X-Git-Url: http://ftp.safe.ca/?a=blobdiff_plain;f=arch%2Fx86%2Fkernel%2Fapic%2Fio_apic.c;h=d836b4d347e60d0587220b5349153770090b4bd8;hb=ffc438366c2660a6a811b94ba33229bf217f8254;hp=0ad3fe77641a7b5c6b797d64e4dc4c6e359f8e13;hpb=b24696bc55f66fecc30715e003f10fc2555a9271;p=safe%2Fjmp%2Flinux-2.6

diff --git a/arch/x86/kernel/apic/io_apic.c b/arch/x86/kernel/apic/io_apic.c
index 0ad3fe7..d836b4d 100644
--- a/arch/x86/kernel/apic/io_apic.c
+++ b/arch/x86/kernel/apic/io_apic.c
@@ -59,12 +59,15 @@
 #include <asm/setup.h>
 #include <asm/irq_remapping.h>
 #include <asm/hpet.h>
+#include <asm/hw_irq.h>
 #include <asm/uv/uv_hub.h>
 #include <asm/uv/uv_irq.h>
 
 #include <asm/apic.h>
 
 #define __apicdebuginit(type) static type __init
+#define for_each_irq_pin(entry, head) \
+	for (entry = head; entry; entry = entry->next)
 
 /*
  *      Is the SiS APIC rmw bug present ?
@@ -115,32 +118,25 @@ static int __init parse_noapic(char *str)
 }
 early_param("noapic", parse_noapic);
 
-struct irq_pin_list;
-
-/*
- * This is performance-critical, we want to do it O(1)
- *
- * the indexing order of this array favors 1:1 mappings
- * between pins and IRQs.
- */
-
 struct irq_pin_list {
 	int apic, pin;
 	struct irq_pin_list *next;
 };
 
-static struct irq_pin_list *get_one_free_irq_2_pin(int cpu)
+static struct irq_pin_list *get_one_free_irq_2_pin(int node)
 {
 	struct irq_pin_list *pin;
-	int node;
-
-	node = cpu_to_node(cpu);
 
 	pin = kzalloc_node(sizeof(*pin), GFP_ATOMIC, node);
 
 	return pin;
 }
 
+/*
+ * This is performance-critical, we want to do it O(1)
+ *
+ * Most irqs are mapped 1:1 with pins.
+ */
 struct irq_cfg {
 	struct irq_pin_list *irq_2_pin;
 	cpumask_var_t domain;
@@ -148,9 +144,6 @@ struct irq_cfg {
 	unsigned move_cleanup_count;
 	u8 vector;
 	u8 move_in_progress : 1;
-#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC
-	u8 move_desc_pending : 1;
-#endif
 };
 
 /* irq_cfg is indexed by the sum of all RTEs in all I/O APICs. */
@@ -182,16 +175,18 @@ int __init arch_early_irq_init(void)
 	struct irq_cfg *cfg;
 	struct irq_desc *desc;
 	int count;
+	int node;
 	int i;
 
 	cfg = irq_cfgx;
 	count = ARRAY_SIZE(irq_cfgx);
+	node= cpu_to_node(boot_cpu_id);
 
 	for (i = 0; i < count; i++) {
 		desc = irq_to_desc(i);
 		desc->chip_data = &cfg[i];
-		alloc_bootmem_cpumask_var(&cfg[i].domain);
-		alloc_bootmem_cpumask_var(&cfg[i].old_domain);
+		zalloc_cpumask_var_node(&cfg[i].domain, GFP_NOWAIT, node);
+		zalloc_cpumask_var_node(&cfg[i].old_domain, GFP_NOWAIT, node);
 		if (i < NR_IRQS_LEGACY)
 			cpumask_setall(cfg[i].domain);
 	}
@@ -212,12 +207,9 @@ static struct irq_cfg *irq_cfg(unsigned int irq)
 	return cfg;
 }
 
-static struct irq_cfg *get_one_free_irq_cfg(int cpu)
+static struct irq_cfg *get_one_free_irq_cfg(int node)
 {
 	struct irq_cfg *cfg;
-	int node;
-
-	node = cpu_to_node(cpu);
 
 	cfg = kzalloc_node(sizeof(*cfg), GFP_ATOMIC, node);
 	if (cfg) {
@@ -238,13 +230,13 @@ static struct irq_cfg *get_one_free_irq_cfg(int cpu)
 	return cfg;
 }
 
-int arch_init_chip_data(struct irq_desc *desc, int cpu)
+int arch_init_chip_data(struct irq_desc *desc, int node)
 {
 	struct irq_cfg *cfg;
 
 	cfg = desc->chip_data;
 	if (!cfg) {
-		desc->chip_data = get_one_free_irq_cfg(cpu);
+		desc->chip_data = get_one_free_irq_cfg(node);
 		if (!desc->chip_data) {
 			printk(KERN_ERR "can not alloc irq_cfg\n");
 			BUG_ON(1);
@@ -254,10 +246,9 @@ int arch_init_chip_data(struct irq_desc *desc, int cpu)
 	return 0;
 }
 
-#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC
-
+/* for move_irq_desc */
 static void
-init_copy_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg, int cpu)
+init_copy_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg, int node)
 {
 	struct irq_pin_list *old_entry, *head, *tail, *entry;
 
@@ -266,7 +257,7 @@ init_copy_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg, int cpu)
 	if (!old_entry)
 		return;
 
-	entry = get_one_free_irq_2_pin(cpu);
+	entry = get_one_free_irq_2_pin(node);
 	if (!entry)
 		return;
 
@@ -276,7 +267,7 @@ init_copy_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg, int cpu)
 	tail		= entry;
 	old_entry	= old_entry->next;
 	while (old_entry) {
-		entry = get_one_free_irq_2_pin(cpu);
+		entry = get_one_free_irq_2_pin(node);
 		if (!entry) {
 			entry = head;
 			while (entry) {
@@ -316,12 +307,12 @@ static void free_irq_2_pin(struct irq_cfg *old_cfg, struct irq_cfg *cfg)
 }
 
 void arch_init_copy_chip_data(struct irq_desc *old_desc,
-				 struct irq_desc *desc, int cpu)
+				 struct irq_desc *desc, int node)
 {
 	struct irq_cfg *cfg;
 	struct irq_cfg *old_cfg;
 
-	cfg = get_one_free_irq_cfg(cpu);
+	cfg = get_one_free_irq_cfg(node);
 
 	if (!cfg)
 		return;
@@ -332,7 +323,7 @@ void arch_init_copy_chip_data(struct irq_desc *old_desc,
 
 	memcpy(cfg, old_cfg, sizeof(struct irq_cfg));
 
-	init_copy_irq_2_pin(old_cfg, cfg, cpu);
+	init_copy_irq_2_pin(old_cfg, cfg, node);
 }
 
 static void free_irq_cfg(struct irq_cfg *old_cfg)
@@ -356,19 +347,7 @@ void arch_free_chip_data(struct irq_desc *old_desc, struct irq_desc *desc)
 		old_desc->chip_data = NULL;
 	}
 }
-
-static void
-set_extra_move_desc(struct irq_desc *desc, const struct cpumask *mask)
-{
-	struct irq_cfg *cfg = desc->chip_data;
-
-	if (!cfg->move_in_progress) {
-		/* it means that domain is not changed */
-		if (!cpumask_intersects(desc->affinity, mask))
-			cfg->move_desc_pending = 1;
-	}
-}
-#endif
+/* end for move_irq_desc */
 
 #else
 static struct irq_cfg *irq_cfg(unsigned int irq)
@@ -378,13 +357,6 @@ static struct irq_cfg *irq_cfg(unsigned int irq)
 
 #endif
 
-#ifndef CONFIG_NUMA_MIGRATE_IRQ_DESC
-static inline void
-set_extra_move_desc(struct irq_desc *desc, const struct cpumask *mask)
-{
-}
-#endif
-
 struct io_apic {
 	unsigned int index;
 	unsigned int unused[3];
@@ -440,13 +412,10 @@ static bool io_apic_level_ack_pending(struct irq_cfg *cfg)
 	unsigned long flags;
 
 	spin_lock_irqsave(&ioapic_lock, flags);
-	entry = cfg->irq_2_pin;
-	for (;;) {
+	for_each_irq_pin(entry, cfg->irq_2_pin) {
 		unsigned int reg;
 		int pin;
 
