srcbus, - m->srcbusirq, m->dstapic, m->dstirq); + apic_pr_verbose("Int: type %d, pol %d, trig %d, bus %02x, IRQ %02x, APIC ID %x, APIC INT %02x\n", + m->irqtype, m->irqflag & 3, (m->irqflag >> 2) & 3, m->srcbus, + m->srcbusirq, m->dstapic, m->dstirq); for (i = 0; i < mp_irq_entries; i++) { if (!memcmp(&mp_irqs[i], m, sizeof(*m))) @@ -279,12 +264,14 @@ static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx) static inline void io_apic_eoi(unsigned int apic, unsigned int vector) { struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(vector, &io_apic->eoi); } unsigned int native_io_apic_read(unsigned int apic, unsigned int reg) { struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(reg, &io_apic->index); return readl(&io_apic->data); } @@ -298,31 +285,20 @@ static void io_apic_write(unsigned int apic, unsigned int reg, writel(value, &io_apic->data); } -union entry_union { - struct { u32 w1, w2; }; - struct IO_APIC_route_entry entry; -}; - static struct IO_APIC_route_entry __ioapic_read_entry(int apic, int pin) { - union entry_union eu; + struct IO_APIC_route_entry entry; - eu.w1 = io_apic_read(apic, 0x10 + 2 * pin); - eu.w2 = io_apic_read(apic, 0x11 + 2 * pin); + entry.w1 = io_apic_read(apic, 0x10 + 2 * pin); + entry.w2 = io_apic_read(apic, 0x11 + 2 * pin); - return eu.entry; + return entry; } static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin) { - union entry_union eu; - unsigned long flags; - - raw_spin_lock_irqsave(&ioapic_lock, flags); - eu.entry = __ioapic_read_entry(apic, pin); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - - return eu.entry; + guard(raw_spinlock_irqsave)(&ioapic_lock); + return __ioapic_read_entry(apic, pin); } /* @@ -333,20 +309,14 @@ static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin) */ static void __ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) { - union entry_union eu = {{0, 0}}; - - eu.entry = e; - io_apic_write(apic, 0x11 + 2*pin, eu.w2); - io_apic_write(apic, 0x10 + 2*pin, eu.w1); + io_apic_write(apic, 0x11 + 2*pin, e.w2); + io_apic_write(apic, 0x10 + 2*pin, e.w1); } static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) { - unsigned long flags; - - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); __ioapic_write_entry(apic, pin, e); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } /* @@ -356,13 +326,11 @@ static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) */ static void ioapic_mask_entry(int apic, int pin) { - unsigned long flags; - union entry_union eu = { .entry.mask = IOAPIC_MASKED }; + struct IO_APIC_route_entry e = { .masked = true }; - raw_spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(apic, 0x10 + 2*pin, eu.w1); - io_apic_write(apic, 0x11 + 2*pin, eu.w2); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); + io_apic_write(apic, 0x10 + 2*pin, e.w1); + io_apic_write(apic, 0x11 + 2*pin, e.w2); } /* @@ -370,95 +338,61 @@ static void ioapic_mask_entry(int apic, int pin) * shared ISA-space IRQs, so we have to support them. We are super * fast in the common case, and fast for shared ISA-space IRQs. */ -static int __add_pin_to_irq_node(struct mp_chip_data *data, - int node, int apic, int pin) +static bool add_pin_to_irq_node(struct mp_chip_data *data, int node, int apic, int pin) { struct irq_pin_list *entry; - /* don't allow duplicates */ - for_each_irq_pin(entry, data->irq_2_pin) + /* Don't allow duplicates */ + for_each_irq_pin(entry, data->irq_2_pin) { if (entry->apic == apic && entry->pin == pin) - return 0; + return true; + } entry = kzalloc_node(sizeof(struct irq_pin_list), GFP_ATOMIC, node); if (!entry) { - pr_err("can not alloc irq_pin_list (%d,%d,%d)\n", - node, apic, pin); - return -ENOMEM; + pr_err("Cannot allocate irq_pin_list (%d,%d,%d)\n", node, apic, pin); + return false; } + entry->apic = apic; entry->pin = pin; list_add_tail(&entry->list, &data->irq_2_pin); - - return 0; + return true; } static void __remove_pin_from_irq(struct mp_chip_data *data, int apic, int pin) { struct irq_pin_list *tmp, *entry; - list_for_each_entry_safe(entry, tmp, &data->irq_2_pin, list) + list_for_each_entry_safe(entry, tmp, &data->irq_2_pin, list) { if (entry->apic == apic && entry->pin == pin) { list_del(&entry->list); kfree(entry); return; } -} - -static void add_pin_to_irq_node(struct mp_chip_data *data, - int node, int apic, int pin) -{ - if (__add_pin_to_irq_node(data, node, apic, pin)) - panic("IO-APIC: failed to add irq-pin. Can not proceed\n"); -} - -/* - * Reroute an IRQ to a different pin. - */ -static void __init replace_pin_at_irq_node(struct mp_chip_data *data, int node, - int oldapic, int oldpin, - int newapic, int newpin) -{ - struct irq_pin_list *entry; - - for_each_irq_pin(entry, data->irq_2_pin) { - if (entry->apic == oldapic && entry->pin == oldpin) { - entry->apic = newapic; - entry->pin = newpin; - /* every one is different, right? */ - return; - } } - - /* old apic/pin didn't exist, so just add new ones */ - add_pin_to_irq_node(data, node, newapic, newpin); } -static void io_apic_modify_irq(struct mp_chip_data *data, - int mask_and, int mask_or, +static void io_apic_modify_irq(struct mp_chip_data *data, bool masked, void (*final)(struct irq_pin_list *entry)) { - union entry_union eu; struct irq_pin_list *entry; - eu.entry = data->entry; - eu.w1 &= mask_and; - eu.w1 |= mask_or; - data->entry = eu.entry; + data->entry.masked = masked; for_each_irq_pin(entry, data->irq_2_pin) { - io_apic_write(entry->apic, 0x10 + 2 * entry->pin, eu.w1); + io_apic_write(entry->apic, 0x10 + 2 * entry->pin, data->entry.w1); if (final) final(entry); } } +/* + * Synchronize the IO-APIC and the CPU by doing a dummy read from the + * IO-APIC + */ static void io_apic_sync(struct irq_pin_list *entry) { - /* - * Synchronize the IO-APIC and the CPU by doing - * a dummy read from the IO-APIC - */ struct io_apic __iomem *io_apic; io_apic = io_apic_base(entry->apic); @@ -468,26 +402,22 @@ static void io_apic_sync(struct irq_pin_list *entry) static void mask_ioapic_irq(struct irq_data *irq_data) { struct mp_chip_data *data = irq_data->chip_data; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); - io_apic_modify_irq(data, ~0, IO_APIC_REDIR_MASKED, &io_apic_sync); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); + io_apic_modify_irq(data, true, &io_apic_sync); } static void __unmask_ioapic(struct mp_chip_data *data) { - io_apic_modify_irq(data, ~IO_APIC_REDIR_MASKED, 0, NULL); + io_apic_modify_irq(data, false, NULL); } static void unmask_ioapic_irq(struct irq_data *irq_data) { struct mp_chip_data *data = irq_data->chip_data; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); __unmask_ioapic(data); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } /* @@ -515,30 +445,24 @@ static void __eoi_ioapic_pin(int apic, int pin, int vector) entry = entry1 = __ioapic_read_entry(apic, pin); - /* - * Mask the entry and change the trigger mode to edge. - */ - entry1.mask = IOAPIC_MASKED; - entry1.trigger = IOAPIC_EDGE; + /* Mask the entry and change the trigger mode to edge. */ + entry1.masked = true; + entry1.is_level = false; __ioapic_write_entry(apic, pin, entry1); - /* - * Restore the previous level triggered entry. - */ + /* Restore the previous level triggered entry. */ __ioapic_write_entry(apic, pin, entry); } } static void eoi_ioapic_pin(int vector, struct mp_chip_data *data) { - unsigned long flags; struct irq_pin_list *entry; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); for_each_irq_pin(entry, data->irq_2_pin) __eoi_ioapic_pin(entry->apic, entry->pin, vector); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) @@ -547,34 +471,31 @@ static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) /* Check delivery_mode to be sure we're not clearing an SMI pin */ entry = ioapic_read_entry(apic, pin); - if (entry.delivery_mode == dest_SMI) + if (entry.delivery_mode == APIC_DELIVERY_MODE_SMI) return; /* * Make sure the entry is masked and re-read the contents to check * if it is a level triggered pin and if the remote-IRR is set. */ - if (entry.mask == IOAPIC_UNMASKED) { - entry.mask = IOAPIC_MASKED; + if (!entry.masked) { + entry.masked = true; ioapic_write_entry(apic, pin, entry); entry = ioapic_read_entry(apic, pin); } if (entry.irr) { - unsigned long flags; - /* * Make sure the trigger mode is set to level. Explicit EOI * doesn't clear the remote-IRR if the trigger mode is not * set to level. */ - if (entry.trigger == IOAPIC_EDGE) { - entry.trigger = IOAPIC_LEVEL; + if (!entry.is_level) { + entry.is_level = true; ioapic_write_entry(apic, pin, entry); } - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); __eoi_ioapic_pin(apic, pin, entry.vector); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } /* @@ -609,28 +530,23 @@ static int pirq_entries[MAX_PIRQS] = { static int __init ioapic_pirq_setup(char *str) { - int i, max; - int ints[MAX_PIRQS+1]; + int i, max, ints[MAX_PIRQS+1]; get_options(str, ARRAY_SIZE(ints), ints); - apic_printk(APIC_VERBOSE, KERN_INFO - "PIRQ redirection, working around broken MP-BIOS.\n"); + apic_pr_verbose("PIRQ redirection, working around broken MP-BIOS.