X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=arch%2Fx86_64%2Fkernel%2Fio_apic.c;h=d70a90316af31271b0df026040f72d0682972a98;hb=refs%2Fheads%2Fvserver;hp=718504a3274161d07feb4dd42449c447f42e7fa2;hpb=9213980e6a70d8473e0ffd4b39ab5b6caaba9ff5;p=linux-2.6.git diff --git a/arch/x86_64/kernel/io_apic.c b/arch/x86_64/kernel/io_apic.c index 718504a32..d70a90316 100644 --- a/arch/x86_64/kernel/io_apic.c +++ b/arch/x86_64/kernel/io_apic.c @@ -21,28 +21,49 @@ */ #include -#include #include #include #include #include -#include #include +#include #include #include +#include +#include +#include +#ifdef CONFIG_ACPI +#include +#endif #include #include #include #include +#include +#include +#include +#include +#include +#include + +static int assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result); + +#define __apicdebuginit __init int sis_apic_bug; /* not actually supported, dummy for compile */ -#undef APIC_LOCKUP_DEBUG +static int no_timer_check; + +static int disable_timer_pin_1 __initdata; + +int timer_over_8254 __initdata = 1; -#define APIC_LOCKUP_DEBUG +/* Where if anywhere is the i8259 connect in external int mode */ +static struct { int pin, apic; } ioapic_i8259 = { -1, -1 }; -static spinlock_t ioapic_lock = SPIN_LOCK_UNLOCKED; +static DEFINE_SPINLOCK(ioapic_lock); +DEFINE_SPINLOCK(vector_lock); /* * # of IRQ routing registers @@ -53,7 +74,7 @@ int nr_ioapic_registers[MAX_IO_APICS]; * Rough estimation of how many shared IRQs there are, can * be changed anytime. */ -#define MAX_PLUS_SHARED_IRQS NR_IRQS +#define MAX_PLUS_SHARED_IRQS NR_IRQ_VECTORS #define PIN_MAP_SIZE (MAX_PLUS_SHARED_IRQS + NR_IRQS) /* @@ -67,35 +88,50 @@ static struct irq_pin_list { short apic, pin, next; } irq_2_pin[PIN_MAP_SIZE]; -#ifdef CONFIG_PCI_USE_VECTOR -int vector_irq[NR_IRQS] = { [0 ... NR_IRQS -1] = -1}; -#define vector_to_irq(vector) \ - (platform_legacy_irq(vector) ? vector : vector_irq[vector]) -#else -#define vector_to_irq(vector) (vector) -#endif +struct io_apic { + unsigned int index; + unsigned int unused[3]; + unsigned int data; +}; + +static __attribute_const__ struct io_apic __iomem *io_apic_base(int idx) +{ + return (void __iomem *) __fix_to_virt(FIX_IO_APIC_BASE_0 + idx) + + (mp_ioapics[idx].mpc_apicaddr & ~PAGE_MASK); +} + +static inline unsigned int 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); +} + +static inline void io_apic_write(unsigned int apic, unsigned int reg, unsigned int value) +{ + struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(reg, &io_apic->index); + writel(value, &io_apic->data); +} /* - * The common case is 1:1 IRQ<->pin mappings. Sometimes there are - * 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. + * Re-write a value: to be used for read-modify-write + * cycles where the read already set up the index register. */ -static void __init add_pin_to_irq(unsigned int irq, int apic, int pin) +static inline void io_apic_modify(unsigned int apic, unsigned int value) { - static int first_free_entry = NR_IRQS; - struct irq_pin_list *entry = irq_2_pin + irq; - - while (entry->next) - entry = irq_2_pin + entry->next; + struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(value, &io_apic->data); +} - if (entry->pin != -1) { - entry->next = first_free_entry; - entry = irq_2_pin + entry->next; - if (++first_free_entry >= PIN_MAP_SIZE) - panic("io_apic.c: whoops"); - } - entry->apic = apic; - entry->pin = pin; +/* + * Synchronize the IO-APIC and the CPU by doing + * a dummy read from the IO-APIC + */ +static inline void io_apic_sync(unsigned int apic) +{ + struct io_apic __iomem *io_apic = io_apic_base(apic); + readl(&io_apic->data); } #define __DO_ACTION(R, ACTION, FINAL) \ @@ -104,6 +140,7 @@ static void __init add_pin_to_irq(unsigned int irq, int apic, int pin) int pin; \ struct irq_pin_list *entry = irq_2_pin + irq; \ \ + BUG_ON(irq >= NR_IRQS); \ for (;;) { \ unsigned int reg; \ pin = entry->pin; \ @@ -119,6 +156,141 @@ static void __init add_pin_to_irq(unsigned int irq, int apic, int pin) FINAL; \ } +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; + unsigned long flags; + spin_lock_irqsave(&ioapic_lock, flags); + eu.w1 = io_apic_read(apic, 0x10 + 2 * pin); + eu.w2 = io_apic_read(apic, 0x11 + 2 * pin); + spin_unlock_irqrestore(&ioapic_lock, flags); + return eu.entry; +} + +/* + * When we write a new IO APIC routing entry, we need to write the high + * word first! If the mask bit in the low word is clear, we will enable + * the interrupt, and we need to make sure the entry is fully populated + * before that happens. + */ +static void +__ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) +{ + union entry_union eu; + eu.entry = e; + io_apic_write(apic, 0x11 + 2*pin, eu.w2); + io_apic_write(apic, 0x10 + 2*pin, eu.w1); +} + +static void ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e) +{ + unsigned long flags; + spin_lock_irqsave(&ioapic_lock, flags); + __ioapic_write_entry(apic, pin, e); + spin_unlock_irqrestore(&ioapic_lock, flags); +} + +/* + * When we mask an IO APIC routing entry, we need to write the low + * word first, in order to set the mask bit before we change the + * high bits! + */ +static void ioapic_mask_entry(int apic, int pin) +{ + unsigned long flags; + union entry_union eu = { .entry.mask = 1 }; + + spin_lock_irqsave(&ioapic_lock, flags); + io_apic_write(apic, 0x10 + 2*pin, eu.w1); + io_apic_write(apic, 0x11 + 2*pin, eu.w2); + spin_unlock_irqrestore(&ioapic_lock, flags); +} + +#ifdef CONFIG_SMP +static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, u8 vector) +{ + int apic, pin; + struct irq_pin_list *entry = irq_2_pin + irq; + + BUG_ON(irq >= NR_IRQS); + for (;;) { + unsigned int reg; + apic = entry->apic; + pin = entry->pin; + if (pin == -1) + break; + io_apic_write(apic, 0x11 + pin*2, dest); + reg = io_apic_read(apic, 0x10 + pin*2); + reg &= ~0x000000ff; + reg |= vector; + io_apic_modify(apic, reg); + if (!entry->next) + break; + entry = irq_2_pin + entry->next; + } +} + +static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask) +{ + unsigned long flags; + unsigned int dest; + cpumask_t tmp; + int vector; + + cpus_and(tmp, mask, cpu_online_map); + if (cpus_empty(tmp)) + tmp = TARGET_CPUS; + + cpus_and(mask, tmp, CPU_MASK_ALL); + + vector = assign_irq_vector(irq, mask, &tmp); + if (vector < 0) + return; + + dest = cpu_mask_to_apicid(tmp); + + /* + * Only the high 8 bits are valid. + */ + dest = SET_APIC_LOGICAL_ID(dest); + + spin_lock_irqsave(&ioapic_lock, flags); + __target_IO_APIC_irq(irq, dest, vector); + set_native_irq_info(irq, mask); + spin_unlock_irqrestore(&ioapic_lock, flags); +} +#endif + +/* + * The common case is 1:1 IRQ<->pin mappings. Sometimes there are + * 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 void add_pin_to_irq(unsigned int irq, int apic, int pin) +{ + static int first_free_entry = NR_IRQS; + struct irq_pin_list *entry = irq_2_pin + irq; + + BUG_ON(irq >= NR_IRQS); + while (entry->next) + entry = irq_2_pin + entry->next; + + if (entry->pin != -1) { + entry->next = first_free_entry; + entry = irq_2_pin + entry->next; + if (++first_free_entry >= PIN_MAP_SIZE) + panic("io_apic.c: ran out of irq_2_pin entries!"); + } + entry->apic = apic; + entry->pin = pin; +} + + #define DO_ACTION(name,R,ACTION, FINAL) \ \ static void name##_IO_APIC_irq (unsigned int irq) \ @@ -128,10 +300,6 @@ DO_ACTION( __mask, 0, |= 0x00010000, io_apic_sync(entry->apic) ) /* mask = 1 */ DO_ACTION( __unmask, 0, &= 0xfffeffff, ) /* mask = 0 */ -DO_ACTION( __mask_and_edge, 0, = (reg & 0xffff7fff) | 0x00010000, ) - /* mask = 1, trigger = 0 */ -DO_ACTION( __unmask_and_level, 0, = (reg & 0xfffeffff) | 0x00008000, ) - /* mask = 0, trigger = 1 */ static void mask_IO_APIC_irq (unsigned int irq) { @@ -151,27 +319,18 @@ static void unmask_IO_APIC_irq (unsigned int irq) spin_unlock_irqrestore(&ioapic_lock, flags); } -void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) +static void clear_IO_APIC_pin(unsigned int apic, unsigned int pin) { struct IO_APIC_route_entry entry; - unsigned long flags; /* Check delivery_mode to be sure we're not clearing an SMI pin */ - spin_lock_irqsave(&ioapic_lock, flags); - *(((int*)&entry) + 0) = io_apic_read(apic, 0x10 + 2 * pin); - *(((int*)&entry) + 1) = io_apic_read(apic, 0x11 + 2 * pin); - spin_unlock_irqrestore(&ioapic_lock, flags); + entry = ioapic_read_entry(apic, pin); if (entry.delivery_mode == dest_SMI) return; /* * Disable it in the IO-APIC irq-routing table: */ - memset(&entry, 0, sizeof(entry)); - entry.mask = 1; - spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry) + 0)); - io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry) + 1)); - spin_unlock_irqrestore(&ioapic_lock, flags); + ioapic_mask_entry(apic, pin); } static void clear_IO_APIC (void) @@ -183,14 +342,6 @@ static void clear_IO_APIC (void) clear_IO_APIC_pin(apic, pin); } -/* - * support for broken MP BIOSs, enables hand-redirection of PIRQ0-7 to - * specific CPU-side IRQs. - */ - -#define MAX_PIRQS 8 -int pirq_entries [MAX_PIRQS]; -int pirqs_enabled; int skip_ioapic_setup; int ioapic_force; @@ -199,120 +350,37 @@ int ioapic_force; static int __init disable_ioapic_setup(char *str) { skip_ioapic_setup = 1; - return 1; + return 0; } +early_param("noapic", disable_ioapic_setup); -static int __init enable_ioapic_setup(char *str) +/* Actually the next is obsolete, but keep it for paranoid reasons -AK */ +static int __init disable_timer_pin_setup(char *arg) { - ioapic_force = 1; - skip_ioapic_setup = 0; + disable_timer_pin_1 = 1; return 1; } +__setup("disable_timer_pin_1", disable_timer_pin_setup); -__setup("noapic", disable_ioapic_setup); -__setup("apic", enable_ioapic_setup); - -#include -#include -#include - -/* Temporary Hack. Nvidia and VIA boards currently only work with IO-APIC - off. Check for an Nvidia or VIA PCI bridge and turn it off. - Use pci direct infrastructure because this runs before the PCI subsystem. - - Can be overwritten with "apic" - - And another hack to disable the IOMMU on VIA chipsets. - - Kludge-O-Rama. */ -void __init check_ioapic(void) -{ - int num,slot,func; - if (ioapic_force) - return; - - /* Poor man's PCI discovery */ - for (num = 0; num < 32; num++) { - for (slot = 0; slot < 32; slot++) { - for (func = 0; func < 8; func++) { - u32 class; - u32 vendor; - u8 type; - class = read_pci_config(num,slot,func, - PCI_CLASS_REVISION); - if (class == 0xffffffff) - break; - - if ((class >> 16) != PCI_CLASS_BRIDGE_PCI) - continue; - - vendor = read_pci_config(num, slot, func, - PCI_VENDOR_ID); - vendor &= 0xffff; - switch (vendor) { - case PCI_VENDOR_ID_VIA: -#ifdef CONFIG_GART_IOMMU - if ((end_pfn >= (0xffffffff>>PAGE_SHIFT) || - force_iommu) && - !iommu_aperture_allowed) { - printk(KERN_INFO - "Looks like a VIA chipset. Disabling IOMMU. Overwrite with \"iommu=allowed\"\n"); - iommu_aperture_disabled = 1; - } -#endif - return; - case PCI_VENDOR_ID_NVIDIA: -#ifndef CONFIG_SMP - printk(KERN_INFO - "PCI bridge %02x:%02x from %x found. Setting \"noapic\". Overwrite with \"apic\"\n", - num,slot,vendor); - skip_ioapic_setup = 1; -#endif - return; - } - - /* No multi-function device? */ - type = read_pci_config_byte(num,slot,func, - PCI_HEADER_TYPE); - if (!(type & 0x80)) - break; - } - } - } -} - -static int __init ioapic_pirq_setup(char *str) +static int __init setup_disable_8254_timer(char *s) { - int i, max; - int ints[MAX_PIRQS+1]; - - get_options(str, ARRAY_SIZE(ints), ints); - - for (i = 0; i < MAX_PIRQS; i++) - pirq_entries[i] = -1; - - pirqs_enabled = 1; - printk(KERN_INFO "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++) { - printk(KERN_DEBUG "... PIRQ%d -> IRQ %d\n", i, ints[i+1]); - /* - * PIRQs are mapped upside down, usually. - */ - pirq_entries[MAX_PIRQS-i-1] = ints[i+1]; - } + timer_over_8254 = -1; + return 1; +} +static int __init setup_enable_8254_timer(char *s) +{ + timer_over_8254 = 2; return 1; } -__setup("pirq=", ioapic_pirq_setup); +__setup("disable_8254_timer", setup_disable_8254_timer); +__setup("enable_8254_timer", setup_enable_8254_timer); + /* * Find the IRQ entry number of a certain pin. */ -static int __init find_irq_entry(int apic, int pin, int type) +static int find_irq_entry(int apic, int pin, int type) { int i; @@ -336,9 +404,7 @@ static int __init find_isa_irq_pin(int irq, int type) for (i = 0; i < mp_irq_entries; i++) { int lbus = mp_irqs[i].mpc_srcbus; - if ((mp_bus_id_to_type[lbus] == MP_BUS_ISA || - mp_bus_id_to_type[lbus] == MP_BUS_EISA || - mp_bus_id_to_type[lbus] == MP_BUS_MCA) && + if (test_bit(lbus, mp_bus_not_pci) && (mp_irqs[i].mpc_irqtype == type) && (mp_irqs[i].mpc_srcbusirq == irq)) @@ -347,6 +413,29 @@ static int __init find_isa_irq_pin(int irq, int type) return -1; } +static int __init find_isa_irq_apic(int irq, int type) +{ + int i; + + for (i = 0; i < mp_irq_entries; i++) { + int lbus = mp_irqs[i].mpc_srcbus; + + if (test_bit(lbus, mp_bus_not_pci) && + (mp_irqs[i].mpc_irqtype == type) && + (mp_irqs[i].mpc_srcbusirq == irq)) + break; + } + if (i < mp_irq_entries) { + int apic; + for(apic = 0; apic < nr_ioapics; apic++) { + if (mp_ioapics[apic].mpc_apicid == mp_irqs[i].mpc_dstapic) + return apic; + } + } + + return -1; +} + /* * Find a specific PCI IRQ entry. * Not an __init, possibly needed by modules @@ -357,10 +446,10 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) { int apic, i, best_guess = -1; - Dprintk("querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", + apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n", bus, slot, pin); if (mp_bus_id_to_pci_bus[bus] == -1) { - printk(KERN_WARNING "PCI BIOS passed nonexistent PCI bus %d!\n", bus); + apic_printk(APIC_VERBOSE, "PCI BIOS passed nonexistent PCI bus %d!\n", bus); return -1; } for (i = 0; i < mp_irq_entries; i++) { @@ -371,7 +460,7 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) mp_irqs[i].mpc_dstapic == MP_APIC_ALL) break; - if ((mp_bus_id_to_type[lbus] == MP_BUS_PCI) && + if (!test_bit(lbus, mp_bus_not_pci) && !mp_irqs[i].mpc_irqtype && (bus == lbus) && (slot == ((mp_irqs[i].mpc_srcbusirq >> 2) & 0x1f))) { @@ -390,30 +479,10 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) best_guess = irq; } } + BUG_ON(best_guess >= NR_IRQS); return best_guess; } -/* - * EISA Edge/Level control register, ELCR - */ -static int __init EISA_ELCR(unsigned int irq) -{ - if (irq < 16) { - unsigned int port = 0x4d0 + (irq >> 3); - return (inb(port) >> (irq & 7)) & 1; - } - printk(KERN_INFO "Broken MPtable reports ISA irq %d\n", irq); - return 0; -} - -/* 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].mpc_srcbusirq)) -#define default_EISA_polarity(idx) (0) - /* ISA interrupts are always polarity zero edge triggered, * when listed as conforming in the MP table. */ @@ -426,12 +495,6 @@ static int __init EISA_ELCR(unsigned int irq) #define default_PCI_trigger(idx) (1) #define default_PCI_polarity(idx) (1) -/* MCA interrupts are always polarity zero level triggered, - * when listed as conforming in the MP table. */ - -#define default_MCA_trigger(idx) (1) -#define default_MCA_polarity(idx) (0) - static int __init MPBIOS_polarity(int idx) { int bus = mp_irqs[idx].mpc_srcbus; @@ -443,38 +506,11 @@ static int __init MPBIOS_polarity(int idx) switch (mp_irqs[idx].mpc_irqflag & 3) { case 0: /* conforms, ie. bus-type dependent polarity */ - { - switch (mp_bus_id_to_type[bus]) - { - case MP_BUS_ISA: /* ISA pin */ - { - polarity = default_ISA_polarity(idx); - break; - } - case MP_BUS_EISA: /* EISA pin */ - { - polarity = default_EISA_polarity(idx); - break; - } - case MP_BUS_PCI: /* PCI pin */ - { - polarity = default_PCI_polarity(idx); - break; - } - case MP_BUS_MCA: /* MCA pin */ - { - polarity = default_MCA_polarity(idx); - break; - } - default: - { - printk(KERN_WARNING "broken BIOS!!\n"); - polarity = 1; - break; - } - } + if (test_bit(bus, mp_bus_not_pci)) + polarity = default_ISA_polarity(idx); + else + polarity = default_PCI_polarity(idx); break; - } case 1: /* high active */ { polarity = 0; @@ -501,7 +537,7 @@ static int __init MPBIOS_polarity(int idx) return polarity; } -static int __init MPBIOS_trigger(int idx) +static int MPBIOS_trigger(int idx) { int bus = mp_irqs[idx].mpc_srcbus; int trigger; @@ -512,38 +548,11 @@ static int __init MPBIOS_trigger(int idx) switch ((mp_irqs[idx].mpc_irqflag>>2) & 3) { case 0: /* conforms, ie. bus-type dependent */ - { - switch (mp_bus_id_to_type[bus]) - { - case MP_BUS_ISA: /* ISA pin */ - { - trigger = default_ISA_trigger(idx); - break; - } - case MP_BUS_EISA: /* EISA pin */ - { - trigger = default_EISA_trigger(idx); - break; - } - case MP_BUS_PCI: /* PCI pin */ - { - trigger = default_PCI_trigger(idx); - break; - } - case MP_BUS_MCA: /* MCA pin */ - { - trigger = default_MCA_trigger(idx); - break; - } - default: - { - printk(KERN_WARNING "broken BIOS!!