-		if (!entry)
-			break;
 		pin = entry->pin;
 		reg = io_apic_read(entry->apic, 0x10 + pin*2);
 		/* Is the remote IRR bit set? */
@@ -454,9 +423,6 @@ static bool io_apic_level_ack_pending(struct irq_cfg *cfg)
 			spin_unlock_irqrestore(&ioapic_lock, flags);
 			return true;
 		}
-		if (!entry->next)
-			break;
-		entry = entry->next;
 	}
 	spin_unlock_irqrestore(&ioapic_lock, flags);
 
@@ -488,7 +454,8 @@ static struct IO_APIC_route_entry ioapic_read_entry(int apic, int pin)
 static void
 __ioapic_write_entry(int apic, int pin, struct IO_APIC_route_entry e)
 {
-	union entry_union eu;
+	union entry_union eu = {{0, 0}};
+
 	eu.entry = e;
 	io_apic_write(apic, 0x11 + 2*pin, eu.w2);
 	io_apic_write(apic, 0x10 + 2*pin, eu.w1);
@@ -518,191 +485,73 @@ static void ioapic_mask_entry(int apic, int pin)
 	spin_unlock_irqrestore(&ioapic_lock, flags);
 }
 
-#ifdef CONFIG_SMP
-static void send_cleanup_vector(struct irq_cfg *cfg)
-{
-	cpumask_var_t cleanup_mask;
-
-	if (unlikely(!alloc_cpumask_var(&cleanup_mask, GFP_ATOMIC))) {
-		unsigned int i;
-		cfg->move_cleanup_count = 0;
-		for_each_cpu_and(i, cfg->old_domain, cpu_online_mask)
-			cfg->move_cleanup_count++;
-		for_each_cpu_and(i, cfg->old_domain, cpu_online_mask)
-			apic->send_IPI_mask(cpumask_of(i), IRQ_MOVE_CLEANUP_VECTOR);
-	} else {
-		cpumask_and(cleanup_mask, cfg->old_domain, cpu_online_mask);
-		cfg->move_cleanup_count = cpumask_weight(cleanup_mask);
-		apic->send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
-		free_cpumask_var(cleanup_mask);
-	}
-	cfg->move_in_progress = 0;
-}
-
-static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq_cfg *cfg)
-{
-	int apic, pin;
-	struct irq_pin_list *entry;
-	u8 vector = cfg->vector;
-
-	entry = cfg->irq_2_pin;
-	for (;;) {
-		unsigned int reg;
-
-		if (!entry)
-			break;
-
-		apic = entry->apic;
-		pin = entry->pin;
-		/*
-		 * With interrupt-remapping, destination information comes
-		 * from interrupt-remapping table entry.
-		 */
-		if (!irq_remapped(irq))
-			io_apic_write(apic, 0x11 + pin*2, dest);
-		reg = io_apic_read(apic, 0x10 + pin*2);
-		reg &= ~IO_APIC_REDIR_VECTOR_MASK;
-		reg |= vector;
-		io_apic_modify(apic, 0x10 + pin*2, reg);
-		if (!entry->next)
-			break;
-		entry = entry->next;
-	}
-}
-
-static int
-assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask);
-
-/*
- * Either sets desc->affinity to a valid value, and returns
- * ->cpu_mask_to_apicid of that, or returns BAD_APICID and
- * leaves desc->affinity untouched.
- */
-static unsigned int
-set_desc_affinity(struct irq_desc *desc, const struct cpumask *mask)
-{
-	struct irq_cfg *cfg;
-	unsigned int irq;
-
-	if (!cpumask_intersects(mask, cpu_online_mask))
-		return BAD_APICID;
-
-	irq = desc->irq;
-	cfg = desc->chip_data;
-	if (assign_irq_vector(irq, cfg, mask))
-		return BAD_APICID;
-
-	/* check that before desc->addinity get updated */
-	set_extra_move_desc(desc, mask);
-
-	cpumask_copy(desc->affinity, mask);
-
-	return apic->cpu_mask_to_apicid_and(desc->affinity, cfg->domain);
-}
-
-static void
-set_ioapic_affinity_irq_desc(struct irq_desc *desc, const struct cpumask *mask)
-{
-	struct irq_cfg *cfg;
-	unsigned long flags;
-	unsigned int dest;
-	unsigned int irq;
-
-	irq = desc->irq;
-	cfg = desc->chip_data;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	dest = set_desc_affinity(desc, mask);
-	if (dest != BAD_APICID) {
-		/* Only the high 8 bits are valid. */
-		dest = SET_APIC_LOGICAL_ID(dest);
-		__target_IO_APIC_irq(irq, dest, cfg);
-	}
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-static void
-set_ioapic_affinity_irq(unsigned int irq, const struct cpumask *mask)
-{
-	struct irq_desc *desc;
-
-	desc = irq_to_desc(irq);
-
-	set_ioapic_affinity_irq_desc(desc, mask);
-}
-#endif /* CONFIG_SMP */
-
 /*
  * 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_cpu(struct irq_cfg *cfg, int cpu, int apic, int pin)
+static int
+add_pin_to_irq_node_nopanic(struct irq_cfg *cfg, int node, int apic, int pin)
 {
-	struct irq_pin_list *entry;
-
-	entry = cfg->irq_2_pin;
-	if (!entry) {
-		entry = get_one_free_irq_2_pin(cpu);
-		if (!entry) {
-			printk(KERN_ERR "can not alloc irq_2_pin to add %d - %d\n",
-					apic, pin);
-			return;
-		}
-		cfg->irq_2_pin = entry;
-		entry->apic = apic;
-		entry->pin = pin;
-		return;
-	}
+	struct irq_pin_list **last, *entry;
 
-	while (entry->next) {
-		/* not again, please */
+	/* don't allow duplicates */
+	last = &cfg->irq_2_pin;
+	for_each_irq_pin(entry, cfg->irq_2_pin) {
 		if (entry->apic == apic && entry->pin == pin)
-			return;
-
-		entry = entry->next;
+			return 0;
+		last = &entry->next;
 	}
 
-	entry->next = get_one_free_irq_2_pin(cpu);
-	entry = entry->next;
+	entry = get_one_free_irq_2_pin(node);
+	if (!entry) {
+		printk(KERN_ERR "can not alloc irq_pin_list (%d,%d,%d)\n",
+				node, apic, pin);
+		return -ENOMEM;
+	}
 	entry->apic = apic;
 	entry->pin = pin;
+
+	*last = entry;
+	return 0;
+}
+
+static void add_pin_to_irq_node(struct irq_cfg *cfg, int node, int apic, int pin)
+{
+	if (add_pin_to_irq_node_nopanic(cfg, node, apic, pin))
+		panic("IO-APIC: failed to add irq-pin. Can not proceed\n");
 }
 
 /*
  * Reroute an IRQ to a different pin.
  */
-static void __init replace_pin_at_irq_cpu(struct irq_cfg *cfg, int cpu,
-				      int oldapic, int oldpin,
-				      int newapic, int newpin)
+static void __init replace_pin_at_irq_node(struct irq_cfg *cfg, int node,
+					   int oldapic, int oldpin,
+					   int newapic, int newpin)
 {
-	struct irq_pin_list *entry = cfg->irq_2_pin;
-	int replaced = 0;
+	struct irq_pin_list *entry;
 
-	while (entry) {
+	for_each_irq_pin(entry, cfg->irq_2_pin) {
 		if (entry->apic == oldapic && entry->pin == oldpin) {
 			entry->apic = newapic;
 			entry->pin = newpin;
-			replaced = 1;
 			/* every one is different, right? */
-			break;
+			return;
 		}
-		entry = entry->next;
 	}
 
-	/* why? call replace before add? */
-	if (!replaced)
-		add_pin_to_irq_cpu(cfg, cpu, newapic, newpin);
+	/* old apic/pin didn't exist, so just add new ones */
+	add_pin_to_irq_node(cfg, node, newapic, newpin);
 }
 
-static inline void io_apic_modify_irq(struct irq_cfg *cfg,
-				int mask_and, int mask_or,
-				void (*final)(struct irq_pin_list *entry))
+static void io_apic_modify_irq(struct irq_cfg *cfg,
+			       int mask_and, int mask_or,
+			       void (*final)(struct irq_pin_list *entry))
 {
 	int pin;
 	struct irq_pin_list *entry;
 