\n"); + max = MAX_PIRQS; if (ints[0] < MAX_PIRQS) max = ints[0]; for (i = 0; i < max; i++) { - apic_printk(APIC_VERBOSE, KERN_DEBUG - "... PIRQ%d -> IRQ %d\n", i, ints[i+1]); - /* - * PIRQs are mapped upside down, usually. - */ + apic_pr_verbose("... PIRQ%d -> IRQ %d\n", i, ints[i + 1]); + /* PIRQs are mapped upside down, usually */ pirq_entries[MAX_PIRQS-i-1] = ints[i+1]; } return 1; } - __setup("pirq=", ioapic_pirq_setup); #endif /* CONFIG_X86_32 */ @@ -649,8 +565,7 @@ int save_ioapic_entries(void) } for_each_pin(apic, pin) - ioapics[apic].saved_registers[pin] = - ioapic_read_entry(apic, pin); + ioapics[apic].saved_registers[pin] = ioapic_read_entry(apic, pin); } return err; @@ -671,8 +586,8 @@ void mask_ioapic_entries(void) struct IO_APIC_route_entry entry; entry = ioapics[apic].saved_registers[pin]; - if (entry.mask == IOAPIC_UNMASKED) { - entry.mask = IOAPIC_MASKED; + if (!entry.masked) { + entry.masked = true; ioapic_write_entry(apic, pin, entry); } } @@ -691,8 +606,7 @@ int restore_ioapic_entries(void) continue; for_each_pin(apic, pin) - ioapic_write_entry(apic, pin, - ioapics[apic].saved_registers[pin]); + ioapic_write_entry(apic, pin, ioapics[apic].saved_registers[pin]); } return 0; } @@ -704,12 +618,13 @@ static int find_irq_entry(int ioapic_idx, int pin, int type) { int i; - for (i = 0; i < mp_irq_entries; i++) + for (i = 0; i < mp_irq_entries; i++) { if (mp_irqs[i].irqtype == type && (mp_irqs[i].dstapic == mpc_ioapic_id(ioapic_idx) || mp_irqs[i].dstapic == MP_APIC_ALL) && mp_irqs[i].dstirq == pin) return i; + } return -1; } @@ -724,10 +639,8 @@ static int __init find_isa_irq_pin(int irq, int type) for (i = 0; i < mp_irq_entries; i++) { int lbus = mp_irqs[i].srcbus; - if (test_bit(lbus, mp_bus_not_pci) && - (mp_irqs[i].irqtype == type) && + if (test_bit(lbus, mp_bus_not_pci) && (mp_irqs[i].irqtype == type) && (mp_irqs[i].srcbusirq == irq)) - return mp_irqs[i].dstirq; } return -1; @@ -740,8 +653,7 @@ static int __init find_isa_irq_apic(int irq, int type) for (i = 0; i < mp_irq_entries; i++) { int lbus = mp_irqs[i].srcbus; - if (test_bit(lbus, mp_bus_not_pci) && - (mp_irqs[i].irqtype == type) && + if (test_bit(lbus, mp_bus_not_pci) && (mp_irqs[i].irqtype == type) && (mp_irqs[i].srcbusirq == irq)) break; } @@ -749,52 +661,16 @@ static int __init find_isa_irq_apic(int irq, int type) if (i < mp_irq_entries) { int ioapic_idx; - for_each_ioapic(ioapic_idx) + for_each_ioapic(ioapic_idx) { if (mpc_ioapic_id(ioapic_idx) == mp_irqs[i].dstapic) return ioapic_idx; + } } return -1; } -#ifdef CONFIG_EISA -/* - * EISA Edge/Level control register, ELCR - */ -static int EISA_ELCR(unsigned int irq) -{ - if (irq < nr_legacy_irqs()) { - unsigned int port = 0x4d0 + (irq >> 3); - return (inb(port) >> (irq & 7)) & 1; - } - apic_printk(APIC_VERBOSE, KERN_INFO - "Broken MPtable reports ISA irq %d\n", irq); - return 0; -} - -#endif - -/* ISA interrupts are always active high edge triggered, - * when listed as conforming in the MP table. */ - -#define default_ISA_trigger(idx) (IOAPIC_EDGE) -#define default_ISA_polarity(idx) (IOAPIC_POL_HIGH) - -/* EISA interrupts are always polarity zero and can be edge or level - * trigger depending on the ELCR value. If an interrupt is listed as - * EISA conforming in the MP table, that means its trigger type must - * be read in from the ELCR */ - -#define default_EISA_trigger(idx) (EISA_ELCR(mp_irqs[idx].srcbusirq)) -#define default_EISA_polarity(idx) default_ISA_polarity(idx) - -/* PCI interrupts are always active low level triggered, - * when listed as conforming in the MP table. */ - -#define default_PCI_trigger(idx) (IOAPIC_LEVEL) -#define default_PCI_polarity(idx) (IOAPIC_POL_LOW) - -static int irq_polarity(int idx) +static bool irq_active_low(int idx) { int bus = mp_irqs[idx].srcbus; @@ -803,125 +679,175 @@ static int irq_polarity(int idx) */ switch (mp_irqs[idx].irqflag & MP_IRQPOL_MASK) { case MP_IRQPOL_DEFAULT: - /* conforms to spec, ie. bus-type dependent polarity */ - if (test_bit(bus, mp_bus_not_pci)) - return default_ISA_polarity(idx); - else - return default_PCI_polarity(idx); + /* + * Conforms to spec, ie. bus-type dependent polarity. PCI + * defaults to low active. [E]ISA defaults to high active. + */ + return !test_bit(bus, mp_bus_not_pci); case MP_IRQPOL_ACTIVE_HIGH: - return IOAPIC_POL_HIGH; + return false; case MP_IRQPOL_RESERVED: pr_warn("IOAPIC: Invalid polarity: 2, defaulting to low\n"); + fallthrough; case MP_IRQPOL_ACTIVE_LOW: default: /* Pointless default required due to do gcc stupidity */ - return IOAPIC_POL_LOW; + return true; } } #ifdef CONFIG_EISA -static int eisa_irq_trigger(int idx, int bus, int trigger) +/* + * EISA Edge/Level control register, ELCR + */ +static bool EISA_ELCR(unsigned int irq) +{ + if (irq < nr_legacy_irqs()) { + unsigned int port = PIC_ELCR1 + (irq >> 3); + return (inb(port) >> (irq & 7)) & 1; + } + apic_pr_verbose("Broken MPtable reports ISA irq %d\n", irq); + return false; +} + +/* + * EISA interrupts are always active high and can be edge or level + * triggered depending on the ELCR value. If an interrupt is listed as + * EISA conforming in the MP table, that means its trigger type must be + * read in from the ELCR. + */ +static bool eisa_irq_is_level(int idx, int bus, bool level) { switch (mp_bus_id_to_type[bus]) { case MP_BUS_PCI: case MP_BUS_ISA: - return trigger; + return level; case MP_BUS_EISA: - return default_EISA_trigger(idx); + return EISA_ELCR(mp_irqs[idx].srcbusirq); } pr_warn("IOAPIC: Invalid srcbus: %d defaulting to level\n", bus); - return IOAPIC_LEVEL; + return true; } #else -static inline int eisa_irq_trigger(int idx, int bus, int trigger) +static inline int eisa_irq_is_level(int idx, int bus, bool level) { - return trigger; + return level; } #endif -static int irq_trigger(int idx) +static bool irq_is_level(int idx) { int bus = mp_irqs[idx].srcbus; - int trigger; + bool level; /* * Determine IRQ trigger mode (edge or level sensitive): */ switch (mp_irqs[idx].irqflag & MP_IRQTRIG_MASK) { case MP_IRQTRIG_DEFAULT: - /* conforms to spec, ie. bus-type dependent trigger mode */ - if (test_bit(bus, mp_bus_not_pci)) - trigger = default_ISA_trigger(idx); - else - trigger = default_PCI_trigger(idx); + /* + * Conforms to spec, ie. bus-type dependent trigger + * mode. PCI defaults to level, ISA to edge. + */ + level = !test_bit(bus, mp_bus_not_pci); /* Take EISA into account */ - return eisa_irq_trigger(idx, bus, trigger); + return eisa_irq_is_level(idx, bus, level); case MP_IRQTRIG_EDGE: - return IOAPIC_EDGE; + return false; case MP_IRQTRIG_RESERVED: pr_warn("IOAPIC: Invalid trigger mode 2 defaulting to level\n"); + fallthrough; case MP_IRQTRIG_LEVEL: default: /* Pointless default required due to do gcc stupidity */ - return IOAPIC_LEVEL; + return true; } } +static int __acpi_get_override_irq(u32 gsi, bool *trigger, bool *polarity) +{ + int ioapic, pin, idx; + + if (ioapic_is_disabled) + return -1; + + ioapic = mp_find_ioapic(gsi); + if (ioapic < 0) + return -1; + + pin = mp_find_ioapic_pin(ioapic, gsi); + if (pin < 0) + return -1; + + idx = find_irq_entry(ioapic, pin, mp_INT); + if (idx < 0) + return -1; + + *trigger = irq_is_level(idx); + *polarity = irq_active_low(idx); + return 0; +} + +#ifdef CONFIG_ACPI +int acpi_get_override_irq(u32 gsi, int *is_level, int *active_low) +{ + *is_level = *active_low = 0; + return __acpi_get_override_irq(gsi, (bool *)is_level, + (bool *)active_low); +} +#endif + void ioapic_set_alloc_attr(struct irq_alloc_info *info, int node, int trigger, int polarity) { init_irq_alloc_info(info, NULL); info->type = X86_IRQ_ALLOC_TYPE_IOAPIC; - info->ioapic_node = node; - info->ioapic_trigger = trigger; - info->ioapic_polarity = polarity; - info->ioapic_valid = 1; + info->ioapic.node = node; + info->ioapic.is_level = trigger; + info->ioapic.active_low = polarity; + info->ioapic.valid = 1; } -#ifndef CONFIG_ACPI -int acpi_get_override_irq(u32 gsi, int *trigger, int *polarity); -#endif - static void ioapic_copy_alloc_attr(struct irq_alloc_info *dst, struct irq_alloc_info *src, u32 gsi, int ioapic_idx, int pin) { - int trigger, polarity; + bool level, pol_low; copy_irq_alloc_info(dst, src); dst->type = X86_IRQ_ALLOC_TYPE_IOAPIC; - dst->ioapic_id = mpc_ioapic_id(ioapic_idx); - dst->ioapic_pin = pin; - dst->ioapic_valid = 1; - if (src && src->ioapic_valid) { - dst->ioapic_node = src->ioapic_node; - dst->ioapic_trigger = src->ioapic_trigger; - dst->ioapic_polarity = src->ioapic_polarity; + dst->devid = mpc_ioapic_id(ioapic_idx); + dst->ioapic.pin = pin; + dst->ioapic.valid = 1; + if (src && src->ioapic.valid) { + dst->ioapic.node = src->ioapic.node; + dst->ioapic.is_level = src->ioapic.is_level; + dst->ioapic.active_low = src->ioapic.active_low; } else { - dst->ioapic_node = NUMA_NO_NODE; - if (acpi_get_override_irq(gsi, &trigger, &polarity) >= 0) { - dst->ioapic_trigger = trigger; - dst->ioapic_polarity = polarity; + dst->ioapic.node = NUMA_NO_NODE; + if (__acpi_get_override_irq(gsi, &level, &pol_low) >= 0) { + dst->ioapic.is_level = level; + dst->ioapic.active_low = pol_low; } else { /* * PCI interrupts are always active low level * triggered. */ - dst->ioapic_trigger = IOAPIC_LEVEL; - dst->ioapic_polarity = IOAPIC_POL_LOW; + dst->ioapic.is_level = true; + dst->ioapic.active_low = true; } } } static int ioapic_alloc_attr_node(struct irq_alloc_info *info) { - return (info && info->ioapic_valid) ? info->ioapic_node : NUMA_NO_NODE; + return (info && info->ioapic.valid) ? info->ioapic.node : NUMA_NO_NODE; } -static void mp_register_handler(unsigned int irq, unsigned long trigger) +static void mp_register_handler(unsigned int irq, bool level) { irq_flow_handler_t hdl; bool fasteoi; - if (trigger) { + if (level) { irq_set_status_flags(irq, IRQ_LEVEL); fasteoi = true; } else { @@ -939,26 +865,26 @@ static bool mp_check_pin_attr(int irq, struct irq_alloc_info *info) /* * setup_IO_APIC_irqs() programs all legacy IRQs with default trigger - * and polarity attirbutes. So allow the first user to reprogram the + * and polarity attributes. So allow the first user to reprogram the * pin with real trigger and polarity attributes. */ if (irq < nr_legacy_irqs() && data->count == 1) { - if (info->ioapic_trigger != data->trigger) - mp_register_handler(irq, info->ioapic_trigger); - data->entry.trigger = data->trigger = info->ioapic_trigger; - data->entry.polarity = data->polarity = info->ioapic_polarity; + if (info->ioapic.is_level != data->is_level) + mp_register_handler(irq, info->ioapic.is_level); + data->entry.is_level = data->is_level = info->ioapic.is_level; + data->entry.active_low = data->active_low = info->ioapic.active_low; } - return data->trigger == info->ioapic_trigger && - data->polarity == info->ioapic_polarity; + return data->is_level == info->ioapic.is_level && + data->active_low == info->ioapic.active_low; } static int alloc_irq_from_domain(struct irq_domain *domain, int ioapic, u32 gsi, struct irq_alloc_info *info) { + int type = ioapics[ioapic].irqdomain_cfg.type; bool legacy = false; int irq = -1; - int type = ioapics[ioapic].irqdomain_cfg.type; switch (type) { case IOAPIC_DOMAIN_LEGACY: @@ -980,8 +906,7 @@ static int alloc_irq_from_domain(struct irq_domain *domain, int ioapic, u32 gsi, return -1; } - return __irq_domain_alloc_irqs(domain, irq, 1, - ioapic_alloc_attr_node(info), + return __irq_domain_alloc_irqs(domain, irq, 1, ioapic_alloc_attr_node(info), info, legacy, NULL); } @@ -995,29 +920,26 @@ static int alloc_irq_from_domain(struct irq_domain *domain, int ioapic, u32 gsi, * PIRQs instead of reprogramming the interrupt routing logic. Thus there may be * multiple pins sharing the same