\n"); - trigger = 1; - break; - } - } + if (test_bit(bus, mp_bus_not_pci)) + trigger = default_ISA_trigger(idx); + else + trigger = default_PCI_trigger(idx); break; - } case 1: /* edge */ { trigger = 0; @@ -591,48 +600,18 @@ static int pin_2_irq(int idx, int apic, int pin) if (mp_irqs[idx].mpc_dstirq != pin) printk(KERN_ERR "broken BIOS or MPTABLE parser, ayiee!!\n"); - switch (mp_bus_id_to_type[bus]) - { - case MP_BUS_ISA: /* ISA pin */ - case MP_BUS_EISA: - case MP_BUS_MCA: - { - irq = mp_irqs[idx].mpc_srcbusirq; - break; - } - case MP_BUS_PCI: /* PCI pin */ - { - /* - * PCI IRQs are mapped in order - */ - i = irq = 0; - while (i < apic) - irq += nr_ioapic_registers[i++]; - irq += pin; - break; - } - default: - { - printk(KERN_ERR "unknown bus type %d.\n",bus); - irq = 0; - break; - } - } - - /* - * 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]) { - printk(KERN_DEBUG "disabling PIRQ%d\n", pin-16); - } else { - irq = pirq_entries[pin-16]; - printk(KERN_DEBUG "using PIRQ%d -> IRQ %d\n", - pin-16, irq); - } - } + if (test_bit(bus, mp_bus_not_pci)) { + irq = mp_irqs[idx].mpc_srcbusirq; + } else { + /* + * PCI IRQs are mapped in order + */ + i = irq = 0; + while (i < apic) + irq += nr_ioapic_registers[i++]; + irq += pin; } + BUG_ON(irq >= NR_IRQS); return irq; } @@ -654,130 +633,276 @@ static inline int IO_APIC_irq_trigger(int irq) } /* irq_vectors is indexed by the sum of all RTEs in all I/O APICs. */ -u8 irq_vector[NR_IRQ_VECTORS] = { FIRST_DEVICE_VECTOR , 0 }; +static u8 irq_vector[NR_IRQ_VECTORS] __read_mostly = { + [0] = FIRST_EXTERNAL_VECTOR + 0, + [1] = FIRST_EXTERNAL_VECTOR + 1, + [2] = FIRST_EXTERNAL_VECTOR + 2, + [3] = FIRST_EXTERNAL_VECTOR + 3, + [4] = FIRST_EXTERNAL_VECTOR + 4, + [5] = FIRST_EXTERNAL_VECTOR + 5, + [6] = FIRST_EXTERNAL_VECTOR + 6, + [7] = FIRST_EXTERNAL_VECTOR + 7, + [8] = FIRST_EXTERNAL_VECTOR + 8, + [9] = FIRST_EXTERNAL_VECTOR + 9, + [10] = FIRST_EXTERNAL_VECTOR + 10, + [11] = FIRST_EXTERNAL_VECTOR + 11, + [12] = FIRST_EXTERNAL_VECTOR + 12, + [13] = FIRST_EXTERNAL_VECTOR + 13, + [14] = FIRST_EXTERNAL_VECTOR + 14, + [15] = FIRST_EXTERNAL_VECTOR + 15, +}; -#ifndef CONFIG_PCI_USE_VECTOR -int __init assign_irq_vector(int irq) +static cpumask_t irq_domain[NR_IRQ_VECTORS] __read_mostly = { + [0] = CPU_MASK_ALL, + [1] = CPU_MASK_ALL, + [2] = CPU_MASK_ALL, + [3] = CPU_MASK_ALL, + [4] = CPU_MASK_ALL, + [5] = CPU_MASK_ALL, + [6] = CPU_MASK_ALL, + [7] = CPU_MASK_ALL, + [8] = CPU_MASK_ALL, + [9] = CPU_MASK_ALL, + [10] = CPU_MASK_ALL, + [11] = CPU_MASK_ALL, + [12] = CPU_MASK_ALL, + [13] = CPU_MASK_ALL, + [14] = CPU_MASK_ALL, + [15] = CPU_MASK_ALL, +}; + +static int __assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result) { - static int current_vector = FIRST_DEVICE_VECTOR, offset = 0; - BUG_ON(irq >= NR_IRQ_VECTORS); - if (IO_APIC_VECTOR(irq) > 0) - return IO_APIC_VECTOR(irq); -next: - current_vector += 8; - if (current_vector == IA32_SYSCALL_VECTOR) - goto next; + /* + * NOTE! The local APIC isn't very good at handling + * multiple interrupts at the same interrupt level. + * As the interrupt level is determined by taking the + * vector number and shifting that right by 4, we + * want to spread these out a bit so that they don't + * all fall in the same interrupt level. + * + * Also, we've got to be careful not to trash gate + * 0x80, because int 0x80 is hm, kind of importantish. ;) + */ + static int current_vector = FIRST_DEVICE_VECTOR, current_offset = 0; + int old_vector = -1; + int cpu; - if (current_vector > FIRST_SYSTEM_VECTOR) { - offset++; - current_vector = FIRST_DEVICE_VECTOR + offset; + BUG_ON((unsigned)irq >= NR_IRQ_VECTORS); + + /* Only try and allocate irqs on cpus that are present */ + cpus_and(mask, mask, cpu_online_map); + + if (irq_vector[irq] > 0) + old_vector = irq_vector[irq]; + if (old_vector > 0) { + cpus_and(*result, irq_domain[irq], mask); + if (!cpus_empty(*result)) + return old_vector; } - if (current_vector == FIRST_SYSTEM_VECTOR) - panic("ran out of interrupt sources!"); + for_each_cpu_mask(cpu, mask) { + cpumask_t domain, new_mask; + int new_cpu; + int vector, offset; + + domain = vector_allocation_domain(cpu); + cpus_and(new_mask, domain, cpu_online_map); - IO_APIC_VECTOR(irq) = current_vector; - return current_vector; + vector = current_vector; + offset = current_offset; +next: + vector += 8; + if (vector >= FIRST_SYSTEM_VECTOR) { + /* If we run out of vectors on large boxen, must share them. */ + offset = (offset + 1) % 8; + vector = FIRST_DEVICE_VECTOR + offset; + } + if (unlikely(current_vector == vector)) + continue; + if (vector == IA32_SYSCALL_VECTOR) + goto next; + for_each_cpu_mask(new_cpu, new_mask) + if (per_cpu(vector_irq, new_cpu)[vector] != -1) + goto next; + /* Found one! */ + current_vector = vector; + current_offset = offset; + if (old_vector >= 0) { + cpumask_t old_mask; + int old_cpu; + cpus_and(old_mask, irq_domain[irq], cpu_online_map); + for_each_cpu_mask(old_cpu, old_mask) + per_cpu(vector_irq, old_cpu)[old_vector] = -1; + } + for_each_cpu_mask(new_cpu, new_mask) + per_cpu(vector_irq, new_cpu)[vector] = irq; + irq_vector[irq] = vector; + irq_domain[irq] = domain; + cpus_and(*result, domain, mask); + return vector; + } + return -ENOSPC; } -#endif + +static int assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result) +{ + int vector; + unsigned long flags; + + spin_lock_irqsave(&vector_lock, flags); + vector = __assign_irq_vector(irq, mask, result); + spin_unlock_irqrestore(&vector_lock, flags); + return vector; +} + +static void __clear_irq_vector(int irq) +{ + cpumask_t mask; + int cpu, vector; + + BUG_ON(!irq_vector[irq]); + + vector = irq_vector[irq]; + cpus_and(mask, irq_domain[irq], cpu_online_map); + for_each_cpu_mask(cpu, mask) + per_cpu(vector_irq, cpu)[vector] = -1; + + irq_vector[irq] = 0; + irq_domain[irq] = CPU_MASK_NONE; +} + +void __setup_vector_irq(int cpu) +{ + /* Initialize vector_irq on a new cpu */ + /* This function must be called with vector_lock held */ + int irq, vector; + + /* Mark the inuse vectors */ + for (irq = 0; irq < NR_IRQ_VECTORS; ++irq) { + if (!cpu_isset(cpu, irq_domain[irq])) + continue; + vector = irq_vector[irq]; + per_cpu(vector_irq, cpu)[vector] = irq; + } + /* Mark the free vectors */ + for (vector = 0; vector < NR_VECTORS; ++vector) { + irq = per_cpu(vector_irq, cpu)[vector]; + if (irq < 0) + continue; + if (!cpu_isset(cpu, irq_domain[irq])) + per_cpu(vector_irq, cpu)[vector] = -1; + } +} + extern void (*interrupt[NR_IRQS])(void); -static struct hw_interrupt_type ioapic_level_type; -static struct hw_interrupt_type ioapic_edge_type; + +static struct irq_chip ioapic_chip; #define IOAPIC_AUTO -1 #define IOAPIC_EDGE 0 #define IOAPIC_LEVEL 1 -static inline void ioapic_register_intr(int irq, int vector, unsigned long trigger) -{ - if (use_pci_vector() && !platform_legacy_irq(irq)) { - if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) || - trigger == IOAPIC_LEVEL) - irq_desc[vector].handler = &ioapic_level_type; - else - irq_desc[vector].handler = &ioapic_edge_type; - set_intr_gate(vector, interrupt[vector]); - } else { - if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) || - trigger == IOAPIC_LEVEL) - irq_desc[irq].handler = &ioapic_level_type; - else - irq_desc[irq].handler = &ioapic_edge_type; - set_intr_gate(vector, interrupt[irq]); +static void ioapic_register_intr(int irq, int vector, unsigned long trigger) +{ + if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) || + trigger == IOAPIC_LEVEL) + set_irq_chip_and_handler_name(irq, &ioapic_chip, + handle_fasteoi_irq, "fasteoi"); + else { + irq_desc[irq].status |= IRQ_DELAYED_DISABLE; + set_irq_chip_and_handler_name(irq, &ioapic_chip, + handle_edge_irq, "edge"); } } - -void __init setup_IO_APIC_irqs(void) +static void __init setup_IO_APIC_irq(int apic, int pin, int idx, int irq) { struct IO_APIC_route_entry entry; - int apic, pin, idx, irq, first_notcon = 1, vector; + int vector; unsigned long flags; - printk(KERN_DEBUG "init IO_APIC IRQs\n"); - for (apic = 0; apic < nr_ioapics; apic++) { - for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) { + /* + * add it to the IO-APIC irq-routing table: + */ + memset(&entry,0,sizeof(entry)); - /* - * add it to the IO-APIC irq-routing table: - */ - memset(&entry,0,sizeof(entry)); + entry.delivery_mode = INT_DELIVERY_MODE; + entry.dest_mode = INT_DEST_MODE; + entry.mask = 0; /* enable IRQ */ + entry.dest = cpu_mask_to_apicid(TARGET_CPUS); - entry.delivery_mode = dest_LowestPrio; - entry.dest_mode = INT_DELIVERY_MODE; - entry.mask = 0; /* enable IRQ */ - entry.dest.logical.logical_dest = TARGET_CPUS; + entry.trigger = irq_trigger(idx); + entry.polarity = irq_polarity(idx); + + if (irq_trigger(idx)) { + entry.trigger = 1; + entry.mask = 1; + entry.dest = cpu_mask_to_apicid(TARGET_CPUS); + } + + if (!apic && !IO_APIC_IRQ(irq)) + return; + + if (IO_APIC_IRQ(irq)) { + cpumask_t mask; + vector = assign_irq_vector(irq, TARGET_CPUS, &mask); + if (vector < 0) + return; + + entry.dest = cpu_mask_to_apicid(mask); + entry.vector = vector; + + ioapic_register_intr(irq, vector, IOAPIC_AUTO); + if (!apic && (irq < 16)) + disable_8259A_irq(irq); + } + + ioapic_write_entry(apic, pin, entry); + + spin_lock_irqsave(&ioapic_lock, flags); + set_native_irq_info(irq, TARGET_CPUS); + spin_unlock_irqrestore(&ioapic_lock, flags); + +} + +static void __init setup_IO_APIC_irqs(void) +{ + int apic, pin, idx, irq, first_notcon = 1; + + apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n"); + + for (apic = 0; apic < nr_ioapics; apic++) { + for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) { idx = find_irq_entry(apic,pin,mp_INT); if (idx == -1) { if (first_notcon) { - printk(KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mpc_apicid, pin); + apic_printk(APIC_VERBOSE, KERN_DEBUG " IO-APIC (apicid-pin) %d-%d", mp_ioapics[apic].mpc_apicid, pin); first_notcon = 0; } else - printk(", %d-%d", mp_ioapics[apic].mpc_apicid, pin); + apic_printk(APIC_VERBOSE, ", %d-%d", mp_ioapics[apic].mpc_apicid, pin); continue; } - entry.trigger = irq_trigger(idx); - entry.polarity = irq_polarity(idx); - - if (irq_trigger(idx)) { - entry.trigger = 1; - entry.mask = 1; - entry.dest.logical.logical_dest = TARGET_CPUS; - } - irq = pin_2_irq(idx, apic, pin); add_pin_to_irq(irq, apic, pin); - if (!apic && !IO_APIC_IRQ(irq)) - continue; - - if (IO_APIC_IRQ(irq)) { - vector = assign_irq_vector(irq); - entry.vector = vector; + setup_IO_APIC_irq(apic, pin, idx, irq); - ioapic_register_intr(irq, vector, IOAPIC_AUTO); - if (!apic && (irq < 16)) - disable_8259A_irq(irq); - } - spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(apic, 0x11+2*pin, *(((int *)&entry)+1)); - io_apic_write(apic, 0x10+2*pin, *(((int *)&entry)+0)); - spin_unlock_irqrestore(&ioapic_lock, flags); } } if (!first_notcon) - printk(" not connected.\n"); + apic_printk(APIC_VERBOSE," not connected.\n"); } /* * Set up the 8259A-master output pin as broadcast to all * CPUs. */ -void __init setup_ExtINT_IRQ0_pin(unsigned int pin, int vector) +static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, int vector) { struct IO_APIC_route_entry entry; unsigned long flags; @@ -787,16 +912,16 @@ void __init setup_ExtINT_IRQ0_pin(unsigned int pin, int vector) disable_8259A_irq(0); /* mask LVT0 */ - apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); + apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); /* * We use logical delivery to get the timer IRQ * to the first CPU. */ - entry.dest_mode = INT_DELIVERY_MODE; + entry.dest_mode = INT_DEST_MODE; entry.mask = 0; /* unmask IRQ now */ - entry.dest.logical.logical_dest = TARGET_CPUS; - entry.delivery_mode = dest_LowestPrio; + entry.dest = cpu_mask_to_apicid(TARGET_CPUS); + entry.delivery_mode = INT_DELIVERY_MODE; entry.polarity = 0; entry.trigger = 0; entry.vector = vector; @@ -805,14 +930,14 @@ void __init setup_ExtINT_IRQ0_pin(unsigned int pin, int vector) * The timer IRQ doesn't have to know that behind the * scene we have a 8259A-master in AEOI mode ... */ - irq_desc[0].handler = &ioapic_edge_type; + set_irq_chip_and_handler_name(0, &ioapic_chip, handle_edge_irq, "edge"); /* * Add it to the IO-APIC irq-routing table: */ spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(0, 0x11+2*pin, *(((int *)&entry)+1)); - io_apic_write(0, 0x10+2*pin, *(((int *)&entry)+0)); + io_apic_write(apic, 0x11+2*pin, *(((int *)&entry)+1)); + io_apic_write(apic, 0x10+2*pin, *(((int *)&entry)+0)); spin_unlock_irqrestore(&ioapic_lock, flags); enable_8259A_irq(0); @@ -820,13 +945,9 @@ void __init setup_ExtINT_IRQ0_pin(unsigned int pin, int vector) void __init UNEXPECTED_IO_APIC(void) { -#if 0 - printk(KERN_WARNING " WARNING: unexpected IO-APIC, please mail\n"); - printk(KERN_WARNING " to linux-smp@vger.kernel.org\n"); -#endif } -void __init print_IO_APIC(void) +void __apicdebuginit print_IO_APIC(void) { int apic, i; union IO_APIC_reg_00 reg_00; @@ -834,7 +955,10 @@ void __init print_IO_APIC(void) union IO_APIC_reg_02 reg_02; unsigned long flags; - printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries); + if (apic_verbosity == APIC_QUIET) + return; + + printk(KERN_DEBUG "number of MP IRQ sources: %d.\n", mp_irq_entries); for (i = 0; i < nr_ioapics; i++) printk(KERN_DEBUG "number of IO-APIC #%d registers: %d.\n", mp_ioapics[i].mpc_apicid, nr_ioapic_registers[i]); @@ -896,21 +1020,17 @@ void __init print_IO_APIC(void) printk(KERN_DEBUG ".... IRQ redirection table:\n"); - printk(KERN_DEBUG " NR Log Phy Mask Trig IRR Pol" - " Stat Dest Deli Vect: \n"); + printk(KERN_DEBUG " NR Dst Mask Trig IRR Pol" + " Stat Dmod Deli Vect: \n"); for (i = 0; i <= reg_01.bits.entries; i++) { struct IO_APIC_route_entry entry; - spin_lock_irqsave(&ioapic_lock, flags); - *(((int *)&entry)+0) = io_apic_read(apic, 0x10+i*2); - *(((int *)&entry)+1) = io_apic_read(apic, 0x11+i*2); - spin_unlock_irqrestore(&ioapic_lock, flags); + entry = ioapic_read_entry(apic, i); - printk(KERN_DEBUG " %02x %03X %02X ", + printk(KERN_DEBUG " %02x %03X ", i, - entry.dest.logical.logical_dest, - entry.dest.physical.physical_dest + entry.dest ); printk("%1d %1d %1d %1d %1d %1d %1d %02X\n", @@ -945,11 +1065,16 @@ void __init print_IO_APIC(void) return; } -static void print_APIC_bitfield (int base) +#if 0 + +static __apicdebuginit void print_APIC_bitfield (int base) { unsigned int v; int i, j; + if (apic_verbosity == APIC_QUIET) + return; + printk(KERN_DEBUG "0123456789abcdef0123456789abcdef\n" KERN_DEBUG); for (i = 0; i < 8; i++) { v = apic_read(base + i*0x10); @@ -963,10 +1088,13 @@ static void print_APIC_bitfield (int base) } } -void /*__init*/ print_local_APIC(void * dummy) +void __apicdebuginit print_local_APIC(void * dummy) { unsigned int v, ver, maxlvt; + if (apic_verbosity == APIC_QUIET) + return; + printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n", smp_processor_id(), hard_smp_processor_id()); v = apic_read(APIC_ID); @@ -979,13 +1107,11 @@ void /*__init*/ print_local_APIC(void * dummy) v = apic_read(APIC_TASKPRI); printk(KERN_DEBUG "... APIC TASKPRI: %08x (%02x)\n", v, v & APIC_TPRI_MASK); - if (APIC_INTEGRATED(ver)) { /* !82489DX */ - v = apic_read(APIC_ARBPRI); - printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v, - v & APIC_ARBPRI_MASK); - v = apic_read(APIC_PROCPRI); - printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v); - } + v = apic_read(APIC_ARBPRI); + printk(KERN_DEBUG "... APIC ARBPRI: %08x (%02x)\n", v, + v & APIC_ARBPRI_MASK); + v = apic_read(APIC_PROCPRI); + printk(KERN_DEBUG "... APIC PROCPRI: %08x\n", v); v = apic_read(APIC_EOI); printk(KERN_DEBUG "... APIC EOI: %08x\n", v); @@ -1005,12 +1131,8 @@ void /*__init*/ print_local_APIC(void * dummy) printk(KERN_DEBUG "... APIC IRR field:\n"); print_APIC_bitfield(APIC_IRR); - if (APIC_INTEGRATED(ver)) { /* !82489DX */ - if (maxlvt > 3) /* Due to the Pentium erratum 3AP. */ - apic_write(APIC_ESR, 0); - v = apic_read(APIC_ESR); - printk(KERN_DEBUG "... APIC ESR: %08x\n", v); - } + v = apic_read(APIC_ESR); + printk(KERN_DEBUG "... APIC ESR: %08x\n", v); v = apic_read(APIC_ICR); printk(KERN_DEBUG "... APIC ICR: %08x\n", v); @@ -1048,12 +1170,14 @@ void print_all_local_APICs (void) on_each_cpu(print_local_APIC, NULL, 1, 1); } -void /*__init*/ print_PIC(void) +void __apicdebuginit print_PIC(void) { - extern spinlock_t i8259A_lock; unsigned int v; unsigned long flags; + if (apic_verbosity == APIC_QUIET) + return; + printk(KERN_DEBUG "\nprinting PIC contents\n"); spin_lock_irqsave(&i8259A_lock, flags); @@ -1078,28 +1202,61 @@ void /*__init*/ print_PIC(void) printk(KERN_DEBUG "... PIC ELCR: %04x\n", v); } +#endif /* 0 */ + static void __init enable_IO_APIC(void) { union IO_APIC_reg_01 reg_01; - int i; + int i8259_apic, i8259_pin; + int i, apic; unsigned long flags; for (i = 0; i < PIN_MAP_SIZE; i++) { irq_2_pin[i].pin = -1; irq_2_pin[i].next = 0; } - if (!pirqs_enabled) - for (i = 0; i < MAX_PIRQS; i++) - pirq_entries[i] = -1; /* * The number of IO-APIC IRQ registers (== #pins): */ - for (i = 0; i < nr_ioapics; i++) { + for (apic = 0; apic < nr_ioapics; apic++) { spin_lock_irqsave(&ioapic_lock, flags); - reg_01.raw = io_apic_read(i, 1); + reg_01.raw = io_apic_read(apic, 1); spin_unlock_irqrestore(&ioapic_lock, flags); - nr_ioapic_registers[i] = reg_01.bits.entries+1; + nr_ioapic_registers[apic] = reg_01.bits.entries+1; + } + for(apic = 0; apic < nr_ioapics; apic++) { + int pin; + /* See if any of the pins is in ExtINT mode */ + for (pin = 0; pin < nr_ioapic_registers[apic]; pin++) { + struct IO_APIC_route_entry 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 ((entry.mask == 0) && (entry.delivery_mode == dest_ExtINT)) { + ioapic_i8259.apic = apic; + ioapic_i8259.pin = pin; + goto found_i8259; + } + } + } + found_i8259: + /* Look to see what if the MP table has reported the ExtINT */ + i8259_pin = find_isa_irq_pin(0, mp_ExtINT); + 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"); + 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"); } /* @@ -1112,108 +1269,38 @@ static void __init enable_IO_APIC(void) * Not an __init, needed by the reboot code */ void disable_IO_APIC(void) -{ - /* - * Clear the IO-APIC before rebooting: - */ - clear_IO_APIC(); - - disconnect_bsp_APIC(); -} - -/* - * function to set the IO-APIC physical IDs based on the - * values stored in the MPC table. - * - * by Matt Domsch Tue Dec 21 12:25:05 CST 1999 - */ - -static void __init setup_ioapic_ids_from_mpc (void) -{ - union IO_APIC_reg_00 reg_00; - physid_mask_t phys_id_present_map = phys_cpu_present_map; - int apic; - int i; - unsigned char old_id; - unsigned long flags; - - /* - * Set the IOAPIC ID to the value stored in the MPC table. - */ - for (apic = 0; apic < nr_ioapics; apic++) { - - /* Read the register 0 value */ - spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(apic, 0); - spin_unlock_irqrestore(&ioapic_lock, flags); - - old_id = mp_ioapics[apic].mpc_apicid; - - if (mp_ioapics[apic].mpc_apicid >= 0xf) { - printk(KERN_ERR "BIOS bug, IO-APIC#%d ID is %d in the MPC table!...\n", - apic, mp_ioapics[apic].mpc_apicid); - printk(KERN_ERR "... fixing up to %d. (tell your hw vendor)\n", - reg_00.bits.ID); - mp_ioapics[apic].mpc_apicid = reg_00.bits.ID; - } - - /* - * Sanity check, is the ID really free? 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 (physid_isset(mp_ioapics[apic].mpc_apicid, phys_id_present_map)) { - printk(KERN_ERR "BIOS bug, IO-APIC#%d ID %d is already used!...\n", - apic, mp_ioapics[apic].mpc_apicid); - for (i = 0; i < 0xf; i++) - if (!physid_isset(i, phys_id_present_map)) - break; - if (i >= 0xf) - 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); - mp_ioapics[apic].mpc_apicid = i; - } else { - printk(KERN_INFO - "Using IO-APIC %d\n", mp_ioapics[apic].mpc_apicid); - physid_set(mp_ioapics[apic].mpc_apicid, phys_id_present_map); - } - - - /* - * We need to adjust the IRQ routing table - * if the ID changed. - */ - if (old_id != mp_ioapics[apic].mpc_apicid) - for (i = 0; i < mp_irq_entries; i++) - if (mp_irqs[i].mpc_dstapic == old_id) - mp_irqs[i].mpc_dstapic - = mp_ioapics[apic].mpc_apicid; +{ + /* + * Clear the IO-APIC before rebooting: + */ + clear_IO_APIC(); - /* - * Read the right value from the MPC table and - * write it into the ID register. - */ - printk(KERN_INFO "...changing IO-APIC physical APIC ID to %d ...", - mp_ioapics[apic].mpc_apicid); + /* + * 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; - reg_00.bits.ID = mp_ioapics[apic].mpc_apicid; - spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(apic, 0, reg_00.raw); - spin_unlock_irqrestore(&ioapic_lock, flags); + memset(&entry, 0, sizeof(entry)); + entry.mask = 0; /* Enabled */ + entry.trigger = 0; /* Edge */ + entry.irr = 0; + entry.polarity = 0; /* High */ + entry.delivery_status = 0; + entry.dest_mode = 0; /* Physical */ + entry.delivery_mode = dest_ExtINT; /* ExtInt */ + entry.vector = 0; + entry.dest = GET_APIC_ID(apic_read(APIC_ID)); /* - * Sanity check + * Add it to the IO-APIC irq-routing table: */ - spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(apic, 0); - spin_unlock_irqrestore(&ioapic_lock, flags); - if (reg_00.bits.ID != mp_ioapics[apic].mpc_apicid) - panic("could not set ID!\n"); - else - printk(" ok.\n"); + ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry); } + + disconnect_bsp_APIC(ioapic_i8259.pin != -1); } /* @@ -1269,7 +1356,7 @@ static int __init timer_irq_works(void) * an edge even if it isn't on the 8259A... */ -static unsigned int startup_edge_ioapic_irq(unsigned int irq) +static unsigned int startup_ioapic_irq(unsigned int irq) { int was_pending = 0; unsigned long flags; @@ -1286,197 +1373,73 @@ static unsigned int startup_edge_ioapic_irq(unsigned int irq) return was_pending; } -/* - * Once we have recorded IRQ_PENDING already, we can mask the - * interrupt for real. This prevents IRQ storms from unhandled - * devices. - */ -static void ack_edge_ioapic_irq(unsigned int irq) +static int ioapic_retrigger_irq(unsigned int irq) { - if ((irq_desc[irq].status & (IRQ_PENDING | IRQ_DISABLED)) - == (IRQ_PENDING | IRQ_DISABLED)) - mask_IO_APIC_irq(irq); - ack_APIC_irq(); -} + cpumask_t mask; + unsigned vector; + unsigned long flags; -/* - * Level triggered interrupts can just be masked, - * and shutting down and starting up the interrupt - * is the same as enabling and disabling them -- except - * with a startup need to return a "was pending" value. - * - * Level triggered interrupts are special because we - * do not touch any IO-APIC register while handling - * them. We ack the APIC in the end-IRQ handler, not - * in the start-IRQ-handler. Protection against reentrance - * from the same interrupt is still provided, both by the - * generic IRQ layer and by the fact that an unacked local - * APIC does not accept IRQs. - */ -static unsigned int startup_level_ioapic_irq (unsigned int irq) -{ - unmask_IO_APIC_irq(irq); + spin_lock_irqsave(&vector_lock, flags); + vector = irq_vector[irq]; + cpus_clear(mask); + cpu_set(first_cpu(irq_domain[irq]), mask); - return 0; /* don't check for pending */ -} + send_IPI_mask(mask, vector); + spin_unlock_irqrestore(&vector_lock, flags); -static void end_level_ioapic_irq (unsigned int irq) -{ - unsigned long v; - int i; + return 1; +} /* - * It appears there is an erratum which affects at least version 0x11 - * of I/O APIC (that's the 82093AA and cores integrated into various - * chipsets). Under certain conditions a level-triggered interrupt is - * erroneously delivered as edge-triggered one but the respective IRR - * bit gets set nevertheless. As a result the I/O unit expects an EOI - * message but it will never arrive and further interrupts are blocked - * from the source. The exact reason is so far unknown, but the - * phenomenon was observed when two consecutive interrupt requests - * from a given source get delivered to the same CPU and the source is - * temporarily disabled in between. - * - * A workaround is to simulate an EOI message manually. We achieve it - * by setting the trigger mode to edge and then to level when the edge - * trigger mode gets detected in the TMR of a local APIC for a - * level-triggered interrupt. We mask the source for the time of the - * operation to prevent an edge-triggered interrupt escaping meanwhile. - * The idea is from Manfred Spraul. --macro + * Level and edge triggered IO-APIC interrupts need different handling, + * so we use two separate IRQ descriptors. Edge triggered IRQs can be + * handled with the level-triggered descriptor, but that one has slightly + * more overhead. Level-triggered interrupts cannot be handled with the + * edge-triggered handler, without risking IRQ storms and other ugly + * races. */ - i = IO_APIC_VECTOR(irq); - v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1)); +static void ack_apic_edge(unsigned int irq) +{ + move_native_irq(irq); ack_APIC_irq(); - - if (!(v & (1 << (i & 0x1f)))) { -#ifdef APIC_LOCKUP_DEBUG - struct irq_pin_list *entry; -#endif - -#ifdef APIC_MISMATCH_DEBUG - atomic_inc(&irq_mis_count); -#endif - spin_lock(&ioapic_lock); - __mask_and_edge_IO_APIC_irq(irq); -#ifdef APIC_LOCKUP_DEBUG - for (entry = irq_2_pin + irq;;) { - unsigned int reg; - - if (entry->pin == -1) - break; - reg = io_apic_read(entry->apic, 0x10 + entry->pin * 2); - if (reg & 0x00004000) - printk(KERN_CRIT "Aieee!!! Remote IRR" - " still set after unlock!