-	for (entry = cfg->irq_2_pin; entry != NULL; entry = entry->next) {
+	for_each_irq_pin(entry, cfg->irq_2_pin) {
 		unsigned int reg;
 		pin = entry->pin;
 		reg = io_apic_read(entry->apic, 0x10 + pin * 2);
@@ -719,7 +568,6 @@ static void __unmask_IO_APIC_irq(struct irq_cfg *cfg)
 	io_apic_modify_irq(cfg, ~IO_APIC_REDIR_MASKED, 0, NULL);
 }
 
-#ifdef CONFIG_X86_64
 static void io_apic_sync(struct irq_pin_list *entry)
 {
 	/*
@@ -735,11 +583,6 @@ static void __mask_IO_APIC_irq(struct irq_cfg *cfg)
 {
 	io_apic_modify_irq(cfg, ~0, IO_APIC_REDIR_MASKED, &io_apic_sync);
 }
-#else /* CONFIG_X86_32 */
-static void __mask_IO_APIC_irq(struct irq_cfg *cfg)
-{
-	io_apic_modify_irq(cfg, ~0, IO_APIC_REDIR_MASKED, NULL);
-}
 
 static void __mask_and_edge_IO_APIC_irq(struct irq_cfg *cfg)
 {
@@ -752,7 +595,6 @@ static void __unmask_and_level_IO_APIC_irq(struct irq_cfg *cfg)
 	io_apic_modify_irq(cfg, ~IO_APIC_REDIR_MASKED,
 			IO_APIC_REDIR_LEVEL_TRIGGER, NULL);
 }
-#endif /* CONFIG_X86_32 */
 
 static void mask_IO_APIC_irq_desc(struct irq_desc *desc)
 {
@@ -850,7 +692,6 @@ static int __init ioapic_pirq_setup(char *str)
 __setup("pirq=", ioapic_pirq_setup);
 #endif /* CONFIG_X86_32 */
 
-#ifdef CONFIG_INTR_REMAP
 struct IO_APIC_route_entry **alloc_ioapic_entries(void)
 {
 	int apic;
@@ -948,20 +789,6 @@ int restore_IO_APIC_setup(struct IO_APIC_route_entry **ioapic_entries)
 	return 0;
 }
 
-void reinit_intr_remapped_IO_APIC(int intr_remapping,
-	struct IO_APIC_route_entry **ioapic_entries)
-
-{
-	/*
-	 * for now plain restore of previous settings.
-	 * TBD: In the case of OS enabling interrupt-remapping,
-	 * IO-APIC RTE's need to be setup to point to interrupt-remapping
-	 * table entries. for now, do a plain restore, and wait for
-	 * the setup_IO_APIC_irqs() to do proper initialization.
-	 */
-	restore_IO_APIC_setup(ioapic_entries);
-}
-
 void free_ioapic_entries(struct IO_APIC_route_entry **ioapic_entries)
 {
 	int apic;
@@ -971,7 +798,6 @@ void free_ioapic_entries(struct IO_APIC_route_entry **ioapic_entries)
 
 	kfree(ioapic_entries);
 }
-#endif
 
 /*
  * Find the IRQ entry number of a certain pin.
@@ -1032,54 +858,6 @@ static int __init find_isa_irq_apic(int irq, int type)
 	return -1;
 }
 
-/*
- * Find a specific PCI IRQ entry.
- * Not an __init, possibly needed by modules
- */
-static int pin_2_irq(int idx, int apic, int pin);
-
-int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin)
-{
-	int apic, i, best_guess = -1;
-
-	apic_printk(APIC_DEBUG, "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n",
-		bus, slot, pin);
-	if (test_bit(bus, mp_bus_not_pci)) {
-		apic_printk(APIC_VERBOSE, "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
-		return -1;
-	}
-	for (i = 0; i < mp_irq_entries; i++) {
-		int lbus = mp_irqs[i].srcbus;
-
-		for (apic = 0; apic < nr_ioapics; apic++)
-			if (mp_ioapics[apic].apicid == mp_irqs[i].dstapic ||
-			    mp_irqs[i].dstapic == MP_APIC_ALL)
-				break;
-
-		if (!test_bit(lbus, mp_bus_not_pci) &&
-		    !mp_irqs[i].irqtype &&
-		    (bus == lbus) &&
-		    (slot == ((mp_irqs[i].srcbusirq >> 2) & 0x1f))) {
-			int irq = pin_2_irq(i, apic, mp_irqs[i].dstirq);
-
-			if (!(apic || IO_APIC_IRQ(irq)))
-				continue;
-
-			if (pin == (mp_irqs[i].srcbusirq & 3))
-				return irq;
-			/*
-			 * Use the first all-but-pin matching entry as a
-			 * best-guess fuzzy result for broken mptables.
-			 */
-			if (best_guess < 0)
-				best_guess = irq;
-		}
-	}
-	return best_guess;
-}
-
-EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector);
-
 #if defined(CONFIG_EISA) || defined(CONFIG_MCA)
 /*
  * EISA Edge/Level control register, ELCR
@@ -1298,6 +1076,64 @@ static int pin_2_irq(int idx, int apic, int pin)
 	return irq;
 }
 
+/*
+ * Find a specific PCI IRQ entry.
+ * Not an __init, possibly needed by modules
+ */
+int IO_APIC_get_PCI_irq_vector(int bus, int slot, int pin,
+				struct io_apic_irq_attr *irq_attr)
+{
+	int apic, i, best_guess = -1;
+
+	apic_printk(APIC_DEBUG,
+		    "querying PCI -> IRQ mapping bus:%d, slot:%d, pin:%d.\n",
+		    bus, slot, pin);
+	if (test_bit(bus, mp_bus_not_pci)) {
+		apic_printk(APIC_VERBOSE,
+			    "PCI BIOS passed nonexistent PCI bus %d!\n", bus);
+		return -1;
+	}
+	for (i = 0; i < mp_irq_entries; i++) {
+		int lbus = mp_irqs[i].srcbus;
+
+		for (apic = 0; apic < nr_ioapics; apic++)
+			if (mp_ioapics[apic].apicid == mp_irqs[i].dstapic ||
+			    mp_irqs[i].dstapic == MP_APIC_ALL)
+				break;
+
+		if (!test_bit(lbus, mp_bus_not_pci) &&
+		    !mp_irqs[i].irqtype &&
+		    (bus == lbus) &&
+		    (slot == ((mp_irqs[i].srcbusirq >> 2) & 0x1f))) {
+			int irq = pin_2_irq(i, apic, mp_irqs[i].dstirq);
+
+			if (!(apic || IO_APIC_IRQ(irq)))
+				continue;
+
+			if (pin == (mp_irqs[i].srcbusirq & 3)) {
+				set_io_apic_irq_attr(irq_attr, apic,
+						     mp_irqs[i].dstirq,
+						     irq_trigger(i),
+						     irq_polarity(i));
+				return irq;
+			}
+			/*
+			 * Use the first all-but-pin matching entry as a
+			 * best-guess fuzzy result for broken mptables.
+			 */
+			if (best_guess < 0) {
+				set_io_apic_irq_attr(irq_attr, apic,
+						     mp_irqs[i].dstirq,
+						     irq_trigger(i),
+						     irq_polarity(i));
+				best_guess = irq;
+			}
+		}
+	}
+	return best_guess;
+}
+EXPORT_SYMBOL(IO_APIC_get_PCI_irq_vector);
+
 void lock_vector_lock(void)
 {
 	/* Used to the online set of cpus does not change
@@ -1559,6 +1395,9 @@ int setup_ioapic_entry(int apic_id, int irq,
 		irte.vector = vector;
 		irte.dest_id = IRTE_DEST(destination);
 
+		/* Set source-id of interrupt request */
+		set_ioapic_sid(&irte, apic_id);
+
 		modify_irte(irq, &irte);
 
 		ir_entry->index2 = (index >> 15) & 0x1;
@@ -1628,58 +1467,70 @@ static void setup_IO_APIC_irq(int apic_id, int pin, unsigned int irq, struct irq
 	ioapic_write_entry(apic_id, pin, entry);
 }
 
+static struct {
+	DECLARE_BITMAP(pin_programmed, MP_MAX_IOAPIC_PIN + 1);
+} mp_ioapic_routing[MAX_IO_APICS];
+
 static void __init setup_IO_APIC_irqs(void)
 {
-	int apic_id, pin, idx, irq;
+	int apic_id = 0, pin, idx, irq;
 	int notcon = 0;
 	struct irq_desc *desc;
 	struct irq_cfg *cfg;
-	int cpu = boot_cpu_id;
+	int node = cpu_to_node(boot_cpu_id);
 
 	apic_printk(APIC_VERBOSE, KERN_DEBUG "init IO_APIC IRQs\n");
 
