X-Git-Url: http://git.onelab.eu/?a=blobdiff_plain;f=arch%2Fx86_64%2Fkernel%2Fio_apic-xen.c;h=5d3f4be19152a07eaa84fe7a89cf49ec646f6a93;hb=refs%2Fheads%2Fvserver;hp=9ba203b24ced14b5b2bb022ea575fe3d3fa8ce08;hpb=76828883507a47dae78837ab5dec5a5b4513c667;p=linux-2.6.git diff --git a/arch/x86_64/kernel/io_apic-xen.c b/arch/x86_64/kernel/io_apic-xen.c index 9ba203b24..5d3f4be19 100644 --- a/arch/x86_64/kernel/io_apic-xen.c +++ b/arch/x86_64/kernel/io_apic-xen.c @@ -26,9 +26,12 @@ #include #include #include +#include #include #include #include +#include +#include #ifdef CONFIG_ACPI #include #endif @@ -41,6 +44,31 @@ #include #include #include +#include +#include + +DEFINE_PER_CPU(vector_irq_t, vector_irq) = { + [0 ... FIRST_EXTERNAL_VECTOR - 1] = -1, + [FIRST_EXTERNAL_VECTOR + 0] = 0, + [FIRST_EXTERNAL_VECTOR + 1] = 1, + [FIRST_EXTERNAL_VECTOR + 2] = 2, + [FIRST_EXTERNAL_VECTOR + 3] = 3, + [FIRST_EXTERNAL_VECTOR + 4] = 4, + [FIRST_EXTERNAL_VECTOR + 5] = 5, + [FIRST_EXTERNAL_VECTOR + 6] = 6, + [FIRST_EXTERNAL_VECTOR + 7] = 7, + [FIRST_EXTERNAL_VECTOR + 8] = 8, + [FIRST_EXTERNAL_VECTOR + 9] = 9, + [FIRST_EXTERNAL_VECTOR + 10] = 10, + [FIRST_EXTERNAL_VECTOR + 11] = 11, + [FIRST_EXTERNAL_VECTOR + 12] = 12, + [FIRST_EXTERNAL_VECTOR + 13] = 13, + [FIRST_EXTERNAL_VECTOR + 14] = 14, + [FIRST_EXTERNAL_VECTOR + 15] = 15, + [FIRST_EXTERNAL_VECTOR + 16 ... NR_VECTORS - 1] = -1 +}; + +static int assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result); #define __apicdebuginit __init @@ -48,17 +76,17 @@ int sis_apic_bug; /* not actually supported, dummy for compile */ static int no_timer_check; -int disable_timer_pin_1 __initdata; +static int disable_timer_pin_1 __initdata; -#ifndef CONFIG_XEN -int timer_over_8254 __initdata = 0; +int timer_over_8254 __initdata = 1; +#ifndef CONFIG_XEN /* Where if anywhere is the i8259 connect in external int mode */ static struct { int pin, apic; } ioapic_i8259 = { -1, -1 }; #endif static DEFINE_SPINLOCK(ioapic_lock); -static DEFINE_SPINLOCK(vector_lock); +DEFINE_SPINLOCK(vector_lock); /* * # of IRQ routing registers @@ -83,13 +111,51 @@ static struct irq_pin_list { short apic, pin, next; } irq_2_pin[PIN_MAP_SIZE]; -int vector_irq[NR_VECTORS] __read_mostly = { [0 ... NR_VECTORS - 1] = -1}; -#ifdef CONFIG_PCI_MSI -#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); +} + +/* + * Re-write a value: to be used for read-modify-write + * cycles where the read already set up the index register. + */ +static inline void io_apic_modify(unsigned int apic, unsigned int value) +{ + struct io_apic __iomem *io_apic = io_apic_base(apic); + writel(value, &io_apic->data); +} + +/* + * 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); +} #ifdef CONFIG_XEN @@ -130,15 +196,118 @@ static inline void xen_io_apic_write(unsigned int apic, unsigned int reg, unsign #define io_apic_write(a,r,v) xen_io_apic_write(a,r,v) #define clear_IO_APIC() ((void)0) +#endif -#else +#ifndef CONFIG_XEN +#define __DO_ACTION(R, ACTION, FINAL) \ + \ +{ \ + int pin; \ + struct irq_pin_list *entry = irq_2_pin + irq; \ + \ + BUG_ON(irq >= NR_IRQS); \ + for (;;) { \ + unsigned int reg; \ + pin = entry->pin; \ + if (pin == -1) \ + break; \ + reg = io_apic_read(entry->apic, 0x10 + R + pin*2); \ + reg ACTION; \ + io_apic_modify(entry->apic, reg); \ + if (!entry->next) \ + break; \ + entry = irq_2_pin + entry->next; \ + } \ + FINAL; \ +} +#endif /* !CONFIG_XEN */ + +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! + */ +#ifndef CONFIG_XEN +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)) @@ -146,7 +315,11 @@ static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask) cpus_and(mask, tmp, CPU_MASK_ALL); - dest = cpu_mask_to_apicid(mask); + vector = assign_irq_vector(irq, mask, &tmp); + if (vector < 0) + return; + + dest = cpu_mask_to_apicid(tmp); /* * Only the high 8 bits are valid. @@ -154,12 +327,11 @@ static void set_ioapic_affinity_irq(unsigned int irq, cpumask_t mask) dest = SET_APIC_LOGICAL_ID(dest); spin_lock_irqsave(&ioapic_lock, flags); - __DO_ACTION(1, = dest, ) - set_irq_info(irq, mask); + __target_IO_APIC_irq(irq, dest, vector); + set_native_irq_info(irq, mask); spin_unlock_irqrestore(&ioapic_lock, flags); } #endif - #endif /* !CONFIG_XEN */ /* @@ -187,27 +359,6 @@ static void add_pin_to_irq(unsigned int irq, int apic, int pin) } #ifndef CONFIG_XEN -#define __DO_ACTION(R, ACTION, FINAL) \ - \ -{ \ - int pin; \ - struct irq_pin_list *entry = irq_2_pin + irq; \ - \ - BUG_ON(irq >= NR_IRQS); \ - for (;;) { \ - unsigned int reg; \ - pin = entry->pin; \ - if (pin == -1) \ - break; \ - reg = io_apic_read(entry->apic, 0x10 + R + pin*2); \ - reg ACTION; \ - io_apic_modify(entry->apic, reg); \ - if (!entry->next) \ - break; \ - entry = irq_2_pin + entry->next; \ - } \ - FINAL; \ -} #define DO_ACTION(name,R,ACTION, FINAL) \ \ @@ -240,24 +391,15 @@ static void unmask_IO_APIC_irq (unsigned int irq) 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) @@ -270,17 +412,6 @@ static void clear_IO_APIC (void) } #endif /* !CONFIG_XEN */ - -static u8 gsi_2_irq[NR_IRQ_VECTORS] = { [0 ... NR_IRQ_VECTORS-1] = 0xFF }; - -/* - * support for broken MP BIOSs, enables hand-redirection of PIRQ0-7 to - * specific CPU-side IRQs. - */ - -#define MAX_PIRQS 8 -static int pirq_entries [MAX_PIRQS]; -static int pirqs_enabled; int skip_ioapic_setup; int ioapic_force; @@ -289,20 +420,18 @@ 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); - -#ifndef CONFIG_XEN static int __init setup_disable_8254_timer(char *s) { timer_over_8254 = -1; @@ -316,140 +445,8 @@ static int __init setup_enable_8254_timer(char *s) __setup("disable_8254_timer", setup_disable_8254_timer); __setup("enable_8254_timer", setup_enable_8254_timer); -#endif /* !CONFIG_XEN */ - -#include -#include -#include -#ifdef CONFIG_ACPI - -static int nvidia_hpet_detected __initdata; - -static int __init nvidia_hpet_check(unsigned long phys, unsigned long size) -{ - nvidia_hpet_detected = 1; - return 0; -} -#endif - -/* 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. - - ... and others. Really should move this somewhere else. - - Kludge-O-Rama. */ -void __init check_ioapic(void) -{ - int num,slot,func; - /* 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_IOMMU - if ((end_pfn > MAX_DMA32_PFN || - force_iommu) && - !iommu_aperture_allowed) { - printk(KERN_INFO - "Looks like a VIA chipset. Disabling IOMMU. Override with \"iommu=allowed\"\n"); - iommu_aperture_disabled = 1; - } -#endif - return; - case PCI_VENDOR_ID_NVIDIA: -#ifdef CONFIG_ACPI - /* - * All timer overrides on Nvidia are - * wrong unless HPET is enabled. - */ - nvidia_hpet_detected = 0; - acpi_table_parse(ACPI_HPET, - nvidia_hpet_check); - if (nvidia_hpet_detected == 0) { - acpi_skip_timer_override = 1; - printk(KERN_INFO "Nvidia board " - "detected. Ignoring ACPI " - "timer override.\n"); - } -#endif - /* RED-PEN skip them on mptables too? */ - return; - case PCI_VENDOR_ID_ATI: - - /* This should be actually default, but - for 2.6.16 let's do it for ATI only where - it's really needed. */ -#ifndef CONFIG_XEN - if (timer_over_8254 == 1) { - timer_over_8254 = 0; - printk(KERN_INFO - "ATI board detected. Disabling timer routing over 8254.