\n"); - if (!entry->next) - break; - entry = irq_2_pin + entry->next; - } -#endif - __unmask_and_level_IO_APIC_irq(irq); - spin_unlock(&ioapic_lock); - } } -static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask) +static void ack_apic_level(unsigned int irq) { - unsigned long flags; - unsigned int dest; + int do_unmask_irq = 0; - dest = cpu_mask_to_apicid(mk_cpumask_const(mask)); +#if defined(CONFIG_GENERIC_PENDING_IRQ) || defined(CONFIG_IRQBALANCE) + /* If we are moving the irq we need to mask it */ + if (unlikely(irq_desc[irq].status & IRQ_MOVE_PENDING)) { + do_unmask_irq = 1; + mask_IO_APIC_irq(irq); + } +#endif /* - * Only the first 8 bits are valid. + * We must acknowledge the irq before we move it or the acknowledge will + * not propogate properly. */ - dest = dest << 24; - - spin_lock_irqsave(&ioapic_lock, flags); - __DO_ACTION(1, = dest, ) - spin_unlock_irqrestore(&ioapic_lock, flags); -} - -#ifdef CONFIG_PCI_USE_VECTOR -static unsigned int startup_edge_ioapic_vector(unsigned int vector) -{ - int irq = vector_to_irq(vector); - - return startup_edge_ioapic_irq(irq); -} - -static void ack_edge_ioapic_vector(unsigned int vector) -{ - int irq = vector_to_irq(vector); - - ack_edge_ioapic_irq(irq); -} - -static unsigned int startup_level_ioapic_vector (unsigned int vector) -{ - int irq = vector_to_irq(vector); - - return startup_level_ioapic_irq (irq); -} - -static void end_level_ioapic_vector (unsigned int vector) -{ - int irq = vector_to_irq(vector); - - end_level_ioapic_irq(irq); -} - -static void mask_IO_APIC_vector (unsigned int vector) -{ - int irq = vector_to_irq(vector); - - mask_IO_APIC_irq(irq); -} - -static void unmask_IO_APIC_vector (unsigned int vector) -{ - int irq = vector_to_irq(vector); + ack_APIC_irq(); - unmask_IO_APIC_irq(irq); + /* Now we can move and renable the irq */ + move_masked_irq(irq); + if (unlikely(do_unmask_irq)) + unmask_IO_APIC_irq(irq); } -static void set_ioapic_affinity_vector (unsigned int vector, - cpumask_t cpu_mask) -{ - int irq = vector_to_irq(vector); - - set_ioapic_affinity_irq(irq, cpu_mask); -} +static struct irq_chip ioapic_chip __read_mostly = { + .name = "IO-APIC", + .startup = startup_ioapic_irq, + .mask = mask_IO_APIC_irq, + .unmask = unmask_IO_APIC_irq, + .ack = ack_apic_edge, + .eoi = ack_apic_level, +#ifdef CONFIG_SMP + .set_affinity = set_ioapic_affinity_irq, #endif - -/* - * Level and edge triggered IO-APIC interrupts need different handling, - * so we use two separate IRQ descriptors. Edge triggered IRQs can be - * handled with the level-triggered descriptor, but that one has slightly - * more overhead. Level-triggered interrupts cannot be handled with the - * edge-triggered handler, without risking IRQ storms and other ugly - * races. - */ - -static struct hw_interrupt_type ioapic_edge_type = { - .typename = "IO-APIC-edge", - .startup = startup_edge_ioapic, - .shutdown = shutdown_edge_ioapic, - .enable = enable_edge_ioapic, - .disable = disable_edge_ioapic, - .ack = ack_edge_ioapic, - .end = end_edge_ioapic, - .set_affinity = set_ioapic_affinity, -}; - -static struct hw_interrupt_type ioapic_level_type = { - .typename = "IO-APIC-level", - .startup = startup_level_ioapic, - .shutdown = shutdown_level_ioapic, - .enable = enable_level_ioapic, - .disable = disable_level_ioapic, - .ack = mask_and_ack_level_ioapic, - .end = end_level_ioapic, - .set_affinity = set_ioapic_affinity, + .retrigger = ioapic_retrigger_irq, }; static inline void init_IO_APIC_traps(void) @@ -1496,12 +1459,7 @@ static inline void init_IO_APIC_traps(void) */ for (irq = 0; irq < NR_IRQS ; irq++) { int tmp = irq; - if (use_pci_vector()) { - if (!platform_legacy_irq(tmp)) - if ((tmp = vector_to_irq(tmp)) == -1) - continue; - } - if (IO_APIC_IRQ(tmp) && !IO_APIC_VECTOR(tmp)) { + if (IO_APIC_IRQ(tmp) && !irq_vector[tmp]) { /* * Hmm.. We don't have an entry for this, * so default to an old-fashioned 8259 @@ -1511,7 +1469,7 @@ static inline void init_IO_APIC_traps(void) make_8259A_irq(irq); else /* Strange. Oh, well.. */ - irq_desc[irq].handler = &no_irq_type; + irq_desc[irq].chip = &no_irq_chip; } } } @@ -1521,7 +1479,7 @@ static void enable_lapic_irq (unsigned int irq) unsigned long v; v = apic_read(APIC_LVT0); - apic_write_around(APIC_LVT0, v & ~APIC_LVT_MASKED); + apic_write(APIC_LVT0, v & ~APIC_LVT_MASKED); } static void disable_lapic_irq (unsigned int irq) @@ -1529,7 +1487,7 @@ static void disable_lapic_irq (unsigned int irq) unsigned long v; v = apic_read(APIC_LVT0); - apic_write_around(APIC_LVT0, v | APIC_LVT_MASKED); + apic_write(APIC_LVT0, v | APIC_LVT_MASKED); } static void ack_lapic_irq (unsigned int irq) @@ -1539,7 +1497,7 @@ static void ack_lapic_irq (unsigned int irq) static void end_lapic_irq (unsigned int i) { /* nothing */ } -static struct hw_interrupt_type lapic_irq_type = { +static struct hw_interrupt_type lapic_irq_type __read_mostly = { .typename = "local-APIC-edge", .startup = NULL, /* startup_irq() not used for IRQ0 */ .shutdown = NULL, /* shutdown_irq() not used for IRQ0 */ @@ -1576,34 +1534,35 @@ static void setup_nmi (void) */ static inline void unlock_ExtINT_logic(void) { - int pin, i; + int apic, pin, i; struct IO_APIC_route_entry entry0, entry1; unsigned char save_control, save_freq_select; unsigned long flags; - pin = find_isa_irq_pin(8, mp_INT); + pin = find_isa_irq_pin(8, mp_INT); + apic = find_isa_irq_apic(8, mp_INT); if (pin == -1) return; spin_lock_irqsave(&ioapic_lock, flags); - *(((int *)&entry0) + 1) = io_apic_read(0, 0x11 + 2 * pin); - *(((int *)&entry0) + 0) = io_apic_read(0, 0x10 + 2 * pin); + *(((int *)&entry0) + 1) = io_apic_read(apic, 0x11 + 2 * pin); + *(((int *)&entry0) + 0) = io_apic_read(apic, 0x10 + 2 * pin); spin_unlock_irqrestore(&ioapic_lock, flags); - clear_IO_APIC_pin(0, pin); + clear_IO_APIC_pin(apic, pin); memset(&entry1, 0, sizeof(entry1)); entry1.dest_mode = 0; /* physical delivery */ entry1.mask = 0; /* unmask IRQ now */ - entry1.dest.physical.physical_dest = hard_smp_processor_id(); + entry1.dest = hard_smp_processor_id(); entry1.delivery_mode = dest_ExtINT; entry1.polarity = entry0.polarity; entry1.trigger = 0; entry1.vector = 0; spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(0, 0x11 + 2 * pin, *(((int *)&entry1) + 1)); - io_apic_write(0, 0x10 + 2 * pin, *(((int *)&entry1) + 0)); + io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry1) + 1)); + io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry1) + 0)); spin_unlock_irqrestore(&ioapic_lock, flags); save_control = CMOS_READ(RTC_CONTROL); @@ -1621,11 +1580,11 @@ static inline void unlock_ExtINT_logic(void) CMOS_WRITE(save_control, RTC_CONTROL); CMOS_WRITE(save_freq_select, RTC_FREQ_SELECT); - clear_IO_APIC_pin(0, pin); + clear_IO_APIC_pin(apic, pin); spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(0, 0x11 + 2 * pin, *(((int *)&entry0) + 1)); - io_apic_write(0, 0x10 + 2 * pin, *(((int *)&entry0) + 0)); + io_apic_write(apic, 0x11 + 2 * pin, *(((int *)&entry0) + 1)); + io_apic_write(apic, 0x10 + 2 * pin, *(((int *)&entry0) + 0)); spin_unlock_irqrestore(&ioapic_lock, flags); } @@ -1634,18 +1593,20 @@ static inline void unlock_ExtINT_logic(void) * 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 modern platforms only. */ static inline void check_timer(void) { - int pin1, pin2; + int apic1, pin1, apic2, pin2; int vector; + cpumask_t mask; /* * get/set the timer IRQ vector: */ disable_8259A_irq(0); - vector = assign_irq_vector(0); - set_intr_gate(vector, interrupt[0]); + vector = assign_irq_vector(0, TARGET_CPUS, &mask); /* * Subtle, code in do_timer_interrupt() expects an AEOI @@ -1654,91 +1615,105 @@ static inline void check_timer(void) * the 8259A which implies the virtual wire has to be * disabled in the local APIC. */ - apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); + apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_EXTINT); init_8259A(1); - enable_8259A_irq(0); + if (timer_over_8254 > 0) + enable_8259A_irq(0); - pin1 = find_isa_irq_pin(0, mp_INT); - pin2 = find_isa_irq_pin(0, mp_ExtINT); + pin1 = find_isa_irq_pin(0, mp_INT); + apic1 = find_isa_irq_apic(0, mp_INT); + pin2 = ioapic_i8259.pin; + apic2 = ioapic_i8259.apic; - printk(KERN_INFO "..TIMER: vector=0x%02X pin1=%d pin2=%d\n", vector, pin1, pin2); + apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", + vector, apic1, pin1, apic2, pin2); if (pin1 != -1) { /* * Ok, does IRQ0 through the IOAPIC work? */ unmask_IO_APIC_irq(0); - if (timer_irq_works()) { + if (!no_timer_check && timer_irq_works()) { nmi_watchdog_default(); if (nmi_watchdog == NMI_IO_APIC) { disable_8259A_irq(0); setup_nmi(); enable_8259A_irq(0); - check_nmi_watchdog(); } + if (disable_timer_pin_1 > 0) + clear_IO_APIC_pin(0, pin1); return; } - clear_IO_APIC_pin(0, pin1); - printk(KERN_ERR "..MP-BIOS bug: 8254 timer not connected to IO-APIC\n"); + clear_IO_APIC_pin(apic1, pin1); + apic_printk(APIC_QUIET,KERN_ERR "..MP-BIOS bug: 8254 timer not " + "connected to IO-APIC\n"); } - printk(KERN_INFO "...trying to set up timer (IRQ0) through the 8259A ... "); + apic_printk(APIC_VERBOSE,KERN_INFO "...trying to set up timer (IRQ0) " + "through the 8259A ... "); if (pin2 != -1) { - printk("\n..... (found pin %d) ...", pin2); + apic_printk(APIC_VERBOSE,"\n..... (found apic %d pin %d) ...", + apic2, pin2); /* * legacy devices should be connected to IO APIC #0 */ - setup_ExtINT_IRQ0_pin(pin2, vector); + setup_ExtINT_IRQ0_pin(apic2, pin2, vector); if (timer_irq_works()) { - printk("works.