-	for (apic_id = 0; apic_id < nr_ioapics; apic_id++) {
-		for (pin = 0; pin < nr_ioapic_registers[apic_id]; pin++) {
-
-			idx = find_irq_entry(apic_id, pin, mp_INT);
-			if (idx == -1) {
-				if (!notcon) {
-					notcon = 1;
-					apic_printk(APIC_VERBOSE,
-						KERN_DEBUG " %d-%d",
-						mp_ioapics[apic_id].apicid, pin);
-				} else
-					apic_printk(APIC_VERBOSE, " %d-%d",
-						mp_ioapics[apic_id].apicid, pin);
-				continue;
-			}
-			if (notcon) {
-				apic_printk(APIC_VERBOSE,
-					" (apicid-pin) not connected\n");
-				notcon = 0;
-			}
+#ifdef CONFIG_ACPI
+	if (!acpi_disabled && acpi_ioapic) {
+		apic_id = mp_find_ioapic(0);
+		if (apic_id < 0)
+			apic_id = 0;
+	}
+#endif
 
-			irq = pin_2_irq(idx, apic_id, pin);
+	for (pin = 0; pin < nr_ioapic_registers[apic_id]; pin++) {
+		idx = find_irq_entry(apic_id, pin, mp_INT);
+		if (idx == -1) {
+			if (!notcon) {
+				notcon = 1;
+				apic_printk(APIC_VERBOSE,
+					KERN_DEBUG " %d-%d",
+					mp_ioapics[apic_id].apicid, pin);
+			} else
+				apic_printk(APIC_VERBOSE, " %d-%d",
+					mp_ioapics[apic_id].apicid, pin);
+			continue;
+		}
+		if (notcon) {
+			apic_printk(APIC_VERBOSE,
+				" (apicid-pin) not connected\n");
+			notcon = 0;
+		}
 
-			/*
-			 * Skip the timer IRQ if there's a quirk handler
-			 * installed and if it returns 1:
-			 */
-			if (apic->multi_timer_check &&
-					apic->multi_timer_check(apic_id, irq))
-				continue;
+		irq = pin_2_irq(idx, apic_id, pin);
 
-			desc = irq_to_desc_alloc_cpu(irq, cpu);
-			if (!desc) {
-				printk(KERN_INFO "can not get irq_desc for %d\n", irq);
-				continue;
-			}
-			cfg = desc->chip_data;
-			add_pin_to_irq_cpu(cfg, cpu, apic_id, pin);
+		/*
+		 * Skip the timer IRQ if there's a quirk handler
+		 * installed and if it returns 1:
+		 */
+		if (apic->multi_timer_check &&
+				apic->multi_timer_check(apic_id, irq))
+			continue;
 
-			setup_IO_APIC_irq(apic_id, pin, irq, desc,
-					irq_trigger(idx), irq_polarity(idx));
+		desc = irq_to_desc_alloc_node(irq, node);
+		if (!desc) {
+			printk(KERN_INFO "can not get irq_desc for %d\n", irq);
+			continue;
 		}
+		cfg = desc->chip_data;
+		add_pin_to_irq_node(cfg, node, apic_id, pin);
+		/*
+		 * don't mark it in pin_programmed, so later acpi could
+		 * set it correctly when irq < 16
+		 */
+		setup_IO_APIC_irq(apic_id, pin, irq, desc,
+				irq_trigger(idx), irq_polarity(idx));
 	}
 
 	if (notcon)
@@ -1832,12 +1683,8 @@ __apicdebuginit(void) print_IO_APIC(void)
 		if (!entry)
 			continue;
 		printk(KERN_DEBUG "IRQ%d ", irq);
-		for (;;) {
+		for_each_irq_pin(entry, cfg->irq_2_pin)
 			printk("-> %d:%d", entry->apic, entry->pin);
-			if (!entry->next)
-				break;
-			entry = entry->next;
-		}
 		printk("\n");
 	}
 
@@ -1846,36 +1693,30 @@ __apicdebuginit(void) print_IO_APIC(void)
 	return;
 }
 
-__apicdebuginit(void) print_APIC_bitfield(int base)
+__apicdebuginit(void) print_APIC_field(int base)
 {
-	unsigned int v;
-	int i, j;
+	int i;
 
 	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);
-		for (j = 0; j < 32; j++) {
-			if (v & (1<<j))
-				printk("1");
-			else
-				printk("0");
-		}
-		printk("\n");
-	}
+	printk(KERN_DEBUG);
+
+	for (i = 0; i < 8; i++)
+		printk(KERN_CONT "%08x", apic_read(base + i*0x10));
+
+	printk(KERN_CONT "\n");
 }
 
 __apicdebuginit(void) print_local_APIC(void *dummy)
 {
-	unsigned int v, ver, maxlvt;
+	unsigned int i, v, ver, maxlvt;
 	u64 icr;
 
 	if (apic_verbosity == APIC_QUIET)
 		return;
 
-	printk("\n" KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
+	printk(KERN_DEBUG "printing local APIC contents on CPU#%d/%d:\n",
 		smp_processor_id(), hard_smp_processor_id());
 	v = apic_read(APIC_ID);
 	printk(KERN_INFO "... APIC ID:      %08x (%01x)\n", v, read_apic_id());
@@ -1916,11 +1757,11 @@ __apicdebuginit(void) print_local_APIC(void *dummy)
 	printk(KERN_DEBUG "... APIC SPIV: %08x\n", v);
 
 	printk(KERN_DEBUG "... APIC ISR field:\n");
-	print_APIC_bitfield(APIC_ISR);
+	print_APIC_field(APIC_ISR);
 	printk(KERN_DEBUG "... APIC TMR field:\n");
-	print_APIC_bitfield(APIC_TMR);
+	print_APIC_field(APIC_TMR);
 	printk(KERN_DEBUG "... APIC IRR field:\n");
-	print_APIC_bitfield(APIC_IRR);
+	print_APIC_field(APIC_IRR);
 
 	if (APIC_INTEGRATED(ver)) {             /* !82489DX */
 		if (maxlvt > 3)         /* Due to the Pentium erratum 3AP. */
@@ -1957,6 +1798,18 @@ __apicdebuginit(void) print_local_APIC(void *dummy)
 	printk(KERN_DEBUG "... APIC TMCCT: %08x\n", v);
 	v = apic_read(APIC_TDCR);
 	printk(KERN_DEBUG "... APIC TDCR: %08x\n", v);
+
+	if (boot_cpu_has(X86_FEATURE_EXTAPIC)) {
+		v = apic_read(APIC_EFEAT);
+		maxlvt = (v >> 16) & 0xff;
+		printk(KERN_DEBUG "... APIC EFEAT: %08x\n", v);
+		v = apic_read(APIC_ECTRL);
+		printk(KERN_DEBUG "... APIC ECTRL: %08x\n", v);
+		for (i = 0; i < maxlvt; i++) {
+			v = apic_read(APIC_EILVTn(i));
+			printk(KERN_DEBUG "... APIC EILVT%d: %08x\n", i, v);
+		}
+	}
 	printk("\n");
 }
 
@@ -2005,6 +1858,11 @@ __apicdebuginit(void) print_PIC(void)
 __apicdebuginit(int) print_all_ICs(void)
 {
 	print_PIC();
+
+	/* don't print out if apic is not there */
+	if (!cpu_has_apic || disable_apic)
+		return 0;
+
 	print_all_local_APICs();
 	print_IO_APIC();
 
@@ -2120,7 +1978,9 @@ void disable_IO_APIC(void)
 	/*
 	 * Use virtual wire A mode when interrupt remapping is enabled.
 	 */
-	disconnect_bsp_APIC(!intr_remapping_enabled && ioapic_i8259.pin != -1);
+	if (cpu_has_apic)
+		disconnect_bsp_APIC(!intr_remapping_enabled &&
+				ioapic_i8259.pin != -1);
 }
 
 #ifdef CONFIG_X86_32
@@ -2328,7 +2188,6 @@ static unsigned int startup_ioapic_irq(unsigned int irq)
 	return was_pending;
 }
 