\n"); - } -#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) -{ - 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; - apic_printk(APIC_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, "... 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); - /* * Find the IRQ entry number of a certain pin. */ @@ -478,9 +475,7 @@ 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)) @@ -496,9 +491,7 @@ static int __init find_isa_irq_apic(int irq, int type) for (i = 0; i < mp_irq_entries; i++) { int lbus = mp_irqs[i].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)) break; @@ -539,7 +532,7 @@ 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))) { @@ -562,27 +555,6 @@ int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin) return best_guess; } -/* - * EISA Edge/Level control register, ELCR - */ -static int EISA_ELCR(unsigned int irq) -{ - if (irq < 16) { - unsigned int port = 0x4d0 + (irq >> 3); - return (inb(port) >> (irq & 7)) & 1; - } - apic_printk(APIC_VERBOSE, "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. */ @@ -595,12 +567,6 @@ static int 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; @@ -612,38 +578,11 @@ 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; @@ -681,38 +620,11 @@ static int 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; @@ -749,64 +661,6 @@ static inline int irq_trigger(int idx) return MPBIOS_trigger(idx); } -static int next_irq = 16; - -/* - * gsi_irq_sharing -- Name overload! "irq" can be either a legacy IRQ - * in the range 0-15, a linux IRQ in the range 0-223, or a GSI number - * from ACPI, which can reach 800 in large boxen. - * - * Compact the sparse GSI space into a sequential IRQ series and reuse - * vectors if possible. - */ -int gsi_irq_sharing(int gsi) -{ - int i, tries, vector; - - BUG_ON(gsi >= NR_IRQ_VECTORS); - - if (platform_legacy_irq(gsi)) - return gsi; - - if (gsi_2_irq[gsi] != 0xFF) - return (int)gsi_2_irq[gsi]; - - tries = NR_IRQS; - try_again: - vector = assign_irq_vector(gsi); - - /* - * Sharing vectors means sharing IRQs, so scan irq_vectors for previous - * use of vector and if found, return that IRQ. However, we never want - * to share legacy IRQs, which usually have a different trigger mode - * than PCI. - */ - for (i = 0; i < NR_IRQS; i++) - if (IO_APIC_VECTOR(i) == vector) - break; - if (platform_legacy_irq(i)) { - if (--tries >= 0) { - IO_APIC_VECTOR(i) = 0; - goto try_again; - } - panic("gsi_irq_sharing: didn't find an IRQ using vector 0x%02X for GSI %d", vector, gsi); - } - if (i < NR_IRQS) { - gsi_2_irq[gsi] = i; - printk(KERN_INFO "GSI %d sharing vector 0x%02X and IRQ %d\n", - gsi, vector, i); - return i; - } - - i = next_irq++; - BUG_ON(i >= NR_IRQS); - gsi_2_irq[gsi] = i; - IO_APIC_VECTOR(i) = vector; - printk(KERN_INFO "GSI %d assigned vector 0x%02X and IRQ %d\n", - gsi, vector, i); - return i; -} - static int pin_2_irq(int idx, int apic, int pin) { int irq, i; @@ -818,49 +672,16 @@ 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; - irq = gsi_irq_sharing(irq); - break; - } - default: - { - printk(KERN_ERR "unknown bus type %d.