legacy IRQ number when ACPI is disabled. */ -static int alloc_isa_irq_from_domain(struct irq_domain *domain, - int irq, int ioapic, int pin, +static int alloc_isa_irq_from_domain(struct irq_domain *domain, int irq, int ioapic, int pin, struct irq_alloc_info *info) { - struct mp_chip_data *data; struct irq_data *irq_data = irq_get_irq_data(irq); int node = ioapic_alloc_attr_node(info); + struct mp_chip_data *data; /* * Legacy ISA IRQ has already been allocated, just add pin to - * the pin list assoicated with this IRQ and program the IOAPIC - * entry. The IOAPIC entry + * the pin list associated with this IRQ and program the IOAPIC + * entry. */ if (irq_data && irq_data->parent_data) { if (!mp_check_pin_attr(irq, info)) return -EBUSY; - if (__add_pin_to_irq_node(irq_data->chip_data, node, ioapic, - info->ioapic_pin)) + if (!add_pin_to_irq_node(irq_data->chip_data, node, ioapic, info->ioapic.pin)) return -ENOMEM; } else { info->flags |= X86_IRQ_ALLOC_LEGACY; - irq = __irq_domain_alloc_irqs(domain, irq, 1, node, info, true, - NULL); + irq = __irq_domain_alloc_irqs(domain, irq, 1, node, info, true, NULL); if (irq >= 0) { irq_data = irq_domain_get_irq_data(domain, irq); data = irq_data->chip_data; @@ -1031,11 +953,11 @@ static int alloc_isa_irq_from_domain(struct irq_domain *domain, static int mp_map_pin_to_irq(u32 gsi, int idx, int ioapic, int pin, unsigned int flags, struct irq_alloc_info *info) { - int irq; - bool legacy = false; + struct irq_domain *domain = mp_ioapic_irqdomain(ioapic); struct irq_alloc_info tmp; struct mp_chip_data *data; - struct irq_domain *domain = mp_ioapic_irqdomain(ioapic); + bool legacy = false; + int irq; if (!domain) return -ENOSYS; @@ -1043,9 +965,19 @@ static int mp_map_pin_to_irq(u32 gsi, int idx, int ioapic, int pin, if (idx >= 0 && test_bit(mp_irqs[idx].srcbus, mp_bus_not_pci)) { irq = mp_irqs[idx].srcbusirq; legacy = mp_is_legacy_irq(irq); + /* + * IRQ2 is unusable for historical reasons on systems which + * have a legacy PIC. See the comment vs. IRQ2 further down. + * + * If this gets removed at some point then the related code + * in lapic_assign_system_vectors() needs to be adjusted as + * well. + */ + if (legacy && irq == PIC_CASCADE_IR) + return -EINVAL; } - mutex_lock(&ioapic_mutex); + guard(mutex)(&ioapic_mutex); if (!(flags & IOAPIC_MAP_ALLOC)) { if (!legacy) { irq = irq_find_mapping(domain, pin); @@ -1066,8 +998,6 @@ static int mp_map_pin_to_irq(u32 gsi, int idx, int ioapic, int pin, data->count++; } } - mutex_unlock(&ioapic_mutex); - return irq; } @@ -1075,26 +1005,20 @@ static int pin_2_irq(int idx, int ioapic, int pin, unsigned int flags) { u32 gsi = mp_pin_to_gsi(ioapic, pin); - /* - * Debugging check, we are in big trouble if this message pops up! - */ + /* Debugging check, we are in big trouble if this message pops up! */ if (mp_irqs[idx].dstirq != pin) pr_err("broken BIOS or MPTABLE parser, ayiee!!\n"); #ifdef CONFIG_X86_32 - /* - * PCI IRQ command line redirection. Yes, limits are hardcoded. - */ + /* PCI IRQ command line redirection. Yes, limits are hardcoded. */ if ((pin >= 16) && (pin <= 23)) { - if (pirq_entries[pin-16] != -1) { - if (!pirq_entries[pin-16]) { - apic_printk(APIC_VERBOSE, KERN_DEBUG - "disabling PIRQ%d\n", pin-16); + if (pirq_entries[pin - 16] != -1) { + if (!pirq_entries[pin - 16]) { + apic_pr_verbose("Disabling PIRQ%d\n", pin - 16); } else { int irq = pirq_entries[pin-16]; - apic_printk(APIC_VERBOSE, KERN_DEBUG - "using PIRQ%d -> IRQ %d\n", - pin-16, irq); + + apic_pr_verbose("Using PIRQ%d -> IRQ %d\n", pin - 16, irq); return irq; } } @@ -1132,10 +1056,9 @@ void mp_unmap_irq(int irq) if (!data || data->isa_irq) return; - mutex_lock(&ioapic_mutex); + guard(mutex)(&ioapic_mutex); if (--data->count == 0) irq_domain_free_irqs(irq, 1); - mutex_unlock(&ioapic_mutex); } /* @@ -1146,12 +1069,10 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) { int irq, i, best_ioapic = -1, best_idx = -1; - apic_printk(APIC_DEBUG, - "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", - bus, slot, pin); + apic_pr_debug("Querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", + bus, slot, pin); if (test_bit(bus, mp_bus_not_pci)) { - apic_printk(APIC_VERBOSE, - "PCI BIOS passed nonexistent PCI bus %d!\n", bus); + apic_pr_verbose("PCI BIOS passed nonexistent PCI bus %d!\n", bus); return -1; } @@ -1196,8 +1117,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) return -1; out: - return pin_2_irq(best_idx, best_ioapic, mp_irqs[best_idx].dstirq, - IOAPIC_MAP_ALLOC); + return pin_2_irq(best_idx, best_ioapic, mp_irqs[best_idx].dstirq, IOAPIC_MAP_ALLOC); } EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector); @@ -1208,17 +1128,16 @@ static void __init setup_IO_APIC_irqs(void) unsigned int ioapic, pin; int idx; - apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n"); + apic_pr_verbose("Init IO_APIC IRQs\n"); for_each_ioapic_pin(ioapic, pin) { idx = find_irq_entry(ioapic, pin, mp_INT); - if (idx < 0) - apic_printk(APIC_VERBOSE, - KERN_DEBUG " apic %d pin %d not connected\n", - mpc_ioapic_id(ioapic), pin); - else - pin_2_irq(idx, ioapic, pin, - ioapic ? 0 : IOAPIC_MAP_ALLOC); + if (idx < 0) { + apic_pr_verbose("apic %d pin %d not connected\n", + mpc_ioapic_id(ioapic), pin); + } else { + pin_2_irq(idx, ioapic, pin, ioapic ? 0 : IOAPIC_MAP_ALLOC); + } } } @@ -1229,31 +1148,26 @@ void ioapic_zap_locks(void) static void io_apic_print_entries(unsigned int apic, unsigned int nr_entries) { - int i; - char buf[256]; struct IO_APIC_route_entry entry; - struct IR_IO_APIC_route_entry *ir_entry = (void *)&entry; + char buf[256]; + int i; - printk(KERN_DEBUG "IOAPIC %d:\n", apic); + apic_dbg("IOAPIC %d:\n", apic); for (i = 0; i <= nr_entries; i++) { entry = ioapic_read_entry(apic, i); - snprintf(buf, sizeof(buf), - " pin%02x, %s, %s, %s, V(%02X), IRR(%1d), S(%1d)", - i, - entry.mask == IOAPIC_MASKED ? "disabled" : "enabled ", - entry.trigger == IOAPIC_LEVEL ? "level" : "edge ", - entry.polarity == IOAPIC_POL_LOW ? "low " : "high", + snprintf(buf, sizeof(buf), " pin%02x, %s, %s, %s, V(%02X), IRR(%1d), S(%1d)", + i, entry.masked ? "disabled" : "enabled ", + entry.is_level ? "level" : "edge ", + entry.active_low ? "low " : "high", entry.vector, entry.irr, entry.delivery_status); - if (ir_entry->format) - printk(KERN_DEBUG "%s, remapped, I(%04X), Z(%X)\n", - buf, (ir_entry->index2 << 15) | ir_entry->index, - ir_entry->zero); - else - printk(KERN_DEBUG "%s, %s, D(%02X), M(%1d)\n", - buf, - entry.dest_mode == IOAPIC_DEST_MODE_LOGICAL ? - "logical " : "physical", - entry.dest, entry.delivery_mode); + if (entry.ir_format) { + apic_dbg("%s, remapped, I(%04X), Z(%X)\n", buf, + (entry.ir_index_15 << 15) | entry.ir_index_0_14, entry.ir_zero); + } else { + apic_dbg("%s, %s, D(%02X%02X), M(%1d)\n", buf, + entry.dest_mode_logical ? "logical " : "physical", + entry.virt_destid_8_14, entry.destid_0_7, entry.delivery_mode); + } } } @@ -1263,30 +1177,25 @@ static void __init print_IO_APIC(int ioapic_idx) union IO_APIC_reg_01 reg_01; union IO_APIC_reg_02 reg_02; union IO_APIC_reg_03 reg_03; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic_idx, 0); - reg_01.raw = io_apic_read(ioapic_idx, 1); - if (reg_01.bits.version >= 0x10) - reg_02.raw = io_apic_read(ioapic_idx, 2); - if (reg_01.bits.version >= 0x20) - reg_03.raw = io_apic_read(ioapic_idx, 3); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - - printk(KERN_DEBUG "IO APIC #%d......\n", mpc_ioapic_id(ioapic_idx)); - printk(KERN_DEBUG ".... register #00: %08X\n", reg_00.raw); - printk(KERN_DEBUG "....... : physical APIC id: %02X\n", reg_00.bits.ID); - printk(KERN_DEBUG "....... : Delivery Type: %X\n", reg_00.bits.delivery_type); - printk(KERN_DEBUG "....... : LTS : %X\n", reg_00.bits.LTS); - - printk(KERN_DEBUG ".... register #01: %08X\n", *(int *)®_01); - printk(KERN_DEBUG "....... : max redirection entries: %02X\n", - reg_01.bits.entries); - - printk(KERN_DEBUG "....... : PRQ implemented: %X\n", reg_01.bits.PRQ); - printk(KERN_DEBUG "....... : IO APIC version: %02X\n", - reg_01.bits.version); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) { + reg_00.raw = io_apic_read(ioapic_idx, 0); + reg_01.raw = io_apic_read(ioapic_idx, 1); + if (reg_01.bits.version >= 0x10) + reg_02.raw = io_apic_read(ioapic_idx, 2); + if (reg_01.bits.version >= 0x20) + reg_03.raw = io_apic_read(ioapic_idx, 3); + } + + apic_dbg("IO APIC #%d......\n", mpc_ioapic_id(ioapic_idx)); + apic_dbg(".... register #00: %08X\n", reg_00.raw); + apic_dbg("....... : physical APIC id: %02X\n", reg_00.bits.ID); + apic_dbg("....... : Delivery Type: %X\n", reg_00.bits.delivery_type); + apic_dbg("....... : LTS : %X\n", reg_00.bits.LTS); + apic_dbg(".... register #01: %08X\n", *(int *)®_01); + apic_dbg("....... : max redirection entries: %02X\n", reg_01.bits.entries); + apic_dbg("....... : PRQ implemented: %X\n", reg_01.bits.PRQ); + apic_dbg("....... : IO APIC version: %02X\n", reg_01.bits.version); /* * Some Intel chipsets with IO APIC VERSION of 0x1? don't have reg_02, @@ -1294,8 +1203,8 @@ static void __init print_IO_APIC(int ioapic_idx) * value, so ignore it if reg_02 == reg_01. */ if (reg_01.bits.version >= 0x10 && reg_02.raw != reg_01.raw) { - printk(KERN_DEBUG ".... register #02: %08X\n", reg_02.raw); - printk(KERN_DEBUG "....... : arbitration: %02X\n", reg_02.bits.arbitration); + apic_dbg(".... register #02: %08X\n", reg_02.raw); + apic_dbg("....... : arbitration: %02X\n", reg_02.bits.arbitration); } /* @@ -1305,11 +1214,11 @@ static void __init print_IO_APIC(int ioapic_idx) */ if (reg_01.bits.version >= 0x20 && reg_03.raw != reg_02.raw && reg_03.raw != reg_01.raw) { - printk(KERN_DEBUG ".... register #03: %08X\n", reg_03.raw); - printk(KERN_DEBUG "....... : Boot DT : %X\n", reg_03.bits.boot_DT); + apic_dbg(".... register #03: %08X\n", reg_03.raw); + apic_dbg("....... : Boot DT : %X\n", reg_03.bits.boot_DT); } - printk(KERN_DEBUG ".... IRQ redirection table:\n"); + apic_dbg(".... IRQ redirection table:\n"); io_apic_print_entries(ioapic_idx, reg_01.bits.entries); } @@ -1318,11 +1227,11 @@ void __init print_IO_APICs(void) int ioapic_idx; unsigned int irq; - printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries); - for_each_ioapic(ioapic_idx) - printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n", - mpc_ioapic_id(ioapic_idx), - ioapics[ioapic_idx].nr_registers); + apic_dbg("number of MP IRQ sources: %d.