-#ifdef CONFIG_X86_64
 static int ioapic_retrigger_irq(unsigned int irq)
 {
 
@@ -2339,27 +2198,124 @@ static int ioapic_retrigger_irq(unsigned int irq)
 	apic->send_IPI_mask(cpumask_of(cpumask_first(cfg->domain)), cfg->vector);
 	spin_unlock_irqrestore(&vector_lock, flags);
 
-	return 1;
+	return 1;
+}
+
+/*
+ * 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.
+ */
+
+#ifdef CONFIG_SMP
+static void send_cleanup_vector(struct irq_cfg *cfg)
+{
+	cpumask_var_t cleanup_mask;
+
+	if (unlikely(!alloc_cpumask_var(&cleanup_mask, GFP_ATOMIC))) {
+		unsigned int i;
+		cfg->move_cleanup_count = 0;
+		for_each_cpu_and(i, cfg->old_domain, cpu_online_mask)
+			cfg->move_cleanup_count++;
+		for_each_cpu_and(i, cfg->old_domain, cpu_online_mask)
+			apic->send_IPI_mask(cpumask_of(i), IRQ_MOVE_CLEANUP_VECTOR);
+	} else {
+		cpumask_and(cleanup_mask, cfg->old_domain, cpu_online_mask);
+		cfg->move_cleanup_count = cpumask_weight(cleanup_mask);
+		apic->send_IPI_mask(cleanup_mask, IRQ_MOVE_CLEANUP_VECTOR);
+		free_cpumask_var(cleanup_mask);
+	}
+	cfg->move_in_progress = 0;
+}
+
+static void __target_IO_APIC_irq(unsigned int irq, unsigned int dest, struct irq_cfg *cfg)
+{
+	int apic, pin;
+	struct irq_pin_list *entry;
+	u8 vector = cfg->vector;
+
+	for_each_irq_pin(entry, cfg->irq_2_pin) {
+		unsigned int reg;
+
+		apic = entry->apic;
+		pin = entry->pin;
+		/*
+		 * With interrupt-remapping, destination information comes
+		 * from interrupt-remapping table entry.
+		 */
+		if (!irq_remapped(irq))
+			io_apic_write(apic, 0x11 + pin*2, dest);
+		reg = io_apic_read(apic, 0x10 + pin*2);
+		reg &= ~IO_APIC_REDIR_VECTOR_MASK;
+		reg |= vector;
+		io_apic_modify(apic, 0x10 + pin*2, reg);
+	}
+}
+
+static int
+assign_irq_vector(int irq, struct irq_cfg *cfg, const struct cpumask *mask);
+
+/*
+ * Either sets desc->affinity to a valid value, and returns
+ * ->cpu_mask_to_apicid of that, or returns BAD_APICID and
+ * leaves desc->affinity untouched.
+ */
+static unsigned int
+set_desc_affinity(struct irq_desc *desc, const struct cpumask *mask)
+{
+	struct irq_cfg *cfg;
+	unsigned int irq;
+
+	if (!cpumask_intersects(mask, cpu_online_mask))
+		return BAD_APICID;
+
+	irq = desc->irq;
+	cfg = desc->chip_data;
+	if (assign_irq_vector(irq, cfg, mask))
+		return BAD_APICID;
+
+	cpumask_copy(desc->affinity, mask);
+
+	return apic->cpu_mask_to_apicid_and(desc->affinity, cfg->domain);
+}
+
+static int
+set_ioapic_affinity_irq_desc(struct irq_desc *desc, const struct cpumask *mask)
+{
+	struct irq_cfg *cfg;
+	unsigned long flags;
+	unsigned int dest;
+	unsigned int irq;
+	int ret = -1;
+
+	irq = desc->irq;
+	cfg = desc->chip_data;
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	dest = set_desc_affinity(desc, mask);
+	if (dest != BAD_APICID) {
+		/* Only the high 8 bits are valid. */
+		dest = SET_APIC_LOGICAL_ID(dest);
+		__target_IO_APIC_irq(irq, dest, cfg);
+		ret = 0;
+	}
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+
+	return ret;
 }
-#else
-static int ioapic_retrigger_irq(unsigned int irq)
-{
-	apic->send_IPI_self(irq_cfg(irq)->vector);
 
-	return 1;
-}
-#endif
+static int
+set_ioapic_affinity_irq(unsigned int irq, const struct cpumask *mask)
+{
+	struct irq_desc *desc;
 
-/*
- * 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.
- */
+	desc = irq_to_desc(irq);
 
-#ifdef CONFIG_SMP
+	return set_ioapic_affinity_irq_desc(desc, mask);
+}
 
 #ifdef CONFIG_INTR_REMAP
 
@@ -2374,26 +2330,25 @@ static int ioapic_retrigger_irq(unsigned int irq)
  * Real vector that is used for interrupting cpu will be coming from
  * the interrupt-remapping table entry.
  */
-static void
+static int
 migrate_ioapic_irq_desc(struct irq_desc *desc, const struct cpumask *mask)
 {
 	struct irq_cfg *cfg;
 	struct irte irte;
 	unsigned int dest;
 	unsigned int irq;
+	int ret = -1;
 
 	if (!cpumask_intersects(mask, cpu_online_mask))
-		return;
+		return ret;
 
 	irq = desc->irq;
 	if (get_irte(irq, &irte))
-		return;
+		return ret;
 
 	cfg = desc->chip_data;
 	if (assign_irq_vector(irq, cfg, mask))
-		return;
-
-	set_extra_move_desc(desc, mask);
+		return ret;
 
 	dest = apic->cpu_mask_to_apicid_and(cfg->domain, mask);
 
@@ -2409,27 +2364,30 @@ migrate_ioapic_irq_desc(struct irq_desc *desc, const struct cpumask *mask)
 		send_cleanup_vector(cfg);
 
 	cpumask_copy(desc->affinity, mask);
+
+	return 0;
 }
 
 /*
  * Migrates the IRQ destination in the process context.
  */
-static void set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc,
+static int set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc,
 					    const struct cpumask *mask)
 {
-	migrate_ioapic_irq_desc(desc, mask);
+	return migrate_ioapic_irq_desc(desc, mask);
 }
-static void set_ir_ioapic_affinity_irq(unsigned int irq,
+static int set_ir_ioapic_affinity_irq(unsigned int irq,
 				       const struct cpumask *mask)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
 
-	set_ir_ioapic_affinity_irq_desc(desc, mask);
+	return set_ir_ioapic_affinity_irq_desc(desc, mask);
 }
 #else
-static inline void set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc,
+static inline int set_ir_ioapic_affinity_irq_desc(struct irq_desc *desc,
 						   const struct cpumask *mask)
 {
+	return 0;
 }
 #endif
 
@@ -2491,87 +2449,19 @@ static void irq_complete_move(struct irq_desc **descp)
 	struct irq_cfg *cfg = desc->chip_data;
 	unsigned vector, me;
 
-	if (likely(!cfg->move_in_progress)) {
-#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC
-		if (likely(!cfg->move_desc_pending))
-			return;
-
-		/* domain has not changed, but affinity did */
-		me = smp_processor_id();
-		if (cpumask_test_cpu(me, desc->affinity)) {
-			*descp = desc = move_irq_desc(desc, me);
-			/* get the new one */
-			cfg = desc->chip_data;
-			cfg->move_desc_pending = 0;
-		}
-#endif
+	if (likely(!cfg->move_in_progress))
 		return;
-	}
 
 	vector = ~get_irq_regs()->orig_ax;
 	me = smp_processor_id();
 
-	if (vector == cfg->vector && cpumask_test_cpu(me, cfg->domain)) {
-#ifdef CONFIG_NUMA_MIGRATE_IRQ_DESC
-		*descp = desc = move_irq_desc(desc, me);
-		/* get the new one */
-		cfg = desc->chip_data;
-#endif
+	if (vector == cfg->vector && cpumask_test_cpu(me, cfg->domain))
 		send_cleanup_vector(cfg);
-	}
 }
 #else
 static inline void irq_complete_move(struct irq_desc **descp) {}
 #endif
 