\n"); + apic_printk(APIC_VERBOSE," works.\n"); nmi_watchdog_default(); if (nmi_watchdog == NMI_IO_APIC) { setup_nmi(); - check_nmi_watchdog(); } return; } /* * Cleanup, just in case ... */ - clear_IO_APIC_pin(0, pin2); + clear_IO_APIC_pin(apic2, pin2); } - printk(" failed.\n"); + apic_printk(APIC_VERBOSE," failed.\n"); - if (nmi_watchdog) { + if (nmi_watchdog == NMI_IO_APIC) { printk(KERN_WARNING "timer doesn't work through the IO-APIC - disabling NMI Watchdog!\n"); nmi_watchdog = 0; } - printk(KERN_INFO "...trying to set up timer as Virtual Wire IRQ..."); + apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as Virtual Wire IRQ..."); disable_8259A_irq(0); - irq_desc[0].handler = &lapic_irq_type; - apic_write_around(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */ + irq_desc[0].chip = &lapic_irq_type; + apic_write(APIC_LVT0, APIC_DM_FIXED | vector); /* Fixed mode */ enable_8259A_irq(0); if (timer_irq_works()) { - printk(" works.\n"); + apic_printk(APIC_VERBOSE," works.\n"); return; } - apic_write_around(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector); - printk(" failed.\n"); + apic_write(APIC_LVT0, APIC_LVT_MASKED | APIC_DM_FIXED | vector); + apic_printk(APIC_VERBOSE," failed.\n"); - printk(KERN_INFO "...trying to set up timer as ExtINT IRQ..."); + apic_printk(APIC_VERBOSE, KERN_INFO "...trying to set up timer as ExtINT IRQ..."); init_8259A(0); make_8259A_irq(0); - apic_write_around(APIC_LVT0, APIC_DM_EXTINT); + apic_write(APIC_LVT0, APIC_DM_EXTINT); unlock_ExtINT_logic(); if (timer_irq_works()) { - printk(" works.\n"); + apic_printk(APIC_VERBOSE," works.\n"); return; } - printk(" failed :(.\n"); - panic("IO-APIC + timer doesn't work! pester mingo@redhat.com"); + apic_printk(APIC_VERBOSE," failed :(.\n"); + panic("IO-APIC + timer doesn't work! Try using the 'noapic' kernel parameter\n"); +} + +static int __init notimercheck(char *s) +{ + no_timer_check = 1; + return 1; } +__setup("no_timer_check", notimercheck); /* * @@ -1758,13 +1733,8 @@ void __init setup_IO_APIC(void) else io_apic_irqs = ~PIC_IRQS; - printk("ENABLING IO-APIC IRQs\n"); + apic_printk(APIC_VERBOSE, "ENABLING IO-APIC IRQs\n"); - /* - * Set up the IO-APIC IRQ routing table. - */ - if (!acpi_ioapic) - setup_ioapic_ids_from_mpc(); sync_Arb_IDs(); setup_IO_APIC_irqs(); init_IO_APIC_traps(); @@ -1773,96 +1743,341 @@ void __init setup_IO_APIC(void) print_IO_APIC(); } -/* -------------------------------------------------------------------------- - ACPI-based IOAPIC Configuration - -------------------------------------------------------------------------- */ +struct sysfs_ioapic_data { + struct sys_device dev; + struct IO_APIC_route_entry entry[0]; +}; +static struct sysfs_ioapic_data * mp_ioapic_data[MAX_IO_APICS]; + +static int ioapic_suspend(struct sys_device *dev, pm_message_t state) +{ + struct IO_APIC_route_entry *entry; + struct sysfs_ioapic_data *data; + int i; -#ifdef CONFIG_ACPI_BOOT + data = container_of(dev, struct sysfs_ioapic_data, dev); + entry = data->entry; + for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ ) + *entry = ioapic_read_entry(dev->id, i); -#define IO_APIC_MAX_ID 15 + return 0; +} -int __init io_apic_get_unique_id (int ioapic, int apic_id) +static int ioapic_resume(struct sys_device *dev) { - union IO_APIC_reg_00 reg_00; - static physid_mask_t apic_id_map; + struct IO_APIC_route_entry *entry; + struct sysfs_ioapic_data *data; 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. - */ + union IO_APIC_reg_00 reg_00; + int i; - if (physids_empty(apic_id_map)) - apic_id_map = phys_cpu_present_map; + data = container_of(dev, struct sysfs_ioapic_data, dev); + entry = data->entry; spin_lock_irqsave(&ioapic_lock, flags); - reg_00.raw = io_apic_read(ioapic, 0); + reg_00.raw = io_apic_read(dev->id, 0); + if (reg_00.bits.ID != mp_ioapics[dev->id].mpc_apicid) { + reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid; + io_apic_write(dev->id, 0, reg_00.raw); + } spin_unlock_irqrestore(&ioapic_lock, flags); + for (i = 0; i < nr_ioapic_registers[dev->id]; i++) + ioapic_write_entry(dev->id, i, entry[i]); - if (apic_id >= IO_APIC_MAX_ID) { - printk(KERN_WARNING "IOAPIC[%d]: Invalid apic_id %d, trying " - "%d\n", ioapic, apic_id, reg_00.bits.ID); - apic_id = reg_00.bits.ID; - } + return 0; +} - /* - * 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 (physid_isset(apic_id, apic_id_map)) { +static struct sysdev_class ioapic_sysdev_class = { + set_kset_name("ioapic"), + .suspend = ioapic_suspend, + .resume = ioapic_resume, +}; - for (i = 0; i < IO_APIC_MAX_ID; i++) { - if (!physid_isset(i, apic_id_map)) - break; +static int __init ioapic_init_sysfs(void) +{ + struct sys_device * dev; + int i, size, error = 0; + + error = sysdev_class_register(&ioapic_sysdev_class); + if (error) + return error; + + for (i = 0; i < nr_ioapics; i++ ) { + size = sizeof(struct sys_device) + nr_ioapic_registers[i] + * sizeof(struct IO_APIC_route_entry); + mp_ioapic_data[i] = kmalloc(size, GFP_KERNEL); + if (!mp_ioapic_data[i]) { + printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i); + continue; + } + memset(mp_ioapic_data[i], 0, size); + dev = &mp_ioapic_data[i]->dev; + dev->id = i; + dev->cls = &ioapic_sysdev_class; + error = sysdev_register(dev); + if (error) { + kfree(mp_ioapic_data[i]); + mp_ioapic_data[i] = NULL; + printk(KERN_ERR "Can't suspend/resume IOAPIC %d\n", i); + continue; } + } + + return 0; +} - if (i == IO_APIC_MAX_ID) - panic("Max apic_id exceeded!\n"); +device_initcall(ioapic_init_sysfs); + +/* + * Dynamic irq allocate and deallocation + */ +int create_irq(void) +{ + /* Allocate an unused irq */ + int irq; + int new; + int vector = 0; + unsigned long flags; + cpumask_t mask; - printk(KERN_WARNING "IOAPIC[%d]: apic_id %d already used, " - "trying %d\n", ioapic, apic_id, i); + irq = -ENOSPC; + spin_lock_irqsave(&vector_lock, flags); + for (new = (NR_IRQS - 1); new >= 0; new--) { + if (platform_legacy_irq(new)) + continue; + if (irq_vector[new] != 0) + continue; + vector = __assign_irq_vector(new, TARGET_CPUS, &mask); + if (likely(vector > 0)) + irq = new; + break; + } + spin_unlock_irqrestore(&vector_lock, flags); - apic_id = i; - } + if (irq >= 0) { + dynamic_irq_init(irq); + } + return irq; +} - physid_set(apic_id, apic_id_map); +void destroy_irq(unsigned int irq) +{ + unsigned long flags; - if (reg_00.bits.ID != apic_id) { - reg_00.bits.ID = apic_id; + dynamic_irq_cleanup(irq); - spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(ioapic, 0, reg_00.raw); - reg_00.raw = io_apic_read(ioapic, 0); - spin_unlock_irqrestore(&ioapic_lock, flags); + spin_lock_irqsave(&vector_lock, flags); + __clear_irq_vector(irq); + spin_unlock_irqrestore(&vector_lock, flags); +} - /* Sanity check */ - if (reg_00.bits.ID != apic_id) - panic("IOAPIC[%d]: Unable change apic_id!