\n", mp_irq_entries); + for_each_ioapic(ioapic_idx) { + apic_dbg("number of IO-APIC #%d registers: %d.\n", + mpc_ioapic_id(ioapic_idx), ioapics[ioapic_idx].nr_registers); + } /* * We are a bit conservative about what we expect. We have to @@ -1333,7 +1242,7 @@ void __init print_IO_APICs(void) for_each_ioapic(ioapic_idx) print_IO_APIC(ioapic_idx); - printk(KERN_DEBUG "IRQ to pin mappings:\n"); + apic_dbg("IRQ to pin mappings:\n"); for_each_active_irq(irq) { struct irq_pin_list *entry; struct irq_chip *chip; @@ -1348,7 +1257,7 @@ void __init print_IO_APICs(void) if (list_empty(&data->irq_2_pin)) continue; - printk(KERN_DEBUG "IRQ%d ", irq); + apic_dbg("IRQ%d ", irq); for_each_irq_pin(entry, data->irq_2_pin) pr_cont("-> %d:%d", entry->apic, entry->pin); pr_cont("\n"); @@ -1362,10 +1271,9 @@ static struct { int pin, apic; } ioapic_i8259 = { -1, -1 }; void __init enable_IO_APIC(void) { - int i8259_apic, i8259_pin; - int apic, pin; + int i8259_apic, i8259_pin, apic, pin; - if (skip_ioapic_setup) + if (ioapic_is_disabled) nr_ioapics = 0; if (!nr_legacy_irqs() || !nr_ioapics) @@ -1375,18 +1283,21 @@ void __init enable_IO_APIC(void) /* See if any of the pins is in ExtINT mode */ struct IO_APIC_route_entry entry = ioapic_read_entry(apic, pin); - /* If the interrupt line is enabled and in ExtInt mode - * I have found the pin where the i8259 is connected. + /* + * If the interrupt line is enabled and in ExtInt mode I + * have found the pin where the i8259 is connected. */ - if ((entry.mask == 0) && (entry.delivery_mode == dest_ExtINT)) { + if (!entry.masked && entry.delivery_mode == APIC_DELIVERY_MODE_EXTINT) { ioapic_i8259.apic = apic; ioapic_i8259.pin = pin; - goto found_i8259; + break; } } - found_i8259: - /* Look to see what if the MP table has reported the ExtINT */ - /* If we could not find the appropriate pin by looking at the ioapic + + /* + * Look to see what if the MP table has reported the ExtINT + * + * If we could not find the appropriate pin by looking at the ioapic * the i8259 probably is not connected the ioapic but give the * mptable a chance anyway. */ @@ -1394,44 +1305,39 @@ void __init enable_IO_APIC(void) i8259_apic = find_isa_irq_apic(0, mp_ExtINT); /* Trust the MP table if nothing is setup in the hardware */ if ((ioapic_i8259.pin == -1) && (i8259_pin >= 0)) { - printk(KERN_WARNING "ExtINT not setup in hardware but reported by MP table\n"); + pr_warn("ExtINT not setup in hardware but reported by MP table\n"); ioapic_i8259.pin = i8259_pin; ioapic_i8259.apic = i8259_apic; } /* Complain if the MP table and the hardware disagree */ if (((ioapic_i8259.apic != i8259_apic) || (ioapic_i8259.pin != i8259_pin)) && - (i8259_pin >= 0) && (ioapic_i8259.pin >= 0)) - { - printk(KERN_WARNING "ExtINT in hardware and MP table differ\n"); - } + (i8259_pin >= 0) && (ioapic_i8259.pin >= 0)) + pr_warn("ExtINT in hardware and MP table differ\n"); - /* - * Do not trust the IO-APIC being empty at bootup - */ + /* Do not trust the IO-APIC being empty at bootup */ clear_IO_APIC(); } void native_restore_boot_irq_mode(void) { /* - * If the i8259 is routed through an IOAPIC - * Put that IOAPIC in virtual wire mode - * so legacy interrupts can be delivered. + * If the i8259 is routed through an IOAPIC Put that IOAPIC in + * virtual wire mode so legacy interrupts can be delivered. */ if (ioapic_i8259.pin != -1) { struct IO_APIC_route_entry entry; + u32 apic_id = read_apic_id(); memset(&entry, 0, sizeof(entry)); - entry.mask = IOAPIC_UNMASKED; - entry.trigger = IOAPIC_EDGE; - entry.polarity = IOAPIC_POL_HIGH; - entry.dest_mode = IOAPIC_DEST_MODE_PHYSICAL; - entry.delivery_mode = dest_ExtINT; - entry.dest = read_apic_id(); - - /* - * Add it to the IO-APIC irq-routing table: - */ + entry.masked = false; + entry.is_level = false; + entry.active_low = false; + entry.dest_mode_logical = false; + entry.delivery_mode = APIC_DELIVERY_MODE_EXTINT; + entry.destid_0_7 = apic_id & 0xFF; + entry.virt_destid_8_14 = apic_id >> 8; + + /* Add it to the IO-APIC irq-routing table */ ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry); } @@ -1454,37 +1360,34 @@ void restore_boot_irq_mode(void) * * by Matt Domsch <Matt_Domsch@dell.com> Tue Dec 21 12:25:05 CST 1999 */ -void __init setup_ioapic_ids_from_mpc_nocheck(void) +static void __init setup_ioapic_ids_from_mpc_nocheck(void) { + DECLARE_BITMAP(phys_id_present_map, MAX_LOCAL_APIC); + const u32 broadcast_id = 0xF; union IO_APIC_reg_00 reg_00; - physid_mask_t phys_id_present_map; - int ioapic_idx; - int i; unsigned char old_id; - unsigned long flags; + int ioapic_idx, i; /* * This is broken; anything with a real cpu count has to * circumvent this idiocy regardless. */ - apic->ioapic_phys_id_map(&phys_cpu_present_map, &phys_id_present_map); + copy_phys_cpu_present_map(phys_id_present_map); /* * Set the IOAPIC ID to the value stored in the MPC table. */ for_each_ioapic(ioapic_idx) { /* Read the register 0 value */ - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic_idx, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) + reg_00.raw = io_apic_read(ioapic_idx, 0); old_id = mpc_ioapic_id(ioapic_idx); - if (mpc_ioapic_id(ioapic_idx) >= get_physical_broadcast()) { - printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n", - ioapic_idx, mpc_ioapic_id(ioapic_idx)); - printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n", - reg_00.bits.ID); + if (mpc_ioapic_id(ioapic_idx) >= broadcast_id) { + pr_err(FW_BUG "IO-APIC#%d ID is %d in the MPC table!...\n", + ioapic_idx, mpc_ioapic_id(ioapic_idx)); + pr_err("... fixing up to %d. (tell your hw vendor)\n", reg_00.bits.ID); ioapics[ioapic_idx].mp_config.apicid = reg_00.bits.ID; } @@ -1493,65 +1396,54 @@ void __init setup_ioapic_ids_from_mpc_nocheck(void) * system must have a unique ID or we get lots of nice * 'stuck on smp_invalidate_needed IPI wait' messages. */ - if (apic->check_apicid_used(&phys_id_present_map, - mpc_ioapic_id(ioapic_idx))) { - printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n", - ioapic_idx, mpc_ioapic_id(ioapic_idx)); - for (i = 0; i < get_physical_broadcast(); i++) - if (!physid_isset(i, phys_id_present_map)) + if (test_bit(mpc_ioapic_id(ioapic_idx), phys_id_present_map)) { + pr_err(FW_BUG "IO-APIC#%d ID %d is already used!...\n", + ioapic_idx, mpc_ioapic_id(ioapic_idx)); + for (i = 0; i < broadcast_id; i++) + if (!test_bit(i, phys_id_present_map)) break; - if (i >= get_physical_broadcast()) + if (i >= broadcast_id) panic("Max APIC ID exceeded!\n"); - printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n", - i); - physid_set(i, phys_id_present_map); + pr_err("... fixing up to %d. (tell your hw vendor)\n", i); + set_bit(i, phys_id_present_map); ioapics[ioapic_idx].mp_config.apicid = i; } else { - physid_mask_t tmp; - apic->apicid_to_cpu_present(mpc_ioapic_id(ioapic_idx), - &tmp); - apic_printk(APIC_VERBOSE, "Setting %d in the " - "phys_id_present_map\n", + apic_pr_verbose("Setting %d in the phys_id_present_map\n", mpc_ioapic_id(ioapic_idx)); - physids_or(phys_id_present_map, phys_id_present_map, tmp); + set_bit(mpc_ioapic_id(ioapic_idx), phys_id_present_map); } /* - * We need to adjust the IRQ routing table - * if the ID changed. + * We need to adjust the IRQ routing table if the ID + * changed. */ - if (old_id != mpc_ioapic_id(ioapic_idx)) - for (i = 0; i < mp_irq_entries; i++) + if (old_id != mpc_ioapic_id(ioapic_idx)) { + for (i = 0; i < mp_irq_entries; i++) { if (mp_irqs[i].dstapic == old_id) - mp_irqs[i].dstapic - = mpc_ioapic_id(ioapic_idx); + mp_irqs[i].dstapic = mpc_ioapic_id(ioapic_idx); + } + } /* - * Update the ID register according to the right value - * from the MPC table if they are different. + * Update the ID register according to the right value from + * the MPC table if they are different. */ if (mpc_ioapic_id(ioapic_idx) == reg_00.bits.ID) continue; - apic_printk(APIC_VERBOSE, KERN_INFO - "...changing IO-APIC physical APIC ID to %d ...", - mpc_ioapic_id(ioapic_idx)); + apic_pr_verbose("...changing IO-APIC physical APIC ID to %d ...", + mpc_ioapic_id(ioapic_idx)); reg_00.bits.ID = mpc_ioapic_id(ioapic_idx); - raw_spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(ioapic_idx, 0, reg_00.raw); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - - /* - * Sanity check - */ - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic_idx, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) { + io_apic_write(ioapic_idx, 0, reg_00.raw); + reg_00.raw = io_apic_read(ioapic_idx, 0); + } + /* Sanity check */ if (reg_00.bits.ID != mpc_ioapic_id(ioapic_idx)) pr_cont("could not set ID!\n"); else - apic_printk(APIC_VERBOSE, " ok.\n"); + apic_pr_verbose(" ok.\n"); } } @@ -1594,10 +1486,9 @@ static void __init delay_with_tsc(void) * 1 GHz == 40 jiffies */ do { - rep_nop(); + native_pause(); now = rdtsc(); - } while ((now - start) < 40000000000ULL / HZ && - time_before_eq(jiffies, end)); + } while ((now - start) < 40000000000ULL / HZ && time_before_eq(jiffies, end)); } static void __init delay_without_tsc(void) @@ -1628,21 +1519,16 @@ static void __init delay_without_tsc(void) static int __init timer_irq_works(void) { unsigned long t1 = jiffies; - unsigned long flags; if (no_timer_check) return 1; - local_save_flags(flags); local_irq_enable(); - if (boot_cpu_has(X86_FEATURE_TSC)) delay_with_tsc(); else delay_without_tsc(); - local_irq_restore(flags); - /* * Expect a few ticks at least, to be sure some possible * glue logic does not lock up after one or two first @@ -1651,10 +1537,10 @@ static int __init timer_irq_works(void) * least one tick may be lost due to delays. */ - /* jiffies wrap? */ - if (time_after(jiffies, t1 + 4)) - return 1; - return 0; + local_irq_disable(); + + /* Did jiffies advance? */ + return time_after(jiffies, t1 + 4); } /* @@ -1663,36 +1549,29 @@ static int __init timer_irq_works(void) * so we 'resend' these IRQs via IPIs, to the same CPU. It's much * better to do it this way as thus we do not have to be aware of * 'pending' interrupts in the IRQ path, except at this point. - */ -/* - * Edge triggered needs to resend any interrupt - * that was delayed but this is now handled in the device - * independent code. - */ - -/* - * Starting up a edge-triggered IO-APIC interrupt is - * nasty - we need to make sure that we get the edge. - * If it is already asserted for some reason, we need - * return 1 to