-#ifdef CONFIG_INTR_REMAP
-static void __eoi_ioapic_irq(unsigned int irq, struct irq_cfg *cfg)
-{
-	int apic, pin;
-	struct irq_pin_list *entry;
-
-	entry = cfg->irq_2_pin;
-	for (;;) {
-
-		if (!entry)
-			break;
-
-		apic = entry->apic;
-		pin = entry->pin;
-		io_apic_eoi(apic, pin);
-		entry = entry->next;
-	}
-}
-
-static void
-eoi_ioapic_irq(struct irq_desc *desc)
-{
-	struct irq_cfg *cfg;
-	unsigned long flags;
-	unsigned int irq;
-
-	irq = desc->irq;
-	cfg = desc->chip_data;
-
-	spin_lock_irqsave(&ioapic_lock, flags);
-	__eoi_ioapic_irq(irq, cfg);
-	spin_unlock_irqrestore(&ioapic_lock, flags);
-}
-
-static void ack_x2apic_level(unsigned int irq)
-{
-	struct irq_desc *desc = irq_to_desc(irq);
-	ack_x2APIC_irq();
-	eoi_ioapic_irq(desc);
-}
-
-static void ack_x2apic_edge(unsigned int irq)
-{
-	ack_x2APIC_irq();
-}
-
-#endif
-
 static void ack_apic_edge(unsigned int irq)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
@@ -2586,11 +2476,8 @@ atomic_t irq_mis_count;
 static void ack_apic_level(unsigned int irq)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
-
-#ifdef CONFIG_X86_32
 	unsigned long v;
 	int i;
-#endif
 	struct irq_cfg *cfg;
 	int do_unmask_irq = 0;
 
@@ -2603,31 +2490,28 @@ static void ack_apic_level(unsigned int irq)
 	}
 #endif
 
-#ifdef CONFIG_X86_32
 	/*
-	* 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
-	*/
+	 * 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
+	 */
 	cfg = desc->chip_data;
 	i = cfg->vector;
-
 	v = apic_read(APIC_TMR + ((i & ~0x1f) >> 1));
-#endif
 
 	/*
 	 * We must acknowledge the irq before we move it or the acknowledge will
@@ -2669,7 +2553,7 @@ static void ack_apic_level(unsigned int irq)
 		unmask_IO_APIC_irq_desc(desc);
 	}
 
-#ifdef CONFIG_X86_32
+	/* Tail end of version 0x11 I/O APIC bug workaround */
 	if (!(v & (1 << (i & 0x1f)))) {
 		atomic_inc(&irq_mis_count);
 		spin_lock(&ioapic_lock);
@@ -2677,9 +2561,46 @@ static void ack_apic_level(unsigned int irq)
 		__unmask_and_level_IO_APIC_irq(cfg);
 		spin_unlock(&ioapic_lock);
 	}
-#endif
 }
 
+#ifdef CONFIG_INTR_REMAP
+static void __eoi_ioapic_irq(unsigned int irq, struct irq_cfg *cfg)
+{
+	struct irq_pin_list *entry;
+
+	for_each_irq_pin(entry, cfg->irq_2_pin)
+		io_apic_eoi(entry->apic, entry->pin);
+}
+
+static void
+eoi_ioapic_irq(struct irq_desc *desc)
+{
+	struct irq_cfg *cfg;
+	unsigned long flags;
+	unsigned int irq;
+
+	irq = desc->irq;
+	cfg = desc->chip_data;
+
+	spin_lock_irqsave(&ioapic_lock, flags);
+	__eoi_ioapic_irq(irq, cfg);
+	spin_unlock_irqrestore(&ioapic_lock, flags);
+}
+
+static void ir_ack_apic_edge(unsigned int irq)
+{
+	ack_APIC_irq();
+}
+
+static void ir_ack_apic_level(unsigned int irq)
+{
+	struct irq_desc *desc = irq_to_desc(irq);
+
+	ack_APIC_irq();
+	eoi_ioapic_irq(desc);
+}
+#endif /* CONFIG_INTR_REMAP */
+
 static struct irq_chip ioapic_chip __read_mostly = {
 	.name		= "IO-APIC",
 	.startup	= startup_ioapic_irq,
@@ -2699,8 +2620,8 @@ static struct irq_chip ir_ioapic_chip __read_mostly = {
 	.mask		= mask_IO_APIC_irq,
 	.unmask		= unmask_IO_APIC_irq,
 #ifdef CONFIG_INTR_REMAP
-	.ack		= ack_x2apic_edge,
-	.eoi		= ack_x2apic_level,
+	.ack		= ir_ack_apic_edge,
+	.eoi		= ir_ack_apic_level,
 #ifdef CONFIG_SMP
 	.set_affinity	= set_ir_ioapic_affinity_irq,
 #endif
@@ -2881,7 +2802,7 @@ static inline void __init check_timer(void)
 {
 	struct irq_desc *desc = irq_to_desc(0);
 	struct irq_cfg *cfg = desc->chip_data;
-	int cpu = boot_cpu_id;
+	int node = cpu_to_node(boot_cpu_id);
 	int apic1, pin1, apic2, pin2;
 	unsigned long flags;
 	int no_pin1 = 0;
@@ -2947,7 +2868,7 @@ static inline void __init check_timer(void)
 		 * Ok, does IRQ0 through the IOAPIC work?
 		 */
 		if (no_pin1) {
-			add_pin_to_irq_cpu(cfg, cpu, apic1, pin1);
+			add_pin_to_irq_node(cfg, node, apic1, pin1);
 			setup_timer_IRQ0_pin(apic1, pin1, cfg->vector);
 		} else {
 			/* for edge trigger, setup_IO_APIC_irq already
@@ -2984,7 +2905,7 @@ static inline void __init check_timer(void)
 		/*
 		 * legacy devices should be connected to IO APIC #0
 		 */
-		replace_pin_at_irq_cpu(cfg, cpu, apic1, pin1, apic2, pin2);
+		replace_pin_at_irq_node(cfg, node, apic1, pin1, apic2, pin2);
 		setup_timer_IRQ0_pin(apic2, pin2, cfg->vector);
 		enable_8259A_irq(0);
 		if (timer_irq_works()) {
@@ -3196,14 +3117,13 @@ static int nr_irqs_gsi = NR_IRQS_LEGACY;
 /*
  * Dynamic irq allocate and deallocation
  */
-unsigned int create_irq_nr(unsigned int irq_want)
+unsigned int create_irq_nr(unsigned int irq_want, int node)
 {
 	/* Allocate an unused irq */
 	unsigned int irq;
 	unsigned int new;
 	unsigned long flags;
 	struct irq_cfg *cfg_new = NULL;
-	int cpu = boot_cpu_id;
 	struct irq_desc *desc_new = NULL;
 
 	irq = 0;
@@ -3212,7 +3132,7 @@ unsigned int create_irq_nr(unsigned int irq_want)
 
 	spin_lock_irqsave(&vector_lock, flags);
 	for (new = irq_want; new < nr_irqs; new++) {
-		desc_new = irq_to_desc_alloc_cpu(new, cpu);
+		desc_new = irq_to_desc_alloc_node(new, node);
 		if (!desc_new) {
 			printk(KERN_INFO "can not get irq_desc for %d\n", new);
 			continue;
@@ -3221,6 +3141,9 @@ unsigned int create_irq_nr(unsigned int irq_want)
 
 		if (cfg_new->vector != 0)
 			continue;
+
+		desc_new = move_irq_desc(desc_new, node);
+
 		if (__assign_irq_vector(new, cfg_new, apic->target_cpus()) == 0)
 			irq = new;
 		break;
@@ -3238,11 +3161,12 @@ unsigned int create_irq_nr(unsigned int irq_want)
 
 int create_irq(void)
 {
+	int node = cpu_to_node(boot_cpu_id);
 	unsigned int irq_want;
 	int irq;
 
 	irq_want = nr_irqs_gsi;
-	irq = create_irq_nr(irq_want);
+	irq = create_irq_nr(irq_want, node);
 
 	if (irq == 0)
 		irq = -1;
@@ -3261,8 +3185,7 @@ void destroy_irq(unsigned int irq)
 	cfg = desc->chip_data;
 	dynamic_irq_cleanup(irq);
 	/* connect back irq_cfg */
-	if (desc)
-		desc->chip_data = cfg;
+	desc->chip_data = cfg;
 
 	free_irte(irq);
 	spin_lock_irqsave(&vector_lock, flags);
@@ -3307,6 +3230,9 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms
 		irte.vector = cfg->vector;
 		irte.dest_id = IRTE_DEST(dest);
 
+		/* Set source-id of interrupt request */
+		set_msi_sid(&irte, pdev);
+
 		modify_irte(irq, &irte);
 
 		msg->address_hi = MSI_ADDR_BASE_HI;
@@ -3344,7 +3270,7 @@ static int msi_compose_msg(struct pci_dev *pdev, unsigned int irq, struct msi_ms
 }
 