\n",bus); - irq = 0; - break; - } - } - BUG_ON(irq >= NR_IRQS); - - /* - * 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, "disabling PIRQ%d\n", pin-16); - } else { - irq = pirq_entries[pin-16]; - apic_printk(APIC_VERBOSE, "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; @@ -884,37 +705,106 @@ 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] __read_mostly; +static u8 irq_vector[NR_IRQ_VECTORS] __read_mostly; + +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, +}; -int assign_irq_vector(int irq) + +static int __assign_irq_vector(int irq, cpumask_t mask, cpumask_t *result) { struct physdev_irq irq_op; - unsigned long flags; - - BUG_ON(irq != AUTO_ASSIGN && (unsigned)irq >= NR_IRQ_VECTORS); + int vector; - spin_lock_irqsave(&vector_lock, flags); + BUG_ON((unsigned)irq >= NR_IRQ_VECTORS); - if (irq != AUTO_ASSIGN && IO_APIC_VECTOR(irq) > 0) { - spin_unlock_irqrestore(&vector_lock, flags); - return IO_APIC_VECTOR(irq); + if (irq_vector[irq] > 0) { + cpus_and(*result, irq_domain[irq], mask); + return irq_vector[irq]; } irq_op.irq = irq; if (HYPERVISOR_physdev_op(PHYSDEVOP_alloc_irq_vector, &irq_op)) return -ENOSPC; - vector_irq[irq_op.vector] = irq; - if (irq != AUTO_ASSIGN) - IO_APIC_VECTOR(irq) = irq_op.vector; + vector = irq_op.vector; + per_cpu(vector_irq,0)[vector] = irq; + irq_vector[irq] = vector; + cpus_and(*result, irq_domain[irq], mask); + + return vector; +} + +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 irq_op.vector; + 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); + #ifndef CONFIG_XEN -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 @@ -922,42 +812,80 @@ static struct hw_interrupt_type ioapic_edge_type; static void ioapic_register_intr(int irq, int vector, unsigned long trigger) { - unsigned idx; - - idx = use_pci_vector() && !platform_legacy_irq(irq) ? vector : irq; - if ((trigger == IOAPIC_AUTO && IO_APIC_irq_trigger(irq)) || trigger == IOAPIC_LEVEL) - irq_desc[idx].chip = &ioapic_level_type; - else - irq_desc[idx].chip = &ioapic_edge_type; - set_intr_gate(vector, interrupt[idx]); + 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"); + } } #else #define ioapic_register_intr(_irq,_vector,_trigger) ((void)0) #endif /* !CONFIG_XEN */ -static 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; + + /* + * 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.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++) { - /* - * 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.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS); - idx = find_irq_entry(apic,pin,mp_INT); if (idx == -1) { if (first_notcon) { @@ -968,34 +896,11 @@ static void __init setup_IO_APIC_irqs(void) 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 = cpu_mask_to_apicid(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)); - set_native_irq_info(irq, TARGET_CPUS); - spin_unlock_irqrestore(&ioapic_lock, flags); } } @@ -1026,7 +931,7 @@ static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, in */ entry.dest_mode = INT_DEST_MODE; entry.mask = 0; /* unmask IRQ now */ - entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS); + entry.dest = cpu_mask_to_apicid(TARGET_CPUS); entry.delivery_mode = INT_DELIVERY_MODE; entry.polarity = 0; entry.trigger = 0; @@ -1036,7 +941,7 @@ static void __init setup_ExtINT_IRQ0_pin(unsigned int apic, unsigned int pin, in * The timer IRQ doesn't have to know that behind the * scene we have a 8259A-master in AEOI mode ... */ - irq_desc[0].chip = &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: @@ -1126,21 +1031,17 @@ void __apicdebuginit 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", @@ -1155,17 +1056,12 @@ void __apicdebuginit print_IO_APIC(void) ); } } - if (use_pci_vector()) - printk(KERN_INFO "Using vector-based indexing\n"); printk(KERN_DEBUG "IRQ to pin mappings:\n"); for (i = 0; i < NR_IRQS; i++) { struct irq_pin_list *entry = irq_2_pin + i; if (entry->pin < 0) continue; - if (use_pci_vector() && !platform_legacy_irq(i)) - printk(KERN_DEBUG "IRQ%d ", IO_APIC_VECTOR(i)); - else - printk(KERN_DEBUG "IRQ%d ", i); + printk(KERN_DEBUG "IRQ%d ", i); for (;;) { printk("-> %d:%d", entry->apic, entry->pin); if (!entry->next) @@ -1336,9 +1232,6 @@ static void __init enable_IO_APIC(void) 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): @@ -1355,11 +1248,7 @@ static void __init enable_IO_APIC(void) /* 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; - 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 the interrupt line is enabled and in ExtInt mode * I have found the pin where the i8259 is connected. @@ -1398,119 +1287,43 @@ 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(); - -#ifndef CONFIG_XEN - /* - * 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; - unsigned long 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.physical.physical_dest = - GET_APIC_ID(apic_read(APIC_ID)); - - /* - * Add it to the IO-APIC irq-routing table: - */ - spin_lock_irqsave(&ioapic_lock, flags); - io_apic_write(ioapic_i8259.apic, 0x11+2*ioapic_i8259.pin, - *(((int *)&entry)+1)); - io_apic_write(ioapic_i8259.apic, 0x10+2*ioapic_i8259.pin, - *(((int *)&entry)+0)); - spin_unlock_irqrestore(&ioapic_lock, flags); - } - - disconnect_bsp_APIC(ioapic_i8259.pin != -1); -#endif -} - -/* - * 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 - */ - -#ifndef CONFIG_XEN -static void __init setup_ioapic_ids_from_mpc (void) -{ - union IO_APIC_reg_00 reg_00; - 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; - - - printk(KERN_INFO "Using IO-APIC %d\n", mp_ioapics[apic].mpc_apicid); - - - /* - * 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; +void disable_IO_APIC(void) +{ + /* + * Clear the IO-APIC before rebooting: + */ + clear_IO_APIC(); - /* - * Read the right value from the MPC table and - * write it into the ID register. - */ - apic_printk(APIC_VERBOSE,KERN_INFO "...changing IO-APIC physical APIC ID to %d ...", - mp_ioapics[apic].mpc_apicid); +#ifndef CONFIG_XEN + /* + * 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) - printk("could not set ID!\n"); - else - apic_printk(APIC_VERBOSE," ok.\n"); + ioapic_write_entry(ioapic_i8259.apic, ioapic_i8259.pin, entry); } -} -#else -static void __init setup_ioapic_ids_from_mpc(void) { } + + disconnect_bsp_APIC(ioapic_i8259.pin != -1); #endif +} /* * There is a nasty bug in some older SMP boards, their mptable lies @@ -1566,7 +1379,7 @@ 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; @@ -1583,107 +1396,19 @@ 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) -{ - move_irq(irq); - if ((irq_desc[irq].status & (IRQ_PENDING | IRQ_DISABLED)) - == (IRQ_PENDING | IRQ_DISABLED)) - mask_IO_APIC_irq(irq); - ack_APIC_irq(); -} - -/* - * 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); - - return 0; /* don't check for pending */ -} - -static void end_level_ioapic_irq (unsigned int irq) -{ - move_irq(irq); - ack_APIC_irq(); -} - -#ifdef CONFIG_PCI_MSI -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); - - move_native_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); - - move_native_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); - - unmask_IO_APIC_irq(irq); -} - -#ifdef CONFIG_SMP -static void set_ioapic_affinity_vector (unsigned int vector, - cpumask_t cpu_mask) +static int ioapic_retrigger_irq(unsigned int irq) { - int irq = vector_to_irq(vector); + cpumask_t mask; + unsigned vector; + unsigned long flags; - set_native_irq_info(vector, cpu_mask); - set_ioapic_affinity_irq(irq, cpu_mask); -} -#endif // CONFIG_SMP -#endif // CONFIG_PCI_MSI + spin_lock_irqsave(&vector_lock, flags); + vector = irq_vector[irq]; + cpus_clear(mask); + cpu_set(first_cpu(irq_domain[irq]), mask); -static int ioapic_retrigger(unsigned int irq) -{ - send_IPI_self(IO_APIC_VECTOR(irq)); + send_IPI_mask(mask, vector); + spin_unlock_irqrestore(&vector_lock, flags); return 1; } @@ -1697,32 +1422,47 @@ static int ioapic_retrigger(unsigned int irq) * races. */ -static struct hw_interrupt_type ioapic_edge_type __read_mostly = { - .