indicate that is was pending. * - * This is not complete - we should be able to fake - * an edge even if it isn't on the 8259A... + * + * Edge triggered needs to resend any interrupt that was delayed but this + * is now handled in the device independent code. + * + * Starting up a edge-triggered IO-APIC interrupt is nasty - we need to + * make sure that we get the edge. If it is already asserted for some + * reason, we need return 1 to indicate that is was pending. + * + * This is not complete - we should be able to fake an edge even if it + * isn't on the 8259A... */ static unsigned int startup_ioapic_irq(struct irq_data *data) { int was_pending = 0, irq = data->irq; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); if (irq < nr_legacy_irqs()) { legacy_pic->mask(irq); if (legacy_pic->irq_pending(irq)) was_pending = 1; } __unmask_ioapic(data->chip_data); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - return was_pending; } @@ -1702,40 +1581,37 @@ atomic_t irq_mis_count; static bool io_apic_level_ack_pending(struct mp_chip_data *data) { struct irq_pin_list *entry; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); for_each_irq_pin(entry, data->irq_2_pin) { - unsigned int reg; + struct IO_APIC_route_entry e; int pin; pin = entry->pin; - reg = io_apic_read(entry->apic, 0x10 + pin*2); + e.w1 = io_apic_read(entry->apic, 0x10 + pin*2); /* Is the remote IRR bit set? */ - if (reg & IO_APIC_REDIR_REMOTE_IRR) { - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + if (e.irr) return true; - } } - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - return false; } -static inline bool ioapic_irqd_mask(struct irq_data *data) +static inline bool ioapic_prepare_move(struct irq_data *data) { - /* If we are moving the irq we need to mask it */ + /* If we are moving the IRQ we need to mask it */ if (unlikely(irqd_is_setaffinity_pending(data))) { - mask_ioapic_irq(data); + if (!irqd_irq_masked(data)) + mask_ioapic_irq(data); return true; } return false; } -static inline void ioapic_irqd_unmask(struct irq_data *data, bool masked) +static inline void ioapic_finish_move(struct irq_data *data, bool moveit) { - if (unlikely(masked)) { - /* Only migrate the irq if the ack has been received. + if (unlikely(moveit)) { + /* + * Only migrate the irq if the ack has been received. * * On rare occasions the broadcast level triggered ack gets * delayed going to ioapics, and if we reprogram the @@ -1754,7 +1630,7 @@ static inline void ioapic_irqd_unmask(struct irq_data *data, bool masked) * with masking the ioapic entry and then polling until * Remote IRR was clear before reprogramming the * ioapic I don't trust the Remote IRR bit to be - * completey accurate. + * completely accurate. * * However there appears to be no other way to plug * this race, so if the Remote IRR bit is not @@ -1763,15 +1639,17 @@ static inline void ioapic_irqd_unmask(struct irq_data *data, bool masked) */ if (!io_apic_level_ack_pending(data->chip_data)) irq_move_masked_irq(data); - unmask_ioapic_irq(data); + /* If the IRQ is masked in the core, leave it: */ + if (!irqd_irq_masked(data)) + unmask_ioapic_irq(data); } } #else -static inline bool ioapic_irqd_mask(struct irq_data *data) +static inline bool ioapic_prepare_move(struct irq_data *data) { return false; } -static inline void ioapic_irqd_unmask(struct irq_data *data, bool masked) +static inline void ioapic_finish_move(struct irq_data *data, bool moveit) { } #endif @@ -1780,11 +1658,11 @@ static void ioapic_ack_level(struct irq_data *irq_data) { struct irq_cfg *cfg = irqd_cfg(irq_data); unsigned long v; - bool masked; + bool moveit; int i; irq_complete_move(cfg); - masked = ioapic_irqd_mask(irq_data); + moveit = ioapic_prepare_move(irq_data); /* * It appears there is an erratum which affects at least version 0x11 @@ -1825,12 +1703,12 @@ static void ioapic_ack_level(struct irq_data *irq_data) * We must acknowledge the irq before we move it or the acknowledge will * not propagate properly. */ - ack_APIC_irq(); + apic_eoi(); /* * Tail end of clearing remote IRR bit (either by delivering the EOI * message via io-apic EOI register write or simulating it using - * mask+edge followed by unnask+level logic) manually when the + * mask+edge followed by unmask+level logic) manually when the * level triggered interrupt is seen as the edge triggered interrupt * at the cpu. */ @@ -1839,7 +1717,7 @@ static void ioapic_ack_level(struct irq_data *irq_data) eoi_ioapic_pin(cfg->vector, irq_data->chip_data); } - ioapic_irqd_unmask(irq_data, masked); + ioapic_finish_move(irq_data, moveit); } static void ioapic_ir_ack_level(struct irq_data *irq_data) @@ -1856,41 +1734,123 @@ static void ioapic_ir_ack_level(struct irq_data *irq_data) eoi_ioapic_pin(data->entry.vector, data); } +/* + * The I/OAPIC is just a device for generating MSI messages from legacy + * interrupt pins. Various fields of the RTE translate into bits of the + * resulting MSI which had a historical meaning. + * + * With interrupt remapping, many of those bits have different meanings + * in the underlying MSI, but the way that the I/OAPIC transforms them + * from its RTE to the MSI message is the same. This function allows + * the parent IRQ domain to compose the MSI message, then takes the + * relevant bits to put them in the appropriate places in the RTE in + * order to generate that message when the IRQ happens. + * + * The setup here relies on a preconfigured route entry (is_level, + * active_low, masked) because the parent domain is merely composing the + * generic message routing information which is used for the MSI. + */ +static void ioapic_setup_msg_from_msi(struct irq_data *irq_data, + struct IO_APIC_route_entry *entry) +{ + struct msi_msg msg; + + /* Let the parent domain compose the MSI message */ + irq_chip_compose_msi_msg(irq_data, &msg); + + /* + * - Real vector + * - DMAR/IR: 8bit subhandle (ioapic.pin) + * - AMD/IR: 8bit IRTE index + */ + entry->vector = msg.arch_data.vector; + /* Delivery mode (for DMAR/IR all 0) */ + entry->delivery_mode = msg.arch_data.delivery_mode; + /* Destination mode or DMAR/IR index bit 15 */ + entry->dest_mode_logical = msg.arch_addr_lo.dest_mode_logical; + /* DMAR/IR: 1, 0 for all other modes */ + entry->ir_format = msg.arch_addr_lo.dmar_format; + /* + * - DMAR/IR: index bit 0-14. + * + * - Virt: If the host supports x2apic without a virtualized IR + * unit then bit 0-6 of dmar_index_0_14 are providing bit + * 8-14 of the destination id. + * + * All other modes have bit 0-6 of dmar_index_0_14 cleared and the + * topmost 8 bits are destination id bit 0-7 (entry::destid_0_7). + */ + entry->ir_index_0_14 = msg.arch_addr_lo.dmar_index_0_14; +} + static void ioapic_configure_entry(struct irq_data *irqd) { struct mp_chip_data *mpd = irqd->chip_data; - struct irq_cfg *cfg = irqd_cfg(irqd); struct irq_pin_list *entry; - /* - * Only update when the parent is the vector domain, don't touch it - * if the parent is the remapping domain. Check the installed - * ioapic chip to verify that. - */ - if (irqd->chip == &ioapic_chip) { - mpd->entry.dest = cfg->dest_apicid; - mpd->entry.vector = cfg->vector; - } + ioapic_setup_msg_from_msi(irqd, &mpd->entry); + for_each_irq_pin(entry, mpd->irq_2_pin) __ioapic_write_entry(entry->apic, entry->pin, mpd->entry); } -static int ioapic_set_affinity(struct irq_data *irq_data, - const struct cpumask *mask, bool force) +static int ioapic_set_affinity(struct irq_data *irq_data, const struct cpumask *mask, bool force) { struct irq_data *parent = irq_data->parent_data; - unsigned long flags; int ret; ret = parent->chip->irq_set_affinity(parent, mask, force); - raw_spin_lock_irqsave(&ioapic_lock, flags); + + guard(raw_spinlock_irqsave)(&ioapic_lock); if (ret >= 0 && ret != IRQ_SET_MASK_OK_DONE) ioapic_configure_entry(irq_data); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); return ret; } +/* + * Interrupt shutdown masks the ioapic pin, but the interrupt might already + * be in flight, but not yet serviced by the target CPU. That means + * __synchronize_hardirq() would return and claim that everything is calmed + * down. So free_irq() would proceed and deactivate the interrupt and free + * resources. + * + * Once the target CPU comes around to service it it will find a cleared + * vector and complain. While the spurious interrupt is harmless, the full + * release of resources might prevent the interrupt from being acknowledged + * which keeps the hardware in a weird state. + * + * Verify that the corresponding Remote-IRR bits are clear. + */ +static int ioapic_irq_get_chip_state(struct irq_data *irqd, enum irqchip_irq_state which, + bool *state) +{ + struct mp_chip_data *mcd = irqd->chip_data; + struct IO_APIC_route_entry rentry; + struct irq_pin_list *p; + + if (which != IRQCHIP_STATE_ACTIVE) + return -EINVAL; + + *state = false; + + guard(raw_spinlock)(&ioapic_lock); + for_each_irq_pin(p, mcd->irq_2_pin) { + rentry = __ioapic_read_entry(p->apic, p->pin); + /* + * The remote IRR is only valid in level trigger mode. It's + * meaning is undefined for edge triggered interrupts and + * irrelevant because the IO-APIC treats them as fire and + * forget. + */ + if (rentry.irr && rentry.is_level) { + *state = true; + break; + } + } + return 0; +} + static struct irq_chip ioapic_chip __read_mostly = { .name = "IO-APIC", .irq_startup = startup_ioapic_irq, @@ -1900,7 +1860,9 @@ static struct irq_chip ioapic_chip __read_mostly = { .irq_eoi = ioapic_ack_level, .irq_set_affinity = ioapic_set_affinity, .irq_retrigger = irq_chip_retrigger_hierarchy, - .flags = IRQCHIP_SKIP_SET_WAKE, + .irq_get_irqchip_state = ioapic_irq_get_chip_state, + .flags = IRQCHIP_SKIP_SET_WAKE | IRQCHIP_MOVE_DEFERRED | + IRQCHIP_AFFINITY_PRE_STARTUP, }; static struct irq_chip ioapic_ir_chip __read_mostly = { @@ -1912,7 +1874,9 @@ static struct irq_chip ioapic_ir_chip __read_mostly = { .irq_eoi = ioapic_ir_ack_level, .irq_set_affinity = ioapic_set_affinity, .irq_retrigger = irq_chip_retrigger_hierarchy, - .flags = IRQCHIP_SKIP_SET_WAKE, + .irq_get_irqchip_state = ioapic_irq_get_chip_state, + .flags = IRQCHIP_SKIP_SET_WAKE | + IRQCHIP_AFFINITY_PRE_STARTUP, }; static inline void init_IO_APIC_traps(void) @@ -1924,14 +1888,13 @@ static inline void init_IO_APIC_traps(void) cfg = irq_cfg(irq); if (IO_APIC_IRQ(irq) && cfg && !cfg->vector) { /* - * Hmm.. We don't have an entry for