 #ifdef CONFIG_SMP
-static void set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask)
+static int set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
 	struct irq_cfg *cfg;
@@ -3353,7 +3279,7 @@ static void set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask)
 
 	dest = set_desc_affinity(desc, mask);
 	if (dest == BAD_APICID)
-		return;
+		return -1;
 
 	cfg = desc->chip_data;
 
@@ -3365,13 +3291,15 @@ static void set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask)
 	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
 
 	write_msi_msg_desc(desc, &msg);
+
+	return 0;
 }
 #ifdef CONFIG_INTR_REMAP
 /*
  * Migrate the MSI irq to another cpumask. This migration is
  * done in the process context using interrupt-remapping hardware.
  */
-static void
+static int
 ir_set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
@@ -3380,11 +3308,11 @@ ir_set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask)
 	struct irte irte;
 
 	if (get_irte(irq, &irte))
-		return;
+		return -1;
 
 	dest = set_desc_affinity(desc, mask);
 	if (dest == BAD_APICID)
-		return;
+		return -1;
 
 	irte.vector = cfg->vector;
 	irte.dest_id = IRTE_DEST(dest);
@@ -3401,6 +3329,8 @@ ir_set_msi_irq_affinity(unsigned int irq, const struct cpumask *mask)
 	 */
 	if (cfg->move_in_progress)
 		send_cleanup_vector(cfg);
+
+	return 0;
 }
 
 #endif
@@ -3426,7 +3356,7 @@ static struct irq_chip msi_ir_chip = {
 	.unmask		= unmask_msi_irq,
 	.mask		= mask_msi_irq,
 #ifdef CONFIG_INTR_REMAP
-	.ack		= ack_x2apic_edge,
+	.ack		= ir_ack_apic_edge,
 #ifdef CONFIG_SMP
 	.set_affinity	= ir_set_msi_irq_affinity,
 #endif
@@ -3496,15 +3426,17 @@ int arch_setup_msi_irqs(struct pci_dev *dev, int nvec, int type)
 	unsigned int irq_want;
 	struct intel_iommu *iommu = NULL;
 	int index = 0;
+	int node;
 
 	/* x86 doesn't support multiple MSI yet */
 	if (type == PCI_CAP_ID_MSI && nvec > 1)
 		return 1;
 
+	node = dev_to_node(&dev->dev);
 	irq_want = nr_irqs_gsi;
 	sub_handle = 0;
 	list_for_each_entry(msidesc, &dev->msi_list, list) {
-		irq = create_irq_nr(irq_want);
+		irq = create_irq_nr(irq_want, node);
 		if (irq == 0)
 			return -1;
 		irq_want = irq + 1;
@@ -3554,7 +3486,7 @@ void arch_teardown_msi_irq(unsigned int irq)
 
 #if defined (CONFIG_DMAR) || defined (CONFIG_INTR_REMAP)
 #ifdef CONFIG_SMP
-static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask)
+static int dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
 	struct irq_cfg *cfg;
@@ -3563,7 +3495,7 @@ static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask)
 
 	dest = set_desc_affinity(desc, mask);
 	if (dest == BAD_APICID)
-		return;
+		return -1;
 
 	cfg = desc->chip_data;
 
@@ -3575,11 +3507,13 @@ static void dmar_msi_set_affinity(unsigned int irq, const struct cpumask *mask)
 	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
 
 	dmar_msi_write(irq, &msg);
+
+	return 0;
 }
 
 #endif /* CONFIG_SMP */
 
-struct irq_chip dmar_msi_type = {
+static struct irq_chip dmar_msi_type = {
 	.name = "DMAR_MSI",
 	.unmask = dmar_msi_unmask,
 	.mask = dmar_msi_mask,
@@ -3608,7 +3542,7 @@ int arch_setup_dmar_msi(unsigned int irq)
 #ifdef CONFIG_HPET_TIMER
 
 #ifdef CONFIG_SMP
-static void hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask)
+static int hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
 	struct irq_cfg *cfg;
@@ -3617,7 +3551,7 @@ static void hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask)
 
 	dest = set_desc_affinity(desc, mask);
 	if (dest == BAD_APICID)
-		return;
+		return -1;
 
 	cfg = desc->chip_data;
 
@@ -3629,6 +3563,8 @@ static void hpet_msi_set_affinity(unsigned int irq, const struct cpumask *mask)
 	msg.address_lo |= MSI_ADDR_DEST_ID(dest);
 
 	hpet_msi_write(irq, &msg);
+
+	return 0;
 }
 
 #endif /* CONFIG_SMP */
@@ -3648,12 +3584,14 @@ int arch_setup_hpet_msi(unsigned int irq)
 {
 	int ret;
 	struct msi_msg msg;
+	struct irq_desc *desc = irq_to_desc(irq);
 
 	ret = msi_compose_msg(NULL, irq, &msg);
 	if (ret < 0)
 		return ret;
 
 	hpet_msi_write(irq, &msg);
+	desc->status |= IRQ_MOVE_PCNTXT;
 	set_irq_chip_and_handler_name(irq, &hpet_msi_type, handle_edge_irq,
 		"edge");
 
@@ -3683,7 +3621,7 @@ static void target_ht_irq(unsigned int irq, unsigned int dest, u8 vector)
 	write_ht_irq_msg(irq, &msg);
 }
 
-static void set_ht_irq_affinity(unsigned int irq, const struct cpumask *mask)
+static int set_ht_irq_affinity(unsigned int irq, const struct cpumask *mask)
 {
 	struct irq_desc *desc = irq_to_desc(irq);
 	struct irq_cfg *cfg;
@@ -3691,11 +3629,13 @@ static void set_ht_irq_affinity(unsigned int irq, const struct cpumask *mask)
 
 	dest = set_desc_affinity(desc, mask);
 	if (dest == BAD_APICID)
-		return;
+		return -1;
 
 	cfg = desc->chip_data;
 
 	target_ht_irq(irq, dest, cfg->vector);
+
+	return 0;
 }
 
 #endif
@@ -3770,6 +3710,8 @@ int arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade,
 	unsigned long flags;
 	int err;
 
+	BUILD_BUG_ON(sizeof(struct uv_IO_APIC_route_entry) != sizeof(unsigned long));
+
 	cfg = irq_cfg(irq);
 
 	err = assign_irq_vector(irq, cfg, eligible_cpu);
@@ -3783,15 +3725,13 @@ int arch_enable_uv_irq(char *irq_name, unsigned int irq, int cpu, int mmr_blade,
 
 	mmr_value = 0;
 	entry = (struct uv_IO_APIC_route_entry *)&mmr_value;
-	BUG_ON(sizeof(struct uv_IO_APIC_route_entry) != sizeof(unsigned long));
-
-	entry->vector = cfg->vector;
-	entry->delivery_mode = apic->irq_delivery_mode;
-	entry->dest_mode = apic->irq_dest_mode;
-	entry->polarity = 0;
-	entry->trigger = 0;
-	entry->mask = 0;
-	entry->dest = apic->cpu_mask_to_apicid(eligible_cpu);
+	entry->vector		= cfg->vector;
+	entry->delivery_mode	= apic->irq_delivery_mode;
+	entry->dest_mode	= apic->irq_dest_mode;
+	entry->polarity		= 0;
+	entry->trigger		= 0;
+	entry->mask		= 0;
+	entry->dest		= apic->cpu_mask_to_apicid(eligible_cpu);
 
 	mmr_pnode = uv_blade_to_pnode(mmr_blade);
 	uv_write_global_mmr64(mmr_pnode, mmr_offset, mmr_value);
@@ -3809,10 +3749,10 @@ void arch_disable_uv_irq(int mmr_blade, unsigned long mmr_offset)
 	struct uv_IO_APIC_route_entry *entry;
 	int mmr_pnode;
 
+	BUILD_BUG_ON(sizeof(struct uv_IO_APIC_route_entry) != sizeof(unsigned long));
+
 	mmr_value = 0;
 	entry = (struct uv_IO_APIC_route_entry *)&mmr_value;
-	BUG_ON(sizeof(struct uv_IO_APIC_route_entry) != sizeof(unsigned long));
-
 	entry->mask = 1;
 
 	mmr_pnode = uv_blade_to_pnode(mmr_blade);
@@ -3876,6 +3816,75 @@ int __init arch_probe_nr_irqs(void)
 }
 #endif
 