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, -#ifdef CONFIG_SMP - .set_affinity = set_ioapic_affinity, +static void ack_apic_edge(unsigned int irq) +{ + move_native_irq(irq); + ack_APIC_irq(); +} + +static void ack_apic_level(unsigned int irq) +{ + int do_unmask_irq = 0; + +#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 - .retrigger = ioapic_retrigger, -}; -static struct hw_interrupt_type ioapic_level_type __read_mostly = { - .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, + /* + * We must acknowledge the irq before we move it or the acknowledge will + * not propogate properly. + */ + ack_APIC_irq(); + + /* Now we can move and renable the irq */ + move_masked_irq(irq); + if (unlikely(do_unmask_irq)) + unmask_IO_APIC_irq(irq); +} + +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, + .set_affinity = set_ioapic_affinity_irq, #endif - .retrigger = ioapic_retrigger, + .retrigger = ioapic_retrigger_irq, }; #endif /* !CONFIG_XEN */ @@ -1743,12 +1483,7 @@ 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 @@ -1759,7 +1494,7 @@ static inline void init_IO_APIC_traps(void) #ifndef CONFIG_XEN else /* Strange. Oh, well.. */ - irq_desc[irq].chip = &no_irq_type; + irq_desc[irq].chip = &no_irq_chip; #endif } } @@ -1846,7 +1581,7 @@ static inline void unlock_ExtINT_logic(void) 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; @@ -1880,8 +1615,6 @@ static inline void unlock_ExtINT_logic(void) spin_unlock_irqrestore(&ioapic_lock, flags); } -int timer_uses_ioapic_pin_0; - /* * 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 @@ -1894,13 +1627,13 @@ static inline void check_timer(void) { 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 @@ -1919,9 +1652,6 @@ static inline void check_timer(void) pin2 = ioapic_i8259.pin; apic2 = ioapic_i8259.apic; - if (pin1 == 0) - timer_uses_ioapic_pin_0 = 1; - apic_printk(APIC_VERBOSE,KERN_INFO "..TIMER: vector=0x%02X apic1=%d pin1=%d apic2=%d pin2=%d\n", vector, apic1, pin1, apic2, pin2); @@ -2005,7 +1735,6 @@ static inline void check_timer(void) panic("IO-APIC + timer doesn't work! Try using the 'noapic' kernel parameter\n"); } #else -int timer_uses_ioapic_pin_0; #define check_timer() ((void)0) #endif /* !CONFIG_XEN */ @@ -2036,11 +1765,6 @@ void __init setup_IO_APIC(void) 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(); #ifndef CONFIG_XEN sync_Arb_IDs(); #endif /* !CONFIG_XEN */ @@ -2061,17 +1785,12 @@ static int ioapic_suspend(struct sys_device *dev, pm_message_t state) { struct IO_APIC_route_entry *entry; struct sysfs_ioapic_data *data; - unsigned long flags; int i; data = container_of(dev, struct sysfs_ioapic_data, dev); entry = data->entry; - spin_lock_irqsave(&ioapic_lock, flags); - for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ ) { - *(((int *)entry) + 1) = io_apic_read(dev->id, 0x11 + 2 * i); - *(((int *)entry) + 0) = io_apic_read(dev->id, 0x10 + 2 * i); - } - spin_unlock_irqrestore(&ioapic_lock, flags); + for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ ) + *entry = ioapic_read_entry(dev->id, i); return 0; } @@ -2093,11 +1812,9 @@ static int ioapic_resume(struct sys_device *dev) reg_00.bits.ID = mp_ioapics[dev->id].mpc_apicid; io_apic_write(dev->id, 0, reg_00.raw); } - for (i = 0; i < nr_ioapic_registers[dev->id]; i ++, entry ++ ) { - io_apic_write(dev->id, 0x11+2*i, *(((int *)entry)+1)); - io_apic_write(dev->id, 0x10+2*i, *(((int *)entry)+0)); - } spin_unlock_irqrestore(&ioapic_lock, flags); + for (i = 0; i < nr_ioapic_registers[dev->id]; i++) + ioapic_write_entry(dev->id, i, entry[i]); return 0; } @@ -2143,26 +1860,253 @@ static int __init ioapic_init_sysfs(void) device_initcall(ioapic_init_sysfs); -/* -------------------------------------------------------------------------- - ACPI-based IOAPIC Configuration - -------------------------------------------------------------------------- */ +/* + * 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; -#ifdef CONFIG_ACPI + 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); -#define IO_APIC_MAX_ID 0xFE + if (irq >= 0) { + dynamic_irq_init(irq); + } + return irq; +} -int __init io_apic_get_version (int ioapic) +void destroy_irq(unsigned int irq) { - union IO_APIC_reg_01 reg_01; unsigned long flags; - spin_lock_irqsave(&ioapic_lock, flags); - reg_01.raw = io_apic_read(ioapic, 1); - spin_unlock_irqrestore(&ioapic_lock, flags); + dynamic_irq_cleanup(irq); + + spin_lock_irqsave(&vector_lock, flags); + __clear_irq_vector(irq); + spin_unlock_irqrestore(&vector_lock, flags); +} + +/* + * MSI mesage composition + */ +#ifndef CONFIG_XEN +#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); + + 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); + + write_msi_msg(irq, &msg); + 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", + .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"); + + return 0; +} + +void arch_teardown_msi_irq(unsigned int irq) +{ + return; +} + +#endif /* CONFIG_PCI_MSI */ +#endif /* !CONFIG_XEN */ +/* + * Hypertransport interrupt support + */ +#ifdef CONFIG_HT_IRQ + +#ifdef CONFIG_SMP - return reg_01.bits.version; +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; + + 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) { @@ -2177,10 +2121,12 @@ 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)) { apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n", @@ -2188,6 +2134,17 @@ int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int a 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 @@ -2198,34 +2155,26 @@ int io_apic_set_pci_routing (int ioapic, int pin, int irq, int edge_level, int a entry.delivery_mode = INT_DELIVERY_MODE; entry.dest_mode = INT_DEST_MODE; - entry.dest.logical.logical_dest = cpu_mask_to_apicid(TARGET_CPUS); - entry.trigger = edge_level; - entry.polarity = active_high_low; + entry.dest = cpu_mask_to_apicid(mask); + entry.trigger = triggering; + entry.polarity = polarity; entry.mask = 1; /* Disabled (masked) */ - - irq = gsi_irq_sharing(irq); - /* - * 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); + entry.vector = vector & 0xff; 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); + triggering, polarity); - ioapic_register_intr(irq, entry.vector, edge_level); + 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(use_pci_vector() ? entry.vector : irq, TARGET_CPUS); + set_native_irq_info(irq, TARGET_CPUS); spin_unlock_irqrestore(&ioapic_lock, flags); return 0; @@ -2254,7 +2203,15 @@ void __init setup_ioapic_dest(void) if (irq_entry == -1) continue; irq = pin_2_irq(irq_entry, ioapic, pin); - set_ioapic_affinity_irq(irq, TARGET_CPUS); + + /* 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); } }