this, - * so default to an old-fashioned 8259 - * interrupt if we can.. + * Hmm.. We don't have an entry for this, so + * default to an old-fashioned 8259 interrupt if we + * can. Otherwise set the dummy interrupt chip. */ if (irq < nr_legacy_irqs()) legacy_pic->make_irq(irq); else - /* Strange. Oh, well.. */ irq_set_chip(irq, &no_irq_chip); } } @@ -1940,26 +1903,23 @@ static inline void init_IO_APIC_traps(void) /* * The local APIC irq-chip implementation: */ - static void mask_lapic_irq(struct irq_data *data) { - unsigned long v; + unsigned long v = apic_read(APIC_LVT0); - v = apic_read(APIC_LVT0); apic_write(APIC_LVT0, v | APIC_LVT_MASKED); } static void unmask_lapic_irq(struct irq_data *data) { - unsigned long v; + unsigned long v = apic_read(APIC_LVT0); - v = apic_read(APIC_LVT0); apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED); } static void ack_lapic_irq(struct irq_data *data) { - ack_APIC_irq(); + apic_eoi(); } static struct irq_chip lapic_chip __read_mostly = { @@ -1972,8 +1932,7 @@ static struct irq_chip lapic_chip __read_mostly = { static void lapic_register_intr(int irq) { irq_clear_status_flags(irq, IRQ_LEVEL); - irq_set_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq, - "edge"); + irq_set_chip_and_handler_name(irq, &lapic_chip, handle_edge_irq, "edge"); } /* @@ -1985,9 +1944,10 @@ static void lapic_register_intr(int irq) */ static inline void __init unlock_ExtINT_logic(void) { - int apic, pin, i; - struct IO_APIC_route_entry entry0, entry1; unsigned char save_control, save_freq_select; + struct IO_APIC_route_entry entry0, entry1; + int apic, pin, i; + u32 apic_id; pin = find_isa_irq_pin(8, mp_INT); if (pin == -1) { @@ -2003,14 +1963,16 @@ static inline void __init unlock_ExtINT_logic(void) entry0 = ioapic_read_entry(apic, pin); clear_IO_APIC_pin(apic, pin); + apic_id = read_apic_id(); memset(&entry1, 0, sizeof(entry1)); - entry1.dest_mode = IOAPIC_DEST_MODE_PHYSICAL; - entry1.mask = IOAPIC_UNMASKED; - entry1.dest = hard_smp_processor_id(); - entry1.delivery_mode = dest_ExtINT; - entry1.polarity = entry0.polarity; - entry1.trigger = IOAPIC_EDGE; + entry1.dest_mode_logical = true; + entry1.masked = false; + entry1.destid_0_7 = apic_id & 0xFF; + entry1.virt_destid_8_14 = apic_id >> 8; + entry1.delivery_mode = APIC_DELIVERY_MODE_EXTINT; + entry1.active_low = entry0.active_low; + entry1.is_level = false; entry1.vector = 0; ioapic_write_entry(apic, pin, entry1); @@ -2044,32 +2006,47 @@ static int __init disable_timer_pin_setup(char *arg) } early_param("disable_timer_pin_1", disable_timer_pin_setup); -static int mp_alloc_timer_irq(int ioapic, int pin) +static int __init mp_alloc_timer_irq(int ioapic, int pin) { - int irq = -1; struct irq_domain *domain = mp_ioapic_irqdomain(ioapic); + int irq = -1; if (domain) { struct irq_alloc_info info; ioapic_set_alloc_attr(&info, NUMA_NO_NODE, 0, 0); - info.ioapic_id = mpc_ioapic_id(ioapic); - info.ioapic_pin = pin; - mutex_lock(&ioapic_mutex); + info.devid = mpc_ioapic_id(ioapic); + info.ioapic.pin = pin; + guard(mutex)(&ioapic_mutex); irq = alloc_isa_irq_from_domain(domain, 0, ioapic, pin, &info); - mutex_unlock(&ioapic_mutex); } return irq; } +static void __init replace_pin_at_irq_node(struct mp_chip_data *data, int node, + int oldapic, int oldpin, + int newapic, int newpin) +{ + struct irq_pin_list *entry; + + for_each_irq_pin(entry, data->irq_2_pin) { + if (entry->apic == oldapic && entry->pin == oldpin) { + entry->apic = newapic; + entry->pin = newpin; + return; + } + } + + /* Old apic/pin didn't exist, so just add a new one */ + add_pin_to_irq_node(data, node, newapic, newpin); +} + /* * This code may look a bit paranoid, but it's supposed to cooperate with * a wide range of boards and BIOS bugs. Fortunately only the timer IRQ * is so screwy. Thanks to Brian Perkins for testing/hacking this beast * fanatically on his truly buggy board. - * - * FIXME: really need to revamp this for all platforms. */ static inline void __init check_timer(void) { @@ -2078,10 +2055,12 @@ static inline void __init check_timer(void) struct irq_cfg *cfg = irqd_cfg(irq_data); int node = cpu_to_node(0); int apic1, pin1, apic2, pin2; - unsigned long flags; int no_pin1 = 0; - local_irq_save(flags); + if (!global_clock_event) + return; + + local_irq_disable(); /* * get/set the timer IRQ vector: @@ -2105,9 +2084,8 @@ static inline void __init check_timer(void) pin2 = ioapic_i8259.pin; apic2 = ioapic_i8259.apic; - apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X " - "apic1=%d pin1=%d apic2=%d pin2=%d\n", - cfg->vector, apic1, pin1, apic2, pin2); + pr_info("..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", + cfg->vector, apic1, pin1, apic2, pin2); /* * Some BIOS writers are clueless and report the ExtINTA @@ -2117,7 +2095,7 @@ static inline void __init check_timer(void) * 8259A. */ if (pin1 == -1) { - panic_if_irq_remap("BIOS bug: timer not connected to IO-APIC"); + panic_if_irq_remap(FW_BUG "Timer not connected to IO-APIC"); pin1 = pin2; apic1 = apic2; no_pin1 = 1; @@ -2136,9 +2114,9 @@ static inline void __init check_timer(void) * so only need to unmask if it is level-trigger * do we really have level trigger timer? */ - int idx; - idx = find_irq_entry(apic1, pin1, mp_INT); - if (idx != -1 && irq_trigger(idx)) + int idx = find_irq_entry(apic1, pin1, mp_INT); + + if (idx != -1 && irq_is_level(idx)) unmask_ioapic_irq(irq_get_irq_data(0)); } irq_domain_deactivate_irq(irq_data); @@ -2149,16 +2127,12 @@ static inline void __init check_timer(void) goto out; } panic_if_irq_remap("timer doesn't work through Interrupt-remapped IO-APIC"); - local_irq_disable(); clear_IO_APIC_pin(apic1, pin1); if (!no_pin1) - apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: " - "8254 timer not connected to IO-APIC\n"); + pr_err("..MP-BIOS bug: 8254 timer not connected to IO-APIC\n"); - apic_printk(APIC_QUIET, KERN_INFO "...trying to set up timer " - "(IRQ0) through the 8259A ...\n"); - apic_printk(APIC_QUIET, KERN_INFO - "..... (found apic %d pin %d) ...\n", apic2, pin2); + pr_info("...trying to set up timer (IRQ0) through the 8259A ...\n"); + pr_info("..... (found apic %d pin %d) ...\n", apic2, pin2); /* * legacy devices should be connected to IO APIC #0 */ @@ -2167,57 +2141,54 @@ static inline void __init check_timer(void) irq_domain_activate_irq(irq_data, false); legacy_pic->unmask(0); if (timer_irq_works()) { - apic_printk(APIC_QUIET, KERN_INFO "....... works.\n"); + pr_info("....... works.\n"); goto out; } /* * Cleanup, just in case ... */ - local_irq_disable(); legacy_pic->mask(0); clear_IO_APIC_pin(apic2, pin2); - apic_printk(APIC_QUIET, KERN_INFO "....... failed.\n"); + pr_info("....... failed.\n"); } - apic_printk(APIC_QUIET, KERN_INFO - "...trying to set up timer as Virtual Wire IRQ...\n"); + pr_info("...trying to set up timer as Virtual Wire IRQ...\n"); lapic_register_intr(0); apic_write(APIC_LVT0, APIC_DM_FIXED | cfg->vector); /* Fixed mode */ legacy_pic->unmask(0); if (timer_irq_works()) { - apic_printk(APIC_QUIET, KERN_INFO "..... works.\n"); + pr_info("..... works.\n"); goto out; } - local_irq_disable(); legacy_pic->mask(0); apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | cfg->vector); - apic_printk(APIC_QUIET, KERN_INFO "..... failed.\n"); + pr_info("..... failed.\n"); - apic_printk(APIC_QUIET, KERN_INFO - "...trying to set up timer as ExtINT IRQ...\n"); + pr_info("...trying to set up timer as ExtINT IRQ...\n"); legacy_pic->init(0); legacy_pic->make_irq(0); apic_write(APIC_LVT0, APIC_DM_EXTINT); + legacy_pic->unmask(0); unlock_ExtINT_logic(); if (timer_irq_works()) { - apic_printk(APIC_QUIET, KERN_INFO "..... works.\n"); + pr_info("..... works.\n"); goto out; } - local_irq_disable(); - apic_printk(APIC_QUIET, KERN_INFO "..... failed :(.\n"); - if (apic_is_x2apic_enabled()) - apic_printk(APIC_QUIET, KERN_INFO - "Perhaps problem with the pre-enabled x2apic mode\n" - "Try booting with x2apic and interrupt-remapping disabled in the bios.\n"); + + pr_info("..... failed :\n"); + if (apic_is_x2apic_enabled()) { + pr_info("Perhaps problem with the pre-enabled x2apic mode\n" + "Try booting with x2apic and interrupt-remapping disabled in the bios.\n"); + } panic("IO-APIC + timer doesn't work! Boot with apic=debug and send a " "report. Then try booting with the 'noapic' option.\n"); out: - local_irq_restore(flags); + local_irq_enable(); } /* @@ -2241,60 +2212,61 @@ out: static int mp_irqdomain_create(int ioapic) { - struct irq_alloc_info info; - struct irq_domain *parent; + struct mp_ioapic_gsi *gsi_cfg = mp_ioapic_gsi_routing(ioapic); int hwirqs = mp_ioapic_pin_count(ioapic); struct ioapic *ip = &ioapics[ioapic]; struct ioapic_domain_cfg *cfg = &ip->irqdomain_cfg; - struct mp_ioapic_gsi *gsi_cfg = mp_ioapic_gsi_routing(ioapic); + struct irq_domain *parent; struct fwnode_handle *fn; - char *name = "IO-APIC"; + struct irq_fwspec fwspec; if (cfg->type == IOAPIC_DOMAIN_INVALID) return 0; - init_irq_alloc_info(&info, NULL); - info.type = X86_IRQ_ALLOC_TYPE_IOAPIC; - info.ioapic_id = mpc_ioapic_id(ioapic); - parent = irq_remapping_get_ir_irq_domain(&info); - if (!parent) - parent = x86_vector_domain; - else - name = "IO-APIC-IR"; - /* Handle device tree enumerated APICs proper */ if (cfg->dev) { - fn = of_node_to_fwnode(cfg->dev); + fn = of_fwnode_handle(cfg->dev); } else { - fn = irq_domain_alloc_named_id_fwnode(name, ioapic); + fn = irq_domain_alloc_named_id_fwnode("IO-APIC", mpc_ioapic_id(ioapic)); if (!fn) return -ENOMEM; } - ip->irqdomain = irq_domain_create_linear(fn, hwirqs, cfg->ops, - (void *)(long)ioapic); + fwspec.fwnode = fn; + fwspec.param_count = 1; + fwspec.param[0] = mpc_ioapic_id(ioapic); - /* Release fw handle if it was allocated above */ - if (!cfg->dev) - irq_domain_free_fwnode(fn); + parent = irq_find_matching_fwspec(&fwspec, DOMAIN_BUS_GENERIC_MSI); + if (!parent) { + if (!cfg->dev) + irq_domain_free_fwnode(fn); + return -ENODEV; + } - if (!ip->irqdomain) + ip->irqdomain = irq_domain_create_hierarchy(parent, 0, hwirqs, fn, cfg->ops, + (void *)(long)ioapic); + if (!ip->irqdomain) { + /* Release fw handle if it was allocated above */ + if (!cfg->dev) + irq_domain_free_fwnode(fn); return -ENOMEM; + } - ip->irqdomain->parent = parent; - - if (cfg->type == IOAPIC_DOMAIN_LEGACY || - cfg->type == IOAPIC_DOMAIN_STRICT) - ioapic_dynirq_base = max(ioapic_dynirq_base, - gsi_cfg->gsi_end + 1); + if (cfg->type == IOAPIC_DOMAIN_LEGACY || cfg->type == IOAPIC_DOMAIN_STRICT) + ioapic_dynirq_base = max(ioapic_dynirq_base, gsi_cfg->gsi_end + 1); return 0; } static void ioapic_destroy_irqdomain(int idx) { + struct ioapic_domain_cfg *cfg = &ioapics[idx].irqdomain_cfg; + struct fwnode_handle *fn = ioapics[idx].irqdomain->fwnode; + if (ioapics[idx].irqdomain) { irq_domain_remove(ioapics[idx].irqdomain); + if (!cfg->dev) + irq_domain_free_fwnode(fn); ioapics[idx].irqdomain = NULL; } } @@ -2303,18 +2275,16 @@ void __init setup_IO_APIC(void) { int ioapic; - if (skip_ioapic_setup || !nr_ioapics) + if (ioapic_is_disabled || !nr_ioapics) return; io_apic_irqs = nr_legacy_irqs() ? ~PIC_IRQS : ~0UL; - apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n"); + apic_pr_verbose("ENABLING IO-APIC IRQs\n"); for_each_ioapic(ioapic) BUG_ON(mp_irqdomain_create(ioapic)); - /* - * Set up IO-APIC IRQ routing. - */ + /* Set up IO-APIC IRQ routing. */ x86_init.mpparse.setup_ioapic_ids(); sync_Arb_IDs(); @@ -2328,19 +2298,22 @@ void __init setup_IO_APIC(void) static void resume_ioapic_id(int ioapic_idx) { - unsigned long flags; union IO_APIC_reg_00 reg_00; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); reg_00.raw = io_apic_read(ioapic_idx, 0); if (reg_00.bits.ID != mpc_ioapic_id(ioapic_idx)) { reg_00.bits.ID = mpc_ioapic_id(ioapic_idx); io_apic_write(ioapic_idx, 0, reg_00.raw); } - raw_spin_unlock_irqrestore(&ioapic_lock, flags); } -static void ioapic_resume(void) +static int ioapic_suspend(void *data) +{ + return save_ioapic_entries(); +} + +static void ioapic_resume(void *data) { int ioapic_idx; @@ -2350,14 +2323,18 @@ static void ioapic_resume(void) restore_ioapic_entries(); } -static struct syscore_ops ioapic_syscore_ops = { - .suspend = save_ioapic_entries, - .resume = ioapic_resume, +static const struct syscore_ops ioapic_syscore_ops = { + .suspend = ioapic_suspend, + .resume = ioapic_resume, +}; + +static struct syscore ioapic_syscore = { + .ops = &ioapic_syscore_ops, }; static int __init ioapic_init_ops(void) { - register_syscore_ops(&ioapic_syscore_ops); + register_syscore(&ioapic_syscore); return 0; } @@ -2367,118 +2344,104 @@ device_initcall(ioapic_init_ops); static int io_apic_get_redir_entries(int ioapic) { union IO_APIC_reg_01 reg_01; - unsigned long flags; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); reg_01.raw = io_apic_read(ioapic, 1); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); - /* The register returns the maximum index redir index - * supported, which is one less than the total number of redir - * entries. + /* + * The register returns the maximum index redir index supported, + * which is one less than the total number of redir entries. */ return reg_01.bits.entries + 1; } unsigned int arch_dynirq_lower_bound(unsigned int from) { + unsigned int ret; + /* * dmar_alloc_hwirq() may be called before setup_IO_APIC(), so use * gsi_top if ioapic_dynirq_base hasn't been initialized yet. */ - return ioapic_initialized ? ioapic_dynirq_base : gsi_top; + ret = ioapic_dynirq_base ? : gsi_top; + + /* + * For DT enabled machines ioapic_dynirq_base is irrelevant and + * always 0. gsi_top can be 0 if there is no IO/APIC registered. + * 0 is an invalid interrupt number for dynamic allocations. Return + * @from instead. + */ + return ret ? : from; } #ifdef CONFIG_X86_32 static int io_apic_get_unique_id(int ioapic, int apic_id) { + static DECLARE_BITMAP(apic_id_map, MAX_LOCAL_APIC); + const u32 broadcast_id = 0xF; union IO_APIC_reg_00 reg_00; - static physid_mask_t apic_id_map = PHYSID_MASK_NONE; - physid_mask_t tmp; - unsigned long flags; int i = 0; - /* - * The P4 platform supports up to 256 APIC IDs on two separate APIC - * buses (one for LAPICs, one for IOAPICs), where predecessors only - * supports up to 16 on one shared APIC bus. - * - * TBD: Expand LAPIC/IOAPIC support on P4-class systems to take full - * advantage of new APIC bus architecture. - */ + /* Initialize the ID map */ + if (bitmap_empty(apic_id_map, MAX_LOCAL_APIC)) + copy_phys_cpu_present_map(apic_id_map); - if (physids_empty(apic_id_map)) - apic->ioapic_phys_id_map(&phys_cpu_present_map, &apic_id_map); - - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) + reg_00.raw = io_apic_read(ioapic, 0); - if (apic_id >= get_physical_broadcast()) { - printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying " - "%d\n", ioapic, apic_id, reg_00.bits.ID); + if (apic_id >= broadcast_id) { + pr_warn("IOAPIC[%d]: Invalid apic_id %d, trying %d\n", + ioapic, apic_id, reg_00.bits.ID); apic_id = reg_00.bits.ID; } - /* - * Every APIC in a system must have a unique ID or we get lots of nice - * 'stuck on smp_invalidate_needed IPI wait' messages. - */ - if (apic->check_apicid_used(&apic_id_map, apic_id)) { - - for (i = 0; i < get_physical_broadcast(); i++) { - if (!apic->check_apicid_used(&apic_id_map, i)) + /* Every APIC in a system must have a unique ID */ + if (test_bit(apic_id, apic_id_map)) { + for (i = 0; i < broadcast_id; i++) { + if (!test_bit(i, apic_id_map)) break; } - if (i == get_physical_broadcast()) + if (i == broadcast_id) panic("Max apic_id exceeded!\n"); - printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, " - "trying %d\n", ioapic, apic_id, i); - + pr_warn("IOAPIC[%d]: apic_id %d already used, trying %d\n", ioapic, apic_id, i); apic_id = i; } - apic->apicid_to_cpu_present(apic_id, &tmp); - physids_or(apic_id_map, apic_id_map, tmp); + set_bit(apic_id, apic_id_map); if (reg_00.bits.ID != apic_id) { reg_00.bits.ID = apic_id; - raw_spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(ioapic, 0, reg_00.raw); - reg_00.raw = io_apic_read(ioapic, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) { + io_apic_write(ioapic, 0, reg_00.raw); + reg_00.raw = io_apic_read(ioapic, 0); + } /* Sanity check */ if (reg_00.bits.ID != apic_id) { - pr_err("IOAPIC[%d]: Unable to change apic_id!\n", - ioapic); + pr_err("IOAPIC[%d]: Unable to change apic_id!\n", ioapic); return -1; } } - apic_printk(APIC_VERBOSE, KERN_INFO - "IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id); + apic_pr_verbose("IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id); return apic_id; } static u8 io_apic_unique_id(int idx, u8 id) { - if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && - !APIC_XAPIC(boot_cpu_apic_version)) + if ((boot_cpu_data.x86_vendor == X86_VENDOR_INTEL) && !APIC_XAPIC(boot_cpu_apic_version)) return io_apic_get_unique_id(idx, id); - else - return id; + return id; } #else static u8 io_apic_unique_id(int idx, u8 id) { union IO_APIC_reg_00 reg_00; DECLARE_BITMAP(used, 256); - unsigned long flags; u8 new_id; int i; @@ -2494,26 +2457,23 @@ static u8 io_apic_unique_id(int idx, u8 id) * Read the current id from the ioapic and keep it if * available. */ - raw_spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(idx, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) + reg_00.raw = io_apic_read(idx, 0); + new_id = reg_00.bits.ID; if (!test_bit(new_id, used)) { - apic_printk(APIC_VERBOSE, KERN_INFO - "IOAPIC[%d]: Using reg apic_id %d instead of %d\n", - idx, new_id, id); + apic_pr_verbose("IOAPIC[%d]: Using reg apic_id %d instead of %d\n", + idx, new_id, id); return new_id; } - /* - * Get the next free id and write it to the ioapic. - */ + /* Get the next free id and write it to the ioapic. */ new_id = find_first_zero_bit(used, 256); reg_00.bits.ID = new_id; - raw_spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(idx, 0, reg_00.raw); - reg_00.raw = io_apic_read(idx, 0); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); + scoped_guard (raw_spinlock_irqsave, &ioapic_lock) { + io_apic_write(idx, 0, reg_00.raw); + reg_00.raw = io_apic_read(idx, 0); + } /* Sanity check */ BUG_ON(reg_00.bits.ID != new_id); @@ -2523,40 +2483,14 @@ static u8 io_apic_unique_id(int idx, u8 id) static int io_apic_get_version(int ioapic) { - union IO_APIC_reg_01 reg_01; - unsigned long flags; + union IO_APIC_reg_01 reg_01; - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); reg_01.raw = io_apic_read(ioapic, 1); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); return reg_01.bits.version; } -int acpi_get_override_irq(u32 gsi, int *trigger, int *polarity) -{ - int ioapic, pin, idx; - - if (skip_ioapic_setup) - return -1; - - ioapic = mp_find_ioapic(gsi); - if (ioapic < 0) - return -1; - - pin = mp_find_ioapic_pin(ioapic, gsi); - if (pin < 0) - return -1; - - idx = find_irq_entry(ioapic, pin, mp_INT); - if (idx < 0) - return -1; - - *trigger = irq_trigger(idx); - *polarity = irq_polarity(idx); - return 0; -} - /* * This function updates target affinity of IOAPIC interrupts to include * the CPUs which came online during SMP bringup. @@ -2567,8 +2501,8 @@ static struct resource *ioapic_resources; static struct resource * __init ioapic_setup_resources(void) { - unsigned long n; struct resource *res; + unsigned long n; char *mem; int i; @@ -2578,7 +2512,7 @@ static struct resource * __init ioapic_setup_resources(void) n = IOAPIC_RESOURCE_NAME_SIZE + sizeof(struct resource); n *= nr_ioapics; - mem = memblock_alloc(n, SMP_CACHE_BYTES); + mem = memblock_alloc_or_panic(n, SMP_CACHE_BYTES); res = (void *)mem; mem += sizeof(struct resource) * nr_ioapics; @@ -2596,6 +2530,24 @@ static struct resource * __init ioapic_setup_resources(void) return res; } +static void io_apic_set_fixmap(enum fixed_addresses idx, phys_addr_t phys) +{ + pgprot_t flags = FIXMAP_PAGE_NOCACHE; + + /* + * Ensure fixmaps for IO-APIC MMIO respect memory encryption pgprot + * bits, just like normal ioremap(): + */ + if (cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT)) { + if (x86_platform.hyper.is_private_mmio(phys)) + flags = pgprot_encrypted(flags); + else + flags = pgprot_decrypted(flags); + } + + __set_fixmap(idx, phys, flags); +} + void __init io_apic_init_mappings(void) { unsigned long ioapic_phys, idx = FIX_IO_APIC_BASE_0; @@ -2608,12 +2560,10 @@ void __init io_apic_init_mappings(void) ioapic_phys = mpc_ioapic_addr(i); #ifdef CONFIG_X86_32 if (!ioapic_phys) { - printk(KERN_ERR - "WARNING: bogus zero IO-APIC " - "address found in MPTABLE, " + pr_err("WARNING: bogus zero IO-APIC address found in MPTABLE, " "disabling IO/APIC support!