+static int __io_apic_set_pci_routing(struct device *dev, int irq,
+				struct io_apic_irq_attr *irq_attr)
+{
+	struct irq_desc *desc;
+	struct irq_cfg *cfg;
+	int node;
+	int ioapic, pin;
+	int trigger, polarity;
+
+	ioapic = irq_attr->ioapic;
+	if (!IO_APIC_IRQ(irq)) {
+		apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
+			ioapic);
+		return -EINVAL;
+	}
+
+	if (dev)
+		node = dev_to_node(dev);
+	else
+		node = cpu_to_node(boot_cpu_id);
+
+	desc = irq_to_desc_alloc_node(irq, node);
+	if (!desc) {
+		printk(KERN_INFO "can not get irq_desc %d\n", irq);
+		return 0;
+	}
+
+	pin = irq_attr->ioapic_pin;
+	trigger = irq_attr->trigger;
+	polarity = irq_attr->polarity;
+
+	/*
+	 * IRQs < 16 are already in the irq_2_pin[] map
+	 */
+	if (irq >= NR_IRQS_LEGACY) {
+		cfg = desc->chip_data;
+		if (add_pin_to_irq_node_nopanic(cfg, node, ioapic, pin)) {
+			printk(KERN_INFO "can not add pin %d for irq %d\n",
+				pin, irq);
+			return 0;
+		}
+	}
+
+	setup_IO_APIC_irq(ioapic, pin, irq, desc, trigger, polarity);
+
+	return 0;
+}
+
+int io_apic_set_pci_routing(struct device *dev, int irq,
+				struct io_apic_irq_attr *irq_attr)
+{
+	int ioapic, pin;
+	/*
+	 * Avoid pin reprogramming.  PRTs typically include entries
+	 * with redundant pin->gsi mappings (but unique PCI devices);
+	 * we only program the IOAPIC on the first.
+	 */
+	ioapic = irq_attr->ioapic;
+	pin = irq_attr->ioapic_pin;
+	if (test_bit(pin, mp_ioapic_routing[ioapic].pin_programmed)) {
+		pr_debug("Pin %d-%d already programmed\n",
+			 mp_ioapics[ioapic].apicid, pin);
+		return 0;
+	}
+	set_bit(pin, mp_ioapic_routing[ioapic].pin_programmed);
+
+	return __io_apic_set_pci_routing(dev, irq, irq_attr);
+}
+
 /* --------------------------------------------------------------------------
                           ACPI-based IOAPIC Configuration
    -------------------------------------------------------------------------- */
@@ -3956,6 +3965,7 @@ int __init io_apic_get_unique_id(int ioapic, int apic_id)
 
 	return apic_id;
 }
+#endif
 
 int __init io_apic_get_version(int ioapic)
 {
@@ -3968,39 +3978,6 @@ int __init io_apic_get_version(int ioapic)
 
 	return reg_01.bits.version;
 }
-#endif
-
-int io_apic_set_pci_routing (int ioapic, int pin, int irq, int triggering, int polarity)
-{
-	struct irq_desc *desc;
-	struct irq_cfg *cfg;
-	int cpu = boot_cpu_id;
-
-	if (!IO_APIC_IRQ(irq)) {
-		apic_printk(APIC_QUIET,KERN_ERR "IOAPIC[%d]: Invalid reference to IRQ 0\n",
-			ioapic);
-		return -EINVAL;
-	}
-
-	desc = irq_to_desc_alloc_cpu(irq, cpu);
-	if (!desc) {
-		printk(KERN_INFO "can not get irq_desc %d\n", irq);
-		return 0;
-	}
-
-	/*
-	 * IRQs < 16 are already in the irq_2_pin[] map
-	 */
-	if (irq >= NR_IRQS_LEGACY) {
-		cfg = desc->chip_data;
-		add_pin_to_irq_cpu(cfg, cpu, ioapic, pin);
-	}
-
-	setup_IO_APIC_irq(ioapic, pin, irq, desc, triggering, polarity);
-
-	return 0;
-}
-
 
 int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
 {
@@ -4031,51 +4008,44 @@ int acpi_get_override_irq(int bus_irq, int *trigger, int *polarity)
 #ifdef CONFIG_SMP
 void __init setup_ioapic_dest(void)
 {
-	int pin, ioapic, irq, irq_entry;
+	int pin, ioapic = 0, irq, irq_entry;
 	struct irq_desc *desc;
-	struct irq_cfg *cfg;
 	const struct cpumask *mask;
 
 	if (skip_ioapic_setup == 1)
 		return;
 
-	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.
-			 */
-			desc = irq_to_desc(irq);
-			cfg = desc->chip_data;
-			if (!cfg->vector) {
-				setup_IO_APIC_irq(ioapic, pin, irq, desc,
-						  irq_trigger(irq_entry),
-						  irq_polarity(irq_entry));
-				continue;
+#ifdef CONFIG_ACPI
+	if (!acpi_disabled && acpi_ioapic) {
+		ioapic = mp_find_ioapic(0);
+		if (ioapic < 0)
+			ioapic = 0;
+	}
+#endif
 
-			}
+	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);
 
-			/*
-			 * Honour affinities which have been set in early boot
-			 */
-			if (desc->status &
-			    (IRQ_NO_BALANCING | IRQ_AFFINITY_SET))
-				mask = desc->affinity;
-			else
-				mask = apic->target_cpus();
+		desc = irq_to_desc(irq);
 
-			if (intr_remapping_enabled)
-				set_ir_ioapic_affinity_irq_desc(desc, mask);
-			else
-				set_ioapic_affinity_irq_desc(desc, mask);
-		}
+		/*
+		 * Honour affinities which have been set in early boot
+		 */
+		if (desc->status &
+		    (IRQ_NO_BALANCING | IRQ_AFFINITY_SET))
+			mask = desc->affinity;
+		else
+			mask = apic->target_cpus();
 
+		if (intr_remapping_enabled)
+			set_ir_ioapic_affinity_irq_desc(desc, mask);
+		else
+			set_ioapic_affinity_irq_desc(desc, mask);
 	}
+
 }
 #endif
 
@@ -4083,7 +4053,7 @@ void __init setup_ioapic_dest(void)
 
 static struct resource *ioapic_resources;
 
-static struct resource * __init ioapic_setup_resources(void)
+static struct resource * __init ioapic_setup_resources(int nr_ioapics)
 {
 	unsigned long n;
 	struct resource *res;
@@ -4099,15 +4069,13 @@ static struct resource * __init ioapic_setup_resources(void)
 	mem = alloc_bootmem(n);
 	res = (void *)mem;
 
-	if (mem != NULL) {
-		mem += sizeof(struct resource) * nr_ioapics;
+	mem += sizeof(struct resource) * nr_ioapics;
 
-		for (i = 0; i < nr_ioapics; i++) {
-			res[i].name = mem;
-			res[i].flags = IORESOURCE_MEM | IORESOURCE_BUSY;
-			sprintf(mem,  "IOAPIC %u", i);
-			mem += IOAPIC_RESOURCE_NAME_SIZE;
-		}
+	for (i = 0; i < nr_ioapics; i++) {
+		res[i].name = mem;
+		res[i].flags = IORESOURCE_MEM | IORESOURCE_BUSY;
+		sprintf(mem,  "IOAPIC %u", i);
+		mem += IOAPIC_RESOURCE_NAME_SIZE;
 	}
 
 	ioapic_resources = res;
@@ -4121,7 +4089,7 @@ void __init ioapic_init_mappings(void)
 	struct resource *ioapic_res;
 	int i;
 
-	ioapic_res = ioapic_setup_resources();
+	ioapic_res = ioapic_setup_resources(nr_ioapics);
 	for (i = 0; i < nr_ioapics; i++) {
 		if (smp_found_config) {
 			ioapic_phys = mp_ioapics[i].apicaddr;
@@ -4150,11 +4118,9 @@ fake_ioapic_page:
 			    __fix_to_virt(idx), ioapic_phys);
 		idx++;
 
-		if (ioapic_res != NULL) {
-			ioapic_res->start = ioapic_phys;
-			ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
-			ioapic_res++;
-		}
+		ioapic_res->start = ioapic_phys;
+		ioapic_res->end = ioapic_phys + (4 * 1024) - 1;
+		ioapic_res++;
 	}
 }