\n", ioapic); +/* + * MSI mesage composition + */ +#ifdef CONFIG_PCI_MSI +static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_msg *msg) +{ + int vector; + unsigned dest; + cpumask_t tmp; + + vector = assign_irq_vector(irq, TARGET_CPUS, &tmp); + if (vector >= 0) { + dest = cpu_mask_to_apicid(tmp); + + msg->address_hi = MSI_ADDR_BASE_HI; + msg->address_lo = + MSI_ADDR_BASE_LO | + ((INT_DEST_MODE == 0) ? + MSI_ADDR_DEST_MODE_PHYSICAL: + MSI_ADDR_DEST_MODE_LOGICAL) | + ((INT_DELIVERY_MODE != dest_LowestPrio) ? + MSI_ADDR_REDIRECTION_CPU: + MSI_ADDR_REDIRECTION_LOWPRI) | + MSI_ADDR_DEST_ID(dest); + + msg->data = + MSI_DATA_TRIGGER_EDGE | + MSI_DATA_LEVEL_ASSERT | + ((INT_DELIVERY_MODE != dest_LowestPrio) ? + MSI_DATA_DELIVERY_FIXED: + MSI_DATA_DELIVERY_LOWPRI) | + MSI_DATA_VECTOR(vector); } + return vector; +} + +#ifdef CONFIG_SMP +static void set_msi_irq_affinity(unsigned int irq, cpumask_t mask) +{ + struct msi_msg msg; + unsigned int dest; + cpumask_t tmp; + int vector; + + cpus_and(tmp, mask, cpu_online_map); + if (cpus_empty(tmp)) + tmp = TARGET_CPUS; + + cpus_and(mask, tmp, CPU_MASK_ALL); - printk(KERN_INFO "IOAPIC[%d]: Assigned apic_id %d\n", ioapic, apic_id); + msix_flush_writes(irq); + vector = assign_irq_vector(irq, mask, &tmp); + if (vector < 0) + return; + + dest = cpu_mask_to_apicid(tmp); + + read_msi_msg(irq, &msg); + + msg.data &= ~MSI_DATA_VECTOR_MASK; + msg.data |= MSI_DATA_VECTOR(vector); + msg.address_lo &= ~MSI_ADDR_DEST_ID_MASK; + msg.address_lo |= MSI_ADDR_DEST_ID(dest); - return apic_id; + write_msi_msg(irq, &msg); + msix_flush_writes(irq); + set_native_irq_info(irq, mask); } +#endif /* CONFIG_SMP */ + +/* + * IRQ Chip for MSI PCI/PCI-X/PCI-Express Devices, + * which implement the MSI or MSI-X Capability Structure. + */ +static struct irq_chip msi_chip = { + .name = "PCI-MSI", + .enable = enable_msi_irq, + .disable = disable_msi_irq, + .unmask = unmask_msi_irq, + .mask = mask_msi_irq, + .ack = ack_apic_edge, +#ifdef CONFIG_SMP + .set_affinity = set_msi_irq_affinity, +#endif + .retrigger = ioapic_retrigger_irq, +}; + +int arch_setup_msi_irq(unsigned int irq, struct pci_dev *dev) +{ + struct msi_msg msg; + int ret; + ret = msi_compose_msg(dev, irq, &msg); + if (ret < 0) + return ret; + + write_msi_msg(irq, &msg); + set_irq_chip_and_handler_name(irq, &msi_chip, handle_edge_irq, "edge"); -int __init io_apic_get_version (int ioapic) + return 0; +} + +void arch_teardown_msi_irq(unsigned int irq) { - union IO_APIC_reg_01 reg_01; - unsigned long flags; + return; +} - spin_lock_irqsave(&ioapic_lock, flags); - reg_01.raw = io_apic_read(ioapic, 1); - spin_unlock_irqrestore(&ioapic_lock, flags); +#endif /* CONFIG_PCI_MSI */ + +/* + * Hypertransport interrupt support + */ +#ifdef CONFIG_HT_IRQ + +#ifdef CONFIG_SMP + +static void target_ht_irq(unsigned int irq, unsigned int dest, u8 vector) +{ + struct ht_irq_msg msg; + fetch_ht_irq_msg(irq, &msg); + + msg.address_lo &= ~(HT_IRQ_LOW_VECTOR_MASK | HT_IRQ_LOW_DEST_ID_MASK); + msg.address_hi &= ~(HT_IRQ_HIGH_DEST_ID_MASK); + + msg.address_lo |= HT_IRQ_LOW_VECTOR(vector) | HT_IRQ_LOW_DEST_ID(dest); + msg.address_hi |= HT_IRQ_HIGH_DEST_ID(dest); + + write_ht_irq_msg(irq, &msg); +} + +static void set_ht_irq_affinity(unsigned int irq, cpumask_t mask) +{ + unsigned int dest; + cpumask_t tmp; + int vector; + + cpus_and(tmp, mask, cpu_online_map); + if (cpus_empty(tmp)) + tmp = TARGET_CPUS; + + cpus_and(mask, tmp, CPU_MASK_ALL); + + vector = assign_irq_vector(irq, mask, &tmp); + if (vector < 0) + return; - return reg_01.bits.version; + dest = cpu_mask_to_apicid(tmp); + + target_ht_irq(irq, dest, vector); + set_native_irq_info(irq, mask); +} +#endif + +static struct irq_chip ht_irq_chip = { + .name = "PCI-HT", + .mask = mask_ht_irq, + .unmask = unmask_ht_irq, + .ack = ack_apic_edge, +#ifdef CONFIG_SMP + .set_affinity = set_ht_irq_affinity, +#endif + .retrigger = ioapic_retrigger_irq, +}; + +int arch_setup_ht_irq(unsigned int irq, struct pci_dev *dev) +{ + int vector; + cpumask_t tmp; + + vector = assign_irq_vector(irq, TARGET_CPUS, &tmp); + if (vector >= 0) { + struct ht_irq_msg msg; + unsigned dest; + + dest = cpu_mask_to_apicid(tmp); + + msg.address_hi = HT_IRQ_HIGH_DEST_ID(dest); + + msg.address_lo = + HT_IRQ_LOW_BASE | + HT_IRQ_LOW_DEST_ID(dest) | + HT_IRQ_LOW_VECTOR(vector) | + ((INT_DEST_MODE == 0) ? + HT_IRQ_LOW_DM_PHYSICAL : + HT_IRQ_LOW_DM_LOGICAL) | + HT_IRQ_LOW_RQEOI_EDGE | + ((INT_DELIVERY_MODE != dest_LowestPrio) ? + HT_IRQ_LOW_MT_FIXED : + HT_IRQ_LOW_MT_ARBITRATED) | + HT_IRQ_LOW_IRQ_MASKED; + + write_ht_irq_msg(irq, &msg); + + set_irq_chip_and_handler_name(irq, &ht_irq_chip, + handle_edge_irq, "edge"); + } + return vector; } +#endif /* CONFIG_HT_IRQ */ + +/* -------------------------------------------------------------------------- + ACPI-based IOAPIC Configuration + -------------------------------------------------------------------------- */ +#ifdef CONFIG_ACPI + +#define IO_APIC_MAX_ID 0xFE int __init io_apic_get_redir_entries (int ioapic) { @@ -1877,17 +2092,30 @@ int __init io_apic_get_redir_entries (int ioapic) } -int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int active_high_low) +int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity) { struct IO_APIC_route_entry entry; unsigned long flags; + int vector; + cpumask_t mask; if (!IO_APIC_IRQ(irq)) { - printk(KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n", + apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n", ioapic); return -EINVAL; } + /* + * IRQs < 16 are already in the irq_2_pin[] map + */ + if (irq >= 16) + add_pin_to_irq(irq, ioapic, pin); + + + vector = assign_irq_vector(irq, TARGET_CPUS, &mask); + if (vector < 0) + return vector; + /* * Generate a PCI IRQ routing entry and program the IOAPIC accordingly. * Note that we mask (disable) IRQs now -- these get enabled when the @@ -1896,63 +2124,66 @@ int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int a memset(&entry,0,sizeof(entry)); - entry.delivery_mode = dest_LowestPrio; - entry.dest_mode = INT_DELIVERY_MODE; - entry.dest.logical.logical_dest = TARGET_CPUS; - entry.trigger = edge_level; - entry.polarity = active_high_low; + entry.delivery_mode = INT_DELIVERY_MODE; + entry.dest_mode = INT_DEST_MODE; + entry.dest = cpu_mask_to_apicid(mask); + entry.trigger = triggering; + entry.polarity = polarity; entry.mask = 1; /* Disabled (masked) */ + entry.vector = vector & 0xff; - /* - * IRQs < 16 are already in the irq_2_pin[] map - */ - if (irq >= 16) - add_pin_to_irq(irq, ioapic, pin); - - entry.vector = assign_irq_vector(irq); - - printk(KERN_DEBUG "IOAPIC[%d]: Set PCI routing entry (%d-%d -> 0x%x -> " + apic_printk(APIC_VERBOSE,KERN_DEBUG "IOAPIC[%d]: Set PCI routing entry (%d-%d -> 0x%x -> " "IRQ %d Mode:%i Active:%i)\n", ioapic, mp_ioapics[ioapic].mpc_apicid, pin, entry.vector, irq, - edge_level, active_high_low); - - if (use_pci_vector() && !platform_legacy_irq(irq)) - irq = IO_APIC_VECTOR(irq); - if (edge_level) { - irq_desc[irq].handler = &ioapic_level_type; - } else { - irq_desc[irq].handler = &ioapic_edge_type; - } + triggering, polarity); - set_intr_gate(entry.vector, interrupt[irq]); + ioapic_register_intr(irq, entry.vector, triggering); if (!ioapic && (irq < 16)) disable_8259A_irq(irq); + ioapic_write_entry(ioapic, pin, entry); + spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(ioapic, 0x11+2*pin, *(((int *)&entry)+1)); - io_apic_write(ioapic, 0x10+2*pin, *(((int *)&entry)+0)); + set_native_irq_info(irq, TARGET_CPUS); spin_unlock_irqrestore(&ioapic_lock, flags); return 0; } -#endif /*CONFIG_ACPI_BOOT*/ +#endif /* CONFIG_ACPI */ + -#ifndef CONFIG_SMP -void send_IPI_self(int vector) +/* + * This function currently is only a helper for the i386 smp boot process where + * we need to reprogram the ioredtbls to cater for the cpus which have come online + * so mask in all cases should simply be TARGET_CPUS + */ +#ifdef CONFIG_SMP +void __init setup_ioapic_dest(void) { - unsigned int cfg; + int pin, ioapic, irq, irq_entry; - /* - * Wait for idle. - */ - apic_wait_icr_idle(); - cfg = APIC_DM_FIXED | APIC_DEST_SELF | vector | APIC_DEST_LOGICAL; + if (skip_ioapic_setup == 1) + return; - /* - * Send the IPI. The write to APIC_ICR fires this off. - */ - apic_write_around(APIC_ICR, cfg); + for (ioapic = 0; ioapic < nr_ioapics; ioapic++) { + for (pin = 0; pin < nr_ioapic_registers[ioapic]; pin++) { + irq_entry = find_irq_entry(ioapic, pin, mp_INT); + if (irq_entry == -1) + continue; + irq = pin_2_irq(irq_entry, ioapic, pin); + + /* setup_IO_APIC_irqs could fail to get vector for some device + * when you have too many devices, because at that time only boot + * cpu is online. + */ + if(!irq_vector[irq]) + setup_IO_APIC_irq(ioapic, pin, irq_entry, irq); + else + set_ioapic_affinity_irq(irq, TARGET_CPUS); + } + + } } #endif