\n"); smp_found_config = 0; - skip_ioapic_setup = 1; + ioapic_is_disabled = true; goto fake_ioapic_page; } #endif @@ -2621,14 +2571,13 @@ void __init io_apic_init_mappings(void) #ifdef CONFIG_X86_32 fake_ioapic_page: #endif - ioapic_phys = (unsigned long)memblock_alloc(PAGE_SIZE, + ioapic_phys = (unsigned long)memblock_alloc_or_panic(PAGE_SIZE, PAGE_SIZE); ioapic_phys = __pa(ioapic_phys); } - set_fixmap_nocache(idx, ioapic_phys); - apic_printk(APIC_VERBOSE, "mapped IOAPIC to %08lx (%08lx)\n", - __fix_to_virt(idx) + (ioapic_phys & ~PAGE_MASK), - ioapic_phys); + io_apic_set_fixmap(idx, ioapic_phys); + apic_pr_verbose("mapped IOAPIC to %08lx (%08lx)\n", + __fix_to_virt(idx) + (ioapic_phys & ~PAGE_MASK), ioapic_phys); idx++; ioapic_res->start = ioapic_phys; @@ -2639,13 +2588,12 @@ fake_ioapic_page: void __init ioapic_insert_resources(void) { - int i; struct resource *r = ioapic_resources; + int i; if (!r) { if (nr_ioapics > 0) - printk(KERN_ERR - "IO APIC resources couldn't be allocated.\n"); + pr_err("IO APIC resources couldn't be allocated.\n"); return; } @@ -2665,11 +2613,12 @@ int mp_find_ioapic(u32 gsi) /* Find the IOAPIC that manages this GSI. */ for_each_ioapic(i) { struct mp_ioapic_gsi *gsi_cfg = mp_ioapic_gsi_routing(i); + if (gsi >= gsi_cfg->gsi_base && gsi <= gsi_cfg->gsi_end) return i; } - printk(KERN_ERR "ERROR: Unable to locate IOAPIC for GSI %d\n", gsi); + pr_err("ERROR: Unable to locate IOAPIC for GSI %d\n", gsi); return -1; } @@ -2708,12 +2657,10 @@ static int bad_ioapic_register(int idx) static int find_free_ioapic_entry(void) { - int idx; - - for (idx = 0; idx < MAX_IO_APICS; idx++) + for (int idx = 0; idx < MAX_IO_APICS; idx++) { if (ioapics[idx].nr_registers == 0) return idx; - + } return MAX_IO_APICS; } @@ -2724,8 +2671,7 @@ static int find_free_ioapic_entry(void) * @gsi_base: base of GSI associated with the IOAPIC * @cfg: configuration information for the IOAPIC */ -int mp_register_ioapic(int id, u32 address, u32 gsi_base, - struct ioapic_domain_cfg *cfg) +int mp_register_ioapic(int id, u32 address, u32 gsi_base, struct ioapic_domain_cfg *cfg) { bool hotplug = !!ioapic_initialized; struct mp_ioapic_gsi *gsi_cfg; @@ -2736,12 +2682,13 @@ int mp_register_ioapic(int id, u32 address, u32 gsi_base, pr_warn("Bogus (zero) I/O APIC address found, skipping!\n"); return -EINVAL; } - for_each_ioapic(ioapic) + + for_each_ioapic(ioapic) { if (ioapics[ioapic].mp_config.apicaddr == address) { - pr_warn("address 0x%x conflicts with IOAPIC%d\n", - address, ioapic); + pr_warn("address 0x%x conflicts with IOAPIC%d\n", address, ioapic); return -EEXIST; } + } idx = find_free_ioapic_entry(); if (idx >= MAX_IO_APICS) { @@ -2754,7 +2701,7 @@ int mp_register_ioapic(int id, u32 address, u32 gsi_base, ioapics[idx].mp_config.flags = MPC_APIC_USABLE; ioapics[idx].mp_config.apicaddr = address; - set_fixmap_nocache(FIX_IO_APIC_BASE_0 + idx, address); + io_apic_set_fixmap(FIX_IO_APIC_BASE_0 + idx, address); if (bad_ioapic_register(idx)) { clear_fixmap(FIX_IO_APIC_BASE_0 + idx); return -ENODEV; @@ -2776,8 +2723,7 @@ int mp_register_ioapic(int id, u32 address, u32 gsi_base, (gsi_end >= gsi_cfg->gsi_base && gsi_end <= gsi_cfg->gsi_end)) { pr_warn("GSI range [%u-%u] for new IOAPIC conflicts with GSI[%u-%u]\n", - gsi_base, gsi_end, - gsi_cfg->gsi_base, gsi_cfg->gsi_end); + gsi_base, gsi_end, gsi_cfg->gsi_base, gsi_cfg->gsi_end); clear_fixmap(FIX_IO_APIC_BASE_0 + idx); return -ENOSPC; } @@ -2791,7 +2737,7 @@ int mp_register_ioapic(int id, u32 address, u32 gsi_base, /* * If mp_register_ioapic() is called during early boot stage when - * walking ACPI/SFI/DT tables, it's too early to create irqdomain, + * walking ACPI/DT tables, it's too early to create irqdomain, * we are still using bootmem allocator. So delay it to setup_IO_APIC(). */ if (hotplug) { @@ -2811,8 +2757,7 @@ int mp_register_ioapic(int id, u32 address, u32 gsi_base, ioapics[idx].nr_registers = entries; pr_info("IOAPIC[%d]: apic_id %d, version %d, address 0x%x, GSI %d-%d\n", - idx, mpc_ioapic_id(idx), - mpc_ioapic_ver(idx), mpc_ioapic_addr(idx), + idx, mpc_ioapic_id(idx), mpc_ioapic_ver(idx), mpc_ioapic_addr(idx), gsi_cfg->gsi_base, gsi_cfg->gsi_end); return 0; @@ -2823,11 +2768,13 @@ int mp_unregister_ioapic(u32 gsi_base) int ioapic, pin; int found = 0; - for_each_ioapic(ioapic) + for_each_ioapic(ioapic) { if (ioapics[ioapic].gsi_config.gsi_base == gsi_base) { found = 1; break; } + } + if (!found) { pr_warn("can't find IOAPIC for GSI %d\n", gsi_base); return -ENODEV; @@ -2841,8 +2788,7 @@ int mp_unregister_ioapic(u32 gsi_base) if (irq >= 0) { data = irq_get_chip_data(irq); if (data && data->count) { - pr_warn("pin%d on IOAPIC%d is still in use.\n", - pin, ioapic); + pr_warn("pin%d on IOAPIC%d is still in use.\n", pin, ioapic); return -EBUSY; } } @@ -2874,45 +2820,49 @@ int mp_ioapic_registered(u32 gsi_base) static void mp_irqdomain_get_attr(u32 gsi, struct mp_chip_data *data, struct irq_alloc_info *info) { - if (info && info->ioapic_valid) { - data->trigger = info->ioapic_trigger; - data->polarity = info->ioapic_polarity; - } else if (acpi_get_override_irq(gsi, &data->trigger, - &data->polarity) < 0) { + if (info && info->ioapic.valid) { + data->is_level = info->ioapic.is_level; + data->active_low = info->ioapic.active_low; + } else if (__acpi_get_override_irq(gsi, &data->is_level, &data->active_low) < 0) { /* PCI interrupts are always active low level triggered. */ - data->trigger = IOAPIC_LEVEL; - data->polarity = IOAPIC_POL_LOW; + data->is_level = true; + data->active_low = true; } } -static void mp_setup_entry(struct irq_cfg *cfg, struct mp_chip_data *data, - struct IO_APIC_route_entry *entry) +/* + * Configure the I/O-APIC specific fields in the routing entry. + * + * This is important to setup the I/O-APIC specific bits (is_level, + * active_low, masked) because the underlying parent domain will only + * provide the routing information and is oblivious of the I/O-APIC + * specific bits. + * + * The entry is just preconfigured at this point and not written into the + * RTE. This happens later during activation which will fill in the actual + * routing information. + */ +static void mp_preconfigure_entry(struct mp_chip_data *data) { + struct IO_APIC_route_entry *entry = &data->entry; + memset(entry, 0, sizeof(*entry)); - entry->delivery_mode = apic->irq_delivery_mode; - entry->dest_mode = apic->irq_dest_mode; - entry->dest = cfg->dest_apicid; - entry->vector = cfg->vector; - entry->trigger = data->trigger; - entry->polarity = data->polarity; + entry->is_level = data->is_level; + entry->active_low = data->active_low; /* * Mask level triggered irqs. Edge triggered irqs are masked * by the irq core code in case they fire. */ - if (data->trigger == IOAPIC_LEVEL) - entry->mask = IOAPIC_MASKED; - else - entry->mask = IOAPIC_UNMASKED; + entry->masked = data->is_level; } int mp_irqdomain_alloc(struct irq_domain *domain, unsigned int virq, unsigned int nr_irqs, void *arg) { - int ret, ioapic, pin; - struct irq_cfg *cfg; - struct irq_data *irq_data; - struct mp_chip_data *data; struct irq_alloc_info *info = arg; + struct mp_chip_data *data; + struct irq_data *irq_data; + int ret, ioapic, pin; unsigned long flags; if (!info || nr_irqs > 1) @@ -2922,45 +2872,47 @@ int mp_irqdomain_alloc(struct irq_domain *domain, unsigned int virq, return -EINVAL; ioapic = mp_irqdomain_ioapic_idx(domain); - pin = info->ioapic_pin; - if (irq_find_mapping(domain, (irq_hw_number_t)pin) > 0) + pin = info->ioapic.pin; + if (irq_resolve_mapping(domain, (irq_hw_number_t)pin)) return -EEXIST; data = kzalloc(sizeof(*data), GFP_KERNEL); if (!data) return -ENOMEM; - info->ioapic_entry = &data->entry; ret = irq_domain_alloc_irqs_parent(domain, virq, nr_irqs, info); - if (ret < 0) { - kfree(data); - return ret; - } + if (ret < 0) + goto free_data; INIT_LIST_HEAD(&data->irq_2_pin); - irq_data->hwirq = info->ioapic_pin; + irq_data->hwirq = info->ioapic.pin; irq_data->chip = (domain->parent == x86_vector_domain) ? &ioapic_chip : &ioapic_ir_chip; irq_data->chip_data = data; mp_irqdomain_get_attr(mp_pin_to_gsi(ioapic, pin), data, info); - cfg = irqd_cfg(irq_data); - add_pin_to_irq_node(data, ioapic_alloc_attr_node(info), ioapic, pin); + if (!add_pin_to_irq_node(data, ioapic_alloc_attr_node(info), ioapic, pin)) { + ret = -ENOMEM; + goto free_irqs; + } + + mp_preconfigure_entry(data); + mp_register_handler(virq, data->is_level); local_irq_save(flags); - if (info->ioapic_entry) - mp_setup_entry(cfg, data, info->ioapic_entry); - mp_register_handler(virq, data->trigger); if (virq < nr_legacy_irqs()) legacy_pic->mask(virq); local_irq_restore(flags); - apic_printk(APIC_VERBOSE, KERN_DEBUG - "IOAPIC[%d]: Set routing entry (%d-%d -> 0x%x -> IRQ %d Mode:%i Active:%i Dest:%d)\n", - ioapic, mpc_ioapic_id(ioapic), pin, cfg->vector, - virq, data->trigger, data->polarity, cfg->dest_apicid); - + apic_pr_verbose("IOAPIC[%d]: Preconfigured routing entry (%d-%d -> IRQ %d Level:%i ActiveLow:%i)\n", + ioapic, mpc_ioapic_id(ioapic), pin, virq, data->is_level, data->active_low); return 0; + +free_irqs: + irq_domain_free_irqs_parent(domain, virq, nr_irqs); +free_data: + kfree(data); + return ret; } void mp_irqdomain_free(struct irq_domain *domain, unsigned int virq, @@ -2973,22 +2925,17 @@ void mp_irqdomain_free(struct irq_domain *domain, unsigned int virq, irq_data = irq_domain_get_irq_data(domain, virq); if (irq_data && irq_data->chip_data) { data = irq_data->chip_data; - __remove_pin_from_irq(data, mp_irqdomain_ioapic_idx(domain), - (int)irq_data->hwirq); + __remove_pin_from_irq(data, mp_irqdomain_ioapic_idx(domain), (int)irq_data->hwirq); WARN_ON(!list_empty(&data->irq_2_pin)); kfree(irq_data->chip_data); } irq_domain_free_irqs_top(domain, virq, nr_irqs); } -int mp_irqdomain_activate(struct irq_domain *domain, - struct irq_data *irq_data, bool reserve) +int mp_irqdomain_activate(struct irq_domain *domain, struct irq_data *irq_data, bool reserve) { - unsigned long flags; - - raw_spin_lock_irqsave(&ioapic_lock, flags); + guard(raw_spinlock_irqsave)(&ioapic_lock); ioapic_configure_entry(irq_data); - raw_spin_unlock_irqrestore(&ioapic_lock, flags); return 0; } @@ -2996,8 +2943,7 @@ void mp_irqdomain_deactivate(struct irq_domain *domain, struct irq_data *irq_data) { /* It won't be called for IRQ with multiple IOAPIC pins associated */ - ioapic_mask_entry(mp_irqdomain_ioapic_idx(domain), - (int)irq_data->hwirq); + ioapic_mask_entry(mp_irqdomain_ioapic_idx(domain), (int)irq_data->hwirq); } int mp_irqdomain_ioapic_idx(struct irq_domain *domain) |